.* • 
 
 h 
 
 J 
 
 .* 
 
 H 
 
 * 
 
ISeRARY 
 
THE PHILOSOPHY OF KANT. 
 
PUBLISHED BY 
 
 JAMES MACLEHOSE AND SONS. GLASGOW, 
 ^oblishers to the anibersitg. 
 
 MACMILLAN AND CO., LONDON AND NEW YORK. 
 
 London, . . . Sivipkin, Marshall, Hamilton, 
 
 Kent, and Co., Liviited. 
 Cambridge, . , AIacviilla7i and Boiues. 
 Editiburgh, . , Do7iglas a7td Fojilzs. 
 
 MDCCCXCVn. 
 
THE 
 
 PHILOSOPHY OF KANT, 
 
 «c 
 
 AS CONTAINED IN EXTRACTS FROM 
 HIS OWN WRITINGS. 
 
 SELECTED AND TRANSLATED' BY ■ ' ' ^. 
 
 JOHN WATSON, LL.D., 
 
 PROFESSOR OF MORAL PHILOSOPHY IN THE UNIVERSITY OF QUEEn's COLLEGE, 
 KINGSTON, CANADA, AUTHOR OF " KANT AND HIS ENGLISH CRITICS." 
 
 NEW EDITION, 
 
 GLASGOW: 
 
 JAMES MACLEHOSE & SONS, 
 
 ^ttlDiishers t^ the ©nibersiti). 
 
 1897. 
 

 WOLOGf 
 hBRARY 
 
 
 • • 
 
 < < t « « 
 
 • • • 
 
 • ••*•• * • 
 
 ^yj^^ ^<fr?^. Uc^. 
 
 t .^K ' i 
 
 - t''- ' » L. 
 
 GