(H.J. Paton) The Categorical Imperative - A Study in Kant's Moral Philosophy

The Categorical Imperative
is
a careful
study of Kant's most important contribution

to
moral
of
is,
philosophy
his
Groundwork
Morals.
It
the
Metaphysic
written
of
however,
not
merely for Kantian scholars

these
are
(though
series
catered
for
in
of
appendices), but for
all
who
are con-
cerned with ethical problems, and not

least for those
who
are dissatisfied with
modern
abolish
relativistic doctrines
tending to
the
distinction
between good
as
clear
and
evil.
In language
it
and
simple as possible
argues that Kant,
as he himself claimed, has succeeded in
formulating the fundamental principles
on which the moral
life
is
based.
It
endeavours to describe and to defend

the function of reason in
human
action,
and
tion
especially in moral action
a func-
commonly
ignored or derided at
the present time.
And
it
is
put
for-
ward, not as a mere historical investigation,
but as a contribution to the

is
moral thinking which

during a period
action
is
indispensable
lawlessness
when
in
mirrored by a corresponding
confusion in thought.
In the course of this study
many
THE CATEGORICAL IMPERATIVE
traditional
misunderstandings
are
re*
moved.
example,
It is
commonly supposed,
in
for
that
Kant's
theory
an
action cannot be good if
we have any
it
natural inclination towards
or derive
;
any pleasure
that a good
from
its
performance
man
will take
no account
of consequences but will be concerned
only with his
own
state of
mind
and
cir-
even that without any regard to
cumstances he will deduce his various

duties
from the bare concept of universal
law.
By
showing such absurdities to

is
be baseless a way
opened up for a
new
and,
interpretation of Kant's doctrines

it
may be hoped,
for a
new advance
in moral philosophy
itself.
BY THE SAME
The Good
Will
AUTHOR
Kant's Metaphysic of Experience
THE SENIOR SERIES
THE
CATEGORICAL IMPERATIVE
A
study in Kant's
Moral Philosophy
by
H.
J.
PATON
White's Professor of Aforal Philosophy

in the University of Oxford Fellow of the British Academy
HUTCHINSON'S UNIVERSITY LIBRARY

47 Princes Gate, London
New
Tork
Melbourne
Sydney
Cape
Town
Printed in Great Britain by The Anchor Press, Ltd.,

Tiptree, F^'sex
CONTENTS
REFERENGESANDABBREVIATIONS
PREFACE
Page
13
15
BOOK
THE APPROACH TO THE CATEGORICAL IMPERATIVE

Chapter
I
THE CRITICAL METHOD

1.
The need
for
understanding Kant's method.
19
2. 3.
4.
5. 6.
7. 8.
Moral judgements are a priori. The marks of a priori judgement. Composite knowledge.
20
21
22
The task of philosophy. The need for pure ethics. The method of analysis.
Analytic and synthetic arguments.
23 24 25 26
27
9.
10. 11.
The Critical method. The method of synthesis. The importance of Kant. Appendix The Divisions
:
29 29
of Kant's Moral Philo31
sophy.
II
THE GOOD WILL

1.
A good will is good without limitation.

Other possible views. Are all goods good without limitation? Are many goods good without limitation? Is a good will good without limitation?
34
35 36 38 39
41
2.
3.
4.
5. 6.
7.
The The
A good will
Goodness
is
8.
highest good. and' its objects. function of reason.
43
44
45
9.
fundamental.
[5]
6
Chapter III
CONTENTS
DUTY A good will

1
.
Page
and duty.
of isolation.
2.
The method
Is
3.
4.
5. 6.
7.
Misunderstandings. duty the motive of a good will? Is goodness the motive of a good will? Are generous emotions the motive of a good will? Conclusion.
46 47 48 50
52 53 55
Appendix: Worth.
Inclinations,
Happiness.,
and Moral
55
IV
THE MAXIM OF MORALITY

1.
2.
3.
Moral value does not depend on results. subjective and objective. Two kinds of maxim formal and material.
Practical principles
58 59
61 61
4.
The maxim
of duty.
V REVERENCE
1
.
Reverence
for the law.
2.
3.
The The
description of reverence. function of reverence.

:
Appendix Reverence
as a Motive.
63 64 65 66
VI
THE LAW
1
.
2.
Law Law
as such.
as
command.
69 70
71 71
3.
4.
5.
The moral motive. Formal maxim and moral law. The categorical imperative.
72
VII
MISUNDERSTANDINGS
1
.
Criticisms.
2.
3.
4.
5.
Kant's formalism. Kant's legalism. The ignoring of consequences. The soundness of Kant's doctrine.
74 74 75 76
77
CONTENTS
BOOK
II
THE BACKGROUND OF THE CATEGORICAL IMPERATIVE

Chapter
VIII
PRACTICAL REASON AND

PRINCIPLES
1.
Page
ITS
SUBJECTIVE
78 79 79
81 81
The
2.
3.
4.
5. 6.
7.
practical function of reason. senses of 'reason The approach to practical reason. Theoretical reason and action. Practical reason. Impulsive action.
Two
82
83 85 87 88
8.
9. 10.
Means and end. The pursuit of happiness. The denial of practical reason.
Morality.
IX PRACTICAL REASON AND PRINCIPLES

1.
ITS
OBJECTIVE
89 90
91
2. 3.
4.
5.
Subjective and objective principles. The principle of skill. The principle of self-love. The principle of morality.
Conditioned and unconditioned principles. Appendix Kant's View of Reason.

:
93 95
2.
3.
Different senses of 'reason'. The category of cause and effect. The schema of regular succession.
96 97
4. 5. 6.
of freedom. Different kinds of concept.

Intuitive understanding.
The Idea
98 99
100 100
X THE GOOD
1.
The good
'Good
for'
in general.
2.
and 'good
at'.
3.
My good.
The moral good. The ideological view of good. The realistic view of good.
4.
5. 6.
103 104 105 107 108 110
8
Chapter
CONTENTS
Page
XI
IMPERATIVES
1.
2. 3.
4.
5. 6.
Imperatives in general. Definition of an imperative. Three kinds of imperative. Rules, counsels, and laws.
The duty
Obligation and goodness. to act morally.
113 114 114 115 1 1 6

117
XII
HOW ARE IMPERATIVES POSSIBLE?

1.
The meaning
of the question.
2.
3.
4.
5. 6.
7.
Analytic propositions. Synthetic propositions. Synthetic a priori propositions.

Difficulties.
120 120 122

1
22
8.
9.
Imperatives of skill are analytic propositions. Synthetic propositions are presupposed. Imperatives of prudence are analytic propositions. Categorical imperatives are synthetic a priori pro-
123 1 23 125 126

127
positions.
BOOK
III
THE FORMULATION OF THE CATEGORICAL IMPERATIVE

XIII
THE FIVE FORMULAE

1.
2. 3.
4.
The five formulae. The relations between the five formulae. The purpose and structure of the argument. The application of the formulae.
129 130 130

131
XIV THE FORMULA OF UNIVERSAL LAW

1.
2.
3.
Formula I. The one categorical imperative.

Universal law.
4.
5. 6.
7.
Maxims.
Material maxims. of moral judgement. Contradiction in the will. The coherence of rational wills.
The canon
8.
9.
10.
The The
rational will as arbiter. permissible and the obligatory.

:
133 134 135 135 137 137 139 139 14 141

142 144
Appendix
The Spontaneity of Mind.

1.
2.
Intellectual spontaneity. Aesthetic spontaneity.
CONTE NTS
Chapter
9 Page
146 147 148 149 150 150 152 152
XV THE FORMULA
1.
OF
THE LAW OF NATURE

duties.
2.
Formula la. Perfect and imperfect
3.
4.
5. 6.
7.
causal law of nature. Teleological law in nature. The perfection of nature. The appeal to teleological law.
The
8.
9.
Kindness. Promises to repay loans.

Suicide. Culture. Practical reason
10. 11. 12.
154 155
The
principles of
:
and purpose. moral action.

of Nature as a
Appendix
The Law
155 156
Type
of the
157 158 159 160
161
Moral Law. 1. The form of law. 2. The problem of 'exhibition'.

3.
4.
5. 6.
The The
Symbolic exhibition. 'type' of moral law.
natural order. Practical exhibition.
162
XVI THE FORMULA OF THE END

1.
IN ITSELF
165 166 167 169 170 171 172 1 73 175
Formula
II.
2. 3.
The nature
of ends. Ends in themselves.

II.
4.
5. 6.
7. 8.
Grounds and ends. The approach to Formula Kinds of duty.
The
9.
Kant's illustrations. soundness of Kant's view. Special characteristics of Formula

:
II.
Appendix
Arguments in Support of Formula II. 1. Argument from the essence of the

2.
categorical imperative. from the rational agents.
175
Argument
nature
of
1
76
3.
Argument from the character of

good
will.
a 177
4.
Argument from the Formula of

Universal Law.
5.
Summary.
177 178
XVII
THE FORMULA OF AUTONOMY

1.
Formula
III.
180
to
2.
The approach
Formula
III.
181
IO
Chapter
CONTENTS
3. 4.
5.
XVII
The
exclusion of pathological interest. Legislating through maxims. The application of Formula III.
Page 182 182

183
XVIII
THE FORMULA OF THE KINGDOM OF ENDS

1.
Formula
Ilia.
Ilia.
2.
3.
4.
5. 6.
7.
The approach to Formula The kingdom of ends. Kingdom or realm. The supreme head.
185 186
187 187 188 188
8.
9.
Dignity and price. The kingdom of nature. The realisation of the kingdom of ends. The application of Formula Ilia.
190
191 192
10.
11. 12.
Moral
progress.
Kant's historical background. Kant's personality.
194 195 197
BOOK IV
THE JUSTIFICATION OF THE CATEGORICAL IMPERATIVE

XIX THE PROBLEM
1.
The
2.
3.
An
question to be answered. alternative question. of a transcendental deduction.
199 201
The purpose
202
4.
5. 6.
different view.
Possible misunderstandings.
Kant's method.
203 204 206
XX FREEDOM AND AUTONOMY

1.
2.
Kant as a pioneer. Freedom as the key

Will as causality.
to the
moral problem.
3.
207 207 208
4.
5. 6.
Freedom and natural
The
Is
necessity. positive concept of freedom.
209
21
1
Freedom and autonomy.

only a good will free? kinds of heteronomy. Degrees of freedom.
212
213 214 215
7. 8.
Two
9.
CONTENTS
Chapter
ii
XXI FREEDOM AS A NECESSARY PRESUPPOSITION

i
.
Page
Freedom and
rational agents.
2. 3.
4.
5. 6.
presupposition of freedom. Theoretical reason and its presupposition. Practical reason and its presupposition. The self-consciousness of reason. The position of the argument.
The
217 217 218 218 220

221
XXII
THE INTELLIGIBLE WORLD

1.
Side
issues.
2.
Moral
interest.
3.
4.
5. 6.
7. 8.
The alleged vicious circle. The way of escape. The two standpoints. The argument from the passivity
Other arguments.
Conclusion.
of sense.
223 223 224 225 226 228
230
231
XXIII
MEMBERSHIP OF THE INTELLIGIBLE WORLD

1.
Inner sense and mental
states.
2.
3.
The noumenal self. The mind affects itself.

Knowledge of mental
activity.
4.
5. 6. 7.
The
activity of reason.
intelligible world.
Membership of the
The
IS
principles of reason.
233 234 235 236 238 24O 240
XXIV HOW
1.
A
?
242 242 243 244 245 246 247 247 249 250
POSSIBLE
2.
3.
4.
5. 6.
7. 8.
9.
deduction. additional argument. conclusion. failure of the deduction. Direct insight into the principle of autonomy. Is the principle of autonomy analytic? The imperative of autonomy. The objective principles of reason. Reason and the unconditioned. Appendix Kant's Additional Argument.
:
The The The The
XXV
SOME FURTHER QiUESTIONS

I.
Further questions to be considered.
2.
3-
The The
real
self.
4-
conflict of reason and desire. Ethics and metaphysics.
253 253 254 255
12
CONTENTS
Page
5. 6.
7.
Chapter
XXV
Moral
interest.
Interest
and obligation.
8. 9.
Practical insight. Modern intellectualism.
256 258 258 259

261
Kant's teleology.
self-consciousness of practical reason. Thought and action.
10.
1 1
.
The
262
12.
Some
practical objections.
263 265
XXVI THE DEFENCE OF FREEDOM

i.
i>.
3.
The antinomy of freedom and The solution of the antinomy. The two standpoints.
necessity.
266 266
267
4.
5. 6.
7.
How
is
freedom possible?
world. no explanation of freedom.
268
269
271 273
Phenomena and nournena. The thought of the intelligible

There
is
8.
Timeless action.
9. 10.
Freedom to act badly. Freedom and necessity.
11.
The
defence of freedom.
273 275 276 277 279
INDEX OF PROPER NAMES

GENERAL INDEX
280
REFERENCES AND ABBREVIATIONS

THERE is unfortunately no generally accepted method for referring to passages in Kant's works. The first number in my references is the page number in the complete edition published by the Berlin Academy. This number is for some of Kant's works given also in the margin of the complete edition published by Meiner in the Philosophische Bibliothek, which on the whole is the most convenient German edition for ordinary use. It is given, for example, in Vorlander's editions of the Grundlegung and the Metaphysik der Sitten in that series. It is also given in the margin of Meredith's translations of the Critique of Judgement. Many English readers are likely to use Abbott's translations of Kant's ethical works, and the second page number in my references, where it is not enclosed in brackets, has been added for their benefit. Here, however, there for the paging of the Groundwork is a difficulty or, as he calls it, Fundamental Principles of the Metaphysic of Morals is different in the small edition published separately and in the larger work which he calls 'Kant's
;
Theory of Ethics\ I have endeavoured to meet this difficulty by making use of the small page numbers which in both works he inserts in the body of the text ; and for the sake of consistency I have adhered to this practice for the other works, or portions of works, which he translates in Kanfs Theory of Ethics, These numbers are taken by Abbott from the complete edition of Kant's works by Rosenkrantz and Schubert, and so may be of use to those who consult that edition. It would be enormously more convenient if all editions and commentaries on Kant could refer to the pages of the original editions. This is already the practice for the Critique of Pure Reason, where the first edition and the second as B. This practice I have folis commonly known as lowed for the Critique of Pure Reason (K.r. V.}, and where such page numbers are readily available for other works I have given them also but in order to mark their character and to distinguish them from the references to Abbott / have enclosed them in brackets. These numbers are printed in the margin of some of the works published in the Philosophische Bibliothek and even in the Academy edition of the Kritik der praktischen Vernunft and the Kritik der Urteilskraft as printed in 1913 (though not earlier). In the case of the First Introduction (Erste Einleitung] to the Critique of Judgement the page references are to Lehmann's edition in the Philo-
sophische Bibliothek.
List of Chief Abbreviations

Volume
in the
Academy
Beobachtungen
= Beobachtungen
Edition
uber
des
Das Ende
Gr. K.p. V. K.r. V.
K. U.
Logik
= Das Ende alter Dinge. Grundlegung zur Metaphysik der Sitten. = Kritik der praktischen Vernunft. = Kritik der reinen Vernunft. = Kritik der Urteilskraft. = Immanuel Kants Logik.
~ Metaphysische Anfangsgriinde der Naturwissenschaft.
das Gefuhl Schonen und Erhabenen.
II
VIII
V V
IV
III-IV
IX
IV VI IV
VI
M.A.d.N.
M.d.S. ProL
Religion
= Metaphysik der Sitten.

Prolegomena û jeder kilnftigen Meta-
= Die
physik. Religion innerhalb der Grenzen der

biossen Vernunft.
Where
and
own
necessary I refer to the edition of the Berlin Academy as Ak. to that of the Philosophische Bibliothek as Phil. Bib. I refer also to book Kant's Metaphysic of Experience as K.M.E.
my
PREFACE
contrived to say something new about morality. This remarkable achievement has compelled every subsequent writer on moral philosophy to examine his views even if only in order to refute his errors. The curious thing is that Kant himself makes no claim to propound a philosophical revolution in moral thinking as he did in speculative thinking. He knew, of course, that he was trying to do something which no one had succeeded in doing before namely, to set forth the first principles of morality apart from all considerations of self-interest and even apart
KANT
problems. Yet he maintained a new formula for the principle by which good men had always judged moral excellence even if they had been unable to make this principle clear to themselves or to separate it of! sharply from other principles concerned with the happiness of the individual and the benefits arising from the moral life. It is impossible to exaggerate the importance of grasping the supreme principle of morality; and because Kant's Groundwork of the Aletaphysic
from
their application to particular that he was only putting forward
human
of Morals (as I call his Gmndlegung zur Metaphysik der Sitten] treats of this topic, and of this topic alone, it is an indispensable book for all who profess to think seriously about moral problems. Yet many readers find it difficult to understand, and the main purpose of my present book is to make understanding easier. This is all the more necessary because as I believe a whole series of misinterpretations has become traditional and stands in the way of an unprejudiced approach. It is indeed a strange thing that so many of those who either explicitly or implicitly regard Kant as a great, or at least an influential, thinker, ascribe to him views which can hardly be considered as anything but silly. Thus he is commonly supposed to maintain that no action can be moral if we have any natural inclination towards it or if we obtain the slightest pleasure from its perfontiance; and again that a good man must take no account whatever of the consequences of his actions but must deduce all the manifold duties of life from the bare conception of moral law as suchwithout any regard for the characteristics of human nature or the circumstances of human life. These doctrines and others equally paradoxical, if they were held by Kant, would not indicate that he had any very profound insight into the nature of morality they can hardly but suggest that his moral philosophy may be dismissed as negligible, if not diseased. It is my hope to show, by a careful examination of the text, that such interpretations are a distortion of his actual teaching, which is always reasonable, even if it may not always be correct. From what I have said it will be clear that I regard the proper inter:
[15]
l6
PREFACE
pretation of Kant's doctrine as of vital importance, not merely to students of Kant, but to all students of moral philosophy, and indeed to all who seek to lead a good life intelligently arid are not content to follow blindly what they consider to be the wisdom of their forefathers. For this reason I have endeavoured in the earlier sections of my book to deal with the more technical aspects of Kant's teaching in appendices, which can be ignored by those whose interests are limited. This device is no longer open to me in the final section Book IV which, since it is concerned with the metaphysical defence of moral principles, cannot be treated without reference to Kant's philosophy as a whole. But this section too can be ignored though not, I think, without some loss by those who are more anxious to understand what morality is than to consider how it is to be defended. For students of Kant I have tried to show how his moral philosophy fits in with other parts of his Critical doctrine. Though I have taken the
Groundwork, so to speak, as my text, I have sought to confirm my interby references to his other works. As a reviewer of my previous book on Kant complained that his eyes were dazzled by having to look down so often to the footnotes, perhaps it is necessary to explain that these are intended, as a rule, for Kantian scholars who may wish to check my interpretation they are not meant to be examined on a first reading, or indeed by anyone who is more interested in the doctrine that is expounded than in the evidence on which it is based. I may add that I have made no attempt to deal with the gradual evolution of Kant's thought on these subjects, since this has already been done most admirably by Proa book which I should fessor P. A. Schilpp in Kanfs Pre-Critical Ethics
pretation
widely used as a companion to my own. Kant's moral philosophy is not nearly so difficult as his theoretical philosophy, and I hope I have presented no insuperable difficulties to any one who is able and willing to do some hard thinking. In writing
like to see
Kantfs Metaphysic of Experience I often felt as if I were fighting my way through a thicket, thankful at times if I could attach any clear meaning to what he was saying. Even in the Groundwork there are passages which
became clear to me only after years of study, but perhaps because I had already written the earlier book I have not found the difficulties so great, and I have been able to use a greater freedom of expression, which I hope will make things easier for my readers. With the same end in view I hope also to publish later a fresh translation of the Groundwork with a few explanatory notes where the *meaning is obscure. I have chosen as my title The Categorical Imperative* because this looms so large, both in Kant's thinking and in the commonly accepted views about him but it will become clear in the course of my book that for Kant the supreme principle of morality is something higher even than the categorical imperative: it appears to us as a categorical imperative only because of our human frailty. It is necessary to grasp this if we are to understand his doctrine in its fulness and to get rid of the view that his
;
moral philosophy is essentially harsh and Puritanic. Perhaps I may be allowed also to confess that when, some twenty
PREFACE
'
17
years ago, I wrote about moral philosophy in The Good Will\ I had not freed myself from the traditional interpretation of Kant, and I supposed that I was dealing with factors in the moral life particularly teleological which he had neglected or overlooked. factors fuller study has shown me my error an error which, I think, is by no means confined to myself. One of the last things which I discovered in my study of him, though it now seems to have been staring me in the face all the time, is that in his application of moral principles Kant takes into account most fully the desires and purposes and potentialities of men, and indeed that it is on a teleological view of man and of the universe that his application of moral principles is based. At one time I intended to add a further section on The Application of the Categorical Imperative'' and to examine the method of application employed in his neglected Metaphysic of Morals but I found that this would make my present book too long, and indeed it requires a separate book to itself. I hope that some student of Kant will one day undertake this very necessary task. I do not doubt that in so doing he will find in Kant's particular judgements considerable blemishes, due in part to personal limitations, but in greater part to the age in which he lived; but he will also find much of permanent value and if he approaches his instead of first attributing to Kant a mass of task sympathetically traditional absurdities and then complaining of inconsistency when he he will throw a Hood of light, not finds these everywhere contradicted only on Kant's method of applying his principles, but on the proper interpretation of these principles themselves. In these days when the pillars of European society are shaken, and when even the British race, after displaying a magnificence in action almost without parallel, is assailed on every side by prophets of unreason for whom moral splendour is so much illusion, I hope that this new interpretation of Kant's doctrine may help men to reflect more deeply on objective principles of action, on the nature of obligation, and on the value of the moral life.
wife for typing the whole of this difficult manuheavy burden laid on housewives by a six years' war and its oppressive aftermath. I must also thank Mr. W. F. R. Hardie, Mr. L. J. Beck, and Mr. C. B. H. Barford for their kindness in reading
I
have
to
thank
my
script in spite of the
the proofs.
Corpus Christi College, Oxford.
August, 1946.
H. J. PATON.
BOOK
THE APPROACH TO THE CATEGORICAL IMPERATIVE

CHAPTER
I
THE CRITICAL METHOD

i.
The need for understanding KanCs method. 1
of the reasons why philosophy is difficult is that in philosophical thinking we ought to know exactly what it is that we arc doing. This does not apply in the same sense to other thinkers it does not apply to the mathematician, to the physicist, or (if we may regard as also a thinker him who is so much more) to the artist. No doubt each of these knows very well what he is doing: he understands his problem, und he alone can solve it. But among the questions that he asks and answers, there is one question not asked the question 'What is mathematics?' or 'What is physics?' or "What is art? These are philosophical questions; and if the mathematician or the physicist or the artist asks these questions, he has taken the first step towards being something more than a mathematician or physicist or artist he is, in short, becoming a philosopher. If it is the part of a philosopher to ask these questions, it is still more the part of a philosopher to ask 'What is philosophy?' According to Kant, what was wrong with philosophy before his time was this that philoso:
ONE
phers blundered along trying to solve philosophical problems without ever asking themselves what it was that they were doing and whether what they were doing was something that could be done. He may have
been unduly hard on his predecessors most philosophers are but in any case this was what he thought. For him a sound philosopher must know clearly and reflectively what he is doing, and only so was philosophy likely to succeed. This is one reason why Kant's philosophy is called the Critical Philosophy. It is also one reason why we must begin by trying to see what Kant was doing as a moral philosopher. In other words we must begin by trying to understand the Critical method, although we are bound to understand it imperfectly until we have followed it in its actual
working.
1
Since this
is
a difficult topic, beginners are

II.
advised on a
first
reading to go
straight
on
to
Chapter
[19]
2O
2.

Moral judgements
are
[I
a priori.
:
We all make moral judgements that is, we judge some men to be good and some bad some actions to be right and some wrong. However much or however little such judgements influence our own actions and surely they do influence our actions to some extent we commonly regard these judgements as capable of being either true or false. Such is certainly the view of the ordinary good man. In this he may no doubt be mistaken, though it seems odd that so many men of so many races and so many ages should have fallen into the same mistake. There have been, and there still are, philosophers who tell us that moral judgements, if true at all, are true only as asserting that at a particular time we happen to experience a particular kind of feeling; or even that they cannot be either true or false, since they assert nothing, but are mere expressions of emotion, compar;
able to a cry of horror or of grief. Kant admits provisionally, at least in the Groundwork, that all moral judgements may be illusory, but he assumes that they at any rate claim to assert something, and he proposes to make clear what is implied in their assertion. Having done this, he hopes to be in a position to justify and defend, not of course every individual moral judgement, but the principles in accordance with which such judgements can be truly made. Even those who are sceptical about all moral judgements should therefore be able to follow his argument. There is a sharp difference between moral judgements and judgements of fact such as 'This carpet is red or 'That table is square.' The
1
judgements are based on sense they can be verified by means of our senses, or by what may be called sensuous experience. Moral judgements cannot be so verified. If we hold, as some sharp-sighted thinkers hold, that only judgements which can be verified by sensuous experience are true, we must either say that moral judgements are neither true nor false, or else that they are true or false only if, in spite of appearances, they are judgements about pleasure and pain and consequently can be verified by our sensuous experience of pleasure and pain. These theories, which may be called positivistic theories, rest on the supposition that only judgements which can be verified by sense are true a supposition which incidentally can itself never be verified by sense. Kant meets the difficulty in another way. Agreeing with the positivists that moral judgements cannot be verified by sensuous experience, but supposing also with the ordinary man that moral judgements can be true or false, he asserts that moral judgements if they are to be true must be a priori judgements. By this he means that they are judgements which are not based on sensuous experience, and he holds that such judgements may be true although they are not verifiable by sensuous experience. Judgements based on sensuous experience (and consequently verifiable by sensuous experience) he calls a posteriori or empirical judgements. To say that judgements are a priori in the sense that they are not based on experience or are independent of experience is thus a negative description. It does not mean that we make moral judgements before experience begins. It does not mean that an infant knows all about moral
latter
:
3]
it
THE CRITICAL METHOD
21
goodness before
begins to see colours and hear sounds. Such absurdities to Kant even by distinguished thinkers, but I do not propose to discuss them. Kant holds expressly that no knowledge is prior to experience in time and that with êxperience all knowledge
have been attributed
begins.
If,
well ask on
priori judgements are not based on experience, we may what they can be based. Are moral judgements possible, and if so, how? It is the main purpose of Kant's philosophy to answer these questions. At present we must ask 'What are the characteristics which lead us to suppose that there are a priori judgements at all?'
however, a
3.
The marks of a. priori judgement.
Empirical judgements arc always judgements of fact. Experience can us what is, and it cannot tell us anything more. Moral judgements tell us what ought to be, or what ought to be done, or what we ought to do. Such judgements are distinct from empirical judgements and cannot be
tell
inferred
fiom empirical judgements no one supposes that men always do what they ought to do; nor arc we justified in arguing as some
:
philosophers have donethat because men are self-seeking, it is therefore a duty to be self-seeking. Words like 'ought' and 'duty' and 'right' and 'good' indicate that the speaker is making not an empirical, but an a priori, judgement, if he is making a judgement at all. Moral judgements arc not, however, the only a priori judgements. Since experience tells us only what is, and not what must be, all judgements of necessity must be a priori they cannot be based merely on sensuous experience. Hence if we can say that a triangle must have its interior angles equal to two right angles or that an event must have a cause, this
:
must be an a priori judgement. Judgements of necessity may also be called truly universal judgements. When we say that an event must have a cause, we can equally say that every event, or that all events, must have a cause. Experience cannot give us such truly universal judgements. We can indeed say loosely that all swans arc white, and this is an empirical judgement. It means, however, merely that all swans hitherto seen have been white and this is only a compendious way of saying 'This swan is white, and that swan is white,' and so on till the list is complete. If it means more than this, it is a generalisation, the result of an inductive inference from the swans we have seen to other swans. The judgement may then be described as general but the sight of one black swan is enough to refute its claim to be universal. 'All swans are white could be a strictly universal judgement only if we could grasp a necessary connexion between being a swan and being white, and this we can never do merely by looking at white swans or indeed in any other way. Hence we may say that all judgements of necessity and all truly universal judgements are a priori and not empirical.
; ;
5
K.r.V.,
Ai, Bi. For a clear distinction between logical and temporal priority see
n.
K.r.V.,
A452
6480
n.
22

Kant
himself, as
[I
we
and so connects the two and the theoretical or scientific. But when we say that we ought to tell the truth, we suppose that it is possible not to tell the truth. Hence to say that men ought to tell the truth and to say that they must tell the truth are two quite different statements. Indeed 'ought' and 'must' seem to be opposed in such a way that when one is present, the other must be absent. Hence even although Kant's view gets some support from the fact that 'must' is sometimes used for 'ought' in ordinary speech, we must for the present assume that 'must' and 'ought' indicate different kinds of judgement. Both these words, however, are used to express judgements which are not based on sensuous experience, and so must be called a priori. The connexion, if there is one, between duty and necessity must be
considered later.
shall sec later, regards duty as a kind of necessity, kinds of a priori judgement the moral or practical,
4.
Composite knowledge.
Some
of our knowledge
is
partly based
on sensuous experience and
is
:
partly not so based. Such knowledge may be called composite knowledge it contains in itself both empirical and a priori elements. For example, we judge that the striking of this match caused a flame. Such a judgement is obviously empirical: it is dependent on sensuous experience. Nevertheless it is also a judgement of necessity it asserts a necessary connexion between the flame and the striking of the match, and such a necessary connexion is not anything that we see or touch. Ultimately the judgement rests on the presupposition that every event must have a cause.
:
Kant sometimes speaks as if experience were merely sensing, because sensing is an essential element in experience. Strictly speaking, however, his doctrine is that experience is composite it contains both empirical and a priori elements. Thus experience, on his view, is not merely a gaping at colours and sounds it is a knowing of real objects such as this red
:
have experience when we hold together square table. before the mind different sensible qualities as qualities of one object.
carpet and
this
We
This could not happen unless these qualities were given to sense. Bat it could not happen unless these qualities, given to different senses at different times, were held together before the mind in accordance with certain principles which cannot be given to sense at all and
equally
so are a priori. Among these principles he includes the principle that what today are called sense-data are changing accidents of a permanent substance, and that every event (including every change in the qualities of a substance) must have a cause.
Some of our ethical knowledge is equally composite. Thus if I can say that I ought not to kill John Smith, who is my enemy, my judgement contains empirical elements. I could not make this statement unless I
had experience both of myself and of John Smith, and unless I knew that John Smith was a man and that men are mortal beings whose lives can be brought to an end by artificial means. Nevertheless my statement
5]
THEGRITICALMETHOD
actually behave.
23
an obligation, and neither obligation in general nor this particular obligation can be known merely from experience of the way in which
asserts
men
5.
The
task
of philosophy.
According to Kant, the task of philosophy is to distinguish from one another the a priori and empirical elements in our knowledge, and to consider what is our justification for accepting the a priori elements. We have, for example, to consider by what right we assume that every event must have a cause. As regards ethics in particular, the task of the philosopher is to seek out, and if possible to justify, the supreme principle of morality. To put the question for a moment in its simplest form, we have to ask what is meant by duty or moral obligation, and what is our
supposing that we have duties at all. Such a question is very far removed from the question whether I ought or ought not to kill John Smith; but unless I can answer it, I am hardly in a position to give a satisfactory answer to the question about John Smith or even to a more general question about the rightness or
justification for
1
wrongiicss of killing. The question as to the supreme principle of morality or the nature of duty as such belongs to the branch of ethics which we may call 'pure or 'unmixed' or 'rational' ethics. The application of the supreme principle of morality to the problems of action presented by human nature may be called 'applied ethics. Such application clearly demands knowledge of
5
5
"anthropology and we should call 'psychology'. Strictly speaking, the Groundwork of the Melaphysic of Morals and the Critique of Practical Reason belong to pure ethics, though they may occasionally bring in problems of applied ethics by way of illustration. Kant's later work, the Metaphysic of Morals., belongs, in great part at least,
nature, which
human
Kant
calls
to applied ethics.
One difficulty in Kant's moral philosophy, as indeed in his theoretical philosophy, is the problem of the line to be drawn between the pure and the applied. Pure ethics, as 1 have said, must strictly be concerned with the supreme principle of morality alone; but Kant has a dangerous tendency to extend it in such a way as to cover what may be called moral laws, such as Thou shalt not lie.' 2 We ought to distinguish (i) moral principles; (2) moral laws, like the ten commandments, which 3 apply to men as men; (3) moral rules, such as the statement that it of a or an executioner to kill and (4) singular soldier may be the duty moral judgements, such as the judgement that I ought not to kill John Smith. Singular moral judgements cannot be a part of philosophy. Moral rules and moral laws alike must belong to applied ethics they all have a reference to human nature, as can be easily seen by considering
c
We shall
see later that this is not an adequate way of putting our question. See, for example, the confused statements in Gr., 389 5-6. Compare M.d.S.* Titgendlehre, 45, 468.
24
in
[I
what sense they could be binding upon non-human rational beings, such as angels, for example, are supposed to be. If angels are immortal, it is ridiculous to say that they ought not to kill one another. Only principles, supposed to be valid for all rational agents as such, can belong to pure ethics. An empirical element must enter in as soon as we begin to derive moral laws from our supreme principles.
6.
The needfor pure
ethics.
Kant is insistent on the need for pure ethics, and again and again he pours scorn on a merely 'popular' philosophy which mixes up the a priori and the empirical and gropes its way by means of examples. Such a procedure merits condemnation as being intellectually confused; but Kant maintains that it is also morally deleterious. The reason for the latter contention is this. A morally good action must not only accord with duty, but it must be willed for the sake of 1 duty. If we fail to grasp the nature of duty in its purity, we may be tempted to act merely for the sake of pleasure or convenience. Actions so grounded may at times accord with our duty, but at times they will be contrary to our duty and in any case they will never be morally good 2 actions, since they will proceed from a non-moral motive. Such a view may appear hard on the ordinary good man, who is
;
unaccustomed to philosophical abstractions. Kant, however, believes that reason in morals, as opposed to speculation, can easily be set on the right path. This is indeed only to be expected, since if we all have a 3 duty to do, it must be possible for us to know what our duty is. On his view, every man ought to have, and indeed has, however obscurely, a 4 pure philosophy of duty. Kant claims to do no more than formulate clearly the moral principle already at work in our ordinary moral judgements. 5 The need for an explicit ethical philosophy arises because the ordinary man may be tempted to let the pure principles of duty be obscured by the attractions of pleasure he may become sophistical about the strict laws of duty, and may seek to cast doubt upon their validity, or at least upon their strictness and purity. 6 This can be corrected only by pure ethics. As to the use of examples, Kant is far from repudiating this in the moral education of the young, provided it is directed towards separating the moral motive from motives of pleasure and self-interest. 7 The danger of basing ethical teaching on examples is this it may give the impression that the concept of duty is a generalisation from experience, a concept of how men actually behave; and this in turn may lead to a confusion of
human
than duty.
speaks rightly of the moral law, and this we shall find to be wider have spoken of duty, because we have not yet heard of the moral law (as distinct from particular moral laws). 2 8 = 25. Gr., 389-90 = 6-7. Gr., 391 = 8 404 4 M.d.S., Rechtskhre. Einl. II, 216 = 16; Tugendlehre, Vorrede, 376 =219. 6 Gr., 405 = 26. *K.p.V., 8 n. = in n. ( = 14 n.) 7 1 n. = 34 n. K.p.V.> 154 flf. = 303 ff. ( Gr., 41 275 ff.)
Kant himself
I
7]
THE CRITICAL METHOD
25
moral with non-moral motives, or even to the view that duty is a mere phantom of the mind. Kant's attack on this use of examples is supported 1 by three main reasons. Firstly, we can have no certain examples of moral
action, since the motives of human action are always obscure. Secondly, the moral law must be valid, not merely for men but for all rational agents as such, and no experience could entitle us to assert such a necessary law. Thirdly, examples must all be judged in the light of a priori principles, and consequently they cannot enable us to dispense with an examination of these a priori principles. Morality is not mere imitation, and examples serve only to encourage us ; that is, they may show our ideal to be not impracticable, and they may make it more vivid to our imagination. Of these arguments the first is at least plausible, if not altogether consistent with the alleged power of examples to show the practicableness of our ideals. The second must be reserved for consideration later. The third is, however, conclusive. Moral judgements are not judgements of fact and cannot be derived from judgements of fact. When we judge a man to be good or an action to be right, we are not making a statement of fact we are appealing to an a priori standard. It is absolutely vital to morality to justify such a standard and to recognise that it cannot be justified by any judgement of fact. We may of course hold that we have no binding duty, that the whole concept of duty is a mere illusion, a relic of primitive tabus, a cover for self-interest, and so on. What is philosophically deplorable is that we should continue to talk and act as if we had duties, and yet should explain them away as due to convention or tradition or self-interest. Only pure ethics can clear away this confusion and justify a belief in duty, if it can be justified at all. Otherwise confusion of mind may easily lead to degeneration in conduct.
:
7.
The method of analysis.
If pure ethics is not based upon experience, on what can it be based? Indeed if we cannot begin our enquiry from experience, how can we even
Undoubtedly we must begin, as Kant does himself, Trom the common use of our practical reason 2 and this means that we must begin with our ordinary moral experience, though Kant himself avoids this usage, presumably because for him experience is confined to experience of facts. It should also perhaps be said that moral judgements are not really ours unless to some extent they influence our actions. We act morally only in so far as we act according to moral judgements, and a justification of moral judgements will also be a justification of moral actions. It may be objected that we are now lapsing back into some kind of empirical ethics. This, however, is a mistake. Our moral judgements have an a priori element in them, and it is with this a priori element that we are concerned. What we want to discover is the a priori principle upon which moral judgements are based. This principle Kant believes to be actually
5
;
begin
it
at all?
Gr.,
406
ff.
ff.
Gr.,
406
28.
26
"
[I
used by the ordinary man as a standard for his judgements, although 1 What Kant it is not conceived by him in its abstract or universal form. seeks to do is to formulate this principle clearly before he considers its
Kant's method may be described as a method of analysis a priori element in ordinary moral judgements is made the whereby explicit and is examined in separation from the empirical element. Such an analysis must be sharply distinguished from other kinds of analysis. It is not to be confused with the analysis whereby we separate out common characteristics found in different individual things and use these common characteristics as the basis for empirical concepts. Still less is it to be confused with an analysis which professes to set forth the successive stages in the supposed development of moral judgements. It is on the contrary concerned only with the a priori element which must be present in moral judgements unless these are mere illusions. The a priori clement in moral judgements is expressed, as we saw
;
justification. Thus far
above, by such words as 'good' and 'evil', 'right' and 'wrong'. We may apply these words wrongly in particular cases but no moral judgement can be valid, unless the concepts corresponding to these words are valid. Hence in separating out the a priori element in moral judgements we are determining the condition, the sine qua non, of the validity of moral
judgements.
We so commonly identify the condition of a thing with its cause and regard the cause of a thing as outside it (either in time or in space or in both) that it may seem absurd to speak as if the condition of the validity of a judgement could be an element in the judgement itself. It certainly raises logical problems, the discussion of which belongs to a study of Kant's logic, not of his ethics. Nevertheless, it is surely obvious both that the concept of 'good' is employed in the judgement that St. Francis was a good man and that the judgement cannot be valid unless the concept of
5
'good
is
valid. 2
8.
Analytic and synthetic arguments.
If by analysis we have separated out the a priori element which is the condition of the validity of a moral judgement, we may be able, at least theoretically, to pass by a similar analysis from this condition to a still prior condition. For example, if we have separated out the concept of moral goodness, we may be able to say that no action is morally good unless it is done for the sake of duty. Acting for the sake of duty is then the precondition of moral goodness and consequently of all judgements in which moral goodness is predicated of individual actions. We can then
1 2
this implication of the analytic method characterises Kant's Critical Philosophy as a whole. Thus he deals in abstraction with categories which are both present in all experience and conditions of the validity of all experience. Failure to
Gr., 403 It should
24.
be noted that
grasp this makes
Kant
unintelligible.
Ig]
go on
to ask
THE CRITICAL METHOD
27
whether there is a still further condition apart from which no action can be done for the sake of duty. An argument of this kind is called by Kant both an 'analytic' and also a 'regressive argument. It starts from something conditioned (such as a particular moral judgement), and it seeks to go back to its condition, and from this to still remoter conditions. It is opposed to a 'synthetic' or 'progressive' argument, which goes in the reverse direction that is, which 1 starts from a condition and goes forward to the conditioned. The word 'regressive' may appear more appropriate than the word 'analytic' for the kind of argument we have described. The word 'analytic'
5 ;
has, however, this merit
it
suggests that the conditions (or series of con-
somehow contained in (or are somehow a part ditions) of) the moral judgements from which we start. How this can be so, and
explicit are
made
step in a regressive argument can be justified, are questions full of difficulty and requiring more examination than they receive from Kant. For example, it is not clear whether each step in the argument requires a separate intuition, nor how such intuition (if any) is to be characterised. Without such examination even the Critical philosopher has not attained his ideal of knowing exactly what it is that he is doing. Nevertheless we have at least the external plan of Kant's argument, and we must understand this plan if we are to follow each step as it is made. All this, however, is only one aspect of Kant's method, and a similar method may be found in other philosophers besides Kant. Indeed it may be doubted whether any other method is possible in moral philosophy. Perhaps, however, it should also be said that in the case of Kant the method involves a gradual clarification of statements which to begin with are relatively vague and obscure. Kant does not believe, as some philosophers do, that we can begin with clear definitions and gradually build a structure on this foundation. He believes, on the contraiy, that in philosophy adequate definitions can be given, not at the beginning, but 2 only at the end, when our analysis is complete.
how each
9.
The
Critical method.
It looks as if the method of analysis, even if we are justified at every step of the argument, can give us no more certainty than that of the original judgements from which we start or at least as if it can give us only the certainty of bringing separate judgements under one principle, the certainty which belongs to a system rather than to isolated judgements. the other hand, there is the possibility that in a regressive series of judgements we might be able to do more than see that the truth of each new judgement is the necessary condition of the truth of the previous judgement we might also have the same kind of insight into the truth
;
On
and 4 (263 and 274-5), This distinction must never be confused the distinction between analytic and synthetic judgements or (2) the distinction between the analysis and synthesis which are present in all judgements. See K.M.E., I, 130 and 219. 15 n.). 113 n. ( 3759; K.p.V., 9 n. 2/T.r.F., A73i
-
See ProL, Vorw.

(i)
with either
28
of each
[I
new judgement as we had into the truth of the judgement from which we started. This is, I think, Kant's view of what actually takes place. Finally we might arrive regressively at a judgement into whose truth we had some special kind of insight, a judgement whose truth was
established with complete certainty. In such a case our whole argument greatly strengthened, and our previous judgements would be should be able to reverse our direction corroborated and justified. and proceed progressively (or synthetically) to our original starting point
would be
We
without falling into a vicious circle. The most obvious way, if not indeed the only way, in which our argument might be thus strengthened would have been found if in our regression we could arrive at a judgement whose truth is self-evident. This may, in a sense, be Kant's own view, but if so, he holds it with a difference. He believes that a philosopher should be very wary of alleged self-evident propositions. 1 They are too often the deliverances of common sense or tradition and are, in his language, the cushion of a lazy philosophy. He admits, it is true, that a priori principles (Grundsatze) are so called, not merely because they contain the grounds (Griinde) of other judgements, but also because they themselves are not grounded in any higher or more universal cognitions of objects. Nevertheless he holds that all such principles (apart from those that are tautologous) must be subject to criticism and require a proof or at least a justification (which he calls a 'deduction'). This applies, in his view, even to the principles of mathematics, which he believes to be grasped by direct intuition and to possess apodeictic certainty. Even so, we have still to ask how knowledge of this kind is possible, and we have still to explain and justify this possibility. 2 How can we offer such an explanation and justification of a priori
knowledge?
istic
It
is
at this point that
we come to
the distinguishing character-
of Kant's philosophy. He holds that when we come to a priori principles which are not grounded in any higher knowledge of objects we must forgive him for making the ground higher than what rests upon it we still have to consider their subjective origin in the nature of reason itself and so to justify them and explain their possibility. This is why his greatest works are called the Critique of Pure Reason and the Critique of
and this is the main ground for calling his philosophy the ; Philosophy. It must not be thought that whenever he arrives at a first principle which appears to be self-evident, he merely asserts that our reason, or some other faculty, is capable of grasping such a self-evident principle. To say this might be true, but it would be unimportant and would carry us no further in the way of explanation or justification. If he is to be successful, we must be able so to understand the necessary working of reason that the possibility of grasping such a first principle is rendered
Practical Reason
Critical
intelligible.
Kant's procedure is so subtle and complicated, and it varies so much with different problems, that no general account of it can be given here.
1
JfC.r.
V.,
A233
= 6285-6.
K.r. V.,
IO]
I
THE CRITICAL METHOD
29
have so far given is unduly simplified though not, I hope, as regards his moral philosophy. All we need say here is that his method has at least an initial plausibility. We certainly cannot go beyond or behind the activity of our own self-critical reason, and it may well be supposed that the working of reason is, in a very special sense, transparent to itself. On the other hand his view has many difficulties, and it apparently introduces us to a new kind of self-evidence, which, however, Kant seems to take for granted rather than to discuss. When we follow an analytic argument of Kant, it is well to remember that amid its apparent twistings and windings it is always moving deliberately towards some possible explanation or justification in the nature of
reason
itself.
Indeed the account
10.
The method of synthesis.

analysis
is
The method of
at a first principle,
we must be
the method of discovery, but if we arrive able to reverse the process ; and if our
first principle is independently established, whether on grounds of selfevidence or otherwise, we have a new kind of argument which, as we have seen above, may be called 'progressive' or 'synthetic'. The synthetic method is the method of exposition rather than of discovery. This in turn raises logical problems, and the very word 'synthetic' suggests that in our descent from a first principle we must bring in further elements not themselves contained in the first principle. It looks as if we re-create the whole which hitherto we have been breaking up into a series of logically dependent parts. This in turn might mean that we cannot make our descent unless we have previously made an ascent. 1 Here it need only be noted that the same subject may be treated by an analytic argument in one book and by a synthetic argument in another; or both methods may be combined in the same book. Thus in the first two chapters of the Groundwork Kant proceeds analytically from common knowledge to the supreme principle of morality the categorical imperative. In the third chapter he proceeds synthetically from the examination of this principle and its sources that is, its sources in practical reason itself to
the common knowledge in which it is employed. Such at least is his own account of the matter, 2 and it is substantially correct, although he does not in fact pay any attention to the lower stages of the descent.
11.
The importance of Kant.

those
who are influenced by modern theories of empiricism and Kant's procedure must inevitably seem to belong to an obsolete positivism rationalism. Historically it is a mistake to consider him as primarily an
To
Compare
the Idea of the

8
Plato's account in the Republic of the way in which dialectic mounts to Good and then descends ; and also Aristotle's account of the difference
between arguing
Gr.,
to
392
an apxî and arguing from an
apx?? (Eth.
jVwr.,
1095 a 34)
9.
3O

:
[IH
exponent of rationalism
his great service
was to break away from a one-
sided rationalism and to do justice to the empirical element in human knowledge. On the other hand he was also endeavouring to answer the one-sided empiricism of Hume in spite of his very great admiration for that philosopher. His central doctrine, as we have seen, is that we must distinguish clearly the empirical and a priori elements in our knowledge in order to do justice to both; and he considers it as much an error to 'intellectualise' empirical appearances with the rationalists as to 'sensify' the concepts of the understanding with the empiricists. 1 It is only by a combination of the empirical and the a priori that we can have knowledge. It is precisely in this that his importance is to be found for the present day. He believes that an unmitigated empiricism is bound to end in a complete scepticism and that the only way to avoid this is to consider the activity which belongs to reason in its own right. At a time like the present, when so-called knowledge is being reduced to the apprehension of tautologies and the reception of sense-data, it is hard to see how the world can be made intelligible even as intelligible as it seems to the
ordinary man. It is still harder to see how we can have any principles of conduct and indeed the problem of moral action tends to be treated or explaining away moral beliefs in terms as a problem of explaining of a theory of knowledge which has been adopted on quite other, and purely theoretical, grounds. The result must inevitably be that our wills, instead of being guided by intelligible principles, are delivered up to mere caprice or self-indulgence or tradition or even fanaticism, the last three being merely particular forms of caprice. And the whole doctrine must lead, as it already did in the case of Hume, to scepticism as regards the very existence of such a thing as a human mind. However much we may respect the thoroughness of empirical philosophers in working out their theory to its logical conclusions, it is not easy to believe that men will long be satisfied with their account of knowledge and above all with their account of moral action. Kant believes that they can be answered only by an examination of the activity of reason itself, and his attempts to do so are at least worthy of serious study. This is particularly true of moral philosophy for apart from questions of purely theoretical interest it is a bleak look-out for the world if we are all to be convinced that reason has no part to play in action. From another and non-empirical point of view Kant is criticised on the ground that his work has been superseded by Hegel and his followers. I have no wish to deny that Hegel and others have found weaknesses in Kant's doctrines (including many of which he was never guilty) or that they have made advances on him in certain respects. On the other hand Hegel in particular has added a good many mistakes of his own, and his school is far from rinding general approbation at the present time. If we want to avoid these mistakes, it may be a good thing to go back to Kant, who started the whole modern movement, and try to see what his doctrine for I agree with Professor Schilpp that many of the absurdities really was
; ;
K.r.V.,
Aayi
6297.
IAPP.]
usually attributed to
THE CRITICAL METHOD

him
3!
are the fault of his interpreters. 1 good understanding of Kant may well be, as it has so often been before, the condition of a new advance, and one of which we are in great need at the present time.
APPENDIX
THE DIVISIONS OF KANT'S MORAL PHILOSOPHY

IT does not matter much what names we give to the different parts of Kant's philosophy, and I have ignored his account of the relations between logic, physics, and ethics but a brief note may be added on the difference between the Groundwork the Critique of Practical Reason, and
;
the Metaphysic of Morals. The Groundwork may be regarded as setting forth the core or kernel of a critique of practical reason it considers only some of the topics peculiar to a critique and ignores others, such as the relation of practical to theoretical reason, a topic not of immediate importance for moral philosophy. Kant believed that practical reason was less in need of a
:
critique than pure reason, and he may have intended to go straight on to a metaphysic of morals. If so, the general failure to understand the Groundwork may have led to a change in his plan. In any case he published
the Critique of Practical Reason three years later (in 1788), while the Metaphysic of Morals was not published till 1797, when he was already 73 years old. Kant commonly regards his two great Critiques as either propaedeutics to, or parts of, the corresponding metaphysic. Neither of these views is satisfactory, since if we suppose metaphysics to be pure a priori knowledge, there is little or no such knowledge outside of the Critiques themselves. The metaphysic is supposed to fill up into a complete system the framework set forth in the critique but such fillings up are made only by bringing in empirical elements. We can, if we like, regard a critique of practical reason as concerned only with the origin of the categorical imperative in practical reason, and a metaphysic of morals as concerned with the different formulations of the categorical imperative. Kant himself suggests the this by his titles for the second and third chapters of the Groundwork second chapter being called a 'transition' to the metaphysic of morals, and the third chapter a 'transition' to the critique of pure practical reason.
;
But such a division is somewhat arbitrary, and is not strictly adhered to in practice. It seems to me better to regard a critique of practical reason and a metaphysic of morals as the same thing.
Kant's
says
1
it
Metaphysic of Morals is something different from what he to be, namely pure ethics. He explicitly recognises that it ought
own
contains empirical elements;
arid
he
calls
*
his
Theory of
Law and
Kant's Pre-Critical Ethics, pp. xiii-xiv.
M.d.S. t Rechtslehre, Vonede, 205.
32

but
[I
APP.
the two parts into which it is divided not a meta'metaphysical rudiments' (Metaphysische Anfangsgrunde) Exactly the same thing happens to the Critique of Pure Reason it is not followed up by a Metaphysic of Nature, but merely by the Metaphysical Rudiments of Natural Science. In both cases the phrase 'metaphysical rudiments' indicates the presence of empirical elements. Kant may have thought that more of his Metaphysic of Morals was pure ethics than is in fact the case. It does contain some ethics which may be called pure, but in the main it is applied ethics. So far, applied ethics is concerned with the moral laws and moral rules which arise when we apply the supreme moral principles to the special conditions of human nature. Kant himself, however, speaks of applied ethics in a second and different sense; or perhaps he uses the term in a confused way covering the two different senses. 1 In the second sense 'applied ethics' is used for a special kind of moral or practical psychology (or anthropology as he calls it) concerned with the conditions which favour or hinder the moral life. It would be extremely useful both for education and for the guidance of our own lives if we could have an authoritative psychology of this kind. In its absence many parents today do not know to take a crude example whether discipline or a complete absence of discipline is the more likely to encourage moral excellence in their children. There is, however, no reason why we should regard such a 2 psychology as practical it is a theoretical examination of the causes of certain morally desirable effects. Still less is there a reason why we should 8 regard it with Kant as a kind of applied or empirical ethics. The application of moral principles is not only dependent on a psychology specially limited to the conditions favouring or hindering the moral life: it is also dependent on general psychology as knowledge of human nature. Without psychology, and indeed without empirical knowledge both of human nature and of the world in which we are, the application of moral principles would be impossible ; and the more of such empirical knowledge we have, the better shall we be able to make sound moral judgements. This does not, however, alter the fact that the knowledge of moral principles, if there is such knowledge, is something quite distinct from empirical psychology, and cannot be derived from
Theory of Virtue
physic,
empirical psychology. To paraphrase Kant pure Ethics cannot be 4 grounded on psychology, but it can be applied to psychology. There is, however, a further question whether the philosophy of moral action should not be preceded by a philosophy, or a philosophical psychology, of action as such. Before we consider what Kant calls a pure will, ought we not to examine the nature of will as such? Kant himself, before he examines what he calls pure thinking that is, the a priori element in our knowledge of objects has before him in formal logic a philosophical account of thinking as such, and indeed his account of pure a priori
Gr., 410 n. == 33 n. ; K.r.V., ASS == 879; M.d.S., Rechtslehre, Einl. II, 217 = 16-17. Applied ethics can be regarded as parallel to the applied logic of jRT.r.F., A52 ff. =677 ff. 8 8 See Gr., 388 = 4. K.U., Erste Einleitung, I, 7. * M.d,S. = 16. t RecktsUhn.* Einl, II, 217
1
IAPP.]
rests to
THE CRITICAL METHOD
33
a considerable extent on the doctrines of formal logic. thinking The parallel may not commend itself to critics, 1 and there may be differences in the two cases; but at least it serves to raise the question. A philosophy of action as such would be comparable to the Philosophic, Practica Universalis of Wolff, which Kant condemns for mixing up empirical and a priori elements 2 but the fact that Wolff botched his job does not show that the undertaking is superfluous. Kant indeed argues that for moral philosophy it is not necessary to consider such topics as the grounds for pleasure and displeasure the difference between the pleasures of the senses and those of taste and moral satisfaction the way in which desires and inclinations spring from pleasures and pains, and in turn, with the 3 co-operation of reason, give rise to maxims for action. Nevertheless he does not wholly ignore such topics in his own moral philosophy, and it can hardly be doubted that he takes for granted a good many doctrines of this kind which it would have been well if he had set forth specifically. He does make contributions to this subject in his Anthropologie and elsewhere but there is much to be said for the view that if he had set forth in detail a sound philosophy of action, this would have been a great help towards understanding his moral philosophy and might have modified it
; ; ;
;
in certain respects.
The
standing.
*
traditional criticisms of Kant on this point were based on complete misunderCompare Dr. Klaus Reich in Die Vollstandigkeit der Kantischen Urteilstafsl a
work which supersedes

Gr.y
390
somewhat
Gr. 9
previous discussions on this topic. M.d.S., Rechtslehre, Einl. IV, has a different character.
all
7.
Th* philosophia practice universails of
427
54-5.
CHAPTER
II
THE GOOD WILL

i.
good will
is
good without
limitation.
KANT begins his argument dramatically. 'It is impossible/ he says, 'to conceive anything in the world, or even out of it, which can be taken as 1 good without limitation, save only a good will.' it This, it must be confessed, although professes to be an ordinary moral judgement, is not the kind of utterance in which the ordinary good man habitually indulges: it is already more like the statement of a principle which the ordinary good man may be supposed to follow in his judgements and actions, even if he does not formulate it explicitly. Nevertheless the question raised is one for ordinary moral insight ; and without further analysis the statement is too vague to be regarded as a satisfactory
moral
principle.
(i)
We
very
much want
limitation'
to be told, for example,
what
is
meant
The first phrase offers little difficulty. To put Kant's assertion in other terms we may say that a good will alone can be good in itself, or can be an absolute or unconditioned good. Words like 'absolute' and 'unconditioned' may today be obscure to the simple and repulsive to the sophisticated, but in the present context they are not difficult to understand. All Kant means is that a good will alone must be good in whatever context it may be found. It is not good in one context and bad in another. It is not good as means to one end and bad as means to another. It is
somebody happens to want it and bad if he doesn't. Its not conditioned by its relation to a context or to an end or to a desire. In this sense it is an unconditioned and absolute good it is good in itself and not merely in its relation to something else. Its goodness is not limited to goodness in this or in that relation it is, in short, good without limitation or qualification or restriction. But what is meant by 'a good will'? This indeed is one of the questions which it is the purpose of our enquiry to answer ; but at present we are concerned with 'a good will' as it is recognised and judged by ordinary
not good goodness
if
is
: :
by 'good without
and
(2)
by
'a
good
will'.
moral consciousness. Hence the conception is necessarily vague, and it is 2 deliberately left so at first by Kant. We might, perhaps not improperly, but moral this 'a describe it as will', might have misleading associations. For example, many people today might regard a moral will as one which is bound by what they call 'conventional morality', and so as intellectually blind and practically misguided. Such a will, however, is then
1
Or. 9 393
1 1.
K.p.V., 62-3
183
10).
[34l
11
2]
THEGOODWILL
35
good but genuinely bad. What we are conwe can regard as genuinely good. If we admit, provisionally, that there can be such a thing and that it can be recognised, or at least conceived, then it is not unplausible to say with Kant that it must be good in whatever circumstances it may be found, and that therefore it must be an absolute and unconditioned good.
regarded as conventionally cerned with is a will which
2.
Other possible views.
Even on this semi-philosophical level there are further questions about the nature of a good will which demand at least a provisional answer. But before considering these, it may be well to examine, first of all, some possible alternatives to Kant's doctrine. Some theories which would render Kant's assertion meaningless or trivial or false must be here passed over. We might say, for example, that since the word 'good' means nothing, 1 Kant's assertion about a good will can mean nothing. This view seems to me an ingenious paradox based on logical prejudices rather than on a disinterested examination of moral judgements. Again we might say that since 'to be good' means *to be pleasant', Kant's assertion is certainly trivial and probably false. This is an ancient fallacy requiring no refutation here. I propose to assume that 'good' means something, and that it does not mean 'pleasant'. On this assumption Kant's statement has two sides, a positive side and a negative side. He asserts positively that a good will is good without limitation, and negatively that nothing but a good will is so. As against his negative assertion, we may affirm that other things than a good will are good without limitation; we may even affirm that all things which are good (in at least one sense of the word 'good') are good without limitation. As against his positive assertion we may deny that a good will is good without limitation; we may even deny that anything is good
without limitation. Let us first consider the position of goods other than a good
3.
will.
Are
all goods
good without limitation?
examination would here be out of place, and it will be sufficient if we can show that Kant's view in opposition to it is at least not unreasonable. 2
possible alternative to Kant's view is that all goods (in one sense of the word 'good') are good without limitation or are absolute and unconditioned goods. This view has very strong support at the present time. brief discussion of this view is hampered both by the fact that in ordinary usage the word 'good' has many meanings, and also by the fact detailed that different philosophers express themselves in different terms.
1 The ablest exposition of this view will be found in Ayer, Language, Truth and Logic, Chap. VI of the first edition. 2 In The Philosophy ofG. E. Moore -I have discussed this topic in an essay entitled The Alleged Independence of Goodness. Compare also my paper on Kant's Idea of the Good in
the Proceedings of the Aristotelian
Society,
1944-5, pp. i-xxv,
36
[H
If we consider various kinds of thing in isolation that is, in abstraction front their context we may properly say of them that they are good. Among these kinds of thing may be included, for example, pleasure, knowledge, and art. Most people would be inclined to agree tKat all IKese kinds of thing, when thus considered in isolation, are good, or even good in themselves. They are at least what Sir David Ross might call
prima facie goods.
The view we are considering maintains that whatever goodness these prima facie goods have in isolation, that very same goodness they must have in any context in which they may be found. In other words all such prima facie goods are absolute and unconditioned goods. In this respect therefore a good will is by no means, as Kant supposes, unique. Such a view is not, I think, consistent in all cases with our ordinary judgements. Thus pleasure is manifestly a prima facie good ; yet in some contexts it is thoroughly bad for example, pleasure in the pain of others seems thoroughly bad. Indeed it is not unreasonable to hold that the actual goodness of these prima facie goods generally varies with their context. It is not merely that art and knowledge are no good to a starving man. In such a context as the burning of Rome, one would have to be a Nero, and not a reasonable man, to think that art was a good thing to pursue. In certain contexts art and knowledge are out of place. It may be objected that even in such contexts it is not art or knowledge, but rather, the pursuit of them, that is bad. Perhaps. But when we say this, are we not merely judging art and knowledge in isolation over again? Whatever these things are in isolation, they are not actually good or bad
;
except as in some sense pursued. The goodness supposed to belong to things considered in isolation is an abstraction; and the only real goodness is the actual concrete goodness of a thing in a concrete situation. However this may be, there is a plausible explanation of the way in which the goodness of things seems to vary with their context, although, as we are told, in reality it does not so vary. The goodness which varies, it is held, is not the goodness of the thing itself, but a quite other goodness which belongs to it as a means, or else the goodness (other than its own) which it contributes to a whole of which it is a part. We must distinguish its intrinsic goodness from (i) its goodness as a means (that is, its utility)
and from
confusing the goodness of a thing with its mere usefulness for something else. It is also true and important to say that the goodness of a whole is greater than the sum of the goodness of its parts taken in isolation from one another and from the whole. The part may contribute to the whole more goodness than it has when it is considered in isolation, that is, when it is not a part. So far so good. But the question is whether the part in thus contributing an extra goodness to the whole is not itself immediately better than it would be if it were not a part, that is, if it were in a different context. It can hardly be doubted that the part in contributing this extra goodness to the whole seems to be more valuable than it would be in
call its contributory goodness. (2) what we It is certainly true that we must be on our guard against
may
II
3]
THEGOODWILL
37
For some purposes it may perhaps be convenient to divide up its apparent goodness into a goodness which belongs to it intrinsically and a goodness which arises from its being a part of this whole although such a division is a matter of theory and does not result from an immediate
isolation.
judgement of value. What
is questionable is the assertion that the extra contributory goodness belongs not to the part but to the whole.The whole is made up of parts, and it has no goodness which is not manifested in its parts (that is, in its parts as parts and not as isolated existents). The goodness contributed by a part to the whole must also be manifested in the parts and above all, one would think, in the part which makes the contribution. This indeed seems the direct utterance of our judgement of value, as is very obvious, for example, of the beauty which is contributed to a poem by the use, in a particular context, of a word which may have no great beauty in itself our aesthetic judgements throw a great light on this problem. Furthermore, if ex hypothesi the part can add an extra goodness to a whole and so to the other parts, it seems impossible to deny on principle that the whole, and so the other parts, can add an extra goodness to the part. That is, the goodness of the part may vary with its context and this, however much we may divide up goodness in theory, seems, as I have said, to be the direct utterance of the judgement of value. In our actual concrete judgements, while we can distinguish the goodness of a thing from its utility (or from the goodness of its consequences), we canrot, I think, so distinguish the goodness which it has in itself from the goodness which it has as part of a whole. Hence the general allegation that all goodness (in one sense) is in1 dependent of its context is not sufficiently grounded to dispose of Kant's view that the goodness of a good will is unique in being independent of its context. When we say that a good will is good without limitation, we do not mean merely that in its actual goodness we can by abstraction distinguish a goodness which would equally belong to it in isolation. Kant,
;
its
at least, hqliisjthat^,a.gop,d will, if it is good withQut.Umitation,Jp[iust have full worth in itself, and that this worth cannot be diminished or
increased, and cannot be outweighed or dimmed, either by any con2 sequences or by the varying contexts in which it may be found. Such a Statement would not be true, for example, of pleasure; for even if we think it proper to attribute to pleasure in a concrete situation a goodness
which would equally belong to it may be more than outweighed by the badness which arises from its The question on which we have to
in isolation, this prima facie goodness the badness of its consequences or by presence in this particular situation. make up our minds is whether or not Kant is right in finding this difference between a good will and all other things which we recognise to be at least prima fade good.
1 1 have assumed that this doctrine means more than an intention to call 'intrinsic goodness* only the goodness which, as present in isolation, is attributed to the thing in whatever context it may be found. If the matter were only one of terminology, Kant would be entitled to use his own. 2 394 =13. See 6 below.
&
38
4. Are

many goods good without
limitation ?
[H 4
Even if we refuse to accept the general principle that all goods (in one sense of the word 'good') are good without limitation, it is nevertheless possible that some goods other than a good willjmay be goo^jvdt|jojit
limitation. KaîFTnTniis^
rejects, the
5
this
possibility. "He
examines, and
view that certain things which 'are without doubt injmany respects goodjand desiraBTe are good without limitation in the same way aFa good will. ^ We are now concerned, not witB their goodness abstraction or their abstract goodness we are concerned with their goodness in
concrete situations or their concrete goodness. The principle upon which Kant selects these prima facie goods for consideration is not altogether clear. 2 He appears to be interested, at least primarily, in goods which are not themselves the product of a good in certain will, although a good will might aim at their development or cases at their attainment. Thus he mentions first gifts of nature. Under these he includes mental talents, such as intelligence, wit, and judgement
(to which we might be allowed to add humour, imagination, and artistic 3 capacity) ; and qualities of temperament, such as courage, resolution, and perseverance. He then passes to gifts offortune, among which he includes power, riches, honour, health, and happiness. Finally, he returns to what are apparently also qualities of temperament, special qualities which he considers peculiarly helpful to a good will and so liable to be regarded as themselves possessing unconditioned value. These are moderation in affections and passions, self-control, and cool deliberation. His criticism of these claimants to absolute goodness is in all cases the same. However good these things may be in some respects, they are not, like a good will, good in all respects and in all relations. On the contrary, when they accompany a bad will, they are themselves bad. They are then bad, not merely as producing bad consequences, but as themselves the source of an additional badness in the whole of which they are parts. 4 The coolness of a villain makes him not only far more dangerous it makes him also immediately more abominable in our eyes than he would have
:
been without
it.
This contention appears to me to be sound; and presumably on Kant's view it would apply equally to other prima facie goods which he does not consider, such as skill in arts and sciences, taste, and bodily 6 agility. None of these things is an absolute or unconditioned good. We may, however, put Kant's doctrine more positively. For him all
Klaus Reich suggests that N.S., Vol. XLVIII, No. 192). 8 See Gr., 435 = 66.
4
*&> 393-4= 8
The
1 1
-i 2.
it
is
due
to the influence of Cicero's
De
Officiis
(Mind,
distinction
is
clearly
made
in this discussion, although
when Kant
discusses
kinds of good abstractly, he uses only the distinction between *good as means' and 'good in itself* presumably on the ground that when anything makes worse the whole of which it is a part, it is then bad immediately (or in itself) in that context. Gr., 414 38. 6 6 Gr. iv n.). 9 394 =12. The italics are mine. Religion) Vonede, 4 n. (
II
5]
THE GOOD WILL
39
these prima facie goods are conditioned goods, and the condition of their 1 goodness is said to be a good will. This would be absurd if it meant that in a bad man nothing not even artistic or scientific activity could be
says that a conditioned good its use is nof contrary to tfie
expressed more accurately when he one which is gqgdQXiJy.on condition that moral law. 2 This would dispose of the view that Nero's fiddling was a good. In the case of happiness perhaps Kant's view varies unconsciously. He always recognises that happiness is in many respects good, 3 and even holds that in man reason has an office which it cannot refuse, the office of seeking happiness in this world and where possible in the next. 4 In the view of reason happiness as a good knows no other limitation than that which springs from our own immoral conduct. 5 But he also says that reason does not approve happiness (however much inclination may desire it) unless it is united with moral behaviour. 6 For the sake of consistency as well as truth we must
good. Kant's meaning
is,
I think,
is
say that
iniquity.
:
all
happiness
is
good except
in so far as
it
is
happiness in
In all this we must remember that Kant is not offering us a general theory of value, but only a theory of goodness. Art and science have their own standards of excellence, but these are standards of beauty and truth, not of goodness. Success in the pursuit of art or science is no doubt a conditioned good; but this, as we have seen, means for Kant that it is good only if it is not contrary to the moral law. No matter how well Nero fiddled, what he did was not good. Indeed it might be argued though I do not attribute this to Kant that the better he fiddled, the worse his action was. To paraphrase Kant the scientific skill and artistic gifts of a villain may not only make him more dangerous they may make him also immediately more abominable in our eyes than he would have been without them.
:
Once the question is put clearly, ordinary judgement would, I think, agree with Kant that even very great goods of the kind we have considered are not necessarily good in any and every context. That is, they are not
good without
limitation.
5. Is
a good will good without limitation?
We can now return to the alleged absolute goodness of a good will. This might perhaps be denied on the general ground that no good can be good without limitation. It might be held that the nature of everything must depend on its context, and that consequently the goodness of
everything must depend on
1
its
context.
From
this it
must follow that even
&> 396
is
but that
1
=15. Their goodness another question.
may
also
be conditioned by our desires and needs;
2 rv n.) . It is stated less clearly when Kant says (Gr., 397 Religion, Vorrede, 4 n. ( 6) that the concept of a good will always stands first in estimating the whole value of any action and constitutes the condition of all other value.
*Gr., 393
*
*
.
A>.K, A8i 4
K.p.V., 61
B8 4 2.
Kj.V.> A8i 3
181
108).
6841,
4>O
[I I
the goodness of a good will must depend on its context, and that it is therefore absurd to speak of a good will as having the same unique and
absolute goodness in any and every context. Since views of this kind have little acceptance at the present time, we may perhaps be allowed to pass over this a priori objection and to consider the goodness of a good will in the light of our ordinary
judgements.
We are concerned, as I have said, only with a genuinely good will. We need not concern ourselves with busy-bodies and killjoys, with scribes
;
good an unbalanced pursuit of moral ideals but these emotions would have found some outlet anyhow, perhaps as in the case of German antiSemites with even worse results. Morality, like religion, may attract to itself unbalanced emotions, but a genuinely good will will seek to moderate and control, not to encourage, such emotions. If we set aside men whose professed goodness springs from moral badness or whose genuine goodness is overlaid with emotional abnordo a great deal of harm which, mality, is it not true that good men may as is commonly said, has to be undone by the wise? And may not even a genuinely good will be at times out of place? It is certainly true that good men may do a great deal of harm and this harm may spring, not from officiousness and vanity (which belong to moral badness), but from mere silliness or stupidity. In intellectual matters Kant shows a very good understanding of stupidity, which he 2 describes as a lack of judgement; but in moral matters he tends to underrate the need for judgement and discrimination in deciding upon 3 the right course of action. I think, however, he would hold that the harm done by a stupid good man was due to his stupidity and not to his will as such cannot issue goodness he certainly seems to hold that a good 4 in wrong actions. In any case it would be paradoxical if we maintained that it is better to be a stupid bad man than a stupid good one and even if the stupid good man may produce some bad results which would not be hold that his good produced by the stupid bad man, Kant would still
in
:
fever; or as the 'fantastically virtuous', who can admit nothing whatever 1 to be morally indifferent. But perhaps we may agree with Kant that even such types belong to the pathology of human nature. The presence will in them means that their excessive emotions are manifested of a
and pharisees, with fanatics and persecutors, so far as their activities are for such men are manifestly inspired by envy and pride and malevolence not possessed of a genuinely good will. These types may indeed shade off Kant as 'moral enthusiasts', who gradually into other types described by allow a lively sympathy with goodness to degenerate into a passion or
258; 433 n. Compare also fC.p.V. 157 M.d.S.> Tugmdlehie, Einl XVII, 409 though perhaps in a some280). Kant himself, however, denies elsewhere ( 23 ( 9). what different sense that actions can be morally indifferent ; Religion, 22 An interesting study of the moral and religious enthusiast is to be found in Heaven's
l
306-7
My
by Thornton Wilder. s = 24. E.g. Gr., 403 A>.K, A 1 33 n. = 6172-3 n. * He the motive of a implies this by arguing that motives other than duty, which is = 34. good will, sometimes issue in right actions and sometimes in wrong. (Jr., 411
Destination,
a
II
6]
THE GOOD WILL
41
considered philosophically it is artificial. A man with a genuinely good will cannot but seek to develop such intelligence as he has especially that intelligence in practical matters which is commonly called wisdom. 2 And a genuinely wise man cannot but seek genuine moral goodness. This is not altered by the empirical fact that a well-meaning clergyman may talk nonsense about economics. We must not regard a good will as a mixture of moral enthusiasm and good will of this kind would be practical folly and personal conceit. out of place, not sometimes, but always. The only kind of situation in which a genuinely good will may reasonably seem out of place is the situation which calls for spontaneous emotion or mutual entertainment without the overriding thought that this is a binding duty. Kant may not have given enough place for such spontaneity few moralists do ; but he was far from supposing, either in theory or in practice, that a good will could not manifest itself in the virtues of social intercourse. 3 It is a mere prejudice to identify a good will with sour-faced morality; and if we abandon this prejudice, it may reasonably be said that of all things in the world a genuinely good will is the only one which is never out of
had a unique and incomparable value which could not be out1 weighed by the evils resulting from his natural defects. The antithesis between goodness and wisdom is useful enough to describe an empirical situation with which we are only too familiar, but
will
place.
It
cribed
remains, however, an open question whether a good will as desby Kant is a genuinely good will.
6.
The highest good.
If a good will is the only thing which is an unconditioned or absolute good in the sense that it must be good in every possible context, can we go on to assert with Kant that it must therefore be the highest good? The phrase 'highest good' is ambiguous. It may mean merely the 4 good which is itself unconditioned and is the condition of all other goods. In this sense 'highest good' and 'absolute good' mean precisely the same thing. But Kant is also making a judgement of value for such a good is to be esteemed as 'beyond comparison higher' than any other good. 5 Its usefulness or fruitlessness can neither add to, nor subtract from, this unique and incomparable worth. There seem to be two questions here 'Are we, first of all, entitled to say that only a highest, and indeed an incomparable, good can be good in every context?' And, secondly, 'Is such an inference confirmed by a direct and enlightened judgement?' It is difficult to estimate an inference of the kind suggested in our
;
:
Gr. t
394
2 3
4
Compare
12-13. M.d.S., Tugendlehre, Einl. VIII, 391
ff.
237
ff.;
and
*
19,
444 ff.
is.
M.d.S., Tugendlehre, 48, 473 ff. =r 246 ( 198) ; Gr., 396 K.p.V., 1 10
15.
Gr. t
394
42
first
THECATEGORICALIMPERATIVE
question. If
[H 6
speaking of something like air, for example, utility for man and other animals, we could not say that air, which is good always, is better than exercise, which is good only sometimes. But manifestly the utility of anything does depend on its context namely, the needs of living organisms and it cannot offer a to our inference about a good will. It is certainly hard to see how parallel a thing could be good in itself and the condition of all other goods without being incomparably more valuable than these conditioned goods. Nevertheless we may be inclined to suspect a snag in this kind of argument and perhaps if its conclusion is not confirmed by independent testimony
we were
whose goodness
consists in
its
even if it is. This brings us back to the question 'Arc we entitled, in the light of ordinary judgements, to rate a good will so far above artistic activity or the pursuit of knowledge? Is Kant's assertion due merely to the personal prejudice of a moralist? It is always difficult to be sure how far any particular judgement of value may be affected by personal prejudices. But it is interesting to observe that Kant himself at one time shared the prejudices, natural to a scholar, which attach the greatest value to knowledge, if not to art and he was converted from this view, not by some grim Puritan, but by Jean Jacques Rousseau. 'I am myself,' he says, 'by inclination a researcher. There was a time when ... I despised the masses, who know nothing. Rousseau has put me right. This blind prejudice disappears; I learn to honour men.' 1 The honour which Kant thus learned to pay to men is, I think, honour for their good will. The passage may be compared with another from the 2 Critique of Practical Reason, in which he says that even against his will his a man, however humble and ordinary, in whom he spirit bows before finds integrity of character in a measure which he is not conscious of finding in himself. This feeling of reverence, as distinct from admiration, is felt for the good man, and for no one else. To the modern intellectual, who tends to value moral goodness mainly as a means necessary to establish satisfactory conditions for the pursuit of art and science, it may seem that Rousseau has put Kant wrong rather than right. Nevertheless if we are discussing the goodness of things, we must do so from the point of view of an agent and not merely of a detached observer. Art and science have their own standards of beauty and truth, but they are also goods which a good man, according to his circumstances, will seek and ought to seek. Nevertheless not only do we regard goodness with a feeling of reverence which we do not give to success in that even art or science 3 but we know in these days almost too well these great goods must at times be sacrificed at the call of duty; and this
:
Fragmente, Phil.
Bib.,
VIII, p. 322. See also Schilpp, Kant's

136).
feel
Pre~Critical Ethics,
pp. 47-8.
A-.AK, 76-7
8
202
Kant holds
that
where we
(=
because of his
duty of
ability, but because he is to us an developing his talents. See O., 401 n.
reverence for a person of great talents, this is not example of one who has performed the 22 n.
II
7]
THE GOOD WILt

most precious.
43
could not be so, unless we held, as so many have held in action, that the goodness of acting well outweighs the goodness even of the things which
we regard
as the
In any case, whatever may be thought of Kant's position, assumption upon which his argument is based.
7.
this is the
good will and
its objects.
It was suggested in 3 above that the goodness supposed to belong to things considered in isolation was a mere abstraction, and it may now be objected that we are falling into this very same abstraction ourselves. good will considered in isolation from its objects is, it may be said, itself a
must have an object; and surely the goodmust be estimated by the goodness of the objects which it attains, or at the very least by the goodness of the objects which it seeks to attain. We cannot even begin to discuss the goodness of a good will before we have discussed the goodness of different kinds of object. Kant regards this objection as a fundamental error, fatal to any sound moral philosophy and indeed to morality itself. 1 We may hope to understand his view better as we advance but it should already be clear that if the objects of a good will arc conditioned goods, and a good will itself is an unconditioned good, the goodness of a good will cannot possibly be derived from the goodness of the objects at which it aims. 2 This contention 3 is in no way affected by the assertion, which Kant fully accepts, that a at or aim ends. must will objects good (like any other) If the goodness of a good will is not derived from the goodness of the ends at which it aims, still less can it be derived from success in attaining these ends. Furthermore, success in action depends on many factors outside our control, and a goodness which varies according to success or
mere
abstraction.
A good will
ness of a
good
will
failure could not
be described
as unconditioned.
And
finally, if
through
some misfortune, such as a stroke of paralysis, a good volition (not a mere wish) had no outer effects at all, it would still retain its unique and 4 incomparable goodness. This last assertion is surely borne out by ordinary
enlightened judgement. To maintain the absolute value of a good will is by no means to deny that other things, and in particular the objects sought or achieved by a good will, may have value. A good will is the highest good, but Kant always rejects the absurd view that it is the only good or the whole 6 good. In order to have the complete good our good volitions must be successful in realising their aim. This is what Kant expresses, not too whole or perfect good (bonum satisfactorily, when he says that in the 6 consummatum) happiness must be included. The common view that Kant underrated the value of happiness is a complete mistake if anything, he
:
rated
1 2
4
it
too high.
ff.
K.p.V., 58
Gr.,
177
ff.
Gr. 9
400 394
19-20; 437
-*3-
= 101 = 68.
198-9).
flf.)
and
especially
62-3
3
5
183-4
K.p.V., 34
=
(
IO~ !
* J
Gr -> 396
146
60).
15-
K.p.V.,
iiD=2 46(=
44
8.

The function of reason.
is
[H8
Kant
will
well aware that his insistence on the absolute value of a
mere
may appear paradoxical. In spite of the support it has from ordinary judgement we may suspect that it has its roots in some unconscious
tendency to the high-flown and the fantastic, which of all things he most abominates. He therefore attempts to corroborate his doctrine by a teleological argument concerning the purpose or function of reason in
human
In
life.
argument he assumes (i) that an organic being is a whole adapted to a purpose or end, namely, life (2) that in such a being every organ is also adapted to a purpose or end which is an element in the total purpose or end and (3) that every such organ is well fitted, and com2 pletely adequate, to attain its purpose or end. Man is such an organic being, and he has as one organ a reason which is practical in the sense that it directs his will. Now if we assume that the natural end of man is a happy life (or the maximum satisfaction of desire), reason is ill adapted to secure such happiness ; and nature would have done better to leave the choice of both ends and means to instinct rather than to reason. If we have to possess reason at all, its function should have been merely to appreciate the excellence of our instinctive nature as a means for securing happiness. Our actual experience shows that on the whole the more we confine the use of reason to the pursuit of pleasure even of intellectual
this
;
pleasure
the
less
On these grounds,
well adapted to the attainment of happiness. The true function of reason on its practical side must be to produce a will good not as a means to something else such as happiness, but good absolutely and in itself. Only on this hypothesis can we understand how reason can interfere with the attainment of happiness and yet not be an exception to the general principle that every organ is well adapted to its end. The argument raises questions about the supposed purposiveness of nature which are too complicated to consider here either in relation to Kant's philosophy or on their own merits and even those who accept a doctrine of teleology in some sense or other may nowadays be doubtful whether all organs are well adapted to their purpose. Indeed they may even regard reason itself as some kind of cosmic mistake. The argument, however, is subsidiary, and its chief interest for our purpose is this that it gives us our first introduction to the part which Kant conceives reason to play in action. He holds that reason does play a necessary part in the pursuit of happiness, both in conceiving the nature of happiness as an end and in guiding our choice of means to this end. The happiness thus sought
;
happy we seem to be. so Kant argues, if reason; like other organs, is to be its end, its end or purpose must be something other than
is,
however, a conditioned good : its goodness is relative, as we have seen, to the goodness of our will, and it is also relative to our desires and needs. Reason must also seek a greater good which is unconditioned by our
1
Gr.,
394
13-15-
See also
JT.r.F.,
6425.
II
9]
THE GOOD WILL

and needs, and only
so can
it
45
its
desires
be adequate to
purpose in spite
of its inadequacy for the pursuit of happiness. Such an unconditioned good can be only a will good in itself, not a will good in so far as it aims at
satisfying certain desires or at attaining the objects of these desires. These doctrines are here adumbrated vaguely, as is to be expected at
this stage
of the argument. We need not at present expect to understand Kant's view of the second and more important function of reason; but it is desirable to note that there are these two distinct functions of reason, and that both aim at a good, the first function at a conditioned good (namely, happiness), and the second at an unconditioned good (namely, a
good
will).
9.
Goodness
is fundamental.
found to be
the examination of Kant's starting-point and and unique goodness of a good will. This requires stressed, because as the argument advances we shall find ourselves talking a great deal more about duty than about goodness. Kant is so
it
We have now completed

in the absolute
commonly regarded
in true perspective
;
we must remember
as the apostle of duty that if we are to get his doctrine that for him goodness is funda-
mental and there is no warrant for supposing that he even entertained the conception of a duty divorced from goodness. On the other hand our conception of a good will is as yet vague and popular and in need of clarification. It is to the task of clarification that
we must now
address ourselves.
CHAPTER
III
DUTY
i,
good will and
duty.
IN order to make clear the nature of a good will Kant proposes to examine the concept of duty. 1 will which acts for the sake of duty is a good will but it must not be supposed that a good will is necessarily one which acts for the sake of duty. On the contrary, a completely good and perfect will would never act for the sake of duty for in the very idea of duty there is the thought of desires and inclinations to be overcome. completely good or 'holy' will, as Kant calls it, would be all of a piece it would manifest itself in good actions without having to restrain or thwart natural inclinations, and so it would not act from a concept of duty at all. We may suppose that God's will is holy, and it would be absurd to speak of Him as doing His duty. But in finite creatures, or at any rate in a finite creature such as man, there are certain 'subjective limitations' man's will is not wholly good but is influenced by sensuous desires and inclinations, which may be hindrances and obstacles to the good will present in him. Hence the good actions in which, but for these obstacles, his good will would necessarily be shown appear to him as duties; that is, as actions that ought to be done in spite of these obstacles. good will under human conditions is one which acts for the sake of duty. Such is. the contention which we have now to examine. It must not be thought that a will is good only because it overcomes obstacles. On the contrary, if we could attain to the ideal of complete goodness, we should have so disciplined our desires that there would be no further obstacles to overcome. Obstacles may serve to make the goodness of a good will more conspicuous, and we may be unable to measure goodness except by reference to such obstacles ; but a good will must be good in itself apart from the obstacles it overcomes. If it were not so, holiness itself would be eclipsed by our imperfect struggles towards
goodness. Our concern, however, is not with a holy will, but with a good will under human conditions and therefore a will which acts for the sake of duty. Hence the phrase 'under human conditions' which it would be tedious to repeat must be read into many of our sentences ; but this must not lead us to forget that much, if not most, of what we say is not to be taken as true of all good wills without exception.
1
Gr. 9 397
6.
M.d.S.> Tugendlehre, EinL X, 397
244.
[46]
Ill
2]
DUTY
The method of isolation.
47
2.
If we are to justify our contention that a good will under human conditions is one which acts for the sake of duty, we must first isolate actions done for the sake of duty and judge whether they possess the supreme worth which we have ascribed to a good will. It is true that actions cannot have precisely the same kind of absolute goodness as a good will, because a good action must be adjusted to a particular situation. Kant himself seems to hint at this difference when he speaks of an action as 'good in itself, but of a will as 'absolutely and in all respects 1 good'. Nevertheless if a good action is to be judged by the good will which is manifested in it, we may avoid fine distinctions and suppose that such an action has something of the same unique goodness as a
good will. I In the case of actions which accord with duty Kant recognises three main types: (i) those done from immediate inclination; (2) those done, not from immediate inclination, but from self-interest; and (3) those done, not from immediate inclination or self-interest, but for the sake of duty. He thinks that actions done from self-interest are not likely to be confused with actions done for the sake of duty but that actions done from immediate inclination are. The reason for this is presumably that in acting for the sake of duty and in acting from immediate inclination we do not look beyond the action itself to some remoter end or result. Hence in order to be quite certain that we are judging the value of actions done for the sake of duty, he asks us to remove the immediate inclination and assess the value of action in its absence. Thus, for example, he asserts that when we have through grief lost the immediate inclination to live and desire nothing so much as death, there still remains the duty of preserving our own life; and that in doing so, not from inclination, but for the sake of duty, the action has for the first time its genuine and unique moral worth./ Kant's method of isolation seems to me eminently rational and his conclusion wholly just. Most men would agree that there is no particular moral value in preserving one's own life merely because one has a desire to live; and most men would agree that the action has its moral value only when it is done for the sake of duty. The action in the first case would accord with duty, but would not be done for the sake of duty.
;
Similarly to take a case of self-interest if a man paid his debt only for fear of arrest, his action would accord with duty, but it would not be done for the sake of duty, nor would it have moral worth.
Such is, as it seems to me, the simple doctrine which Kant expounds. !His contention is that an action has moral worth only so far as it is done for the sake of duty^He may be mistaken, but at least there is nothing absurd in what he says. Nevertheless his argument has laid itself open to the most
,
strange interpretations which are now commonly accepted as an essential part of Kant's moral theory. It is only fair to add that the complication,
1
K.p.V., 62
182
109).
Gr.,
397-9
== 16-19.
48
[HI 3
and perhaps the
looseness, of his language have not unnaturally led to
these interpretations.
3.
Misunderstandings.
How
known
early these misunderstandings arose verses of Schiller: 1
can be seen from the well-
Gladly
I serve
I
Hence
my friends, but alas I do it with pleasure. am plagued with doubt that I am not a virtuous person.
answer
is
To
this the
given
And
Sure, your only resource is to try to despise them entirely, then with aversion to do what your duty enjoins you.
This is poor poetry and worse criticism. The common interpretation of Kant along these lines may be put more precisely as follows
:
1.
An
action has
if
present or
no moral worth if any inclination to do the action is any pleasure results from the satisfaction of this inclination
;
and
2. An action has no moral worth if any satisfaction arises from the consciousness of doing one's duty.
These two points are quite distinct, although they are sometimes confused. Both are absurd in themselves and mistaken as interpretations of Kant. It would be possible to fill a volume with citations 2 to show how remote such views are from Kant's doctrine. Here we must be content to state briefly Kant's own teaching on these topics. Some of the evidence for our assertions will be found in an appendix to this chapter. Kant holds that in order to have moral worth an action must not merely accord with duty, but must be done for the sake of duty. In holding this he holds also that so far as an action is done merely from inclination or even from such a motive as a rational desire for happiness it has no moral worth. In this there is nothing in the least paradoxical. Can anyone maintain that an action must be good, or that it must be a duty, merely because we have an inclination to do it? In order to establish his doctrine Kant adopts a method of isolation. He considers actions done solely from inclination without any motive of duty and says they have no moral worth. Similarly he considers actions done solely for the sake of duty without any inclination and says they have
1 Uber die Grundlage der Moral, 6. The translation, which I take from Rashdall, The Theory of Good and Evil, Vol. I, p. 120, is by A. B. Bullock. 1 Professor Schroeder has collected a certain number in Some Common Misinterpretations of the Kantian Ethics; The Philosophical Review, July, 1940.
Ill
3]
1
DUTY
49
moral worth. To use such a method of isolation is by no means to assert that where an inclination is present as well as a will to do one's duty, there can be no moral worth in an action. Kant means only (i) that an action is good precisely in so far as it springs from a will to do one's duty (2) that we cannot confidently affirm an action to be good except in so far as we believe that the will to do one's duty could by itself have been sufficient to produce the action without the support of any inclination ; and (3) that such a belief is hazardous except in the absence of direct inclination to
;
perform the action. Furthermore, in determining one's duty Kant starts as we shall see from his formulation of the imperative of duty with an action suggested to us by our inclinations. He then asks whether we can at the same time will an action of this kind as compatible with the universal moral law. If we decide upon our actions by this criterion, we are acting for the sake of duty, and our action is good. Kant's doctrine is that the motive of duty must be present at the same time as inclination and must be the determining factor, if our action is to be good. It is therefore a distortion of his view to say that for him an action cannot be good if inclination is present at the same time as the motive of duty. On the other hand he certainly holds that in determining our duty we must take no account of our inclinations or even of our happiness (which is the maximum satisfaction of our inclinations). By this, however, he means that we cannot affirm an action to be a duty merely because we happen to have an inclination to do it or because we think it would make us happy. If we assert an action to be a duty on these irrelevant grounds, we are self-deceivers. We all know that to follow duty may mean the sacrifice of inclination and even of happiness. Nevertheless we have, on Kant's view, a right, though not a direct duty, to pursue our own happiness in our own way, so long as this is not incompatible with the moral
law.
Kant recognises that inclinations have a part to play in the moral life. The advantages of virtue and the disadvantages of vice have to be used in order to prepare untutored minds for the paths of virtue. Some inclinations, such as natural sympathy, will greatly help us to perform our duties of benevolent action, and they ought to be cultivated for this reason. Even the happiness and the advantages which follow from a good life are properly stressed to counterbalance the attractions of vice and so to give the motive of duty a better chance to exert its influence. What we have to avoid is the substitution of the motive of personal happiness or personal advantage for the motive of duty. To do this is to undermine morality, for an action is not good in virtue of being done for the sake of personal happiness it is good only if it is done for the sake of duty. Apart from the everyday happiness which we naturally seek and which may be attained in leading a good life, there is also and here we come to our second point a special satisfaction which springs from
:
men may want
1 As Kant says in K.p.V.> 156 = 305^-6 ( = 279), the purity of the moral principle can be strikingly shown only by removing from the motives of action everything that
as belonging only to happiness.
5O
consciousness
THE CATEGORICAL IMPERATIVE of having acted well. We may not call it pleasure,
[HI 4
since
it
does not arise from the satisfaction of a particular inclination ; and we may not call it happiness, since it does not involve the maximum possible satisfaction of our inclinations taken together; but it is something analogous to these. We may call it contentment or self-approval, and we might even say though Kant himself at times denies this that it may be an important element in happiness. Kant is far too sensible to say that the good man will necessarily be happy. Because we are sensitive finite creatures, the good man will not be happy on the rack and it is a paradox to say, with Groce, that if this is so, it must be because he is not a sufficiently good man. Kant recognises that even if we have the consolation and inner peace which comes from knowing that we have acted well, we may nevertheless in our human weakness continue to be distressed at the results of our good action or at the loss which we might have avoided by means of a bad action. 1 If we do an action merely to gain this feeling of moral self- approval, we shall lose the feeling for the feeling arises from the recognition that we have acted, not to obtain a feeling, but for the sake of duty. An action done merely for the sake of this feeling would be a self-centred action without moral worth. It should be added that on Kant's view it is a mark of genuine goodness if we do our duty with a cheerful heart. He deplores a 'Carthusian' or monkish morality of self-mortification. From all this it is clear that for him an action does not cease to have moral worth if it is accompanied by pleasure or even by a desire for pleasure it ceases to have moral worth if it is done only for the sake of pleasure or only to satisfy an inclination.
:
These thoroughly sensible doctrines are completely distorted by epigram and by other interpretations of the same kind. Such interpretations arise from misunderstanding two quite different doctrines of Kant, (i) that virtue is most easily and surely recognised where duty is opposed to inclinations, and (2) that inclination must not be taken into account when we are trying to determine what our duty is.
Schiller's
4. Is duty the motive
of a good will?
Having set aside these misunderstandings, we are now in a position to consider the truth or otherwise of Kant's doctrine. And first we must ask whether acting for the sake of duty has the same unique goodness which we have ascribed to a good will. Is what we have called a genuinely good will manifested in acting for the sake of duty? Our answer will depend on what is meant by 'acting for the sake of duty'. If we take the man who acts for the sake of duty to be a man who is unceasingly worrying about what his duty is, so that there is never ease
or spontaneity in his actions, then this
is
not what
we regard
good man or a genuinely good

1
will. Still less
do we regard
as a genuinely as showing a
K.p.V.,
87-8
216
156-7).
Ill
4]
will the
DUTY
man who
51
approaches every situation with a series genuinely good of rigid and iron rules which he proceeds conscientiously to apply. Perhaps this is too harsh a judgement; for such men may be trying to do their best and so may be worthy of respect. Nevertheless they are at least
acting stupidly,
would say that
and they ought to know better. And certainly their actions shine like jewels for their own sake.
no one
Some of us may believe that what Kant admired was this wooden or mechanical or rigid kind of goodness. Certainly Kant was in some ways an austere, and even at times a rigid, moralist. Nevertheless he was essentially a humane man; and I believe that the types we are considering would be condemned by him as 'fantastically virtuous' or as 'moral enthusiasts'. 1 If we remember his rigid rule against lying, we should
also
remember
that in applying ethical rules
we must
leave
room
for the
free play of choice, 2 a certain 'latitude', where judgements of prudence and not of morality are decisive. 3 I do not say that he lays adequate
on the spontaneity of moral action, but he always insists that 4 morality is not a matter of wooden imitation or of mechanical habit. 5 His is no slavish morality and no one could insist more strongly on the freedom and independence of the moral will. The good man, on his view, must obey only laws of which he is himself the author. How then are we to understand 'acting for the sake of duty' if we are to interpret Kant at his Lest? We may get some guidance if we consider the man who is prudent rather than good. The prudent man is not the man whose mind is so concentrated on his personal advantage that he loses all spontaneity in action, nor is he the man who makes a rigid plan of his future conduct and applies it mechanically in all circumstances. Such a man is only superficially prudent, but genuinely imprudent ; for this is no way to have a happy or successful life. Genuine prudence is rather a controlling influence which is permanently in the background, ready to emerge at any moment in order to check acts of folly and to see that varied spontaneous impulses work together harmoniously instead of thwarting one another. So it is, surely, with the motive of duty in the good man. His life is controlled by the idea of duty; he is watchful against unruly impulses which conflict with duty, and against insensitiveness to the moral claims arising from his situation ; duty is the determining factor in his conduct, for there is always present the will to do what is right and to avoid what is wrong ; but granted this fundamental scrupulousness, there is a great place left for the free play of spontaneous impulses and a good man, like every other, has the right, as Kant always holds, to seek his happiness in his own individual way, so long as this is not in conflict with duty. If we interpret duty in some such way as this, it should be clear except to those who regard duty as an illusion that Kant is right in holding a good will to be manifested in acting for the sake of duty. Even where an action seems to spring from immediate inclination or from
stress
;
See Chapter II
5.
4
M.d.S.> Tugendlehre, Einl VII, 390

Gr.,
M.d,S., Tugendlehre, 10, 433 n. M.d.S. } Tugendlehre, EinL XV, 407
409
235.
31.
256.
52

is
[HI 5
the factor determining the action; for the good prudence, duty man would not perform the action if he believed that it was not in
still
accordance with duty,
5. Is goodness the motive
of a good will?
If we agree that a good will is shown as, I take it, most men would in acting for the sake of duty, there is still a possibility that it might also be shown in acting from other motives. In particular a good man might be good in acting, not for the sake of duty, but for the sake of goodness
itself. Good actions may be done, it is said, sub ratione boni; and Kant is harsh in stressing the motive of duty and the obstacles to be overcome. He forgets the good man's delight in goodness and good actions. Indeed goodness for its own sake, and not duty, is the motive of the really good
man.
Perhaps there
is
some truth
different
we do not have two
in this, but it must be observed that here moral theories what we have is two
:
different views of the empirical facts. Kant knows very well that the perfectly good or holy will must act spontaneously, and yet necessarily, from a sheer love of goodness, and not for the sake of duty. This he regards as the ultimate moral ideal to which we should for ever strive,
but to which under human conditions we can never attain. What he doubts is the actual presence of such a holy will in himself and in his
fellow men.
It is, however, only right to say that Kant has a high idea of holiness. For him a holy will would be wholly good it would always issue in good actions and would never be tempted to sin. 1 Such a will may belong to God, but not to man; and if the goodness of an action is derived from the goodness of the will which wills it, there are no human actions which can be regarded as holy and so divorced from the idea of duty. It is because of the height of his standards that he refuses man admittance into the company of the Saints. If we adopt lower standards, we shall no doubt be able to take in more recruits. It may be said that it is only reasonable to adopt lower standards and to recognise that at least some actions even of imperfect men may be inspired by a love of goodness without any thought of duty and may in this sense be holy. I confess that at times I have had sympathy with this view; and it may be held even if we accept Kant's belief that we can see too little into the secrets of our own hearts to be sure that such a motive, and not the 'dear self, inspires even the best of human actions. But even if we accept this, we ought to recognise the sound practical common sense which here, as so often, lies behind Kant's doctrine. He is rightly afraid that if we slur over the motive of duty in the interests of love of goodness, we shall lay ourselves open to all sorts of moral enthusiasm or Schwarmerei, to a windy, high-falutin, fantastic attitude of mind which will regard moral action as the meritorious manifestation of a bubbling
;
Gr.,
414
37; 439 SB 70. K.p.V., 32
144
57-8)-
Ill
6]
DUTY
53
;
may easily lead to vanity, self-complacency, and arrogance and although at first it may attract the young, it tends to produce a violent reaction. Our moral teaching should be manly, and not melting or sentimental or flattering or grandiose. We are not volunteers but conscripts in the moral struggle, and our state of mind therein is at the best virtue and not holiness. 1 Kant's view may be austere and even rigorous, but in the main it is sound. It was certainly a very necessary warning against some of the weaknesses of the German temperament, though such weaknesses are by no means confined to Germans. Nor should it be thought to involve any harsh separation between duty and goodness. For him virtue, even as steadfastness in the struggle to do one's duty, is something beyond price, casting into the shade, if we can see it in its true nature, everything that attracts our sensuous inclinations. 2 In any case the differences of opinion here are concerned, not with principles, but with empirical facts and with the practical side of moral teaching. Kant fully recognises that with the attainment of perfect goodness there would be no place left for duty. 3
heart. This
6.
Are generous emotions
the motive
of a good will?
Some thinkers believe that actions may have moral worth when they spring neither from devotion to duty nor from love of goodness, but simply from generous natural emotions, such as /sympathy and love. Here we come to a real difference of principle; Tor Kant denies this
outright.
kinship with virtue and so receive its name. He recognises that these adopted virtues have great similarity to true virtue, since they involve the feeling of an immediate pleasure in kind and benevolent deeds. To say of Kant that he treats generous emotions as on the same footing with the crudest animal impulses is downright false it can at best be excused as a result of ignorance. In spite, however, of his admiration for goodheartedness, actions which spring from such motives are to him worthy of love, praise, and encouragement, but they are not properly described as having that unique worth which we have called moral. They are in 3ome ways like actions which spring from desire for honour, a desire which, although it is on a lower level, may also lead us to acts which accord with virtue 5 and may itself be said to have about it a 'glimmer' of virtue. 6 On a still lower level perhaps though Kant does not say so
:
have seen above that such generous impulses and emotions are, in Kant's view, a help to the moral life and ought to be cultivated for this reason. In one of his earlier works 4 he even calls sympathy and obligingness 'adopted virtues' ; they are, so to speak, ennobled by their
We
(?r., 2
See especially K.p.V.> 84 411 n. 34 n.
ff.
21
ff.
150
ff.)
157
ff.
306
ff. (
280
150).
ff.)
Gr. t
*
5
2it ( K.p.V., 84 64-5 ; 426 n. 53 n. 435 Beobachtungen, 217-18. See also Schilpp, Kant's Pre-Critical Ethics, p. 53. * Gr,, 398 =17. Beobachtwigen, 218.
54
[III
they are like acts which spring from the enviable quality of animal courage, which, though it greatly helps a man to be brave, is still distinct from the genuine virtue of courage. It is not harsh or unreasonable to make a distinction between the man who is good and the man who is good-natured or good-hearted. Indeed the good-hearted are themselves often eager to disclaim any kind of moral merit. This may perhaps only show how good-hearted they are; but perhaps it also marks a real truth. Moreover, natural goodheartedness may lead in some circumstances to very wrong actions, either because it is by itself an inadequate motive or because it is a wrong motive. Good-hearted mothers have spoilt many children through lack of other requisite qualities; and the man whose amiability is shown in being 'easily led' is far from a model of moral worth. Even with the still higher motive of love, although it is admirable to act for love of another, the man who needs the influence of a good woman to keep him straight may not be a bad fellow, but he does not inspire us with respect. It may, however, seem harsh to say of such people, whether their actions accord or fail to accord with duty, that they are self-centred or that they are really seeking their own happiness. It would be harsh, if by this we meant that their generous impulses have no more value than quarrelsomeness or greed but there is no hint of such a view in Kant. Such people may at times act more generously, and even be more selfforgetful, than people who are conscientious. Nevertheless, although they are not selfish, they are in a sense self-centred. Their conduct depends less on the claims of the situation than on the accidental presence of certain impulses and if a particular situation arouses in them a wrong kind of impulse there is nothing more in them to which we can appeal. They will do the right thing in certain situations because this happens to fit in with their inclinations but equally they will not do the right thing in other situations, because this does not so fit in. In living by impulse, their will is centred upon the self and its emotions, even although many of these emotions may happen to be generous they act generously only because certain emotions happen to be theirs. If we are to judge the goodness of an action, as Kant does, by the goodness of the will manifested in the action, it is not unreasonable to say that even a generous action may be without moral goodness, though it has a certain value of its own. Kant makes no attempt to grade such types of action or at any rate
; ;
he makes no attempt to grade them systematically on any clearly formulated principle. We cannot complain of this, since the topic does not belong to ethics as he understands it and he has given us so much that we have no justification for blaming him because of the books he did not write. It is easy, however, to see that he estimates the value of certain impulses and emotions by the extent to which they can contribute to our happiness and facilitate our obedience to duty. But he also attaches to actions which spring from such motives at least when they accord with duty a certain aesthetic value. It is, he says, very beautiful to do good to men out of-love and sympathy; 1 and again that if love to mankind
;
JT./.F.,
82
209
146).
Ill
7]
DUTY
55
were lacking, a great moral adornment would be absent from the world. 1 In the latter case he is perhaps thinking more of practical benevolence than of natural sympathy but it may be suspected that for him it is the emotional or sensuous side of such benevolence which makes it
;
beautiful.
7.
Conclusion.
Kant's doctrine, when rightly understood, accords with our ordinary enlightened moral judgements. Under human conditions a good will is one which acts for the sake of duty, and only the actions in which such a will is manifested have moral worth. Hence we have taken our first step in the regress which seeks to establish the conditions under which our judgements of good actions can be true. Although our discussion has been more abstract than is customary in ordinary moral judgements, we are still 011 the level of what Kant calls 'common rational knowledge' of
now to move to a more philosophical level by passing concept of duty to the highly abstract concept of 'the law'. This further movement will be considered in the following chapters. If there are any who still feel dissatisfied with the argument on the ground that some good actions spring not from the motive of duty but from sheer love of goodness, this consideration need not disturb them.
morality.
We
have
from the
common
The 'law' to which we are moving is the same for the holy will as for the merely virtuous will; and it can be reached without difficulty by the same type of argument, whether we start from the virtuous will which seeks to do its duty or from the holy will which is above duty altogether.
AP PENDIX
INCLINATIONS, HAPPINESS,
AND MORAL WORTH
KANT'S
it is
doctrine on these matters has been so wildly misunderstood that necessary to give some specific references in order to make his position
clear.
Let us consider first our natural inclinations (or Habitual desires) and our desire for happiness {or the maximum possible satisfaction of our though he sometimes speaks as if a inclinations). According to Kant 2 it would rational man would wish to be wholly free from inclinations
1
Gr. t
M.d.S., Tugendlehre, 35, 458. 56; K.p.V., 118 428
256
212).
56
not only be
to root

f
[III
APP.
would also be harmful and blameworthy, to want Pure practical reason does not demand that we should renounce all claims to happiness, but only that the moment duty is in question we should not take it into account. Indeed my own happiness is included in the complete good which it is my duty to seek and in which virtue is combined with a proportionate happiness; 2 and it would be
futile, it
them
out. 1
self-contradictory if every man always sacrificed his own happiness, his own true needs, to further those of others. 3 This does not mean that it is a direct duty to seek our own happiness, since, according to Kant, we seek this by nature; 4 still less does it mean that it is a duty to seek the happiness of others on the ground that this will promote our
own.
That moral action is compatible with the presence of inclination is, however, best shown in the very formulation of the imperative of duty itself. According to Kant every action has a maxim that is, a principle actually at work in it and this maxim arises normally from the co6 operation of reason and inclination. The good man will act only on maxims which he can at the same time will to be universal laws. When Kant says that in a morally good action the motive of duty must always be present at the same time as maxims and so as inclinations, it is remarkable that he should be interpreted as saying that it must never be present at the same time as inclinations. 6 It is of course true that in some cases maxims have to be rejected because they are incompatible with duty. Nevertheless when they are not rejected they are present along with the motive of duty. There is no inconsistency in saying this and yet in holding that only in so far as the motive of duty is by itself sufficient to determine our action are we entitled to attribute to the action moral worth. It may be added that on Kant's view various inclinations and also the desire for happiness, though they can never determine what our duty
have nevertheless their part to play in the moral life. He recognises that in order to bring an untrained and unmanageable spirit into the path of virtue we must at first attract it by a view of its own advantage or alarm it by a fear of loss. 7 The life of virtue may be combined with so many charms and satisfactions that it may be worth following from the point of view of prudence, and it may even be advisable to join this prospect of a cheerful enjoyment of life with the moral motive which is already by itself sufficient to determine action but this must be done only as a counterpoise to the attractions of vice, not in order to substitute
is,
*
8
Religion,
58
K.p.V., 129-30
69).
271
233-4; Einl. IV, 386 230. 205 ( 118); 79 141). *K.p.V., 67 189 ( 8 Similarly when Kant says that sensa can be given to us only under the forms of space and time, he is often supposed to mean that they must be given apart from the forms of space and time and subsequently brought under them
Tugendlehre, Einl. VIII 2, 15, and Tugendlehre, 27, 451. * M.d.S., Tugendlehre, Einl. VB, 388
8
M.d.S. t
234); M.d.S., Tugendlehre, 27, 451. 240: see also Rechtslehre, Einl. 393
II,
216
/T./.F., 152
300
271).
Ill APP.]
this
DUTY
57
motive for the motive of duty. 1 Furthermore, a natural inclination to that which accords with duty (e.g. to benevolence) can greatly facilitate the effectiveness of the moral maxim, although by itself it can produce no such maxim. 2 Indeed he even goes so far as not only to say that since we have a duty to act benevolently, we have an indirect duty to cultivate natural sympathy as a means to this end ; but also to add as a reason for this that such sympathy is one of the impulses given to us by nature in order to do what the thought of duty alone would not do. 3 In all his writings Kant attaches great value to good-hearted inclinations and regards it as their chief function to counterbalance bad inclinations and so to give the motive of duty a better chance to determine action. 4 As to any suggestion that on Kant's view a good man must always be unhappy, this is pure nonsense. He always insists that the moral life 6 brings with it its own peculiar satisfaction or contentment (^jufriedenheif) It is true that he interprets it rather negatively as a special kind of satisfaction in doing without, and regards it as something analogous to, but
.
distinct from, happiness, which is the maximum satisfaction of inclination. 6 It is a kind of consolation or inner peace which a man may have
well. 7 In his later writings he goes farther well be called happiness. 8 may He also speaks of moral feeling as a capacity for finding pleasure in obedience to the law; 9 and describes a moral pleasure going beyond mere contentment with oneself (which may be merely negative). 10 Furthermore, he insists that cheerfulness of heart in the doing of one's 11 duty is the sign of the genuineness of a virtuous sentiment. If he rejects alike the Epicurean view that to seek happiness is to be virtuous and the Stoic view that to be conscious of virtue is to be happy, this only shows
even in
distress, if
he has acted
it
and recognises that
his
distinct
sense. 12 This moral satisfaction, it must be remembered, is from the everyday happiness which every man has a right, and even an indirect duty, to seek in his own way, so long as this is not incompatible with duty. Can it be wondered that Kant vigorously denies that his doctrine leads to any 'Carthusian' spirit of self-mortification? 18 Kant is certainly in his way an austere moralist, but we shall never be able to understand him unless we free our mind from caricatures of his
common
doctrine.
16. K.p.V., 88 217 ( 158). See also MAS., Rechtslehre, Einl II, 216 K.p.V., 118 256 ( 212-13). 3 M.d.S., Tugendlehre, 35, 457. This last statement should be only a concession to human weakness, but sec the even stronger statement in Das Ende, 388, where he asserts that the motive of love is necessary to make up for the imperfection of human nature and that without the assistance of this motive we could not in fact count very much on the motive of duty. Perhaps Kant was mellowing with old age.
2 1
= =
= =
4
6
7
O., 396 = 15. K.p.V. , 88 = 216
Compare
10 11
12
Schilpp, Kanfs Pre-Critical Ethics, p. 53. * 256 ( 212). K.p.V., 117 8 M.d.S., Tugendlehre, Vorrede, 377. ( 157). MAS., Tugendlehre, Einl. XII, 399 246. M.d.S., Tugendlehre, Einl. VII, 391 237.
Religion,
K.p.V.,
Religion,
in
24 n.
= =
18
23 n.
25 n. ( = 1 1-12 n.). 248 ( = 200). = 24 n. ( 10 n.). The whole passage should be studied.
CHAPTER
IV
THE MAXIM OF MORALITY

i
.
Moral
value does not depend on results.
WE have seen that the moral value of an action depends on its being done for the sake of duty, not on its being done to satisfy any inclination. We must now add that the moral value of such an action does not depend on results sought or attained. 1 The results we seek to produce by our action are not necessarily the same as the results we in fact produce; and it might even be the case, as in a sudden stroke of paralysis, that a morally good volition would produce no external results at all. Kant holds that a morally good action does not depend for its value either on the results it produces or even on the results it seeks to produce. In
holding this he rejects all forms of utilitarianism. Kant's conclusion follows directly from what has already been said for if an action depended for its moral value on results sought or attained, it would have this value even if it were done only from an inclination to
;
produce these
results
and
this
is
a possibility which
we have
rejected.
suggests also another argument against deriving moral value from results produced. 2 Other causes might produce the same results as a good will they might, for example, produce happiness for ourselves or others. If a good will derived its absolute value merely from producing these results, other causes producing these results would have the same absolute value. If so, a good will could not have the unique value which,
:
Kant
we have argued, it must have. Indeed if a good will had its value only as producing results, it would have value, not in itself, but merely as means to an end. Kant's conclusion has moreover been already established, as we saw in Chapter II 7, by what has been said about a good will. If a good will is an unconditioned good, and if its possible objects are conditioned goods, a good will cannot derive its absolute goodness from the merely relative goodness of the objects which it attains or seeks to attain. This conclusion still holds, although we have now added that, under human conditions, a good will is one which acts for the sake of duty and it applies also to actions done for the sake of duty, if the value of these is to be estimated by the goodness of the will which is manifested in them. Kant is not so foolish as to deny that an action done for the sake of duty will produce results and will seek to produce results. It always seeks to produce results, and normally it succeeds in doing so. All Kant is
;
(Jr.,
399-400
=19.
[58]
Gr. 9 401
21.
IV
2]
is
THEMAXIMOFMORALITY
59
that its distinctively moral value does not depend on the results saying sought or the results attained.
2. Practical principles
subjective
and
objective.
We must now try to make this doctrine more positive. If an action does not derive its value from results sought or attained, it must derive and this motive must be its value from its motive (the motive of duty) other than a mere desire to produce certain results. Kant expresses this by saying that an action derives its value from its maxim and although he does not put this too clearly 1 that the maxim is not a maxim of producing results. To see the relation between a maxim and a motive we must understand some of Kant's technical terms. A maxim is a particular kind of principle it may be defined as a subjective principle of action. In order to understand this we must first consider what is meant by 'a principle'. 2 It is difficult to state this precisely without going into too much detail, and the present account must be taken as provisional and over-simplified. 8 A principle is in general a universal proposition which has under it other propositions of which it is the ground for this reason it is also called a 'ground proposition' (Grundsatz). Strictly speaking, a principle should have no higher ground (except in so far as it is grounded in the nature of reason itself) it is then an absolute or supreme principle. Kant's Critical 4 method, as we have seen, is concerned with the discovery and justification of such supreme principles, and the object of our present enquiry is the discovery and justification of the supreme principle of morality. But the word 'principle' is also used, in a relative sense, for universal propositions which are not supreme principles, but yet have under them other propositions of which they are the ground. In this relative sense the major premise of a syllogism may be called a principle. In the strict sense the term 'principle' should be confined to such an ultimate principle as the law of non-contradiction. We are here, however, concerned only with practical principles or principles of action. Kant believes that human action is distinguished from animal behaviour by the fact that men act in accordance with he says, 'works in accordance with principles. 'Everything in nature,' laws. Only a rational being has the power to act in accordance with his idea 5 oj laws, that is, in accordance with principles.' Practical principle? are still universal propositions; they are, however, the ground, not only of other propositions or of judgements, but ultimately of actions. In them also we must distinguish between strict or absolute or supreme principles, and principles, in a looser sense, which are relative and subordinate.
,
:
1 2
is said to be a beginning (Anfang) Logik, 34, the Greek word apxtf8 6356-9. Compare, for example, K.r.V.> A299~3O2 5 4 See Gr. 412 Chapter I 9.
;
A principle
Gr.,
400
19-20.
1
10. It
is
a translation of
= 36.
6O
[IV
In practical principles there is, however, another and more immediately important difference. There are, first of all, principles actually at work in our action, principles which are the real ground of our action.
calls 'subjective principles Subjective principles are valid only for the individual subject or agent as the principles on which he chooses to act. Against these we must set 'objective principles' ; that is, principles on which any rational agent would act if reason had full control over his passions. 1 Objective principles are thus valid for every rational agent, and they may be called 'principles of reason'. subjective principle, in order to be such, must be acted upon:
.
These principles Kant
a subjective principle only if it is a principle on which we act. we do not always act on objective principles, and objective principles are still objective whether we act on them or not. We may, however, act on an objective principle, and when we do, this becomes a subjective principle as well as an objective one. Kant's technical term for a subjective principle is a 'maxim'. 2 Although a maxim differs from an objective principle in being, qua maxim, valid
it is
Needless to say,
only for the individual agent (not for
all
rational agents),
it
also differs
from a motive in being more general than a motive; and this is why it is called a principle. An animal might be said to have a motive so far as it is moved on a particular occasion by a particular impulse of hunger or a particular smell of food; but it could not be said to generalise its motive and so to have a maxim. Only a rational agent can have a maxim. Thus if I commit suicide because my life offers more pain than pleasure my maxim is 'If life offers more pain than pleasure, I will commit suicide.' 3 Here not merely an impulse or motive, but a general principle, is supposed to be at work in my action, a principle which I would apply to any similar situation. My maxim, as it were, generalises my action, including my motive. 4 My maxim is the principle which is in fact the determining ground of my action; but it does not profess, like an objective principle, to be valid for any one else, and it may be good, or it may be evil. 5 All this raises questions about the extent to which such subjective principles or maxims are consciously formulated by a rational agent. Nevertheless Kant is trying to mark a real difference between human conduct and animal behaviour. In acting, a human being does not, unless in very exceptional circumstances, respond blindly to impulse. He knows
1
Gr.,
400 n.
n.
Compare Webb, Kanfs Philosophy of Religion, p. 95. 'The origin of this use of the word is to be explained as follows. Every properly human that is, deliberately willed
act is done for some reason, subsumed, as it were in a syllogism, under some general major premise or major propositio. That to which in any individual case an9 act is ultimately referred is thus the ultimate major premise, maxima propositio or maxim.
8
23-4 = 25 (= 11-12). In this passage, as in others, Kant asserts a connexion between acting on maxims and freedom of the will. A free will cannot be determined to action by any sensuous motive except in so far as a man has taken up the motive into his maxim. In Logik, Einl. Ill, 24, Kant understands by a maxim the inner principle of a choice between different ends. 8 A maxim is not to be confused, as it often is by beginners, with what we call *a copy book maxim*. It is something purely personal to myself.
4
0.422=
48.
Religion,
IV
3]
;
THEMAXIMOFMORALITY
;
6l
what he is doing he recognises the quality of his action and he could not do this without some concept, however vague, of the principle on which he acts. 1
3.
Two
kinds of maxim
formal and material.
inclinations,
Kant speaks at times as if all maxims were grounded on sensuous and consequently as if a divine or holy will could have no maxims. 2 A divine or holy will would have no maxims which were not
also objective principles;
this is not to deny that it acts in interpret 'maxims' to mean principles manifested in action. It is all-important to recognize that while maxims are commonly based on inclinations (as in the case of suicide given above), it may nevertheless be possible to act on maxims which are not so based. have seen that a morally good action cannot have as its determining motive any mere inclination to produce certain results. If this is so, and if the maxim of an action is, as it were, a generalisation of the action and its motive, it follows at once that the moral maxim is not based on any mere inclination to produce certain results it holds irrespective of the ends which the action is intended to produce. 3 In saying this we are not altering or modifying our previous contention we are only putting it in a more technical way. In order to distinguish the maxims based on inclination from the maxims not so based we require to use certain technical terms. Maxims based on sensuous inclinations Kant calls empirical or a
but to say
if
accordance with maxims,
we
We
posteriori
maxims they depend on our experience of desire. Maxims not based on sensuous inclinations he calls a priori maxims that is, they do
: :
not depend on experience of desire. Empirical maxims are also called material maxims they refer to the desired ends which the action attempts to realise, and these ends are the matter of the maxim. 4 A priori maxims are also called formal maxims. At present formal maxims are characterised only negatively: they do not refer to desired ends which the action attempts to realise. What they are positively we shall have to consider later. Since in all actions the will is determined by a principle and so has a maxim, it must be determined either by a material maxim or by a formal maxim. There is no other possibility open.
:
4.
The maxim of duty.
We
shows
can now
say, in
Kant's technical terminology, that a
man who
good will in acting for the sake of duty is acting on a formal, and not on a material, maxim. In less technical terms, a good man does
his
1 might adapt a famous assertion of Kant's by saying that principles without content are empty, impulses without concepts are blind. See Tf.r.F., A$i 675. 2 /T./>.F., 67 205 ( 141). 189 ( 118); 79
We
Gr, 3
400
20.
Compare
Gr.,
436
66,
62

to desire.
[IV 4
not act merely on the principle of producing certain results which he
happens
causes so
This doctrine
much
already established by our previous argument, yet it difficulty to many readers that we may summarise the
is
argument
at this stage.
action done for the sake of duty does not derive its moral value either from the fact that it produces certain results or from the fact that it seeks to produce certain results. If it does not derive its value from the fact that its intention or motive is to produce certain results, it does not derive its value from a material maxirn that is, from a maxim or principle
An
of producing such results. Consequently its value must be derived from a formal maxim, a maxim independent of the desired ends which the action seeks to produce. The maxim of an action done for the sake of duty must therefore be formal and not material. All this may appear so negative as to be almost unintelligible, and we must give a more positive account of the moral maxim, if our position is to be made clear. Yet Kant's doctrine can be put quite simply, and when so put becomes almost obvious. The good man's maxim, in virtue of which his action has moral worth, is not 'This is the kind of thing I will do if I happen to have an inclination for it.' His maxim is T will do my duty, whatever my duty may be.' This is what we mean when we say that a man acts for the sake of duty. This is what we mean when we say that he is a conscientious and good man. The formal maxim of duty, and not the material maxim of inclination, is what determines his conduct and gives
it its
young Victoria was brought down from her bedroom in the early morning to be told that she was now the Queen, she is reported to have exclaimed I will be good. Although no philosopher, she was able to express in its simplest form what Kant believed to be the formal maxim of a good life. The difficulty of Kant's doctrine, so far, arises partly from his abstract way of putting it, but mainly from a whole series of presuppositions which we have been accustomed to read into it. We must try to rid our minds of such presuppositions. We have still to consider the positive side of his doctrine, but his argument, so far as we have gone, is surely sound.
C
5
When
value. the
CHAPTER
REVERENCE
1.
Reverence for the law.
LET
us
now
try to describe moral actions

out
more
positively.
To
act for the
sake of duty
is to act
of reverence for
the law.
Here we get two new points. Firstly, moral action has an emotional aspect, and the emotional aspect of it may be called 'reverence'. Secondly, in moral action we are seeking to follow a law which we reverence, and
therefore the moral maxim must be '/ will follow the law.' This assertion about our moral maxim continues the direct path of our argument and is essential to it. The discussion of reverence is not strictly necessary for the argument but it is necessary if our view of moral action is to be complete and if we are to meet the objection that emotion cannot be excluded from moral action. Let us consider first the nature of this
;
2 consciousness that we have tried to do our duty. We must now see that there is a specific emotion which Kant believes to be present in actions done for the sake of duty. It is obvious to any attentive reader that Kant himself feels most intensely this emotion of reverence for the law, and that both from his description and from the language he uses the feeling in question is something almost akin to religious emotion. For this reason I have translated the German word Âchtung" by 'reverence' and not by the word 'respect', which is commonly used by 3 English translators. This may be considered a mistake, and I may be told that the German word does not suggest any very profound emotion
:
emotion. It is a mistake to regard Kant's attitude to morality as cold and heartless. We have already seen that in his view contrary to the common belief inclination or emotion may be present in a moral action, although the action will have its distinctively moral value only so far as it is done for the sake of duty. We have seen that a good man has a right to seek his own happiness and a duty to seek the happiness of others. We have seen also that a special contentment or peace of mind may arise from the
indeed commonly used by railway porters when they wish you to get out of the way, and in this connexion is equivalent to 'Look out* or the
it is
1 208 ( 145). I substitute this for Kant's unsatisfactory Compare K.p.V., 81 20: Duty is the necessity of acting out of reverence for the third proposition in Gr., 400 law. 2 Chapter III, especially 3 and the Appendix. 8 After deciding on this translation I was pleased to find that it is also used by Edward Caird in The Critical Philosophy of Kant.
[63]
64

'Attention'.
[V
'Respekt',
French
Nevertheless
Kant himself
translates
it
it
word
*reverentia\
and he expressly
distinguishes
from
by the Latin which
has in it an additional element, namely, fear. 1 He compares his emotion towards the moral law with his emotion towards the starry heavens. 2 In the Critique of Judgment he connects Achtung with our feeling for the sublime. 8 These and many other passages suggest that Achtung is akin to reverence, or even awe, and is very remote from a strictly limited emotion
word 'Achtung' as a technical emotional atmosphere other than that in ordinary German usage. The feeling in question has no doubt degrees. 4 When, as in ordinary moral action, we are not greatly moved, it may well be akin to respect; but when it is present in its full force, it seems much more closely akin to reverence. 6
term in Kant which
like respect. are, I think, entitled to treat the
We
may have an
2.
The
description
of reverence.
For Kant the emotion of reverence is unique. 6 It is not directed to any object given to sense nor is it connected with the satisfaction of our natural inclinations. The emotion of reverence arises because I am conscious that my will is subordinated to a law without the intervention of any object of sense. In Kant's language, it is a feeling 'self-produced through a concept of reason'. Its similarity to religious emotion is shown, I think, by the fact that in it I feel at once humbled and also uplifted or exalted. 7 On the one hand the moral law is a check to my inclinations and a blow to my selfconceit, so that in being humbled the other hand as we shall see
On
my feeling
more
of reverence
I
is
clearly later
am also uplifted
akin to pain.
from
consciousness that the constraint imposed on my inclinations comes my own free and rational will. On this side my feeling is akin to pleasure. Yet it is so little a feeling of pleasure that I would gladly avoid feeling reverence for a good man or even for the law itself my vanity stands in the way. And it is so little a feeling of pain that, once I have laid aside my vanity, I cannot weary of beholding the splendour of the law the soul is itself uplifted as it sees the holy law uplifted above itself and its frailty. Kant also speaks of reverence as analogous to fear and inclination ; but as inclination is not strictly a feeling, he must mean by it something more like attraction or delight.
by
XII, 402 = 250 13, 438. The whole passage bears out my contention. * = he defines There K.U.> 27, 257 ( Achtung as a feeling of the inadequacy 96-7). of our power for the attainment of an Idea which for us is law. * = 13)Compare K.p.V., 73 = ( * Kant himself recognises in Gr,, 442 = 74 that feelings, and apparently also moral feelings, vary infinitely in degree.
Sn
,
Tugendlehre, Einl
161
312
288).
W-8
21 n. 401 n. See the long account given in JC.p.V.* 71 ff. i26ff.). Kant*s lan195 ff. ( guage at times may lack mathematical precision, but what he says does not lack
Gr.,
*
psychological insight
V3]
REVERENCE
65
All this belongs, not to ethics, but to moral psychology. It may have little or no meaning to those who have not been fortunate enough to experience the emotion in question; but even the little I have said is surely enough to dispose of the common belief that Kant is an unemotional moralist. He never seeks to substitute unction for philosophical analysis ; but that is a merit and not a defect.
3.
The function of reverence.
On the supposition that we have experienced this emotion Kant expects us to agree that we cannot feel reverence for any mere product of our action or for any mere inclination in ourselves or in others. I can feel reverence only for something which determines my will and does not serve my inclinations, but rather outweighs them or at least takes no account of them in determining my choice. This something we have hitherto known as duty. We are now told that it must be law in and for
itself.
1
must, however, remember that we may also feel reverence for those people and actions in which the law is exemplified. 2 And we must not forget that a being whose will was holy would not feel reverence for the law, but something more akin to love. For the moral philosopher the most important point in our discussion is this. On Kant's view we feel reverence because we recognise that the law is binding upon our wills. The great error of the moral sense school is to suppose that the law is binding because we feel reverence. No feeling can be the basis of a binding moral law, but the moral law may be the ground of a specific moral emotion. For Kant, to act out of respect for the law is the same thing as to act out of duty or for the sake of duty or for the sake of the law itself: 3 It is very different from any attempt to gratify an emotion ; and for this reason Kant classifies believers in a moral
sense- as unconscious,
We
even
if
4 well-meaning, hedonists.
0., 400 = 20. In Kant*s later writings he speaks of Achtung as owed to all men simply as men (in virtue of their capacity for morality) and from this important duties follow. The Latin
1
translation of 'Achtung' in this connexion is not 'reverentia* but 'observantia', that or 'consideration*. See M.d.S., Tugendlehre,"^, 462. * Gr -> 442^3 74207 (= 144). K.p.V., 81
is,
'respect*
66
[V APP.
APPENDIX
REVERENCE AS A MOTIVE
THERE
action,
are certain difficulties about the function of reverence in moral
and we may even be tempted to ask whether it is possible to regard reverence as a motive of good action on Kant's theory. He has certainly described good action as action done out of reverence for the law but he has also described good action as action done for the sake of
;
duty. He indeed appears to regard these expressions as equivalent. Nevertheless reverence, on his view, seems to result from, or even to be, consciousness that our will is determined by the law. 3 How can it also be a motive inducing our will to determine itself by the law? At first sight the difficulty might be got over by means of Kant's distinction between objective and subjective principles of action. 4 Kant first recognise the law as an objective might mean something like this. principle that is, as a principle which any rational agent would follow if reason fully controlled his inclinations, and so as a principle which we ought to follow in spite of our inclinations. This recognition arouses our feeling of reverence. Reverence in turn induces us to adopt the law as our subjective principle or maxim and so to act for the sake of duty or the sake of the law. Reverence would then be the connecting link between our recognition of the law as an objective principle and our adopting it as a subjective principle or maxim. Kant may have had something of this kind in view he describes reverence as a motive for making the law our maxim. 5 There appears, however, to be another doctrine when he insists that the moral law must determine the will immediately without the intervention of a feeling of any kind. 6 Here it might indeed be argued that he does not intend to exclude reverence as understood by him, but only any feeling (including reverence) considered as the ground of the law and not as its result. But he also says expressly that the moral law by itself alone must be not only the objective ground determining our action, but also at the same time a sufficient subjective ground determining our action. The law by itself, without the aid of any feeling, seems to be the motive necessary for good moral action. 7 This is a stumbling block to those who hold that the will can be moved to action only by means of feeling.
1 2
We
1 The contrast is sharper in the German, since a Achtung* (out of reverence) and'ausPflicht* (out of duty).
good action
is
done both
'aus
**./>.
8 *
5
Gr.,
F.,8i 401 n.
= 207 = 144). = 21 n. Compare also K.U.,

(
5,
210
15).
See Chapter K.p.V., 76
objectively
81
=
7
and of reverence
(
207
135). He also speaks of the law as determining our will 20 and 1C. p. F., as determining it subjectively ; Gr., 400 144). In the latter passage, however, this seems to mean something
IV
201
2.
(
rather different.
K.p.V., 71 K.p.V., 72
= 195 = 126-7). = 195-6 = 127).

(
(
See also
Gr.,
419
= 45,
'The
will
is
here deter-
mined without any other motive merely by the
law.*
V APP.J
REVERENCE
67
Perhaps Kant himself was not conscious that he appears to be speaking of two different kinds of motive. The word 'motive' is itself ambiguous. A motive is what moves us and many people seem to assume that it must
;
therefore be something that, as it were, pushes or shoves us from behind, as a feeling may be said to do. They forget that we may also be moved by something that pulls or attracts us, as perhaps the idea of the law might be said to do. These are crude metaphors, and we may ask ourselves whether the
Kant in a highly metaphysical way. We may attempt to treat them on a humbler common-sense level. First of all we can take an external and scientific view of action. On this view, which Kant regards as legitimate and even necessary, we 'explain' action as follows. First of all we apprehend something, whether it be a binding moral law or a glass of wine. This gives rise to a feeling, which in turn gives rise to an impulse, which in turn (in co-operation with reason) gives rise to an action (in accordance with a maxim). The whole process is explained as a chain of causes and effects. The only difference would be that the moral law is not, like the glass of wine, apprehended by means of sense. But even on a humbler and common-sense level we have also a very different point of view, the point of view of the agent acting, a point of view which sees the action from within, not from without. From this point of view we feel the previous account to be totally inadequate it omits the fundamental fact of our free choice. From this point of view willing is something other than a causal event, and an action cannot be explained as the effect of a previous cause. It is rather the direct product of our free will, whether we drink for the sake of pleasure or obey the law for the sake of obeying the law. 2 Hence it may be the case that from an external or psychological point of view our motive is the feeling of reverence, whereas from the internal or practical point of view our motive is simply the moral law, the law of our own free and rational will, without the intervention of any kind of feeling. We may perhaps say that from one point of view reverence is the cause of our action, but from another point of view the moral law is its ground. It should be added that if, as Kant holds, the feeling of reverence is, in finite human beings, bound up inseparably with apprehension of the moral law, 3 then to act for the sake of the moral law is also to act out of reverence for the law. We have indeed no duty to feel reverence for if we did not feel reverence, we should be blind to duty, and so should
enquiry, are treated by
1
:
whole language of motive is not crude and metaphorical. The difficulty seems to arise from the fact that we regard actions from two different points of view. These points of view, as we shall see towards the end of our
Gr.,
450
83-4.
recognises these two kinds of action as springing respectively from practical reason or pure practical reason, from will or pure will, and, in his later books, from Willkur (liberum arbitrium) or Will* (will in a technical sense).
Kant
A*./.F.,
80=
206
(=
142).
68
[V APP.
have no duty at
is
1 all. But the law compels our reverence, and reverence the emotional side of our recognition that the law is binding. We cannot excuse ourselves from acting for the sake of the law on the ground that reverence is a feeling which has to be conjured up. As we are always free to obey the law because we recognise it to be the law, so we are always free to act out of reverence for the law. The distinctively ethical command is not merely to do what accords with duty, but to do it for the sake of
duty.
Compare
M.d.S., Tugendlehre, Einl. XII,
402=250
a passage primarily concerned
with a special kind of reverence. 8 M.d.S,, Tugendlehre, Einl VII, 391
236,
CHAPTER
VI
THE LAW
i.
Law
as such.
WE have now to leave the comfortable levels of ordinary moral judgement, where we have hitherto maintained an uneasy footing. We climb beyond this level to the giddy heights and rarer atmosphere of philosophy when we say that the motive of the good man is to obey law as such. What is this mysterious law as such? And how can we maintain that a good will has its unique and absolute value only if it is determined by the thought of law as such? 1 Universality is the essential characteristic of law as such. A law, in the strict sense of 'law', must hold for all cases and admit of no exceptions. A law of nature, for example, must hold of all events in time without exception. If the principle that every event must have a cause is a law of nature, then there can be no exceptions to it; and if we were convinced that any exceptions were possible, we should at once deny this principle to be a law of nature. So it is also with what Kant calls the law of freedom' that is, the law in accordance with which a rational agent would act if reason had full control over his inclinations. This law of freedom, or moral law, cannot have exceptions without ceasing to be law. There cannot be one moral law for me and another for you. The law must be the same for all. 2 In Kant's technical language, universality is the form of law. Whatever a law may be about that is, whatever may be its matter it must have the form of universality for unless it is universal, it is not a law at all. Laws of freedom and laws of nature, in spite of fundamental differences, share in the common form of universality. In spite of its abstractness Kant's view still finds an echo in ordinary moral consciousness. If we try to consider duty in general, duty apart from the particular duty to do this and that, what can we say about it? We can say that duty appears to us as a law which must be the same for all and can admit of no arbitrary exceptions in our favour or in favour of anyone else. This is only another way of saying that if there is such a
;
1 2
O., 402
It
is
=
A
22.
extraordinary how early the human mind seems able to grasp the universality of moral law. small boy of five, not specially conspicuous either for goodness or for intelligence, was presented on a flag-day with several flags. One of these he was kind enough to give to me. Later he gave another to his sister, who rewarded him with a sixpence. Whereupon surely on the assumption that his sister's action was a manifestation of universal law (even if this was not without advantage to himself) -he asserted, 'If G. gives me a sixpence, the Professor will have to give me a sixpence too.'
[69]
7O
[VI
thing as morality, there must be an objective moral standard; for an objective standard is a universal standard, one valid for all rational agents independently of their desires for particular ends. So far from being paradoxical, Kant is merely stating the minimum that can and must be said about morality: morality must have the form of a universal law valid for rational agents as such. In all this we are talking about the one fundamental moral law of duty. When it comes to particular moral laws or rules, we have to face another problem. In a looser sense of 'law' we may say that 'Thou shalt not kilP is a law, but that nevertheless in some circumstances it may be our duty to kill. This does not mean that the moral law itself admits of exceptions. If it is the duty of a rational agent to kill in certain circumstances, then it is the duty of every rational agent to kill in these circumstances. It is fundamentally immoral to regard killing as a special privilege
of my
2.
own from which

Law
as command.
other
men
are excluded.
It
that
it is
might be suggested that a second characteristic of law as such is a command and Kant says that mere law in itself can be an
;
and therewith a command. 1 This, however, is not a characteristic of law as such, nor need it be true of the moral law. A law of nature is not a command and to a holy will the moral law would not appear as a command, although the holy will, because of its essential rationality, would necessarily follow the moral law. It is a very common mistake to say of Kant that for him the moral law is always a command or an imperative. On the contrary, we must make a sharp distinction between the moral law and the moral imperative. The moral law appears
object of reverence
;
human conditions as a command or imperative, because in us reason has not full control over the inclinations; but this characteristic does not belong to the moral law as such. For the will of a perfect being the moral law is a law of holiness; for the will of every finite rational being it is a law of duty.* This is the other side of Kant's statement that human beings reverence the law rather than love it. To say this is not to deny that it is possible to love the law. Only because of this distinction between moral law and moral command is Kant able to say of the moral law, as a law of action, that it must be the same for all rational agents that is, for all beings capable of action and for men only as rational agents. 8 This does not mean that we
to us under
need ascribe to Kant, as Schopenhauer does, an unhealthy interest in the morals of the angels. It means, as we shall see later, that the moral law holds for us solely in virtue of our rationality, not because as human beings we happen to desire certain ends, such as happiness. We can say
1
Gr.,
400
K.p.V., 82
= 20. =
08
(=
practical principle
and a
146). Compare also the distinction between a 56. categorical imperative in Gr., 428
supreme
*Gr. t 408
30-1.
VI
3]
THE LAW
71,
indeed that the moral law must appear to us as an imperative because of our character as finite and sensuous beings whose rational will has imperfect control over inclinations; but the moral law is not based on this character. Indeed the moral law is in no way based on the special nature of man. The special nature of man has to be taken into account, not in determining the essence of the moral law, but solely in determining
its
particular applications.
3.
The moral
motive.
If a good action is one which is done for the sake of law as such, this must mean that the moral motive, the motive of duty, must be lawabidingness, and this must serve as the principle of our will unless duty is
a purely chimerical concept. 2 To state it thus may appear paradoxical, but it is not difficult to see what Kant means. The morally good man seeks to obey the moral law as such whatever it may enjoin. If he obeys it only so far as he happens to desire some of the actions it enjoins, he is not a morally good man. Equally, if he recognises that the moral law is
binding upon
all
is
rational agents
and yet
seeks to
make
exceptions in his
own
favour, he
not a morally good man.
4.
Formal maxim and moral law.

are
connexion between law as such and our that the maxim of the good man must be formal that is, it is not a maxim or principle of producing certain results. How do we pass from this negative statement to the positive statement that the maxim of the good man is to obey law as such? At first sight the argument looks incredibly weak. The maxim of the good man is a formal principle that excludes all reference to the desired ends which are the matter of a maxim. But if so, Kant seems to argue, it can be only the form of a principle, and this is the same as the form of law or law as such. The only characteristic of a formal principle from which all matter is excluded is its universality. Hence the formal principle of moral action is to act in accordance with universal law. 'Since I have robbed the will of every inducement which could accrue to it from the following of any particular law, there remains nothing over except the conformity of actions to universal law as such, and this alone must serve the will as a principle.' 4 'The idea of the law in itself, so far as it, and not the expected result, determines the will, can alone constitute
to establish the
How
we
previous argument?
Kant has maintained 3
5 good which we call moral.' Kant has an unfortunate tendency to make his crucial transitions in arguments that are much too hurried and condensed. The weakness of
that pre-eminent
the present argument is as follows. principle is certainly universal; and a maxim or subjective principle, which is valid only for the individual
1
Gr.,
Gr.,
412 402
= 35. = 22.
Gr.,
402
22.
6
See Chapter
Gr.,
401
= 21.
IV
4,
72
agient, is
[VI 5
universal in the sense that it is applicable to different situations of thje same kind. The form of universality in this sense belongs to every maxim; but how can we pass from this to a quite different sense of universality in which a principle is universal as being valid, not merely for the individual agent, but for all rational agents without exception? Put in this way the argument is merely fallacious. But I doubt whether it is fallacious when we consider it against the whole background of cannot clear this up until we have examined, as we Kant's thought. shall later, the part which he believes reason to play in human action ;
We
but we can give a provisional interpretation of his line of thought. He is considering material maxims as a product of practical reason working 1 *at the service of inclination'. Practical reason, as it were, generalises our actions together with the inclination which is their motive; and although these maxims make no profession to be valid for anyone but the agent, "nevertheless even they would be valid for every rational agent if he were seeking to satisfy this inclination. 2 When we exclude from our maxim all reference to the satisfaction of inclinations and the attainment of results, and when we nevertheless suppose that our action must have a maxim, a purely formal maxim, which is the product of a practical reason no longer at the service of inclination, what could such a maxim be? All it would retain, as is indeed inevitable if it is a product of pure reason without reference to particular inclinations, would be its validity for
every rational agent as such, validity no longer qualified by an 'if. would be merely a maxim of being reasonable; it would be completely impartial as between my inclinations and yours and it would be acted on by every rational agent except in so far as his reason was thwarted by formal maxim individual desires. It would in short be a universal law. can be only the maxim of following universal law as such. To this topic we shall in due course return. But at least we now know where we are, and we should observe that however abstract Kant's language and however obscure his argument, his conclusion at least is in line with ordinary moral judgement. The moral law, if valid at all, must be valid for every rational agent and the good man is a man who that is, to act in accordance with seeks to obey such a universal moral law an objective standard which is independent of his personal desires and ends. Whatever we may think of the argument, the conclusion stands
It
;
firm*
5.
The
categorical imperative.
Even if we accept Kant's conclusion, we may still hold that it is empty and useless. How can we pass from the empty concept of lawabidingness to the manifold duties and virtues of the moral life?
Gr., 413 n. 45) Religion, 45 n. 37 n. ; K.p.V., 25 133 ( 52 n. ( 50 n.}. This holds onfy so far as the agent acts rationally in seeking to satisfy this inclination ; but even when he does not, his action is rational within arbitrarily imposed limits ;
;
maxim is
rational, for example, if he is seeking to satisfy some other inclination. conditionally valid for every rational agent. See Chapter IX x.
Here too the
VI
5]
THE LAW
is
73
certainly a difficult problem, but we need not make it more difficult than it is. Kant at once makes it clear that there is no question as is sometimes supposed of deducing particular duties merely from the empty form of universal law. On the contrary, we have to consider the matter which has to be fitted into this empty form. The matter consists of our ordinary material maxims based on inclination for definite objects; and what we have to do is to accept or reject these maxims by the principle of universality. Thus we can express the ultimate principle of duty more concretely than we have hitherto done. '/ ought never to act except in such a 1 way that I can also will the maxim of my action to be a universal law.' Emphasis should be put on the word 'also', because it is so commonly ignored. This is a formulation of what Kant later calls the categorical imperative, although here it is put in a negative form. We shall have to examine it in more detail later. At present we need note only that Kant speaks as if by the help of this principle alone we could easily solve all our moral problems. This is manifestly untrue ; we require to bring in many further suppositions, as Kant does himself when he comes to work out the application of the moral law in the Metaphysic of Morals. Nevertheless the principle itself is entirely sound, whatever we may think about Kant's method of applying it. To judge our own actions by the same universal standard which we apply to the actions of others is an essential condition
This
of morality.
CHAPTER
VII
MISUNDERSTANDINGS
i.
Criticisms.
may be
thought that Kant's doctrine

tricked
somehow been
We began with a good We with a merely formal maxim, a mysterious reverence for empty law as such, a vague principle of law-abidingness, and an unworkable test of universality for the maxims of our actions. All this formalism and legalism
may leave us cold. Furthermore, it may be said, we have been argued gradually into a manifestly ludicrous view, the view that in determining our duty no account whatever is to be taken of the results sought or attained by our action. These criticisms can be answered only by a re- examination of our previous argument and indeed of Kant's argument as a whole, but we may attempt to make brief comments on some of the points raised.
2.
is paradoxical, and that we have subtlety and complexity of the argument. will 'shining like a jewel for its own sake'. end
by the
Kant''s formalism.
the theoretical side there is little justification for complaining of We ought not to expect an abstract philosophical analysis of moral goodness to arouse the same warm emotion as may be aroused by the spectacle of a good man or a good action ; nor is it the to quote business of a moral philosopher to 'emotionalise the district' the phrase used by a disappointed American visitor to one of Professor Cook Wilson's lectures. Kant's terminology may be technical, but it was familiar in his time, if not in ours. It is well adapted to the expression of his meaning; nor is it on acquaintance more rebarbative than that of
On
Kant's formalism.
many modern
It is
philosophers.
hard to see why we should blame a philosopher for being too formal in dealing with the form of anything, even the form of morality. We do not blame Mr. Bertrand Russell because his logic is too formal, though some people may wish that he would write another kind of logic as well. Why should we complain that Kant's ethics is too formal, especially as he has written another kind of ethics, his Metaphysic of Morals not to mention his Lectures? In the Groundwork Kant, as he says, is dealing with the supreme principle of morality he is dealing with the a priori part of ethics in abstraction and considering the form of moral action apart from the matter. When Kant sets before himself a programme of this kind he is in the habit of sticking to it. It is hard to see why he should be blamed
y
:
[74]
VII
3]
MISUNDERSTANDINGS
75
for keeping to his subject and excluding irrelevancies. It is still harder to see why he should be charged with forgetting that moral action has a matter as well as a form, an empirical as well as an a priori element, and
an object
as well as a
his readers to
supreme remember.
principle.
Kant does not
forget.
He expects
3.
Kant's legalism.
replied that this answer does not really meet the criticism, concerned more with the formalism or legalism of his moral attitude than with his method of exposition. No doubt, it may be said, good men of a certain type may take a pride in obeying the law for the
It
may be
is
which
sake of law, and in controlling inclinations from the standpoint of a detached and unmoved reason ; but it is a great mistake to identify such a type with the moral life at its best. It is true that a man's philosophy, and especially his moral philosophy, takes a certain colour from his own individual moral attitude. This applies
to Kant as to any one else, and may help to explain certain idiosyncrasies of emphasis and perspective; but I do not think it vitiates his analysis, and it certainly does not excuse us from meeting his arguments. Kant himself was a gentle and humane man with a passion for freedom and a hatred of intolerance there is no evidence whatever to suggest that he was either cold or domineering or ascetic. I have tried to expose, in the course of my discussion, some of the misconceptions on which this charge of legalism is based ; but there is one special point which must be added in the present connexion. The main ground for charging Kant with legalism is the belief that he bids us perform our moral actions for the sake of a vague abstraction called the law, and thereby forbids us to perform moral actions for their own sake. Since this view is completely opposed to Kant's doctrine, and
:
yet may easily be read into his language, I will try to make his position clear. According to Kant every action aims at a result or end or object. In non-moral behaviour we perform the action because we desire the object ; we then have what Kant calls a 'pathological' interest in the object, and our interest in the action is mediate that is, it depends on our interest in the object. In moral behaviour we perform the action because the action, aiming as it does at certain results, is an embodiment of the moral law ; but it must not be supposed that the action is then willed only as a means f the contrary, we take an to an empty abstraction called the law'. immediate interest in the action itself 'when the universal validity of its maxim is a sufficient determining ground of the will'. 1 One of Kant's strongest convictions is that we take an immediate interest in moral actions. This is the reason why on his view actions done out of immediate inclinations, such as sympathy and benevolence, are more difficult to distinguish from moral actions than are actions done from self-interest, where there is no immediate inclination to the action. This immediate moral interest is
On
Gr. 9
460
n.
95
n.
See also 413
n.
37
n.
76

1
[VII 4
indeed another name for
'reverence', which we may feel for actions, and still more for persons, in whom the law is embodied. I do not believe that there need be any real inconsistency in Kant about this he is entitled to say both that we take an immediate interest in a moral action and that we do the action for the sake of the law. The category of means and end is inadequate to action, and grossly inadequate to moral action. The law is not for Kant an end to which the action is a means it is the form or principle of the action itself. Although it is the condition of the action's goodness, it is nevertheless an element in the
:
action
itself.
4.
The
ignoring of consequences.
I
Nothing,
suppose, will ever get rid of the illusion that for
Kant
good man must take no account of consequences in some sense which means that a good man must be a perfect fool. This interpretation rests on the ambiguities of language. 3 There is a sense in which the good man will take no account of consequences in deciding what he ought to do. He will not begin with the consequences and say that because an action will have certain consequences which he desires, therefore he will regard the action as his duty. He knows that it may be his duty not to produce results which he may greatly desire. Kant is right in saying that the expected consequences cannot be the determining ground of an action if it is to have moral worth. Nevertheless the good man begins with the maxim of a proposed action and asks himself whether the maxim can be willed as a universal law and the maxim is always of the form 'if I am in certain circumstances, I will perform an action likely to have certain
;
consequences'. How could we propose to steal or to kill or to act at all, if we ignored the fact that an action has consequences? Nevertheless we must not judge the action to be right or wrong according as we like or dislike the consequences. The test is whether the maxim of such an action is compatible with the nature of a universal law which is to hold for others as well as for myself. good man aims at consequences because of the law he does not obey the law merely because of the consequences. Such is the simple and obvious truth so often caricatured. If Kant had said merely that we must not allow our desires for particular consequences to determine our judgement of what our duty is, he would have avoided a great deal of misunderstanding.
<7r., 2
401 n.
22 n.
K.p.V. t 109-10
I 7, at the end. What Kant says about the highest good in 244-5 ( == 196-7) applies to any action or object. I will summarise and simplify. If our volition is determined by the thought of an action or object independently of the law, then our action is not moral. The moral law is the only determining ground of the pure will. But it goes without saying (Es versteht sich von selbst) that if the moral law is included in the concept of such an action or object as its supreme condition, then the concept of such an action or object and of its realisation by our will is at the same time the determining ground of the pure will. 8 It is, however, also encouraged by a tendency in Kant to exaggerate the generally sound principle that in many cases remoter consequences ought to be ignored.
Compare Chapter
VII
5]
MISUNDERSTANDINGS
1
77
5.
The soundness of Kant
s doctrine.
One
of the great merits of Kant's doctrine
is
the sharp distinction
which he makes between the a priori and the empirical, between duty and inclination. Since he wrote, there is no longer any excuse for the muddled thinking which confuses my good with the good and consciously or unconsciously substitutes for the moral motive mere desire for our own personal happiness either in this world or the next. A veiled and unconscious hedonism is as corrupting as it is confused. The primary aim of a good man is not to satisfy his own inclinations, however generous, but to obey a law which is the same for all, and only so does he cease to be selfcentred and become moral. There is no more fundamental difference than that between a life of prudence or self-love and one of moral goodness. We cannot give a general description of moral action by reference to its objects, both because the objects of moral action vary indefinitely and because they may be produced by action which is riot moral. Moral action must be described, not by its objects, but by its motive or principle or maxim; and this principle or maxim, for the same reasons as before, cannot merely be one of producing certain objects. The only possibility is that it should be a maxim of obeying a law which is the same for all in Kant's language, it must be a formal maxim. A man who is guided by the formal maxim of morality must not be conceived as acting in a vacuum. In the light of this maxim he selects and controls his ordinary maxims of self-love and inclination. In this way he resembles the prudent man, who selects and controls his maxims of
:
inclination in the light of the maxim of self-love. 1 The behaviour of the prudent man is familiar to us all, though we do not describe it in Kant's language ; and it too is the work of practical reason. The work of reason in moral action is not very much more difficult to conceive than the work of reason in prudential action. I will add one more point in anticipation of what is to follow. One of the reasons why Kant ascribes absolute value to a good will is that in obeying law for its own sake a good man is raised above the stream of events which we call nature he is no longer at the mercy of his own natural instincts and desires. good man is free in so far as he obeys the formal law which is the product of his rational will instead of being pulled about by desire, and it is this freedom which arouses Kant's veneration. 2 Whatever be our judgement of this, we do well to note that Kant's view of the formal character of the moral law is necessary to his doctrine of freedom. So far his philosophy has at least the merit of being
:
consistent.
Compare Chapter
III
4.
*Gr.,
426
= 53.
BOOK
II
THE BACKGROUND OF IMPERATIVE CATEGORICAL THE
CHAPTER
VIII
PRACTICAL REASON
AND
i
.
ITS SUBJECTIVE PRINCIPLES
The practical function of reason.
1 especially in moral action. Two distinct practical functions have been ascribed to reason: (i) the pursuit of happiness and (2) the 2 must now attempt to explain and justify this pursuit of goodness. and our task is the more necessary if, on Critical principles, assumption ; we have to trace the concept of duty, and indeed of goodness, to its origin in reason itself. 3 The view that reason has a part to play in action is not an innovation
T has been taken for granted hitherto that reason plays a part in
human
action
and
We
of Kant's: on the contrary, it has been the predominant doctrine of philosophy at least since the time of Plato. The widespread rejection of the doctrine at the present time is partly due to a distrust of reason as such. With this we have no concern, though it may be observed that if the distrust of reason professes to be based on reason, it is self-contradictory ; and if it is not so based, then it is admittedly irrational, and it would be waste of time to argue about it. Such a thorough-based irrationalism or scepticism must inevitably make all values rest on mere feeling and so make them arbitrary it must deny all objective standards of value. We are, however, supposing for the sake of argument that moral judgements, whether they are illusions or not, can be analysed as if they were objective we are endeavouring to discover their implications and the conditions which would have to be fulfilled if there were to be a possibility of their being true. One of the conditions is manifestly that such judgements must be a product of reason, which is the only source of truth ; and we have to face the contention that while reason may be concerned with truth, it
:
:
1 1
1
have discussed Chapter II 8.
this question
more
generally in Can Reason be Practical? 8 36. Chapter I 9 ; Gr. t 412
[78]
VIII
2]
SUBJECTIVE PRINCIPLES
79
cannot have any connexion with value, in particular with moral value, nor can it have any bearing on action. This view, in part at least, seems to rest upon misunderstandings of Kant's doctrine.
2.
Two
senses of'reason\
One of these misunderstandings is the common assumption that reason

be regarded only as a power of inferring or reasoning. If this were so, seems obvious enough that we cannot infer or demonstrate any particular action to be good or to be a duty ; and indeed we cannot infer or demonstrate that there is such a thing as goodness or duty at all. Where could we possibly get the premises for such an inference? There is a sense in which practical reason, on Kant's view, is connected with reason as a power of inference, although the connexion is very different from the simple-minded connexion we have just sug1 gested. But at present we must deal with this question on a lower level.
is
it
to
Reason, in traditional usage, is the higher faculty of cognition and as such is opposed to sense and imagination. 2 Aristotle defines man as a rational animal; and in this definition rationality or the possession of reason is what differentiates man from the higher animals, who share with him the possession of sense and perhaps even, in some degree, of
imagination.
sense
The
cognitive
is
power which man
possesses in addition to
and imagination
wider than a power of making inferences, though
this is included.
In the wide traditional sense reason covers (i) the power of entertaining concepts (or 'understanding in Kant's terminology) ; (2) the power of applying concepts to given objects (the power of 'judgement') and (3) the power of making mediate inferences (the power of reasoning or of 'reason* in a narrower sense). To say that reason in the wide sense plays no part in human actions is palpably untrue. All we need to consider is what is the precise part which such reason does play in human action.
;
3.
The approach
to
practical reason.
On Kant's view reason or rationality may be displayed in action as well as in thought. Reason, so far as it influences or determines action, is called by Kant 'practical reason' as opposed to 'theoretical reason' ; and he believes that practical reason and theoretical reason are the same power manifested in different ways. Full treatment of the identity of reason in its different manifestations belongs properly to a Critique of Practical Reason? shall have to be content with a more elementary
We
discussion.
*
It
1 f *
is
obvious enough that
we can discover the part played by reason in

this
K.r.V. y A835 Gr. t 391 8.
See Appendix to Chapter IX. 6863. Kant also uses 'understanding' in
very general sense.
80
[VIII 3
action only by considering the nature of action itself. This is Kant's own view, but he is sometimes misunderstood on this point. He says, for example, that we must derive (ableiten) moral laws from the universal 1 concept of a rational being as such. This might be taken to mean that we start with the concept of a rational being, in the sense of a being who thinks or even of a being who makes inferences; and that by mere analysis of this concept we arrive at certain propositions asserting, for example, that a rational being ought to act for the sake of duty, or even that he ought to fulfil such particular duties as the payment of debts, the display of gratitude, and the like. Such a procedure is manifestly impossible, nor is there any trace of such an argument in Kant. He explicitly rejects the
fundamental assumption on which
this
procedure
is
based ; for he asserts
that a practical proposition (or moral law) connects a volition immediately with the concept of the will of a rational being as something which is not
contained in that concept?
understand theoretical reason because we are beings capable of and we understand practical reason because we are beings capable of acting. Why should there be more difficulty in the one case than in the other? The only difference is that in understanding theoretical reason we are thinking about thinking, whereas in understanding practical reason we are thinking about something which is not itself thinking, but acting. Nevertheless in both cases we have, so to speak, an inside view of
thinking,
We
what we are thinking about. At times Kant identifies practical reason with the
he speaks of reason
:
will.
At other times
determining the will. The former terminology is more satisfactory it suggests that our willing is as rational as our thinking and is not merely something blind and unconscious which per impossibile is causally affected by our conscious thinking. Nevertheless when we speak of reason as determining the will, we indicate that volition has a cognitive aspect, which can be considered in abstraction. Similarly thinking has a volitional aspect which can also be considered in abstracas
3
tion.
admittedly some danger in speaking of powers and faculties the whole man who thinks and feels and wills. Nevertheless it is often convenient to speak of powers of thinking, feelingj and willing, when we wish to discuss different functions within man's total activity. In fact, it is almost impossible to avoid doing so, and there is no harm in this so long as we remember that what we are trying to understand is always an element in a whole rational life.
There
is
at
all. Strictly, it is
1
*
Gr.,
Gr.,
71. To connect the con54 and 440 46 n. See also O., 426-7 cept of a rational being with the moral law or the categorical imperative is to show how a categorical imperative is possible; and to do this, as we shall see rhore clearly later, is to justify a synthetic a priori proposition, not an analytic proposition. See
n.
412 420
35.
Chapter XII
Gr.,
9.
412=36.
VIII
4]
and
action.
8l
4.
Theoretical reason
Even
in
its
theoretical aspect reason
makes a vast difference toôur
actions. It is only by theoretical reason that we can be aware of the situation in which we must act and to which our action must be adapted ;
and in this situation we must include not only physical nature and the nature of other rational agents, but also our own nature, and especially our desires and needs. To the outside observer it may seem that all events happen, all organisms function, all animals behave, and all men act in one and the same world but from the point of view of the agent men act in the world as it is known to them. This remains true although the agent knows that he knows incompletely the world in which he must act. Hence it may be said that for the agent the world, or the situation, in which he acts varies according to the extent of his knowledge. Since men act differently in different situations, our theoretical knowledge must affect the character of our actions. If, for example, we do not know the cup to be poisoned, we may drink ; but if we do know it to be poisoned, we will not drink, unless it is our intention to commit suicide. The last point illustrates very clearly the special importance of knowledge about causes and effects. Without such knowledge all action would be impossible. But although theoretical knowledge is a precondition of action, and although action is always willed in the light of theoretical knowledge, knowledge does not thereby cease to be theoretical and become practical. Practical reason must be manifested in something other than the acquisition of such theoretical knowledge as may be useful or even necessary for action.
;
5.
Practical reason.
If our reason is practical, we must be able to will our individual actions as instances of a concept, just as our theoretical reason, in the wide sense, knows individual objects as instances of a concept. Kant holds that on the side of action reason manifests itself in this way. Only a rational 1 being, he says, has the power of acting in accordance with his conception of laws that is, in accordance with principles. And he contrasts such a being with things in nature, which work in accordance with laws, but do not act in accordance with their conception of laws. Everything in nature works in accordance with laws so far as its
working is governed by the laws of cause and effect; but although a stone, for example, falls in accordance with the law of gravitation, it does not (so far as we know) fall in accordance with its conception of this law. The movements and functions of man's body, and even according to Kant
of his mind, are equally governed by laws of nature, including the law of cause and effect. But in action something more than this is present. We have already seen2 that every action has its maxim or subjective principle;
1 Gr. t
412
= 36.
Chapter IV
2.
82
[VIII 6
and this is the first thing Kant has in mind when he says that a rational being acts in accordance with principles, or in accordance with its conception of laws, and so only has it a will. To have a maxim of action is something different from being aware of the law in accordance with which events happen to us or even in us. The physicist who falls from an aeroplane may reflect that he is now falling in accordance with the law of gravitation, but the law of gravitation cannot be regarded as the maxim of his action of falling nor indeed in such a case is his falling an action. If we were in any precise sense moved by irresistible impulses, something of the same kind would happen to us as Kant himself suggests, 1 we might be able to understand, and even to admire, the working of our own nature; but in this there would be neither a maxim nor an action. Although some psychologists speak as if all action were merely the result of a play of forces which they alone are able to understand presumably even when, if ever, they are acting themselves their view carries little conviction if we regard action from the agent's point of view. Even the murderer who pleads irresistible impulse seems generally to maintain that at one stage of the proceedings everything went black, and that only when he came to himself again did he find that the lady had been strangled. But however this may be, if we merely understood the law in accordance with which our nature worked, there would be no maxim and no action. Practical reason is shown, not in understanding a law of our behaviour, but in willing in accordance with a principle or maxim or, as I should be inclined to put it, in willing an action as an instance of a concept or
;
:
rule.
At
least in finite rational beings, there
is
no volition and no action
without a maxim; and it is only because we act in accordance with maxims that we can be said to have a will. Indeed Kant defines a will as a power to determine oneself to action in accordance with the conception of certain laws 2 that is, in accordance with maxims. And the same definition applies to practical reason since practical reason is identified with will.
6. Impulsive action.
It follows from all this that practical reason, with a maxim or subeven in action jective principle, is present in every kind of human action that we call impulsive provided that it is consciously willed. This is
what what
In
distinguishes human action from animal behaviour, or again from called 'reflex action', which we do not regard as our action at all. impulsive action the generality of the maxim, as a rule which we might
is
;
apply to similar situations, is far from being conspicuous and may easily be overlooked but even in such a case we are aware of the quality of our action and will it as an action having this quality. This would, I imagine,
395 14. O., 477 = 55. Compare Religion, 21 n. = 22 n. ( 7 n.), where Kant says that apart from a maxim no determining ground of free choice can or ought to be adduced.
Gr->
1
VIII
7]
83
be regarded by a British jury as the test of our responsibility or the condition apart from which the action could not be called ours, and indeed could not be called an action. In this there may well be degrees; and it may be hard to decide whether, for example, the avoidance of a sudden blow is an action or a mere reflex. But if it is an action, and if we are responsible for it, we are aware of its character, and we will it as having this character; and if this is so, our action has its maxim or principle. Kant unfortunately does not consider a philosophy of action to be 1 necessary for ethics, and consequently he does not discuss these questions in any detail. Nevertheless he gives us a good many hints as to his view. In a rational being an animal inclination, in so far as it is conceived through reason, becomes what he calls a 'pathological interest and on this interest a material maxim is based. 2 As he considers the pathological interest to be directed to the object, or intended results, of the action, he may possibly regard every material maxim as setting forth the kind of action 3 necessary to secure the object. If so, every material maxim may be as the to an end, the means being the action and means regarded stating the object being the end. In his actual usage, however, maxims express
9
primarily the will to act in a certain kind of

situation.
way
in a certain kind of
All this is important mainly in order to guard against misunderstandings. It is too commonly assumed that Kant opposes impulsive actions to actions based on a principle or maxim ; he is said to fall into a dualism of reason and inclination and he is alleged to recognise no difference in action other than that between pure moral action on the one hand and mere animal behaviour on the other. It is true that Kant recognises, as we all must, an antagonism between animal inclinations and pure reason in human action; but he also recognises that animal inclination can never issue in human action unless through the activity of reason it is transformed into an interest which gives rise to a maxim or rational principle of action. Here, as so often, the absurdities attributed to Kant arise only in the imagination of his interpreters. Animal inclination or impulse is never for Kant a motive of human action except in so far as 'taken up' by 4 practical reason into its maxims.
;
7.
Means and
end.
Practical reason
must be shown
in adjusting action to different
and
changing situations as these become known to us. This side of action Kant assumes rather than discusses he is not professing to offer us a general
See Chapter 79 95 n.
2
1
K.p.V.,
I,
Appendix (towards the end).
205
(=141); 67=189 (=
118);
Gr.,
413
n.
37 n.; 460
n.
The object may be the result which the action seeks to produce (that is, a change in the actual world) or it may be the satisfaction of inclination (that is, a change in the
self).
4
Religion,
23-4
= 24-5 =
(
12).
84
[VIII
philosophy of action beyond what is necessary for his ethical theory. He may conceivably regard this side of action as falling under the concept of means and end if I adjust myself to rain by taking shelter, this might be regarded as applying the means necessary to my aim of keeping dry. Anyhow he is interested primarily in the function of reason as looking forward and using means towards desired ends. This is soihething which is present in human beings, but not in the higher animals. Many of our maxims must take the form / will do as a means to T.' In this there must be all sorts of gradations and even a kind of continuity with animal life. An animal may take shelter from the rain and in this case, as in many others, such as eating and drinking, animal behaviour resembles human action, though we have no reason to suppose that the animal recognises the quality of what it is doing or that it regards its actions as means to a desired end. In human action, taking shelter may be relatively casual, or it may be part of what I call a 'policy', as when a man deliberately and systematically avoids any risk of getting wet. The most notable use of means to ends is in the exercise of skill, and it is in
:
this that It
Kant
is
particularly interested.
I
should be noted that while
may
act
on many particular maxims
concerning the best means to my various ends, these actions may be considered as having a more general maxim, the maxim '/ will use the most effective means to any end I may have.' This may be regarded as the controlling maxim when I use any particular means to a particular end. By itself this more general or higher maxim is empty, and it may not be consciously formulated but we may nevertheless be said to act in accordance with it if we refuse to act on particular maxims which are opposed to it. In such a case we might be said to act only on maxims which fall under the higher maxim of using the most efficient means to our ends. Needless to say we do not act on this higher maxim in vacuo we must also desire a particular end and be aware of a particular means. This account is far from adequate as a complete description of action which, however, it does not profess to be. We might question the category of means and end altogether, since it may be doubted whether we will part of an action as a means and part as an end the distinction seems rather a theoretical one applied to actions either already done or
;
contemplated. If we set this objection aside, we may still urge that the unity of an action, and still more of a policy, is not exhausted by conformity to any maxim, and in particular that it is not exhausted by conformity to the maxim of seeking the most effective means to our ends. The maxim is general, the action is individual. Because of this there must be more in the action than in the maxim, and the unity of an action or of a policy resembles in some ways the unity of a work of art rather than the unity of conforming to a concept. This, however, raises problems which cannot here be considered. 1
else
1 The Hegelians in particular charge Kant with considering only abstract universals and failing to take into account the individual, or what they call the 'concrete universal*. These charges are often exaggerated and are at times due to misunderstanding, but they are not without some element of truth.
VIII
8]
85
Whatever be the deficiencies of Kant's doctrine, the distinction of means and end is commonly accepted, and he is dealing with a recognisable aspect of action which will be important for his later argument.
Practical reason is manifested in maintaining the conscious unity of our actions and policies, and this unity may be described, although inadequately, as conformity to the concept of means and end.
8.
The pursuit of happiness.
selves,
not enough to maintain the unity of actions and policies in thempractical reason manifests itself also in maintaining their unity in relation to one another in a whole organised life. We have many needs to meet, many desires to satisfy, many ends to attain, and all these must be fitted into a whole life or policy of life. This task Kant regards as the pursuit of happiness and as the principal function of practical reason below the level of morality. His account of the pursuit of happiness is not altogether satisfactory, nor is it wholly consistent with itself. At times he takes a hedonistic view and seems to regard happiness as little more than the greatest possible amount of continuous or uninterrupted pleasure throughout the whole of 1 This he considers to be the final end which all men seek. He recoglife. nises indeed that reason, even on this level, must estimate happiness, not by the passing sensation, but by its influence on our whole existence and our contentment therewith. 2 Yet in the main the function of practical reason on this view is to will the appropriate means to this clearly envisaged end. Here again the category of means and end seems inadequate for the description of action, and Kant himself combines this view inconsistently with another view, according to which we have no determinate and sure 3 concept of happiness as an end which we seek. is then regarded as the total satisfaction of our needs and Happiness 4 inclinations, and what was formerly considered as a means to happiness is now considered as an element in happiness. Among such elements there appear to be reckoned such things as riches, knowledge and insight, long 5 that is, objects which we desire and ends which we seek. life, and health, Practical reason is no longer concerned primarily with the means to a known end namely, happiness or continuous pleasure. Practical reason is above all concerned with what constitutes the happiness which we all seek as our end it must aim at satisfying as many as possible of our needs in an organised life or (as Kant puts it) at bringing our natural inclinations into harmony with one another in a whole called happiness. 6 In so doing practical reason manifests itself as prudence or rational self-love, and it has the task of determining the constituents of happiness as well
It
is
and
1 8
6
E.g. K.p.V., 22 Gr. 9 399 =19. Gr. t 418 42-3.
129
= 40).
* *
K.p.V., 61
Gr.,
= 181 = = 26. = 70). Religion, 58

(
107).
405
86
[VIII
it is
as the task of prescribing means to the attainment of these constituents: 1 shown in a choice of ends as well as of means. 2
This second view is much more satisfactory than the first, though it is not without its own difficulties. Rational beings seek to realise their multifarious ends within an organised and systematic life. This comprehensive rational end is perhaps unfortunately called 'happiness', but there seems to be no other more satisfactory word. It is absurd to suppose that the only object we desire and the only end we seek is continuous pleasure and the avoidance of pain. The principle of rational self-love is not so much a principle of using the means to continuous pleasant feeling, but is rather a principle of integrating our ends, of which pleasant feeling is only one, into a single comprehensive whole. If we were writing a philosophy of action, this account would require a great deal of expansion and qualification. Because Kant is concerned primarily with the rational element in action, he makes no effort to consider the element of spontaneity or creativeness, and indeed of arbitrarito a lesser extent ness, which is present in our search for happiness and even in our use of means towards particular ends. To suppose on this ground that he ignored or denied such an element would be completely unjustified. There are many indications that he distrusted a too nicely calculated method of life see, for example, his obvious sympathy with the gouty man who risks his extra glass of port 3 and in thinking as well as in art he is always on the side of creativeness and spontaneity. Indeed his whole philosophy may be described as one of spontaneity and freedom. At present, however, we are concerned only with the function of practical reason in controlling and organising the fulfilment of our
desires.
The pursuit of happiness, if it were concerned only with the attainment of pleasure, might well be described as selfish; but if it is concerned
with the fulfilment of our desires, it should be described as self-centred rather than as selfish, since, as Kant always recognises, we have otherregarding as well as self-regarding desires. The centre on this level is always the self, but the circumference, so to speak, may cover many things, including the happiness of others. So far from rejecting the life of prudence or self-love, Kant maintains consistently that every man has a right to pursue his own happiness so long as this does not conflict with the moral law. He even maintains that we may have an indirect duty to do so, since wretchedness may easily lead us into immoral
action.
The maxim of prudence

happiness. This
9
is
or rational self-love is I will seek my own better interpreted, not as a maxim dealing with means
:
and end
1
('/ will use the
means
to secure the
maximum of pleasant feeling')) but
Religion,
45
n.
= 52 n. = 50
(
n.).
M.d.S., Tugendlehre, Einl. Ill, 385 6828, where 230. See also K.r.V., A8oo Kant says that in the doctrine of self-love the whole task of reason consists in uniting all ends set us by our inclinations into the one end, happiness, and in harmonising the means for its attainment.
2
Gr.,
399=
'9-
VIII
as a
9]
of integration ('/ will aim at
the satisfaction
87
desires in
a of my 9 whole organised and systematic life ). Indeed even the maxim of skill may be interpreted, not as concerned merely with the application of means to a particular end, but as concerned with the integration of an action or a policy as a limited whole. However we interpret it, the maxim of prudence commonly overrides the maxims of skill, though it does not supersede them. may reject the most effective means to a desired end, we may even reject a desired end itself, if such a means conflicts with our happiness as a whole. In this way the maxim of prudence may be our controlling maxim, and we may be said to act on particular maxims only as they fall under the maxim of prudence. But once again we do not act upon this controlling maxim in vacuo it has its content only so far as we desire particular ends and seek to make use of particular means.
maxim
We
9.
The denial of practical
reason.
that practical reason plays this part in human life seems to plainly false. The denial owes such plausibility as it possesses to an assumption of the following kind. Reason is purely theoretical; it reveals to us, although only in part, the situation in which we have to act ; and we act differently in different situations. But what we call the will is itself a blind force, or even a combination of blind forces, which manifests itself differently in different situations as known to us. Reason affects action only in presenting to us the situation and, so to speak, the external stimulus to action. 1 This assumption, though in one place it seems to receive a modified 2 approval from Kant himself, is surely mistaken. It is not the result of a but of dogmatic and even metaphysical prejudice, empirical observation, natural enough to those whose chief occupation is thinking. There are certainly occasions on which we think first and act afterwards, but far more often we think in acting the action is not preceded by intellection, but is itself intelligent. could a blind force, or combination of blind forces, even seem to respond differently and intelligently to each different known situation? 3 Reason is shown, not merely in understanding the situation or in recognising the quality of the completed action, but in as adjusted to the situawilling the action as an action of a certain kind tion, as making use of the best means, and as contributing to the agent's
me
To deny
How
happiness.
behaviour as
The character of human volition is as human apprehension of the situation

never merely theoretical.
y
different
is
different
from animal from what
we may suppose to be animal apprehension. There is far more plausibility in the pragmatic
is
always practical and

1 8
is
The truth is
contention that reason surely that
a K.U. Erste Einleitung, I, 1-9. See 4 above, It may be objected that instinct seems to do something like this. Yet we know instinct and not reason only because of its lack of flexibility or intelligence.
it is
88
[VIII
IO
reason is always both, although at one time the theoretical, and at another time the practical, aspect is the more prominent of the two. 1
10.
Morality.
There is some ground for saying that men sometimes act, or at least believe themselves and others to act, on maxims of morality. Kant's view that the maxim of morality has a formal character has already been
discussed. 2
For some points in
this section I
am
indebted to 'a letter from Proiessor H. H.
Price.
*
See Chapter IV, especially
3-4.
CHAPTER
IX
PRACTICAL REASON AND ITS OBJECTIVE PRINCIPLES

i.
Subjective
and
objective principles.
HITHERTO we
have considered practical reason so far as its principles are in fact manifested in human conduct and can in that sense be called even 'subjective*. Our contention has been that reason is practical since apart from morality it does in fact influence human action it is present in the material maxims of skill and self-love by which human action is very often determined. To deny that reason is practical is to assert that reason is never manifested either in the exercise of human skill or in the pursuit of happiness, and such an assertion is manifestly absurd. On the other hand we know only too well that even on this level reason may fail to determine action. We may be led by some sudden passion into actions which may make the attainment of a desired end impossible and may even wreck the happiness of our whole life. Our human nature is far from being such that we must necessarily act in accordance with rational maxims of skill and self-love. In Kant's technical language these rational maxims are not 'subjectively necessary' ; they are
:
'subjectively contingent*. equally we may not.
That
is
to say,
we may
act
upon them, but
It must not be thought that in asserting this we are abandoning our previous contention that reason is manifested in all human action. When moved by passion we may still recognise the quality of our action, and we may will our action as an instance of yielding to fear or anger, as the case may be if we ceased to do this, our action would become mere animal behaviour rather than human conduct. But in stupidly impulsive action the influence of reason is narrowed we act on a maxim rational within its own limits, but irrational in relation to our wider view of the
:
end we
seek.
We may
be rational enough
to recognise this irrationality,
not merely afterwards, but even at the time, and only so do we impute folly to ourselves. If reason were not present as well as passion, our behaviour would be merely unfortunate and not foolish. We can conceive a rational agent so rational that he would never act foolishly. In such an agent reason would have full control over the passions. He would never allow passion to distract him from acting in accordance with the principles of skill and self-love. To yield to passion in despite of reason would be contrary to his rational nature, and we can say that for him the maxims of skill and self-love would be subjectively
1
ft.,
413*-* 36.
[89]
9O
[IX
fully necessary and not as in our own case subjectively contingent. rational agent could not but behave rationally. Such a conception even apart from the difficulties that may be felt about the freedom of rational agents may appear to be valueless since
The conception are to consider the character of objective 1 principles; for objective principles, as we saw above, are principles on which a rational agent would necessarily act if reason had full control over his passions. Objective principles are valid for every rational agent, while subjective principles or maxims as such make no pretence to be valid for any one but the agent who is acting on them. In the previous chapter we considered the principles of skill and of
acquaintance with fully rational agents.
useful if
we have no
may, however, be
we
merely as maxims on which rational agents sometimes act and which practical reason is manifested. But on Kant's view they are more than mere rnaxirns even if they were never acted upon at all, they would still be objective principles, though of a special kind. Indeed provided we are right in regarding them as manifestations of practical reason it would be surprising if they had not some kind of validity for every rational agent. Kant regards the principles of skill and self-love as principles on which any rational agent would necessarily act if reason had full control over the passions. To say this is to say that they are objective principles; and
self-love
in
again, in Kant's language, that they are objectively necessary, even 2 may be subjectively contingent. What is peculiar about them is that although they are objective principles, they are so only subject to a condition. The meaning of this we must now examine.
although they
2.
The principle of skill.
Let us consider first the principle of skill, and let us stick to Kant's terminology of means and end. A rational agent who seeks any particular end will so far as reason controls his passions necessarily make use of the most effective means in his power. Men indeed vary in power and consequently may have to use different means; but if we regard the agent's power as part of the situation in which he has to act, we can say that in the same situation every rational agent, qua rational, will necesthat is, he will necessarily sarily act in accordance with the same maxim use the most effective means to this particular end, if he seeks the end. It is possible to raise various objections which cannot here be considered in detail. Sometimes a rational agent may appear, quite rationally, to choose some means other than the most effective, as when a climber takes the most difficult way up a hill but in such a case his end is something other than merely getting to the top. Sometimes he may dislike the means more than he likes the end and so may give up the whole project. Always the principle of skill may be rationally overridden by
;
Chapter IV 2. For a fully rational agent they would be both objectively and subjectively necessary; for an imperfectly rational agent like man they are objectively necessary but subjectively contingent. Gr., 412-13 = 36.
2
IX
3]
OBJECTIVE PRINCIPLES
91
the higher principle of prudence, not to mention the still higher principle of morality. But within its own limits it is conditionally valid for every
rational agent.
There must always be considerable 'latitude or 'play-room' in the application of particular maxims of skill. An iron shot by Bobby Jones will differ in some ways from an iron shot by Harry Vardon even under the same conditions. This is still more obvious in the fine arts (which are not here in question save as regards technique) 'There are nine and sixty ways of constructing tribal lays, and every single one of them is right.' I see no ground for holding that Kant expects his principles to be applied mechanically and identically in every case. On the contrary, he speaks as if the essential character of a maxim were to require judgement in its application and to leave room for 'latitude'. 1 He asserts this primarily in regard to maxims of morality but in the light of his treatment of art, and also in the light of his marked contempt for imitation as opposed to creation, we are justified in extending his view to cover maxims of skill. In all action a rational agent not only has a maxim, but also sets before himself an end, 2 if we can interpret 'end' to cover actions done for their own sake and not merely for the results they produce. 3 The general principle of skill the principle of using the most effective means is objectively valid for any rational agent. Particular principles of skill (which
:
are only applications of the general principle) are also objectively valid, but only subject to the condition that some particular end is sought. Their conditional character does not detract from their objectivity. There is also another sense in which principles of skill are conditioned. No rational agent, even if he sought to attain a particular end, would necessarily, so far as he was rational, use means, however effective, which would destroy his happiness as a whole. Principles of skill are conditioned, not merely by the end sought, but also by their compatibility with the higher principle of prudence or self-love. This, however, in no way means that they cease to be objective principles.
3
.
The principle of self-love.
The same considerations apply to the principle of rational self-love or prudence. According to Kant, a rational agent, so far as reason has full control over his passions, will necessarily seek his own happiness. Since man, as belonging to the sensible world, is a being with needs, reason has so far an office which it cannot refuse', the office of serving the interests of sensibility and of seeking happiness in this world and where possible in the next. 4 To seek one's own happiness is not merely a maxim on which
c
many men
is
act
it is
also
an objective principle of practical reason.
is objectively necessary for a rational agent, it not thereby also subjectively necessary for an imperfectly rational agent 1 M.d.S., Tugendlehre, Einl. VII, 390 = 235 Einl XVIII, 411.
Although
this principle
229. 55; M.d.S., Tugendlehre, Einl. Ill, 385 This is assumed by Kant, although his language often suggests that an end merely a product or result.
Gr.,
427
is
*K.p.V.,
61=
181
(=
108).
Q2
[IX
such as man.
passion or
Many men in fact wreck their own happiness through weakness. To say this is not to deny that there is some ration-
ality even in their imprudence. Imprudent actions have, however, a lower rationality, since in them reason, so to speak, takes account of Jess than all the facts which it has before itself.
As we saw in the previous chapter, 1 Kant interprets happiness in two ways. We shall find his view here also more convincing if we take happiness to be, not a maximum of pleasant feeling throughout life, but
satisfaction of desires or the maximum integration of ends. should not, however, take him to mean that the prudent agent will necessarily, so far as he is prudent, plan his whole life in advance. The part of prudence is not to destroy spontaneity, but to prevent one passing desire of the moment from thwarting many other desires permanently. The philosophy of prudence might be elaborated indefinitely, 2 but it is not immediately relevant to our purpose. Probably for this reason Kant himself tends to neglect it. We must, however, recognise that maxims of prudence admit of great latitude in their application. While
the
maximum
We
agents will, so far as they are rational, necessarily seek their happiness, they may find it in very different ways. The desires of the same individual vary within limits according to his circumstances and his experience ; and an early decision, such as the choice of a career, may cause him to find his happiness in one way of life when he might equally well have found it in another. The essential point for our present purpose is that particular principles of self-love, though objective, are still conditioned. They are conditioned first of all by the character and desires of the particular agent WG could not reasonably expect Mahatma Gandhi to find happiness in the same kind of life as Winston Churchill. In the second place if Kant is right the principles of self-love are conditioned by their compatibility with the principles of morality. On his view a rational agent, so far as reason had full control over his passions, would necessarily refuse to plunge a continent into war however much such a catastrophe might contribute to
all rational
own
his
own
It
happiness.
should be added that although (as we saw in the previous section) the principle of self-love may override principles of skill in particular cases, there is no general incompatibility between the two kinds of principle. So far as an agent is rational, he will seek to acquire and use the skill necessary for the realisation of those ends in which his happiness 3 is to be found.
1
Chapter VIII
8.
elaboration would be necessary to meet charges of over-simplification, but I must leave the reader to fill in the kind of qualifications which he may deem necessary. I tried to deal with some of these problems in The Good Will, especially Chapter VIII. 3 This is, I think, brought out better by the German word '/Clugheit', which I have translated as 'prudence* or 'self-love'. Kant defines it in Gr., 41 6 n. 41 n. as *the sagacity to combine all one's purposes to one's own lasting advantage*. He recognises a second kind of *Klughei? namely, Weltklugheit (or worldly wisdom) which is subordinate to the first and is more explicitly concerned with means ; for it is described as skill in influencing others so as to use them for one's own purposes.
Much
IX
4]
The principle of morality.
93
4.
So far we have recognised that while the principles of practical reason are objective, they are at the same time conditioned by the desires and character of particular agents. Many of us, perhaps most of us, may be content with such principles and may regard an enquiry into further principles as a mere waste of time. Nevertheless the question must arise whether there can be principles of practical reason which are objective without being conditioned. Such principles would be unconditioned or absolute; and on Kant's view the function of conceiving the unconditioned which appears to be implied by our knowledge of the conditioned is one which belongs to what he calls the power of reason in
such concepts of the unconditioned is 'Idea* of 'Idea Reason'. 1 Kant always insists that we find it elaborately necessary to conceive such Ideas and that when properly understood they can play a useful part in our thinking. Nevertheless he holds and this is the essential point that, so far as speculative reason is concerned, they cannot give us knowledge of reality. Human beings can have knowledge only so far as their concepts refer to objects given to the senses. Since all objects given to sense are conditioned, and since the unconditioned is not given to sense, the Ideas of reason do not refer to possible objects of experience and cannot give us knowledge. Our present concern, however, is not with speculative, but with practical, reason. We are discussing, not the ultimate nature of reality, but the principles on which a rational agent would necessarily act if reason had full control over his desires. The conditioned objective principles which we have hitherto examined have suggested to us the Idea of an unconditioned objective principle; and such a practical Idea might conceivably play in action a part similar to what Kant calls the 'regulative', as opposed to the 'constitutive', function of speculative Ideas in our thinking. It might set before us an ideal which we might continually strive to approach, even if it were beyond our attainment. An unconditioned objective principle would omit all reference to the desires and character of particular agents it would omit all reference to particular ends and even to the comprehensive end of the agent's own 2 happiness. Such a principle, as we have seen, could only be the form of a a or formal the principle, principle principle, so to speak, of having an objective principle and so of being reasonable. Kant has described it as a principle of law-abidingness, a principle of acting in accordance with universal law as such.
or
his special technical sense. The technical name for
more
Such a principle may appear completely empty, and we must be on our guard as Kant himself well knows3 of omitting the conditions which
1
use the
word
When 'idea* with a small V is used,

often as a translation
8
'Idea' with a capital

it is
'I'
meant to be taken of the German word 'Vorstellung

'.
as translation of the German word 'Idee? in the ordinary English sense

s
Chapter IV
3.
K.r.V.,
A674
Byoa,
94
[IX 4
alone render a concept intelligible, and then imagining that what is left will be supremely intelligible and will give us some very special sort of insight. But where, as in this case, a principle is practical, and not theoretical, we may be able to fill up the empty form in action. If it is possible or even conceivable for a rational agent to act on such a principle in virtue of his rationality, then the principle itself will not be without
meaning.
All this discussion would be completely in the air, were it not for our practical acquaintance with moral ideals and moral standards. Unless morality is an illusion, we must say that a morally good man is not merely one who is skilful in attaining his ends or prudent in co-ordinating a whole series of ends in an organised life. morally good man is one who acts upon a law holding equally for himself and for other rational agents and who follows the principles of skill and of self-love only so far as these are compatible with such a universal law. For him the principles of skill and self-love are conditioned, not merely by reference to his desires, but also by reference to an absolute and unconditioned law. No doubt there are many today who regard such moral beliefs as mere illusions to be explained away by various ingenious hypotheses. Apart from the extremer zealots who use their reason only to deny reason in action and even in thought, even a moderate man may feel considerable qualms. He may find no difficulty, since he is himself a rational agent, in understanding that a rational agent, so far as reason has full control over his passions, will necessarily act in accordance with
conditioned principles of skill and self-love. But lie may well ask what is our warrant for saying that a rational agent, so far as reason has full control over his passions, will necessarily act in accordance with the unconditioned principle of morality. The difficulty arises, not only because the concept of an unconditioned principle is harder to understand, but also because in action we are conscious of acting more from motives of skill and self-love than from motives of morality. Nevertheless if we conceive a rational agent in a world of rational agents, and if we suppose that in him practical reason were not at the service merely of his own inclinations and his own happiness, what can we say of the principle on which he would necessarily act? I do not see that we can give any other answer than Kant has done namely, that he would necessarily act on a principle or law valid for every rational
agent as such. Admittedly this would have no meaning for us unless we were acquainted with what at least seems to be the ideal of moral action. It acquires a meaning because down the ages men have come more and more to recognise no doubt amid much confusion that a good man, whatever else he does, acts or endeavours to act on this principle, and only so is a good man. Thus our analysis of practical reason brings us to the same result as .our previous analysis of the implications of ordinary morality. We have been led to conceive a special kind of practical reason which Kant calls
'pure practical reason'
satisfying desire,
reason not simply functioning for the sake of but determining action independently of desire. Such a
IX
5]
OBJECTIVEPRINCIPLES
95
pure practical reason would necessarily, if it had full control over desire, act in accordance with an unconditioned and objective principle such as appears to be manifested in moral action a principle or law valid for every rational agent as such irrespective of his particular desires for
particular ends.
5.
Conditioned and unconditioned principles.
I have followed Kant in speaking of the principles of skill and selflove as conditioned and of the principle of morality as unconditioned, but it may be necessary to add certain qualifications. It is clear enough that particular principles of skill are objectively valid only if a certain end is desired, and that particular principles of self-love are valid only if the character of the agent is such that he will find happiness in acting according to these principles. But these particular principles are all applications of a general principle, of skill or of self-love as the case may be. The general principle of using the most effective means to a desired end is not itself conditioned by desire for a particular end only its application is so conditioned. And the general principle of seeking the maximum satisfaction of desires or the maximum integration of ends is not itself conditioned by the particular character of the agent only its application is so conditioned. On the other hand the general principle of skill is, so to speak, taken up into the general principle of self-love. Self-love has in view a more comprehensive end, and in reference to that end it will use the most effective means. Here also it is rather particular principles of skill that are conditioned by the principle of self-love. The general principle of skill is similarly taken up into the principle of morality a good man will use the most effective means to the attainment of his moral ends, and it is only particular principles of skill which are conditioned by the principle of
:
morality. Similarly particular principles of self-love are conditioned by the principle of morality, but it is harder to say whether the general principle of self-love is so also. Certainly on Kant's view the general principle is not opposed to morality: the good man has a right, and even at times an indirect duty, to seek his own happiness. At times, however, a good man, so far as he is good, will necessarily be prepared to sacrifice his whole happiness and his whole life, and this looks as if the general principle of self-love, and not merely its particular applications, is conditioned by the principle of morality. Even if this is so, it seems clear that the general principles of skill and self-love (if we can distinguish these from their particular applications) are not conditioned by particular human desires, 1 although their application is so conditioned. So far both general principles seem to have an unconditioned character, and to have it in virtue of being objective
principles.
When we
1
turn to the unconditioned moral principle

the fact that
we
find that have
its
They are, however, conditioned by
men, as
finite beings,
desires.
96
[IX APP.
applications are not conditioned by the agent's desire for any particular end or even for happiness. Nevertheless it would be a mistake to suppose that its applications were in no sense conditioned by human desires. Kant holds that the good man will necessarily act, for example, on the principle of seeking the happiness of others, and that the way in which he does this must depend, partly at least, on the desires of others. Kant may consider the different ways of making others happy to come within the 5 'play-room or 'latitude , where prudence decides and not morality ; but even so a good man, as a good man, will necessarily display such prudence. I do not wish to obscure the clear difference, which it is Kant s great merit to have emphasised, between morality on the one hand and skill and self-love on the other ; but if we are entitled to distinguish between principles and their application, the general principles of skill and prudence are in a sense unconditioned and the application even of the unconditioned moral principle must be in a sense conditioned, although not conditioned by the agent's desires. 1 It may be added that one principle is called higher than another because it takes a more comprehensive view. Self-love takes into account more ends than the one considered by mere skill and morality takes into account other agents and their desires, not merely the desires and ends of the one agent considered by self-love. This may not be the whole difference, but it is a difference ; and it helps us to understand how the lower principle may be conditioned by the higher.
5 5
:
APPENDIX
KANT'S VIEW OF REASON

i.
Different senses of 'reason*
IN Chapter VIII 2 two different traditional senses of 'reason were mentioned, two senses which are accepted by Kant. But in addition he has a special usage of his own, and this must be grasped, at least in a simplified form, if we are to follow adequately his argument, and especially his argument in Chapter III of the Groundwork. 'Reason', as we saw, was traditionally used for the general power of thinking, as opposed to merely sensing or imagining. This general power of reason which we may call reason in general 2 was supposed to display itself in three main activities, and accordingly is given the name of three different powers. Firstly it conceives or entertains concepts, and is then called understanding. Secondly it applies concepts to given objects, and is
1 On the other hand, even the general principles of skill and self-love rest on the supposition that we seek ends because of our sensuous desires, and they have no meaning apart from this supposition. The general principle of morality does not rest upon, nor is it conditioned by, such a supposition. 1 Or 'reason as such1 .
IX
APP. 2]
97
then called the power ofjudgement. Thirdly, it makes mediate inferences, and is then called reason in a narrower sense. These activities may be described as the logical use of reason. Broadly
speaking, concepts so far as we ask after their origin are supposed to be derived by abstraction from objects given to sense, and they are applied to objects given to sense. But in addition and this is Kant's special doctrine reason in general is alleged to produce certain concepts as a result of its own activity, or as a result of consciousness of its own activity, concepts which can in no way be derived from sense, or from objects considered merely as sensible. Such concepts are therefore a priori that is, they are not derived from sense. They are products of what Kant
calls generally 'pure reason'.
So
we may
far as these concepts give us knowledge of objects, or claim to describe their use as the real use of reason.
1
do
so,
But here we get into terrible complications, most of which must be passed over. It is enough to say that the power of entertaining these a priori concepts is called 'pure understanding', and these concepts themselves are called 'categories of the understanding*. Furthermore, by an activity of inference working on the categories, we arrive at another kind of a priori concept called 'Ideas of reason', and the reason which entertains these concepts is called 'pure reason in a special and narrower sense than that used above. The application of all these concepts in judgement we may here ignore.
1
2.
The
category
of cause and
effect.
perhaps follow this special Kantian doctrine more easily, consider a category which is of direct importance for his moral philosophy. Reason in general as is seen even in a hypothetical judgement as well as in a syllogism necessarily makes inferences from grounds to consequents, and as it becomes explicitly or reflectively conscious of its own working, it conceives or entertains the concept of ground and consequent. The concept of ground and consequent is certainly not given to sense. It is therefore an a priori concept ; it is entertained by pure understanding ; and it may be called apure category of the understanding. Yet when we call it a pure category of the understanding, we imply, according to Kant, that the concept in question can be applied, or can be thought to be applied, to some sort of object. This is why we spoke above of a 'real' use as opposed to a merely 'logical' use. Can we apply the concept of ground and consequent, not to propositions, or parts of proposithat is, to individual things which can be given to tions, but to objects sense (or given through sense) ? Kant does not believe we can know any objects other than those that can be given to sense, and he maintains that when we apply the pure category of ground and consequent to objects, we use it no longer as the pure category of ground and consequent, but
if
We may
we
Some
of these complications are necessary, but
Kant also adds

identical
unnecessarily to our
'reason*.
difficulties
by using the word 'understanding* as if it were
with
98
as
[IX APP.
the category (sometimes called by commentators the schematised category) of cause and effect. The reasons behind this transition are so difficult that those of my readers who are unacquainted with the Critique of Pure Reason would be well advised to take this transition for granted, to skip my attempt at
explanation, and to go straight on to 4. It is sufficient to see that a cause is a ground which necessarily precedes its consequent in time, and an effect is a consequent which necessarily succeeds its ground in time. cannot understand cause and affect except by reference both to temporal succession and to the concept of ground and consequent. This is one reason why modern empiricists seek to abolish the concept of cause and effect or to reduce it to invariable succession.
summary
We
3.
The schema of regular
succession.
On Kant's view, if we are to apply the pure category of ground and consequent to sensible objects, we can do so only by supposing that where there is a regular succession of A followed by B, then A is to be taken as the ground and B as the consequent. 1 But it looks as if we had no more
right to suppose that there must be regular succession in the sensible world than we have to suppose grounds and consequents in it. This is where Kant displays his greatest ingenuity, which, whether it is successful or not, is seldom adequately appreciated by modern thinkers. Kant argues that if we are to distinguish an objective succession of events in time from our own successive apprehensions 2 and we all do this in practice we can do so only on the supposition that in an objective
succession each kind of succeeding event is always necessarily preceded 3 by the same kind of preceding event. More simply, there must be regular succession in events, if they are to be distinguished as objective events find our in time (and not merely as a succession in apprehension). object for the category of ground and consequent in the regular successions then have a preceding whose necessity Kant believes he has proved.
We
We
event which is a ground, and a succeeding event which is a consequent ; or in other words we have a cause and an effect. All this is too summary to be convincing or even clear, and it cannot be precise, since it slurs over difficulties and complications. Furthermore, we cannot deal with one category in isolation our statement really presupposes the category of substance. Nevertheless we may see or at least assume that necessary succession in accordance with a rule is what Kant calls the schema which justifies us in applying the pure category to objects,
This statement requires qualification. In looking round a house we apprehend successively a series of what we believe to be simultaneous events or existents, while in watching a ship going down stream, what we successively apprehend namely, the ship's changes of position we believe to be also objectively successive. We always distinguish objective successions from merely subjective ones, and without this distinction there could be no human experience. 8 This does not mean that in every objective succession of events the earlier event is necessarily the cause of the later it may be or it may not, and the test, speaking roughly, is regularity or repetition. But every objective succession must be causally determined.
*
:
IX
APP. 4j
99
and which in combination with the pure category gives us the schematised category of cause and effect. The category of cause and effect necessarily applies to all objective events in time, if we can establish this schema (as Kant thinks he can) as necessary to experience of objective events in time. I would add, since it will be highly relevant to later discussions, that if the mind thus imposes its own categories on the objects which it knows,
clearly can know reality only as it appears under human conditions, and not reality as it is in itself. This Kant regards as already established, since on his view time, and also space, are merely forms under which we must intuit reality, forms which our sensibility imposes upon what is given to our senses, and therefore forms which need have no counterpart in
it
reality as
it
really
is.
4.
The Idea offreedom.
If we are to follow Kant further, we must assume him to have proved, he claims he has, that every event must have a cause. The category of cause and effect having been established, what does pure reason make
as
out of it? In its logical use reason not only infers consequents from grounds, but it also, according to Kant, seeks to find grounds for supposed consequents. The conclusion of a syllogism may follow from its premises but reason is not satisfied thereby till it in turn finds the grounds for the premises, and the grounds for these grounds and so on. Reason must at least conceive the Idea of the totality of the grounds, a totality which, as a totality, could have no further ground, and so would be itself ungrounded, unconditioned, or absolute. Something like this must happen to the category of cause and effect. Every event must have a cause, but the cause itself (or at any rate what Kant calls its 'causality', i.e. its causal action) is also an event which must have a further cause, and so on ad infinitum. Reason must therefore conceive the totality of causes for any given event, and this totality of causes, because it is a totality, cannot itself be caused. So we come to the necessary Idea of an unconditioned or uncaused cause, a spontaneous action which produces effects, but is not caused to do so by anything external to itself. The concept of such an unconditioned absolute spontaneity is the transcendental Idea of freedom, a concept not of pure understanding but of pure
;
reason.
The Idea of freedom is a concept which pure reason cannot but enteryet if we suppose, as many do, that this Idea can give us knowledge of any reality, we fall into illusion. The supposedly real use of pure reason in this way is very natural, and even irresistible, but it does not thereby cease to be illusory. Our reason is discursive; that is, it can know objects
tain
;
by means of its concepts only if it can apply these concepts to objects given But the absolute spontaneity of a cause cannot be given to sense; and if it could be so given, it would itself be an event in time which would necessarily have a preceding cause, and so could not be spontaneous. Indeed since all objects of experience must be conditioned, and since all
to sense.
IOO
[IX APP. 5
Ideas of reason are concepts of the unconditioned, no Idea of reason can have any corresponding object of experience. We can even describe an Idea as a concept which can have no corresponding object. It is, however, not unthinkable that such an Idea might refer to something in the world as it really is, though not to the world as we experience it, if it is true, as we have argued, that we cannot experience the world as it really is. But this is no help to us in the theoretical use of reason, since we have no knowledge of the world as it is, but only knowledge of the world as it appears to us in experience. The only use of the Idea of an absolute that is, totality of causes is, for theoretical reason, a purely regulative one it does not help us to know objects (and so it is not what Kant calls 'constitutive*} but it encourages us, when we have discovered a cause, to seek for a further cause, and so on indefinitely. Nevertheless if practical reason necessarily acts on the supposition that it is free, and if it thereby acts in accordance with the Idea of freedom, it may turn out to be important that there is nothing self-contradictory in the Idea. As we have said, it is at least not unthinkable that the Idea might refer to something not in the world as we experience it, but in the world as it really is. For the purpose of action this might be enough. Even from this imperfect attempt to summarise the Critique of Pure Reason in a few pages we can see at least that Kant is consistent in ascribing to pure practical reason the Idea of an unconditioned good, the Idea of an unconditioned principle (or law), and the Idea of an unconditioned
1
or categorical imperative.
5. Different kinds
of concept.
It
may be
and
useful to
add parenthetically a
:
list
of the different kinds of
concept recognised by Kant 1. Empirical concepts drawn from experience, such as the concepts
of'red'
2. 3.
'cat'.
Arbitrary concepts, such as the concept of 'chimera'. Mathematical concepts, such as the concept of 'triangle*.
These he
regards as a special kind of arbitrary concept. 4. Categories of the understanding, such as the concepts of 'substance' and of 'cause and effect', which must apply to all objects of
experience.
5. Ideas of reason, such as the concepts of 'God', 'freedom', and immortality', which cannot apply to objects of experience.
6. Intuitive understanding.
An Idea of reason is based on our knowledge of something conditioned

is a concept of the totality of its conditions. Such concepts are regarded by Kant as concepts of a systematic whole, or of a system as opposed to a mere aggregate. In a system the whole is the unconditioned
and
IX
APP.
6]
IOI
condition of the parts, and the Idea of the whole is logically prior to the 1 parts, while in an aggregate the parts are prior to the whole. We should not think of reason, in Kant's technical sense, as a mysterious faculty, or of Ideas as recondite concepts familiar only to philosophers. All human thinking has a natural tendency to pass from given parts to the conception of a wider whole, and ultimately of the widest possible whole in which the parts would become fully intelligible. In science, for example, the Idea of a complete system is at work in our thinking, like a kind of seed or germ, although science may have to proceed a very long way before the Idea can be seen in a clear light. 2 Once we have grasped the Idea, and with it the purpose and form of science as a whole, we should be able, according to Kant, to determine the necessary articulations of science into its various branches by a priori reasoning. An Idea of reason, though it is the Idea of a complete systematic whole and so of something individual, is still a concept, and there is always less in a concept than in the individual object known by means of it. If by means of our Idea we are to know, and not merely to conceive, an individual object, this object must be given to sense, and in fact no complete and unconditioned individual whole can be given to sense. This suggests to Kant the thought of an intelligence different from ours, one in which there would not be this divorce between concepts and the individual reality as given to sensuous intuition. Such an intelligence would not require to await sensuous intuitions given to it from without. It would possess what Kant calls 'intellectual intuitions' or an 'intuitive
understanding' its thinking, if we can call it such, would be intuitive, not discursive like our own and in its thinking it would know, without further aid from sense, an individual and intelligible reality. For such an intelligence there would no longer be a distinction between thought and
;
;
its universals would also be individuals, or in Hegelian language would be concrete and not abstract. 3 For it, so far as I can see, there would also be no difference between thought and action, since its thinking would be essentially creative of reality; 4 but this conclusion does not appear to be drawn explicitly by Kant himself. We have no means of knowing whether an intelligence of this kind is even possible, let alone actual. Kant merely uses the concept of such an intelligence in order to bring out by contrast the limitations inseparable from our finite human understanding. Nevertheless he presumably believes that such an intelligence would have to be ascribed to an infinite
sense:
being.
surely there
Speculations of this kind may appear to be merely fantastic. Yet must be a curious limitation in my human understanding if I can know only a world in space and time 'whose margin fades for ever
1 8
B86o~i. Logik, Einl. IX, 72 ; K.r.V., A832-3 K.r.V., A834-5 6862-3. Kant does not use these terms but contrasts instead the 'analytically universal' with
the 'synthetically universal'; K.U., 77, 407 ( 348-9). 4 This might also help to explain how God could act without sensuous desires besides knowing without sensuous intuitions.
IQ2

for ever
[IX APP. 6
when I move' ; and it seems natural enough to conceive an understanding which would grasp the whole of reality at once and would find it intelligible as it is not intelligible to us. I mention this doctrine, however, only in order to suggest that on his own principles Kant appears at times to claim too much for human reason. He speaks of it as if it could go altogether beyond sensibility, 1 and gives us the impression that on his view what is absolutely good for human reason must be good even in the eyes of God. 2 Is he entitled to do this on his own premises? Our Ideas of reason are only the categories of the
and
infinite
understanding pushed as far as the unconditioned; and the categories, even if they are not, in their pure form, bound up with space and time, are merely principles for combining given sensuous intuitions in accordance with certain rules and could have no significance in relation to the 8 knowledge of an intuitive understanding. What is more, so far as the Ideas of reason can be applied at all to individual objects, this can be done only through the medium of the categories. 4 In spite of all Kant's ingenuity it is hard to sec how he can attain to the Idea of an absolute
goodness which
reason.
is
not in some
way
relative to the limitations of
human
Gr., 452 85-6. K.U., 76, 266 (
= 338)
6V.,
439
70.
7,
K.r
V.,
6145.
M.d.S., Rechtslehre,
253.
CHAPTER
THE GOOD
i
.
The good
in general.
PRACTICAL reason besides having principles has also objects, and these, 2 according to Kant, are the good and the bad. The good is defined as 'a of the of necessary object power appetition (Begehrungsvermogeri) in accordance with a principle of reason'. To get a definition of the bad we have only to substitute 'the power of aversion (Verabscheuungsvermogeri) for 'the power of appetition'. The principles of reason here mentioned are certainly objective principles, not mere maxims otherwise the good and bad would be merely what we willed or rejected. 3 A power of appetition which is determined in accordance with a principle of reason is simply a rational will. 4 The good is therefore what a rational will, so far as it had complete control over the passions, would necessarily will, and the bad is what it would 6 necessarily reject. In more technical language we may say that the good
3
:
is a necessary object of a rational will in accordance with an objective principle of practical reason. It is all-important for Kant that the good falls under an objective
under a concept. Because of this we can say that certain be good we are not left to judge each individual instance by immediate feeling, as we are in estimating the pleasant and even the beautiful. As Kant says in the Critique of Judgement,* 'What pleases by means of reason through the mere concept is good/ Our apprehension of the good is accompanied by an emotional satisfaction of approval or esteem or even (in the case of moral goodness) of reverence but this satisfaction is the result of our apprehension and not its ground ; and the apprehension itself is by means of concepts or principles, principles
principle
and
so
kinds of thing must
1 1 have discussed the subject of this chapter at greater length in Kant's Idea of the Good (Proceedings of the Aristotelian Society, 19445, pp. ixxv). *K.p.V, 9 58=*= 177 (= 101). In K.p.V.> 57 =176 (= 100) Kant speaks, perhaps loosely, as if an object of practical reason were simply *an effect which can be produced
through freedom*. 3 See also the definition of the practically good in Gr.î 3 = 37. The whole passage Gr. 9 412-13 is presumably concerned with the good in general, although Kant may have in mind particularly the moral good. A maxim would at most give us only *the seeming
good'.
*K.U.,
6
4,
209 (=13-14)
necessity in question is what beings also a subjective necessity.
The
Kant
calls objective necessity
it is
not in
human
4,207(=io).
t
I0 3]
IO4
[X
of practical, not of theoretical, reason. Unless we had an internal acquaintance with action, the word 'good' would have no meaning for us; and every volition, perhaps every wish, is an implicit assertion of goodness or badness, which may, however, be mistaken. Kant's definition is inadequate even from his own point of view. If we took seriously the suggestion that the object of a rational will is merely an
effect which it produces (or endeavours to produce), this would deny goodness to actions, and still more to wills goodness would be confined to the states of affairs which we were able to produce. This is not at all Kant's view. He is far more interested in the goodness of volitions and actions, and of the will itself, than in the goodness of mere products. Clearly we must extend the meaning of 'object' to cover actions. This is reasonable, since an action may be regarded, though not without danger of error, as a change in the world produced by our will. But this would not be enough, since on Kant's view it is the will itself which above all must be esteemed as good. On his view a will which acts on objective principles is a good will; an action which is willed in accordance with objective principles is a good action and what a good will would seek to produce or use in accordance with objective principles is a good state of affairs. Even this may be an over-simplification and we must remember that in order to be good, or at any rate to be morally good, an action must not merely accord accidentally with objective principles it must be willed in accordance with these principles. Nevertheless we may say that to be good is to be in accordance with objective principles of practical reason. To understand this fully we must see how Kant works out his
:
principles in
more
detail.
that we are so far concerned only with the concept of good in general. As we have already recognised three kinds of objective principle, there must be at least three different kinds of good or, perhaps better, different senses of good. To these we must now turn.
In
all this
we must remember
2.
'Good for' and 'good at\
The objective principle of using the most effective means towards an end gives us the concept of the 'good as means' to something else. 1 Things which are good as means may be described as 'useful' or as 'good for' something, as a natural product like coal is good for burning, and an artificial product like a knife is good for cutting. Actions which are good as means may be described as 'skilful', though this term is rather too
narrow, since
it fails
to cover such qualities as perseverance in action,

:
which obviously are good as means. Actions may also, at least in some cases, be described as 'good for' something for example, walking is good
for health.
When we turn
to the
man who
at'
to the will manifested in these actions or, better, wills these actions, we may describe him as 'skilful' or as
'good
good
l
certain things. may also say that he is good as a cricketer or as a soldier, or more simply still that he is a good cricketer, or a good
414
We
Gr. t
= 38.
X3]
THE GOOD
105
soldier, or even a good poisoner. All these phrases indicate that the goodness in question holds only within a limited sphere. It is not an unconditioned, but a conditioned, goodness. 1 Although good in this sense is conditioned and relative, it does not to be It is true that a cease be no knife objective. thereby may good to me, if I do not want to cut but this does not alter the fact that it is good for cutting and that it would necessarily be so regarded by any rational agent
;
who understood what is meant by 'cutting'. Some means are clearly more effective than
others,
and
so far
means
can be regarded as better or worse. There are thus degrees of goodness. But all this tells us nothing about the goodness of ends; although things (including actions and even men) have a limited goodness or badness merely as means, they have another quite different goodness or badness
according as the ends to which they are means are themselves good or bad.
3.
My good.
We might expect that the objective principle of happiness would be concerned with the goodness of ends, or at least with the relative goodness of ends. Unhappily Kant's vague and varying views about happiness
produce a corresponding vagueness on
this side of his doctrine. certainly regards happiness as my weal, rny welfare, my wellbeing, and as a good which men naturally seek. So far as he takes happiness to be merely a state of continuously pleasant feeling, other things (including actions) are lumped together merely as means to this state. Their goodness is then merely goodness as means, and though happiness itself is
Kant
my
here means only difference in the goodness of things as a means to happiness, rather than to other ends, is that happiness is an end which in fact all men seek, however much they may fail to get it because of ignorance or folly or passion. 2 If we regard the pursuit of happiness as requiring a choice of ends as well as of means, and indeed as an attempt to integrate our different ends within a whole organised life, the position is different. This view becomes more prominent in Kant's later works, but it is not wholly absent even from the Groundwork. If it is true, as it surely is, that a rational agent, so far as reason has full control over his passions, will necessarily seek to integrate his various ends, then happiness as a comprehensive end is his good, and separate ends have their goodness as elements in this comprehensive end. They are not merely means to happiness, but elements in happiness, or as they are sometimes called 'constituent means'. The goodness of these ends is also reflected back, as it were, on the means to these ends, so that they
clearly regarded as good,
it
may seem doubtful whether 'good'

The
anything other than 'pleasant'.
Gr.,
438
= 68.
this view of happiness, however, Kant presumably regards continuous pleasant feeling as an end which a rational "finite being moral considerations apart would necessarily pursue, if reason had full control over passion. If so, happiness is an objective good.
2
Even on
IO6
[X
too can be judged with reference to the agent's good. They may still be abstractly good as a means to his particular ends, but they may be such that he may, or may not, be able to find his happiness or his good in making use of them. In this way we arrive at the concept of 'my good' or 'the good for me'. This is not to be taken merely as what I happen to think good for me. I may be mistaken in what I think, and although this good is relative to my character and desires and needs, it does not thereby cease to be objective. On the other hand, what I think good for me is a factor in determining what is good for me. Apart from the fact that my thinking something to
clear to myself, I am on the whole unlikely to find good and happiness in things which other people, but not I, think will be good for
be good for
make
me may arise from some unconscious need which I my
can hardly
my
always bitter against those who seek to make some one else way and not in his. Many things may be described as good to me or for me, and in them I may in a sense find my good but things proper (as opposed to actions) appear to be good only as means to the ends which constitute my happiness. This is particularly obvious in the case of things I do not really like thus rice pudding may be good for me because it is a means to the health of my body and even criticism may be good for me because it is a means to the health of my soul. In the case of gifts of fortune, like health and wealth, power and prestige, it is the possession or use of these things which constitutes my good and the same is true of the gifts of nature, powers of mind like intelligence and judgement, or qualities of temperament like animal courage or natural moderation. 1 Broadly speaking, I find my good in activities rather than in things, in work and play, in love
me. Kant
is
happy
in their
and
their
affection, in philosophy and art. Different people will naturally find good in different activities, and it is obvious that the same may
man
find different activities
more or
less
good. Here too there are degrees of
are not in the habit of describing a man as good good at integrating his ends or securing his own happiness. We describe him rather as sensible, competent, prudent, or even wise. There seems to be no precise word in English for this kind of goodness perhaps it is covered better by the German word 'Klugheif but there is no reason to doubt that a man of this character has a will which is 'in some sense good'. 2 Happiness in the narrower sense of continuous pleasant feeling may be one of the ends which a rational agent, qua rational, must necessarily seek ; but it is certainly not the only end, and experience tends to show that the less we deliberately pursue it, the more likely we are to attain it. On the other hand, if a man could attain in an organised life all the ends he set before himself and yet could experience no feeling of pleasure, we should not call him happy. Such a state of affairs may be impossible, and if it were possible it would surely indicate the presence of some desperate mental disease. In any case pleasure is not only one element in happiness,
goodness. Curiously enough,

is
we
merely because he
Compare
Gr. t
393-4 =11-12.
Gr.,
414
= 38.
4]
THEGOOD
IO7
a criterion of happiness, and this is one reason why it is so often mistaken for happiness itself. Even at the cost of repetition I must emphasise that my happiness and the elements in my happiness are objective goods, although they are relative to the agent. good is not merely what seems good to me. It is not even what would seem good to any rational agent who put himself dramatically in my place. In order to see my good another rational agent must indeed put himself dramatically in my place he must make himself conscious in imagination of my needs and my desires and must judge in the light of such knowledge; he must also remember that in these matters there is great latitude for free and even arbitrary choice, and that the free choice of the individual is a potent factor in determining the possibility of happiness. But when all this is said, then, in spite of the uncertainty which shrouds these matters, it remains true that there are some activities which an agent of a particular type can combine in an organised life, and there are others which he cannot. I can find my good in some activities and not in others, in some kinds of life and not in others. Most important of all, although my good is relative to my needs and desires, and is in that sense conditioned, there is another sense in which my good and your
it is
My
good,
as
a good
my happiness and your happiness, must be regarded simply as by any impartial rational agent, and in that sense may be described an unconditioned good. 1 This I believe to be fully accepted by Kant,
but when we get to this point we are, I think, passing over to the moral point of view ; and from this point of view my good may cease to be a good when it conflicts with the good of others. In that sense my good is still a conditioned good.
4.
The moral good.
The unconditioned objective principle of practical reason is concerned with the unconditioned or moral good, and indeed it was by an analysis of the moral good that we originally came to entertain this
we have considered an agent to be good in limited senses, (i) he acted reasonably, i.e. skilfully, in satisfying his desires and so attaining his ends, and (2) so far as he acted reasonably, i.e. wisely or prudently, in pursuing his own happiness and so attaining the maximum satisfaction of his desires. In all this the prior condition of reasonable action has been the agent's own desires and needs. But nobody (unless he is momentarily corrupted by a false philosophy) pretends that such a reasonable agent is necessarily a good man. A good man is one who acts
so far as
concept. Hitherto
reasonably, not for the sake of satisfying his desires, but for the sake of
1 This is the other side of the statement made in Chapter IX 5 that the general principle of self-love like that of skill is not conditioned by the desires of the particular agent, however much its applications" may be so conditioned. Compare also the passage quoted in Chapter IX 3, in which it is said that reason has an office which it cannot refuse the office of seeking happiness.
I08

so of satisfying his reason. This
is
[X
in his technical as of of the way goodness speaks principle perhaps repellently a purely formal principle of following universal law as such : it leaves out needs as its prior condition. But in holding reference to my desires and this Kant surely accords with ordinary moral judgement. To ask whether an action is morally good is not to ask whether it enables the agent to satisfy a desire or to realise his ideal of happiness ; nor do we require to know the answer to the last questions before we can answer the first. Kant may put this view so strongly that to many it appears paradoxical, but this is the simple and obvious truth which he is endeavouring to teach. Nevertheless it is impossible, at least for man, to act reasonably in a vacuum, and one of the commonest arguments against Kant is that it is useless to speak of moral goodness unless we suppose that other things are good besides the mere will to act well or to act reasonably. This is why I have had to insist so strongly, and with even more detail than Kant himself, that he does recognise the other objective, although conditioned, goods which we have discussed. Kant knows no man better that a good man will necessarily seek the happiness of others as well as of himself, but he is surely right in saying that he will not seek the happiness of others for
acting reasonably and
why Kant
my
the sake of his own. There are certain ends which reason must set before own sake or, as Kant puts it, for the sake of the law as such. 1 And a good will is not good merely because it produces or seeks to produce the happiness of the agent or even of other agents it has a value far above that of mere happiness, and for the sake of that value even happiness and
itself for its
:
life itself
may have
to
be
sacrificed.
obvious that moral goodness does not characterise things, but actions and, above all, persons. Nor should it surprise us that an action is morally good only so far as the agent acts on a principle. The same thing is true even of skilful and prudent actions. We distinguish between a skilful action and a mere fluke ; and if a man found happiness in a wife whom he had won in a lottery, we should count him fortunate, but not wise.
It
is
5.
The
teleological
view of good.
It may still be thought that it is not enough to recognise, besides the moral good, the good as the useful and the good as happiness or as an element in happiness (even if happiness is taken in a wide and Aristotelian sense). This may well be so, though Aristotle himself got on not so badly with no more. Certainly I do not pretend that Kant supplies an adequate philosophy of action or of the good, and he seems to me to rest too much on the category of means and ends. What I maintain is that he is not so
inadequate as
1
is
commonly supposed.
morally good action is, on Kant's view, an action done for the sake of law as such, but it is not good merely as a means to an abstraction called the law : it is, on the contrary, good in itself, good in virtue of the objective and unconditioned principle embodied in it and not in virtue of the desires which it satisfies or the ends at which it aims. Moral interest, as we have seen in Chapter VTI 3, is an immediate interest in the action itself.
X5]
It
THEGOOD
109
may well be the case that the value of things like philosophy, science, and art cannot properly be estimated in terms of human happiness, however widely this may be interpreted. These activities have standards of their own which are neither eudaemonistic nor moral, as Kant recognises. Nevertheless he might maintain that their goodness, as opposed to their validity or truth or beauty, must be relative to the objective principles of practical reason, and that they have their goodness as elements in a moral, and not merely in a happy, life. There is, however, in Kant a further strain which may be described as Aristotelian or 1 teleological. We have already met this in his view that reason in man a a have and function must special special end ; and he appears to hold that the development and use of the powers which distinguish man from a mere animal have a value different from happiness inasmuch as they are realisations of distinctively human ends. This comes out clearly when he distinguishes between the perfection which the good man will seek for himself and the happiness which he will seek for others. 2 This special value or goodness is spoken of only in a moral context, but nevertheless its
implications may be far-reaching. Such a teleological view was too easily accepted in the Eighteenth Century. It is perhaps too easily rejected now, since Darwin propounded his doctrines of evolution. All we need note here is that although Kant may make an uncritical use of it in his moral philosophy, his doctrine of teleology is not to be interpreted in a crude way. Although we have to consider some things as if they had been made for a purpose, and although we are unable to understand them in any other way, this does not mean that we profess to know they have been made for a purpose, or even to think it. The concept of purposiveness in nature is what he calls 'regu3 lative', not 'constitutive'. Even modern mechanistic zoologists sometimes had a purpose, although they would strenously various as if organs speak deny this to be the case. Perhaps the assumption that practical reason has a function or purpose and that the fulfilment of this function or purpose must be good is the root
assumption of Kant's whole moral philosophy, and indeed of almost all Western moral philosophy. Perhaps, when properly understood, this is a legitimate and even necessary assumption, like the assumption that in thinking we must trust our own reason. This is opposed, as I understand, to great Eastern philosophies which regard the only good as the complete annihilation of all activity. It would be ill to speak lightly of views of which one has little knowledge, out we may perhaps observe tentatively that when this view is maintained in the West, it seems to be often an expression of the natural desire for rest after toil, or at times even a reaction from an unsuccessful effort to use practical reason in the pursuit of enjoyment whether on the higher level of art or the lower level of sensuality. Kant himself believes that reason is a poor instrument for seeking happiness and that for this purpose it must inevitably fail ; but he concludes from this
1
Chapter II
K*. (7.,
8.
See also Chapter
XV
4-6.
2
8
M.d.S., Tugendlehre, Einl. IV, V,

76,
and IX, 385-94
230-40.
404(^34 4)-
IIO
[X6
that it must serve some higher purpose than the satisfaction of desire, not that we should abandon reason altogether in favour of instinct or even of unconsciousness. It should be added that even if certain human activities have a goodness other than that of constituent elements in human happiness, and even
goodness is not to be regarded as moral goodness (which may have been Kant's view), this in no way detracts from the supremacy of moral goodness. The goodness of specifically human activities is still a conditioned goodness in the sense that in a concrete situation they cannot be 1 good unless they are compatible with moral law.
if this
6.
The
realistic
view of good.
as to the defects of Kant's account of a certain attractiveness in his attempt to see the different senses of good as constituting a developing series under a common concept. Nor is it difficult to understand how on his view the higher good must at times override the lower for the principle of skill takes into account only the one desire and the one end the principle of prudence takes into account all the desires and all the ends of one agent ; the principle of morality, whatever else it does, takes into account the desires and ends of all agents, so far at least as they may be affected by the agent's action. Views of this kind find little favour today, and we must recognise that even to those who do not reject reason in favour of sense or emotion Kant's whole attempt to define good in terms of practical reason and its principles seems a perverse effort to put the cart before the horse ; and they take the plausible view that the good is not good because it is, or would be, willed by a rational agent, but that on the contrary it is, or would be, willed by a rational agent because it is good. These are difficult matters, and for the appreciation of Kant's philosophy it may be sufficient to recognise there is a necessary and reciprocal connexion between goodness and a will, such that a rational
Whatever may be thought

is
goodness, there
had full control over the passions would necessarily choose good, and what is good would necessarily be chosen by a rational will. 2 But such a view may be interpreted in various ways. First of all there is the possibility that the good may be defined in terms of a rational will and its objective principles, and this I take to be the view of Kant, though he may not have been wholly clear about it. This view cannot, I think, be accused of what Professor Moore calls the 'naturalistic fallacy' at least it is not confusing the good with a natural 3 object, for an objective principle of practical reason is not a natural object. Furthermore, goodness in the subordinate senses is not arbitrary, but depends on the intrinsic properties of things (including the desires
will
if it
what
is
Compare Chapter
II
and
4.
omit the qualifications necessary, many of which are well brought out in The Philosophy ofG. E. Moore, pp. 608-11 and 615 ff. 8 See G. E. Moore, Principia Ethica, p. 13,
2 I
X6]
THEGOOD
III
and needs of the agent). The main difficulties arise in regard to the moral and unconditioned good, if we take this to mean that such goodness has absolutely no reference to the intrinsic properties of things (including the desires and needs of rational agents). This I do not believe to be Kant's view, because, in spite of his insistence on the formal character of goodness and of the moral principle, he is well aware that every form must have a matter. 1 Nevertheless it may be thought by some that on this view goodness disappears altogether and is swallowed up in a so-called rational will which is merely arbitrary, if it is not completely empty.
A second possibility is that a rational will may

2
goodness.
This seems to
me
be defined in terms of inadequate as an account of a rational will,
and it is open to the same objections as the third possibility which we have to .consider. The third possibility is that a rational will and goodness can neither of them be defined in terms of the other, but that we can recognise a necessary and reciprocal connexion between them. This view has a great attractiveness and attempts, so to speak, to do justice to both sides of the equation. But can we understand the two factors separately, and above all can we
make
if we cannot understand the principle of our own activity, how can we understand anything? In the case of goodness we must make up our mind as to the manner of its apprehension. If it is apprehended through sense or emotion, it is not easy to understand how our judgements about it can claim to be valid for any one but ourselves ; 3 and it seems impossible to understand how we can be entitled to say that all things or all actions of a certain kind must be good, or again that the good would necessarily be chosen by a rational will. If on the other hand it is apprehended through a concept not derived from sense or emotion, how can we acquire such an a priori concept except from knowledge of the necessary working of our own minds or our own wills? If we set this question aside as improper, can we attach any meaning to a concept of some real quality which is not known by sense or emotion, a real quality of which we can say only that it is what it is and it is nothing other than it is? Above all, how in these circumstances can we be entitled to say that all things of a certain kind must have this quality, or again that a rational will, qua rational, must have this quality? On this view these judgements must be synthetic a priori judgements, not 4 analytic ones ; and it is difficult to accept such judgements as true unless we can explain how such acceptance can be justified. There may be a fourth possibility that a rational will and goodness may be definable only in relation to one another, like 'right' and 'left', or
:
them? We can understand a because we have, and indeed in a sense are, a rational will
rocal connexion between
intelligible the possibility
of understanding a necessary and reciprational will
66 and 454 88. See for example 6?r., 436 view akin to this (substituting 'right' for 'goodness') seems to be suggested in The Philosophy of G. E. Moore, p. 616. But perhaps his view is more like that of the third possibility mentioned below.
1
This distinction
is
discussed briefly in Chapter
XII
2-5.
I J
2
5
[X
'above and 'below'. This view is clearly worthy of exploration, and it is alien to Kant than the second and third views ; but even this view he would probably consider a mistake since it treats goodness as co-ordinate with a rational will instead of subordinate to it and derived from it. I have no wish to be dogmatic on this problem, the discussion of which has been greatly developed since Kant's time. But we can say at least that his view does not make goodness unreal or arbitrary, nor does it make goodness depend on the momentary attitude of the agent. What difficulties there are, arise, as I have said, chiefly in regard to the moral good or the absolute good. This is the most important sense of the word, and we are not really in a position to estimate Kant's doctrine until we have more knowledge of the way it works out in detail.
less
CHAPTER
XI
IMPERATIVES
i.
Imperatives in general.
KANT'S account of imperatives can be seen in its proper perspective only when we have grasped his theory of the subjective and objective principles
of practical reason and of their relation to different kinds of goodness. Otherwise imperatives will bulk far too large in our view of his philosophy, which will then become so distorted as to appear harsh and
forbidding.
We have seen that the objective principles of practical reason need not also be subjective principles that is, they need not be the maxims on which we act. This painful paradox of our practical experience arises from the fact that our inclinations may be 'obstacles and hindrances' to practical reason. This happens in regard to the objective principles of skill or selflove as well as in regard to the objective principles of morality. Many men are led by passion to act in a way which they know will interfere with the realisation of a desired end and will even be detrimental to their happiness. They are still more obviously led by passion to act in a way which they know to be morally bad. It is a plain fact that men do not always do what would necessarily be done by a rational agent if reason had full control over passion. This fact makes all the objective principles of practical reason, which are always principles of some kind of goodness, appear to us as principles of obligation, and so as commands or imperatives. Instead of being the necessary laws of our inner rational nature, they seem to be something almost external or alien to us, something which constrains or compels or
our only partially rational will. very important in this connexion to distinguish between 'necessity' and 'necessitation'. 1 completely good or holy will would necessarily, although spontaneously, manifest itself in good actions. An imperfectly good will may, because of passion, feel reluctance and difficulty in following an objective principle of goodness, and the principle then seems to exercise pressure or constraint almost against our will. The principle, though recognised as objectively necessary, is not subjectively necessary; and the good action, even if we do it, seems to be, not necessary, but
'necessitates'
It is
necessitated.
36. The German for 'necessitation* is 'Ntitigung', and the Latin 'neces413 This term, as unfortunately are too many others, is obscured in the translation of Abbott, who renders it as 'obligation*.
1
Gr.,
sitatio*.
["3]
114

1
[XI
This seemingly almost external necessitation is brought out by saying that an objective principle appears as a command or imperative. No doubt this comparison is inexact (as comparisons always are) ; for a command is wholly external, whereas the objective principles of practical reason are manifestations of our own rational nature. No one has insisted on this more than Kant, who always rejects the view that morality can be mere obedience to the commands of the State or even of God. Nevertheless the affinity between obligation and command is shown by the common tendency to express moral and other obligations by using the
imperative mood.
2.
Definition
of an
imperative.
:
is
Kant's definition of an imperative is as follows The conception of an objective principle , so far as it is necessitating for a will, called a command (of reason)) and the formula of the command is called an
'
92
imperative.
This definition should now offer little or no difficulty, but certain may be noted. (1) Kant seems to make no use of the distinction between a command and an imperative, and it can therefore be ignored. 8 (2) Where a practical reason unhindered by desire would say simply *I will', a practical reason hindered by desire has to say l ought'. (3) The principles of goodness thus appear in our finite human condition as principles of obligation. This is true even where the principle in question is one of skill or rational self-love and not of morality. Men are not wholly rational in the pursuit of happiness or even in the adoption of means to ends.
points
e
3.
Three kinds of imperative.
have recognised three kinds of objective principle in action and three corresponding kinds (or senses) of good ( i ) the useful, or the good as means, (2) the good for me, or my good, and (3) the morally good. For an agent whose will does not necessarily conform to these objective principles, the three kinds of principle must appear as three kinds of imperative; and corresponding to the three kinds of good action there 4 will be three kinds of action in some sense obligatory.
:
We
There is a passage in Mr. H. G. Wells's Mr. Polly which describes very well the seemingly external character of moral necessitation. The necessitation, as Kant says, is less marked in imperatives of skill or self-love; see Gr. t 416 41. 1 38: 'Imperatives are only formulae for Gr., 413 36. Compare also Gr. 9 414 expressing the relation of objective laws of willing in general to the subjective imperfection of the will of this or that rational being, e.g., the human will/ The relation in question is that of necessitation or constraint. * We may, if we choose, consider the imperative as the philosophical formula for what appears to the agent as a command.
1
.,
414 ff.
XI
4]
IMPERATIVES
115
the objective principle of practical reason is conditioned by an * end, the imperative is hypothetical. It takes the form Every rational agent, to he wills a certain will the action end, ought if good as a means to this end.' The command of reason is here conditioned by the end; and as the ends vary, the action enjoined by reason will also vary. Kant recognises two forms of hypothetical imperative. If the end is merely what one might will, the imperative is a problematic imperative, an imperative of skill. Where the end is what everyone naturally wills, namely happiness, the imperative is an assertoric imperative, an imperative of prudence or rational self-love, called also a 'pragmatic' imperative. Where the objective principle of practical reason is not conditioned by any end, the action is enjoined for its own sake, as good in itself without reference to any further end. The imperative is then categorical: that is to say, it is not conditioned by the hypothesis that some particular end is desired. It takes the general form 'Every rational agent ought to will the action good in itself.' 1 This imperative Kant calls an apodeictic im2 perative, and it is the imperative of morality. Kant's terminology is not wholly satisfactory, and it is not unreasonable to say that he may have been unduly influenced by a supposed parallel
Where
with problematic, assertoric, and apodeictic judgements. Elsewhere 8 he himself suggests that a problematic imperative is a contradiction in
is 'technical imperative' or 'imof The skill'. perative pragmatic imperative of self-love, he adds, is also a kind of technical imperative, but it requires a special name. The two reasons he gives for this are (i) the old reason namely, that our own happiness cannot be accounted as a merely arbitrary end like the ends of skill ; and (2) a new reason namely, that self-love is concerned, not merely with the method of attaining an end already presupposed, but also with determining what constitutes the end itself.
terms,
and that the proper expression
4. Rules, counsels,
and laws.
The different kinds of imperative exercise different kinds of necessitation or constraint. 4 Kant marks this difference by his terminology. He 9 opposes 'the rules of skill' and 'the counsels of prudence to 'the commands (or
9
may be clear and definite, but we are bound by so far as we wish to attain a particular end. The counsels of prudence or self-love are in some ways more uncertain, yet they are also more binding. They are more uncertain because different counsels may hold for different individuals it is difficult or even impossible to be sure wherein a particular individual will find his happiness, and this will depend
them only
:
laws) of morality . The rules of skill
1 We must not forget that an action which is good in itself is one which is done for the sake of universal law as such. Compare Chapter VII 3. 8 Note also that Kant sometimes opposes the word 'practical* in the sense of 'moral* to 'pragmatic* in the sense of 'prudent*.
K.U., Erste Einleitwg,
I,
8 n.
Gr. y
416
41.
Il6

believes to
[XI
partly on what the individual
be necessary
for his happiness.
Kant tends to be pessimistic about the chances of attaining happiness by nicely calculated conduct. Indeed temperamentally he seems almost to dislike the calculations of prudence when compared with the spontaneity of inclination. Yet in spite of their uncertainty the counsels of prudence are more binding than the rules of skill, since it is mere folly to wreck one's happiness, an end which is very far from being arbitrarily
chosen.
A holy will would act necesand spontaneously in accordance with moral law, though in this there would be no consciousness of obligation. To imperfectly rational agents the moral law appears as an obligation or command because of their imperfection and this obligation is still unconditioned the obligation in no way depends on the fact that we happen to seek a particular end or even on the fact that by acting in accordance with the law we are likely to attain happiness. On the contrary, the law will still hold even if it is opposed to our inclination and even if it is detrimental
tion) valid for every rational agent as such.
sarily
;
:
Both rules of skill and counsels of prudence are opposed to the commands of morality, which alone are strictly entitled to be called 'laws*. 1 The word 'command' serves to show that moral obligation does not depend on our inclinations and may even be opposed to them there is no question here of cajolery or persuasion, The word 'law', it must be 2 remembered, does not necessarily carry with it the idea of 'command' it expresses here, as Kant says, an unconditioned necessity (not necersita:
to our happiness.
To deny this is to deny morality altogether. If we consider that morality as ordinarily understood is an illusion, by all means let us say so; but at least let us avoid the muddled thinking which retains the ordinary terms of morality, but interprets them as referring only to selfinterest or even to mere passing and personal emotions.
5.
Obligation and goodness.
Kant's three kinds of imperative enjoin three kinds (or senses) of
good action. For him obligation is inseparable from some kind of goodness, though there may be goodness where there is no obligation, as in the case of a holy will. Thus goodness is more fundamental than obligation, which arises only because of our human imperfection. Apart from some kind of goodness there is no kind of obligation. This general principle holds in the case of moral obligation. The moral imperative enjoins moral goodness it bids us to act morally that is, as we have seen, to act for the sake of the law or for the sake of duty.
:
1 Gr., 416 41. Earlier in Or. t 413 =*= 36 Kant speaks of all imperatives both as laws and as commands. This may be defended so far as the general principles of skill and self-love (though not their application) are unconditioned. See Chapter IX 5. 1 See Chapter VI 2, Gr., 413 3$. 37 ; 414
XI
The
63
IMPERATIVES
command
is
117
universal ethical
'Act in accordance with duty for the
sake of duty'. 1 If we take the universal moral imperative to enjoin action for the sake of duty, we must recognise that as it is applied to particular situations, it enjoins also some particular kind of action, such as the paying of debts where money is owed. How such application is possible, we are not here concerned to ask. Granted the possibility of such application, Kant insists that the moral imperative bids us pay our debts for the sake of duty; that is, it enjoins not merely a kind of action, but the doing of this kind of action from the moral motive of duty. Indeed on his view this is what differentiates mere law that is, State law from
morality. If we
motive,
do an action of the kind enjoined but do it without the moral call it an action done in accordance with duty 2 (pflichtmassig). We have no precise adjective for this in English, but we it call the action for sake of 'due'. When the the action is done might
we can
describe it not merely as 'due', but also as 'dutiful' (pflichtvoll) ; so is it a morally good action. Thus on Kant's view the moral imperative enjoins actions which are not only due, but also
duty
we
and only
dutiful.
merely a matter of verbal convenience whether or not we say has done the action which is due has done his duty. He has done his duty in one sense, even if the moral motive is not present; but in another sense he has not done his duty, for his action is not dutiful and so has not fulfilled the moral imperative. We come, however, to a profound philosophical difference when the view that we ought to act for the sake of duty is explicitly rejected. 8 This doctrine is so strongly maintained today that it can hardly be passed
It is
that a
man who
over in silence.
6.
The duty
to act
morally.
It is impossible to deal adequately with this question here, but there are two main objections to Kant's view. The first objection is that since we cannot summon up motives at will, it cannot be our duty to act on them. On this view motives are regarded as feelings, and it is true that we cannot summon up feelings
M.d.S., Tugendlehre, Einl. VII, 391 = 236: 'Handle pflichtmassig aus Pflicht'. Kant even accepts the commands 'Be holy' and 'Be perfect', though he recognises with his usual common sense that this is an ideal to which we can only approximate progres21 and 22, sively: the duty is one of imperfect obligation. See M.d.S., Tugendlehre, 446-7. 1 The word 'right' is sometimes used for this, but (i) the original meaning of 'right* seems to be 'fitting* rather than 'obligatory* (2) in its ordinary usage it means 'permissible rather than 'obligatory'; and (3) it'is not without the suggestion of some kind of value, which ought here to be excluded. 3 See Sir David Ross, The Right and the Good, pp. 4-6.
1
;
5
Il8
fXI 6
of benevolence or affection, which are sometimes regarded as moral motives. Kant himself (so far as he takes a motive to be a feeling) regards the feeling of reverence as the only moral motive in man, but for him it is the necessary emotional accompaniment or consequence of my recognition of duty. Hence it does not need to be 'summoned up' ; if it were absent, I should recognise no duties, and I should be neither moral nor immoral, a mere animal and not a man. As he himself says of moral feeling, 'no man is wholly destitute of moral feeling; for if he were totally unsus-
would be morally dead'. 1 There is, however, a more subtle objection. We are told that to do act A from a sense of duty is to do act A from a sense that it is our duty to do act A. The moral imperative therefore, if it bids me to do act A from a sense of duty, asserts that it is my duty to do act A from the sense that it is my duty to do act A. But in this there is a contradiction. The* contradiction lies in the fact that while the whole assertion affirms that it is my duty to do act A from a certain motive, the final clause affirms that it is my duty to do act A simply, that is, altogether apart from any
ceptible to this sensation, he
motive. If I try to amend the final clause so as to make it harmonise with the whole assertion, I merely repeat the contradiction in a different form equally requiring amendment, and so fall into an infinite regress. It becomes my duty to do act from a sense that it is my duty to do act and so on from a sense from a sense that it is my duty to do act
A
5
ad
infinitum.
a Kantian point of view the use of the phrase 'sense of duty is unfortunate, not only because he firmly denies that there is any such 2 sense, but also because it suggests, not merely a feeling, but a judgement; and since in morally good action we must have judged act A to be our duty, it is easy to suppose from this that the moral motive must be concerned with this particular duty. But surely while our intention is to do the particular duty which we have called act A, our motive is to do our duty as such. To say that we ought to do act A for the sake of duty is a completely different thing from saying that we ought to do act A for the sake of this 3 duty. To act for the sake of this duty would seem to indicate a personal preference among duties which is distinct from, if not opposed to, the motive of morality. It is in any case clear that the objection based on an alleged infinite regress assumes that we cannot act for the sake of duty as such. This 4 assumption seems to me erroneous, and it results in the paradox that we have no duty to act morally. It would be hard to believe this even if the
M.d.S. t Tugendlehre> Einl. Xlla, 400 247. If we take the moral motive to be the or duty itselfsee Chapter V, Appendix it is still more clear that we can always act on the motive of duty we have only to act on the principle of doing nothing which we recognise to be incompatible with moral law.
1
From
law
itself
M.d.S., Tugendlehre, Einl Xlla, 400 247. Similarly to say that we ought to kill this German soldier because he is a German soldier is a completely different thing from saying that we ought to kill him because he is
8
this
4 If
German soldier. we cannot act

is?
for the sake of duty as such,
why
should
we even
ask
what our
duty
XI
6]
IMPERATIVES
Iig
objections were stronger than they are. At any rate, if we are to understand Kant, we must recognise that for him the one supreme duty is to act for the sake of law as such or duty as such, not for the sake of a particular law or a particular duty. In other words our supreme duty is to act morally, and if this were not so, we should have no particular duties
at
all.
CHAPTER
XII
HOW ARE IMPERATIVES POSSIBLE?

i.
The meaning of the
question.
THE next question to be considered is 'How are the different kinds of 1 imperative possible?' This question should not be misunderstood, as it sometimes is. The question is not a psychological one it is not concerned with the way in which consciousness of obligation gradually develops in the individual or in the race; it is not even concerned with the way in
:
which imperatives give
rise to
emotion and so
affect action,
though Kant
:
elsewhere discusses at some length the way in which reverence is aroused 2 by the categorical imperative. Nor is the question a metaphysical one it does not ask how reason, and especially pure reason, can be the cause of events in the phenomenal world of experience. Such a question Kant expressly declares to be unanswerable, at least as regards the categorical
imperative. The question may be described as logical or epistemological, though even these terms are not wholly suitable. Imperatives are propositions, although they are practical and not theoretical. Our task is to show how these propositions can be justified or how they can be valid. The question and its answer find their parallels in the Critique of Pure Reason, where Kant attempts to answer the questions 'How is pure mathematics possible?',
3
'How is pure physics possible?', and 'How is metaphysics possible?'. In order to understand Kant's answer we must first get clear about
the logical difference between analytic and synthetic propositions, and especially about the logical character of what Kant calls 'synthetic
a priori propositions'. 4
2.
Analytic propositions.
In an analytic proposition the predicate is contained in the subjectconcept and can be derived by analysis of the subject-concept.
1
Gr., 41 7
Gr. t
is
4589
how it
4
perative : of the will which
41, that the question is not possible to conceive the accomplishment of the action enjoined by the imthe question is how it is possible to conceive that necessitation (or obligation)
98.
= 41 = 94; 461 =
.
See Chapter
He also says,
in Gr., 41 7
2.
is
Kant in using the terms 'proposition' as equivalent. Some philosophers deny that there are such things as judgements, while others deny that there are such things as propositions; but Kant is dealing with a real problem which may be expressed in other terms but cannot be ignored save at our peril.
See K.r.V., A6ff.
expressed in the imperative. Bioff. I here follow
and 'judgement'
[120]
XII
2]
HOW ARE IMPERATIVES
POSSIBLE?
121
An example of an analytic proposition is 'All bodies are extended.' 'Being extended' not merely characterises all bodies, but is contained in the concept of body as such. You cannot conceive a body except as extended. It is all-important in this connexion to guard against a too common that is, the concept or notion of the subject error. The subject-concept should on no account be confused with the subject, which is usually not a concept but a thing or class of things. Manifestly in every true proposition (whether analytic or synthetic) the predicate must be contained in the subject in the sense that it must characterise the subject. It by no means follows from this that in every true proposition the predicate must be contained in the subject-concept. The concept, as universal, contains in itself much less than do the particulars which fall under it. Thus when I truly say 'This body is made of gold', 'being made of gold' may not too happily be said to inhere in this body. But it is ridiculous to suppose that 'being made of gold' is contained in the concept of 'body', which is here on Kant's view the subject-concept, as indeed it must be if concepts are always general. In spite of this obvious distinction the subject-concept is often confused with the subject, not only by beginners, but also by masters, in philosophy (including interpreters of Kant). Indeed according to Kant the whole intellectual system of Leibniz rests on this elementary error. 1 All analytic propositions are necessary and universal and therefore a priori. This is true even where the subject-concept is an empirical concept, e.g., the concept of 'body'. To know what a body is, is to know that it is extended, and there is no need of any further appeal to experience. Nor, granted that we possess complex concepts, is there any difficulty in explaining how analytic propositions are possible. We may of course be challenged to show that there are real objects
we have used in a parproposition. An analytic proposition, though this is not always recognised, professes to be a statement about objects, and not merely about a concept, as is obvious in the example I have given. Kant was the first to make the distinction between analytic and
corresponding to any particular concept which
ticular analytic
synthetic propositions, and if we are to understand him, we must take the distinction as he made it. The view sometimes held that analytic propositions are about concepts or even about the meaning of words is a statement that a concept has certain parts distortion of his doctrine.
analytic than a statement that a fiddle has is that a possible subject-concept has statement that a word is, or will be, used been turned into a subject. by me or by others to mean a particular kind of thing, or as an equivalent for certain other words, is quite obviously not an analytic statement in
would be
for
him no more
all
certain parts
that has
happened
See K.r.V., Aa8i
8337. This
is
lectures
is
which
hope will be published.
On Leibniz* theory to say that the predicate is in
well brought out
by the
late
Mr. Joseph
in
the subject of a true proposition (praedicatum messe subjecto verae propositions) is to say that it contained in some way in the notion of the subject. It is easy to see how this must result in a theory of monads.
122
[XII 3
his sense at all; and indeed such a usage seems extraordinarily farfetched. It is of course open to his successors to say that the propositions which he dubbed analytic in a precise sense have in fact a quite different
they continue to apply to them the term 'analytic , can lead to nothing but confusion, so far at least as the interpretation of Kant is concerned. It should be added that analytic propositions need not be confined to those which assert a relation of subject and predicate. As we shall see, hypothetical imperatives are for Kant analytic propositions, although they are most naturally expressed by an 'if clause. In order, however, to bring out their analytic character he translates them into the subjectpredicate form.
character; but
if
presumably on
historical grounds, this
'
3. Synthetic propositions.
In a synthetic proposition the predicate is not contained in the subjectconcept and cannot be derived by analysis of the subject- concept. It may be, and it often is, derived by an analysis of the subject that is, of a thing or things conceived in the subject-concept and experienced by us
through the senses. An example of a synthetic proposition is 'All bodies are heavy.' 'Being heavy is on Kant's view no part of the concept of body. There is no contradiction in supposing that an astral body so-called or a resurrected body should be without weight. Most synthetic propositions are empirical. 1 They are known through
5
experience of the thing conceived in the subject-concept. We are aware that bodies are heavy because we have experience of bodies. And if we do not enquire too closely into the nature of experience, there is no difficulty in understanding how such synthetic propositions are possible.
4. Synthetic
a priori
propositions.
universal
5
synthetic propositions, according to Kant, are necessary and and are therefore a priori that is, they cannot be derived from 2 experience. For example, the proposition 'Every event must have a cause is not empirical and it is also not analytic, for the concept 'event' does not contain in itself the concept 'being caused As it stands, it is a
;
5 .
Some
synthetic a priori proposition.

is
The most difficult, and "also the most important, task of philosophy how synthetic a priori propositions are possible that is, how they can be valid or how they can be justified. Kant holds, not only that
to explain
1
Mr. Ayer,
in Language, Truth and Logic

facts
(first
edition), p. 103, avoids difficulties
by
defining synthetic propositions as empirical.

is
determined by the
8
*A proposition ... is synthetic when its validity of experience.' But this begs important philosophical
questions.
See Chapter I
3.
XII
5]
HOW ARE IMPERATIVES
POSSIBLE?
123
certain synthetic a priori propositions to be valid, but also that if no such propositions were valid, we could understand neither how experience is possible nor how empirical synthetic propositions are
possible.
we know
5. Difficulties.
That there are many difficulties in Kant's distinctions should not be ignored particularly the difficulty of determining precisely what is contained in a subject-concept. 1 Modern logicians may reject some or all of Kant's examples and may seek to state his problem in other terms.
Nevertheless these distinctions are far from dead in modern philosophy. 2 Purely logical difficulties are beyond the scope of this discussion, but some readers may feel a special difficulty in understanding the difference between an analytic proposition and a synthetic a priori proposition ; for in both cases there is asserted a necessary connexion between the subject and the predicate. The difference is roughly that in the case of an analytic proposition you cannot conceive the subject without also conceiving the predicate, however obscurely ; while in the case of a synthetic a priori proposition you can conceive the subject without conceiving the predicate. This may be grasped most easily from an example. In the analytic proposition 'Every effect must have a cause', you cannot conceive an effect without conceiving that it has a cause; for an effect is simply an event that has a cause. In the synthetic a priori proposition 'Every event must have a cause you can conceive an event without conceiving that it has a cause ; for an event is merely a changing of something in time or a coming to be of something in time, and whether or not this must necessarily have a cause is a matter for further consideration. In the case of analytic propositions a necessary connexion can be established by a mere analysis of the subject-concept by itself. In the case of synthetic
5
some further evidence, some 'third term', between the subject and the predicate.
a priori propositions, as indeed in all synthetic propositions, we require to establish the connexion
6.
Imperatives of skill are analytic propositions.
We can now return to the question 'How are the different kinds of imperative possible?' It must be admitted that Kant's treatment of hypothetical imperatives is over-simplified and would be much too summary for a philosophy of action. All he is doing is to clear them out of the way and so get on to the moral problem. According to Kant the possibility of imperatives of skill requires no
1 8
Compare K.M.E.,
I,
82
ff.
very similar problems arouse the interest of Professor G. E. Moore and are sometimes expressed by him in very similar terms. See The PhiloofG.E. Moore, pp. 660 ff., especially pp. 663 and 667.
It is instructive to see
how
124
[XII 6
1 special explanation, since they are all analytic propositions. He seems to assume that the end is always an object or result, that there is only
to the end, and that this means is always some possible action of the agent. We could then say that in the concept of willing an end (as opposed to merely desiring it or wishing it) there is contained the concept of willing the means to the end. Hence the proposition 'to will the end is to will the means' is a theoretical analytic
one means
proposition.
Kant himself
is
sketchy about the
way
in
which
this
theoretical
analytic proposition becomes a hypothetical imperative and so a practical must try to fill up the gaps. First of all we have analytic proposition. the objective principle of practical reason, 'Any rational agent who wills the end will necessarily so far as reason has a decisive influence over his actions will the means which are in his power.' This proposition, which is still analytic, appears as an imperative to us because reason, though present in us, has no such decisive influence. It then takes the form 'If any rational agent wills the end, he ought to will the means.' All
We
this must be taken within its own limits, wider considerations of prudence or morality being excluded. It may be objected to this slightly tidier version of Kant's argument that there is in it a contradiction. If to will the end is to will the means, how can it be said that a rational agent ought to will the means? Does not the latter statement imply that in fact it is possible to will the end and yet not to will the means? Perhaps we should meet this by speaking, as Kant himself does almost immediately afterwards, of willing the end 'completely' and not just of willing the end. Perhaps, however, it is better to keep the statement as it stands ; for it at least serves to bring out what Kant calls, in another connexion, the 'contradiction' in imperfectly rational wills, a contradiction which is analogous to a contradiction in thought but is better called an 'antagonism'. It arises from the fact that, as we have seen, a principle which is objectively necessary is not thereby also sub3 jectively necessary. And it is surely only too true to experience that we can both will and yet not will an end, e.g., when our attention is distracted in a greater or less degree. The assumption of all this appears to be that we actually possess a rational will, but that in action this rational will may be opposed by
what Kant calls 'inclination'. Perhaps it would be better to say, at least Where the inclination prevails, that a rational volition is overcome by a
less rational volition. It is the possibility of such defeat that gives rise to imperatives, and unless both types of volition were present in us, or at least possible in us, there would be no imperatives. Into the difficulties of this we cannot enter. Kant's fundamental
l
Gr.,
417
= 42.
;
Kant himself does not begin with this proposition but a more
same
effect
is
to the
given subsequently as the reason
why
elaborate proposition the objective principle of
practical reason is analytic. 8 See Gr. 9 424 =51. Compare
IX
i.
7]
HOW ARE IMPERATIVES

is
POSSIBLE?
like
125
assumption clearly
contradict
act, as
itself.
1
A rational agent, so far as he is rational,
that practical reason,
theoretical, cannot will necessarily

if he is
irrational
he will necessarily think, coherently; and he ought so to act enough to be tempted to do otherwise.
7.
Synthetic propositions are presupposed.
We have so far considered only the general principle of imperatives of skill. Even this, it should be observed, already presupposes knowledge of synthetic propositions; it presupposes knowledge that events have causes and that we may have power to cause events for to will the means to an end is to will the cause of a desired event. These synthetic propositions are theoretical ; and many more synthetic theoretical propositions are needed if we are to apply the general principle to particular cases. 2 They are needed, as Kant says, to determine the means to a proposed end. Take, for example, the particular imperative 'If I want to see properly, I ought to wear glasses.' No such imperative could be formulated without knowledge that the wearing of glasses will counteract my defects of sight; and to know this is to apprehend a synthetic proposition. Nevertheless, Kant insists, the imperative itself remains an analytic proposition so far
;
as willing
I
is
concerned.
mean something like this. All sorts of synthetic propositions are required if we are to know the possible causes of events that we desire. These propositions are purely theoretical. But once we have this theoretical knowledge, the principle governing rational action, and so the principle of hypothetical imperatives of skill, is simply the proposition that to will the end so far as reason has decisive influence on action is to will the means. This is an analytic proposition ; and so far as action is concerned it remains so, no matter how many synthetic propositions are necessary in order to specify what are the particular means to a particular end
take him to
3 proposed in a particular situation. In the Critique of Judgement Kant speaks of all technical rules as being mere corollaries or consequences of purely theoretical knowledge. They differ from theoretical knowledge only in their formulation, not in their content, 4 It is not clear how far he regards this view as modifying the view of the Groundwork; but we have still to explain the difference in formulation, and it is hard to see how this could be done without bringing in the analytic proposition that to will the end is to will the means.
1 That is, it cannot do so 'objectively* or qua reason. Subjectively that is, as influenced by feeling and desire it can contradict itself or at least oppose itself. *Gr., 417 43. 8 Though the necessary means are not contained in the concept of an end, willing whatever means are necessary is contained in the concept of willing an end. 4 /T./., Erste Einl. I, 4; also EM, I, 172 ( xni).
126

8.
[XII 8
Imperatives of prudence are analytic propositions.
For imperatives of prudence or self-love the same explanation holds as for principles of skill. 1 The only new difficulty is that here our concept of the end is so vague and imprecise. Everybody wants happiness, but no one knows, or can know, exactly what it is that he wants. Neverwhen, so far as we can, we have determined, in synthetic theoretical means to our own happiness, we can infer the corresponding imperatives of prudence or self-love on the supposition that practical reason, like theoretical, is necessarily coherent with itself. The
theless
propositions, the
general principle of the imperatives of prudence is the same as the principle of the imperatives of skill, namely, that a rational agent who wills the end must, so far as he is rational, necessarily will the means, and ought to do so if he is irrational enough to be tempted to do otherwise. This
practical principle remains analytic however many synthetic propositions may be required in order to determine theoretically the means to our
happiness.
This sounds simple enough, but even if our end were happiness considered as the maximum possible amount of pleasant feeling, there would still be a new problem the problem of combining the various means in an organised life. This problem did not arise in our oversimplified account of the imperative of skill. Furthermore, although Kant persists in regarding the problem merely as one of means and ends, his language, as we have seen, betrays another view, which becomes the view that the problem is fully explicit only in his later writings concerned also with the constituents of the end itself. What other ground could there be for saying that the concept of the end is vague and indeterminate' ? Manifestly the imperatives of prudence must be grounded on another principle, if we are concerned with a choice of ends as well as of means.
Perhaps we might say that any rational agent who wills a total end will necessarily so far as reason has a decisive influence over his actions will the constituents of that end as well as the means to it. This proshould still have to consider how the position would be merely analytic.
We
of determining the total end may the problem of satisfying the maximum possible of the agent's desires or of realising the maximum possible of the agent's powers. It remains theoretical, even if it can be solved, however imperfectly, only by experiment, and even perhaps if it must take into account as an important factor choices that are to a large extent arbitrary. There is, however, a further practical assumption that since man, as belonging to the sensible world, is a being with needs, reason has so far 'an office which it cannot refuse', the office of serving the interests of sensibility and of seeking happiness in this world and where 2 possible in the next. Kant's imperatives of self-love are not derived merely from the fact that all men happen to seek happiness. They are derived
total
end
is
determined.
The problem
be regarded as a theoretical problem
Gr.,
417-19
42-4.
/r..F., 61
152
108).
XII
g]
HOW ARE IMPERATIVES
POSSIBLE?
12?
from the assumption that this is what a finite rational agent, qua rational, would necessarily do. Only so can the principle be an unconditioned principle (except in so far as it may be overridden by a higher) and only so can happiness on Kant's theory be an objective and genuine good. It is, however, by no means obvious that this principle is an analytic
;
proposition.
waverings and even confusions in regard to this topic happiness is an uncertain Ideal of the imagination, the ideal of a whole of satisfaction which can never be determined by reason with precision. He is very far from the crudely 'scientific principles of hedonists and utilitarians. He is also sound in insisting that for this reason prudence can give us only counsels and not commands, mere rough generalisations about the average ways in which men can be happy. Incidentally it is interesting to observe that he personally thinks, as an oldish man, that happiness is most likely to be found in diet, frugality, politeness, and reserve This rather pathetic conclusion suggests that he is perhaps sound in thinking reason to be but a poor instrument for the attaining of happiness as our only end.
In spite of
is
his
Kant
sound in
his insistence that
9.
Categorical imperatives are synthetic
a priori
propositions.
The main point of Kant's discussion has been to provide a sharp contrast with the categorical imperative. Categorical imperatives do not bid us will the means to an end, and so are not conditioned by will for an end already presupposed this is why we say that they are unconditioned,
:
categorical imperative says simply and categorical. unconditionally 'Every rational agent ought to will thus and thus.'
absolute,
and
Hence
the explanation given of hypothetical imperatives can in no

it.
way
apply to
The very concept of a categorical imperative might appear fantastic were we not acquainted with the seemingly unconditioned claims of morality. Kant, however, insists that we can in no way establish the categorical imperative by an appeal to experience. He knows too well that seemingly categorical imperatives may conceal a motive of personal interest. So-called moral action may have some secret self-interest as its basis, as many philosophers maintain today. We have the difficult task of
establishing the possibility of a categorical imperative, not merely the task of explaining a possibility which we already take to be established. 2 Even the latter task, needless to say, could not be completed by an appeal to experience, since experience cannot tell us what ought to be,
categorical imperative is not an empirical, but an priori, practical proposition. The special difficulty of dealing with the categorical imperative arises from the fact that although it is an a priori proposition, it is not analytic.
but only what
is.
1
2
Gr.,
Gr. 9
419 420
is
= 44.
is
perative question
about the categorical im45. Thus in the Groundwork our question more akin to the question 'How is metaphysics possible?' than it is to the
*How
mathematics possible?'
JSr8
fXII 9
By merely analysing the concept of 'rational wilP or 'rational agent' we 1 cannot, Kant always insists, arrive at the obligation to will in a certain that is we The connect the obligation directly or immediately way. paradox with the Concept of a rational will as such (that is to say, we do not connect it mediately through the supposition that some end is already sought) ; and yet this obligation is not contained in the concept of 'rational 2 3 will'. categorical imperative is a practical synthetic a priori proposition, and to explain how synthetic a priori propositions are possible is always an undertaking of the greatest difficulty. As in all synthetic a priori propositions we shall require some 'third term' to establish a necessary connexion between a subject and a predicate which is not contained in the concept of the subject. This 'third term' we shall find to be the Idea of
freedom. This problem must be reserved till later. For the present we must be content to determine the character of the categorical imperative more
precisely
and
to
examine
its
various formulations.
Compare Compare
Gr. 9
Gr. t
Gr.,
426 = 54 and 440 =71. 420 n. = 46 n. For a closer adherence to Kant's actual Words
see Chapter VIII

8
420
3. = 45-6.
BOOK
III
THE FORMULATION OF THE CATEGORICAL IMPERATIVE

CHAPTER
XIII
THE FIVE FORMULAE

i
.
The Jive formulae.
We might have expected Kant to be content with one formulation of the categorical imperative. Instead he embarrassesûs with no less than five different formulae, though, curiously enough, he tends to speak as if there were only three. 1 If we are to see where we are going, it may be well to set forth all five formulae at the outset. For ease of reference I propose
number (or a number + a letter), and also a based on its key words. The system of numeration adopted is intended to bring out the special connexions between different formulae and to conform, as far as possible, to the view that there are three main formulae. Formula I or the Formula of Universal Law Act only on that maxim through which you can at the same time will that it
to give each formula a
title
:
should become a universal law. 2
UNIVERSAL
So
Formula la or the Formula of the Act as if the maxim ofyour action were
Law
to
of Nature
LAW OF NATURE*
II or the
become through your will a in Itself:
Formula
Formula of the End
act as to use humanity, both in your own person and in the person 41 other9 always at the same time as an end, never simply as a means.
of every
III or the Formula of Autonomy So act that your will can regard itself at the same time as making universal law through its maxim. 5 Formula Ilia or the Formula of the Kingdom of Ends So act as ifyou were always through your maxims a law-making member in a universal kingdom of ends*
:
:
Formula
= 66 and 437 = 67. = 47. 5 worth noting Gr., 434 = 63. It

*
Gr., 436 Gr., 42 1
Gr.,
Or.,
42 1 = 47429 == 57*
is
Formula III the phrase

Gr.,
438
= 69.
'at the
same
that alike in Formula I and Formula II time' cannot properly be ignored.
and
130

2.
[XIII
The
relations between the Jive formulae.
We can follow the relations between these formulae only as we study both them and the arguments by which Kant passes from one to another. It is, however, obvious that there is a close connexion between Formula I and Formula la, and again between Formula III and Formula Ilia, though Formula Ilia is also closely connected with Formula II. The system of numeration adopted, together with the classification which it implies, does not profess to be more than a convenience. It may also be observed there is at first sight a close resemblance between Formula I and Formula III, so much so that the difference between them is sometimes ignored. If to ignore this difference were not clearly contrary to Kant's intention, we might identify Formula III with Formula I and take Formula I as occupying a special position above all the others. We could then take Formula la, Formula II, and
Formula Ilia as subordinate formulae, intended to help us in the application of Formula I by bringing the general and supreme principle of morality (Formula I) nearer to intuition and so to feeling. There is at least some warrant in Kant himself for treating Formula la, II, and Ilia 1 along these lines, but we must not carry this so far as to ignore Formula III altogether. In the Critique of Practical Reason it is Formula III, and not Formula I, which takes pride of place. 2
3.
The purpose and

the face of
it
structure
of the argument.
five formulae is carried task of articulating the different aspects and implications of the categorical imperative he regards as appropriate to a 8 metaphysic of morals. Nevertheless we must not suppose that he is his the contrary, by leading us up of losing sight argument as a whole.
On
Kant's examination of the
on
for its
own
sake,
and the
On
to the principle of autonomy and the principle of a kingdom of ends he is preparing the ground for the argument in the last part of the book he is establishing principles which will be connected later with the
:
concept of freedom and the concept of an intelligible world. From this point of view Formula III and Formula Ilia are the most important. They constitute, as it were, the main hinge on which the argument of the Groundwork turns. It will be remembered that Kant regards his argument at the present stage as analytic or regressive that is, as an argument from the conditioned to its condition. 4 This must not, however, be taken too literally: it can hardly apply precisely to every step in the argument. Thus in moving
1 2
Gr, t
436-7
7,
K.p.V.,
same time be
8
( 54) valid as a principle
30
= 66-7. = 141 =
making
'So act that the maxim of your will can always at the universal law.'
See Chapter I 10 and Appendix. See Gr. t 392 = 9.
XIII
4]
THE FIVE FORMULAE
13!
from Formula I to Formula II we are moving from the form to the matter of moral action, while Formula Ilia combines both form and matter. 1 Such a movement cannot be merely from the conditioned to its condition. Furthermore, Kant is prepared to pass directly from the essence of a categorical imperative to Formula IP and also to Formula III, 8 4 Hence we cannot say that just as he does in the case of Formula I.
each successive formula stands to its predecessor in the relation of condition to conditioned. 5 What we can say is that throughout Kant is accepting the concept of duty as implied in our moral judgements and is attempting to analyse its conditions and implications.
4.
The
application
of the formulae.
of the present discussion is to formulate precisely the supreme of We are not endeavouring to justify this principle morality. principle all attempt at justification is reserved till later. Above all, we are not seeking to work out a system of morality or to show how the supreme principle articulates itself in a moral code. This task is expressly rejected 6 by Kant and belongs, not to the Groundwork, but to his later Metaphysic
:
The aim
of Morals. Nevertheless Kant has always in mind the actual concrete moral judgements of the good man, and he attempts to throw light on his supreme principle by giving a few examples of its application. This is a natural and reasonable thing to do but if we concentrate on this aspect of his exposition, we shall get a very misleading impression. His formulae might be completely sound, even if his method of application were totally fallacious. Moreover, even his method of application cannot be judged on the basis of a few examples given only for the sake of illustration. In actual fact his systematic application of the categorical imperative in the Metaphysics of Morals differs greatly from the sketchy illustrations given in the Groundwork. To reject Kant's principles merely on the ground that they do not give us some easy and infallible criterion for moral judgement is an absurd method of criticism and one which we should never dream of applying to any other philosopher. This type of criticism, it must be admitted, is encouraged by some of Kant's own expressions. 7 His great strength as a philosopher is his per;
and more general principles behind any accepted doctrine. His weakness is a tendency to pass too quickly from his higher principles to the empirical doctrines which fall under them witness, for example, his passage from general principles which underlie all
sistence in seeking higher
experience and
1
all
8 physics to the special doctrines of Newton. In spite of

*
Gr.,
Gr.,
8
7
436 = 66. 60. 431-2 See Chapter I 7-8. = 24-5. E.g., Gr., 403-4
Gr.,
*
fl
Gr.,
Gr.,
428-9 420-1
392
= 9.
= 56-7. = 46-7.
See, for example, K.r.V., 6225.
132
[XIII 4
showing a caution conspicuous feature
and common sense which is not always the most of German philosophers, and which a Scotsman may
be pardoned for attributing to his paternal ancestry, Kant exhibits this tendency to a too hurried application in some of his moral utterances. Nevertheless the value of his formulae is one thing, and the method of
their application
is
another.
CHAPTER XIV
THE FORMULA OF UNIVERSAL LAW

i.
Formula
I.
'Act only on that maxim through which you can at the same time will that it should become a universal law.* 1 This injunction is already familiar to us. found it, though in a negative form, as the result of analysing the implications of ordinary
We
moral judgements. 2 On a higher philosophical level we found it again though we did not make it fully explicit in analysing the objective principles and imperatives of practical reason, and in particular of pure 3 practical reason. In this analysis we were concerned only with the an unconditioned objective principle and what is a What is question we did not ask expressly 'What is it that a categorical imperative?' categorical imperative enjoins?' Nevertheless the answer to this last 4 question was implicit in our discussion for, as Kant tells us, in discovering what a categorical imperative is we have also discovered what it enjoins. An unconditioned objective principle is one which every rational agent irrespective of his particular desires for particular ends would necessarily obey if reason had complete control over his passions, and one which he ought to obey if he is irrational enough to be tempted to do
6
otherwise. The categorical imperative formulates the obligation or command to obey this unconditioned principle and a principle excluding
;
reference to particular ends can be only the form of a principle, or a formal principle, or universal law as such. \What the categorical imperative bids us do is to act for the sake of law as such ; and this means that the maxim of our action (whatever else it may be) should be to obey universal law as such. 5 There is thus only one categorical imperative; and we may
call it the categorical imperative'. I This formulation will now seem
(
little less
jejune
if
we remember
that, according to
Kant, the categorical imperative also bids us perform

;
which are good, not as a means to some further end, or to the some particular desire, but in themselves for an action good in itself is one which every rational agent as such would perform if reason had complete control over his passions. The unconditioned principle of pure practical reason is the principle of action good in itself, although
actions
satisfaction of
the categorical imperative expresses it as a principle of obligation as one which in our imperfect rationality we ought to obey. And
1
that
is,
we must
8 *
Gr. 9 421 47. See especially Chapter
IX
4.
Gr.,
Chapter VI 5, 420-1 = 46-7.
Compare Chapters IV 4 and VI
4.
[133]
134
[XIV
not think of universal law as a principle outside the action or as a further end for the sake of which the action ought to be done. On the contrary, it is the principle of the action, the formal principle which is embodied in the action and in virtue of which the action is good. 1
2.
The one
speaks of 'the categorical imperative' and asserts that he has in mind the principle of all particular categorical one, only imperatives ; just as the hypothetical imperative If you will any end, you 2 ought to will the means' is the principle of all hypothetical imperatives. Particular hypothetical imperatives are applications of the hypothetical
there
is
'
When Kant
imperative; and particular categorical imperatives, like 'Thou shalt not are applications of the categorical imperative. In Kant's language, they are 'derived' (abgeleitef) from it as from their principle, but this need not mean, and in fact does not mean, that they are 'deduced'. 3 Kant holds that to conceive a categorical imperative is to know what it enjoins, but that to conceive a hypothetical imperative is not to know what it enjoins. 4 This view requires some qualification. To conceive the principle of all hypothetical imperatives If you will any end, you ought to will the means' is to know what it enjoins Kant himself holds that it is an analytic practical proposition. What we do not know is the particular hypothetical imperatives in which it is applied. To
kill',
{
:
know
these we require to know something else namely, what end we seek and what are the means to it. If we have this knowledge, we know all that is necessary to establish a particular hypothetical imperative. As regards the principle of all categorical imperatives, to conceive this is to know what it enjoins, but it is not to know the particular categorical imperatives in which it is applied. So far there is no difference between the categorical imperative and the hypothetical imperative. The difference lies in this that in the case of the categorical imperative knowledge of the end we seek and of the means to this end does not determine how the imperative is to be applied. 5 The categorical imperative, unl ke the hypothetical, can and must be applied independently of our particular desire for a particular end.
1
Compare Chapters VII

imperative
and
4.
Strictly speaking, the principle

rests ultimately
need not be expressed in the form of an imperative.

principle that a rational agent
The
on the
who
wills
an end
will necessarily, qua rational, will the means. * Gr.y 421 47. Abbott actually translates it as 'deduced', although an examination of Kant's usage will show that it seldom or never means this, and although it is not the
ordinary German word for 'deduced*. Kant habitually opposes the derived or derivatory (abgeleitet) to the original (urspriinglich) without any suggestion that one is deduced from the other. For this usage as applied to intuition see K.r.V., 672; as applied to good see K.p.V., 125 267 ( 226); as applied to possession see M.d.S. f Rechtslehre, 6, 251.
* Gr. t
8
is indeed embodied in the material maxim whose conformity with the categorical imperative has to be judged; but our moral judgement must not be determined by the known fact that we seek or desire a particular end. See 5 below.
420 = 46. This knowledge
XIV
3]
UNIVERSAL LAW
that
135
by an act of will we can decide to live a life of obedience to moral law as opposed to one of self-interest and can thereby already obey the categorical imperative. But we could equally decide to live a life of prudence rather than one of impulse, so that there seems to be no real difference here,
Kant may perhaps think
3.
Universal law.
Kant's insistence on the duty to act for the sake of universal law as such is repugnant to many people ; but when we remember that he is here concerned only with the form of moral obligation, we shall see that a great deal of what he says is common to most, if not to all, moral philosophy which does not regard duty as purely subjective or as a matter of selfinterest. He is assuming, as we all must, that there are, or at least may be, other rational agents besides ourselves, and he is saying that the principle of moral action must be the same for every rational agent. No rational agent is entitled to make arbitrary exceptions to moral law in favour of himself or even in favour of his friends. To say that the ultimate moral law
must be universal is to say that every particular moral law must be objective and impersonal, that it cannot be determined merely by my desires, and that it must be impartial as between one person and another. In this there is surely nothing to cavil at, even if we believe that we have a direct intuition of an unanalysable quality of goodness (or obligation) and a direct intuition of the kinds of action in which such an unanalysable
quality
is
moral law already 1 implies reciprocity of obligation between person and person. It implies that I am not morally entitled to treat you on one principle and yet to claim that you should treat me on another ; or again that if I can claim to be treated by you in one way, I must be prepared to treat you in the same way. The importance of this for morality can hardly be exaggerated. It is made more explicit by Kant in his later formulae. need not further insist on Kant's view that the moral law holds for every rational agent, and that it appears as a command to finite human beings only because they are not completely rational. 2
Perhaps
necessarily manifested. we may add that the universality of
We
4.
Maxims.
A more distinctive feature of Kant's doctrine is his introduction of maxims as intermediaries between the abstract universal moral law and the concrete individual action. So far from embarking on a futile attempt
to deduce particular moral laws from the bare form of law as such he bids us begin with an action done or contemplated and consider what is the principle actually manifested in it. The principle or maxim in question is the material maxim of the action,* as, for example, 'When I am in need
1
Compare M.d.S. Rechtskhre, Chapter IV 2-4.

9
8, 256.
Chapter
VI
a.
136

I will
[XIV
borrow it on a promise of repayment which I do not of money, intend to keep'. 1 We must then ask whether we can will the action, not merely as falling under a principle upon which we propose to act, but also as falling at the same time under a principle valid for every rational agent. More simply, is our maxim merely a principle on which we choose to act or is it one which we can at the same time regard as valid for a rational agent as such? If we adopt or reject the maxim according as the answer to the second alternative is 'Yes' or 'No', we are acting on the 2 formal moral maxim 'I will obey universal law as such Such a procedure is surely in conformity with common sense. The formal maxim of morality does not act in a vacuum, but it selects from in the same kind of way as does among our proposed material maxims 3 the general maxim of prudence. In judging any particular case it is
5 .
always well to consider the concrete action, to formulate the principle manifested in it, and then to ask whether this principle can be regarded as a moral law or a moral rule. This is wiser than approaching each situation with a set of ready-made rules. The latter procedure may indeed serve well enough when the moral situation is simple, but we cannot always be sure beforehand that the moral situation will be simple. And it is important to remember that in judging the validity of moral laws, such as 'Thou shalt not kill', or of moral rules, such as 'It may be the duty of a soldier to kill', 4 we shall judge best if we have concrete instances
5 before our mind. We need not further labour the point that in dutiful action the material maxim, which may embody both consequences aimed at and motives for arriving at them, is present at the same time as the formal maxim. 6 The formal maxim is present if we would have rejected the material maxim had we thought it incompatible with universal law. Perhaps Kant wishes to emphasise the interpenetration, as it were, of the formal and the material maxim, when he uses the rather curious preposition 'Act only on that maxim through which you can at the same time will that it should become a universal law.' Obviously it will not always be easy to say whether or not an action has its sufficient determining ground in the formal maxim of morality. It may not even be easy to say what is the material maxim of an action, and in this there lie great possibilities of self-deception, which the genuinely good man will necessarily seek to eliminate. This question belongs to a discussion of the way in which Kant's formula is to be to observe that for different applied in detail. It is, however, important it is legitimate and necessary to consider different maxims for
purposes the same action.
What
1 Gr. 9
mean by
this is that if
is
wish to consider only what
we take a narrow view of duty and 7 'due', we shall omit from the maxim all
*
Compare Chapter IV 4. 48. 422 * Chapter I 5. Chapter VIII 8 at end, and Chapter III 4. 8 Moral judgement has certain resemblances to what is called 'intuitive induction* such as is found in mathematics, although there are also great differences between them, 7 See Chapter III, Appendix. Chapter XI 5.
5]
UNIVERSAL LAW
is
137
questions of motive. This
perfectly legitimate, but the truly conscientious
man
in judging his own actions will take his maxim as generalising the motive as well as the action. He will regard it as right to pay his debts, but wrong to do so from the motive of spiting his children ; and he will endeavour to diminish any feeling of bitterness in his heart and to prevent
.it
from influencing
5.
his actions.
Material maxims.
is
by the way of material maxims that circumstances are considered moral judgements. Material maxims are of the form 'When life offers me more pain than pleasure, I will commit suicide', 'When I am in a difficulty, I will tell a lie', and so on. These examples might indeed be intended to indicate the motive rather than the circumstances of the action, but even in the most moral action it may be necessary to specify the circumstances. Thus a thoroughly moral maxim might be 'When I see a person drowning, I will pull him out of the water.' If my maxim were 'Whenever I see any one in the water, I will pull him out', this would be a maxim, not of duty, but of insanity. Similarly it is by the way of material maxims that ends and consequences are considered in our moral judgements. 'When I can cause any one pain, I will do so' is a thoroughly immoral maxim. 'When I can cause anyone pleasure, I will do so' is a moral maxim if it is accompanied by the proviso that the pleasure caused must not be incompatible with moral law, as would be, for example, pleasure in the pain of others. Kant is surely right in saying that an action is not good or obligatory merely because it aims at so-called good consequences like pleasure or even the pleasure of others we have to take into account the whole context in which so-called goods may be real evils. And he is also right in saying that there may in principle be some cases in which remote good consequences must be ignored. It is wrong to refuse to pay your debt to a rich man on the ground that the money would more profitably be spent on the education of your children. All this does not alter the fact that a material maxim is concerned with ends and consequences, and so alone can it be judged. A further advantage arising from the introduction of material maxims is that the maxim, however specific, is abstract and general, while the
It
in our
action is individual and concrete. This means or 'play-room' in the application of the maxim. latitude where the maxim has to be rejected as maxim is accepted as moral, it is important to that there is still some place for common sense.
6.
that there
is
a latitude
There is indeed no such immoral; but where the

recognise, as
Kant
does,
The canon of moral judgement.
It
explain
1
may be objected that even if all this be true, it does nothing to how some material maxims can be regarded as fitted to be
Chapter IV 2 towards the end,
138
what every universal rational agent as such would do, and this may sound very fine indeed; but how are we to find out what every rational agent as such would do? From a logical or theoretical standpoint this objection is sound. cannot extract theoretical propositions from the mere form of law or universality. If we are given a series of theoretical propositions, we can indeed say which are universal in form and which are not. The logical form of universality could not, however, tell us which universal propositions were true and which false, though we might be able to decide this without further ado in the case of analytic propositions. Is Kant asserting that, given the two propositions 'All rational beings ought to tell the truth' and 'All rational beings ought to tell lies', we can say by mere inspection that one is false and the other true? He himself holds that these are not analytic propositions.
THE CATEGORICAL IMPERATIVE laws and others not. We may be bidden to do
[XIV 6
We
From
this
standpoint
it
seems as
is
if
we
shall
have
to fall
is
On this point Kant gives us no very clear light beyond saying on occasion that certain moral principles 'leap to the eye' ;* and it may be that he would not be altogether averse from admitting some element of what may be called direct intuition. But attention must be called to another distinctive feature in his doctrine namely, that the propositions in question are not theoretical, but practical, propositions and must be treated as such. There is a special 'canon' for practical moral propositions we must be able to will that a maxim of our action should become a universal law. 2 From the standpoint of action Kant's principle takes on a very different colour. In actual practice the value of the principle can hardly be exaggerated. The attempt to stand outside our personal maxims and estimate impartially and impersonally their fitness to be principles of action for others than ourselves is the necessary condition of all moral judgement and may throw a most unwelcome light upon our own actions and our own character. It must also be remembered that all of us, even very young children and it may be conjectured even sceptical philosophers, have a very acute and precise sense of what is unfair or unjust in regard to ourselves. If men were willing to take these principles,
:
direct intuition that telling the truth rational agent as such would do.
good or obligatory or
back on a what every
to universalise them, to apply them to others, and above all to act upon them when so universalised, the world would be a much better place than it is. Even bad men have this clear conviction ofjustice and injustice,
though they fail to universalise it. Hitler himself was probably genuinely shocked at any injustice inflicted upon his country or compatriots by
foreigners.
explain away all such convictions in philosophy; we may that they are mere illusions; but we cannot get rid of them in practice. Yet in making claims on our own behalf to standards which we implicitly assert to be objective, we are appealing to Kant's principle of
We may
show
Gr.,
E.g. M.d.S., Tugendkhre, 50, 424
8, 427.
Compare
also Gr. 9
423-4
50.
XIV
7]
UNIVERSAL LAW
139
truly rational will must surely, as Kant maintains, will coherently, as a rational mind must think coherently. cannot without contradiction claim that all rational agents as such should treat us in accordance with certain universal principles, and claim at the same time that we are not bound by these principles ourselves.
a universal law.
We
7.
Contradiction in the will.
The phrase 'able to will' is manifestly the main clue to the understanding of Kant's doctrine, but it is unfortunately very hard to interpret. It appears to mean 'able to will without contradiction'. We have already found that a contradiction arises in imperfectly rational wills which act 1 contrary to their own hypothetical imperatives. We are now told that a contradiction arises also in regard to categorical imperatives. 2 So far Kant is a supporter of the coherence theory of goodness, a theory which has had wide support from Plato onwards but is difficult to make
sufficiently precise.
clearly a contradiction in willing that a maxim should be a and willing at the same time that we should make arbitrary exceptions to it in our own favour. According to Kant we shall find that this is what happens in our actual infringements of the law. do not and on his view we cannot will that our immoral maxim should be universal law. the contrary, we recognise that the law stands; but we say to ourselves 'This is a very special occasion' or 'I am a very special person',
There
is
universal law
We
On
so proceed to establish an exception to our own advantage. He adds that although from the point of view of reason there is a contradiction in this, in action it is rather an antagonism between reason and inclination, an antagonism in virtue of which we sophistically take the moral law to be general rather than truly universal. Nevertheless, he maintains, we recognise the validity of the categorical imperative, and only permit ourselves, as we pretend, some inconsiderable exceptions to it. All this is sound enough practically, though we may add that the antagonism would be better described as one between a rational volition and a less rational volition. But it supposes as I supposed in the case of injustice to ourselves that we know what the law is it does not explain how it is possible to know it. If we can assume certain maxims to embody the law, it is not too difficult to judge our consistency or inconsistency in their application. The difficulty is to see how the universalisation of certain maxims immediately brings out their self-contradictory character. In the next chapter we shall have to consider Kant's further attempts to explain
and
this.
8.
The
coherence
of rational
wills.
There
it is
1
is,
made more
Chapter XII
however, already another side to Kant's doctrine, although explicit in his later formulae. The form of law is a form
6.
2
Gr.,
424
51.
14<>
[XIV
which requires a matter, and this matter consists of the inclinations and ends of rational agents. There is a suggestion that some matter can, and some cannot, be adjusted to this form. The Idea of our pure rational will stands to our will as affected by sensuous desires in roughly the same
relation as the categories stand to our sensuous intuitions. 1 This doctrine is already expressed in the statement that we must be able to will our material maxims as universal laws. Kant brings out this side of his doctrine most clearly in the Critique 2 of Practical Reason. There he maintains that only if our actions are
determined by a formal principle can there be a harmony or coherence (Einstimmigkeit] of rational wills. He expresses astonishment that intelligent men should have imagined maxims of self-interest, because they are universally acted on, to be therefore suitable to constitute a universal practical law. The result of adopting such maxims would produce only the greatest disharmony; for in that case the will of all has not one and the same object, but every one has his own, and any harmony between
them
is
harmony,
purely accidental. as when Francis
The harmony
I
remarked 'What
of self-seekers is really a dismy brother Charles wants
(Milan), I
want
too.
Kant's view is clearly that coherence of rational wills can be based only on obedience to one and the same universal law as such, and that without this there can be no genuine coherence. 3 This principle applies more obviously to external legislation, but according to Kant it applies equally to internal legislation and presumably to our duty to ourselves as well as to our duty to others. If this is so, in seeking to obey universal law as such we are seeking to realise the condition of coherence among rational wills. It is not yet clear how far Kant is taking into account the possible coherence of rational wills in determining what maxim is or is not suited to be a universal law. This question belongs to the discussion of our maxims as possible laws of
nature.
9.
The
rational will as arbiter.
Whatever the difficulties in Kant's view and they are many his fundamental assumption is that the will is as rational in action as intelligence is in thinking. We know what is good or obligatory, not in abstract thinking, but in virtue of actual or imagined willing. If we abandon mere self-seeking and adopt the standpoint of universal law, the will with no great difficulty will decide in a concrete situation what it ought to do. He is not attempting to introduce any new principles of morality but merely to formulate with precision the principle actually at work in the moral actions of ordinary men. 4
Gr. 9 454 88. More will be said about this in the Appendix to this chapter. K.p.V., 4, 27 ff. 137 ff ( 6839. 49 ff.) ; see also K.r.V., A8i i 8 See also Ernst Cassirer, Kants Leben und Lehre, pp. 256 ff. in n. ( 14 n.). *K.p.V., 8n.
8
1
XIV
10]
UNIVERSAL LAW
14!
Kant may exaggerate the ease and certainty with which the rational will decides upon its course of action, but it seems unfair to charge him with offering us a mechanical criterion of moral action and unreasonable
1 complain that he has failed to do so. He himself compares his principle 2 of a but a to possession compass does not absolve us from the compass; out where we are of in relation to our destination, nor finding necessity even from the necessity of adjusting ourselves to the lie of the land as we proceed in that direction. We cannot base a great deal on a mere metaphor, but it looks as if we ought to be told more about the best way to use the compass we have been offered.
to
formulae.
To some extent the information required is supplied by Kant's other What is surprising is that he speaks as if the present formula by itself were sufficient and as if in using it we were following the strictest method. 3 One would rather have imagined that by itself it was incomplete.
10.
The permissible and
the obligatory.
There is a further point of interpretation which demands attention. Kant's doctrine is commonly supposed to be that where we can will a maxim to be a universal law, we ought to act on that maxim. That is to say, the possible universality of a maxim imposes on us a positive duty. Yet it seems not too difficult to will that the maxim of playing games in one's spare time should be universal law, although few would suggest that every one ought to play games in his spare time even if he feels that he can get more satisfaction or more relaxation in other ways. It is quite clear that on Kant's view to act on maxims which will not meet the requirements of universality is to act wrongly. It is also clear
permissible (and in that sense 'right') to act on maxims which requirement. But what is the ground for saying that he regards it as a positive duty to act on every such maxim? It will be noted that in the Groundwork Kant is prepared to express the 4 first formula of the categorical imperative negatively. Even in the variant we have given 5 the phrase used is 'Act only on that maxim through which .' The word 'only' has a rather negative flavour which may be
that
it is
meet
this
intended primarily to exclude maxims which do not satisfy this demand. It is true that the word 'only' is sometimes omitted; 6 but against this it may be observed that while a good man has the duty at all times of acting on maxims which can be willed as universal laws, it does not follow that in any particular situation he is obliged to act on a particular maxim which may meet this test not if there are other suitable maxims which
it equally well. should also be noted how Kant says expressly 7 that the universal validity of our maxim as a law must be the limiting condition of our actions. This phrase, which is more commonly used of the Formula of the End
will
meet
It
Compare
Gr.,
Gr.,
Gr.,
4
6
404 = 24-5. 402 = 22; 440 436-7 = 67.
his rejection of
a universal criterion of truth in K.r. V. t A.$8

*
ff.
B8a
ff.
Gr.,
Gr.,
Gr.,
436-7
421
71.
7
449
= 47. = 92.
67.
142
THE CATEGORICAL IMPERATIVE [XIV APP.
in itself, 1 indicates a limiting condition of freedom and is clearly a negative condition. It entirely fits in with the doctrine that every man has a right (but not a duty) to pursue his own happiness except in so far as this is incompatible with universal law. 2 If this interpretation is correct, it puts an entirely different complexion on Kant's doctrine. The categorical imperative, so far, prohibits certain actions ; but where actions are not prohibited, we have every right to go forward as we please in accordance with our inclinations. The young man, so often quoted, who is in love with a young woman, can go ahead and marry her without asking whether it is his duty to do so. All he has to consider is whether there is any just impediment in the way. It may be objected that nevertheless Kant manages to extract from this formula positive duties, like the duty of developing talents and of helping others. reply is that even here Kant at least professes to establish these duties by showing that to will the opposite is to fall into self-contradiction. 3 Furthermore, in such cases our duty is not to perform certain definite actions but rather to adopt certain maxims of action which
My
must be applied with prudence. So far as I can see, the Metaphysic of Morals bears out my interpretation. There we are told that according to categorical imperatives some
actions are permissible or non-permissible, while others or their opposites are obligatory. 4 The positive duties which are not merely the other side of a prohibition seem to depend on something more than mere form, on ends which are also duties, and so on matter. 5 All this is connected with the difference between perfect and imperfect duties, and also with the difference between legal and ethical duties. It is not easy to find one's way through the complications of Kant's doctrine, but without doing so
no one
is
entitled to
be dogmatic in
his interpretations.
APPENDIX
THE SPONTANEITY OF MIND

i.
Intellectual spontaneity.
KANT'S ethical doctrine maintains that pure practical reason can spontaneously order and regulate, in accordance with its own principle, the human maxims and actions which are based ultimately on desire.
1 *
Gr., 431 = 59. See also the use of the word 'permitted* (erlaubt) in Gr., 439 70. 4 M.d.S, 9 Rechtslehre, Einl. IV, 221-22. Gr., 424 50. *M.d.S. f Tugendlehre, Einl. II, 382 f. 226 ff.; 4, 419.
XIV APP.
doctrines
l]
UNIVERSAL LAW
143
This doctrine can best be understood
which he propounds in
when we have grasped similar regard to the spontaneity of mind as
manifested in other spheres of activity. In the sphere of knowledge we should, according to Kant, know nothing either of objects or of ourselves apart from the spontaneous activity of imagination and understanding. Mere sensations as passively received, and in that sense given, have in themselves no reference either to objects or to a self; nor could knowledge either of self or of objects ever arise from any loose association of images or even from memory of
in so doing give rise to transcendental schemata corresponding to certain principles of combination inherent in the understanding. 1 These principles, in accordance with which the understanding works as, for example, the principle of ground and consequent when they are consciously conceived and applied through the schemata to sensations are the categories; and through them we
successive sensations. sensations in one time
The imagination must

and space and
spontaneously combine
experience a world of objects, a world of permanent substances, the succession of whose accidents is causally determined. Without the spontaneous activity of mind working in accordance with its own principles there could be no knowledge of an objective world nor could there be any knowledge of the knowing self. Thus in knowledge the mind must function in accordance with its own principles just as the will must function in accordance with its own 2 principles in moral action. There is, however, a fundamental difference between the two. In knowledge the mind is applying more than the mere 3 concept of law or universality as such. According to Kant, in the ordering of given sensations it has at its disposal a whole series of specific categories, and it has also before it the forms of time and space, through which these categories can be applied. In action the will works only with the form of law or universality, and it is consequently much more difficult to understand how its principle can be applied in the ordering of actions based on
4 given desires. There is a further difference between the two cases. The categories are concepts of the understanding; they refer to sensible objects, apart
from which they have no objects at all; and they are, so to speak, confirmed by their actual use in experience. The supreme moral principle is an Idea of reason which can have no object in sensuous experience and so cannot be confirmed by experience: it belongs and this is Kant's view to another world than the world of experience, though we can
give
1 8
it
a kind of object so far as
we
act in accordance with
it.
Chapter IX, Appendix 3. Gr. t 454 = 88. Kant indicates that the parallel is only a rough one by using the word 'wgefdhr* (roughly), which Abbott perversely translates as 'precisely*. 8 Kant himself rather slurs over the difference by saying that the categories by themselves mean nothing other than the form of law in general ; but surely they mean specific forms of law. 4 Kant himself, however, develops his categories also into categories of freedom, and he seems to consider the absence of reference to space and time to be a positive advantage; K.p.V., 65 -6 = 186-8 (= 115-17).
144
THE CATEGORICAL IMPERATIVE [XIV APP

2. Aesthetic spontaneity.
Kant recognises also a further kind of spontaneity the spontaneity of imagination in the experience of beauty. In knowledge imagination is at the service of concepts its business is to exhibit concrete instances falling under concepts. Just as imagination produces transcendental schemata corresponding to the categories when it combines sensations in one time and space, so it produces schemata on a humbler level in making pictures or images which might be tactual as well as visual corresponding to the empirical concept of dog or the mathematical concept of triangle. 1 These schemata are sometimes spoken of as themselves wavering diagrams, 2 but they seem more correctly to be the imaginative apprehension of the method necessary to construct such images and in this way to be kinaesthetic quite as much as visual. In all this we have a co-operation or harmony between understanding and imagination, the work of imagination always corresponding to concepts entertained by the understanding. In aesthetic experience Kant recognises a further activity of imagination not directed to the construction of instances falling under a concept. Here the concept is irrelevant: a picture is not necessarily beautiful because it is the picture of a madonna. The distinctive feature of Kant's
:
is that he considers successful aesthetic activity to be still a kind of co-operation between imagination and understanding, although without a specific concept. Presumably the work of imagination possesses that kind of unity in difference which might be necessary for the 'exhibition' of a concept, but is here produced independently of a concept, the individual image being portrayed for its own sake. The success of this co-operation between imagination and understanding is judged, not by reference to any concept, but by a special kind of feeling, the aesthetic feeling or the feeling of beauty. Here too we have a kind of parallel between the free activity of imagination in harmony with understanding and the free activity of the rational will in harmony with the Idea of law. On this ground Kant himself speaks of beauty as a symbol of morality. The freedom of the imagination is 'represented' here presumably meaning 'felt' in the appreciation of the beautiful as harmonious with the law-abidingness of the understanding; and the freedom of the will is conceived in moral judgement as an agreement of the will with itself in accordance with universal laws of reason. 3 The moral feeling of satisfaction in the performance of duty corresponds in some ways with the aesthetic feeling of
aesthetic theory
beauty.
Ai40~a
and
8
59, 35*
Bi8o-i. See also K.U.,
57, Anmerkung I, 341
ff.
239
ff.),
(=254 #)
ssa
259). In the previous paragraph Kant, in speaking of aesthetic gives itself the law in regard to objects of so pure a satisfaction, as reason does in regard to the faculty of desire (Begehrungsvermogen.) This, however, just is only one of the many ways in which beauty is a symbol of morality.
K.U., 59, 354 ( judgement, says that
it
XIV APP.
2]
UNIVERSAL LAW
145
Nevertheless here also there are fundamental differences between the
two cases. The aesthetic judgement is based on feeling, but to Kant it is anathema to say that the moral judgement is based on feeling the precise contrary is true, for the moral feeling must be based on the moral judgement, and the moral judgement rests on the concept of universal law as 1 such, while the aesthetic judgement rests on no concepts whatsoever. Hence there are difficulties in regard to moral judgement which do
:
not arise in regard to aesthetic judgement.
1 This is why in aesthetic judgement the harmony the moral judgement it is 'conceived* or 'thought*.
is
'represented* or
'felt',
while in
CHAPTER XV
THE FORMULA OF THE LAW OF NATURE

i.
Formula
la.
A UNIVERSAL LAW OF NA TURE. 1
Act as if the maxim of your action were
to
become throughyour will
The difference between this formula and the previous one is sharp, and should not be slurred over. Up till now we have been concerned with a universal law of freedom one on which any rational being would act so far as reason had full control over passion. The best, if not the only, way
make such a law vivid in our imagination is to picture to ourselves a world in which everybody in fact acted in accordance with it. This is the eminently sensible procedure which Kant now commends to us. It is one which is commonly followed by ordinary men. 2 The duty of firewatching, for example, was sometimes pressed home by the question 'What would happen if everybody refused to do it?' To ask questions of this kind is to consider a maxim as if it were to become through our will a universal law of nature. Here, as so often, Kant is close to the logic of action, which must be distinguished from the logic of abstract speculation. We can will that the maxim of our action should become at the same time a universal law of freedom in the sense that we can will our action as an instance of a principle valid for all rational agents and not merely adopted arbitrarily for ourselves. We cannot, however, will that our maxim should become a universal law of nature: that- is a project far 8 beyond our power. Hence in this new formula Kant very properly says .' Act as if In using this formula we put ourselves imaginatively in the position of the Creator and suppose that we are making a world of nature of which we ourselves are a part. 4 In the Critique of Practical Reason^ Kant puts his formula in a more elaborate way 'Ask yourself whether you could regard
to
<
.
. :
your proposed action as a possible object ofjour will if it were to take place in accordance with a law of nature in a system of nature of which you were yourself
9
apart.
the illustrations given in the Groundwork* it is clear that in formula we assume empirical knowledge of nature (parapplying
this
* Gr. t 421 K.p.V., 69 192 (= 122-3). 47. 66 and K.r.V., ASoy believe this to be Kant's view see also Gr. t 436 8835 to have the universality if will maxims could he as we at times though speaks loosely
I
From
I 1
of a law of nature ;
*
e.g., Gr.,
424
50.
*
Religion, Vonede, 5
Gr., 421
ff.
= 48ff.
(=
vra).
[I 4 6]
K.p,V., 69
192
(=
122).
XV
2]
THE LAW OF NATURE
147
ticularly of human nature) and its general laws. This shows again the absurdity of the view that Kant proposed to apply the moral law without taking into account any empirical facts. The moral law remains a priori even when applied to men, but if we are to bring particular cases under it and enable it to affect human wills, we must have judgement sharpened
by experience.
apply Formula I by means of Formula la, though Kant speaks as if Formula I could be even better applied by itself. 2 In the Critique of Practical Reason* he asserts that Formula I can be applied only through Formula la just as a pure 4 category can be applied only through a transcendental schema. If Kant had regarded the Formula of the Law of Nature merely as a useful, or even as a necessary, practical device in estimating the suitability of our maxims to be moral laws, there would be little or nothing to boggle at. The weakness of his position is the apparent suggestion that by this means we can reach a purely intellectual and nonmoral criterion of moral law at least in the case of what he calls 'perfect
illustrations in the Groundwork all
duties'.
The
2.
Perfect
and imperfect
duties.
attaches great importance to the distinction between perfect duties, but he seems nowhere to define the distinction clearly, although he admits that he is not following the ordinary usage of the schools. 5 The ordinary usage regarded duties as perfect if they could be enforced by external law, and as imperfect if they could not be so enforced. 6 This would presumably mean that perfect duties could be only duties to others. Kant holds that we can have perfect duties also to
Kant
and imperfect
ourselves.
inclination,
Perfect duties, he here informs us, admit of no exception in favour of and this would suggest that imperfect duties do admit of
such exceptions. Such a suggestion is surprising from Kant, and he is 1 probably speaking more precisely in the Metaphysic of Morals, when he indicates that the exceptions in question are rather the limitation of one maxim of duty by another as when the duty of benevolence to a neighbour is limited by a similar duty to one's parents. Kant's view is apparently that some duties, as, for example, the paying of debts, are not so limited we are not entitled to refuse payment of a debt on the ground that we need the money to assist our parents; and the payment of a debt must be regarded as a perfect duty.
:
= 35; 41 on. = 33 n.; M.d.S., Rechfslehre, Einl. Gr., 389 == 6. See also Gr. 412 216-17 = 1 6. The first passage cited shows signs of Kant's confusion between moral law and moral laws. Compare Chapter I 5. * = 193 ( = 123). K.p.V., 70 Gr., 436-7 67. * See Chapter IX, Appendix 3; Chapter XIV, Appendix; and also the Appendix
1
t
II,
to the present chapter.

6
Crr.,
421 n.
= 47 n.
=
*
7
Mellin, Worterbuch der Kritischen Philosophic, IV, 562. 235-6. M.d.S.> Tugendlehrc> Einl. VII, 390
148
[XV
That there is an important distinction between such duties as the payment of debts on the one hand and such duties as benevolence on the other must be readily admitted; but we may doubt whether even the 1 an duty of paying a debt is to employ a term Kant uses elsewhere unconditioned duty in the sense that it cannot be overridden by any other. As Plato pointed out long ago, it would not be a duty to return a 2 weapon to a man who had become a homicidal maniac. Kant marks his distinction also in another way. In the case of perfect duties we are obliged to perform a definite act for example, to pay 5 gs. 6d. which we owe. In the case of imperfect duties we precisely the are bound to act only on a maxim although we ought to act on the maxim
:
to our discretion to decide whom we ought to help, and to what extent we ought to help. There is thus a 'latitude' or 'play-room' in the case of imperfect duties, which are also called 'broad' or 'meritorious', as opposed to 'narrow' or 'strict' or 'rigorous', duties 3
of benevolence,
it is left
last three adjectives appear to be equivalent to 'perfect' when they are applied to duties. If we have perfect and imperfect duties both to ourselves and to others, duties are thereby classified into four main kinds. The object of Kant's illustrations is to show that each kind of duty falls under the Formula of the Law of Nature. Maxims opposed to perfect duties cannot even be conceived as laws of nature without contradiction. Maxims opposed to imperfect duties can be conceived as laws of nature, but could not be willed as such without contradiction.
The
3.
The causal law of nature.
In its strictest sense Kant usually takes a law of nature to be a causal law, and it is essential to such a law that it should have no exceptions the same cause must always produce the same effects. Hence it may be thought that if we are to find contradiction in a maxim when it is conceived as a law of nature, this must be because it would assert that the same cause could produce different effects. It is perhaps possible to interpret Kant's discussion of suicide on this
:
basis.
Here the maxim
a principle of self-love, and we assume empirical knowledge that self-love described here oddly as a sensation has the 'determination' (Bestimmung) to work for the furtherance of life. Hence if we universalise this maxim of a law of nature, we are supposing that self-love, which is the cause of life, should in certain circumstances be the cause of death. We are, in short, conceiving a law of nature to admit of arbitrary exceptions, and so we are falling into a contradiction. From this the inference is made that the maxim must be contrary to moral law: it is a breach of a perfect duty to
offers
is
me more
is supposed to be 4 pain than pleasure'. This
'I
will
commit
suicide, if life
oneself.
1
Das Mag
fop* 33 *
in der Theorie richtig sein;

*
c.
Gr.,
424
Ak. VIII, 300 n. 50-1.
Gr.,
421-2
= 48,
XV
4]
is
THE LAW OF NATURE
149
right in saying that there can be no arbitrary exceptions either to a moral law or to a law of nature: but it is manifestly impossible to find, by this method, breaches of a law of nature from which we are assert no breach of a law of entitled to infer breaches of a moral law. nature if we say that food, which ordinarily causes life, may in special circumstances for example, in certain kinds of illness cause death.
Kant
We
there be any breach of causal law if self-love, which ordinarily should in special circumstances cause death. 1 Indeed we may say generally that any attempt to make the causal law of nature a test of moral law is foredoomed to failure. It is, however, abundantly clear, if we look at the discussion of the other kinds of duty, that this is not the correct interpretation of the argument. Kant is not concerned with causal laws nor with finding breaches in them his arguments, if interpreted in this way, are completely broken-backed. In every case he appeals to teleological considerations; and there is no possibility of even beginning to understand his doctrine, unless we realise that the laws of nature he has in mind are not causal, but
causes
life,
:
Nor need
teleological.
4.
Teleological
law
in nature.
Nature is the totality of phenomenon governed by law. Phenomena are the matter of nature ; but the form in virtue of which they constitute nature and not mere chaos is the form of law. This law is primarily causal law, which covers not only the mechanical causation of physical bodies, but also the instinctive behaviour of animals, and even from one point of view all human action and experience. Nevertheless in nature causal law is not the only kind of law recognised by Kant. Even in the understanding of physical nature we may have to use another concept besides that of causal law the concept, namely, of purpose or end. This concept seems to be necessary for the study of organisms. To say this is not to say either that organisms and their organs are the product of conscious purpose or that they themselves have a conscious purpose it is rather to say that we must consider them as if they had a 2 purpose and see whether in this way we can understand them better. For the understanding of human nature the concept of purpose or end is still more necessary; for it is an essential characteristic of human nature to set purposes before itself. As we shall see later, 3 morality even seems to demand that we should act as if nature itself were purposive and had a final end. Teleological laws, it is true, do not constitute nature in the same
:
Incidentally it would be interesting to discover what ingenious gentleman first interpreted this passage as meaning that suicide could not be a law of nature because, if it were, there would be nobody left to commit suicide This interpretation is repeated so frequently as to be almost orthodox ; yet it has no justification in this passage nor, so far as I know, in any other. 8 They seem to have what Kant calls *purposiveness without a purpose* that is, without a conscious purpose.
!
See Chapter
XVIII
8.
150
sense as

do causal laws
:
[XV
perhaps even stretching a point to speak of them as laws of nature. Nevertheless in using the law of nature as a means of applying the moral law Kant is going behind causal law to the more fundamental abstract idea of law as such, which constitutes the form of nature. Why he then proceeds to take teleological laws to embody law as such we shall consider in a moment.
it is
5.
The perfection of nature.
the background of Kant's doctrine we must remember the teleological view of an organism is not only that as a whole it is adapted to a purpose or end, but that every organ is also adapted to a purpose or end which is an element in the total purpose or end. 1 In the Critique of Pure Reason he declares that reason in considering living beings must necessarily accept the principle that no organ, no faculty, no impulse, indeed nothing whatsoever, is either superfluous or disproportionate to its use, but that everything is exactly adapted to its
that for
To understand
him
21
purpose (Bestimmung) in life. That he was deeply affected emotionally by 3 this view is brought out by a story which, according to Wasianski, he used himself to tell as an old man. In a cool summer, when there were few insects, he discovered that the swallows themselves were casting some of the young ones out of the nest so that there should be sufficient food for the others. At first, Kant said, he could not believe his eyes and added with strong emotion 'My mind stood still there was nothing to do but
;
fall
down and worship.' Whether or not his observations were correct, his whole attitude may seem strange today. Eighteenth-Century optimism about nature was shattered by the advent of Darwinism, and we tend to regard nature as 'red in tooth and claw', careless of the race as she is careless of the
individual. The correctness or incorrectness of Kant's biological presuppositions, however, is not here in question. What is important is this that on Kant's view to conceive human nature as governed by teleological law is to suppose a complete harmony of ends both within the race and within the individual. can consider human nature as if there
We
were such a systematic harmony of ends in accordance with a law of nature and we can ask whether any proposed maxim, if it were made a law of nature, would fit into such a systematic harmony. Some maxims would destroy such a systematic harmony, while others would merely fail to foster it, and this seems to be the basis of the distinction between perfect and imperfect duties.
;
6.
The appeal
to teleological
law.
may
1
Kant's appeal to a teleological, rather than to a causal, law of nature at first sight seem arbitrary, but this is far from being the case.
See Chapter II 8. K.r. V., 6425. Immanuel Kant in seinem let&n Lebensjakren, p. 41 1, in Immanuel Kant, by Alfons Hoffmann, which also contains the accounts of Kant by Jachmann and Borowski.
*
XV
6]
THELAWOFNATURE
15!
are asked to conceive a proposed maxim as a law of nature, we must conceive it as a teleological law of nature; for it is a maxim of action, and action as such (quite apart from moral considerations) is essentially purposive. Furthermore, we are asked to conceive it primarily as a law of human nature, even if we are setting it against the background of nature as a whole ; and human nature must be regarded as essentially purposive. All this was apparently so much taken for granted by Kant that he fails to state it explicitly, and so tends to mislead his
readers.
When we
When we
ask whether
we can
will a
proposed
maxim
as if it
were to
become thereby a law of nature, we are asking whether a will which aimed at a systematic harmony of purposes in human nature could
maxim as a law of human nature. Clearly without empirical knowledge of the needs, desires, and powers of men, but it is no part of Kant's doctrine that the moral law can be applied without any regard to empirical knowledge of the facts of
consistently will this particular
we cannot do
this
human
In
life.
we must be clear what Kant is doing. He is putting forward the doctrine that the ideal coherence of human purposes and human wills is the test or criterion, but not the essence, of moral action. This is why the moral law must be distinguished from the law of nature. This distinction may be seen most clearly as regards the duty of kindness or benevolence. Here we are aiming at the happiness of other men. The happiness of the individual may be regarded as the systematic satisfaction of his desires and the systematic harmonisation of his ends. In normal action it is our duty to promote this happiness in individuals
all this
own comfort. We may do it because we happen to be good-hearted people whose generous emotions happen to be peculiarly strong. Such conduct may be prudent; it may even be admirable and praiseworthy and amiable; and it would be chimerical to ignore these
tribute most to our
2
so far as it is not inconsistent with similar happiness in others. 1 No doubt in each case our duty is towards one individual or a group of individuals, but this must be subject to the controlling maxim of promoting the general happiness ; and in this sense we may be said to be aiming at the systematic harmonisation of human ends. But this we may do for various reasons. may do it for the purely selfish reason that in this way we may con-
We
auxiliaries to morality. Nevertheless a man is not morally good merely in virtue of the fact that he is pursuing the systematic harmonisation of human ends. For moral goodness something more is necessary. good
not moved merely by an emotion, however amiable: he is seeking to obey a law valid for rational agents as such, binding upon him and upon others even in the absence of generous emotions, and indeed even in the presence of natural dislike. Nevertheless, since such obedience is the only way in which the purposes of men can be fully harmonised, we may take it that when a maxim universalised as a law of nature could lead to such a harmony of human purpose, that maxim is fit to be adopted also as a moral law. When it is so adopted, we perform benevolent actions
is
1
man
here omit further qualifications.
Compare Ak. XIX, Nr.
6560.
152
[XV
because they are in themselves good and because they are our duty, not merely because of a desire for their results. By the help of the law of nature we can decide what we ought to do ; but the law of nature tells us nothing about the spirit in which we ought to do it. Hence, as I have said, we have here a test of moral action, but not its essence.
Kindness.
shall see
7.
We
trations.
Kant's doctrine more clearly
if
we
look at his
illus-
He admits further that men could go about their purposes even if nobody ever helped any one else mutual refusal of help could perfectly well be conceived as a law of nature. There would, however, be inconsistency in a will which willed this to be a universal law ; for since each of us at some time is bound to seek help for himself, he would thereby will an exception to this law, and consequently he could not will it to be a law. This way of putting the matter stresses too much the need of the individual agent for help and so gives rise to the mistaken view that Kant is concerned merely with self-interest. It is, however, true to say that our kindness will be less mixed with condescension if we remember that we also are subject to human needs and it is important to emphasise that we can reasonably claim help from others only if we are willing to regard the principle of helping others without thought of reward as a
will ever help us.
:
As regards the duty of kindness or benevolence, 1 Kant assumes empirical knowledge that men, because of their weakness, at times need and desire each other's help. He does not argue from this prudentially that we must help others if we wish to be helped ourselves nor does he make the improbable suggestion that if we refuse help to others, nobody
;
law which is binding equally upon ourselves. Nevertheless, as Kant shows in another discussion of kindness, 2 the argument does not turn on the fact of my own needs it holds even if I am strong enough to do without the help of others. The argument turns rather on the fact that human beings are in need of mutual help, and that only by means of mutual 8 help can the systematic harmony of their purposes be attained.
:
8. Promises to repay loans.
As regards the keeping of promises to repay borrowed money, 4 Kant assumes empirical knowledge that the purpose of such promises is to produce trust and so to get out of financial difficulties; and further that the universal breaking of such promises would make the attainment of
Gr., 398-9 = 17-18. There Kant himself says helping those in need is a duty because we must regard each other as fellow men subject to needs and as united by nature in one dwelling for purposes of mutual help. O., 422 = 48-9.
1
Gr.,
423
a* 49-50.
Compare
M.d.S,, Tugendlehre,
30, 453.
XV
this
8]
THE LAW OF NATURE
153
purpose impossible. Here again Kant does not proceed to argue 1 prudentially that I ought to keep my promises if I want to be believed, nor does he make the improbable assumption that if I break this promise,
the result will be that I shall never be believed again. What he argues is that if my maxim were to be a universal law of nature, so that this everyone in financial difficulties made similar false promises, this would defeat the very purpose of such false promises. The law would cancel itself out and could not even be conceived, let alone willed, as a law of nature. Here again, it must be remembered, a teleological law of nature, on Kant's view, must assert the adequacy of every organ for its purpose and could not admit of purposes which defeated themselves. What Kant says is true enough so far as it goes, but it does not offer a satisfactory basis for moral judgement unless we make the further assumption that the keeping of such promises and the mutual confidence thereby aroused are essential factors in the systematic harmony of human
an assumption which Kant makes is shown by his I think, read it into the present argument if we are to find it other than artificial. Furthermore, without assumptions of this kind we might find ourselves compelled to regard honour among thieves simply as a virtue. It may be said that when we talk about mutual confidence we are
purposes.
That
this
is
later discussion, 2
and we must,
simply judging the action by its results, and that the action is still wrong, even if it does not disturb mutual confidence. This might very well happen if the lender were to die suddenly without leaving any record of the loan. In one sense we are not judging the action by its results firstly, because Kant does not pretend that universal disturbance of Confidence will be the result of my action, and secondly, because Kant is not arguing that because I happen to want the prevalence of confidence among men, therefore I ought not to break a promise. Such an argument would be prudential and not moral. What he is doing is to take the principle of our action and to ask whether if universalised it is compatible with a systematic harmony of purposes in society. The answer is clearly 'No', and this answer remains, even if in this particular action I might be able to escape
discovery.
In another sense he is taking into account the consequences. endeavouring to see the action as it fundamentally and essentially is
is,
He
is
that
as
an action destructive of mutual confidence and of any systematic

;
harmony of purposes
sphere) it is the killing of a human being. This seems legitimate, although it might be said that he lays too much stress on the destruction of confidence, and too little on the fact that the breaking of such a promise is in itself the treacherous disruption of a systematic harmony of wills,
1 8
just as in considering
in a particular sphere (and ultimately in every murder one must take into account that
Gr. 9
Gr.,
He also makes the further assumption see M.d.S., Tugendlehre, that the power of communicating our thoughts has as its essential and natural to the discussion of lying as a telling of the truth ; but this assumption belongs sin against oneself, not as a breach of our duty to others.
429 end the
9,
402 = 22-3. 58. 429-30
154

it is
[XV
whether
clear
discovered or undiscovered.
The second
point he makes
when he deals with men as ends in themselves. Perhaps his emphasis on the relatively external side of the action is due to the fact that he has in mind a contract which can be enforced by State law, since perfect duties to others were traditionally regarded as falling under the law of the State, and the State must necessarily take an external view of
action.
9. Suicide.
When Kant considers duties towards oneself, he does not test maxims their fitness to produce a systematic harmony of purposes among men if they were to become universal laws of nature. He does, however, test
by
them by reference to harmony of purpose, a harmony between the ends proposed by the maxim, when universalised as a law of nature, and what he calls 'purposes of nature'. 1 In the case of suicide, when Kant says that the 'determination' 2 (Bestimmung) of self-love is the furtherance of life, he means that this is its purpose or function and not merely its effect. If I conceive myself as having created man and given him self-love with this end in view, can I will, or even conceive, it to be a law of nature that this self-love should in certain circumstances aim at producing death? Kant's answer is 'No'; but it may be conjectured that he gives the answer because he already assumes suicide to be wrong. Why should it not be a merciful dispensation of Providence that the same instinct which ordinarily leads to life might lead to death when life offered nothing but continuous pain? This is the weakest of Kant's arguments. It might be maintained as against him that the principle of self-love, which he usually regards as a rational principle of reason at the service of desire, would be in contradiction with itself if it did not vary in its effects according as pleasure exceeded pain or vice versa. Kant here calls it a 'sensation', meaning by that presumably some sort of instinct of self-preservation. 3 But this is not one of the powers which distinguish man from the brutes, where Kant's method of argument is more plausible and unless we have an exaggerated idea of the perfection of teleology in nature, unless indeed
;
we commit
this
ourselves to
some theory of the working of divine Providence,

little
argument can carry
conviction except to those already con-
vinced.
plausible if we were to maintain that offered more pain than pleasure is at variance with the function of reason as aiming at absolute good; for it is to withdraw oneself in the interests of comfort from the duty of leading
The argument would be more

suicide only because
to
commit
life
the moral
1
life.
The
essential thing for
our present purposes
is,
however,
59. Compare the phrase 'the essential ends of humanity* in his E.g., Gr., Lectures on Ethics; see, for example, Vorlesung, Menzer, 161 ; Infield, 136.
*
430 =
Gr.,
The
a law
4
48. rational principle of self-love, when considered not as a of nature, is reduced to the level of instinct.
Gr.,
422
law of freedom but as
Compare
429
57; M.d.S., Tugendkhre,
6,
422-3. See also Chapter
XVI
7.
XV
IO]
THE LAW OF NATURE

on which the argument
is
155
not the plausibility of the particular argument, but rather the general
principle
based.
10. Culture.
In the case of developing talents particularly those talents which Kant is on stronger ground. Here his distinguish man from the brutes teleology is more explicit, and he even puts it almost in the language of religion ; for he says that a rational being must necessarily will the development of his powers, not only because they serve, but also in the second edition because they have been given, for all sorts of possible purposes. This, he holds, remains true, although there is no contradiction in
conceiving (as opposed to willing) a law of nature such that all men live the life of lotus-eaters. Apart from the theological language, if I conceive myself as having created men with all sorts of talents, I should certainly feel myself to be willing inconsistently if I willed it to be a law of nature that these talents should never be developed or used. This argument is miles apart from the argument that I ought to develop my talents because I shall find this profitable or advantageous to myself. The powers which Kant has in mind include the powers of the body, but they concern primarily the powers which distinguish men from the brutes the power of scientific thinking, the power of aesthetic apprecia1 tion, and above ail the power of leading a rational or moral life. In all his discussions of duty to oneself there is a marked strain of teleological or Aristotelian ethics, but this holds especially as regards the development and exercise of powers whereby men are distinguished from animals. Here we cannot start, as in the case of duty to others, merely with the purposes which men naturally have and proceed to ask how they can be combined in a systematic harmony. We have to argue that man has a duty to himself as a man to use his powers for the purpose inherent in themselves, and above all to develop and use the powers in virtue of which alone he is a man. This is a doctrine not fashionable today, and if it is to be reinstated it will have to be reformulated in the light of modern evolutionary theory ; but to sweep it aside as silly, or still worse as silly against an Eighteenth-Century background, is to show a lack of historical and philosophical insight.
11. Practical reason
and purpose.
Kant's doctrine becomes intelligible only when his law of nature is interpreted teleologically as concerned with the harmony of human purposes. Behind all he says there is a fundamental assumption that practical reason, and indeed that pure practical reason, is concerned with the realisation of human purposes and of a systematic harmony 2 among them. Kant puts this very clearly in the Metaphysic of Morals
:
M.d.S,, TugendUhre, Einl., VA, 386-7 =*= 231-3; M.d.S., Tugendkhre> Einl. IX, 395 242.
19,
444-6.
156
*

for
[XV
12
of men to themselves and others can be an end, that pure practical reason ; for it (that is, reason) is a power of ends as such; to be indifferent in regard to these, that is, to take no interest in them, is therefore a contradiction.' And he drives this home by adding that if practical reason were indifferent to ends, it would not determine maxims for actions (since these always involve an end) ; and therefore it would not be practical reason at all. For Kant the view that practical reason ought to aim at complete detachment and ultimate oblivion would be self-contradictory and immoral a breach of the logic of action. And it is perhaps fair to say
What in the relation

an end
is
is anything in his doctrine that willing is as rational as then for the understanding of his doctrine we must adopt the thinking, standpoint, not of mere observers, but of agents or in his own language we must be, not world-observers, but world citizens. 1 You can no more judge moral action without at least trying to act morally yourself than you can judge thought without thinking yourself.
that if there
12.
The principles of moral

see
action.
We may now
of Universal
actions.
First of all the
more
clearly the
way
in
Law and
of the
Law
of Nature
may
which the two Formulae together be a guide to
as the sole guide to

tively, or, as
good man must set aside the principle of self-interest conduct and must subordinate it to a wider impersonal
it,
the principle of acting reasonably and objecacting on a law valid for all rational agents. This is a fundamental change of attitude in the will, the adoption of a new spirit, the very essence and principle of the moral life. It is expressed in the Formula of Universal Law. Kant may hold, perhaps too easily, that a will which adopts this attitude can by its own nature find its way clear before it in actual living in somewhat the same way as does the artist who gives himself to art or the thinker who sets aside prejudice and gives himself unreservedly to the dispassionate and free working of his own thought. Nevertheless Kant recognises that the empty form of law must be filled, and that the purposive actions of men must be brought under the law by a consideration of the principles or maxims at work in them. In order to do this we must imagine these maxims to become laws of nature as a result of our volition. We can test them so far as they affect others, by considering how far the universal adoption of these maxims would further, or fail to further, or would actually destroy, a systematic harmony of purposes among men. We can test them so far as they affect ourselves by considering whether as universal laws of nature they would further, or fail to further, or would actually destroy, a systematic harmony of purpose in the individual, it being assumed that his powers have a natural purpose which can be recognised and that these powers, and especially the powers which are the differentia of man,
and impartial principle Kant puts
XV,
ii,Nr. 1170.
XV APP.
l]
THE LAW OF NATURE
157
must be furthered and not destroyed, if this systematic harmony of purpose is to be realised. All this is the mere bare bones of Kant's view ; but unless we grasp this elementary anatomy, we shall never begin to understand the body of
his doctrine.
harmony, not merely because of his own desires, however generous, but because he was at the same time endeavouring to obey a law valid for all rational agents. This must be, not his only, but his controlling and overruling, motive. If the moral law is included in the concept of the
systematic
test of moral action, not its essence. You might have such a in a colony of ants, but there would be no morality there. You might conceivably have it, or something like it, in a society governed entirely by self-love as a result of action and reaction, but there would be no morality there. Nor would there be morality even if everybody in such a society aimed at this harmony of purposes on the ground that he happened to like it or that it would make things more comfortable for himself. Morality would be embodied in such a systematic harmony of purpose only so far as each member of the society sought to further this
is
The test of the maxims is thus speaking broadly a systematic harmony of purpose such as might accord with a law of nature. But this
only a
harmony
harmony of
purposes,
then
it
goes without saying
this
concept and the idea that we can help to bring about the existence of such a harmony is at the same time the determining ground of a pure moral
will. 1
that not even in the case of perfect duties is Kant test. The reason why certain maxims intellectual a purely applying cannot be conceived as laws of nature is that to will in accordance with such maxims is not merely to fail in furthering a systematic harmony of purposes, but it is to disrupt and destroy that systematic harmony which not only is essential to the concept of a teleological law of nature, but is also the outward, if ideal, manifestation of obedience to the moral law.
It
may be added
APPENDIX
THE LAW OF NATURE AS A TYPE OF THE
MORAL LAW
i.
The form of law.
IN the Groundwork Kant gives us hardly any explanation of his reasons for introducing Formula la as a supplement to Formula I. In Formula I we are told that our maxims ought to conform to the universality of law as such. Kant then informs us that the universality of law constitutes the
form of nature. Nature, whatever be its matter,
1
is
characterised essentially
Compare
K.p.V.,
109-10 =
244-5
158
THE CATEGORICAL IMPERATIVE [XV APP.
by the
law
however much the laws of nature and the laws of freedom differ. On this ground Kant apparently assumes that the universal law imposed by the categorical imperative can be translated in terms of a universal law of nature; and we thus pass from Formula I to Formula la. 1 This appears to mean that there is some sort of analogy between the universal law of freedom (to which all actions ought to conform) and the universal law of nature (in accordance with which all phenomena are determined). It is in virtue of this analogy that we can substitute 'universal law of nature' for 'universal law* in our formula. This analogy is explicitly mentioned in the useful summary of the argument provided towards the end of Chapter II of the Groundwork. 2 A view of this kind manifestly requires considerable expansion, which
fact that in it all events take place in accordance with universal particularly the law of cause and effect. Nature and moral action thus in a sense have the same form the form, namely, of universal law
Kant
avoids in the Groundwork, perhaps as being too technical.
2.
The problem of
9
'*
'exhibition*'.
expressed more technically in the Critique of the problem of applying a concept to individual 4 objects or of 'exhibiting' an object corresponding to a concept. By means of our concepts we can know objects only if we can exhibit the corresponding objects in some sort of sensuous intuition. When we exhibit the objects of a concept, we are often said to exhibit the concept itself.
Kant's problem
It
is
Practical Reason.
is
There is no difficulty in exhibiting objects for our empirical concepts. These concepts are abstracted from sensuous experience, and we can reconstruct objects for them in imagination the imagination in so doing follows very much the same procedure as it follows in combining sensations when we are actually seeing the object. Imagination can also construct objects for our arbitrary mathematical concepts in pure intuition, that the expression is not Kant's we can in imagination is, to speak roughly carve them out of space. 5 In being aware of the imaginative procedure necessary for each concept we have what Kant calls a 'schema' of the 6 concept. If a sensed or imagined object fits the schema, it falls under the
concept.
There is more difficulty in exhibiting objects for a pure category of the understanding, like ground and consequent. Nevertheless, since in order to have knowledge of empirical objects we must combine them in
one consciousness, we have in knowing objects to combine them 1 = 66. Gr., 421 = 47. The same transition is apparently made also in Gr., 436
2
to universal
67. Compare also Gr., 431 = 60, where the conformity of moral actions 437 law is said to be similar to an order of nature. 8 1 K.p.V., 67 ff. 189 ff.9 ( 19 ff.). * in Latin 'exhibere*. *Darlegm or 'darstellen * In K.U., 62, 365 ( as 276), a mathematical object is said to be possible only by
Gr,,
=
9
'determining* space.
*
See also Chapter XIV, Appendix
2.
XVAPP.
3]
THE LAW OF NATURE
159
one time and space. This means, according to Kant, that we must combine or at least seek to combine sensuous intuitions in certain ways, and the necessary procedure of imagination here gives us what are called 'transcendental schemata corresponding to the different pure categories. In particular it gives us the transcendental schema of necessary succession in accordance with a rule; and where this is present, as it must be if we are to be aware of objective change, we are entitled to apply the pure category of ground and consequent, which thus becomes the category of cause and effect. Necessary succession in accordance with a rule is the transcendental schema in virtue of which we can exhibit objects falling under the category of the understanding. 1 The difficulty about our concept of the unconditioned and absolute law of morality is that it is an Idea of reason and therefore ex hypothesi 2 it can have no corresponding object in sensuous experience. The actions which we wish to bring under the moral law are from one point of view mere events subject to the law of nature and not to the law of freedom. They cannot be adequate to the Idea of an unconditioned law, and we have no schema, transcendental or otherwise, whereby we can exhibit an object for such an Idea of reason. 3
imaginatively in
5
:
3.
Symbolic exhibition.
X, where X is the unknown something in it cannot be like any human inclination. A father's love for his child is therefore a symbol whereby through an 6 analogy we can represent to ourselves God's love for human beings. Kant holds all our knowledge of God or at least our way of conceiving God to be symbolic in this sense. If we ascribe to Him human
as
There is, however, a second kind of exhibition possible, exhibition, not directly by means of schemata, but indirectly by means of symbols, and this is the method by which we supply sensuous intuitions for Ideas of reason, although no sensuous intuition can be adequate. Symbols are not objects falling directly under a concept, but they are objects which we use in order to attach some sort of meaning to a concept. They do so in virtue of an analogy. 4 In this they differ from mere arbitrary signs, such as the letters of the alphabet. An analogy, according to Kant, expresses, not an imperfect likeness of two things, but a perfect likeness of two relations between things wholly unlike. Thus A (promotion of a child's happiness) is to B (a father's love)
C (human
welfare)
is
to
God which we call love,
although
human understanding and human will, we are into falling anthropomorphism. Nevertheless by using these human characteristics as symbols, although we cannot know God as He is, we can conceive His relations to us in terms of relations with which we are
inclinations, or even
* See also Chapter IX, Appendix 3. See Chapter IX, Appendk K.r.V., A66s = 8693. **./., 59, 351 ff. ( 6706 255 if.). Compare alio K.r.V., A6y8 87241!.
5.
Compare
Pro/.,
58, 357.
l6o
familiar;
invisible

1
[XV
APP. 4
and this may be sufficient for purposes of action. For man the must be represented by something visible or sensible; and indeed for purposes of action it must, as it were, be made sensible by means of an analogy. 2 Religious rites and sacraments can be justified as
such symbols. When they are practised as having value in themselves, 3 they become mere superstitions. Is it possible for us to exhibit actions as objects falling under the Idea of moral law, if we make use of the symbolic or indirect method of exhibition by means of an analogy?
4.
The
'type*
of moral law.
I have already pointed out4 that there is an analogy between the universal moral law and the universal law of nature. The preceding section suggests that it would be more correct to speak of an analogy between the two sets of objects conceived to fall under the two laws, moral wills on the one hand and temporal events on the other; and this is confirmed to this extent that Kant speaks of an analogy between the 6 need not, however, kingdom of ends and the kingdom of nature. spend time on subtleties, though we could wish that Kant had articulated his doctrine more fully. The two types of law play similar parts in two different systems, and we may say that there is a kind of analogy between them. This relation between the two laws Kant expresses by 6 saying that the law of nature can be called a 'type' of the moral law. The word 'type* is commonly used in theology in more or less the same way as Kant uses the word 'symbol it is that by which something is symbolised or figured. Thus the people of Israel are said to be a type of God's people, and the Paschal lamb is said to be a type of Christ. Kant's application of the word to the law of nature is a natural extension
We
of this usage.
The law of nature, as a concept of the understanding, has always a schema or schemata in virtue of which it applies to sensible objects, the schema of causal law in particular being necessary succession in accordance with a rule. The law of freedom, as an Idea of reason, can have no schema whereby we can exhibit objects for it directly in intuition. Kant's suggestion is that we can exhibit objects for it indirectly or symbolically: in virtue of the analogy between the moral law and the law of nature,
1 K.U. 9 59, 352-3 ( 82 n.). For 257-8); ProL, 58, 358; Religion, 65 n. ( beauty as a symbol of morality see also Chapter XIV, Appendix 2, 1 8 4 i above. Religion, 192 ( 299). Religion, 192 ff. ( 299 ff.) f Gr., 439 SB 69. His earlier statement of the analogy in Gr., 437 67 is extremely elaborate, but it is nearer asserting an analogy between the two laws than between the sets of objects conceived to fall under them. K.p.V., 69 192 ( 193 ( 124), that we can 122). He also says, K.p.V., 70 use the nature of the sensible world as the type of an intelligent nature, but he means here naturaformaliter spectata, which in turn is the form of law as such. It may also be relevant that he speaks of 'exhibition* as a hypotyposis ; K.U., 59, 352 ( 255). iiroTvrraxns Latin, adumbratioi& a kind of sketch or outline. Compare Aristotle, Eth. JVtc. 1098 a a i and Plato, Timaeus, 76
XVAPP.
5]
THELAWOFNATURE
l6l
which share the common form of universality, we can treat events governed by the law of nature as symbols for objects conceived to fall under the law of freedom. 1 If we can do this, we use the law of nature as a type of the moral law. It must be remembered that in all this Kant has in mind a contrast between the sensible world and the intelligible world, the world of phenomena and the world of noumena, the world of appearances and the world of things-in-themselves. Furthermore, it is no accident that the Idea of the moral law has to be applied through the law of nature and the categories of the understanding. While understanding with its concepts is always directed through schemata to sensible objects, reason, alike in its logical and in its transcendental use, is always directed to the concepts of the understanding, and only through them to sensible objects. 2 There is always more in Kant's proceedings than meets the eye but without going
;
into the full implications of his doctrine it is possible to see that events in nature might be used to symbolise moral actions, and that the order in nature might be used to symbolise the moral order, in somewhat the same way as the love of a parent for his child might be used to symbolise the love of God for men.
5.
The natural
order.
In modern times we may feel more compunction in using the natural order as a symbol of the moral order. This is not only, as I suggested 3 above, because the impact of Darwinism dealt a shattering blow to Eighteenth- Century optimism: it is also because modern physics leaves a different emotional impression on the modern mind than that conveyed by the discoveries of Newton to his contemporaries and successors. We have become acutely conscious especially since the famous Romanes lecture of T. H. Huxley that the order of nature is compatible with the wildest convulsions, with the collision of stars and the annihilation of worlds. We believe that our own planet must some day become destitute of life and mind, and that the very universe as a whole may be gradually running down. All this has made us less ready to see in nature the working of a divine purpose or any concern for the ideals of men. It is put personally and bitterly by a modern poet
:
Tor
nature, heartless, witless nature,
know What stranger's feet may find the meadow And trespass there and go,
Will neither care nor
Nor
ask amid the dews of morning If they are mine or no.'
men
1 8
In contrast with this the discoveries of Newton seem to have left with a feeling of the order and harmony of the universe as the
K.p.V., 69
Chapter
= 192 XV 5.
122).
"See
>.?.,
A$o*
3359; A664-6
162
THE CATEGORICAL IMPERATIVE [XV ApP.
manifestation of a divine wisdom. This
hymn of Addison
is expressed in the well-known 'The spacious firmament on high' with its interesting denial of the music of the spheres, perhaps on the ground of the unreality
:
of secondary qualities
'What though in solemn silence all Move round the dark terrestrial ball? What though no real voice, nor sound, Amidst their radiant orbs be found?
In Reason's ear they
all rejoice,
;
And
For
utter forth a glorious voice ever singing, as they shine,
"The hand
that
made
us
is
divine."
'
have only to turn to the too much quoted passage about the starry heavens and the moral law in order to see that Kant, in spite of his ruthless insistence upon causal law as the only constitutive law of
We
was deeply affected by this view of the physical universe; and indeed that the two laws, the natural and the moral, however much he might insist on their difference, were closely connected in his emotional experience. This is brought out by the way he couples together Newton and Rousseau Rousseau who had first opened his eyes to the true nature of moral value. 'Newton was the first to see order and regularity combined with simplicity and since then comets move in geometrical paths. the deeply hidden nature of man and Rousseau was the first to discover the concealed law in accordance with which Providence is justified
nature,
. .
. .
. .
through
his observations
since
It is in the light of this attitude,
Newton and Rousseau God is justified.' 1 common to Kant and to many of
symbol for the moral order is to be understood. In the light of this attitude we can also understand more easily how Kant can pass, without any statement of his grounds, from the formal law of nature to teleological laws of human nature as a basis for the Tightness or wrongness of our maxims. He is justified in doing this, as I have suggested, because he is dealing with human maxims and human character, which must be regarded as purposive. But this legitimate appeal to a systematic harmony of human purposes he may have found all the easier because he tends to regard the order of nature as itself a systematic harmony, and perhaps even a systematic harmony of purpose, however much he may reject the claim of such beliefs to be treated as scientific knowledge. 2
6. Practical exhibition.
his contemporaries, that the use of the natural order as a
The theory of the use of symbols as an indirect method of exhibiting concepts and the emotional attitude of the period to the law of nature must be taken into account in understanding Kant's doctrine. But in
1
Fragmnte, Phil. Bib., VIII, p. 329. This note is probably rather early. Sec, for example, JT.r.K, A6go ff. 6719-20. 6649 ff.; ASgi-a
XVAPP.
6]
THE LAW OF NATURE

we
163
are not concerned with the theoretical use of symbols, and still less are we concerned with the symbolic expression of emotion. Our concern is with action 'Act as if the maxim of your action were to become through your will a universal law of nature.' This is the test of the maxims to be adopted or rejected by a man whose motive is always, at the same time, to obey a universal moral law valid for all rational agents. The problem is to determine for purposes of action which maxims are to be regarded as moral or immoral indeed it is to exhibit in action itself, and not merely in theory, examples or instances of obedience to the moral law. The problem is one of practical, and not theoretical, exhibition. 1 Kant's solution is that the actions which can and ought to be willed
dealing with the categorical imperative
in obedience to moral law (or 'exhibited' symbolically in practice as instances of such obedience) are those whose maxims, if conceived as a law of nature, would further a systematic harmony of purposes among men, or at least would do nothing to destroy such a systematic harmony. This is a very broad statement: it omits the different ways in which systematic harmony is to be understood according as we are dealing with duties to self or duties to others ; it obviously requires a great deal of working out in detail in order to carry full conviction ; but in itself it is at least not unreasonable, and it is very far from being the kind of nonsense commonly attributed to Kant. Indeed it may well be doubted whether
it is
possible to
is
basis,
even although
in
work out a systematic moral philosophy on any other we may hold that Kant's own attempt to work it
out
many
what is done. The morally good man must have another motive than merely the desire to produce certain results, even such a desirable result as general happiness or a systematic harmony of purpose among men. We also get rid of mystical theories which take as a schema what is merely a symbol and suppose that the moral motive springs from non-sensuous intuitions of an invisible kingdom of God. The danger from mystical theories is less, partly because these are compatible with the purity of the moral law, and partly because they are contrary to the natural thinking of mankind. Empirical theories, on the other hand, destroy
to
1
an action aiming at a systematic harmony of purposes is not thereby necessarily moral. In his technical language such actions are symbols, not examples, of the moral law; and so far as empirical or scientific knowledge is concerned, they can be nothing more. In order to have genuine goodness these actions must be willed, at the same time, for the sake of the moral law itself. When this motive is present and we cannot be sure that it is we can in action produce examples which are, at least approximately, examples of obedience to the moral law and so objects which in this case alone can be said to be concrete manifestations, however inadequate, of an Idea of reason. In this way, according to Kant, we avoid empirical theories of moral philosophy, which suppose, falsely, that actions are good merely according
Furthermore, Kant
respects faulty. is right in saying that
K.p.V.> 71
194
125).
164
THE CATEGORICAL IMPERATIVE [XV ApP.
morality at its roots; for they substitute for duty a merely empirical 1 interest based ultimately upon inclination. It should be noted further that for Kant, here as always, obligation is
and goodness original. He starts from the notion of a good obey the moral law as such, and he tells us the kinds of action which would necessarily be performed by such a good and rational will. If we are so irrational as to be tempted to act otherwise, these kinds of action must appear to us as duties, but this is possible only because of the presence of a good will in us and we cannot say we are acting morally except in so far as we can say that we have performed our duties
derivative
will seeking to
;
for the sake of duty or for the sake of moral
law
as such.
*/r./>.F.,
70-1
193-4
124-6).
CHAPTER XVI
THE FORMULA OF THE END IN ITSELF

i.
Formula
II.
other,
'So act as to use humanity, both in your own person and in the person 91 always at the same time as an end, never simply as a means.
of
every
This new formula may be said to enjoin respect for personality as such. This is necessary to supplement the first formula, which forbids us to be respecters of persons in the sense of discriminating arbitrarily and unfairly between them. The new formula, like the old, should, once it is understood, receive the immediate approval of ordinary moral
judgement.
Strictly speaking, this formula, like all others, should cover rational beings as such; but since the only rational beings with whom we are acquainted are men, we are bidden to respect men as men, or men as rational beings. This is implied in the use of the term 'humanity* the essential human characteristic of possessing reason, and in particular of possessing a rational will. It is in virtue of this characteristic that we are bound to treat ourselves and others, never simply as a means, but always at the same time as ends. The words 'at the same time' and 'simply' must not be overlooked they are absolutely necessary to Kant's statement. Every time we post a letter, we use post-office officials as a means, but we do not use them simply as a means. What we expect of them we believe to be in accordance with their own will, and indeed to be in accordance with their duty. Considerations of this kind do not arise in regard to the stamp which we stick on our letter or the post-box to which we entrust it they arise only in regard to persons and not to things. So far as we limit our actions by such considerations, we are treating persons 'at the same time' as an end, though we may also be using them as a means. The formula applies to the agent's treatment of himself as well as of others. Some thinkers hold that all duties are social, and that even our duties to ourselves are duties to a society of which we are members. There are others who hold that all duties are personal, and that if only we respect our own personality, our duties to others will immediately follow. Kant takes the middle path between these one-sided views, but he holds 2 that we could have no duty to others unless we had a duty to ourselves. as means' Kant has It should be added that by 'using persons merely in mind the using of them as means to the satisfaction of inclination or
: :
Gr. 9
429
= 57.
M.d.S., Tugendlehre>
2,
417-18.
l66

on
inclination.
1
[XVI
to the attainment of ends based
To
sacrifice one's life in
the performance of duty is not to use oneself merely as a means. As we have seen, 2 it is an error to regard even a good action as a means to the realisation of the law : it is a realisation of the law. It would be a still greater error to suppose that in dying for the sake of duty I was using my rational will as a means to a further and merely relative end. Moral sacrifice is a problem not to be dealt with lightly, but at least we can say that sacrifice in the way of duty is not the use of my rational will as a means to the satisfaction of inclination it is rather the subordination of my inclinations, and even of my life, to an end whose value is incomparable with that of any sensuous satisfaction the manifestation of a good will
:
as
an end in
itself.
2.
The
nature
of ends.
If
we
are to understand Kant's formula,
we must
consider what
is
in so doing we must look more closely at an aspect of the will hitherto taken for granted. One essential characteristic of a will we have already discussed at length a rational agent wills or
meant by an end; and
pursue and
kill. Man knows what he is doing in purposive action, and this means that we must have a concept of the object to be produced, although naturally images may, and perhaps must, be present as well. Hence Kant says also that an end is the object of a concept so far as the concept is 7 regarded as the cause of the object (the real ground of its possibility). Perhaps we may put this otherwise by saying that here, as always, our
is expressly stated in Gr., 436 = 66. Chapter VII 3. 8 I have discussed this problem in The Good Will, pp. 397 ff. 4 Gr., 427 sss 55 and 412 = 36. See also Chapter VIII 5. 6 Gr., 437 = 67; K.p.V., 58-9= 178 ( = 103); M.d.S. 9 Tugendkhre, 385 = 229; Einl. VIII, 392 = 238. 6 M.d.S. = 224; Einl. Ill, 384 = 229. t Tugendlehre, Einl. I, 381 7 K.U., 10, 220 (= 32) ; Einl. IV, 180 (= xxvm).
himself to act in accordance with his conception of laws that is, in accordance with principles. 4 There is, however, another essential characteristic to be considered namely, that the willing of a rational 6 agent is always directed towards an end which he sets before himself. All willing has an end or purpose as well as a principle. The concept of 'end* is familiar to us, but it is not altogether easy to define precisely. An end is ordinarily taken to be an effect which the will seeks to produce, but it must be said further that the idea of this effect, or the idea of producing this effect, determines the will it does so most obviously in inducing us to adopt certain means towards the production of the end. On this level we may with Kant define an end as an object of a free will (Willkur), the idea of which determines the free will to an action whereby the object is produced. 6 The 'idea' of an object must here be a concept it is more than a mere image or sensum, as when the smell or sight of a rabbit induces a dog to
sets
:
This
Einl.
Ill,
XVI
3]
THE END
maxim
IN ITSELF
167
or principle, the principle of producing an object conceived as being of a certain kind. As Kant says elsewhere, ends are 1 always 'determining grounds of the will in accordance with principles'. Needless to say this does not exclude either images of the objects or desires for the object. Kant tends to speak of an end as if it could only be an effect of the the action itself, action, like a house or picture, but clearly it may also be 2 as in the case of playing games. Furthermore, where the sight or thought of an apple is the starting point of a purposive action, the end of the action is the eating of the apple, not the apple itself, so that here too it looks as if the action itself is our end. Even where our action aims at it may be that our a producing something, as in the building of house, end is not the house itself, but rather the possession of the house. But complications of this kind must here be ignored. For our present purposes, the important point is the connexion between ends and principles. We will the end freely, but in willing this in accordance with a particular maxim, so particular end we must will that the end may be said to be the ground of our maxim, and thus to be the ground which determines our will in accordance with a principle. So far we have spoken only of maxims or subjective principles; but we must add and this is the absolutely essential point that our chosen ends are the ground also of objective principles, the ground of hypothetical if he wills the end, ought to will the means; imperatives. A rational agent, are ends this sense in and grounds which determine our rational wills3 in accordance with principles which are objective, although hypothetical. This may be part of what Kant means when he defines an end most for the selfobscurely as that which serves as an objective ground 4 that ends to out on rate he At will. the of point determination goes any material ends which are arbitrary products of our will are only relative that their worth is relative to the special constitution of the agent
action has a
;
should say, they are 'good for' him) and finally that they can be the ground only of hypothetical imperatives. They cannot be the are not conditioned by the ground of categorical imperatives since these a particular effect. of the desire to fact that we happen production
(or, as I
;
3.
Ends
far
in themselves.
So
we
are
is
we have
to face
on comparatively familiar territory. The new problem the relation between ends and categorical imperatives.
There must be such a relation, since every categorical imperative enjoins 6 action and every action must have an end.
iJT.j&.F., a
8
59also
178 (=103).
i. Chapter i -2. See Chapter IX, especially he Gr,, 427 = 55. By 'objective ground'
Compare
may
possibly
_
mean
here merely a ground
in objects'.
*M.d.S., Tugendlehre, Einl.
I,
381
224; Etnl. Ill, 385
229.
l68
If this
is so,

there
[XVI
must be ends given us by mere reason itself, not by reason at the service of inclination. These ends must be valid for every rational being; they must be objective ends, not merely subjective ones; they must be absolute and not relative; they must have an absolute and not a relative worth, being good in themselves and not merely good for a particular kind of agent. To say this is to say that they must be ends in
themselves. 1
may be thought that we know the solution to this problem already. open to Kant to say that the ends enjoined by the categorical imperative are simply moral actions willed for the sake of duty, which he recognises to be good in themselves. It is also open to him to say that the effects which such actions seek to produce such as one's own natural and moral perfection and the happiness of others must by extension be regarded as objective ends, at least in those contexts where it is our duty to pursue them. In his own terminology these are ends which are
It
It
is
also duties. 2
is sound Kantian doctrine, but in his eyes it is apparently not This may be partly because the products of moral action are enough. not absolutely good, good in any and every context; and because even morally good actions, though good in themselves, are not absolutely good, good in every respect and the supreme condition of all good, such absolute 3 goodness belonging only to a good will. An action is morally good because it is the manifestation of a good
All this
and the categorical imperative in enjoining morally good action in accordance with a universal law is enjoining that a good will as such should be manifested and not thwarted by mere inclination. A good will maniwill,
is the end enjoined by the categorical imperative ; must follow from the very nature of a categorical imperative that I ought to respect the rational wills of all moral agents including myself and not subordinate them to mere inclination. Another reason for Kant's view appears to be that he is looking for an absolute end which can itself be the ground of categorical imperatives in somewhat the same way as relative ends are the ground of hypothetical imperatives. If anything could be the ground of categorical imperatives, it would fall under Kant's definition of an end; for it would be a ground determining our will in accordance with principles', which in this case would be not only objective principles, but dso categorical principles
festing itself in action
and
it
valid for every rational agent as such. 4 In using this language Kant is manifestly extending the meaning of the word 'end'. An objective and absolute end could not be a product of our will; for no mere product of our will can have absolute value. An end in itself must therefore be a self-existent end, not something to be produced by us. Since it has absolute value, we know already what it
1
Gr.,
427-8
= 55. =
M.d.S. 9 Tugendkhre, EinL III-IV, 384 ff. 229 fF. 182 (=: 109). ./. F., 62 4 It would presumably also serve as an objective ground for the self-determination of the will, according to the obscure definition in O., 427 55.
XVI
4]
THE END
1
IN ITSELF
169
must be namely, a good will. This good or rational will Kant takes to be present in every rational agent, and so in every man, however much it may be overlaid by irrationality. Hence man, and indeed every rational agent as such, must be said to exist as an end in itself, one which should never be used simply as a means to the realisation of some end whose value is merely relative. It may seem arbitrary to speak of existent things as ends, yet it is
true that the existence of persons does determine a rational will, qua rational, in a way analogous to the way in which the adoption of an end determines a rational will. Furthermore, since the existent things in question have rational wills, it is possible to further, or at least to refrain from overriding, their rational volitions. When we do this, not from mere inclination only, but also at the command of reason, we are treating rational agents, not as mere means, but also as ends in themselves, and the phrase 'end in itself becomes less inappropriate than at first sight it seems. No doubt it is difficult to make this conception theoretically precise, but we may say provisionally with Kant that the will of a rational
is not to be subjected to any purpose which cannot accord with a law which could arise from the will of the person affected himself. 3 We must distinguish sharply between this duty to respect the will of others and a mere prudential adjustment to them. We may further the inclinations and happiness of others because this satisfies our own inclinations or administers to our own happiness to do this is to use them as a means. We may also yield prudentially to their pressure or threats, and in this we are treating them like any other obstacle which we have to overcome or avoid. There is no difference in principle between running away from a bandit out of concern for our own safety and running away from an avalanche or a man-eating tiger. In prudential adjustment to others we treat them as means or as obstacles to our satisfaction, as we may treat inanimate objects and animals. This is something quite different from treating them as ends in themselves.
person
4.
Grounds and ends.
It may still seem puzzling that Kant should regard rational agents both as the objective ends enjoined by categorical imperatives and also as the grounds of such imperatives. There is no difficulty in seeing that the end of an action is also its
ground and so the ground of the maxim embodied in the action. Nor is there any difficulty in seeing that in hypothetical imperatives of skill the end chosen is the ground of these imperatives, though these imperatives enjoin only the means to the end and not the end itself. In the case of hypothetical imperatives of self-love their ground is the happiness of the agent, and it is also the end enjoined so far as we take these imperatives
1
Kant
establishes this afresh
inclinations themselves,
nature; Gr., 428

*/r./>.F.,8 7
= 56. = 2i 5 (-
by eliminating the claims of objects of inclination, and mere irrational things whose existence is a product of Compare Gr., 400 = ao.
156).
I7O
to
[XVI
be concerned, not with the means to happiness in the sense of continuous pleasure, but with the harmonisation of the ends set before us by
inclination. 1
Granted that a categorical imperative enjoins the treatment of rational agents as ends-in-themselves, in what sense can we say that rational agents are also the grounds of a categorical imperative? There seem to be three senses in which this can be said. Firstly, it
is
because rational agents exist that a categorical imperative must enjoin respect for their rational wills. Secondly, it is because rational agents exist with wills which can be thwarted or furthered in different ways
that
to
recognise particular categorical imperatives we ought not wills by fraud or violence, and we ought to further their happiness. Thirdly, it is only because rational agents exist that there can be such a thing as a categorical imperative at all. Because agents are rational, their will necessarily manifests itself in universal laws. Because they are imperfectly rational, these universal laws must appear as categorical imperatives. The categorical imperative has its ground in the 2 will of rational agents who are not completely rational. It is the first two senses that Kant presumably has in mind when he speaks of man as the ground of a possible categorical imperative and of particular categorical imperatives. But the ultimate reason why man can be such a ground is because he is a ground in the third sense it is because the categorical imperative has its origin in his rational will that his rational will ought not to be subordinated to any meaner end but is itself an end which the categorical imperative must bid us to further and not to thwart.
we must
thwart their
5.
The approach
to
Formula H.
Kant's approach to Formula II is over-subtle, and there are four strands to be distinguished in his argument. There is first of all the argument from the essence of a categorical imperative, the ends which it imposes, and the grounds on which it rests. This is the argument we have just considered. It is supplemented by a second argument based on the way in which rational agents as such must conceive themselves namely, as agents capable of acting in accordance with rational laws of freedom and therefore not to be subjected as a mere means to the satisfaction of inclination. 3 In the final summary we have a third argument starting from the conception of an absolutely good will and its objects, instead of from the essence of a categorical imperative, but otherwise following in
main
1 1 See Chapter VIII 8. They contain 'the rule of prudence in the choice of ends ; M.d.S. y Tugendlehre, Einl. Ill, 385 230. 8 This accords with Kant's general view that while a priori principles cannot have their ground in objects, they can have their ground and origin in reason ; K.r.V., 148-9 Bi88. Similarly categorical imperatives cannot, like hypothetical ones, have their ground in objects, but they can have their ground and origin in practical reason. An imperative so grounded must enjoin respect for rational wills as ends, and rational wills thereby become grounds in a derivative sense. * Gr., 429 57 as supplemented by Gr, 9 447-8 = 80- 1
XVI
6]
THE END
same
line as the first
IN ITSELF
1
l*]t
brief the
argument. This is followed by a fourth argument perhaps the easiest of all which maintains that Formula II is already implicit in Formula I, if we consider action on its purposive
side. 2
This multiplicity of arguments

are
is
somewhat complicated,
risk
I reserve
typical of Kant's method, but, as they them for treatment in an appendix.
of over-simplification we may say that they all rest on one the principle, principle that a good will has a unique and absolute value. Granted that this is so, it must be wrong to subordinate it as a mere means to any end of lesser value, such as the satisfaction of personal inclinations. It must indeed be a duty, not merely to refrain from thwarting its manifestations in action, but also to further these manifestations so far as it is in our power to do so. This is a principle on which any rational agent, whose reason had full control over passion, would necessarily act,
At the
and it is one on which he ought to act if he is irrational enough to be tempted to do otherwise. There is a further point. As Kant well knows, men are not saints, and for this reason it may at times be necessary to refrain from furthering their ends and even to thwart their wills. This must introduce qualifications into applications of our principle, especially when we have to consider the function of the State. Nevertheless and this is a fundamental conviction of Kant a good will is present in every man, however much it may be overlaid by selfishness, and however little it may be manifested in action. 3 Because of this he is still entitled to respect and is not to be treated as a mere instrument or a mere thing. 4 As a being capable of moral action, a man, however degraded, has still an infinite potential value; and his freedom to work out his own salvation in his own way must not be restricted except in so far as it impinges on the like freedom
of others. We shall never understand Kant aright unless we see him as the apostle of human freedom and the champion of the common man.
6.
Kinds of duty.
which
Like Formula I, Formula II is a supreme practical principle from all other laws of the will may be derived. 6 In the light of our discussion it is easy to see how this principle gives rise to the distinction between perfect and imperfect duties. So far as we take the principle negatively it forbids us to use rational agents merely as a means and so to override the rational wills of moral agents in order merely to satisfy our own inclinations. This is the basis of perfect duties, and it forbids such wrongs as murder, violence, and fraud, as also suicide and lying. It lies at the root of Kant's philosophy of legal obligation. But
1
67-8. 437 See also Chapter

Gr.,
Gr.,
1 1.
437-^
= 68.
Kant
See also
holds,
Gr.,
431
= 59.
XXV,
2, 10,
and
4 M.d.S. t Tugendlehre, 39, 463. For subjected to shameful punishments.

*
this,
reason,
men
should never be
Gr.,
429
= 57.
172

also take
:
[XVI
our principle positively it bids us to act on the maxim of furthering the ends of rational agents. Here, it must be remembered, there is a place or 'play-room' for discretion. This positive interpretation
we must
is
for
Kant
the basis of positive
and
ethical, as
opposed to
legal, obliga-
tions.
We transgress perfect duties by treating any person merely as a means. We transgress imperfect duties by failing to treat a person as an end, even
though we do not actively treat him merely as a means. The difference between duties to self and duties to others is commonly recognised, though in some respects it is not easy to account for such a
Kant pushes this difference very far in regard to imperfect by insisting that our duty to ourselves is to seek as an end our own natural and moral perfection but not our happiness; and that our duty to others is to seek as an end their happiness but not their perfection. It may, however, be our duty to seek our own happiness as a means to our moral welfare and it may be our duty to seek the moral welfare of others in a negative sense. We ought not to tempt them to courses of action which 1 might be likely to cause them pangs of remorse.
difference.
duties
7.
Kant's
illustrations.
As might be expected in the case of a formula dealing specifically with ends, Kant's illustrations 2 bring out more clearly and in some
more satisfactorily the teleological implications which we saw in the previous chapter to be involved in his appeal to a universal law of nature. 1. Suicide. The man who commits suicide because the disagreeable prospects of life seem to overbalance the agreeable ones is making pleasure and the avoidance of pain his final end and in him practical reason, which is capable of realising absolute moral worth, is being subordinated as a mere means to the relative end of avoiding discomfort. If apart from all questions of duty to others there can be a right to commit suicide, this can be justified only on the ground that there is no longer any possibility of living a moral life and manifesting moral worth. Such cases may arise when pain is unendurable or insanity certain. Suicide cannot be justified lightly on the ground of mere discomfort, however gently we may judge those whose misfortunes have brought them to so desperate a state of
cases
;
interpret it with undue rigidity. 2. Promises to repay loans. In extracting money from others by false promises it is obvious that we are using them as mere means and not as ends. As Kant says, this is still more obvious in crimes of violence. What is interesting is Kant's view that to treat others as ends in themselves is to treat them in such a way that their rational wills can be in agreement with ours and that they must be able 'to contain in themselves' the end
mind. Kant's principle do not say that he does
is
fundamentally sound even
if
he
may
and
of our action towards them.

1
M.d.S. 9 Tugendlehre> Einl. VIII, 394
240.
Gr. }
429-30
= 57-9.
XVI
8]
THE END
As
IN ITSELF
173
to the imperfect duty of developing one's talents, failure not actively to use oneself only as a means we are still maintaining humanity in ourselves, and not destroying it, but we are failing actively to further humanity in ourselves as an end. Here again
3. Culture. to do this is
:
bids us assume that it is the purpose of nature to develop humanity that is, to develop in us especially those powers which distinguish us from the brutes. 1 In this argument 'humanity' covers, not only our rational will, but all our rational powers as manifested in art and science, and indeed our bodily powers so far as these are necessary for the leading
in us
Kant
of a
human
life.
In the case of imperfect duties towards others here 4. the duties primarily of kindness or benevolence Kant recognises that the natural end of man (which is to be distinguished from the end or 3 purpose of nature) is happiness. If we are to harmonise positively, and
Kindness.
not merely negatively, with humanity in others, as an end in itself -that is, if we are to further it actively, and not merely to refrain from violating it we rnust make this natural end of others, as far as possible, our own. This means that we must, as far as possible, further the relative ends of others. To treat other men as ends in themselves, if this idea is to have
its full effect,
must be
to
make
their ends, their relative
and personal
ends,
as far as possible
our own.
'as far as possible' presumably indicates two things: that it is not within our powers to further the ends of all men equally, and all that is enjoined is the adoption of a maxim which must be applied with discretion ; and (2) that we ought to further the ends of others only so far as they are not manifestly foolish or incompatible with the moral law.
(i)
The phrase
8.
The soundness of KanCs
view.
It may be objected to Kant's principle that having first of all told us that a rational agent ought to act on the principles on which every rational agent would act (if reason had full control over passion), he is now telling us that every rational agent ought to be treated in the way every rational agent would will to be treated (if reason had full control over our passion). To tell us this is to tell us precisely
nothing.
Such a view
in the
first
place complains that
when Kant
sets
out to
state the form of moral action, he does state the form of moral action, and not its matter. It is hard to see why he alone among philosophers should
be blamed for being consistent. If we are not interested in the form of moral action, there is no reason why we should study Kant's moral
philosophy. In the second place such a view forgets that
1
Kant
is
trying to state
also Gr., 438-9 58-9. M.d.S., Tugendlehre, Einl VIII, 391-2
Compare
= 237-8;
19,
444-5.
74
[XVI 8
the form of moral action, the supreme principle of a rational wilt. He is trying to formulate clearly the principle upon which, however vaguely envisaged, good men have acted, and upon which we ought to act. If we consider his principle from the point of view of agents, and not of logicians, we shall find that it throws a flood of light upon what we ought to do and what we ought not to do. The difference between the good man and the average sensual man is surely that the former recognises the infinite and unique value of all moral agents and treats them in accordance with this value, while the latter does not. The adoption of this principle in action constitutes a moral revolution in the soul of man, as does the adoption of the Formula of Universal Law. 1 It may be said that this is a mere excuse for vagueness of thought, the substitution of an emotional attitude for the clear analysis incumbent on a philosopher. What is the use of telling us to adopt an attitude, a spirit, a principle of action, if you don't know and can't say what the
principle is?
The answer is that we do know what the principle is, although, like many other things which we know, it may be extremely difficult to formulate with precision. Kant has done his best to formulate it, and it
If the complaint formulation does not contain in itself the criterion of its own application, this is a perfectly outrageous demand to make. There are no moral principles, there are not even moral laws or moral rules, which can be applied mechanically or by any method of logical deduction without practical judgement and moral insight. If Kant thought that there were and I do not believe he did he may be blamed for so thinking ; but we cannot both blame him and at the same time accept this thought as the basis of our own criticisms. What we can say is that there is a vast field open for discussion as to the way in which Kant himself applies his principles. This, however, cannot be done without a careful examination of his Metaphysic of Morals. In the Groundwork he attempts no more than to show by way of illustration how certain types of accepted virtues and vices fall under his principle. In this he succeeds so far as it is possible to succeed without entering into an elaborate discussion of various kinds of circumstances which might form the context of action. And once more he has shown clearly enough that the natural desires and powers and purposes of men constitute, on his view, the matter which has to be organised within the framework of
may be doubted whether anyone has formulated it better.

about lack of precision means that
his
moral
must indeed recognise that the systematic application of Kant's 2 principles will offer more difficulty than he anticipates, and that it may not be easy to formulate, let alone to justify, the further presuppositions on which his exposition rests. But these considerations belong to another
enquiry.
We
principle.
See Chapter'XV
12.
Gr.,
392
9.
XVI
9.
9]
THE END
IN ITSELF
175
Special characteristics of Formula II.
It should be observed that the Formula of the End in Itself has no subordinate formula to be used in its application, whereas the Formula of Universal Law has a subordinate formula (Formula la), in which the law of nature is used as a 'type' for the application of the moral law. This difference is more apparent than real. The Idea of 'humanity which, strictly speaking, ought to be the Idea of 'personality' or of a 'rational will' is applied through the concept of 'man' as a rational animal with distinctive powers and personal desires and needs and the ideal systematic harmony of human purposes is the test of moral action even more clearly
9
it should also be noted that Kant speaks of the that those affected by my action should themselves 'contain' possibility the end of my action ; and says also that I should make the ends of other rational agents, as far as possible, also my own. Here too we have a conception which requires a lot of working out. There is clearly more in question than mere approval; but it is not easy to express precisely the sense in which different agents may have the same ends so far as their actions are willed in accordance with the same universal law. The Formula of the End in Itself is in Kant's mind closely connected with freedom. Men can compel me to perform actions which are directed towards certain ends, but they cannot compel me to adopt any end as my own. 1 This is why ethics proper is concerned with purposes (and so with motives), while the theory of law is concerned only with external
than it was before. In this connexion
actions. This is also one reason why Kant can pass from this present formula to the Formula of Autonomy (Formula III). If Formula I is supposed to deal with maxims in relative isolation though this may be doubted Formula II marks a real step forward in bidding us consider individual persons as wholes. It must, however, be itself inadequate, unless we consider the individual agent more explicitly as a member of society. This Kant proceeds to do in Formula Ilia, but he does so by first passing through Formula III.
APPENDIX
ARGUMENTS
The main strands
i.
IN
SUPPORT OF FORMULA
of the
II
:
of Kant's argument
the essence
may be worked out as follows
Argument from
First of all there
imperative We may put

1
itself,
it
is an argument from the essence of the categorical the argument examined in Chapter XVI, 3-4 above. here summarily for the sake of completeness.
I.
M.d,S., Tugendkhn, Einl.
381 == 224.
176
If there
is
THE CATEGORICAL
M P E R A TI VE [XVI ApP.
a categorical imperative, it must enjoin upon us objective ends. Since these ends must have absolute worth, they cannot be the relative ends which we seek to produce: they must be rational agents or, for practical purposes, men. Without this there could be no absolute worth and so no categorical imperative. Hence the categorical imperative must bid us treat men as absolute ends or as ends in themselves. 1 Worked in with this conclusion is the view that men, considered as ends in themselves, are the ground^ both of a possible categorical imperative (that is, of Formula II), and also of particular categorical
and absolute
imperatives.
2.
Argument from
the nature
of rational
agents.
In a different sense of 'ground' the ground of our present formula is be 'Rational nature exists as an end in itself.' 3 Kant seems to take an independent starting point; and it is supported by a new and obscure argument, which may be called the argument from the nature of rational agents as such. Every man, he says, necessarily regards his own existence as an end in itself; and since every other man does the same as regards his existence on the same rational ground which is valid also for me^this is not merely a subjective principle, but also an objective principle valid for every one it is, in fact, the basis or ground of our present
said to this as
:
formula.
is too brief to be intelligible, and it looks like a fallacy worthy of Mill himself. Kant might be taken as arguing that because Stuart John each man takes a self-centred interest in his own welfare, therefore all men ought to take an interest in the welfare of all. We must, however, note Kant's assertion that the ground of the judgement which other men make about their own existence is 'the same rational ground which is valid also for me'. For the justification of this assertion we are referred, rather vaguely, to Chapter III of the Groundwork. I take him to mean by this the section on the necessary presupposition of freedom, which begins by saying that it is not enough to ascribe freedom to our own rational
This
will
on whatever ground: we must show
it
to
rational agent as such. 4 If the passage about freedom is the key to the present passage, Kant means that when I regard myself as an end in myself, I am regarding myself as a moral agent subject to moral law and so of infinite value. I must do this in virtue of nature as a rational agent, and so must every other man. Hence this principle is an objective principle valid for all
be presupposed by every
my
rational agents as such

1
and applying
Gr.,
to ail rational agents as such. In

*
Gr.,
Gr. 9
428-9 = 55-7. 429 a* 57. See also
428 = 56.
Gr. 9
Or.,
427-8 447-8
= 55 and 428-9
80-1.
56-7,
XVI
APP.
3]
THE END
IN ITSELF
177
and treat all rational nature as such I must regard 1 persons (including myself) as moral agents. If this is the correct interpretation, it is similar to Kant's insistence that every man regards himself as subject to moral law, even in trans2 gressing it. Perhaps there is also the suggestion that what is an end to itself must also be an end in itself, and therefore necessarily an end for
virtue of
my
3 every man.
3.
Argument from
the character
of a good will.
In the final summary of his argument 4 Kant as is his habit seems break new ground. Here he starts neither from the essence of a categorical imperative nor from the essential nature of a rational agent, but from an absolutely good will, which is the fundamental basis of his whole moral philosophy. Every rational will must in its actions set before itself an end ; but in considering a good will we have to abstract from the ends it seeks to produce, since a good will, as we have seen, cannot derive its absolute goodness from the attainment of such ends. Kant assumes, however, that even thus considered in abstraction a good will must still have an end in a sense an abstract end and that this end must therefore be an already existent end. This means that it is an end only in a negative sense, something against which a good will must never act, and therefore 5 something never to be treated merely as a means. He adds, as before, that this end must be the subject of all possible ends that is, a good or rational will itself. The fundamental contention which one might have thought could stand by itself without all this preparation is that an absolutely good will, and even the human being capable of manifesting such a will, cannot be subordinated as a means to any object of merely that is, without a breach of rational relative worth without contradiction and coherent willing.
to
4.
Argument from
the
Formula of Universal Law.
Kant's final and most obvious argument 6 is that Formula II is already implicit in Formula I, if we take into account the purposive as well as the formal side of human action. This argument may also be regarded as following from the essence of the categorical imperative; for Formula I 7 merely states the essence of the categorical imperative.
1
beings marks
*
56, that the nature of persons as rational Compare also the statement, in Gr., 428 them out already as ends in themselves that is, as something which ought
not to be used merely as a means.

Gr.,
424
= 51
8
.
Compare
Gr.,
428
= 57.
M
Gr., 437-8 67-8. 8 In this negative statement

6
Gr.,
437-8
= 68.
he appears to
forget imperfect duties.

7
Gr.,
420-1
= 46-7.
178
THE CATEGORICAL
M PE R A TI VE [XVI APP.
material maxims are always maxims of using means to some end. Formula I bids me act on these maxims subject to a limiting condition that they can be valid as laws for every rational agent. Since these laws are laws of freedom, laws in accordance with which every rational agent as such can will, this means that the will of every rational agent is a limiting condition in my use of means to an end ; and this means in turn that my adoption of maxims is subject to a limiting condition, that of not using rational agents merely as means. I will try to put this less technically. Formula I bids me act only on maxims which can be universal laws for all men. Since these laws are laws of freedom, this means that in determining my actions I have to take into account the rational wills of other men I ought to act only in such a way that as rational beings they can act on the same law as I.
My
my actions and must not be arbitrarily to say, I ought not to use them merely as means to the satisfaction of my desires. Similarly I ought not to use my own rational will merely as a means to the satisfaction of my
their rational wills limit
is
Hence
overridden by me. That
desires.
In this argument Kant concentrates both on the law, and on the existence of rational agents, as limiting conditions of my arbitrary will. But the fact that we also have imperfect duties, which are positive, shows this treatment to be unduly negative. If the aim of a good man is the realisation of an objective and universal moral law for its own sake, then as a rational agent he must be concerned, not only to refrain from thwarting, but also positively to further, all actions of others as well as of himself so far as these actions accord with the moral law. No doubt we are more immediately concerned with the realisation of our own maxims; but from a moral, and so impersonal, point of view, this is not because they happen to be our maxims, but because their realisation is more under our control. The good man is concerned with the realisation of good as such, whether it is realised through himself or through another ; and if this is so, he can no more subordinate the rational will of others to his own inclinations than he can subordinate his own rational will to his inclinations.
5.
Summary.
The multiplicity of Kant's arguments and the subtlety of his thought tend to weaken the force of a doctrine which can be stated with comparative simplicity.
Morality can be regarded in various ways as obedience to a categorical imperative, as the manifestation of a good will, and as the expression of our rational nature in action; and Kant com-
plicates his argument by bringing in all these different ways. But his essential point may be expressed thus: if morality, however regarded, is the control of inclination by a rational will obedient to universal law, it
cannot but be immoral to use a rational will in ourselves or others merely as a means to the satisfaction of inclination (or to the attainment of ends set up by reason solely at the service of inclination). This does not exclude
XVI
APP. 5]
THE END
IN ITSELF
179
the pursuit of relative ends which satisfy the inclinations of ourselves and others; indeed under certain conditions it positively enjoins the pursuit of such ends ;* but it does mean that these relative ends must be limited by, and subordinate to, an ultimate end the realisation of a good
will in
myself and in others.
Compare O., 433
=* 62.
CHAPTER XVII
THE FORMULA OF AUTONOMY

i.
Formula HI.
So act that your will can regard itself at the same time as making universal law through its maxim. 91 This formula is based on the principle that a rational will makes, or the principle of autonomy. In gives itself, the laws which it obeys the Critique of Practical Reason 2 it is described as the fundamental law of pure practical reason. It is there expressed as follows 'So act that the maxim ofyour will can always at the same time be valid as a principle making universal law?
:
At first sight this new formula seems to add nothing to Formula I. In bidding us act only on a maxim through which we can at the same time will that it should become a universal law, Formula I is already suggesting that a rational will is its own law-maker or law-giver. Nevertheless in Formula I the emphasis is on the objectivity of the moral law
and the necessitation or compulsion which it exercises on an imperfectly rational will. Formula III brings out what is only implicit in Formula I. It does not indeed deny necessitation or compulsion ; for it is a formula of duty. But it insists that the necessitation or compulsion is exercised by
our own rational will. We make the law which we obey. The will is not merely subject to law: it is so subject that it must also be regarded as making the law, and as subject to the law only because it makes the law. 3 Autonomy is the source of the unconditioned or absolute worth which belongs to moral persons as making laws and not merely obeying them. 4 It should be noted that where I speak of 'making universal law', translators and others often speak of 'universal legislation'. This means
nothing at all. 'Universal law-making' or more literally 'universal does not mean 'universal legislation'. The word 'universal' law-giving' qualifies the law, not the making just as in English an antique furniture dealer is not an antique gentleman who deals in furniture but a gentleman who deals in antique furniture.
1
0., 434
*K.p.V. 9 30
as 'the
8
= 63. = 141 =
(
54).
In
Gr.,
440
71 the
autonomy of the
4
will
is
described
supreme principle of morality*.

431
Gr.,
= 60.
Gr.,
439-40
= 7i
[180]
XVII
2.
2]
AUTONOMY
to
l8l
The approach
Formula
III.
As is his wont, Kant arrives at Formula III in a number of ways which he does not sharply distinguish from one another. He suggests that Formula III follows from combining Formula I and Formula II. 1 But he also suggests, as I have already done, that Formula III can be derived directly from Formula I 2 and again that it can come out of Formula II. 8 In yet another passage he speaks as if the inference were from Formula I to Formula II, and from Formula II to Formula III 4 Finally, as in other cases, he proposes to derive Formula III from the
;
.
essence of a categorical imperative as such. Kant would have helped the reader considerably if he had been more methodical in setting forth his arguments. Nevertheless we must remember that he is writing a treatise on moral philosophy, not on mathematics. Any clever man who adopts the modern practice of taking words as if they were mathematical symbols which can be substituted for one another in
accordance with grammatical rules without any regard to what is being can make nonsense of any kind of moral philosophy. What we must remember is that Kant always has his eye very closely on what he is talking about, and that he is endeavouring to bring out the different aspects essentially involved in moral action as such. The really important point is to see why Kant lays such stress on
said
Formula
wills as
II in his passage to
Formula
III.
In enjoining respect for rational such Formula II already suggests that the moral law must originate in rational wills as he says, they are the ground both of a categorical im5 perative and of particular categorical imperatives. But there is perhaps also a further reason the setting of ends before oneself is the essential mark of freedom. By force or threats I can be compelled to actions which are directed as means to certain ends; but I can never be compelled by others to make anything my end. 6 If I make anything my end, I do so of
reason
:
The main
is
clearly this.
my own free will ; and if duty or the law enjoins, as it does, the adoption of certain ends, and in particular if it enjoins the treating of all rational persons as ends in themselves, then the compulsion or necessitation
always present in duty must spring from my own free and rational will. I must be the source and author of the law to which I am subjected. Hence Formula II, in enjoining the pursuit of ends, implicitly asserts the autonomy of the will in making the laws which it ought to obey. Or, if we prefer it, we may say that it is the combination of the Idea of law (as expressed in Formula I) and the Idea of an end in itself (as expressed the making of in Formula II) which gives rise to the Idea of autonomy universal laws whereby I impose ends on myself.
1
the
Gr., 431 = 60. Compare kingdom of ends. Gr.,
also Gr.,
436
=
3
6
66,
Gr.,
though there he passes directly to
* 6
Gr.,
434 =63. 68. 438
MAS.,
Tugendlehre,
Einl
I.,
381
= 224.
64 and 434 See Chapter XVI
Gr.,
4.
435
= 65.
182

3.
[XVII
The
exclusion
ofpathological
interest.
There is a further strand of argument in Kant's attempt to derive Formula III directly from the essence of a categorical imperative as
such. 1
Because the moral imperative is categorical and unconditioned, every formulation of it must implicitly exclude interest, that is, pathological interest: it is not, like a hypothetical imperative, conditioned by the presence of a pathological interest in us. Formula III explicitly excludes interest ; for to say that a moral will is autonomous, that it makes its own law, is to say that it is not determined by any interest. To be determined
by
interest,
and
so
by
desire
and
inclination,
is
to
be heteronomous,
to
be
subject to a law not of our own making, and so ultimately to a law of nature, which here must be a law of empirical psychology, the law of our own needs. 2 Hence we can say 'If there is a categorical imperative, the moral will which obeys it must not be determined by interest, and therefore must itself make the universal laws which it is unconditionally bound to obey.' This is the principle set forth explicitly in Formula III. It should be noted that while the principle of autonomy can thus be established by analysis of the concept of morality and the concept of a categorical imperative, it cannot, according to Kant, be established by analysis of the concept of 'rational being' or 'rational will'. It remains a
synthetic,
and not an
analytic, practical proposition.
4. Legislating through
maxims.
is another side of Kant's doctrine which is apt to be overlooked. not saying merely that in every man there is present, however obscurely, a pure practical reason which necessarily wills in accordance with law as such and imposes this as an ideal by reference to which the maxims of action are to be selected and rejected. He is saying that a rational agent ought to act in such a way that he can regard himself at the same time as making universal law through his maxims.* This serves to bring out Kant's view of spontaneity or freedom. On the one hand he insists in a way which leads almost inevitably to mis-
There
is
Kant
that the free and autonomous and moral will must be influenced by no needs, no desires, no pathological interests, in determin5 ing its duty. On the other hand he does not forget that our sensuous motives must be 'taken up' into our maxim, if they are to influence 6 action; nor does he forget that if we are to have moral action, our
understanding
XX = 421 47. See also Gr., 439 = 70 and 432 = 6z. See, for example, Gr., 441 = 72-3. = 12). See also Chapters IV Religion, 23-4 = 24-5
*
Gr.,
Gr.,
431-2 439 =
= 60-1.
Compare
8 8. Gr., 440 70. But see also Chapter 71-2. the curious use of the preposition 'through' even in Formula I ; Gr .,
*
8
and VIII
5.
XVII
5]
AUTONOMY
iSg
into our rational volition of law as such. Because of this it is possible for moral men to make, or to give, particular moral laws through their maxims. The Idea of law as such has its origin in pure practical reason, but the making of particular moral laws constitutes the dignity and prerogative of man as a rational animal ; for the maxims through which he legislates are based upon his sensuous nature and not
maxims must in turn be taken up
1 only upon pure reason.
The common criticisms of Kant that his ethics are purely formal, that he forgets the need for every form to have a matter, that he ignores the sensuous and empirical element in morality all spring from a curious blindness to this side of his doctrine. Even when it is observed that in practice Kant always brings in sensuous and material considerations, there is a tendency to say 'You see how inconsistent he is, and how
impossible it is to work out a purely formal ethics But on this point Kant is not inconsistent. The alleged inconsistency arises only because first of all a one-sided view is imposed on Kant by the critic, and then all deviations from this view #re regarded as mere lapses. As a consequence Kant's views are made to appear doubly ridiculous. Once we adopt the hypothesis that Kant means what he says, and that he is engaged in analysing the formal side of moral actions which have always both a form and a matter, his whole doctrine becomes reasonable and intelligible, and a vast mass of
!
traditional difficulties begin to disappear.

difficulties, as all theories
His theory will
still
have
difficulty of explaining
we shall no longer have to face the why Kant came to hold moral beliefs which could
have; but
not conceivably have been held by any ordinary common sense.
man
of moderate intelligence and
5.
The
application
of Formula HI.
All this is very important if we are to understand Kant's view of the application of the categorical imperative. It must be applied not by thinking in abstraction, but by action. Since human reason is discursive, every Idea, every principle, and indeed every concept, must be abstract: it must contain less than the 2 instances which fall under it. In the case of theoretical principles, where we are able to apply them by the have concrete can instances, they observation of nature. Thus while we can say a priori that every event must have some cause, we cannot say a priori what the cause will be ; but we can, as it were, fill up our abstract formula by the study of regular sequences in the phenomenal world. This method of procedure is not open to us where we are dealing with Ideas, including practical Ideas ; for ex hypothesi Ideas can have no instances which it is possible to observe.
Yet, as
Kant
438
cannot be derived (abgeleitef) explicitly recognises, the particular
from
1
the universal by itself?

Gr.,
68-9.
(
A kingdom of ends is possible only in accordance with maxims or
self-imposed rules.
*
K.U.,
77, 407
= 348-9).
K-V-*
77> 4<>6
348).
184
[XVII
In Kant's technical language we cannot directly 'exhibit' an instance or instances of the categorical imperative, but we can exhibit such
instances 'symbolically' in action so far as we act upon maxims through which we may regard ourselves at the same time as making universal laws. 1 These maxims, provided we can act as if they were to become through our action teleological laws of nature, can be regarded as what he calls 'types' of the moral law. As we freely and spontaneously adopt such maxims in action on the ground of their fitness to be universal laws, we thereby make them to be particular moral laws for ourselves and others. These particular laws are not derived by deduction from an empty formula they are rather enacted in the living of a good life. can understand this best if we compare it with the activity of the artist who creates beauty, which, like goodness, cannot be deduced from 2 any empty formula. Yet we must not forget that the good man, unlike in is accordance with concepts the concepts of universal the artist, willing law and of a universal and teleological law of nature.
:
We
face in
Apart from the metaphysical problem of freedom, which we must due course, there is here no difficulty other than those we have already met; and Kant thinks it unnecessary to illustrate this principle
with further examples. 3
Formula I, must be applied by reference to teleoand when we pass to the next principle, the principle of the Kingdom of Ends and its analogy with the Kingdom of
Formula
III, like
logical laws of nature;
Nature, we shall find that these laws are concerned with a complete system of ends. It is therefore natural to regard our next formula as the means through which Formula III is applied. For this reason I call it Formula Ilia and not Formula IV. This fits in with Kant's own final summary of the different formulae, where the concept of law-making and that of a kingdom of ends as a kingdom of nature are combined in one formula, which he himself describes as his third. 4
1 *
a *
See Chapter See Chapter

SB*
XV, Appendix,
XIV, Appendix
especially
2,
6.
Gr.,
&> 432 n. 61 n. 436 ~ 66.
CHAPTER
XVIII
THE FORMULA OF THE KINGDOM OF ENDS

i.
Formula Ilia.
'
So act as ifyou were through your maxims a law-making member of a kingdom
is perhaps the simplest of the many versions of Formula Ilia. however, incomplete in so far as it makes no mention of the 'kingdom of nature', which Kant, in his arbitrary way, does not bring in till he comes to his final summary. We must supplement it as follows 'All maxims which spring from your own making of laws ought to accord with a possible kingdom of ends as a kingdom of nature.' 2 In this recond version the kingdom of ends is parallel to the 'universal law' of Formula I (with the addition of the 'ends in themselves' of Formula II), while the kingdom of nature is parallel to the 'universal law of nature' of Formula la. The use of the word 'kingdom' makes it clear that the laws in question are not to be considered in isolation but as part of a system of laws in both cases. Formula Ilia is thus the most comprehensive of all Kant's formulae. It expressly mentions both the form (universal law) and the matter (ends in themselves) of moral action. It shows that we are dealing, not with isolated laws or isolated ends, but with a system of laws and a system of ends. It correlates the law of freedom with the law of nature; and like the Formula of Autonomy it renders explicit the freedom with which the morally good man makes his own laws through his maxims. Kant puts some of these points more technically some may think more pedantically with reference to the categories of unity, multiIlia is said to be a 'complete determination' plicity, and totality. Formula of the maxims of moral action. 3 In Formula I we recognised that moral we recognised actionhas one form (the form of universal law) In Formula that it has for its matter many objects or ends. Finally in Formula Ilia we have reached the conception of all rational beings as ends in themselves united in one complete system under one universal law. 4 This formula combines the other two in itself. 5 In these circumstances it may seem odd that Kant should regard Formula Ilia as in some sense inferior to Formula I (and presumably
of ends. This
It
is,
= 69. See also Gr. 434 = 63 and 439 = 69. = 66. 8 Ibid. To say this to say that it combjnes form and matter. 4 Gr. 436 = 67. For a similar treatment of the categories see K.r.V., Bin Gr.> 436 = 66.
1
2
O., 438 Gr. 436

9 9
is
l86
to
[XVIII
Formula III). He appears to adopt this view precisely because Formula Ilia does not restrict itself to the form of morality, but brings in the matter as well namely, persons as ends in themselves and as having at the same time^their own subjective ends. In so doing it presents us with the attractive prospect of a world where, as a result of moral action, the ends of all free agents are realised (so far as they are compatible with freedom
under universal law), and where in consequence happiness is proportionate to virtue. This supplies a strong motive for morality, 1 but therein lies its danger for we may be tempted to seek the realisation of such a world only because we believe that in it our many interests might more easily find their satisfaction. The actions enjoined by morality must be done for their own sake as embodiments of the law and not as a means to the furtherance of our own interests. The superiority of Formula I and lies in its insistence that action is still more perhaps of Formula III morally good in virtue of its motive, not in virtue of its objects or results. Kant has by now made it plain that the moral law enjoins the pursuit of certain ends but he still maintains that a man is not morally good merely in virtue of pursuing these ends if he is to be morally good he must have a motive other than self-interest or the satisfaction of natural
;
impulses. In the formalism of Kant, even when it seems pushed to the verge of paradox, there is almost always a core of common sense. Kant does not fail to recognise the advantages of bringing in also the matter of moral action the persons whose happiness we seek and the order of society which it is our purpose to establish. It is necessary to do this in order to bring our Idea of the moral law nearer to intuition and so to feeling 'by means of a certain analogy'. 2 But we must not blind ourselves to the truth that in moral action there must be a moral motive and not merely the pursuit of a right end.
2.
The approach
to
Formula Ilia.
The approach
establish
it
to our
new formula
is
comparatively easy.
We
could
from any of Kant's previous formulae, but the simplest way is to start from Formula III. In Formula III every rational agent is enjoined to look upon himself as the maker of universal laws through his maxims, and to adopt this 3 standpoint for the purpose of criticising himself and his actions. This leads straight to the concept of a kingdom that is, of a self-governing society, a connected system of rational agents under common, selfimposed, and yet objective, laws. It leads to the concept of a kingdom of
1
Gr.,
Gr.,
for the question of 67. See the appendix to Chapter analogy. The Formula of the Law of Nature seems also to fall under this assertion 66 under indeed there primarily we get our analogy though it is given in Gr., 436 5 the head of 'form , not 'matter*. 8 62 ; also Gr., 438 Gr., 433 68-9. This is the 'standpoint* from which we
439 436
= 70. = 66 and 437 = =
XV
regard ourselves as free; Gr., 450
= = 83.
XVIII
3]
THE KINGDOM OF ENDS
187
ends because, as stated in Formula II, the laws enjoin that every member should treat himself and all others, never merely as means, but at the same time as an end. If we combine the principle of universal law with the principle of ends in themselves, we must conceive a kingdom of ends and a kingdom of ends is said to be 'possible' that is, conceivable in accordance with our 1 previous principles. Such a kingdom is very unlike our actual society,
;
and
it is
admittedly only an Ideal.
3.
The kingdom of ends.

Ideal
is
a whole or system conceived by an Idea of reason, 2 and the kingdom of ends is a whole or system of all the ends which we ought to seek. These ends are not only rational agents as ends in themselves, but are also the individual ends which each rational agent may set before himself? This latter assertion may seem at first sight inconsistent with what Kant also says at the same time (and indeed in the same sentence) namely, that we must abstract from the personal differences of rational beings and from all the content of their private ends. By this, however, he presumably means that, when we act as law-giving members of a kingdom of ends, our actions cannot be determined by personal differences or private ends as such the kingdom of ends is concerned with private ends 4 only so far as they are compatible with universal law. The system of a kingdom of ends governed by self-imposed, objective laws is the framework within which the private ends of ourselves and others ought to be realised. Such a framework by its apparent emptiness leaves room for the creativeness, in a sense the arbitrary creativeness, of
:
An
human will.
4.
Kingdom
is
or realm.
one point of terminology to be noted. For a time I was 5 persuaded by an Oxford colleague to abandon the phrase 'kingdom of ends'. This was on the ground that the German word 'Reich" does not 6 strictly mean a 'kingdom' but a 'realm.' A 'Reich* may be a kingdom (Konigreich) or an empire (Kaiserreich) or a mere tyranny like the so-called
There
'Third Reich',
is
now fortunately deceased.
These are strong arguments, and in some ways the word 'realm'
a more exact translation.
1
On
the other
hand
it is
not always possible
* * K.r.V. y As74 = 6602. Gr., 433 = 62. 433 = 62. His statements avoid inconsistency because they are based on the premise that 'laws determine our ends by reference to their universal validity*. That is to say, the good man seeks ends which are universally valid because permitted or enjoined by universal laws. This must be so, if maxims are the maxims of actions aiming at ends, and if the good man acts only on maxims which he can at the same time will to be universal
Gr>>
laws.
6
Mr. W. D.
Falk.
Or
perhaps a 'commonwealth*.
l88
[XVIII
to get precise equivalents in different languages. The word 'realm', when taken strictly, is a trifle pompous and archaic, whereas the word 'Reich* 5 is neither. What is worse, the word 'realm', like the word 'sphere , is apt to be taken colourlessly, as when we speak of 'the realm of fancy' or 'the
sphere of industry'. The conclusive consideration is, however, this that the word 'Reich* is here manifestly reminiscent of 'Das Reich Gottes* (The Kingdom of God). If the Germans use the word 'Reich' as a translation of the Greek Bao-iXeia (kingdom), it cannot be wrong for us to speak of a 'kingdom of ends'.
5.
The supreme head.
Kant assumes that the kingdom or realm of ends would have a supreme head. 1 Such a supreme head would be the author of the law, but he would not be subject to the law, as are the members of the kingdom of ends. This is possible only in the case of a being who is holy a holy being is not indeed above the moral law, for in virtue of his rational nature he
:
necessarily act rationally; but he is above a law limiting his will, is above duty and obligation, and in this sense is not subject to the law. 2 Hence he must be a being who is completely (and not just partly) rational. Such a being could have no needs, since these may give rise to desires opposed to a moral will; and this means that he must be a com-
would and so
pletely independent,
and so presumably infinite, being. The power of such a being would be adequate to his holy will. Kant adds that the supreme head would not be subject to the will of any other, thereby implying that a law-giving member of the kingdom of ends would be subject to the will of others. Perhaps he has in mind the external compulsion which is justly imposed on finite members of human society so far as they attempt to interfere with the freedom of
others.
the last part of the Groundwork. On Kant's view the existence of God is a postulate of pure practical reason, and moral law must lead to religion that is, to the recognition of all duties as divine commands. He presumably brings in this doctrine here because without it the concept of a kingdom of nature would not be intelligible.
The doctrine of a supreme head is here introduced without any argument or defence. The argument for it belongs to the Critique of Practical Reason^ though its metaphysical background is adumbrated in
6.
Dignity and price.

as distinct from the supreme head, of the subject to the law, they are nevertheless subject
2
ff. (
Although the members,
kingdom of ends are

1 *
Gr.,
433-4
K.p.V.> 124
= 63. = 265
ff.
= 223
K.p.V., 32
144
= 58).
ff.).
XVIII
to laws
6]
THE KINGDOM OF ENDS

their
189
rational will. The kingdom of ends is imposed by 1 possible only through the autonomy, or the freedom of will, of its members. This autonomy is the ground of their absolute value, their 'dignity or 'prerogative', their inner value or worth or worthiness. 'Dignity or 'worthiness' is a technical term borrowed from the Stoics and is opposed to 'price'. 2 The fact that everything in the kingdom of ends has either a price or a dignity shows how concretely, indeed how prosaically, Kant is at times prepared to interpret his kingdom
5
5
own
of ends.
value in exchange a thing has a price if we can put something place as an equivalent. Dignity or worthiness is above all price and has no equivalent or substitute. 3 The interesting thing about this doctrine (which apart from the terminology is already familiar to us) is that Kant distinguishes between a 'market price and what, in spite of misleading associations, we may perhaps call a 'fancy price (Affektionspreis) meaning thereby a value for fancy or imagination. The market price, according to Kant, is determined by the universal inclinations and needs of men the fancy price depends on taste and not on any previous need. If both these are properly called 'price', and if the things which have such a price can have an equivalent, then the price is in both cases determined by supply and demand, and there is no real difference between them. But Kant appears to be feeling his way towards a distinction between the economic value of a thing, which alone has an equivalent even in the case of works of art, and its aesthetic value, which strictly has no equivalent and yet is here regarded by Kant as 'relative Perhaps he means that music has no value for a deaf man, nor painting for a blind one ; nor indeed has any work of art any value for a philistine. A distinction of this kind would place aesthetic value somewhere between economic value and moral value, which is absolute and unique, valid for all rational agents as such, and incomparable with any other value. It is only through morality that a rational being can be a law-making member of a kingdom of ends and consequently can be an end in himself! From this Kant concludes that only morality, and humanity so far as it is 4 capable of morality, can have dignity or worthiness or inner worth, and so can be an object of reverence. 5 All other things, even our aesthetic that is, good under the condition activities, are only conditionally good 6 that their employment is not contrary to the moral law. It is worth noting that throughout his discussion Kant is concerned with the value of human activities. As examples of what has a market price he mentions skill and industry, presumably having in mind the doctrine
Price
is
:
else in its
Gr.,
434
= 63.
= 64. = 64 and 440 =
opposed
to pretium see Seneca, Ep. t 71, 37.
2 8
For
Gr., Gr.,
Gr.,
dignitas as
434 435
436
71
See also M.d.S., Tugendlehre,
i,
434
ff. ;
and
38-9.
462
ff.
5 6
= 66.
4
n.
(
Religion, Vorrede,
iv-v
n.).
Compare Chapters
II
6 and
IX
5.
igO
of
[XVIII
Smith that labour is the real measure of exchangeable value. 1 The examples of what has a fancy price are wit and humour and
Adam
imagination.
7.
The kingdom of nature.
Up to this point I have treated the kingdom of ends on a common 1 sense level, as I believe Kant expects us to treat it at this stage. The ideal set before us is a community of rational persons, obeying the same moral law for its own sake, respecting each other's freedom, and in this way striving to realise a harmonious system of ends such as can be realised in no other way. The duty of a good man is to act as a law-making member of such an ideal community. When Kant, rather late in the day, 2 tells us that the maxims of a good man ought to accord with such a possible kingdom of ends as a kingdom of nature, he in a way corroborates this common-sense treatment; yet at the same time the distinction between a kingdom of ends and a kingdom of nature offers us a glimpse of his own metaphysical doctrines and gives rise to metaphysical problems which go beyond the limits of moral philosophy in its ordinary sense and belong strictly to the final section of the Groundwork so far as they can be considered at all in a book whose range is deliberately restricted. 3 When we think of a world of rational agents which is completely in accord with moral laws, as from the moral point of view it can and ought to be, we are thinking of a moral world very different from the world of nature and of men as we know it in experience. Kant describes it therefore as an intelligible world in conceiving which we abstract from the hindrances to morality and even from all ends considered as conditions of obligation. The concept of such a world is a practical Idea which can and ought to influence the world of sense or experience we must not take it
:
as giving
knowledge of some supersensible
object.
This concept Kant identifies with the concept of a kingdom of ends, to be taken in abstraction from its manifestations in action. He tells us that in virtue of the autonomy of rational wills a world of rational agents, described for the first time as an intelligible world in the sense of 'conceivable' 5 as a (mundus intelligibilis), is 'possible kingdom of ends. If it is to be possible in another sense that is, if it is to be capable of realisation this can be so only through an analogy with a 6 kingdom of nature. We have already met with a similar analogy between the universal moral law and the universal law of nature. 7 There, however, we were concerned primarily with the problem of 'practical exhibition', the problem of discovering in what kinds of action obedience to the moral law would be manifested. Kant's answer was speaking broadly that
which then appears
See M.d.S.,
Gr.,
391
Rechtslehre,
31, 1, 289.
Gr.,
8-9.
8
7
Compare
Gr.,
433
= 62
K.r.V.,
Gr.,
and
2 above.
6.
A8o8 = 8836. 438 =69.
436
66.
See Chapter
XV,
Appendix, especially
XVIII
8]
THEKINGDOMOFENDS
IQI
the actions in question were those whose maxims, if conceived as laws of nature, would further a systematic harmony of purposes among men, or at least would do nothing to destroy such a systematic harmony. This continues to hold when we are dealing with the kingdom of ends and the kingdom of nature. But in his discussion of the two kingdoms Kant seems to be concerned with a different, though allied, problem even if we know what kinds of action are obligatory on one who seeks to realise a kingdom of ends as a kingdom of nature, what are the conditions under which alone it is possible for this ideal to be realised ? Before we consider Kant's answer we must observe that he emphasises the sharp difference between the moral laws of the kingdom of ends and the mechanical laws of nature. The kingdom of ends is possible only as governed throughout by the self-imposed laws of its members, while nature as such is governed by external and mechanical laws, above all by the laws of cause and effect. Nevertheless nature as a whole, although regarded as a machine, can also be called a kingdom of nature so far as it is directed to rational beings as its end. The analogy between the kingdom of ends and the kingdom of nature holds only in so far as nature can be regarded as purposive, and indeed as directed to a final end. This corroborates our contention that the laws of nature provide a useful analogy with the moral law only so far as they are taken to be teleological. It would be extremely arbitrary to regard nature and its laws as
:
we had some basis for this in human experience. I have already pointed out 1 that to Kant and his contemporaries the discoveries of Newton seemed to come as the revelation of a divine purpose in the universe and even apart from this it is one of Kant's essential doctrines in the Critique of Judgement that we must presuppose a 2 that is to say, the scientist must presuppose logical purposiveness in nature that the empirical laws of nature constitute an intelligible system, and he must proceed as if nature were in this sense adapted to human understanding. But Kant's fundamental point is that when we examine organic life in nature, we must make use of the theoretical Idea of a kingdom of ends in order to explain what actually exists. The moralist, on the other hand, is using the practical Idea of a kingdom of ends as a kingdom of nature in order to bring into existence something which does not exist but can to some extent be made real as a result of our efforts. 3 The concept of teleology in nature offers us a bridge between a natural world where everything is mechanical and a moral world where everything is
teleological unless
;
;
free.
8.
The
realisation
of the kingdom of ends.
We are concerned, however, not with the completeness of Kant's system nor with the subtleties of his teleological doctrines, but with the ends in moral action. Such a possibility of realising the kingdom of
1 9
Gr. f
Chapter XV, Appendix 436 n. = 66 n.
5.
K.U., Erste Einleitung, V, 23.
ig<2
[XVIII 9
kingdom could become actual only if the maxims prescribed to rational 1 agents by the categorical imperative were universally followed. Obviously it could become actual only if all rational agents always acted in accordance with such maxims and in this sense constituted, as it were, a kingdom of human nature; but even this would not be enough. A kingdom of ends could become actual only if nature itself were a kingdom governed by teleological law and were so constituted as to promote or guarantee the success of our moral volitions. In spite of his insistence that a good will is good in itself, Kant recognises that the effective manifestation of good will in the world depends, not only on the co-operation of other men, but also on the co-operation of nature itself. This is inconceivable unless nature can be regarded as manifesting a divine purpose and so as a kingdom of nature under a divine head. A good man in endeavouring to realise a kingdom of ends in this world is acting as if nature were created and governed by an all- wise and beneficent ruler for the ultimate purpose of realising the whole or perfect good (bonum consummatum) in which virtue is triumphant and is rewarded with the happiness of which it is
worthy.
ideal
2
the absolute obligation to aim at an kingdom of ends and the kingdom of nature are united under one divine head. 3 On Kant's view it is far beyond the powers of human reason to prove that such a condition can be fulfilled. The utmost that theoretical reason can do is to show that such a supposition is neither self-contradictory nor excluded by the character of our experience. Nevertheless this supposition is a necessary 'postulate' of pure practical reason and is inseparable from an unconditionally
The paradox of
which can be
morality
is
realised only if the
4 binding moral law. It may perhaps be thought that Kant puts his point too crudely, after the fashion of the Eighteenth Century, by speaking as if the main function of the deity were to add a purely external happiness to the human achievement of virtue; but however we may express it, the fact remains that the state of affairs at whose realisation the good man holds himself obliged to aim cannot be brought about by his own efforts. In this struggle to attain an ideal which is beyond his unaided powers and yet is imposed on a good man by the law of his own reason lies, according to Kant, the 'sublimity' of the moral life. A good will retains its unique and absolute value whether or not its efforts are crowned with success.
9.
The
application
of Formula Ilia.
As in the case of the other formulae the Formula of the Kingdom of Ends is useless for those who seek to determine moral duties by a process
of theoretical demonstration, but it is illuminating to those who are willing to adopt it as a principle of action. In some ways, as Kant recog1 8
Gr.,
Gr.y
438 439
= 69. = 70.
K.p.V.y 122
See Chapter II 7. 263 (
= 220).
XVI 1 1
9]
THE KINGDOM OF ENDS

human and the most moving of all his war brings home even to the most unreflective
193
principles. One is that men are
nises, it is
the most
thing that greatly moved and uplifted by playing their part, without thought of self-interest, in a great common enterprise under a great leader. It is in the light of such an experience, and not as a problem of purely logical analysis, that Kant's doctrine has to be judged. This is true, but it does not exempt the philosopher from the duty of endeavouring to articulate his practical ideal in theory, difficult though such a task must be. He must always remember how hard it is to explain in words even such simple actions as swinging a golf- club or playing a game, and how meaningless such explanations are unless we have direct experience of such actions. This is still more true of attempts to describe the moral life or the moral ideal, and it may be that Kant should have given more attention to this question on the highest abstract level. Nevertheless he does give us some guidance. He speaks of a corpus mysticum of rational beings in the world so far as their free will ( Willkur) under
moral laws has thorough-going systematic unity both with itself and with the freedom of every other. 1 Such a systematic unity of purposes or ends has to be understood, on his view, in the light of our experience of systematic unity of purpose in an individual will we must conceive all actions of rational beings so to happen as if they sprang from one supreme will which grasped all private willings (Willkur] in itself or under itself. 2 And indeed, as we have seen, the moral ideal of the kingdom of ends cannot be realised, unless there is such a single and supreme will. If there is to be complete unity of purpose among different wills, the
:
supreme
eternal. 3
will
must be all-powerful,
all-knowing,
omnipresent,
and
With the
but
it is
theological implications of this we are not here concerned, surely obvious that the systematic unity of different wills can be
when we have first described the systematic unity of one individual will. This task is unfortunately neglected by Kant, perhaps because he fails to recognise the need for a philosophy of action as such. 4 On the whole we can say that his analysis of what is meant by the systematic unity of different individual wills is incomplete. Attempts, whether satisfactory or not, to carry further this analysis have been made by various philosophers belonging to what is sometimes called the idealadequately described only
istic
school.
spite of this limitation
it is abundantly clear that on Kant's view framework of systematic unity has to be filled up by reference to the actual needs, desires, powers, and purposes of men. The common assertions to the contrary rest only on misunderstandings which have now become traditional and ingrained. We have already seen in connexion with the Formula of the Law of Nature how actions must be judged according as they further, or at least do nothing to destroy, such a systematic harmony of purposes among men and we have also seen that an action aiming at such a systematic harmony of purposes is not thereby
In
this abstract
K.r.V.,
IC.r.V.,
A8o8 = B8 3 6. A8i5 = 6843.
K.r.V.,
A8io
= B8 3 8.
I,
See Chapter
Appendix, at end.

necessarily moral.
action.
1
XVIII
o]
Coherence
is
the
test,
but not the essence, of moral
There
is,
however, a further question which Kant has not adequately
treated, the question how far in action a good man has to modify his conduct to meet the conditions of an imperfect world. This question arises in regard to all his formulae, but it arises most conspicuously in regard
to the
Formula of the Kingdom of Ends. At times his language suggests that a good man should
his
in
no way
modify
conduct to meet the actual
very often his doctrine doctrine would be manifestly absurd, and any one who has taken the trouble to study the applications of Kant's theory in detail ought to be aware that in practice Kant is very far from such extreme crudeness, 2 very noticeably indeed in his philosophy of politics. What Kant's view is careful consideration than it receives in this matter requires much more it really demands a book for itself alone ; but we may say with confidence that it is far more humane and sensible than is commonly believed. In particular we should never forget that for him there is in ethics what he calls latitude or play- room in the application of our moral maxims, and that he is always antagonistic to those who think otherwise. Nevertheless it may well be said that this is a problem which ought to be tackled fairly and squarely on the highest level, and that on this level Kant's treatment is inadequate and perhaps at times even misleading.
evil present in the world, and is interpreted in this rigoristic sense. Such a
10.
Moral progress.
in certain respects Kant's analysis
However inadequate
must recognise
provided we are willing to refrain from impossible demands such as are made on no other philosopher that he has succeeded in setting forth supreme principles of morality which are on quite a different level from a mere classification of particular duties. Without such supreme principles morality must be reduced to a hard and dogmatic legalism, when it is not dissolved into mere prejudice or emotion. One great merit of Kant's system is that it puts into a true perspective the spirit, as opposed to the letter, of the moral law. If we ask ourselves whether there is still something lacking in Kant's moral doctrines, the answer at first sight might be that as so often in
See Chapter XV, especially the Appendix 6. Apart from Kant's insistence on punishment, which would have no place in a community of saints, we may note the following points. He recognises that what is abstractly right in itself may, for reasons of subjective convenience, not be right in the eyes of the law. He expressly asserts that where there is no organised civil society, we are not obliged to act in the same way as if there were such a society. He declares that an institution, such as a hereditary nobility, may be permissible for a certain age and may even be necessary according to circumstances. See M.d.S., Rechtslehre, 36-40, 296 ff. 42, 307 ff.; and Anhang, 8 c., 369. In his lectures on Physical Geography Einleitung 5 he goes so far as to say that while the first principle of civil society is a universal law, particular laws are relative to the soil and inhabitants of a particular region.
2
;
may
be,
we
XVIII
Il]
THE KINGDOM OF ENDS
195
Eighteenth -Century philosophy they are too rigid and static and fail to give an account both of moral progress and of the empirical conditions under which such progress must take place. It is true enough that in his ethical writings Kant devotes little attention to these matters, though even here the short sections on methodology both in the Critique of Practical Reason and in the Metaphysic of Morals give us some inkling of his doctrines. He has a habit of dealing with one subject at a time, and it should not be assumed from this that he fails to see the relation between one subject and another. So far from being uninterested in history and moral progress he is miles ahead of his time in these matters as in so many others the late Professor Collingwood, a good authority and no Kantian, used to assure me that Kant's views on the philosophy of history were the soundest he knew on that subject. Kant's observations on moral progress are scattered throughout his writings, but the simplest, clearest, and shortest account of them is to be found in his Idea for a Universal History, published in 1 784 that is, in the 1 year before the publication of the Groundwork itself. Here we need note a of only couple points. In the first place Kant is well aware that what he calls 'culture', in which morality must be included, evolves in the first instance, not on the basis of a priori thinking, but as a result of give and take. This evolution he ascribes chiefly to the conflict of men in society, to what he calls their 'unsocial sociability'. In this way the first real steps are taken from barbarism to culture all talents are gradually developed, taste is formed, and through progressive enlightenment the beginning is made in forming an attitude of mind which can in time change the rough natural disposition for making moral distinctions into determinate practical principles and can thus in the end transform a pathologically extorted harmony into a moral whole. 2 In the second place Kant is equally aware that even when men have risen to the apprehension of moral principles, progress in moral insight is neither mechanical nor the result of purely intellectual ratiocination. Reason, he says, does not work instinctively, but requires experiment, practice, and instruction in order to progress gradually from one level of insight to another. 3 We may think that he ought to have developed his views on such points at greater length, but we must not imagine that he was blind
:
to these
fundamental
truths.
11.
Kanfs historical
background.
and If, as Kant holds, the moral philosopher formulates, clarifies, systematizes the moral principles already presupposed in moral judgements and moral actions, we may ask what is the historical background with reference to which Kant's own doctrines are to be understood.
1
It
is
Idee
z.u einer
contained in the eighth volume of the edition of the Prussian Academy. 8 allgemeinen Geschichte, Satz 4, 20-1, Op. cit., Satz 2, 19.
196
[XVIII
1 1
This is a further question deserving detailed study, but perhaps we can say simply that the two greatest factors in Kant's historical background
are, firstly, the Christian religion in its Protestant form, and, secondly, the influences which made for the American and French Revolutions. The whole of Kant's moral philosophy might almost be described under the title of one of his last books as 'religion within the bounds of reason alone'. For him religion is primarily the Christian religion purified, not only from the dogmas of an authoritarian church, but also from miracles and mysteries and from what he regards as the substitution of historical beliefs for rational ones. His Formula of Universal Law, insisting as it does on the spirit as opposed to the letter of the moral law, is his version of the Christian doctrine that we are saved by faith and not by works. His Formula of the End in Itself is his way of expressing the Christian view that every individual human being has a unique and infinite value and should be treated as such. His Formula of the Kingdom of Ends as a Kingdom of Nature is quite explicitly his rational form of recognising a church invisible and visible, the Kingdom of God which has to be made manifest on earth. The influences behind the American and French Revolutions, which can be summed up as the 'Enlightenment', may be regarded as a continuation of the Reformation, though the French Revolution was more bloody, and more violent against religion, because there the Reformation had been defeated. These influences are shown most markedly in Kant's passionate insistence upon freedom as the basis of all progress and all morality. The more direct and obvious influences come out, as is natural, in his political philosophy and especially in his hatred of despotism in any shape or form. It may perhaps be objected to what I have said about religion that Christianity is a religion of love and that in this respect Kant misses its very essence. Even if this were wholly true, it would in no way alter the fact that other central doctrines of Christianity find in Kant their philosophical expression. But without denying that there is here some partial truth, it may be conjectured that the sickly, unmanly, and almost meaning1 sentimentality which, in Kant's day as in our own, is sometimes preached to us under the sacred name of love, is very far removed from the original doctrine of dycbrr/ ; and indeed it may be suggested that Kant is very much nearer to that original doctrine than is commonly
less
supposed.
There are many nowadays who will think that such historical facts invalidate Kant's doctrines from the start, since they are thereby shown to be a mere 'rationalisation' of something arrived at on wholly other grounds. Even on a higher level there are many distinguished thinkers whose interest in the history and background of philosophy has led them to the view that the philosophy of any period sets forth merely the presuppositions consciously or unconsciously accepted at that period and has no bearing on other periods nor any claim to express what is rather contemptuously called 'eternal' truth. may indeed rightly be modest
We
Compare
K.p.V., 157
= 307 = 280).
(
XVIII
12 J
THE KINGDOM OF ENDS
197
about claiming truths to be eternal, and we must recognise that philosophy is more intimately bound up with its background than is such a study as pure mathematics. This gives rise to difficult problems which cannot be considered here. But it is a monstrous perversion of the truth to conor depreciate philosophical doctrines on the ground that, besides being the product of an unusually able individual mind, they also, as it were, bring into focus the thought and experience of an age or even of many ages. Such a view is too absurd to merit discussion. Unless we are complete sceptics in which case we can say nothing about anything we must to the best of our ability judge every philosophy on its own merits. What we judge to be its errors we may properly seek to explain as the result of merely contingent circumstances. But it is a far cry from this to the doctrine that a theory must be false if it has a historical back-
demn
ground. This is particularly true as regards the moral doctrines of Christianity. One of the disadvantages of asserting that these doctrines must be true because they are given by divine revelation is that those who deny revelation often infer by an obvious non sequitur that the doctrines must be false. If these doctrines are, as is contended, the result of merely
human
thought and experience and endeavour through
many
ages, this
in itself warrants a strong presupposition that they contain at least a core of truth. No doubt the moral and religious thought of different ages (including our own) tends to be cluttered up with a lot of extraneous and accidental nonsense ; but it is the mark of the great thinker to set aside the nonsense, to get at the core of truth underneath, and from this starting point to develop further the moral insight of men. This is what Kant has attempted to do; and it would be completely unreasonable to dismiss his efforts as foredoomed to failure on the purely a priori ground that doctrines which are the fruit of a long and laborious historical evolution must be unworthy of serious consideration.
12. Kant's personality.
Men are apt to judge a philosophy by the personality of its creator, and alongside the popular caricature of Kant's doctrines there has grown up a similar travesty of his character. Unfortunately, though we know a great deal perhaps too much of his life as an old and even a dying man, we know very little of his youth and not very much of his middle age. Hence a picture of him has become popularly accepted in which he
a figure of fun, the traditional philosopher, unable to think when the begin to obscure the face of the town clock, determined at all costs to avoid the slightest perspiration, governed at every moment of the day by a punctual and unvarying routine, the slave of what he called 'maxims', without spontaneity or heart, an intellectual machine, and not a man. Like all caricatures this has some slight basis in reality, at least in his
is
trees
198
1
[XVIII
12
.* He was outstandingly good at cards, but he had to give them up because he could not bear the slowness with which his partners played. He was fond of the society of ladies as they were fond of his ; and he enlivened every party at which he was present by his graceful, though dry, and at times caustic, but never pedantic, wit. Until he was a very old man he never dreamed of dining alone, but had guests at his own table, 'never less than the number of the graces or more than the number of the muses'. In his lectures he could enthral the rough youth
9
old age; and indeed without a fixed routine it would have been quite impossible for him to get through the immense amount of teaching and writing, let alone thinking, which he did when he was already elderly, if not old. But we must remember that there was a time when Kant was young, when in spite of his pocket stature and his pigeon breast he was known as
Magister
*der schone
of Konigsberg, and could apparently at will excite them to laughter and

even, it is said, to tears. And when, in later middle age, he had been appointed to a full professorship, the wiseacres could still shake their heads and say that he was too much of a dilettante to be much good at
philosophy.
Even when he was old and withered, he did not lose his simple and kindly and courteous character. When he was so weak that he fell in the street and could not rise till two unknown ladies helped him up, he presented one of them with the rose which he happened to be carrying. Perhaps the most illuminating story is of a visit paid him by his doctor nine days before his death. The old man, already almost blind and incapable of speaking clearly, so feeble that he could hardly stand, struggled to his feet and with a great effort remained standing, mumbling
some
unintelligible
words about
'posts', 'kindness',
and
'gratitude'.
The
doctor was unable to understand, but Kant's friend Wasianski explained that he was trying to thank the doctor for coming in spite of many other claims upon him, and that he would not sit down till his guest was seated. At first the doctor doubted this, but was soon convinced and moved almost to tears, when Kant, gathering together all his forces, said with a supreme effort 'The feeling for humanity has not yet left me.' These are little things, but they help to reveal character; and they should be set beside the little things which show Kant in an absurd, or even an unpleasing, light. A truer view of Kant's life will show him as essentially a humane and kindly man in spite of his single-minded devotion to philosophy. And a truer view of Kant's ethics will show him as the philosopher, not of rigorism, but of humanity.
1 For some of it I have not been able to discover any evidence. Thus it is commonly repeated that he ate too many cakes at tea, and that his only moral problems arose in regard to this matter. I know no evidence for this, and the whole story is most surprising, as it is well known that he took only one meal a day, and this meal was certainly not tea. 8 Perhaps this can be best rendered in the Scottish tongue as 'the bonny M.A.*
BOOK
IV
THE JUSTIFICATION OF THE CATEGORICAL IMPERATIVE

CHAPTER XIX
THE PROBLEM
i.
The
question to be answered.
perative,
are to understand Kant's attempt to justify the categorical imwe must keep in mind its place in the whole argument. In Chapter I of the Groundwork Kant started from an analysis of ordinary moral beliefs, and he argued from them that the condition of moral
IF
we
action- at least for finite human beings is obedience to law for its own sake or obedience to a categorical imperative. In Chapter II he formulated the categorical imperative in five different ways. For our present purpose, and for the Critique of Practical Reason, the Formula of Autonomy is the most important 'So act that your will can regard itself at the same time as making universal law through its maxim.' Autonomy, or the making of the universal laws which we are bound to obey, is thus the principle behind ordinary moral beliefs and purports to be the condition of moral action. The question we have now to face in Chapter III of the Groundwork is
:
'Can
this principle
be
justified?'
does not always keep the distinction clear, we should not forget that the principle of autonomy need not take the form of a that is, it categorical imperative it expresses the essence of moral law is the principle on which a rational agent as such would necessarily act if reason had full control over passion. As we shall see, 1 Kant attempts to justify the principle as a moral law and only thereby to justify it as a if it is a principle on which a fully rational agent categorical imperative would necessarily act, it must also be on his view a principle on which
Though Kant
an imperfectly rational agent ought to act, if he is tempted to do otherwise. We can pass without difficulty from an unconditioned objective principle
of action to a categorical imperative. Since the Formula of Autonomy sets forth the essence of the categorical imperative, our question may be put in the form 'How is a categorical imperative possible?' This question we have hitherto burked on the ground of its difficulty, a difficulty arising from the fact that a categorical
1
See Chapter
XXIV
7.
[199]
2OO
[XIX
imperative, unlike a hypothetical one, is a synthetic a priori practical 1 proposition. The possibility of hypothetical imperatives could easily be made intelligible because it rested on the purely analytic proposition that to will the end is to will the means. Hence any rational agent, so far as reason had decisive influence over his inclinations, would necessarily will the means to his chosen end, and ought to do so if he is sufficiently irrational to be tempted to do otherwise. But according to Kant no analysis of the concept of 'rational being' or 'rational agent can give us a categorical imperative affirming that every rational agent ought to act in accordance with the principle of autonomy. 2 Here we have an a priori proposition which cannot be derived from mere analysis of the subject-concept and is consequently synthetic. As in the case of all synthetic a priori propositions we require some 'third term* to establish a necessary connexion between the subject and the predicate. This 'third term' we shall find speaking broadly to be the Idea of
5
freedom.
The problem which we have here to face is therefore a special form of the general problem which is always for Kant the fundamental question of philosophy, namely, 'How are synthetic a priori propositions possible?' Kant's answer varies greatly in dealing with special problems he is never satisfied with merely mechanical solutions but it may be said that in every case it has one common characteristic it seeks to justify a synthetic a priori proposition by tracing its origin to the nature of mind as such and in particular to the activity of reason itself. Such a justification is what Kant calls a 'transcendental deduction ; and it belongs to a critique or criticism of reason by itself. Since the synthetic a priori propositions with which we are at present concerned are practical propositions, stating not what must be, but what 3 ought to be, our argument belongs to a critique of practical reason. It does not, however, profess to be a full critique, but aims only at setting forth the main features of such a critique so far as these are necessary for our present purpose. 4 Such a summary treatment, taken by itself,
:
is
bound
to over-simplify
and
so give rise to difficulties. For
an adequate
understanding of Kant's doctrine

works.
we
require knowledge of his other
It is clear from all this that Chapter III of the Groundwork is on quite a different footing from the first two chapters. Chapters I and II may be influenced by Kant's doctrines, but they at least profess to offer an analysis of the implications of our ordinary moral judgements: this analysis might stand, and I suggest does stand, whatever we think of the Critical Philosophy as a whole. In Chapter III we have to make an incursion into Kant's own metaphysical doctrines, his views of freedom and necessity, and his distinction between the phenomenal and the noumenal world. Nevertheless it would be a mistake to suppose even here that the questions raised and the difficulties involved are
1
See Chapter XII Gr., 440 = 72.
9.
4
Gr. 440 O., 445

t
= 71-2 420 n. = 45-6 n. = 77 391 * 8-9.

; ;
449
= 82.
XIX
2]
THE PROBLEM
2OI
peculiar to Kant's philosophy. His way of stating and of solving may be peculiar to himself, but this should not blind us to the fact that he is dealing with real problems, not artificial ones.
problems
2.
An
alternative question.
Can we justify the proposition that a rational agent as such ought to act in accordance with the principle of autonomy? This is the question which Kant has asked and the question which in fact he attempts to
answer in his argument. Unfortunately, however, he puzzles us by propounding also what seems to be a quite different question, though he shows little sign that he is conscious of the difference. His second question is 'Can we justify the synthetic a priori proposition that an absolutely good will must be one which acts in accordance with the principle of
autonomy?' It must not be thought that an absolutely good will is here one which is holy, and so one which is not under a categorical imperative, although it necessarily acts in accordance with self-imposed laws; for in this connexion Kant has said rather loosely that the principle of an 2 absolutely good will must be a categorical imperative. Neither an absolutely good will nor a rational agent as such need be under a categorical imperative ; but both of them alike under human conditions, where they have to meet the resistance of passion, are under a categorical imperative; and Kant is still concerned with them under such human
conditions,
although
this
qualification
is
Otherwise
possible?'
his question
could not be
'How
is
not continually repeated. a categorical imperative
Curiously enough, Kant has already said that analysis of the concepts of morality can perfectly well show that the principle of autonomy is the 3 only principle of morality. If so, presumably analysis of the concept of a moral will must equally be able to establish the principle of autonomy; and this suggests a distinction in connotation, if not in denotation, between an absolutely good will and a moral will. There are also further compli4 cations into which we need not enter. The simplest way out of these complications is to assume that a rational will and an absolutely good will are regarded by Kant as identical, and this might be maintained on the basis of his definition of goodness, 5 provided it were suitably amended. Certainly it seems reasonable to maintain that a good will can be only a will which acts rationally; but here again we have to ask whether this is an analytic or a synthetic proposition. If it is a synthetic proposition, how is it to be justified? If
1
Gr., 440 = 72. 444 = 77. 397 = 16, for example, the coricept of duty is said to contain in itself the concept of a good will, although under certain subjective limitations and hindrances, 6 See i Chapter
2
50
= 82.
Gr.,
4
Gr. t
447
79; 444
=77. The two
questions
seem
8
to
be connected in Gr. 9 449-
In
Gr.y
2O2
it is
[XIX
an analytic proposition, it must surely rest, not on an arbitrary definition, but on some kind of direct insight. In spite of the importance of the connexion between an absolutely good will and the principle of autonomy, we must take Kant's main question to be concerned with the relation between a rational agent as such and the principle of autonomy. His position appears to be that if this main question can be answered satisfactorily, the question about an absolutely good will offers no further difficulties. Kant's insistence that the Idea of freedom is necessary to connect an absolutely good will with the principle of autonomy1 shows clearly his own emotional attitude it is in freedom that he finds the unique worth and sublimity of finite human beings. We must not assume that this necessarily introduces an irrational and merely subjective element into his argument, but it is reasonable to say that in this respect his argument
:
requires careful scrutiny.
3.
The purpose of a
transcendental deduction.
is
In some cases a transcendental deduction

the
necessary only to justify
known
it
is
cases
possibility of a synthetic a priori proposition, whereas in other necessary also to establish it. Thus, according to Kant, the
:
known
synthetic a priori propositions of mathematics and physics are already to be possible because of the success of these sciences our only task is to consider how they are possible. In the case of metaphysics, however, we are offered so many contradictory statements that we have to ask whether its synthetic a priori propositions are possible before we begin to ask how they are possible. In the Groundwork Kant treats the synthetic a priori principles of moral philosophy as more akin to those of metaphysics than to those of mathematics or physics. Over and over again he insists that he is not maintaining their truth, still less pretending that he has a proof of them. 2 All he has done is to determine by a regressive or analytic argument the ultimate conditions or presuppositions of accepted moral beliefs. If this is so, we cannot without a vicious circle proceed to justify ordinary moral beliefs
progressively from these ultimate conditions, unless we can in some way establish these ultimate conditions independently by a transcendental deduction. 3 Hence we require a transcendental deduction of the principle of autonomy. Its purpose is to establish the possibility that moral judgements may be valid, and not merely to justify a possibility taken as already
established.
illusions.
At present
it is
an open question whether moral
beliefs,
and
with them the presuppositions on which they are grounded,
may
not be
Kant
1
dental deduction.
Gr., Gr.,
certainly regards the Groundwork as offering us such a transcenThus he talks about the Tightness of his deduction ; 4
See, for example, 447 79. 450 =s 83 ; 453 *= 86-7. See also Chapter I 9.
Gr.,
444-5
==
77.
Gr.,
454
88.
XIX
4]
THE PROBLEM
about the deduction of the concept of freedomTrom pure practical reason ;* and about his deduction of the supreme principle of morality. 2
4.
different view.
We must remember that by a 'deduction' Kant does not mean a deduction in the ordinary sense, but rather a 'justification' based on the insight of reason into its own rational activity. Nevertheless the view of the Groundwork is open to grave objections. In the first place, the moral law may seem to be more certain than its alleged justification we have at least as much assurance of being under moral obligations as of being free. In the second place, even if this were not so, it is surely inconceivable that the moral law could be justified by anything other than
:
itself.
have become aware of both these objections. There are an awareness even in the Groundwork, though the obvious interpretation is the one I have given but it is in the Critique of Practical Reason that there emerges a view which is fundamentally
to
Kant seems
traces
of such
different.
The difference between his two views may be put roughly as follows. In the Groundwork he seems to think that the moral law is both justified
and established by an independent and necessary presupposition of freedom. In the Critique, on the contrary, it is our consciousness of the moral law which leads to the concept of freedom and in such consciousness Kant no longer finds difficulty. 'We can become conscious of pure practical laws just as we are conscious of pure theoretical principles.' 3 The moral law is, as it were, a 'fact of pure reason', of which we have a priori knowledge and which is apodeictically certain. 4 Only on the basis of the moral law can we justify the presupposition that the rational will must be free; and the moral law is even described as a principle for the deduction of freedom. We are expressly said to seek in vain for a deduction of the moral 5 law, which requires no deduction. We need not, however, exaggerate the difference between the two views or suppose that the argument of the Groundwork is a mere waste of breath. Even if we hold the moral law and the categorical imperative to be apodeictically certain, there still remains the problem of freedom; and if freedom were merely an illusion, this could not but suggest that our belief in moral obligation and moral responsibility must also be an illusion. Hence both the Groundwork and the Critique of Practical Reason have at least to show how a categorical imperative is possible by refuting the dogmatic determinism which affects to prove that freedom is
;
Gr.,
447
79.
163 ( 81). 82-3). The moral law does, however, receive some sort that is, apparently, some additional and purely theoretical of 'credentials' (Kreditiv) ground for being accepted from the fact that it gives positive content to the Idea of freedom which pure theoretical reason can and must conceive only negatively.
1T./.K,
30=
140
Gr.,
463
53).
(
K.p.V.,
47-8
163-4
*K.p.V., 47
100.
204

lies
[XIX
impossible. This problem
behind the whole of the argument we are about to consider, though Kant does not make it explicit till he is within 1 sight of the end. We must keep an open mind and continue to judge the
argument on its merits. It may, however, seem disappointing if we are no longer to expect proof of the validity of the moral law or of our obligation to obey it. But
disappointment, as Kant recognises even in the Groundwork* may spring from false expectations and from complete misunderstanding o the moral law itself. The metaphysician cannot prove the validity of the moral law from some principle which is not itself moral. If Kant ever thought that he could, he was mistaken; but it is doubtful how far he 3 regards freedom in the Groundwork as a non-moral principle.
this
5. Possible misunderstandings.
We must be clear at the outset as to the meaning of the question 'How a categorical imperative possible?' There are various ways in which this question may be misunderstood. What we are asking is whether a proposition is or is not valid. Ordinary moral judgements affirm that we ought or ought not to do certain actions (or kinds of action) We are not now concerned with the truth or falsehood of these judgements in particular cases. The point is that, so far as they are moral and not prudential, they do not assert that we ought to
is
.
we happen to want something else what they assert is or ought not to do something, whatever we may happen to want. Thus every moral judgement presupposes the possibility of asserting a simple, unqualified, or categorical 'ought'. If Kant's analysis of this assertion is sound, all moral judgements presuppose the ultimate proposition that every rational agent as such ought to act in accordance with the principle of autonomy. What we must now ask is whether we are justified as rational agents in accepting and acting on this proposition. Is the proposition the result of mere prejudice or is it valid for every rational agent as such? If the proposition is not valid, then all our moral judgements without exception are illusory we cannot say, for example, that the sadists of Belsen acted wrongly, but only that we happen to dislike their conduct especially if it is directed towards ourselves. To many people it may seem that Kant's proposition is already justified if the rejection of it means the denial of all moral judgements it is as much justified as a theoretical proposition of which it can be truly asserted that unless you think this, you cannot think at all nor can there be any such thing as truth. But even if this were so, even if on this ground we did not doubt that Kant's proposition is valid and in that sense 'possible', we do not doubt the propositions of might still ask 'How is it possible?'
do something
that
if
:
we ought
We
89 ff. 29 sas 140 ( positive concept of freedom

Gr.,
ff.
455
In
/T./J.F.,
word
'diese*
52) he hints that an unconditioned moral law and the may be the same thing, but his language is careless. The in line 27, p. 29, of the Academy Edition ought surely to be 'dieses'.
Gr. 9
463
= 99-1 oo.
XIX
5]
THEPROBLEM
we may
still
2O5
mathematics, but
ask how these propositions are possible
and give an answer.

mathematics are
It would, for example, be an answer though on Kant's view a false answer if we could show that the propositions of
all analytic.
The question we are asking in no way concerns the historical development of moral ideas. For enquiries of this kind Kant always shows the greatest respect but although they might explain how we came to enter;
would do nothing to justify our acceptance of even a tendency nowadays to assume without further ado that they would show our concepts to be illusory. We are concerned solely with the validity of propositions, just as we are when we ask how pure mathematics is possible. No doubt there are important differences between theoretical and practical propositions as regards validity but if we turn back to Kant's discussion of the possibility of hypothetical imperatives, 2 we shall see that there too he deals with the
tain certain concepts, they
them. 1 There
is
validity of a proposition
and
indicates clearly
enough that the question
about the possibility of a categorical imperative is of the same kind as the question about the possibility of hypothetical imperatives. Hence it is a mistake to suppose we are concerned with psychological questions. We are not asking how a categorical imperative can manifest itself in action by giving rise to some emotion which can act as a motive.
true that in the Critique of Practical Reason* Kant does attempt to how consciousness of our subjection to moral law can arouse the feeling of respect or reverence which is the emotional side of moral action
It
is
explain
may be regarded as its motive ; but this explanation is not concerned with the possibility of categorical imperatives. It is true also that in his present argument Kant speaks of the 'interest' attaching to moral Ideas; but he takes a different view from that in the Critique and maintains that
and
moral interest
It
is is totally inexplicable. greater mistake to suppose that Kant is trying to propound a speculative theory about the way in which a categorical imperative can have effects in the phenomenal world. He has, it is true, to deal with metaphysical questions in order to show that the possibility of freedom is not excluded by the very nature of our experience. But this is subsidiary to his main task, which is to justify a synthetic a priori proposition by showing that we are entitled to affirm a necessary connexion between namely, the Idea of subject and predicate in virtue of a 'third term'
still
Teedom. 4
It is sheer error to suggest that Kant is trying to explain how pure 5 reason can be practical or how freedom can be possible. These questions Kant has not only refrained from answering he roundly asserts that they 6 ire beyond the power of human reason to answer.
:
1 8
Compare
K.p.V., 71 Gr. y 447
K.r.V.,
ff.
A86-7
ff. (
196
= Bi 18-19. Chapter XII = 128 See also Chapter V 2.

a
6 and 8.
ff.).
5 It
fessor 6
that even the criticisms made by so good a scholar as the late ProHoernle* in the Personalist (October 1939) rest upon this error.
79.
seems to
461
me
Gr.,
97-8; 458-9
94.
206

6.
[XIX
Kanf s
this
method.
important discussion, as in so many others, Kant's arguments more than once and are apt to be modified in the process. This is almost inevitable since of the 5 Sections into which Chapter III of the Groundwork is divided, the first three are preparatory; the fourth gives a summary outline of the whole argument; while the fifth both clarifies what has gone before and adds a variety of new points. We shall have in so difficult a matter to follow Kant's divisions pretty closely, while trying to avoid some of the repetition. His method, we must remember, he expects us to get fuller insight as we proceed. Though is deliberate this is not without its advantages, it offers considerable difficulty to the reader, at least if he is mathematically minded and expects a quite different kind of argument.
are repeated
:
In
CHAPTER XX
FREEDOM AND AUTONOMY

i
.
Kant as a pioneer.
IN the discussion of freedom Kant's work is that of a pioneer. The Greeks never really came to grips with the subject and did little to carry it beyond limited questions of legal responsibility. In mediaeval philosophy there was a real advance, but the problem was considered in theological terms how was human freedom to be reconciled with divine omnipotence and omniscience? Kant separated the problem of freedom from its legal and theological setting and asked simply how freedom can be compatible with the causal law which prevails throughout nature, and apparently
:
throughout human nature. is breaking new ground, we need not be surprised if we find a certain roughness and obscurity, perhaps even inconsistency, in his doctrines. This will not deprive him of the credit of having posed the question so sharply that thereafter it can no longer be neglected. It seems probable that there was a considerable development, not only in Kant's 1 terminology, but in his views: we have already noted a fundamental difference between the Groundwork and the Critique of Practical Reason in this respect. A full discussion of Kant's theory of freedom demands a book to itself: here we can attempt only to understand the argument of the
also
As Kant
Groundwork in the light of his other writings.

2.
Freedom as
the key to the
moral problem.
is
Kant begins his discussion by suggesting that the concept of freedom the key to the explanation he should rather have said 'justification' of the principle of autonomy. If we could justify the concept of freedom by tracing it to its origin in practical reason, we should then be able to show how a categorical imperative, as expressed in the Formula of Autonomy, is first task, however, is merely to show that a free will is possible. Our autonomous will and so to a will under moral law. equivalent to an This equivalence must be established by stating what freedom is ; that is, by definition of freedom or of a free will. giving an acceptable Granted that we can define a free will correctly, the proposition that
a free will
is
equivalent to an autonomous will

2
free will,
analytic proposition. though in our analytic proposition only implicit before.

i
It
is on Kant's view an can be established by analysis of the concept of a
we make
explicit
what was
Chapter
XIX
4.
Gr.,
447
79; K.p.V., 28-9
138-40
= 51-2).
[207]
2O8
[XX
Although a free will is equivalent to an autonomous will, we apparently arrive at these two concepts by different routes. Our concept of an autonomous will was obtained by an analysis of the conditions or presuppositions of morality. Our concept of a free will is to be obtained by a definition which presumably must be based on insight into the nature of something other than moral beliefs. This last view appears to be explicitly 1 rejected in the Critique of Practical Reason, yet it is not altogether without
interest
and
plausibility.
3.
Will as causality.
if it characterises
Assuming that freedom,

a
will,
anything, must characterise

c
begins with a new definition of wilP. Hitherto we have known 'will' as 'the power of a rational being to act in accordance with 2 its conception of laws, i.e. in accordance with principles'. We are now 3 told that 'the will is a kind of causality belonging to living beings so far as they are rational*. Will is regarded as the power of a rational being
Kant
to
produce
effects in the
world.
The power
will,
to act
phenomenal world, and primarily in the physical would commonly be regarded as a power to
produce such
effects.
Our
however,
may
also
produce changes in our own mental
world, the world of inner sense as when we decide, for example, to think about a particular topic. rational being capable of doing this might be considered to possess a will even if he were unable to influence physical events. The point is of some importance, for, as we shall find later, Kant regards our judgements, and not merely actions in the ordinary sense, as taking place under the Idea of freedom. 4 Kant has also indicated the possibility that we might will (not merely wish) and yet through some special disfavour of fate might produce no external result or at least not the result intended. 5 He may have in mind some sudden misfortune, such as a stroke of paralysis but quite apart from this it is only too common an experience to find that what we do is in fact different from what we willed to do every missed stroke at tennis is an example of this. Perhaps it was with reference to some such distinction between willing and doing that Kant defined the will as 'a power to determine oneself to action in accordance with our conception of certain laws'. 8 Problems of this type are, however, generally passed over by Kant. They belong not to moral philosophy but to the philosophy of action. For purposes of moral philosophy we are perhaps entitled to consider only what may be described as the normal case namely, that we produce the
effect
which we
will to produce.
If the will is a power to act or to set oneself to act in accordance with one's conception of laws, willing must be a conscious, and indeed in some degree a self-conscious, activity. To think of rational beings as
1
446 6 Gr. 9 394

Gr.,
K.p.V., 29-30
= 78. = 12.
140
= 53).
Gr. 9 412 Gr. s 448

Gr.,
= 36. = 81 427 = 55.
452
86.
XX
41
FREEDOM AND AUT C/N O M Y
2OQ
will is to think of them as possessing 'consciousness of their causality in regard to actions'. 1 One last point should be noted. The word 'causality' is commonly used c by Kant in two senses, (i) It may mean a power to produce effects' ; and it mean 'causal When he action'. (2) may says that the will is a kind of causality, he means that it is a* power to produce effects. When he speaks
endowed with a
cause as being 'determined to causality' by something is determined to causal action that it is itself caused to act causally. Willing may be described as causal action, but 'the will' is merely the power to act causally that is, to produce effects. We may
efficient
2
of an
else,
he means that it
fail to
understand Kant
if
we
neglect to notice this ambiguity.
4.
Freedom and natural
necessity.
If we conceive the will to be free, we must mean in the first place that is a power to produce effects without being determined or caused to do so by anything other than itself. Freedom is a quality belonging to a special kind of causality. Perhaps it would be simpler to say that it characterises a special kind of causal action. It is opposed to 'natural necessity' or 'the necessity of nature', a quality characterising all causal action in nature. 3 may for brevity state the contrast simply as one between freedom and necessity, but we must remember that there may be
the will
We
other kinds of necessity
for example, logical necessity.
What
is
meant by
this 'necessity'
which characterises causal action in
nature? Let us take a crude example. If a billiard ball strikes against a billiard ball which is at rest, it will cause the second ball to move. But the first ball does not spontaneously cause the second ball to move it causes the second ball to move only because it was itself driven against the second ball by a billiard cue. It does produce an effect namely, the movement of the second ball but its causal action in so doing was itself caused by something other than itself namely, by a blow from a cue. In Kant's more technical language its causal action was necessary we might almost have said necessitated and not free. If to act is to produce effects, then all action is causal action. Hence we can omit the qualification 'causal' and say that all action in nature is necessary. In nature there is no spontaneity and no freedom only an endless chain of causes and effects. The necessity of nature is causal it is a necessity in accordance with which every event must be caused by a
:
preceding event.
The necessity of nature is not confined to inorganic objects. It is because of an external influence that moths fly into a light. Even on the level of the higher animals Kant conceives of a dog as moved to pursuit by the sight or smell of a rabbit. To say this is not to say that a dog is merely a material machine. We may with Leibniz call it a spiritual, and
i 8
Gr.>
Gr. t
449 = 81. 446 =s 78. The word
Gr. 9
446
= 79.
O
'action'
is
here used widely and not restricted to dis-
tinctively
human
action.
2IO

is
[XX
it falls
may cover psychological as well as 1 purely physical laws. If the will of a rational agent is conceived as free, this must mean that we regard his causal actions, or more precisely his volitions, as not determined by causes external or alien to himself. Under external causes we must here include, not merely physical forces, but also the sensa given us from without, the images suggested by these sensa, the emotions aroused by sensa and images, and the desires stimulated by emotions. It must never be forgotten that on Kant's view the whole succession of events in inner sense, and in particular the succession of sensa, images, emotions, and desires, is as much governed by natural necessity as is the movement of
stocks
not a material, automaton, for it the mechanism of nature, which
moved by
ideas; but
under
and
stones. 2
This does not mean that Kant recognises no differences between men and animals any more than it means he recognises no differences between animals and things. If we look at the question purely from a psychological point of view, animals differ from things in being moved by ideas as well as by physical forces, and men differ from animals in being moved by reason as well as by ideas of sense and imagination. It is an empirical fact that men have a power to rise above immediate sensuous impressions and impulses through concepts of what is useful or harmful in a remoter way. This remains an empirical fact even if the reason which thus prescribes laws for action is itself determined in turn through influences from elsewhere. Hence Kant ascribes to animals an arbitrium brutum and to men an arbitrium liberum, terms which unfortunately have no satisfactory English 3 We may dislike the terminology, we may even suspect equivalent. confusion of thought to lie behind it, but there is no doubt whatever that this is Kant's doctrine, and that in it he is at least recognising
empirical facts which it would be stupid to ignore. This empirical characteristic of man by which reason plays some part a fact recognised equally by Hume 4 Kant in determining his actions sometimes calls, not too happily perhaps, 'psychological' or Comparative' freedom ; but he knows perfectly well that it may all fall under the head of 'natural necessity'. Man might still be a spiritual automaton, even if he were moved, not by instinct, but by a reason which revealed to him remoter objects of desire and the means to attain them his so-called freedom might be rather like that of a turnspit which moves of itself
:
5 up. The freedom which we have conceived as belonging something quite different it can be present only if the will is a power to produce effects without being determined by anything other than itself. As so described the freedom which is opposed to necessity is a negative
once
it is
wound
to a will
is
A549-50 = 6578-9; K.p.V., 95-7 - 226-8 (= 171-4). 8 = 8562. The nearest we can get A>.F., A802-3 = 8830-1 A534
*
iK.p.V., 96-7
TT.r.F.,
227-8
172-4)-
expression for arbitrium (Willkur]

is
is
perhaps
'choice', or 'free choice'
to an English where the arbitrium
liberum.
4
*
See also 8 below.
Compare Mrs. R. M. Kydd,

K.p.V., 96-7
227-8
Reason and Conduct in Hume's Treatise p. 115.

-,
172-4)-
XX
5]
211
concept, an Idea of reason derived from reflection upon necessity itself. We may call it the 'transcendental Idea' of freedom, a purely theoretical 1 concept not based upon any moral considerations and in itself empty.
if we totally reject this negative concept, it will be impossible to justify a positive concept of freedom. Kant claims to have proved in the Critique of Pure Reason, not that there corresponds to this negative concept any actual, or even possible, object, but only that the concept is neither self2 contradictory nor necessarily excluded by the nature of our experience.
Yet
5.
The positive
concept
offreedom.
if he
Kant has to go beyond the negative sense of freedom to a positive one is to show that freedom is equivalent to autonomy. He attempts to do this by means of the concept of causality, having defined will as free causality, that is, as a power of free causal action. The concept of causality,
he
asserts,
implies the concept of law; and this must hold whether here used presumably in the sense of causal action is determined by natural necessity or is free. 3 The grounds alleged for this crucial assertion are curiously inadequate. The concept of causality, we are told, implies the concept of laws (Gesetze) in accordance with which through something which we call 'cause' something else namely, the effect must be posited (gesetzt). Hence a causality characterised, not by necessity, but by freedom, cannot be lawless, but must accord with unchanging laws of a special kind. Otherwise a free will would be a logical absurdity (ein Unding}. The word 'posited' in English is always obscure, and with us there is no magic in the connexion between 'Gesetz? (law) and 'gesetzf (posited). The law of which Kant speaks appears, by his own account, to be a law connecting cause and effect so that like causes necessarily have like effects. But this applies only to natural necessity. It is hard to see how we can be entitled to pass from this to a law of freedom, which so far from connecting causes and effects is a law for causal action considered in 4 itself. The law or principle of autonomy, as we have seen it hitherto, in no way asserts a necessary connexion between causes and effects. There is rather more force in the contention that a lawless free will would be an absurdity. Such a view, however, is not derived from any necessary connexion between causality and law. It arises because a lawless free will would be governed merely by chance and so could not properly be described as free. If Kant's doctrine turned merely on the present argument, we should have to dismiss it as fallacious. The curious thing is that the argument is as superfluous as it is weak. There is no need whatever to make the connexion between free will and law depend on the concept of causality.
causality
2 See Chapter IX, Appendix 4. K.r.V., ^557-8 6585-6. 218 ( Gr.> 447 160) Religion, 35 39 ( 32). 78. See also K.p.V., 89 5 4 Perhaps Kant has in mind the account of what he calls 'character in the Critique 6577 ff.). But I cannot see of Pure Reason (see especially A539 6567 and A549 fF. 1
= =
that
it justifies
his present assertion.
212
[XX
of will as previously given was 'the power of a rational being to act in accordance with its conception of laws, i.e. in accordance with principles'. 1 If so, a lawless free will would not be a will at all it would be a contradiction in terms (a mere Unding). As we shall see, Kant's doctrine rests on insight into the necessary activity of reason, and in particular of practical reason. The present statement puts forward a rough account of what he hopes to justify, and
:
The very definition
supports it by an unconvincing argument, which were better omitted. He is not trying to prove that the will is free, but merely showing how a if there were such a thing would have to be conceived. We free will need not labour the point that a will governed by the laws of natural necessity would not be free ; for by such laws, whether physical or psychological, causal action must itself be caused in the long run by something other than the agent, and no one would regard this as freedom. If a free will cannot be lawless, its laws must be of a different kind from the laws of nature.
6.
Freedom and autonomy.
are we to distinguish the laws of nature from what we may now laws of freedom? In nature the causal action of an efficient cause is itself caused by something else: it is not spontaneous. This means, according to Kant, that the law governing causal action in nature is not self-imposed but is imposed by something else. This is what he calls 'heteronomy'. Hence if we are to distinguish the laws of freedom from the laws of nature, we can do so only by supposing that the laws of freedom are self-imposed. The spontaneous causal action of a free will must therefore take place in accordance with self-imposed law. But this is just what we mean by 'autonomy' ; and a free will must be conceived as acting under the principle of autonomy that is, as capable of acting on maxims which can at the same time be willed as universal laws. Since we have discovered this to be the principle of moral action, we can say that a free will and a will under moral laws are one and the same thing. 2 The argument, to say the least, is hurried. We have to make several assumptions, each of which demands careful consideration. We have to assume (i) that free willing must have its own special kind of law; (2) that since all law must be either self-imposed or other-imposed, the law of freedom must be self-imposed; and (3) that self-imposed law can be only the law of following law as such. In spite of some help from preyious discussions, these assumptions are not to be lightly made. The argument might be more hopeful if it could be tackled, so to speak, from the other end. If we could show that practical reason must assume itself to be capable of acting in accordance with its own rational
call the
How
we might be justified in holding that it must be free and also autonomous in the sense required. 3 This, as we shall see, is much more like the argument which Kant actually uses, though in it the Idea of 2 1 3 Gr. 447 = 79. See Chapter IX 4. Or., 412 =5 36.
principles,
}
XX
y]
less prominent as the link connecting the subject and predicate of the categorical imperative. We must take the present argument as one of the preliminary and incomplete discussions with which Kant is wont to preface his central arguments. Incidentally, it is not wholly clear, even in the present passage, whether the positive concept of freedom is itself the 'third term' necessary for Kant's purpose. Freedom, we are told, furnishes (schaffi) this 'third term', but it is also said to 'point' to this 'third term' (worauf uns die Freiheit weiset). Sometimes Kant speaks as if the 'third term' were not freedom but membership of the intelligible world. On the whole, however, these two concepts are so closely interconnected that he may not have been conscious of any sharp difference between them.
freedom becomes
7.
Is only a
good will free?
There are further difficulties about the meaning of Kant's doctrine. he maintaining that a morally good will alone is free, while a morally bad will is determined? It may be thought that the answer must be in the affirmative, since Kant has just said that a free will and a will under moral laws are one and the same thing. This, however, would be a mistake. Kant expressly distinguishes between a will 'under moral laws' and a will which always 1 obeys moral laws. To be under moral laws is to recognise the categorical imperative, but not necessarily to obey it. Apart from the merely linguistic question there is no doubt that Kant believes us to assume freedom in the mere recognition of the categorical imperative: 'I ought' implies 'I can'. Indeed the examples by which he illustrates his doctrine of freedom are nearly always examples of bad moral action, such as lying and thieving. 2 The bad man could have acted otherwise. Unless this is so, there is no meaning in saying that he ought to have acted otherwise, and there is no justification in blaming him for what he has done. Whatever may be the case with other philosophers, it is not Kant's view that we are responsible for our good actions and not for our bad ones: we are responsible for both alike. In the Groundwork itself, while he declares very sensibly that we are not responsible for our inclinations and impulses, we are responsible for the indulgence which enables them to influence our maxims to the prejudice of moral law. 3 In his later works Kant sets forth what seems to be a development of his present position. There he makes a distinction between will (Wille) and arbitrium (Willkur). Will in this technical sense is concerned only with the law and so seems to be equivalent to pure practical reason it is said to be neither free nor unfree. 4 Arbitrium on the other hand, is the source
Is
:
K.U., 87, 448 n. ( 422 n.). 8 Gr. t 458 1 7 1 ) K.r K, 554 B 5 8 2 K.p. V., 95 226 ( 93. 4 M.d.S., Rechtslehre, EinL, IV, 226 27. In this he may perhaps be forgetting that such a will may be creative, but perhaps he 'considers that in man it is so only in conjunction with arbitrium. God's will is free because spontaneous and not predetermined 57 n. ( 58-9 n.) (there being no temporal succession in God) ; Religion, 50 n.
*
.
= =
214
[XX
only of maxims, and so far may be identified with practical reason on its it alone is free. We are free in adopting maxims, and behind our maxims it is useless to seek further determining grounds of our actions. 1 It must not be thought from this that Kant is deliberately falling back on the liberty of indifference. He insists that we cannot define freedom merely as a power of choosing to act for or against the law, although 2 experience gives us many examples of the latter. The power of choosing to act against the law is not a necessary characteristic of freedom. He apparently thinks that such a power in a rational being is unintelligible, as indeed it is; and he even declares it not a power, but rather an incapacity. Freedom in relation to the inner law-making of reason is alone properly a power. All this is highly abstract and difficult, and it may be confused but
subjective side:
:
it all
serves to
show
that for
;
of obeying the moral law merely of saints, but of all apart from this there could be no recognition of duty and no moral
;
Kant we are free so far as we are capable and this on his view is a characteristic not men, and indeed of all rational agents for
responsibility for failure.
8.
Two
kinds of heteronomy
It may be objected that Kant has identified natural necessity with 3 heteronomy, and he has also asserted that heteronomy is present wherever the motive of action is other than purely moral. 4 If non-moral, and still more if immoral, actions are instances of heteronomy, can it be denied that they must fall under natural necessity? How can we maintain that they may be free? To answer this we must anticipate a little. According to Kant there are two points of view from which actions may be regarded. We may call these provisionally the point of view of the observer and the point of view of the agent. From the point of view of the observer all actions, moral and immoral alike, are instances of natural necessity. From the point of view of the agent these same actions, both moral and immoral, must be regarded as free, since he recognises, even in doing immoral actions, that he could obey the moral law and ought to do so. Kant does not hold that the very few morally good actions if there are such are wholly free, whereas all other actions are wholly determined. Nor could there be much to recommend such a theory. To this it may be replied that even if all actions from one point of view may be regarded as determined and so as heteronomous, there must be some sense in which from one and the same point of view presumably the point of view of the agent non-moral actions are heteronomous, while moral actions are not.
1 *
Religion, 21 n,
22 n. ( 7 n.). M.d.S., Rechtslehre, Einl. IV., 226-7
28-9.
*
Gr.,
446
==
79.
Gr. t 441
=-.=
72.
XX
9]

is
215
means that Kant, whether just, consciously or not, is using 'heteronomy' in two senses. Heteronomy in the case of inanimate objects means that their causal action is wholly
Such an observation
but
it
determined from without. This is not the case as regards human action from the point of view of the agent. In all human action will is active and spontaneous it does not indeed act for the sake of law as such, except in the case of morally good actions but it is said to go beyond itself and seek the law which is to determine it in the character of some
:
look at Kant carefully, we shall see that human never pathologically determined 2 or necessitated by sensuous motives: it is always merely affected or influenced, and this is the main reason why it is called arbitrium liberum or free
1 object or other. If
we
arbitrium or choice
is
choice. 3
Arbitrium is attributed to animals, presumably because they are moved by sensuous impressions and not merely by physical forces. Their
arbitrium,,
however,
is
not liberum but brutum, because their action
is
deter-
mined or necessitated, and not merely influenced or affected. There are indeed grave difficulties in supposing that from one point of view human action may be wholly free, and from another wholly
are trying determined. But these difficulties are not here in question. to discover only what Kant's doctrine is ; and there is no doubt whatever that for him at least from one point of view the heteronomy of non-
We
moral actions is compatible with freedom. 4 Human action, whether moral or immoral, is worlds apart from the motions of inanimate bodies or even the behaviour of animals and this difference is due to the presence in man of practical reason.
;
9.
Degrees offreedom.
Even if we decline to attribute to Kant the absurd view that moral action is wholly free and all other action wholly determined, there is still the question whether in all our actions we are equally free. Are there degrees of freedom? On this point Kant does not appear to have a clear and consistent view. At times he adopts a rigidity of moral attitude such as is fitting, if at all, in judging ourselves and not in judging others. may judge from one point of view that the telling of a wicked lie is completely determined by a man's unhappy nature, wretched circumstances, and past life. Yet from another point of view we none the less blame the agent we judge him on the supposition that the past can be wholly set aside; that in performing the action, reason is completely free and self-sufficient,
We
Compare Gr., 458 = 93. 6830; K.p.V., 32 = 144 ( = 57); M.d.S., RechtsI, 144; Religion, 49 n. = 56-7 n. ( = 58 n.). This freedom is different from the so-called psychological freedom in which reason affects actions by its concepts of what is remotely useful or harmful. See 4 above. 4 Presumably we must freely allow ourselves to be influenced by heteronomous laws.
1
Gr.,
441
K.r.V., lehre, Einl.
A534
= 6562; A8o2 = 213 = 12; Anthr. 9,
72.
2l6

;
[XX
whatever be the sensuous obstacles

guilt
is
and that in the moment of lying the

:
1 entirely his.
Here apparently there are no degrees of freedom as soon as we attain the age of discretion, we are all completely free. But quite a different, and much more human, view stands cheek by jowl with this. 2 According to this view the genuine morality of actions, their merit or guilt, is entirely concealed from us. We can judge only what Kant calls a man's empirical character but how much of this is the pure effect of freedom, and how much has to be ascribed to mere nature, to a fortunate or defective
;
temperament, no man can fathom and consequently no man can judge with complete justice. 3 Here surely there is a possibility of degrees of freedom. It is far too
;
to look at the rigid side of Kant and to ignore the humaner side, always differs equally compatible with his central doctrine. consciously from the doctrine of the Stoics and allows much both for
common
is
which
He
human ignorance and for human weakness. command of duty is 'Be perfect' and even 'Be
It
is
true that for

;
him
the
holy*
but
we must remem-
ber that although the ideal is absolute, yet as regards degree this can only be an imperfect obligation because of the frailty of human nature. It is our duty to strive after this ideal, but not (in this life) to attain it. 4 If we could progress continuously towards our ideal, we should have done our
duty.
K.r.V.,
K.r.V.,
A554-5 A55i n.
= 6582-3. = 6579 n.
Tugendlehre,
See also K.p.V., 99-100 231 ( 178-9). 8 See also M.d.S., Tugendlehre, 48, 474. 21-2, 446-7.
CHAPTER XXI
FREEDOM
i.
AS A NECESSARY PRESUPPOSITION
agents.
Freedom and rational
we have now described it could be established, then, according to Kant, the principle of autonomy, and so the supreme principle of morality, would follow by a mere analysis of the concept of freedom. If, however, we are to justify the moral principle, it is not enough
IF freedom as
merely to define freedom or to describe the characteristics which must necessarily be present in a will if it is to be regarded as free. It looks as if we have to show that every rational agent is, and indeed must be, free in the sense we have explained; 1 and this task certainly cannot be
regarded as an easy one. Apparently anything short of
this
would
fail
to
give us that justification of moral principle for which we seek ; for we have seen that the moral law must be valid for all rational agents, and for all men only in virtue of the fact that they are rational agents. 2
It is useless to appeal to any alleged experience of human nature; for experience of freedom, if it were possible (which it is not), would give us only a fact and not a necessary connexion of the kind which we seek. The plain man may indeed say that he feels himself to be free in action, and especially in moral action. But we are not entitled to derive from mere feeling a necessary connexion between being human and being free, and still less to extend our observation so as to cover all rational agents as such. Furthermore, on Kant's view, what is revealed to us by experience of our own successive states of mind is not freedom but necessity. Here as elsewhere a transcendental deduction must rest, not on experience, but on the insight of reason into its own necessary activity. 3
2.
The presupposition offreedom.

to
prove that a rational agent as such must be knowledge he regards as far beyond our human limitations. We might, however, be able to show that a rational agent as such can act only on the presupposition that he is free (only under the Idea of freedom) From the practical point of view of the agent, not the observer, this would be sufficient ; for the laws which would hold for a hold also for a being Known theoretically to be free must for purposes of action is free* The establishment of he that who the must act on being presupposition such a necessary presupposition would be enough to justify the moral
free.
Kant makes no claim

Such a claim
to theoretical
1
*
Gr.>
447
80.
Compare
K.p.V.,
29 =
Gr. t
143
52-3).
Gr. }
408 448
= 80 and footnote.
30-1
41 i-ia == 35.
[217]
2l8
[XXI
law and so to complete our task as moral philosophers. If a rational agent must act on the presupposition that he is free, he must act on the presupposition that he is under the principle of autonomy. In all this Kant is following out his doctrine that no Idea can give us knowledge of an actual or even possible reality. What it does is to prescribe a rule which reason by its very nature must obey and to set forth an ideal at which it must aim.
3.
Theoretical reason
this
and
its
presupposition.
At
point the argument takes a surprising turn.
Kant
bases his
case, not
merely on the nature of practical reason, but on the nature of

possibly conceive of a reason as being consciously directed
theoretical reason.
'We cannot
from outside in regard to its judgements.' 1 If a rational being were conscious of any such external influence, he would regard his judgements as determined, not by reason, but by impulse. Reason must if it is to be reason
regard itself as the author of its own principles independently of external influences. This is a strong argument, though it is seldom used in discussions about freedom. 2 It applies most obviously to a judgement which is the conclusion of an argument. If every judgement is determined solely by previous mental events and not by rational insight into a nexus between premises and conclusion independent of temporal succession, there can be no difference between valid and invalid inference, between reasoning and mere association, and ultimately there can be no truth. In that case determinism itself could not be accepted as true, nor could the arguments in its defence be accepted as valid. Kant seems to be correct in saying that any reason which is conscious of itself as reason must regard itself as reasoning (or as forming its own conclusions) in accordance with its own rational and objective laws or principles, and not by the influence of any external cause or bias. This
at all
remains true even although it may be reasoning about something given it from without, as, for example, so-called sense-data. That is to say, reason must regard itself as the author of its own principles and as capable of
functioning
according to these principles independently of external influences. This means, in Kant's terminology, that reason must regard itself as free, both negatively and positively, in the act of reasoning.
4.
Practical reason
and
its
presupposition.
If Kant's
argument holds of
that
is,
practical reason
1
theoretical reason, it holds equally of of a rational will or of reason as exercising
The italics are mine. E. Taylor, Freedom and Personality, Philosophy, and Personality Again, Philosophy, XVII, 65.
Gr.,
448
= 81.
Compare A.
XIV, 55 and Freedom

;
XXI
4]
FREEDOM
AS A
PRESUPPOSITION
causality. Here too a rational agent as such must in action presuppose his rational will to be the source of its own principles of action and to be capable of functioning in accordance with these principles. In other words he must in action presuppose his rational will to be free both negatively and positively that is, to be free from determination by desire and free to obey its own rational principles. To say this is to say that a rational agent can act only on the presupposition that he is free he must act under the Idea of freedom. This is the doctrine which we set out to establish and from which the principle of autonomy is said to follow analytically.
:
This doctrine need not be based merely on an inference from theoretical to practical reason though Kant may think that it is. It may rest also on the same kind of rational self-consciousness as does the previous may perhaps say that our insight into theoretical reason argument.
We
an insight into reason as such and consequently must cover practical reason as well; but the same insight into reason as such is surely found again in our insight into practical reason, and indeed must be found again if our conclusion is to be justified. I do not want to make too much of this, but I believe Kant to be saying more than that in acting we necessarily conceive ourselves to be free. Action is not a blind something which is preceded and succeeded, or even accompanied, by thought. Action is as intelligent and as rational as thinking. What distinguishes human action from animal behaviour, and still more from physiological functioning or physical movement, is that
is
also
we will in accordance with principles. I take Kant to be saying that a rational agent can act, just as he can think, only on the presupposition of freedom he must think and act as if he were free. The presupposition of freedom is as implicit in his acting as in his thinking; and unless we can act on this presupposition there is no such thing as action, and there
:
is
no such thing as will. As Kant himself puts it, 'The will of a rational 1 being cannot be a will of his own except under the Idea of freedom'. Human action cannot differ from animal behaviour merely in being accompanied by a conception of freedom if it differs at ail, it must
:
differ
by being
itself rational.
A rational agent must will his actions under
the Idea of freedom, just as he must will his actions as instances of a 2 particular principle or maxim. are there doubt No many today who will be content to deny that there is any such thing as action or as will, if by this is meant rational action and rational will. This view can be reasonably held, as Kant recognises, if we take up the standpoint of observers. But if we regard thinking from the same external point of view and it is arbitrary not to do so we shall equally be compelled to deny that there is any such thing as thinking. There are some who are willing to accept even this conclusion ; but if they do, it seems foolish to try to convince others of it by
argument.
Kant makes a point of considerable value by his appeal to the selfconsciousness of theoretical, as well as of practical, reason ; for we cannot and reject it in the other. lightly accept its verdict in the one case
1
Gr.,
448
= 81.
See Chapter VIII
5.
220

5.
[XXI
The
self-consciousness
of reason.
Kant's argument turns on what may be called the self-consciousness of reason in its own activities. Such self-consciousness is for him no mystical insight into the nature of some mysterious pure ego, as it is sometimes called, which can be known apart from its activity. It rests rather on a rational understanding of the principles manifested in the activity of reason as such. These principles are manifested, however imperfectly, in our every-day thoughts and actions; but we are able, on Kant's view, (i) to conceive them in abstraction from their empirical concomitants, and
(2) to understand both their internal necessity and their necessity as conditions of the whole activity from which they are abstracted. That is to say, we can, in virtue of our own rationality, understand both how these principles are necessary manifestations of reason as such, and also how they are necessary to our theoretical and practical judgements. In the case of theoretical reason these principles may for the present be regarded as what Kant calls the Principles of the Understanding 1 especially the three Analogies of Experience; but behind these, and behind the categories applied in them, there is the more ultimate concept of universal law as such. 2 This is the concept bound up with, and correlative to, that ultimate self-consciousness of reason which Kant calls the transcendental unity of apperception. If the self-consciousness of reason is inseparable from consciousness of the rational principles manifested in its own activity, what are these principles in the case of practical reason? The principles in question must obviously be what we have called 'objective principles of practical 3 reason', principles in accordance with which a rational agent, qua rational agent, would necessarily act, if reason had full control over passion. For the present purpose Kant ignores the objective principles of skill and self-love, whether because they are irrelevant to his argument or because they are regarded as merely hypothetical and as subordinate to desire. are left with the unconditioned objective principle of morality, the principle of acting for the sake of universal law as such. To agents only partially rational this principle must appear as a categorical imperative and in particular as the imperative of autonomy. If this is so and in the Critique of Practical Reason Kant recognises that it is so we cannot ascribe to rational agents a necessary presupposition of freedom independently of their recognition of the categorical imperative. If the Ideas of freedom and of moral autonomy are really distinct, the inference is not from freedom to autonomy but vice versa. T ought'
We
implies I can.'
1
Duty
implies freedom.
to
The
principles of
well.
2
See Chapter
be called the principle of objec88. This is identical with what Gr., 454 tivity in knowledge or the concept of an object in general. 8 See Chapter IX. Yet we must also remember that for Kant at least in his later
XXIV
Formal Logic and the Ideas of reason have

8.
be considered as
may
works
freedom
is
manifested in acting upon maxims.
XXI
6]
FREEDOM
AS A
PRESUPPOSITION
221
Kant may indeed be right in asserting that reason as such, and in particular theoretical reason, must ascribe to itself freedom to function in accordance with its own principles independently of necessitation from without. This may give us ground for suspecting that practical reason must ascribe to itself a similar freedom ; but the nature of this freedom is not intelligible without grasping the unconditioned objective principle of action that is, the principle of morality. It is a mistake to suppose as Kant may have supposed in the Groundwork that morality could be justified by a non-moral concept of freedom established without regard to any moral considerations. Our consciousness of the fundamental principle of moral autonomy can according to the Critique of Practical Reason be called a Tact of reason' it cannot be extracted by subtle argument from antecedent data of reason, such as consciousness of freedom ; for this is
:
not antecedently given. 1
6.
The position of the argument.
If
directly
on these grounds we must abandon the attempt to derive freedom from the self-consciousness of reason as such and thereby to
produce a transcendental deduction of the categorical imperative, how do we stand? Are we simply to say that Kant's moral philosophy, in spite of all its ingenuity, has ended in failure? It seems unnecessary to take up so pessimistic an attitude. If, as Kant 2 claims, he has given a correct analysis of the implications of all moral judgements, if apart from the categorical imperative and the principle of autonomy we must abandon all belief in moral goodness and in binding
judgement as it is to set aside the fact of theoretical judgement. A philosophy which can show that you must accept its principles if you think at all, is on strong ground and a philosophy which can show that you must accept its principles if you make moral judgements at all, is on almost
;
and in acting obligations, this in itself would justify us in accepting upon his doctrine. However much we may disagree with particular moral judgements, it is almost as hard for us to set aside the fact of moral
equally strong ground. Kant does not, however, rest his case merely on the fact that his philosophy gives the correct account of the implications or conditions of ordinary moral judgement. In the regress which he has made, every step seems to involve insight not merely into the relation between the conditioned and the condition, but also into the truth of the judgement which expresses the condition. 3 In particular he holds that as rational agents we have direct insight into the truth that a rational agent as such would necessarily act in accordance with a universal principle valid for
1 K.p.V., 31 143 ( 55-6). Kant still insists that the principle of autonomy is a synthetic a priori proposition, grounded neither on pure nor on empirical intuition. It would be analytic presumably in the sense that it would follow analytically if we could presuppose the freedom of the will. For this, however, as a positive concept we should require an intellectual intuition such as we cannot presume to have.
Compare
K.p.V., 32
143
56),
See Chapter
9.
222

This doctrine
[XXI
all rational agents. 1
self-consciousness of reason,
The
presupposition
still turns on what I have called the to this topic we shall have to return later. of freedom remains a necessary presupposition for
and
rational agents, although it is now based on recognition of the categorical imperative. It is possible to act on presuppositions which we do not know to be true ; but it is not possible to act on presuppositions which we know not to be true. Hence the presupposition of freedom is in need of defence; we must show at least that it is neither self-contradictory nor excluded by he experience. This is the task to which Kant now sets himself, although is relevance whose of considerations a in it variety by bringing complicates
not altogether
clear.
See Chapter
IX
4.
CHAPTER
XXII

i.
Side issues.
sensible (or
have now to concern ourselves with Kant's distinction between the phenomenal) world and the intelligible (or noumenal) world. The main reason why we must do so is that without this distinction we shall be compelled, on Kant's view, to deny the possibility of freedom and therefore to reject, not only his own ethical doctrine, but also the very possibility of morality. This seems to be the view of the Critique of Practical Reason. In the Groundwork, curiously enough, he does not make this explicit till after he has completed his arguments and, perhaps because he is seeking mistakenly to derive the imperative of autonomy from the presupposition of freedom, he approaches the problem of the intelligible world in quite a different way. He raises two apparently interconnected difficulties: (i) the question of what he calls 'moral interest'; and (2) the question of the vicious circle. These seem to be side issues at
;
WE
the best.
2.
Moral
interest.
His first question is 'Why should I be moral?' Granted that as a rational agent I must act on the presupposition of freedom, and granted that from the presupposition of freedom there follows, for finite creatures like men, the principle of autonomy as a categorical imperative, why should should I subject myself as a rational agent to this principle? actions to those whose maxims can be willed at the same time I limit
Why
as universal laws? And why should I attach so supreme a value to conduct of this kind that in comparison the claims of pleasure become negligible? 1
It may be said that to these questions Kant has given no sufficient answer. This objection he recognises, but his language does not make it clear whether he regards it as a sound criticism or not. He indicates once more that it would be no answer to say I was driven by some pathological interest: pathological interest cannot give us a categorical imperative. Nevertheless I must take an interest, a moral interest, in duty, and I must
my
be able to understand how this can happen for 'I ought' is equivalent to 'I will' for a rational agent as such, and appears as *I ought' only because 2 of the hindrances of passion. He adds that it is no answer to say that in fact I do take such a moral interest, since this interest, which he has else;
Gr.,
449
82.
See also Gr., 414
38.
[223]
224

itself follows
[XXII
where equatetf with reverence, 1 moral law is binding.
from the supposition that the
How
far
Kant
takes this objection seriously cannot be easily deter-
mined by the subsequent course of the argument. He does not return later to this argument explicitly, except at the very end of the book; and there he seems to say that these questions are really due to misunderstanding. The only answer to the question 'Why ought I to do my duty'? must be 'Because it is your duty. To expect any other answer is to deny the essence of duty and to suppose that we ought to do our so-called duty for some other reason, presumably from self-interest.
5
We can indeed say that a rational agent as such will necessarily act in accordance with a universal law equally valid for all rational agents, and that he ought so to act if he is irrational enough to be tempted to do otherwise. But this is the answer Kant has already given, and it rests on
the ultimate insight of reason into its own principles, beyond which we cannot go either as regards thinking or as regards action. Kant may believe he is going to add to this answer by saying that in so acting a rational agent is being his real and intelligible self, or at least is acting in accordance with his Idea of his real and intelligible self. This line of thought, however, though familiar to us in later idealism, is at least not conspicuous in Kant. It is in any case more satisfactory to regard his doctrine of our membership in the intelligible world as a metaphysical defence of the ethical doctrine already expounded and not as an attempt to prove an ethical doctrine on purely metaphysical grounds.
3.
The
alleged vicious
circle.
There is a second objection interwoven with the previous one. The objection is that in appealing to the presupposition of freedom we have have argued that we must be free because fallen into a vicious circle. we are subject to a categorical imperative ; and we have then proceeded to argue that we must be subject to a categorical imperative because we are free. 2 This circle is of the kind which inevitably arises if we follow up a 8 regressive argument by the same argument stated progressively. In the Kant a indicates, may establish necessary present case the argument, as and reciprocal connexion between the concept of being under the categorical imperative and the concept of being free ; but it does nothing to establish either the validity of the categorical imperative or the neces-
We
sity
it
is
of presupposing freedom. Why should we not reject both? need not suppose that Kant takes this criticism too seriously: 4 is the kind of objection which may be thought to arise. Nevertheless it but should not mention that he should make so much it, only strange
We
* *
See Gr., 401 n. See Chapter I
^2 n.
('it
Gr. 9
449
= 82
450
= 83
and 453
= 86-7.
illusory
9.
The use of the phrase *es schein? character of what is being said.
appears') often indicates in
Kant the
XXII
4]
225
of it, and should apparently admit that only by an appeal to the doctrine of an intelligible world can he succeed in escaping from the circle. In plain fact the objection totally misrepresents his argument. He never argued from the categorical imperative to freedom, but at least professed, however mistakenly, to establish the presupposition of freedom by an insight into the nature of self-conscious reason quite independently of moral considerations. Perhaps when he came to the objection he was beginning to see dimly that the presupposition of freedom of the will did really rest on moral considerations ; but it is surely unusual for a man to answer the sound argument which he has failed to put and to overlook the fact that this is irrelevant to the unsound argument which alone has
been
explicitly stated.
4.
The way of escape.
are now introduced to the doctrine of the two standpoints as a out of our alleged vicious circle. In order to escape from this we must ask Do we take one point of view of ourselves when we think of ourselves as free causes working according to our own laws, and another point of view when we think of our actions as observed events in the sensible world?' 1 The suggestion here is that there may be two equally legitimate points of view from which we may regard ourselves. From one point of view
We
way
regard ourselves as belonging to an intelligible world and so as while from another point of view we may regard ourselves and our actions as belonging to the world of nature. This may be a way of reconciling moral freedom with causal necessity, but how is it going to help us get out of our circle? There is clearly no hope of escape unless we can claim an independent insight into our membership of the intelligible world. The insight claimed must rest neither on the presupposition of freedom nor on the presupposition of morality for otherwise we shall merely have added another link to the alleged circle, which will remain incorrigibly vicious. As we shall see, Kant does to some extent appear to escape from the circle by concentrating his attention on the theoretical function of reason and thereby attempting to establish a necessary characteristic of reason as
free,
;
we may
such. But this is exactly what he had already in fact done in justifying the presupposition of freedom. Furthermore, it is by no means clear that he deliberately wishes to rest his case on insight into the activity of theoretical reason. At the end of his argument 2 he does suggest that a rational being must consider himself &y intelligence (presumably theoretical intelligence) to belong to the intelligible world, and therefore to act under the Idea of freedom, and therefore to be subject to the categorical 3 imperative of autonomy. Yet immediately thereafter, in claiming to have escaped from the circle, he does so on the ground that when we think of ourselves as free, we transfer ourselves as members to the intelligible
world.
*
Gr.,
450
83-4.
Gr.,
452
86.
Gr.,
453
87.
226
[XXII
In the section which follows and purports to give the answer to the question 'How is a categorical imperative possible?' we find the same appearance of hesitation. He begins again by maintaining that a rational being must as intelligence count himself as belonging to the intelligible world, and seems to infer that only so can he consider himself to exercise 1 causality and manifest a free will. Yet at the end of his whole argument he sums it up by saying that categorical imperatives are possible since the 2 Idea of freedom makes me a member of the intelligible world. Is he arguing from membership of the intelligible world to freedom or vice versa? Or is he merely establishing a universal and reciprocal connexion between the concept of membership in the intelligible world and the concept of freedom, a connexion which would do nothing whatever to break the alleged
circle. 3
not merely a technical point which it is a waste of time to the contrary, it raises a fundamental issue of principle. If Kant really supposes that he can start from theoretical reason and infer from this to membership of the intelligible world, and from this to freedom, and from this to a justification of the binding character of the moral law, then however much he may be escaping a vicious circle he is falling
This
is
consider.
On
into a fundamental error. It is manifestly impossible to deduce moral obligation from purely metaphysical or ep'stemological considerations which have nothing to do with morality. Metaphysical considerations about freedom and about membership of the intelligible world may be absolutely necessary to defend a moral
principle taken to be independently established, but they cannot by themselves establish it. Whatever be his view in the Groundwork, Kant himself was well aware of this when he came to write the Critique of Practical Reason. Our main task must therefore be to consider how far his argument is successful as a defence of freedom and so of morality.
5.
The two
standpoints.
The doctrine of the two standpoints takes us for the first time into the heart of the Critical Philosophy. Here we must face the difficulty that knowledge of the Critique of Pure Reason is necessary for a full understanding of what Kant is now trying to say. The main doctrines of that work and the arguments without which they are barely intelligible cannot be expounded in a short book devoted to ethics. On the other hand it cannot be assumed that readers interested in ethics have a detailed grasp of the and any commentator who tries to Critique of Pure Reason. Hence Kant follow him has to give a semi-popular and admittedly inadequate 4 exposition of his metaphysical doctrines. He does so partly by making based statements on his dogmatic previous work and partly by offering some rather simple arguments in support of his position. The result cannot be wholly satisfactory and must be taken only as an introduction
1
Gr.,
453
Compare
87. 3 above.
Gr., Gr.,
453
451
88.
84.
XXII
to the
5]
THEINTELLIGIBLEWORLD
227
whole philosophy in the light of which the argument for freedom must be judged. Broadly speaking, Kant's fundamental doctrine, his Copernican revolution, is the result of an attempt to work out a view commonly held in the Eighteenth Century and present in such diverse thinkers as Leibniz and even Hume, 1 though Kant alone made a serious and sustained effort to think out its ramifications in detail. This view is that the human mind contributes something to the world as known to it. If this were so, and if the human mind were simpler and more easily known than the world, then a philosophical study of it would enable us to discover certain fundamental characteristics which the world as known to us must necessarily have. It would at once follow, although this consequence was not clearly grasped before Kant, that we should have no reason to consider the world as known to us to be identical with the world as it is in itself. On the one hand we should have the world as it appears to limited and human finite minds, the world which may be described as the world of appearance or the phenomenal world. Since this world is revealed to us only through our senses, we may also call it the sensible world. This, ex hypothesi^ is the only world we could know but we should have to conceive, as it were behind it as its ground, the world as it really is in itself, the world of things as they are in themselves. This world may be described as the world of things -inthemselves or again as the intelligible or
;
noumenal world. This

tion,
last description clearly requires further justifica-
but at least it is not inappropriate so far as we may be said to conceive such a world in our thought, though it can never be known to us through our senses. Indeed, since on Kant's view all knowledge depends on a combination of thought and sense, we must say that the world of thingsin-themselves, though it can and must be conceived, cannot be known
Such a view assumes both that human minds function in the same sort of way and that the world which appears to them is one and the same world, however differently it may appear to different persons. The first assumption, however hard to justify, is one which every philosopher must make in the act of communicating his thought. But why should we assume that the different phenomena which appear to different observers are grounded in one and the same real world, even if it can never appear to them as it really is? The doctrine of the thing-in-itself is impatiently swept aside both by those who hold that there is no reality other than sens a (and perhaps sensibilia) and by those who hold that the human mind must know reality as it really is. Kant's whole philosophy is a continuous argument against the latter view, which he calls transcendental realism. As against the phenomenalists, who are content to regard sensa as the only reality, he is an empirical realist that is, he believes with Newton and common sense that space is filled with permanent substances which act causally upon one
:
by
us at
all.
Compare
Price, Hume's Theory of the External
Wodd,
p. 9.
228
THE CATEGORICAL IMPERATIVE,
[XXII
another ; and indeed one of his main aims is to demonstrate that this so. But while he holds that the world as it appears to human observers who possess both sense and intelligence must appear as a world of bodies in space and time, he regards this world of bodies as itself only an appearance and he seems to take for granted rather than to argue explicitly that it could not appear to us as one common world unless, in addition to there being minds which functioned in the same sort of way,
must be
there were also one and the same real world which appeared to them all, the world as it is in itself and not as it appears under the limitations of
minds. Without assuming a world as it is in itself, even if it cannot us, it is at least difficult for the phenomenalist to justify his if he does reject it; and if he is prepared to accept rejection of solipsism the existence, in some sense, of individual minds, it at least looks as if for him each individual mind must have its own private world, out of which it can never escape to a common or public world.
finite
be known by
The sensible world is better called the phenomenal world, since for Kant it is an appearance, not to sense alone, but to sense and thought; and if we are speaking Greek, we should contrast it with the noumenal
world. But if we prefer to speak English, we may contrast the sensible with the intelligible world, so long as we remember that the sensible world is not merely sensible. This terminology should not lead us into the error of thinking that on Kant's view there are two worlds. For him there is only one world though we can know the world only as it appears to us, what we know is the one real world as it appears to us. We must indeed conceive this one world (i) as it appears to our sensuous experience and (2) as it is in itself. 1 This is the reason why Kant speaks of two standpoints or points of view (i) the point of view from which we regard things as phenomena, and (2) the point of view from which we conceive things as noumena or thingsin-themselves. The phrase 'point of view' is not intended, as is sometimes thought, to indicate that either phenomena or noumena or both are unreal it is used only to guard against the supposition that there are two distinct and separate worlds. There is only one world, but it can be looked at from two different points of view.
: : :
6.
The argument from
the passivity
of sense.
In order to render
this doctrine intelligible to uninstructed readers
Kant supports it by an argument which he thinks likely to find acceptance

on the level of common sense or at least common judgement. 2 This argument is not used in the Critique of Pure Reason, and by itself it is, at most, persuasive rather than convincing. It fails to do justice, not only to the arguments for his doctrine, but also to the doctrine itself. We may call it the argument from the passivity of sense.
1
Gr.,
457
92; K.r.l
7
.,
xxvi;
A 38 =
655.
*Gr. 450-1
84-5
XXII
6]
On Kant's view our knowledge of an objective world depends on the 1 co-operation of two cognitive powers: (i) an active power of thinking and a means of by (2) concepts; passive capacity of sensibility that is, a capacity to receive what are today called sensa or sense-data. Kant himself calls them 'sensations' (Empfindungeri) to indicate that they are neutral 2 apart from thinking, they reveal to us neither the characteristics of physical objects nor the states of our own mind. Though there are
:
objections to this usage, it is perhaps no more question-begging than any other. He also calls them ideas or representations or presentations
though this is a term of wider application. physical objects only through sensations. Thus we think for without thinking we can know no physical object that the brown which we sense is the quality of a penny; and similarly with its other qualities. However active we may be in thinking, we do not make our sensations they are passively received and in that sense given. As naive realists we find no difficulty in assuming that the qualities of objects are thus given to us directly ; but a little reflection and a very little physiology is apt to disturb this assurance. Our sensations, we discover, are the result of a chain of causes emanating from the physical object and effecting changes in our nerve-endings and ultimately in our brain. As soon as we realise this, it seems as if there is no reason for supposing our sensations to give us qualities of the object, or even to be like the object in any respect. Yet since we know objects only through sensations, it looks as if we cannot know objects at all. What we call physical objects in space (including our own brain) become mere mental constructions based on sensations; and with this result the science of physiology seems to commit
(Vorstellungeri),
We know
suicide.
It may be objected that physiology cannot legitimately destroy the fundamental assumptions on which it itself is based. Nevertheless this is what it seems to do, and some philosophical solution of the difficulty is demanded. There may be different solutions. Kant's solution is that the world of bodies extending through infinite time and space is in fact a mental construction based on our sensations, but that these sensations must themselves be grounded on a reality which we can never know. Only on this supposition, and on the supposition that different minds function on the same principles, can we have before us one phenomenal world of physical bodies (including our own brain) in causal interaction with one another. Kant himself puts the position much more simply. He merely urges that since sensations come to us in ways we cannot control, they can give us knowledge of objects only as the objects affect us, and not as they are
1 1 pass over the power of imagination which is also essential both for 'taking up* or 'apprehending* sensations and for combining them with image?. Those who argue against Kant that sensation must be also active forget that the active side of sensation, the taking up of the given into consciousness, is recognised by him, but is attributed to
imagination.
Kant equates
we have no adequate translation for the neutral word 'Empfindung* with 'Sensation* only when it at the same time attracts attention to the state of the subject. See Anthr. 15, 153.
2
Strictly speaking,
it
230
[XXII
in themselves. 1 This argument, however, is telescoped, unless we distinguish clearly between (i) the sensations, (2) the physical object revealed by the sensations and supposed to cause them, and (3) the thing as it is in itself, which is the ground of the sensations and so of the physical object
The physical object affects our brain, and both the and our brain are phenomenal objects. The thing-in-itself if it affects anything affects our mind but this is not a causal relation. The thing-in-itself is the ground, and not the cause, of our sensations
revealed by them.
physical object
; ;
or,
better,
it
is
the
unknown
which we call a penny. 2 not by itself a convincing argument to maintain that our sensations, because they are passively received apart from any activity or choice of ours, must be given from some other source, and must reveal the way in which we are affected by something which can never be
It
is
limitations, appears to us as a region of space- time, the physical object
reality which, because of our human hard, brown, circular object in a particular
known by
sensations.
finite
minds whose knowledge
is
necessarily dependent
on
the other hand it seems reasonable to assume, at least as a hypothere is some reality which is other than the individual self and is what it is independently of our knowing it. It is difficult, if not impossible, to believe that this reality is composed only of sensa, or even of sensa and sensibilia. If Kant can also show, as he claims, that bodies
thesis, that
On
(or physical energy)
occupying space-time cannot be such an independent
reality, it at least looks as if reality as it is in itself cannot be known by minds ; nor does there appear to be anything antecedently finite
human
improbable in such a supposition.

Other arguments.
7.
are to understand Kant's doctrine properly, we must have in the three lines of argument used to support it in the Critique of Pure Reason. The first two rest on the contention that the human mind
If
we
mind
contribute something to reality as known to it. The first concerns space and time. Space cannot, as Newton supposed, be a real thing in its own right independently of what is in it. But equally
must
cannot be a mere quality of physical things, as Leibniz held (though with considerable qualifications) The only alternative is that it must be a form of our sensibility that is, a form under which alone our sensations, and consequently physical objects revealed by these sensations, can be given to our human minds. Our mind imposes space on what we sense in somewhat the same way that blue spectacles impose blue colour on
it
.
what we
It
is
Similar considerations apply to time. strange that Kant regards this argument as absolutely conclusive. If, however, he is right in holding that these are the only three possibilities, the plausibility of his argument is very great.
see.
we
451 =84. This does not mean that there must be a separate thing-in-itself for everything that choose to regard as a distinguishable physical object.
Gr.,
2
XXII
8]
231
The second line of argument rests on the necessity of thinking, if we are to know not mere sensa, but real objects, especially physical objects. In order to do this we must make use of certain categories, such as
substance and accident or cause and effect. These categories are certainly not given to our senses and cannot be derived by abstraction from what is so given. They are imposed by the nature of our thinking on objects
to us. 1 If both these lines of argument are sound, and if we assume that in knowing we are concerned with a reality other than ourselves which is not of our creating, then clearly we have no reason whatever for assuming that reality as it is in itself is characterised either as spatial or as temporal the other hand or as subject to such categories as cause and effect. reality as it appears to us must be so characterised, and a reality not so
as
known
On
characterised
we can
never know.
third line of argument goes farther and sets forth reasons for denying that reality as it is in itself can be so characterised. For if we suppose that the world which we know as spatial and temporal and governed by causal law is reality as it is in itself, then we fall inevitably
The
into a series of contradictions called by Kant 'Antinomies'. These contradictions disappear only if we suppose that the world which we know by means of our categories is not reality as it is in itself, but is only reality
must appear to human minds. This third line of argument should certainly be received with caution, though many of the so-called refutations of it are based on misunderstandings of what Kant's doctrine is. If this argument were sound, it would really clinch Kant's case; and it is, as we shall see later, very especially concerned with the defence of freedom. Incidentally this last type of argument clearly assumes that reality as it is in itself cannot be self-contradictory contradictions can characterise only appearance. Perhaps Kant even assumes positively that reality must be some sort of coherent system. He never denies that we must conceive reality as it is in itself and must do so by means of the pure 2 categories which have no reference to time: we have indeed no other there is no But for us means of thinking at all. knowledge without sense as well as thought, and it is impossible for us to sense a reality which is nontemporal and non-spatial. Hence we cannot know that there is, or even that there can be, such a reality. Though we must conceive such a reality by an Idea of reason, our conception cannot be elevated to the dignity
as
it
:
of knowledge.
8.
Conclusion.
In the light of these arguments we can understand better what Kant means by his two standpoints. Suppose we are asked 'Is the world governed throughout by causal
1
2
For a fuller statement see Chapter IX, Appendix. See Chapter IX, Appendix 3.
232
law?'
[XXII
The answer is 'Yes, certainly' if we are interested in a point of view concerned only with the world as it must appear to human minds. If, however, we are interested in a point of view concerned with the world as it is in itself independently of human knowing, the answer must be or at the very least that we have no reason for thinking 'Certainly not' so, and every reason for thinking the opposite. The two answers cease to be contradictory only when it is understood that they are given from different points of view. There is at least no contradiction in supposing that both answers may be true. It may be hard to give an exact parallel for such differences in point of view, and so far as I know Kant does not give any. But it may be
supposed that the physicist, the chemist, and the biologist might give seemingly contradictory answers to a question about living organisms, and yet the contradiction would disappear if we took into account the differences in their point of view. Similarly, if we were asked whether bodies were or were not solid, we might give one answer from a macroscopic, and quite another from a microscopic, point of view, and do so without any contradiction.
CHAPTER
XXIII
MEMBERSHIP OF THE INTELLIGIBLE
WORLD
i.
Inner sense and mental
states.
we may
to consider Kant's suggestion that from one point of view regard ourselves as belonging to the intelligible world and so as under the law of freedom, while from another point of view we may regard ourselves as belonging to the sensible world and so as under the causal law of nature. Unfortunately Kant's doctrine of self-knowledge is the most obscure and difficult part of his philosophy, and for the present purpose it has to be treated very summarily. Since all knowledge requires a combination of thought and sense, our knowledge of self must be based on what Kant calls 'inner sense' 1 in order to distinguish it from the outer senses, through which we become aware of physical objects, including our own body. Inner sense may be identified with the power of introspection, as it is sometimes called today. At every moment we are directly and immediately aware of our own states of mind. These states of mind may be, and perhaps always are, both cognitive and affective and conative. Our thoughts, our feelings, our inclinations, our desires, our resolutions, and also, quite explicitly, our ideas of physical objects all these belong to inner sense. 2 Inner sense is appropriately described as a kind of sense (though unlike the outer senses it has no special sense-organ) ; for it is always an immediate awareness of something individual given here and now. Starting from what is given to inner sense, we are able in thought to reconstruct our whole mental history in somewhat the same way as, starting from what is given to outer sense, we can in thought reconstruct the history of the physical universe. Our knowledge of ourselves is as empirical as our knowledge of the world, and it is equally based on what is directly given and passively received, not made by our knowing of it. If we accept Kant's previous argument from the passivity of sense, it follows that we can know ourselves only as we appear to human minds, and not as we are in ourselves. And since we know our mental states only as succeeding one another in time, the same conclusion follows if we 3 accept the argument that time can be only a form of our sensibility.
*Gr., 451
a
WE
have now
=85.
A357-8; B6y; B xxxix
n.;
214
8
/r.r.K,
13.
A34
= B5o;
M.d.S.,
Rechtslehre,
EM.
I,
See Chapter
XXII
6 and
7.
[233]
234

The
[XXIII
inner sense
succession of mental states which is all we are aware of through may be called the phenomenal self, the self as it appears. It must, however, be remembered that such a succession cannot be known through inner sense in itself; for this we require also imagination, memory,
and thought. Since thought, on Kant's view, must work by means of its categories, the phenomenal self as an object of knowledge must fall under the categories, particularly the category of cause and effect, which
own
concerned with objective succession in time. 1 Hence from one point of view we must regard ourselves that is, our phenomenal selves as belonging to the sensible world and so as subject to the causal law of nature.
is
2.
The noumenal
far
self.
So
appearances.
we know ourselves through inner sense only as a succession of Kant holds, as he did in regard to outer experience, that we
must assume something else to be the ground of these appearances. That 2 something else must be the I-as-it-is-in-itself. This must be conceived; but since on his principles it cannot be sensed, it can no more be known than can the thing-in-itsclf. We may call it the intelligible or noumenal self, and we must regard it as belonging to the non-sensible, and in that sense to the intelligible, or noumenal, world. At this point the difficulties become acute, and we can almost hear
the positivists shouting 'Take away that bauble.' This is not the place to expound, and still less to defend, Kant's doctrine in detail ; but here also, it may be observed, Kant's view, if strangely expressed, bears some resemblance to what may be described as the assumptions or prejudices of common sense. Besides the long mental history of ideas, feelings,
desires,
and
not be an
feelings, least difficult to avoid.
which is known as my phenomenal self, must there which knows this history, and which has these ideas, these and these desires? If this be a superstition, it is one which is at
volitions
I
must, however, admit that Kant's treatment of this subject is necessarily inadequate in the Groundwork, and also that it is far from clear even in his other writings. All we can attempt is to clarify, if possible,
We
some of the statements

1
in the passage
we
are
now
considering.
See Chapter XXII 7. There is some doubt as to the way in which Kant conceived the category of substance to apply to the self. We are not, he says, adequately informed either through experience or through inferences of reason whether man has a soul (in the sense of a spiritual substance dwelling in him distinct from the body and capable of thinking independently of it) or whether life may not rather be a quality of matter. 'Life* here is *the power of a being to act in accordance with its own ideas'. And again he says that reason as a theoretical power could quite well be the quality of a living corporeal being. See M.d.S., Tugendlehre, 19, and Rechtslehre, Einl.> 211=9. See also 3-4, 418 112 n. ( K.p.V., 9 n. i6n.) for 'life'. If there is nothing permanent in mental life, as Kant holds, and if the phenomenal self is to be known under the category of substance and accident, it is hard to see how he can regard mental states as other than qualities of a permanent material body. Compare K.r.V., 6275 ff.
O., 451 =85.
XXIII
3.
3]

affects itself.
1235
The mind
According to Kant if man knows himself through inner sense, he knows himself only 'through the way in which his consciousness is affected'. Here again we have the assumption that since mind is passive in sense, it must be affected by something. In dealing with the objects of outer sense we supposed that our mind must be affected by a thing-initself which is other than our self. In self-knowledge we come up against the strange paradox that the mind must somehow be affected by itself. Kant's efforts to explain this paradox are difficult to follow. 1 He 2 and it is gives ordinary 'attention' as an example of what he means
;
easy to see that attending to something does affect our mental consciousness and brings a different content into the phenomenal self. Indeed all our thinking and all our willing may be said to do the same. It may be objected that we are now passing from the realm of paradox into that of downright contradiction. We have said that all our thoughts and volitions arc states of mind which we know through inner sense. How can we also regard them as activities which by affecting consciousness make the phenomenal self what it is? There is certainly a difficulty here, but is it arbitrarily created by Kant? If we consider, for example, a historian, it looks from one point of view as if he stood above the battle as a timeless observer assigning to each historical event its place in a causal series which he understands. Yet when we turn to consider what he has done in composing his history, this too appears as an historical event in the same causal series, and in no way different from any other historical event. But in saying this we seem again to be ourselves above the battle only to find on reflection that our judgement is in turn something to be judged as on the same level with its object. Every man is, as it were, the historian of his own life, and it looks as if Kant were right in saying that we must regard ourselves from two different points of view. What we know as an object is always the phenomenal self, but the knowing of it seems to be something different; and when in turn we regard this knowing as an event at a particular time, we do so in virtue of a fresh knowing which seems, as it were, to be above events and above
And so it goes on through some infinite regress in which we can never get rid of the knowing which is itself not known as an event, although it makes a change in our consciousness, a change which can be known to farther reflection as a mental state. This curious contrast runs through all our thinking, which can always be considered from two different points of view. The scientist contemplates nature almost as if he were a god; yet from another point of view all his science is only a part of nature. The psychologist contemplates human
time.
1 a
Compare
K.r.V.,
K.r.V.,
6155
ff.
mind.
in
it
may be directed either to outer objects or to states of of the difficulties about introspection is that the effort of attention present tends to alter the consciousness which we seek to introspect.
6156-7
n.
Attention
One
236
[XXIII 4
nature and human thought as if he alone were a divine judge; yet from another point of view all his psychology is just another piece of human
that although we can turn back and contemplate then ceases to be thinking, and becomes merely thought. It is no longer thinking, but an object thought about and different from the thinking which thinks about it. In the terminology of Gentile it has ceased to be pensiero pensante and has become pensiero pensato. Hence our thinking appears to be an activity which affects our consciousness and is thereby the ground of a state of mind which can be known through inner sense (together with fresh thinking) but as known through inner sense it has become a state of mind which is no longer an an object known, rather than itself an act of thinking. activity Perhaps this is not so surprising as it seems. We can see everything within the range of our vision except the place from which we see. If we step back a little, we can see this place also, but we do so again from another place which we cannot see.
is
thought. The strange thing
our
own
thinking,
it
4.
Knowledge of mental
activity.
It may well be felt that there must be something wrong about all this. Kant's view, it may be said, is perhaps just intelligible as regards knowledge of outer things, but as regards self-knowledge it is merely fantastic. Let us try to put the matter by means of a rather crude analogy. According to him we have before us in outer sense some sort of semitransparent screen on which images are thrown from the other side and can be seen by us. We must suppose that something throws these images on to the screen, perhaps some kind of cinematograph but as we cannot get to the other side of the screen, it is reasonable enough to maintain that we can never know what this something really is. In the case of selfknowledge the situation is quite different; for there the cinematograph (or whatever it is) is on the same side of the screen as I am, and indeed I am supposed to be the person who is turning the handle. How then can it be maintained that I still know only what appears on the screen and can have no knowledge of the reality which is its ground? 1 We may put the same point without the aid of analogy. If my activity 'affects' my consciousness and thereby is the ground of a state of mind which is known through inner sense, surely I must know something of my own activity as well as of the resulting state of mind. Kant does not overlook this point. He says that pure activity in a
;
agent comes into consciousness immediately and not through 2 affecting the senses. This looks a little like what Professor Alexander called 'enjoyment' as opposed to 'contemplation'. It seems to be a special
rational
1 Perhaps to complete the analogy we should add that what is thrown on the screen from the far side is not visible till something is done from my side. But this is doubtful, since Kant holds that it is possible to have outer sense as in the case of animals
without inner sense (and consequently without consciousness of time) *Gr., 451 =85.
XXIII
4]
MEMBERSHIP OF THE INTELLIGIBLE WORLD 237
kind of self-evidence of which we should like to have a much fuller account. 1 The name which Kant gives to this kind of awareness is 'pure 2 apperception'. Kant does not believe that this pure apperception can by itself give us knowledge of ourselves. In all knowledge of objects, including knowledge of ourselves as objects, there must be present both sense and thought, both passivity and activity. Thought is the source of conceptions, and sense of intuitions. The self-consciousness of thinking is only a consciousness of the principles (or conceptions) in accordance with which it functions. We can separate out by abstraction these principles of thinking, but what we have then is logic and not psychology, an awareness of the
timeless principles in accordance with which thinking as such necessarily function. If we seek to know anything individual, we
;
must must
bring in the empirical element always present in actual thinking but as soon as we do this we bring in sense, and in particular inner sense ; and consequently knowledge of ourselves as individuals, and of our thinkings as individual events, is always consciousness of the self as it appears to us in time, and not as it is in itself. It looks as if the knowing self is in a curious way different from the self as known to be a succession of mental states in time. Although they are the same self regarded from different points of view, the first self seems to be the reality in which the second self is grounded. 3 Yet, when we seek to know the knowing self, we seek in vain, and always apprehend
only some
eyes,
new mental states instead. Similarly if we seek to see our own we do so in vain, and always see something else instead. Nevertheless
conceive our self as a power of binding together given sensa and in accordance with certain principles whose necessity we underimages stand. This is why Kant speaks of the synthetic unity of apperception as the highest point to which all use of the understanding must be attached. 4 In all this we must remember that Kant is concerned with the kind of unity or identity the self must have in order to know that is to experience
the physical world and its own mental history. This problem should not be confused with another and quite different problem that of discovering the kind of unity or identity which the self must have in order to be known
as a
self.
5
we must
Kant does not conceive of as an eternal substance 6 pure activity of synthesis in accordance with certain necessary principles. One difficulty of this is that pure activity, as merely conceived in abstraction from all its sensible accompaniments, must be conceived as timeless and must be manifested only in (or even identified
self
The knowing
but as an
activity, a
I 9. K.r.V., Ai 15-16; 6153; Anthr. 7, 141-2. contents of the phenomenal self are also partly grounded in a reality, a thingin-itself, other than the self. 4 K.r.V., Bi34n. The central defect of Caird's interpretation is that he always assumes the analytic unity of apperception to be the highest point and equally groundidentifies this analytic unity with the tautological proposition 'I am F. lessly 5 Compare my paper on Self-Identity; Mind, Vol. XXXVIII, N.S. No. 151, especially
1
See Chapter
The
pp. 328-9. 6 By a pure activity
Kant means one unmixed with
sense.
238
[XXIII 5
with) timeless principles of synthesis whose necessity we understand. The proposal that we should conceive of an activity as timeless is one which is not unnaturally nowadays accepted with some reluctance. 1
hand our actual empirical thinking is successive and be successive as Kant recognises; 2 it appears to us as a succession of mental states in time. But on Kant's view there is present in all empirical thinking certain necessary principles of unity which are themselves timeless, although they are successively manifested in time because they are, as it were, embedded in empirical data. The difficulty, if we approach the problem from this angle, is that it is hard to see why such principles (together with the fundamental principle that there must be one necessary synthetic unity in all our thinking) should be regarded
the other
to
is
On
known
as themselves timeless activities.
To
this
paradox we must return
later. 3
5.
The
activity
of reason.
In virtue of whatever pure activity may come immediately into conman must account himself as belonging to a world which Kant here describes, not merely as the intelligible world, but as the 'intellectual' world one, that is to say, which is intelligent as well as 4 intelligible, perhaps one which is intelligible because it is intelligent. This is a big step forward; but before we consider it, we must be clear about the particular pure activity which Kant has in mind. The pure activity on which he lays stress is the theoretical activity of reason, as it was in his previous discussion of freedom; 5 but this time theoretical reason is taken in its special Critical sense as a power of Ideas. 6
sciousness
Why does Kant single out this special function of theoretical reason concerned with Ideas of the unconditioned? Reason in its most general sense as the higher faculty of cognition 7 has its own principles, which Kant seems to find in the forms of judgement and of syllogism recognised by the traditional logic, but which 8 might be described in a more convincing way. In this general sense reason must regard itself as the author of its own principles and as functioning in accordance with these principles, and this was our ground for 9 saying that it must function on the presupposition of freedom. On Kant's view these principles (as manifested in the forms of judgement) determine certain necessary characteristics which all objects must have if they are to be known as objects. 10 They are then called the categories of the understanding, the word 'understanding* being here used in its
as
1 3
8
7
Gr., 451 = 85. See also Chapter 4. Chapter XXVI 8. 6 Chapter IX, Appendix 4 and 5. Chapter XXI 3. Chapter VIII 2. See K.M.E., I, 553-4, and especially Reich, Die Vollstdndigkeit der kantischen
4
Compare K.r. V., Atf
654.
K.r. V.,
Aa 10
6255.
XXV
Urteilstafel
9
Gr.,
10
448 = 81. Chapter IX, Appendix
2,
and Chapter XXII
7.
XXIII
5]
239
Understanding must conceive its activity to be spontaneous, and consequently must function under the presupposition of freedom; but although its categories are the product of pure rational activity and are not derived by abstraction from what is given to sense, they seem in Kant's eyes to suffer from some defect for purposes of the
special Critical sense.
present argument. This defect concerns, not their origin, but their application they serve only to bring sensuous ideas under rules and thereby to unite them in one consciousness. 1 As an example of the rules in question we may take the principle that every event must have a cause, a principle which applies, and applies only, to sensible events in time. All this is familiar to students of the Critique of Pure Reason and must here be stated dogmatically. Thus understanding, however active or spontaneous, and however distinct from sense, which is passive, is nevertheless closely bound up with sense and is directed to the knowing of sensible objects. Reason suffers from no such limitation. In its special Critical sense reason shows a spontaneity so pure that by means of its Ideas it goes far beyond anything which sensibility can offer it as an object. It conceives the unconditioned and is able to show that if we think of the sensible world as unconditioned, we are bound to fall into contradictions. These contradictions it resolves by treating the sensible world as conditioned and ascribing the unconditioned to the intelligible world and thereby it confirms the distinction, which we must make on the level of understanding, between the sensible and intelligible worlds. In so doing it performs its highest office and is able to set limits for the understanding. 2 It is not quite clear what Kant gains by his concentration on this special function of pure reason indeed he himself goes on immediately to speak as if his argument were based on reason in its most general sense, in which it is opposed to our lower powers, certainly those of sense and perhaps also those of imagination. Perhaps he is influenced by the fact that the Idea of the unconditioned forces us beyond the limits of experience and compels us to suppose that the unconditioned must be found in thingsin-themselves. 3 Perhaps in this respect reason is shown most conspicuously 4 to be an activity wholly distinct from sense. Perhaps a further ground of Ideas is the closest approximaas a reason that theoretical be power may tion to pure practical reason and so makes the passage from the theoretical to the practical more easy to follow. But in spite of all this the argument seems to hold, if it holds at all, for theoretical reason in all its forms; for all reason must assume itself to be a power of spontaneous activity in accordance with its own principles, whatever be the objects to which
:
it is
directed.
Or., Gr.,
452 452
4,
= 85. = 86;
xxi.
Appendix
8
K.r.V.>
K.r.V., B xx-xxi (including the and Chapter XXII 7. *
footnote). See also
Chapter IX,
6.
But see Chapter IX, Appendix
24O
6.

Membership of the
intelligible world.
[XXI II 6
Kant's conclusion
intelligence
This,
it
is that a rational being must regard himself qua as belonging, not to the sensible, but to the intelligible, world. 1 must be remembered, is only a necessary presupposition of think-
knowledge. It should also be noted that the self, both here and elsewhere, 2 seems to be regarded as belonging to the sensible world, not only so far as it is known through inner sense, but also so far as it is capable of receiving sensations passively. Hence a rational being has two standpoints from which to consider himself. From both standpoints he can know the laws this emphasis on laws should be noted of the use of his own powers and consequently the laws of all his actions. 3 So far as he regards himself as belonging to the sensible world, these laws are the causal laws of nature. So far as he presupposes himself to belong to the intelligible world, these laws are principles of reason which are not empirical and are independent of
ing, not a claim to
nature. If we confine ourselves to laws or principles, Kant's contention seems to be wholly sound. The difficulty lies in the alleged membership of the
conceived as something more than being under the principles of reason. From this membership he goes on, whether by inference or by further insight, to maintain that his doctrine applies equally to a rational being assumed to exercise causality in virtue of his reason. Man as a rational being must conceive himself to will, as well as to think, in accordance with independent principles of reason. This means, as we have seen already, that he must conceive himself as acting under the laws of freedom and so under the principle of autonomy and the categorical imperative of morality. We need only add that it is as belonging to both the sensible and the intelligible worlds that the principles on which as rational beings we should necessarily act appear to us as imperatives on which we ought to act. Here again the self seems to be reckoned as belonging to the sensible world, not only as an object of inner experience, but also as possessed of lower capacities in this case the capacity for feeling and desire as
intelligible world, if this is
opposed to a rational
will.
7.
The
principles
of reason.
Kant has still some further points to add to his argument, and we must consider these before we can attempt to assess its value. But whatever we may think about the argument as a whole, he is on very strong ground in maintaining that a rational being must suppose himself able both to think and to act in accordance with principles of reason which are other
l
Gr.,
452
86.
to pass already
Here Kant seems
from
85. Gr.,4 5 i theoretical to practical activity, apparently
by way of inference.
XXIII
7]
MEMBERSHIP OF THE INTELLIGIBLE W O R L D

effect. It
241
than laws of cause and
that we could, as it were, infer ples of logical thinking; and if Kant's doctrine is to be upheld, there must be as sure an insight into the principles of action as into those of
would indeed be a mistake to suppose the principles of morality from the princi-
thought. Nevertheless the presuppositions of moral action receive strong confirmation if similar presuppositions are present in all thinking as such. In particular we are given a possible answer to the scientist or psychologist who asserts that all this talk of rational moral principles must be nonsense since the conduct of man is determined by causal law just as much as the movement of the planets. Kant's answer is that from one point of view this contention is correct, but that there is another point of view ; and furthermore that this other point of view is necessarily taken by the scientist himself even in asserting the truth of determinism. For the determinist is himself assuming that his own assertion is made in accordance with principles of reason and not merely as a result following from a causal series of sensa and feelings and desires. It is only on this supposition that his assertions have any claim to be true and if a supposition of this kind is legitimate for the thinker, it cannot be dogmatically set aside as necessarily illegitimate for the moral agent.
;
CHAPTER XXIV
HOW
i.
IS
POSSIBLE?
The
deduction.
E have now come to the end of Kant's preparatory exposition, and we naturally expect to be given his transcendental deduction of the categorical imperative. We are confirmed in this expectation by the heading given to the section we are about to study. It is entitled 'How is a categorical imperative possible?'
1
follows
Kant, as is his habit, gives us a preparatory exposition and up by what may be called an authoritative exposition, he expects us to read the authoritative exposition in the light of the preparatory one; and at times the authoritative exposition would be hardly intelit
When
2 by itself. This is what happens in the present case. He a brief such and almost gives perfunctory summary of his previous argument that one can hardly avoid a feeling of anti-climax. He does, how-
ligible if it stood
ever, add one seemingly new point, and he follows up his whole argument, as usual, by an appeal to ordinary moral judgement for confirmation.
2.
The
additional argument.
3 obscurely expressed, but its aim is apparently arguments. Having defined freedom in such a way as to identify it with autonomy, he has maintained that a rational being must think and act under the presupposition of freedom. This presupposition, he has insisted, is neither self-contradictory nor excluded by experience once we grasp the necessity of considering reality rational agent must from one point from two different points of view. of view regard himself as belonging to the intelligible world and as acting under the Idea of freedom. From another point of view he must regard his actions as belonging to the sensible world and as subject to the causal laws of nature. If I were merely a member of the intelligible world, my actions would conform to the principle of autonomy. If I were merely a member of the sensible world, my actions would conform to the heteronomy of nature.
His
new
contention
is
to
clamp together
his previous
1
2
Gr.,
453
87.
of this is the second and authoritative exposition of the transcendental deduction of the categories in the first edition of the Critique of Pure Reason; Ai 15 ff. 8 The obscurities are dealt with in an appendix.
An example
[242]
XXIV
3] IS
CATEGORICAL IMPERATIVE POSSIBLE?
243
two points of view were equal and co-ordinate, it is hard to see where this would take us. Kant's new argument seems intended to show that one point of view must be subordinate to the other. The intelligible
If the
world
is the
ground of
the sensible
world and of
its
laws. 1
Hence
my
free will
acting under the principle of autonomy can, as belonging to the intelligible world, be none the less the ground of actions which, as belonging to the sensible world, are under the causal laws of nature. The relation between my free will and my determined actions is in no way unique, but is only a special case of the generic relation which must be conceived to hold between the intelligible and the sensible world.
3.
The
conclusion.
The
rest of Kant's argument follows the beaten track. 2 In spite of the fact that I must recognise myself as a being belonging
to the sensible world, I must also recognise myself, qua intelligence, as 3 'subject to the law of the intelligible world, that is, to the law of reason, which in the Idea of freedom contains the law of the intelligible world'.
This law is identical with the principle of autonomy, and all my actions would necessarily conform to it, if I were only a member of the intelligible world. I must, however, regard myself as also a member of the sensible world and so as affected and hampered by desires and inclinations.
It follows that I
imperative for
me and must
must regard the principle of autonomy as an consider actions which accord with this
principle to be duties. This doctrine is not new to us, 4 but I believe though with some that Kant holds it to be fully intelligible only when we have hesitation seen that the intelligible world is the ground of the sensible world and of its laws, and consequently that a rational will can be the ground of actions which from one point of view are causally determined. Kant might seem to make a new point in so far as he appears in this passage to identify the law of the intelligible world with the law of reason or of freedom. But he must not be thought to claim knowledge that the intelligible world is necessarily characterised by freedom in the positive sense. He is not arguing that since we conceive ourselves as belonging to the intelligible world and consequently as subject to its known laws, we must therefore conceive ourselves to be under the law of freedom. know nothing of the law of the intelligible world apart from our knowledge of the necessary principles of our own reason. 5 The most we can be said to know or at least to think of the intelligible world is that it cannot
We
1 Gr.y 453 = 87. See also G>., 461 97, where subordinated to the character of the thing-in-itself. 2
it is
said that appearances
must be
Gr., 453 It is possible to omit the words 'the law of as some translators do. See Gr.y 414 and compare also Gr., 455 89. 5 In regard to his pure activity a rational agent must account himself as
3
== 88.
= 378
=
belonging
to the intellectual world, of which, however, he has no further knowledge; Gr. 3 451 See also K.p.V., 43 74). 157-8 (
85.
244
[XXIV4
be under the law of nature, and this conception is purely negative. Kant may indeed hold we are justified in believing that the intelligible world is also an intellectual world, but this is based, mainly at least, on our moral insight and so cannot be used to justify an ethical doctrine.
The failure of the
deduction.
4.
It seems clear enough that Kant's argument has failed as a deduction of the supreme principle of morality. Certainly there can be no question of a deduction in the common meaning of that word we cannot by inference derive morality from the presupposition of freedom, and still less can we by inference derive the necessity of presupposing freedom in the positive sense from the presupposition that we are members of an intelligible world. Even in Kant's sense of a 'justification', we cannot justify morality by anything other than itself: there can be no higher principle in the functioning of a rational agent than the principle of
:
morality. This objection by itself is fatal, but in any case we have no independent insight into the alleged necessity for presupposing freedom. Kant is indeed ingenious and, I think, sound in suggesting that freedom (in the sense of a power to accord with the principles of reason) is a necessary presupposition of all thinking. This may serve as a defence of the presupposition of a similar freedom in action, but it is not sufficient to justify
can
this presupposition. The justification for presupposing freedom of the will rest only on our insight into the principle of moral action, which
therefore cannot be derived from the presupposition of the freedom of the will. Even if the presupposition of freedom could be established independently of moral insight, our difficulties would not be over. What we are trying to show is that a rational agent as such would necessarily act in accordance with the principle of autonomy, and ought so to act if he is irrational enough to be tempted to do otherwise. This is a synthetic a priori proposition, and we therefore require a 'third term' to justify our statement that there must be a necessary and universal connexion between
being a rational agent and acting only on maxims which can at the same time be willed as universal laws. If it is surprising that a mere Idea, like the Idea of freedom, should function satisfactorily as a 'third term' for this purpose, Kant is perhaps justified in asserting that from a practical point of view the same law must be valid for a being who acts under the Idea of freedom as would be valid for a being known to be in fact free. But even if this is so, his attempt to establish the requisite connexion between the Idea of freedom and action in accordance with a self-imposed universal law suffered, as we saw, 1 from very obvious weaknesses. If we are to defend him, we should have to show that these weaknesses could be overcome, and it may be doubted whether this is possible. The chief difficulty, however, concerns the very assertion that a
1
Chapter
XX
5.
XXIV
5] IS
PO
S S
BLE
245
must as such think and act under the presupposition of freedom. Even if we do not doubt and I see no reason to doubt the truth of this assertion, we must still ask what is its logical character. Have we here a new synthetic a priori proposition? If so, our original difficulty is merely postponed, and we have to look for yet another 'third
rational being
term' to connect the subject and predicate of this new proposition. This, for reasons already given, we are hardly likely to find in our Idea of membership in the intelligible world ; and even if we could find it there,
not merely have provided yet a fresh synthetic a priori proposito seek still another 'third term' in order to connect a rational being with the Idea of membership in the
should
tion,
we
and should we not be obliged

world?
intelligible
unless
would go on indefinitely, stage, to direct insight into the necessary activity of a rational being as such ; and in that case we should have 1 ultimately to rely on a proposition which in some sense is self-evident.
It looks as if this process
we can come,
at
some
5. Direct insight into the principle
of autonomy.
If we admit the possibility of direct insight into the necessary activity of a rational being as such, do we really require Kant's elaborate machinery in order to establish the principle of autonomy? As he himself 2 says, the human mind does not always at first find the shortest way towards its goal. May we not have direct insight into the principle of autonomy itself? And is not this really assumed by Kant's argument? If we consider theoretical reason, a rational being must have direct insight into the principles of rational thought and must conceive himself to be capable of thinking in accordance with these principles. Furthermore, he must conceive these principles to be valid for all rational beings as such: whatever these principles may be in detail, they must be conceived as having universal validity. Unless we accept this, there is an end to all rational discourse and indeed to anything that can properly be called doctrine applies equally to practical reason. A rational agent must have direct insight into the principles of rational action and must conceive himself to be capable of acting in accordance with these principles. Furthermore, he must conceive these practical principles to be valid for all rational beings as such they too must be conceived as having universal validity. Yet to say this is to say that a rational agent as such will necessarily act on a principle universally and unconditionally valid for all rational agents as such. This proposition is identical with the principle of autonomy which Kant has discovered by analysing the implications of ordinary moral judgement; for the principles of rational action, like those of rational thought, are not imposed on a rational being
thinking.
3
The same
M.A.d.N., 476 n. theory conceive an intuitive understanding different from our own, but at least we must suppose the principles of rational thought to be valid for all human rational beings, if not also for all finite rational beings. See Chapter IX, Appendix 6.
I
*
We may on Kant's
See Chapter
9.
246

principles of his
[XXIV
from without but are the necessary and in that sense are self-imposed.
If this contention
is
own
rational activity
admitted as regards theoretical reason and denied
We might well be content with Kant's principle of autonomy, if it could be shown to be the necessary condition of all moral judgements and so of all moral actions. Nevertheless the principle becomes greatly strengthened if we can legitimately claim that as beings who are themselves rational agents we have a direct insight into the necessity of the principle when considered in itself.
6. Is the principle
as regards practical reason, the onus of proof surely lies least appear to be making an arbitrary distinction.
on those who at
of autonomy analytic?
Even if we claim as rational agents to have direct insight into the principle of autonomy, we have still to face the question whether the principle itself is synthetic or analytic. In spite of Kant's repeated assertions to the contrary we should not reject without consideration the possibility that the principle of autonomy may be an analytic proposition. Certainly the proposition that a rational agent as such will necessarily act rationally seems to be analytic. It is perhaps not so clear whether we must regard as analytic the proposition that to act rationally is to act in accordance with a principle unconditionally valid for all rational agents as such. The latter proposition certainly seems to rest on direct insight into the concept of rational activity as such, and if this is enough to constitute an
analytic proposition, then an analytic proposition it will have to be. If, it is claimed that the subject-concept present in every analytic proposition must be purely arbitrary, then we must insist with Kant that our proposition is not analytic, but synthetic, and also that, since it states a necessity, it must be a priori it rests, not on the analysis of an arbitrary concept, but ultimately on direct insight into the necessary activity of a rational being such as we ourselves are.
however,
discussed.
question raises problems of logic which cannot here be can only say that if this is an analytic proposition, it is an 1 analytic proposition about reality, The really essential point which should not be obscured is that in the
The whole
We
Kant claims a special and direct insight into the necessary activity of a rational being as such, no matter what be the logical character of the propositions in which this insight is expressed. In this respect, as he himself recognises elsewhere, our insight into the principles of practical
last resort
is on the same footing as our insight into the principles of theoretical reason. Indeed our consciousness of pure practical reason and of pure theoretical reason alike springs from our consciousness of the necessary and unconditioned principles which a rational being as such must follow in action and in thought. 2
reason
See Chapter XII
a.
K.p.V., 30
140
= 53).
XXIV
I
7] IS
PO
S S I
B L E ? 247
need hardly add that direct insight into the principles on which a must act could not be attained without personal experience if we may call it so of ordinary moral actions and of ordinary moral judgements. We have gradually to separate these principles from their empirical concomitants but when we have successfully disentangled them, we may find in them a necessity and intelligibility such as could not characterise any mere empirical generalisation. In this respect also the parallel with theoretical reason is complete.
rational being as such
;
7.
The imperative of autonomy.
So far we have considered the principle of autonomy only as a principle on which a rational agent would necessarily act, if reason had full control over his passions. It may be suggested that we have still to explain how this principle can be an imperative, a principle on which a rational agent ought to act if he is irrational enough to be tempted to do otherwise. Is it
its special character as a categorical imperative that Kant describes this principle as a synthetic a priori proposition? The answer to this question is in the negative. Once we have accepted the principle of autonomy as one on which a rational agent as such would necessarily act, we require no further synthetic a priori proposition in order to assert that it is also a categorical imperative for imperfect rational
perhaps in
agents.
Kant's whole procedure bears out this view, which is also confirmed his treatment of hypothetical imperatives. Once he has established the principle that a rational agent as such, if he wills the end, must necessarily will the means, Kant finds no difficulty perhaps he should have found more in turning it into a hypothetical imperative he takes it for granted that if anything is what a rational agent as such would necessarily do, it is also what a rational agent ought to do, should he be tempted to do otherwise. Exactly the same assumption is made in the case of the categorical imperative, and the passage from a moral principle to a categorical imperative is in no need of further justification, although it may be defended by the contention that the intelligible world is the ground of the sensible world. If a hypothetical imperative is an analytic proposition, this is because it is founded on a principle which is analytic and if a categorical imperative is a synthetic a priori proposition, this is because it is founded on a principle which is itself a synthetic a priori proposition.
by
8.
The
objective principles
of reason.
Kant's justification of the categorical imperative is reduced in the end that a rational agent as such has direct insight into the principle of autonomy as a principle on which -a rational agent, with full control over his passions, would necessarily act. That is the fundamental contention from which the categorical imperative follows. This contention is
to this
248
[XXIV
supported by a similar insight of theoretical reason into its own necessary principles, and the whole doctrine is claimed to be the only one which can render ordinary moral judgements intelligible. The latter claim would receive strong confirmation if the doctrine were worked out into a complete system, though this confirmation by itself would be by no means a certain proof. 1 We need not be surprised at this ultimate claim to direct insight it accords both with Kant's general view that reason must be transparent to itself, and also with his moral view that the common reason of the plain man is an adequate guide to conduct. In the case of practical reason the principles into whose necessity we have direct insight are the principles which are at the base of our five
:
formulations of the categorical imperative, and in particular the principle of autonomy. In the case of theoretical reason it is not so easy to specify the principles. I believe them to cover the necessary and Universal forms of thought, which Kant identified partly with the table of judgements and of syllogisms as given in traditional logic (somewhat modified by himself) They cover also the principles of the understanding (though these are complicated by reference to time and space) and presumably the supreme principle of the necessary synthetic unity of apperception. Finally, though this need not exhaust the list, they cover the Ideas of reason, the principles of reason in Kant's own technical sense. 2 Such an appeal for corroboration to the complicated doctrines of the Critique of Pure Reason can carry no weight with those who have not read that work and may only encourage in their scepticism those Who have; nor is it any part of my contention that Kant's pioneer labour in this difficult field is not in need of the most drastic revision. But we may, I hope, see at least the plausibility of his general doctrine if we consider one Idea of reason the Idea of a system. As a thinker every rational being as such must necessarily aim and knows that he must necessarily aim at a complete system: 3 he cannot rest in contradictions and inconsistencies nor can he let himself be satisfied with loose ends, though he may have practical grounds for doing both. Such a system of thought must be based on sense, but it is not given to sense. Can this necessary ideal of a rational thinker as such rest on anything other than the insight of theoretical reason into its own necessary
.
activity?
such must necessarily aim and complete system he cannot be satisfied with limited ends or with antagonism and conflict between desire and desire, between man and man, between State and State. Such a system must be a system of actions whose ends are based on the desires of men, but it is the aim of a rational will, not the object of any desire or aggregate of desires. Can the necessary ideal of a rational agent as such rest on anything other than the insight of practical reason into its own necessary activity? The five formulae of Kant are nothing but a philosophical
As an agent every rational being knows that he must necessarily aim
as
at a
Gr.,
392 9. See Chapter IX, Appendix
Gr.,
452
= 85-6.
6.
XXIVg]lS
PO
S S I
BLE
249
analysis of such a necessary that is, objectively necessary system of action, beginning with the universal law which is its form and ending with the complete ideal of a kingdom of ends to be realised in time. This parallel between theoretical and practical reason might be elaborated indefinitely, and the differences might be as illuminating as the resemblances. All I suggest is that this is a line of thought not to be rejected lightly without due consideration. Unless we accept the ideal, in practice if not in theory, of an objective principle of reason, the result in the long run is lunacy. Unless we accept the ideal of objective principles of practical reason, the result in the long run is criminal lunacy and recent events suggest that the run need not be very long.
;
9.
Reason and
is
the unconditioned.
one further point, which is brought in at the very end of the Kant's view practical reason must conceive and seek to realise an unconditioned moral law, which for human beings must be an unconditioned or categorical imperative. 1 Apart from this there can be for us no unconditioned or absolute good. As rational agents we can all understand that if we will the end Z, we ought to will the means Y. This gives us a conditioned obligation and a conditioned good, a good as means. But as rational agents we cannot be content to pursue only goods which are good as means practical reason cannot be satisfied unless we can pursue a good which is unconditioned and good in itself, and without this all our pursuit of means is
There
Groundwork.
On
sheer
futility.
this again Kant sees a close parallel with theoretical reason. The function of theoretical reason, when it finds itself confronted as it does in experience with the conditioned, is to conceive and to seek the un-
In
conditioned.
For example, we suppose that every event is necessary and that we understand its necessity by discovering its cause. In this way we explain a conditioned necessity by discovering its condition. But to understand
the necessity of the cause,
so ad infinitum.
However
far
we must back we
discover the cause of the cause, and go, we never come to anything other
than a conditioned necessity, and this is no more satisfying to human reason than the conditioned necessity with which we started. Theoretical reason must conceive the totality of causes, which, because it is a totality of causes, cannot itself be caused. It is this Idea, as we have seen, 2 which of an uncaused cause or a free gives us the conception, however empty, cause. In the same way theoretical reason must conceive the totality of conditions for every conditioned necessity, a totality which must itself
be an unconditioned necessity, if there is to be any necessity at all. The conception of the moral law as unconditionally necessary is only a further
example of the activity of reason (here of practical reason) in conceiving and seeking to realise an unconditioned necessity.
1
Gr. }
463
= 99-100.
Chapter IX, Appendix
4.
250
its
THE CATEGORICAL IMPERATIVE [XXIV ApP.
If we suppose that we can understand a necessity only by stating condition, then manifestly we cannot understand an unconditioned necessity: to explain it by stating its condition involves us in direct contradiction. Hence Kant declares, with a rather unnecessary, appearance of paradox, that the unconditioned necessity of moral law is in-
comprehensible, but we can comprehend its incomprehensibility. It to say that we understand it well enough as a necessary ideal of reason, though we cannot without self-contradiction profess to
would be truer
understand
it by discovering its condition. Those who ask why we should do our duty are falling into this contradiction. They are assuming that we should do our duty only if we want something else, such as happiness in this world or the next. They are in short asking what is the condition under which we should obey an unconditioned imperative. This is merely to deny that there can be an unconditioned imperative or to show that they do not understand what a
categorical imperative is. The theoretical Idea of a totality of conditions or an unconditioned necessity cannot give us knowledge; for no object of this kind can be given to us in experience. can, however, act in accordance with our Idea of an unconditioned moral law, and thereby we can realise in time an unconditioned and absolute good.
We
APPENDIX
KANT'S ADDITIONAL ARGUMENT

WHAT I have called Kant's additional argument consists of three proposiof which the second is supposed to follow from the first, and the supposed to follow from the second. The intelligible world contains the ground of the sensible Proposition A world. Presumably Kant has in mind here the doctrine that the thing-initself is the ground of the sensible world so far as it is the ground of the matter (sensa) of the sensible world. This was brought out in the argument from the passivity of sense. 2 But he may also have in mind the doctrine of the Critique of Pure Reason, namely, that the I-in-itself is the ground of the form of the sensible world sensibility is the ground of the forms of space and time, while understanding is the ground of the categories. We must not forget that in experience the mind must contribute something to the world
tions,
1
third
is
as
known by
it.
Therefore the intelligible world contains also the ground of the Proposition laws of the sensible world. This new assertion is misleading unless we take into account both
:
Gr. 3
453 = 87-8. See also Chapter Chapter XXII 6.
XXIV
above,
2.
3
Chapter
XXII
7.
XXIVAPP.]lS A CATEGORICAL IMPERATIVE POSSIBLE? 25!

the doctrines mentioned above. According to Kant it is the I-in-itself or the understanding and sensibility of the knower which is the ground of the ultimate and necessary and universal and formal laws of the sensible world, and in particular of the ultimate law that every event must have a cause. So far as the laws of the sensible world are empirical, they may be said to have their ground in the thing-in-itself; for the thing-in-itself is the ground of the sensa on the basis of which these empirical laws are formulated by science. Yet empirical laws, though they neither are nor can be deduced from the ultimate a priori laws, are nevertheless specifications of them we could not, for example, discover that an absence of Vitamin C leads to scurvy except on the presupposition that every event must have a cause. 1 2 Proposition C Hence the intelligible world is, and must be thought to be, the source of laws immediately^ governing my will (which belongs wholly to the
: :
intelligible
world]
This statement so palpably fails to follow from the previous ones that it is hard to see what kind of telescoped argument Kant has in his mind. Granted that the intelligible world is the ground of the laws of the sensible world, we cannot infer that it is the source of the laws immediately governing my will; for my will, as he expressly says, belongs, not to the sensible world, but to the intelligible world. We may try to reconstruct his argument in the light of the passage immediately preceding these three propositions. He has just recapitulated
preparatory exposition and affirmed that so far as we consider ourbelong to the intelligible world, we must conceive all our actions to accord completely with the principle of autonomy; and that so far as we consider ourselves to belong to the sensible world, we must conceive our actions to be wholly under the heteronomy of nature. 4 Hence he has already taken it as established that my will from one point of view must be conceived as governed by the laws of the intelligible world, and he has no need whatever to re-establish it by a new inference. What he has to do is to meet the objection that it is impossible both to conceive my will as belonging to the intelligible world and as governed by its laws, and also to conceive my actions as governed by the quite different laws of the sensible world the laws, namely, of cause and effect. His answer is that there is nothing unique or miraculous about such a double conthe intelligible ception, since quite apart from the case of my will world must always be conceived as the ground of the sensible world and
his
selves to
nature of particular empirical laws as a system is worked out fully in the See especially Erste Einleitung III, 15 ff. 2 Strictly speaking, all this should surely be a question only of what must be thought to be we are here concerned only with Ideas, not with knowledge of fact. 8 This means that these laws do not govern my will through the medium of an end
1
The
Critique of Judgement.
:
already desired, as do hypothetical imperatives. 4 This is identified with the law of nature governing desires and inclinations that is, with a law of psychological, not physical, causation. This is in turn identified with the of this is" that the principle of happiness cannot be principle of happiness. The difficulty a complete account of the psychological laws of human nature ; for men in fact often act in ways which conflict with the attainment of happiness.
252
THE CATEGORICAL IMPERATIVE [XXIV
APP.
of the laws particularly the causal laws by which it is governed. In spite of the difficulty of recognising two kinds of law the relation between my will and my actions is only a special case of the generic relation which must be conceived to hold between the intelligible and the sensible world. If this reconstruction is regarded as too hazardous, it may be noted that the first two propositions, which assert that the intelligible world is the ground of the sensible world and of its laws, are emphasised by special type, which is not used for the third and most difficult proposition. The argument would seem to run smoothly enough if the third proposition were simply omitted ; but it seems to me that the general sense of my
interpretation
is
assumed in what
follows.
CHAPTER XXV
SOME FURTHER QUESTIONS

i
.
Further questions
to be considered.
So far we have considered only what may be called the logical character of Kant's attempt to justify the categorical imperative. It may perhaps seem disappointing if his deduction is reduced to a claim to have direct insight into the principle of autonomy as one which is valid for all rational agents as such and is the ground of moral goodness as well as the basis of the categorical imperative. Even so, Kant is in no worse a position than those philosophers who claim to know by direct insight that pleasure is the only good or again that certain kinds of thing are good or that certain kinds of act are right. Indeed he is in a better position ; for the hedonistic claim to speak dogmatically is both paradoxical and false ; while the claim to have a series of independent and unrelated intuitions into the goodness of certain kinds of thing or the rightness of certain kinds of act
is
unsatisfactory unless we can grasp some single intelligible principle in the light of which these intuitions can be brought into relation with one another and applied to particular situations where different goods or different obligations may appear to be in conflict with one another. There are, however, other difficulties in Kant's ethical doctrine,
its
particularly
connexion with
his metaphysics, its relation to feeling,
and above
cal insight.
all its insistence
on a difference between
to these topics
theoretical
and
practi-
Without some reference
our discussion cannot
be complete.
The
2.
real self.
to justify the categorical imperative by saying accords with the will of my real self as a member of the intelligible 1 world, and this is why I ought to obey it. If there is anything of this in the present argument, it is certainly not emphasised Kant's emphasis is on the rationality of the moral will, and not on its supposed metaphysical 2 he reality. It is true that in his appeal to ordinary moral judgement claims that even a bad man wishes to be a better one, that in so doing he transfers himself to another order of things than the sensible, and that in thus transplanting himself into the intelligible world he is conscious of a good will in himself which constitutes the law for his bad will ; but even here there seems no suggestion that "he should obey his rational will
Kant may be thought

it
that
See also Chapter
XXII
2.
Gr.,
454-5
- 88-9
[253]
254
because
it is

his real will rather
[XXV
than because it is his good or rational will. In his more philosophical account 1 Kant is far from laying any weight on the reality of the pure rational will. On the contrary, he says that in addition to my sensuously affected will there comes the Idea of the same
pure will belonging to the intelligible world. In other passages Kant identifies a man's intelligence and his rational will with his 'proper' self. 3 This must be so, since it is the possession of intelligence which distinguishes man from other animals and makes him a self. Only as a rational agent can man be subject to the categorical imperative (or indeed to any other), and it is not clear whether Kant is saying more than this when he declares that man is bound by the categorical imperative because he is a member of the intelligible world. One of the passages, however, does suggest that Kant is at least reinforcing the appeal to the rationality of the self by ascribing to the rational self a special metaphysical status. 4 He has insisted that the moral law is valid for us as men because it springs from our will as intelligence, that is, from our proper self; and he goes on to say with all the emphasis of special type that what belongs to mere appearance is necessarily subordinated
character of the thing as it is in itself. implications of this are somewhat obscure, but it looks like saying that intelligence i? real, desires are mere appearance, and therefore a rational agent will subordinate his desires to his intelligence.
will as a
by reason
to the
The
This kind of metaphysical argument for morality makes little appeal modern man, and the reason why desires ought to be controlled is surely because they may lead to irrational action, and not because they
to the
are themselves unreal.
We may properly hold
as
Kant does 5
that because
man is a rational
being, he will always be dissatisfied and in conflict with himself so long as he devotes his reason merely to the pursuit of pleasure and even of happiness. Since his reason has another and more essential function, his proper self cannot be realised in a life of self-seeking, and such a life must inevitit is,
ably produce a feeling of frustration. This contention, if true, as I believe seems to be in no need of metaphysical support, though it may be in need of metaphysical defence.
3.
The
conflict
of reason and
desire.
The moral law must appear
to us as
rational agents with a rational will
desire. In some ways such a conflict ways we understand it only too well. Kant seeks to make it more intelligible by distinguishing the self considered as a member of the intelligible world from the self considered as a member of the sensible world.
:
an imperative because we are which may come into conflict with may be hard to understand in other
88. relation between the Idea and sensuously affected will the relation of the categories to sensuous intuition.
Gr. t
1 8
453
The
Gr.y
Gr.,
my
is
compared with
457 = 92-3 ; 458 460-1 ~ 96-7.
93
461
= 96-7
'Das
eigentliche Selbsf.
6
Gr.,
394-6
13-15.
XXV
If
4]
$55
take this as a metaphysical theory (not merely as a practical Idea on which we are entitled to act), it does not seem to make the conflict of reason and desire more intelligible. It is hard to see how there could be a conflict between the self as it is and the self as it appears. It is hard to see how the self as it is could be in any way affected by the self as it appears. It is in short hard to see how desires could be either a help or a hindrance to a rational will, if such a will belongs to an intelligible and non-sensible world. If this is so, the distinction between the intelligible and sensible worlds does not help us to understand how the principles of autonomy must appear to us as a categorical imperative indeed it makes difficult what is otherwise comparatively easy. The objection may be due to the crudity of our interpretation in taking a practical Idea as a speculative theory, but it was already raised in Kant's time, and it is perhaps worth while merely to note his answer. 1 Nature, he says, does not help or hinder freedom; but nature as an appearance helps or hinders the effects of freedom as appearances in the sensible world. 2
:
we
4.
Ethics and metaphysics.

is not based on in spite perhaps of occasional lapses would be truer to say that his metaphysics, so far as
Kant's ethics
his
metaphysics
it
we
take this to be concerned with a supersensible reality, is based primarily on his ethics. Whatever confusion or error there may be on this topic in the Groundwork is to a great extent cleared away by the
Critique
of Practical Reason. in the Groundwork the somewhat hesitating attempt to justify the principle of autonomy by the necessity of presupposing freedom in all rational activity is not wholly an appeal to non-moral considerations. For Kant freedom is essentially a moral ideal. What arouses his awe and admiration is the spectacle of a minute rational creature, thinking freely
Even
and acting freely, in accordance with his own principles, in an endless physical universe governed throughout by iron law. This spectacle is only the concrete embodiment of the Idea of law for its own sake; but perhaps Kant's emotional attitude towards it may be put down, at least for freedom, a passion which is also shown partly, to his personal passion in his political and religious thinking. In any case it is clear, and it becomes
LV n.). K.U.> Einl. IX, 195-6 n. ( For the sake of completeness we must add the rest of his answer, which perhaps raises more problems than it solves. Even the causality of freedom (that is, of pure and practical reason) is, he says, itself the causality of a natural cause subordinated to freedom (that is, of the subject as a human being and consequently considered as an appearance.) The intelligible which is conceived under the Idea of freedom contains the ground of the determination of this latter causality in a way which is admittedly inexplicable. In a somewhat telescoped parenthesis he indicates further and we may compare that a similar relation holds between the intelligible, conceived as the super2 above sensible substratum of nature, and the series of causal events which constitute nature as
1
an appearance.
256
clear to

Kant himself later,
[XXV
that the necessity of presupposing freedom of the will rests on direct insight into the objective principles of morality. Once that is seen, his ethics becomes manifestly independent of his metaphysics and must be judged on its own merits.
justifies
beliefs about God and immortality which Kant on the basis of his ethics lie beyond the scope of the present work. They do not alter the content of his ethics, nor can they add either to the supreme value of the moral will or to the binding nature of the categorical
The metaphysical
1 imperative. Nevertheless, as Kant recognises, it is a great stimulus to moral effort and a strong support to the human spirit, if man can believe that the moral life is something more than a mortal enterprise in which he can join with his fellow men against the background of a blind and indifferent universe until he and the human race are blotted out for ever. Man cannot be indifferent to the possibility that his puny efforts towards moral perfection may, in spite of appearances, be in accord with the purpose of the universe, and that he may be taking part in an eternal enterprise under a divine leader. There may still be some today, and they are fortunate, who can honestly claim to know by pure theoretical reason that this is not a mere possibility but a reality. Most men who are at all touched by the modern spirit would be well content if with Kant they could hold this as a reasonable faith based on their moral convictions; but the discussion of this topic belongs to Kant's philosophy of religion, not to his ethics. There are some who hold that Kant's belief in a timeless and intelligible world undermines morality, since it makes the moral struggle unreal. This may be so, if we take him to be propounding a scientific theory about two wholly different worlds. It is not so, if we regard him as teaching that our task as men is to manifest in time eternal principles of reason on which we may rationally believe that this mysterious universe is governed. And indeed his view is not so different from that of some of the theologically minded who gibe at it, inasmuch as they too regard the activity of God as timeless and the temporal universe as the manifestation of a time-
less will.
5.
Moral
interest.
being it is obvious enough will be convinced of moral by the purely intellectual recognition that a fully rational agent would necessarily act in accordance with a universal law unobligation
No human
conditionally valid for all rational agents as such. This should not surprise nor should it be used as a ground for rejecting Kant's philosophy. Even the purely intellectual ideal of system would mean nothing whatever to a man who had not thought, and very little to a man who had not thought a great deal; but it is no less a necessary ideal of thought because of that. If Kant's analysis of the moral ideal is to mean anything to us, we
us,
1
Gr.,
462-3
= 99.
257 must not only have a capacity for highly abstract thinking we must also have a certain experience of moral action. Perhaps it is the apparent thinness of the moral ideal as analysed by philosophy which makes Kant so insistent on the question 'How can I take an interest in morality?' 1 He returns to this question at the 2 very end of his book, and declares that we can no more explain this than we can explain how freedom is possible or how reason can be
5]
:
XXV
practical.
on Kant's view always belongs to beings who are partly and partly sensuous. Moral interest, so-called, is identified by him with the feeling of reverence, 4 and it is natural enough that he should turn to consider the feeling element in morality after he has established the categorical imperative. It may seem strange, it must seem strange, that so strong a feeling should be aroused by a morality which has been so
'interest'
An
rational
abstractly analysed.
How is it that man believes himself to feel his personal
worth in obedience to the categorical imperative and to estimate the worth of mere pleasure as nothing in comparison with this? 5 Kant may even be returning to the second form of his main question about the the form which is concerned possibility of a categorical imperative with an absolutely good will rather than \vith a rational agent. 6 Or at least he may be asking how we can explain the feelings which accompany the judgement that in obedience to the categorical imperative there is manifested an absolutely good will. The feeling concerned he regards in this passage as one of pleasure or satisfaction at the fulfilment of duty, though that is clearly an in7 adequate account of reverence. He adds little to what he has said already beyond perhaps making it clear that we have to consider a relation between a mere Idea and an emotion, not a relation between two events, which alone can be 'explained', if the explanation of anything is a statement of its cause. He denies that it is possible to understand a priori how a mere thought can produce a feeling of pleasure or pain a view which is
considerably modified in the Critique of Practical Reason? It would be impossible from a philosophical account even of such a thing as skill to understand the feelings aroused by an outstanding exhibition of skill. The same is even more true as regards a philosophical account of art. Kant is perfectly right in insisting that if we are to feel the supreme value of moral goodness we must separate it in thought from 9 all regard to a man's own advantage in this world and the next ; but it must be added that for partly sensuous and partly rational beings like men this feeling will be aroused in its full intensity only if they see or imagine goodness manifested in individual men and so brought nearer to
intuition. 10
1
8
Gr.,
449-50
See Chapter
Gr.,
6
7
449-50
= 82.
XXV
82-83.
7.
Gr.,
459-60
401 n.
Gr.,
= 95-6. = 22 n.
2.
Chapter
XIX
See Chapter V, especially the Appendix. 8 K.p.V.> 73 198 ( 130)10 Gr, = 66 and 437 67. 9 436
*Gr., 411 n.
= 34 n.
258
6. Interest

and
obligation.
[XXV
Kant once more warns us, as so often, against supposing the moral to be binding upon us because it arouses our interest or affects our emotions. On the contrary, it arouses our interest and affects our emotions because it is binding and is recognised as binding. This is a truth which is overlooked by those who today bid us to be moral because of the higher satisfaction which it will bring. Once we have
law
ceased to believe that the moral law is binding and that there is supreme goodness in the life of obedience to the moral law, these alleged higher satisfactions will melt into thin air. Apart from this it is always a denial of morality to bid men pursue it for what they will get out of it whether this takes the old form of promising that God will reward them or the new form of promising that they will find a higher happiness in their present life. The older form is the more plausible of the two inasmuch as by postponing our reward to a future life it is less likely to be contradicted by obvious facts. It is also less removed from a moral standpoint, and makes a stronger appeal to the human heart, inasmuch as it bids us face, at least for a season, the effort and the trials without which no great thing can be done. Nevertheless to cajole or threaten men into being good is always to assert implicitly that goodness is of less value than feelings of pleasure or satisfaction and that the moral law has in itself no binding other than the not too estimable quality of force. There is no morality prudence unless there is, independently of our emotions, an unconditioned goodness and an unconditioned obligation.
7.
Practical insight.
difficult, and yet the most valuable, element in Kant's remains to be considered, hard as it is to express this clearly and unambiguously. There is a difference between theoretical and practical insight, and it is the special characteristics of practical insight with which we are now concerned. Hitherto I have dealt with the resemblance rather than with the difference between these two kinds of insight. In both of them we have gradually to separate out from their empirical concomitants the principles at work in ordinary judgements, and to grasp the necessity of these principles alike in themselves and as conditions of the judgements in which they are embodied. In both of them the apprehension of principle 1 is ascribed by Kant to the self-consciousness of reason. But in moral insight the self-consciousness in question is a self-consciousness of practical reason ; it is not merely self-conscious thinking but is at the same time a self-conscious attitude of the rational will. This attitude of the will must be present in ordinary moral judgements
The most
still
doctrine
See Chapter
XXI
5.
XXV
will.
8]
SOME FURTHER QU E S T I O N S
259
as well as in philosophical ones. Thus even a bad man is said to be conscious of a good will in himself which constitutes the law for his bad
very principle of autonomy itself brings out this kind of selfa rational agent knows that, considered as a rational agent, he would necessarily act always as a law-making member of a kingdom of ends. The moral I ought' is always a necessary I wilP of a person who is conscious of himself as a rational agent. Kant says even of the bad man that 'he believes himself to be this better person' at least when he adopts the point of view from which he regards himself as free and as a member of the intelligible world. 1 Our thought of the good will present in us is, it is true, spoken of as an Idea. 2 An Idea is not the concept of any actual object in space and time ; and Kant, as we have seen, is the last man to claim possession of a 3 good or holy will either for himself or for others. Nevertheless this does not mean that we have here merely a theoretical conception divorced from all actual volition. Every Idea, though it is not constitutive of actual objects, is regulative that is, it is a principle actually employed, however feebly, in the thinking of the rational being who conceives it. And the Idea of a good will is a practical Idea, one which is regulative for our actual will and employed, however feebly, in the volition of the rational agent who conceives it. If we were not conscious of, as it were, the stirrings of a good will in ourselves, if the Idea were wholly without influence at least on our wishes, morality would mean nothing at all to us. We should be brutes and not men, and to speak of duty in such a context would be absurd. This view is fundamental to Kant's conception of morality. The bad man recognises the authority of the law even in transgressing it. 4 We all recognise the validity of the categorical imperative even when we sophistically reduce its universality to mere generality, and pretend we are only making an inconsiderable exception to it under the pressure of circumstances. 5 Our feeling of reverence for the moral law, a feeling which we cannot eradicate without ceasing to be men, arises from consciousness that our will is subordinated to the law. 6 All these statements must be taken to imply that a good will is in some degree present in us.
consciousness
:
The
8.
Modern
intellectualism.
This doctrine of Kant is unfortunately seldom taken seriously today. This general neglect seems to me largely due to the modern tendency to split things up instead of trying to see them as a whole. Analysis is more popular than synthesis. Thus even the modern defenders of reason in morals seem to regard reason as a purely intellectual faculty whose function, so far as it is not concerned solely with inference, is to apprehend immediately a number of
1
8 8
454-5 = 89. Chapter III 5. Gr., 424 =51.

Gr.,
2
.
Gr.,
454
Gr.,
401 n.
Gr., 455 89. 21 n. See also Chapter
= 88. =
2-3.
26O
XXV
separate and isolated truths truths, for example, about the kinds of thing that are good or the kinds of act that are right. When reason is thus divorced from volition, it is hard to believe it capable of apprehending truths of this kind, which seem to be bound up with emotional and volitional attitudes in a way strangely different from the truths of mathematics or of logic. The fact that such apprehensions are claimed as knowledge, and therefore as infallible, does not make them easier to stomach by an age which is only too conscious of the relative element in moral judgement; and there is a widespread and not unplausible tendency to maintain that purely intellectual apprehensions can have no influence over the passions. The further development of this tendency is a general revolt against reason. This is marked even in theoretical matters, as is indeed inevitable once the synthetic activity of theoretical reason is ignored or denied. It is still more marked in morals, where undiluted relativism has passed from the scholar's study to the chit-chat of general conversation. Moral judgements tend to be regarded as mere expressions of personal emotion or of likes and dislikes, and moral action so-called is viewed as mere convention or habit, or as based at most on intelligent self-interest. Thus the two views which Kant described as dogmatism and scepticism are as prevalent today as ever, scepticism, however, being well in the lead. They must arise if we split the whole personality of man into separate and unrelated faculties, theoretical reason on the one hand and a mysterious entity called the will, which operates blindly after thinking has taken place and produces actions which may be objects of subsequent, thinking. It is easy to see that such an entity could not be a will at all, and certainly not a free or rational will so far as it operated, its operations would be the result of some preceding intellectual activity or apprehension. It is almost inevitable that the very existence of such an entity should be denied, and that the so-called will should be reduced to a mere complex of impulses, having, perhaps, their origin in the unconscious. This last theory is still intellectualistic it is formed from the point of view, not of an agent, but of a spectator which is what most philosophers, and perhaps even most psychologists, tend to be from the very nature of their
:
profession.
putting my own interpretation on interpretation at that. I am not the doctrine that of and unrelated faculties is explicitly asserting separate 1 taught my complaint is rather that it appears to be taken for granted, and that the rationality of will receives no adequate discussion. Expressly though Kant repudiates any attempt to construct a philosophy of action, his account of practical reason on its different levels is at least an effort to consider the way in which human volitions are themselves rational. Until philosophers recognise with him that willing may be as rational
In contrasting these views
am
modern movements, and a summary

:
as thinking,
it
is
hard to see how we can have any
satisfactory
moral
philosophy.
1
This
is
as true of
modern psychoanalysis
as
it is
of other doctrines.
XXV
9.
g]
26l
Kant's teleology.
Kant's treatment of the problem may be sketchy and his solutions inadequate, but if we accept his view of willing as itself* a rational activity, we begin to see the relations between skilful and prudential and moral action ; we begin to understand the emotional factors which undoubtedly accompany moral judgement ; above all we begin to realise that duty is neither a mystery nor an illusion, but an intelligible conception and a necessary ideal in the whole context of our practical experience. The adoption of Kant's view of willing need not commit us to claims of a special metaphysical status for the real and rational self, 1 though it raises metaphysical questions about the possibility of freedom. On the other hand, if we regard rationality as the distinguishing character of man and of his will, it does commit us to the doctrine that the end of man is to realise rationality in thought and action, not indeed as an isolated failure individual, but as a member of a community of rational beings. to seek the realisation of this end must inevitably result in a feeling of
frustration.
It may be objected that the suppression of natural desires will equally lead to feelings of frustration, and that these must have their place in life as well as practical reason. This is true, but it is irrelevant as a criticism of Kant. He always rejects a 'monastic' asceticism which goes to work with 'self-torture and crucifixion of the flesh' and can in the end produce only a secret hate against the commands of virtue. What he calls 'ethical gymnastic' consists only in resistance to natural desires to the extent necessary to become their master in cases where they may threaten danger to morality ; and this makes a man hardy and also cheerful in the 2 consciousness of his recovered freedom. The Idea of my pure moral will stands to my will as affected by sensuous desires in somewhat the same 3 that is, as a principle of relation as the categories stand to my intuitions of we must a not as Furthermore, suppression. principle organisation, remember Kant's teleological principles in applying the moral law: 4 everything in human nature has its proper function and purpose, and, on Kant's view as on Plato's, it will perform its proper function only under the governance of a rational will. Nor must we forget that a rational will must concern itself, not only with my desires, but also with the desires of others, and that a harmony of human purposes is possible 5 only on the condition that all men seek to obey the same moral law. a has and even an indirect man that both holds Kant always right, duty, to seek happiness so far as his doing so does not transgress the moral law, and that only so far as all men subordinate their search for happiness to the fulfilment of their duty can happiness be widely attained though even then happiness would to some extent depend on other considerations which are not under human control. For him the realisation of a rational
'
See 2 above.
Gr. t
454
= 88.
137-8
(
M.d.S., Tugendlehre,
*K.p.V.,
28-
= 50-1).
Chapter
XV
53, 485.
6.
262
will
is

not opposed
to,
[XXV
IO
realisation of human
but is rather the condition powers and human desires.
of,
the fullest possible
10.
The
self-consciousness
of practical
reason.
When we have cleared away misunderstanding, Kant's justification of the categorical imperative is seen to rest, not on purely theoretical argument or insight, and not on some emotional experience, still less a mystical one, but on what may be called the necessary self-consciousness of practical reason. Does this contention really amount to anything, or are we merely using fine phrases as a sort of cloak under whose disguise amiable aspirations may be able to pass muster as serious philosophical theories? 1 If we are to do justice to Kant's view, we must see it in the context of a whole philosophy in which the self- consciousness of reason holds the central place. According to him, to entertain any concept, or at least any complex concept, is to be conscious of the plan or pattern of one's own mental activity (including the activity of imagination). Thus to entertain an empirical concept is to be conscious of the plan or pattern followed in combining sense-data into a physical object. To entertain a geometrical concept is to be conscious of the plan or pattern followed in constructing a geometrical figure. To entertain a logical concept like that of ground and consequent is to be conscious of the pattern of the necessary activity of reason as such. To entertain a category like that of cause and effect is to be conscious of the pattern of a necessary activity of the imagination in combining objects in one time and space and of its correspondence with the concept of ground and consequent. To entertain an Idea of reason is to be conscious of the pattern of the necessary activity of reason in passing beyond the conditioned to the totality of conditions and so to the unconditioned. 2 have a similar consciousness or self-consciousness of the pattern of the activity of practical reason. In our maxims we are conscious of willing our actions as having a certain character or as conforming to a pattern or a rule ; and this consciousness is a self-consciousness of practical reason, not merely a theoretical consciousness super-added to a blind and unconscious volition. In our objective principles of skill and prudence we have a similar self-consciousness of the pattern or rule of a necessary activity of practical reason, of the pattern or rule which a rational agent would necessarily follow if he had full control over passion, and ought to follow if he is irrational enough to be tempted to do otherwise. But in all this the pattern or rule is subject to a condition, the condition that a particular end is sought. In moral principle we have a similar selfconsciousness of the pattern or rule of the necessary activity of practical reason, but in this case the principle is no longer subject to the condition that we happen to be seeking a particular end. All this is far too summary to be precise, and it omits many necessary qualifications, but it may be sufficient to indicate the general lines of
We
Compare
Gr.,
453
= 87.
See Chapters IX, Appendix, and
XXIII
5.
XXV
Il]

will, if
263
to under-
Kant's thinking to a reader of good stand than to refute.
he
is
more anxious
The central point, and the central difficulty, is that as regards the objective principles of practical reason we are still supposed to have a self-consciousness of the necessary activity of practical reason, not merely a theoretical consciousness super-added to a blind or unconscious will. Yet we cannot suppose, as we do ex hypothesi in the case of a maxim, that the objective principles are in fact being followed. Hence the necessary self-consciousness of practical reason appears to be a misnomer. can we say that a good will must be present in men as rational agents, when it so very manifestly is nothing of the kind?
How
11.
Thought and action.
It is extraordinarily difficult to think out clearly the relation between moral judgement, especially philosophical moral judgement, and moral action, if we are dissatisfied with the view that moral judgement is a purely theoretical activity not fundamentally different from logical or mathematical judgements. I have suggested that our philosophical grasp of moral principles is in some ways analogous to our philosophical grasp of logical principles. The latter must be present in our thinking and yet, especially before we have formulated them, they may be almost as obscurely and imperfectly present in our thinking as moral principles are in our acting. This is particularly true of what Kant calls Ideas of reason, which set before us the ideal of seeking for the condition of every condition until we have completed the totality of conditions, which itself must be unconditioned. This theoretical ideal is no more capable of complete attainment than is the moral ideal, which is also an Idea of reason, although of practical reason. Yet in both cases the ideal, when abstractly formulated, is seen to be a necessary ideal of reason, and not merely a principle actually at work, even if imperfectly, in rational
thinking or rational action.
reason
The difference is, however, that in conceiving our Idea of theoretical we are ourselves thinking; the thinking thought about does not
kind from the thinking which thinks about it ; and consequently there seems to be less difficulty in claiming direct insight into the necessity of this Idea. If, however, in conceiving an Idea of practical reason we are merely thinking, then what we are thinking about seems to be something different in kind from the thinking which thinks about it ; and consequently there seems to be more difficulty in claiming direct insight into the necessity of the practical Idea. Is this greater difficulty due to our incorrigible habit of thinking about action as if it were something essentially unintelligent, a mere object thought about and not an activity of practical reason? Ought we not to deny the divorce between thinking and acting and to say that the apprehension of the necessity of the moral ideal is something more than mere theory and somehow contains in itself an activity of will? Is this apprehension in short properly described as selfdiffer in
264
[XXV H
consciousness of the necessary activity of practical reason, a self-consciousness which is volitional as well as theoretical, as indeed it must be even in our moral actions? Such a view is certainly difficult and hard to make precise in words, yet our experience of moral judgement seems to suggest that there is expressed in it something more than theory, some attitude of the will. This aspect of moral judgement or of what is sometimes thought to be
moral judgement is brought out by some positivists, 1 although they do not go on to ask whether this attitude of the will may be rational or irrational they seem to assume that whatever is not purely theoretical must necessarily be irrational, an assumption which seems to me in need
:
of some justification.
Perhaps
am making
too
many
difficulties,
and
it
may
be
sufficient
One thing at least is certain. A moral philosopher cannot argue men into being good. All he can do is to separate out the principles by which good men think and act, principles which are not nearly so obscure as some modern intellectuals pretend ; and he can claim as nearly all moral philosophers do that he has a direct insight into the necessity of these principles. Kant goes further and claims that such insight is an insight on the part of reason into its own necessary practical activity, and that it is intelligible when we consider the nature of reason as a self-conscious activity, not only in thinking but
of moral volition in ourselves.
also in acting.
all we can say to a man who asks us why he something like this 'We can set before you in a more or less clear and systematic form certain principles by which good men have judged and to some extent acted down the ages ; but if you want to be convinced that it is your duty to act on these principles and that such action is a pre-eminent good and the one source of human dignity, then you must act on them yourself. You cannot attain moral insight without acting morally any more than you can attain logical insight without thinking correctly. Indeed in either case you must take a decision and make a venture of faith. You must give yourself up to the activity which you wish to understand, and this is as true about morality as it is about logic, or indeed about science or about art. Do not make the absurd mistake of supposing that you can understand and judge the moral life without any personal experience, a mistake which you would never dream of making in regard to any other kind of human activity. Whatever be the theory of the matter, this is the practical answer, and it appears to be in accordance with Kant's doctrine. To many such a view may appear nonsensical and absurd, the product of a moralistic prejudice ; but they ought at least to ask whether their condemnation may not spring from an intellectualistic prejudice, which may be just as misleading as a moralistic one.
to say here that we cannot expect to grasp the necessity of moral principles, or even to make moral judgements, unless we are conscious of some sort
If this
is
the proper view,

is
should be moral
See The Philosophy of G. E. Moore, pp. 71
ff.,
for
an
interesting article
by G. L.
Stevenson.
XXV
12.
12]

objections.
265
Some practical
No adequate discussion of the theoretical difficulties in this view can be attempted here, but it may be well to note some objections on the
any kind of life will be judged good chosen and persistently followed, so that this method of seeking moral insight will result in purely arbitrary judgements. This, however, is true only within limits and mainly as regards what is good for the individual the moral life I claim to be good in itself and not merely good for me to be good in itself even when, in a narrow sense, it is not good for me. It may also be thought that a great deal of spurious morality has been advocated, and may be advocated, along precisely these lines. In this there is nothing surprising, for men grasp moral principles slowly and clutter them up with nonsense ; yet the strength of their advocacy may rest on the genuine moral principles mixed up with the nonsense, and the nonsense may be cured by further living and by further thought. The fundamental principle of Kant's moral philosophy is not that we should live blindly, but that we should live intelligently. Finally it may be thought by some that the doctrine is insufficient: the moral decision must be supplemented by the grace of God. If this is true, it belongs to the philosophy of religion and not of morals ; and in any case the moral decision is essential, and by itself it will carry us a
if it is freely
:
level of practice. It may be thought that almost
long way.
CHAPTER XXVI
THE DEFENCE OF FREEDOM

i
.
The antinomy offreedom and necessity.
we regard the principle of autonomy as following from the presupposition of freedom or consider the presupposition of freedom to follow from our recognition of a categorical imperative, it is necessary to defend freedom against the charge that it is incompatible with the causal necessity which we know to prevail in nature. This fundamental problem has hitherto been kept in the background ; and only after Kant has explained how a categorical imperative is possible does he turn explicitly to the task of reconciling spiritual freedom with the causal 1 necessity which governs the phenomenal world. in the light of the preceding The new argument is comparatively easy discussion. Freedom and necessity seem to be equally necessary, the first for action, the second for science; and yet they seem to be contradictory. Since we cannot afford to give up either, we must suppose that there is no real contradiction between them, and we must try to see whether the seeming contradiction can be removed ; otherwise of the two it is freedom which we should have to give up. The contradiction will be removed only if we are able to show that freedom and necessity can, and indeed must, be conceived as combined in man. It is manifestly impossible to hold that man can be both free and causally determined, if we conceive man in the same sense, or in the same relation, in both cases. Freedom and necessity can be reconciled only if man can be conceived in two senses or in two relations (that is, as thing-in-himself and as phenomenon). It is the task of speculative reason to resolve the antinomy by establishing this double standpoint (or double point of view) and so to defend practical reason and moral beliefs against all possible attack. Practical reason legitimately requires this service from theoretical reason, and in so doing it is not going beyond its proper limits.
2.
WHETHER
The solution of the antinomy.
solution which Kant offers for the antinomy 2 is the one we have already found in our attempts to escape from the alleged vicious circle of the argument. 8 There is no contradiction in supposing a thing as it is in itself to be
The
Gr.,
455 ff. 89 ff. Chapter XXII 3 and 4

[266]
Gr. 9
457-8
=*
92-3.
XXVI
3]
367
independent of the laws to which the same thing is subject as an appearance or as belonging to the sensible world. Man must look upon himself from both these points of view. He must regard himself, qua intelligence, as independent of the causal laws which govern his sense-impressions his reasoning must be determined by grounds of quite another kind. Similarly though here again it is not clear whether Kant regards this as an inference or as due to independent insight he must conceive himself as an intelligence possessed of a will and so as exercising causality in the phenomenal world; and here too he must regard his rational willing as determined by quite other grounds than the causal laws which must govern him considered as a temporal series of impressions, feelings, desires, and actions. Hence he must regard himself from two different points of view he must consider himself, as a rational being and a rational agent, to be a thing-in-itself and a member of the intelligible world under the laws of freedom; and he must also consider himself, as a sensuously affected being and an object of inner sense, to be an appearance in the sensible world completely governed by the laws of cause and effect. Once we understand that man must regard himself from these two points of view the supposed contradiction disappears.
: :
3.
The two
standpoints.
Kant seems to be right in saying that there are two standpoints from which we must regard ourselves. The two standpoints I have described on a humble level as the standpoint of the agent himself and the standpoint of the observer. In the light shed by Kant's account of theoretical and practical reason this description is seen to be too simple. The two standpoints are rather the standpoint of the self as consciously thinking and acting, and the standpoint from which the self is considered as something thought about and acted upon. Both these standpoints seem to be necesanything we can call self-knowledge. standpoint of the self as thinking and acting does seem to be one which in a curious way is outside appearances and even outside time. 1 Yet by itself such a view is quite unsatisfactory. Not only can we know our thinking and acting as events in time, but our thinking and acting, besides always taking place at a time, also take place through a time, as Kant himself admits. 2 Even if in our thinking we seem to survey all time in the light pf timeless principles, we do so from a particular point in time, and one which is continually changing. Our view of the future is not the same as our view of the past. Our acting is based on such a view and sucli a survey, but it seems even more closely bound up with time; for however timeless be its principles, its very essence is, as it were, to thrust out of eternity into time
sarily present in
The
Compare Chapter XXIII
3-4. See also Gr., 458
93 for a standpoint 'outside
appearances'.
268
[XXVI
4 a
and
to affect the course of changing events as these pass before us at
particular moment.
It is hard to describe the double standpoint from which we seem necessarily to regard ourselves, but it would surely be unwise to suggest that there is no double standpoint at all, and that Kant's problem is artificial. The main troubles arise when we begin to consider his attempt to justify this double standpoint and to treat it as falling under a more
general principle. What Kant maintains is that this special double standpoint is only a particular case of a more general standpoint from which we must necessarily regard every object of knowledge. Everything must be conceived as a thing-in-itself and as an appearance to us. All things as appearances must be governed by the causal laws of the sensible world, but the very same things as they are in themselves may be exempt from these laws.
4.
How
is freedom
possible?
It is not unnatural to suppose that Kant, in propounding this doctrine, offering us a theory about the intelligible world which will explain how freedom is possible. There are two main and obvious objections to such a theory. The first is that he seems to claim knowledge of the intelligible world to which he is not entitled; and the second is that even if this knowledge were admitted, his theory would offer a very poor
is
explanation of freedom indeed. As to knowledge of the intelligible world, considered as the totality of things-inthemselves, we may perhaps grant him, on his premises, that the intelligible world cannot be conceived as temporal or as subject to causal law, which is essentially temporal. But he seems to go very much further than this. Even to call the world as it is in itself an 'intelligible'
world is unnecessarily ambiguous, if by 'intelligible' he means only that although it must be conceived, it is inaccessible to our senses. He has, 1 however, already described it as an 'intellectual' world, and so presumably as one which is actively intelligent. There seems to be little doubt that he conceives the intelligible world as 'the whole of rational beings as thingsin-themselves'. 2 Man, as he tells us in the present passage, 3 knows nothing more of the intelligible world than this that in it reason alone, and indeed pure reason independent of sensibility, is the source of law and it is added that the laws of the intelligible world apply to man categorically because only in that world is he his proper self. 4 All this suggests that Kant claims a surprising amount of knowledge about the intelligible world and is prepared to base his present defence of freedom, and even his justification of the categorical imperative, on such knowledge. It is hard to see why he permits himself to use language which he knows to be misleading. Let us suppose, however, for the sake of argument, that we know
;
Gr.,
451
= 85.
For objections to
Gr., 458 94. this doctrine see Chapter
XXIV
Gr.,
457
93.
9,
5J
ultimate reality to be a society of rational beings engaged in timeless activity. In what way are we in a better position to deal with the problem of freedom? It is admittedly impossible to understand how such a society of timeless beings could be the ground of a changing world in space and time governed by causal law. What is more, since timeless activity is inconceivable to us, a free timeless activity is equally inconceivable, and could not possibly explain our freedom to act here and now in accordance with the moral law. Yet it is precisely this freedom which we are attempting to defend. It is unnecessary to elaborate further criticisms of such a theory: they can be found in almost any book in Kant. The only defence of Kant, if there is one, is that in spite of his admittedly misleading language he is not offering such a theoretical explanation at all. It is at least worthy of remark that he himself explicitly denies the possibility of explaining how
freedom
is
possible.
5.
Phenomena and noumena.
In trying to interpret the distinction between

sensible
Kant we must take seriously his insistence phenomena and noumena, or between
that the
and intelligible worlds, is to be understood in a negative sense. 2 holds that the thing-in-itself which we are bound to conceive as the reality which is the ground of appearance is not an object of our senses and consequently cannot be known by us. If we suppose that it can be known by means of our conceptions, this on Kant's view is to suppose that we know it by some kind of intuition which is not, like ours, sensuous and passive, but intellectual and active. We then take 'noumenon* in a positive sense as an object known by means of a non-sensuous in-
He
to do this is always illegitimate. basis of this distinction some commentators tend to regard the thing-in-itself as unnecessary to Kant's system ; what we cannot know we
tuition
;
and
On the
cannot consider to be in any sense real. Kant is far too much of a realist to accept this. For him appearances are always appearances of a reality; and I see no reason to doubt that he always regarded the thing-in-itself as real in itself but unknown by us. 3 With the difficulties of this view such as the meaning of the 'real in this connexion we are not here concerned. What should be noted is that the concept of a noumenon is attained by making complete abstraction from our sensuous intuitions under the forms of time and space. 4 This
5
method of abstraction, when applied to objects of experience in general, leaves us with the pure concept of an object in general or of a thing in general. When applied, as Kant applies it, to particular classes of objects, it leaves us with different concepts of noumena, and these at first sight
seem
i
to
have
little
to
do with
things-in-themselves,
*
Gr.,
458-9=94-
K.r.V.,
/T.r.K,
*
Sec also K.M.E., Chapter LVI 3. 98 for the Idea of an intelligible world. Compare Gr. t 462
.
B xx
8311; 8306-7.
270
[XXVI
There are many examples in Kant of such a procedure. Thus when we abstract from the temporal reference in the category of cause, we are left with the concept of causa noufnenon (cause as a noumenon), that is, of a ground. When we abstract from the sensible characteristics of man, we are left with the concept of homo noumenon, which is at least often identified with humanity or personality. This tendency becomes more marked in Kant's later works. For example, legal possession is described as possessio noumenon inasmuch as the concept abstracts from the empirical conditions which belong to the physical occupation or use of some object in time and space. 1 If we abstract from all the sensible and temporal characteristics of moral action, what are we left with? All we are left with is the concept of a timeless universal moral law considered as the ground (not the cause) of the action. When we conceive a moral action in this way we are considering action to be free in so far as it can have its ground in such a
timeless law. The action is considered, not in relation to temporal antecedents, but in relation to its rational ground.
This seems a much more satisfactory way of conceiving free moral action it does not involve any unintelligible theory that a moral action is in reality the timeless activity of an unknown I-in-itself. Here we have one point of view from which actions must be regarded and the same of course applies to thoughts. Has this point of view anything to do with the point of view whereby objects are conceived as grounded in unknown things-inthemselves ? It is easy enough to see how commentators may hold that the thingin-itself is superfluous and that Kant requires no more than the concept of action as grounded in a universal moral law. But it is also easy to see why Kant if I may be dogmatic rejects such a contention. For him, if the sensible world were the sole reality, we should not be entitled to conceive actions in this way ; the laws of cause and effect as necessarily applied to our observed actions would exclude such a possibility. If, however, all sensible objects must be conceived as grounded in an unknown reality not governed by cause and effect, then, and then only, are we entitled to the presupposition on which as rational agents we must act, the presupposition that our actions can be willed in accordance with a universal and timeless moral law. And it is only on the same hypothesis that we are entitled to think as we do on the presupposition that our judgements and inferences can be made in accordance with the timeless principles of theoretical reason. Hence if Kant is to think of a man as acting freely in accordance with timeless moral laws, and in that sense as a member of the intelligible world, he has at the same time to think of that world as the world of things as they are in themselves and not as they appear to our senses. But this involves no claim to know the world as it is in itself. Still less does it involve a claim to understand how an intelligible or intellectual world, considered as a thing-in-itself, can through its timeless activity be mani:
fested in temporal actions.

1
Af.d.S.>
Rechtskhre,^Q and 10, 251 apd 259.
XXVI We
6]

also
27!
is
must
remember
that for
Kant freedom
practical Idea,
and that
this is
something very different from a speculative theory of
reality. All he has to show is that there is no theoretical reason why we should not be entitled to act as the categorical imperative enjoins in accordance with a universal moral law. This he does by arguing that we could exclude the possibility of such action only if the sensible world, with its causal laws, were the sole reality. If he can justify the view that it is not to be conceived as the sole reality, there is no need for him to go
farther.
6.
The thought of the
intelligible world.
It may be objected that all this is special pleading illegitimate thus to water down Kant's crude statements appeal to a vague general conspectus of his doctrine.
and that it is by making an
To this the answer is that Kant himself immediately proceeds to water down, or rather to make more precise, his doctrine, and that he does so along the lines suggested. This is his first approach to the question of the limits of practical reason with which the whole of the final section of the Groundwork is supposed to concern itself. As his exposition
and the subject controversial, the passage requires careful study. Practical reason, he maintains, does not go beyond its limits when it conceives the intelligible world and thinks itself into the intelligible world. It does go beyond its limits when it wants to intuit or sense itself into the intelligible world. In other words it must not claim to know the intelligible world or to know itself as belonging to the intelligible
is difficult 1
world.
regard this positive power? He is very careful to insist that it is only a power to act in accordance with a universal moral principle. If we go further and try to hale some motive, some object of the will, out of the intelligible world, practical reason is going beyond its limits and is claiming acquaintance with something of which it knows nothing. This should exclude the possibility that we can know our real self to be some sort of timeless intellectual activity in the intelligible world and can for this reason regard ourselves as bound by the categorical imperative.
1
2
conceive or think oneself into the intelligible world is only a thought the thought that the sensible world does not give laws to a rational will. 2 The thought of self as belonging to the intelligible world becomes positive only in one point we conceive ourselves, not only negatively as exempt from the laws of the sensible world, but also as possessing a positive faculty or power, the power which we call a will, the power of reason to cause events in the sensible world. How does Kant
negative
:
To
Gr.,
458
= 93-4.
Compare
also Gr.,
462
98.
as
Here the intelligible world is conceived as a noumenon in the negative sense that is, not known to our sensuous intuition, and consequently as not under the causal laws of
the sensible world.
272

am
[XXVI
If I not to abandon consciousness of myself as intelligence that is, as thinking and acting in accordance with rational principles I must so far conceive thoughts and actions as not determined by sense in accordance
with merely causal laws. In order to do this I must conceive myself as a member of a non-sensible world. My concept of a non-sensible or as Kant calls it intelligible world is thus said to be only a standpoint which reason sees itself compelled to adopt outside appearances in order to
conceive itself as practical. say this is not to
To say that the intelligible world is unreal. The concept of the sensible world is as much a standpoint as the concept of the 1 intelligible world. Incidentally to say that the concept of the intelligible world is only a standpoint is not to say that the intelligible world itself is only a standpoint. On the contrary, unless the intelligible world were real, although unknown by us, we should not be justified in claiming to act, and to judge our actions, from this standpoint, as I have already
tried to explain. 2
The thought of myself as a rational agent acting in accordance with a universal moral law does, Kant admits, carry with it the Idea of another order and another kind of law than that which governs the sensible world it makes it necessary for us to conceive the intelligible world, and to conceive it as the totality of rational beings considered as things-inthemselves. It is very clear from this that Kant is far from abandoning the thingin-itself. It is equally clear that his concept of a totality of rational beings considered as things-in-themselves is, like his belief in God and im:
it is not an independent and mortality, based on his ethics metaphysical conception and still less is it metaphysical knowledge from which his ethics can be deduced. There follows a difficult passage 4 in which Kant disclaims all knowledge of the intelligible world. There is not the slightest pretension to do more than think in accordance with the purely formal condition of such a world. But what is this formal condition? The formal condition is nothing other than the principle of morality itself, the principle of acting only on maxims which are valid as universal laws, or in other words the principle of autonomy which alone is compatible with freedom. What we know is only the ultimate universal law on which a rational agent as such would necessarily act. We cannot conceive a rational agent as so acting unless we distinguish the intelligible from the sensible world and suppose him to be from one point of view a member of the intelligible world. But this formal principle gives us no knowledge of any object, least of all of an intelligible world. On the contrary, all laws which give us knowledge of a determinate object are the causal laws of nature; and these govern only the sensible world and constitute what Kant calls
:
'heteronomy'.
1
2 See 3 above. 86. 452 This concept, however, receives support from considerations urged in the Critique
Gr.,
of Judgement.
*Gr., 458
= 94.
XXVI
7.
7]

is
There
no explanation offreedom.
Any
its
attempt to explain
how freedom is possible

is
1 by Kant. To explain anything
by showing that it is caused by something other than itself; and indeed the Idea of freedom is ex hypothesi not the concept of an object of possible experience, and so not the concept of any object in nature falling under causal law. To explain how freedom is possible would be the same thing as to explain how pure reason can be practical that is, how pure reason can have effects in the phenomenal world. In order to do this we should have to intuit pure reason as well as to conceive its formal law. This assumes that we can know the intelligible world by intuition as we know the phenomenal world; and indeed it treats the intelligible world as if it were the phenomenal world, and as if it were under the laws of the phenomenal world. To attempt a task of this kind would be indeed to go beyond the limits of practical philosophy. All we can do is to defend freedom against attack by showing that it must be conceived to belong to man as rational and so as a member of the intelligible world and not as a member of the sensible world. Once we have understood that two standpoints are necessary and that from one of them we must conceive things-in-themselves to be the hidden ground of phenomena, there is no contradiction in supposing man to be both free and determined, free as a member of the intelligible world and determined as a member of the sensible world. It should be observed also that this passage comes very much nearer to the doctrine of the Critique of Practical Reason. Apart from its insistence on the defence of freedom, the Idea of freedom is based on consciousness of the moral law. The Idea of freedom is said to be valid only as a necessary presupposition of reason in a being conscious of a rational will as a power of determining himself to act in accordance with laws of
reason.
especially the causal laws. More cause. Obviously we cannot explain free action
is definitely excluded to bring it under the laws of nature, simply, to explain anything is to state
8.
Timeless action.
I have done my best to show the reasonableness of Kant's doctrine by keeping strictly within the limits he has set himself. A rational agent as such would necessarily act in accordance with principles unconditionally valid for all rational agents and would regard such principles as imperatives if, because of non-rational elements within him, he were tempted to act otherwise. He would also necessarily presuppose himself
to be free to act in accordance with such principles. All this is borne out by our moral consciousness, and it is true of our thinking as well as of
1
Gr.,
458-9
==
94-5.
274
[XXVI
our acting. It makes a claim that there is more in the world than energy or motion governed by physical laws and that a rational agent is something more than such energy or motion, and even something more than a series of mental events succeeding one another in accordance with the causal laws of psychology. Kant's metaphysical doctrine of the thing-initself cannot be adequately expounded, and still less defended, here ; but at least it is a reasoned attempt to justify the view that there is such a 1 something more. It gives us an answer to these dogmatists who, on the
their special insight into the nature of reality, are boldly 2 can always say to such a to deny the possibility of freedom. prepared man with justice 'There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy/
ground of
We
Furthermore, the relation between the principles of reason and our changing thoughts and actions is not one of temporal succession or of cause and effect. No doubt we must think the premises before we think the conclusion, but the validity of the argument is independent of this temporal succession: an argument is sound because it accords with timeless principles. Similarly we must conceive both the actual situation and also the maxim of a proposed action, as well as the moral law, before we act morally, 3 but an action is not moral because it follows, or is caused by, a previous act of thinking it is moral because it is willed in accordance with timeless principles. Unfortunately, because of this timeless element in temporal action
;
Kant
as if
and
this
is
specially
marked
in the Critique of Pure Reason*
speaks
our actions in the intelligible world were timeless. In so doing he gives a handle to his critics of which they take full advantage. Such a view, we must remember, was propounded during a period in which thinkers were more ready than they are now to explain events as due to the timeless activity of God. Nevertheless a doctrine of this kind is surely contrary to Kant's own principles it is an example of that 'swarming about' or 'fluttering about' in the intelligible world which he consistently 5 repudiates. The objection to it is not merely that it cannot explain our
:
this Kant would certainly admit. The objection is rather that timeless activity is itself inconceivable, at least by us and on Kant's own principles we could defend it, if at all, only as a piece of mythology. What we can conceive is only the timeless principles in accordance with which as rational beings we think and act ; and indeed
temporal actions:
no one
insists
on
this
more than Kant

selves
himself.
On the other hand it is only fair to Kant to remember that he repudiates a doctrine of
*
two
such as
we
2
are apt to attribute to him, not
<>., 459 = 95. moral law in acting upon it, but equally in inferring we must grasp the principles in accordance with which we draw the conclusion. Both thinking and acting, successive though they are, yet rise in a curious way above succession; and in both an obscure grasp of principles may be present in the activity itself before these have become clear to conscious reflection.
8
We
Gr.,
462
= 98.
must
also conceive the
Gr.j
462
98.
The German word
is
'herumschwarmerf.
XXVI
9]
275
unnaturally, because of the language which he employs. If we are to extract any satisfactory doctrine from him, we must insist that there are not two selves, but one self considered from two points of view. Similarly there are not two actions, one temporal and the other timeless; there is only one action which has both temporal and timeless aspects. The truth which he is setting out is that an action may be a temporal event and may 1 yet be grounded on a timeless principle.
9.
Freedom
to act badly.
So far freedom is freedom to act in accordance with the principles of 2 autonomy. But, as we have seen, Kant supposes man to be free when he acts badly. If this were not so, man would not be responsible for his bad actions; nor could he regard the moral law as an imperative if it were not possible for him to obey the moral law. Man, as Kant says, is not responsible for his desires but he is responsible for the indulgence with which he allows his desires to influence his maxims to the detriment of the moral
;
law. 3
This view is fully in accord with common sense. The difficulty of it for Kant's philosophy is that the ground of immoral action must somehow lie in rational agents as members of the intelligible world. There must be
an I-in-itself for each separate rational agent as known to experience, and each I-in-itself must somehow contain in itself, not only a pure rational will, but also some kind of irrationality which is manifested in
its
actions.
In
this
way
irrationality
is
introduced into the intelligible
world.
however, concerns the origin of evil, and perhaps to plead that this can have no intelligible ground. 4 All that any one can say of evil is that it must spring from irrationality in a rational will ; but Kant's theory seems to make this contention more difficult rather than more easy. If we are free in acting badly, then we must be free, not only in so far as we act according to objective principles of reason, but also in so far as we act on subjective principles, or maxims, even when they are opposed to the moral law. This is Kant's explicit view, especially in his later works. 5 He holds that no desire or interest can influence our actions unless it is taken up into our maxim by an act of free
This
difficulty,
Kant may be allowed
choice. 6
Such a view seems

ence, with all
1
its difficulties,
to bring us back very near to the liberty of indifferin spite of Kant's denial that freedom can
law)
2
K.p.V., 99 s= 230 ( 177). 'When the law of our intelligible existence (the moral is in question, reason recognises no differences of time.* Chapter 178-9). 7. Compare also K.p.V., 99-100 231 ( Gr.,4 5 8 93. 21-2 ( Religion, 43 49 ( 7-8). 46-7) ; 21 6 22 n. ( 25 ( 12). Religion, 24 Religion, 21 n. 7 n.).
XX
276
THE CATEGORICAL IMPERATIVE" [XXVI

this
1
10
way. Nevertheless his recognition of two senses of not altogether clear, seems necessary at lâst as a first approach to the problem. On the one hand, freedom seems merely a power to act rationally with a full knowledge of the circumstances in which we act. On the other hand, if this were all, the extent to which we were able to act rationally might be wholly determined by something other than ourselves ; and this seems to be a denial of freedom. It looks as if we must be free to act both rationally and irrationally, or at least to act both more rationally and less rationally: without this there could be no responsibility for our actions and apparently no meaning in the statement that we ought to pursue a particular course of action. If so, to adopt any maxim or principle of action is to act freely, and behind this we cannot go. It still remains true that only rational actions actions in accordance with maxims can be free ; and we must not forget that every maxim has some rationality, even if this rationality be displayed only in relation to a set of circumstances which seems to be arbi2 trarily limited. It is hard to understand why we should regard ourselves as free even in limiting the scope of our own rationality, but the fact remains that we do. Indeed unless we do, responsibility and obligation
if it is
be defined in freedom, even
seem to disappear. Although Kant recognises freedom even in acting badly, he does not cease to regard a free will as a power to produce effects without being determined to do so by anything other than itself; and although our will is often said to be influenced, not determined, by desires, this influence is apparently possible only through our free
alike
choice.
This view seems to deny that there are degrees of freedom, but other 8 passages suggest the contrary. Clearly this is a topic which requires more discussion than Kant has given it.
10.
Freedom and
necessity.
Although from one point of view all actions are free, we must not moral and immoral forget that from another point of view all actions alike are determined. I believe Kant to be completely sound in saying that we can and do, and even must, take these two points of view. I even think he must be right in maintaining that the first point of view must in some way have a deeper insight into reality than the second it is at least looking for something without which there could be neither action nor thought, not even
:
the thought of determinism. Such a belief, however, cannot be upheld except by an elaborate metaphysic. The Critical Philosophy is a remarkable effort to supply such a metaphysic, although the attempt to put
1
M.d.S., Rechtslehre, Einl. IV, 226
28.
s
See Chapter
IX
i .
See Chapter
XX
9.
XXVI
Il]

intelligible
2?7
freedom into the
world and necessity into a phenomenal one
breaks down if we take the doctrine to mean that there are two different kinds of action in two different kinds of world. It is also a question whether it is possible to take these two points of view without allowing either to influence the other. The possibility of degrees of freedom must suggest that there are also degrees of determinism, or at least that a rational will may be influenced to a greater or lesser extent by nature in a way not wholly dependent on a free choice. Kant indeed maintains that what he calls our empirical character as an appearance is grounded in our intelligible character as a thing-in-itself ; but he also says of our empirical character that no**rnan can fathom how much should be ascribed to mere
nature. 1 On the other
way
hand it is also hard to believe that a free will can in no influence or deflect the course of events. According to Kant, if we
had enough knowledge of a man's nature and environment we could calculate his future behaviour with as much certainty as an eclipse of the sun or moon. 2 In these calculations we must admittedly take into account the man!s 'empirical' character; and so far as this is grounded on his intelligible character, he may still be said to influence events,
only in a way that could be foretold with sufficient knowledge. But even his empirical character must itself be regarded as the effect of events happening before he is born ; and if this is so, it is hard to see how any remnant of freedom can be saved. Kant never properly faces this difficulty. It would surely be fantastic to suggest that the intelligible character of each individual prepared the way for his empirical character by being the ground of events happening before he was born.
if
n. The
defence
offreedom.
It is impossible within the limits of the present book to discuss adequately Kant's defence of freedom. This is only a part of the vast metaphysical structure for which he clears the ground in his discussion of theoretical reason, but which is itself based primarily upon his ethics and is for him a matter of faith and not of knowledge. I have sought only to indicate the method of approach to the problem, to clear away some misconceptions, to suggest certain difficulties, and above all to show that Kant's ethics, so far from being deduced from the Idea of freedom, is itself the ground on which the Idea of freedom is based. Hence Kant's ethics is independent of his metaphysics and has a claim to our acceptance, or at least to our consideration, no matter what be our attitude towards his metaphysics. But he is surely right in maintaining that there
can be no morality, as
1
it is
understood, however obscurely, by ordinary

n.
JT.r.F.,
A39 = 6567
K.p.V., 99
= 230
and ^551
177)
;
Jf.r.F.,
= 8579 n. A549~5 o =* 8577-8.
278
XXVI
1 1
men, without the presupposition of freedom. The defence of freedom is necessary for the defence of morality, and if Kant's defence is not to be regarded as successful, at least he has shown us the character of the problem and perhaps even suggested some of the lines on which it
may
be solved.
INDEX OF PROPER NAMES

ABBOTT,
II3n., 1340.,
KANT,
I.,
197
2 ion.
Addison, 162 Alexander, S., 236

Aristotle, agn., 108, i6on.
Kydd, Mrs. R. M.,
Ayer, A. J,,
3511., 12211.
LEIBNIZ,
121, 2O9, 227,
230
BAR FORD, G.
Beck,L.J,, 17
B. H., 17
MEL LIN,
Mill,
John
I47n. Stuart, 176

E.,
Borowski, i5on. Bullock, A. B., 48n.
Moore, G.
no,
NERO,
cAiRD
63n., 237n. Cassirer, Ernst, i4on. Churchill, Winston, 92 Cicero, 38n. Collingwood, R. G., 195
,
36,39
131, 161, 162, 191, 227,
Edward,
Newton,
230
PLATO,
Price,
H. H.,
29n., i6on., 26in. 88n., 227n.
Cook Wilson, 74
Croce, 50
DARWIN,
FALK, W.
Francis
I,
09
Canon, 48n. Reich, Klaus, 33n., 38n., 238n. Ross, Sir David, 36, 1 1 7n. Rousseau, J. J., 42, 162 Russell, Bertrand, 74
RASHDALL,
D., i87n.
140 Francis, St., 26
SCHILLER,
48, 50 Schilpp, P. A., 1 6, 30, 42n., 53n., 57n.
Schopenhauer, 70
GANDHI, Mahatma,
Gentile, G., 236
92
Schroeder, Professor, 48n. Seneca, i8gn. Smith, Adam, 190 Stevenson, C. L.,
HARD IE, W.
Hegel, 30 Hitler, 138 Hoernle", 2O5n.
F. R., 17
TAYLOR,
A. E., 2i8n.
Hoffmann, Alfons, ison. Horatio, 274
v A R D o N Harry, 91 Victoria, Queen, 62

,
Hume, David,
30, 210, 227
Huxley, T. H., 161
WASIANSKI,
Webb, C. C.
150, 198
J.,
6on.
JACHMANN,
I50H.
Wells,
H. G., ii4n.
Jones, Bobby, 9 1 Joseph, H. W. B., 12 in.
Wilder, Thornton, 4011. Wolff, Christian, 33
279
GENERAL INDEX
ABGELEITET,
80 Absolute, 34
Ableiten,
134
Contradiction in
wills,
24,
39
Culture, 155, 173, 195
Action, philosophy of, 32-3 , impulsive, 82 , moral, matter of, 73, 75, 185-6
, timeless, 273 Activity, mental, knowledge of, 236
DARWINISM,
,
,
150, l6l
Deduction, 28, 203

failure of,
244
ff.
transcendental, 200, 202, 242
pure, 237
, timeless, 237-8, 269, 273 Agent, rational, 128, 200-1, 246 Agents, rational, 70, 89-90, 176 Analogy, 158, 159, 160, 190-1
Definitions, 27 Dignity and price, 188
Due, 117
Duties, perfect
and
imperfect,
147,
150,
171-2
to self
Analysis,
method
of,
256
and
others, 148, 172
Antagonism, 124, 139 Anthropology, practical, 32

Antinomies, 231
Dutiful, 117
Duty and a good will, 46 and necessity, 22

ff.
,
Antinomy, of freedom and necessity, 266

Apperception, pure, 237
Arbitrium, and will, 2 1 3 brutum, 210, 215
liberum, 210,
motive
sense
of,
1 1
50
8
ff.,
18
of,
215
EMOTIONS,
generous, motive
of,
53
ff.
Argument, analytic or regressive, 130 Arguments, analytic and synthetic, 26-7,

29
, progressive Attention, 235
Empiricism, 30 Ends, i66fT. and grounds, 167, 169

in themselves, 167
,
and
i
regressive, 27,
of,
29
9
ff.
kingdom
of,
9
185
ff.
Autonomy, imperative
,
247 246
realisation of, 191
principle
of,
180, 199, 243,
private, 187
Enlightenment, 196
Enthusiasts, moral, 40, 51
BEAUTY, 144, 184 Behaviour, animal, 219 Bonum consummatum, 43, 192
CATEGORIES,
143, 231, 238-9 of the understanding, 97, 100 Category, pure, 97 schematised, 99 , Causality and law, 2 1 1 two senses of, 209 ,
Epicurean, 57 Ethics and metaphysics, 255 , applied, 23, 32 , Aristotelian, 155 ff. , pure, 23 Examples, use of, 24-5 Exhibition, 158
, ,
practical, 162
symbolic, 159, 184 Experience, 22
Cause and
effect,
97
Christianity, 196-7 Circle, vicious, 28,
FALLACY,
224
1
ft'.
naturalistic,
1 1
Feeling, moral, 57, 118

1
Coherence of wills,
39, 151,
94
Concept, different kinds of, 100 Condition, limiting, 141-2, 178

Conditions, human, 46 Conflict of reason and desire, 254 Consequences, 153 , ignoring of, 76
Formalism, 74, 186 Freedom, 214
and autonomy, 212 and ends, 175, 181 and maxims, 214 and necessity, 209, 276
,
-,
defence
degrees
of, of,
266
ff.,
277
Contentment, 50, 57
215, 276, 277
280
INDEX
Freedom, Idea of, 99, 211, 271 law of, 69, 211, 212 no explanation of, 273
positive concept of, 211 possibility of, 268
Imagination, 143, 144, 22gn. Immortality, 256 Imperative, categorical, 115, 127 , , application of, 131, 192 fF.
, , ,
, ,
184-5,
presupposition
to act badly,
of,
217
possibility of, 127, 199,
204
275
definition of, 114 hypothetical, 1 1 5
kinds
-,
of,
114
of,
GOD,
',
,
,
Imperatives, categorical, 127
256
conception
existence
of,
ground
of, of,
159 188
188, 196
hypothetical, ground in general, 1 1 3

fF. fF.
168, 170, 176, 181 of, 168, 169
Kingdom
at,
Good,
,
absolute, 34
120 , possibility of, Inclinations, 49-50, 55
104 104
conditioned, 39
for,
IndifFerence, liberty of, 214, Insight, direct, 245 fF.

, practical, 258 Intellectualism, modern, 259 Interest and obligation, 258
275-6
fF.
highest, 41 in general, 103 moral, 107
,
,
, ,
my, 105
realistic
fF.
perfect; 43, 192
,
,
view of, teleological view
no
of,
mediate and immediate, 75 moral, 223, 256 pathological, 75, 83, 182
4
fF.
'I
flf.
108
ought' and
I will',
of,
114, 223,
259
Isolation,
method
47
fF.
unconditioned, 34 without limitation, 34 Good-heartedness, 54 Goodness and obligation, 116

,
,
JUDGEMENT,
moral, canon
fF.
of,
37
contributory, 36-7
is
fundamental, 45 motive of, 52 Goods, firima facie 36, 38-9 Gymnastic, ethical, 261
, ,
Judgements, a priori, 20 -, moral, 20
KINDNESS,
fF.
152, 173
Knowledge, composite, 22
HAPPINESS and
,
,
duty, 49, 55
85, 92,
1
pursuit
of,
85
of,
two views
06, 126
LATITUDE,
Law and
as
91, 92, 137, 148, 194
Hegelians, 84
Law-abidingness, 71
universality,
Heteronomy, 212
9
69
tvô kinds of,
214
of,
command, 70
69
of,
History, philosophy Holiness, 52
195
, ,
as such,
Honour, 53 Humanity, 165, 198
157 moral, and moral imperative, 70 of freedom, 69, 211, 212 of nature, 69, 146 fF.
teleological, 149
fF.
form
universal, 133
ff.
IDEA,
93n.
fF.,
of reason, 93, 97, 99 of system, 248

,
143, 263
Laws, moral, 23, 70 Legalism, 75

Legislation, universal, 180
, ,
practical, 190, 263, 271 practical and theoretical, 191, 263 use of, regulative and constitutive,
Liberty of indifFerence, 214, 275-6 Life, 234n.

,
organic, 191
100 Ideal of imagination, 127 of reason, 187
Loans, promises to repay, 152, 172 Love, 65, 196 , motive of, 54, 57n.
282
INDEX
73, 75,
1
MATTER of moral action,

Maxim, 59
,
85-6
Principles, application of, 17, 131, 174, 183

,
, ,
ff.,
6on.
formal, 72, 77 of duty, 6 1 -2 Maxims, 135 and material, 61 , -formal 182 , legislating through, , material, 137 Means and end, 83 , constituent, 105 Method, the Critical, 19 ff., 27
a priori, 28 conditioned and unconditioned, 95 objective, of reason, 247 ff.

practical, subjective
ff.,
and
objective,
59
89, 113
Progress, moral, 194 Propositions, analytic, 120 ff, 246

, ,
self-evident,
28
ff.
Mind
affects itself,
235
ff.
1 1
synthetic, 122 a priori, 122, 127 , synthetic Prudence, 51
Morality, commands laws of, 1 1 6 ,

, ,
of,
paradox
of, 192 principle of, 93

of,
Morals, metaphysic
,
31
ff.
Motive of reverence, 66 two senses of, 67 Motives, 117-18
counsels of, 115 imperatives of, 126 Psychology, general, 32 , practical, 32 Purposiveness, logical, in nature, 191 without a purpose, i49n.
, ,
RATIONALISM,
Realm
N A T u R E kingdom
,
30
or kingdom, 187
of,
of,
185, 190
Reason, activity
238
law
,
,
69, 146 ff. perfection of, 150

of,
purposes
of,
154
,
Necessitation,
Necessity and , natural, 209
3 duty, 22
ff.
1 1
the unconditioned, 249 as a power of Ideas, 239 as such, 219, 221 different senses of, 79, 96
fact of, 221
and
,
,
function
logical
of,
44-5, 78
real use of,
Noumenon,
different senses of,
269
, ,
, ,
and
97
OBJECTS of a good will, 43 Obligation and goodness, 1 1 6 Obstacles, 46

PERMISSIBLE and
Pflichtmassig
practical reason practical principles of, 240, 247 ff. pure, fact of, 203 self-consciousness of, 220, 258
see
ff.
ff.,
262
, ,
,
theoretical,
,
,
and
action, 81
and
presupposition, 2 1 8 principles of, 200, 238, 248

its
obligatory, 141
1 1
and pflichtvoll*
of,
Phenomena and noumena, 269

Philosophy, the task
Pleasure, 106-7
Politics,
23
194
Religion, 188, 196 Respect, 64, 65n. Results of action, and moral value, 58 Reverence, 42, 63 ff., 224
philosophy
of,
Revolution, Copernican, 227 Rules, moral, 23
Practical reason, 80, 81

.
and its presupposition, 2 1 8 and purpose, 155-6

,
SACRIFICE,
166
succession,
critique of, 31
Schema of regular
98
,
, ,
,
denial of, 87 function of, 78

limits of, 271
Schemata, 144, 158

, transcendental, 143, 144, 159 Schwarmerei, 52
principles of, 59
ff.,
89, 113,
Self,
220, 240, 248 , pure, 94

,
9 ,
phenomenal and noumenal, 234 and substance, 234n.

,
real,
253
self-consciousness of, 262
ff.
Self-evidence, 237
Self-love, principle of, 91
Prerogative, 189
Price, 189
Sense, inner, 233
Principle,
59
moral, 65
INDEX
Sense, passivity of, 228 ff. Skill, imperatives of, 123 ff. ^ principle of, 90
,
283
239
intuitive,
UNDERSTANDING,
,
100
ff.
rules of, 115
Universal, concrete, 101 Utilitarianism, 58
Soul, 23411.
Space, 230 Spontaneity, aesthetic, 144 , intellectual, 142

Standpoint, 272 Standpoints, two, 225
Stoic,
ff.,
VALUE,
i8 9
aesthetic,
economic, and moral,
Virtuous, fantastically, 40, 51

240, 267
57
216
Stoics, 189,
WILL,
and
subjective,
<
82, 103, 166, 208, 211, 212, 219,
Succession, objective Suicide, 154, 172
260
,
Symbols, 159
Synthesis,
method
of,
of,
29
,
,
System, Idea
248
arbitrium, 213 good, 34 ff, 177, 201 holy, 46, 52 individual, unity of, 1 93 supreme, 193 under moral laws, 2 1 3
and
World,
44, IQ$, IQI, 26l Term, third, 128, 213, 244-5 Theories, empirical and mystical, 163
,
intellectual,
TELEOLOGY,
, ,
9
intelligible, 190,
,
238 223 ff. and sensible, 227 ff., 243 9 ground of sensible world,
positivistic,
20
250
,
ff.
, , 9
Thought and
action, 263 Thing-in-itself, 269 ff.
knowledge of, 243, 268, 272 membership of, 233 ff.

thought
of,
Things~in-themselves, 227
ff.
9
,
271
ff.
Time, 230 Type, 1 60
noumenal and phenomenal, 227 Worth, moral, 47 ff.

(H.J. Paton) The Categorical Imperative - A Study in Kant's Moral Philosophy

Uploaded by

Copyright:

Available Formats

(H.J. Paton) The Categorical Imperative - A Study in Kant's Moral Philosophy

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

(H.J. Paton) The Categorical Imperative - A Study in Kant's Moral Philosophy

Uploaded by

Copyright:

Available Formats

The Categorical Imperative

study of Kant's most important contribution

merely for Kantian scholars

appendices), but for

cerned with ethical problems, and not

are dissatisfied with

argues that Kant,

as he himself claimed, has succeeded in

formulating the fundamental principles

on which the moral

endeavours to describe and to defend

especially in moral action

the present time.

ward, not as a mere historical investigation,

but as a contribution to the

moral thinking which

In the course of this study

THE CATEGORICAL IMPERATIVE

action cannot be good if

natural inclination towards

of consequences but will be concerned

only with his

even that without any regard to

cumstances he will deduce his various

from the bare concept of universal

showing such absurdities to

interpretation of Kant's doctrines

Kant's Metaphysic of Experience

THE SENIOR SERIES

White's Professor of Aforal Philosophy

HUTCHINSON'S UNIVERSITY LIBRARY

Printed in Great Britain by The Anchor Press, Ltd.,

THE APPROACH TO THE CATEGORICAL IMPERATIVE

THE CRITICAL METHOD

understanding Kant's method.

THE GOOD WILL

A good will is good without limitation.

highest good. and' its objects. function of reason.

DUTY A good will

THE MAXIM OF MORALITY

for the law.

description of reverence. function of reverence.

THE BACKGROUND OF THE CATEGORICAL IMPERATIVE

PRACTICAL REASON AND

IX PRACTICAL REASON AND PRINCIPLES

Conditioned and unconditioned principles. Appendix Kant's View of Reason.

of freedom. Different kinds of concept.

103 104 105 107 108 110

Obligation and goodness. to act morally.

113 114 114 115 1 1 6

HOW ARE IMPERATIVES POSSIBLE?

Analytic propositions. Synthetic propositions. Synthetic a priori propositions.

120 120 122

123 1 23 125 126

THE FORMULATION OF THE CATEGORICAL IMPERATIVE

THE FIVE FORMULAE

129 130 130

XIV THE FORMULA OF UNIVERSAL LAW

Formula I. The one categorical imperative.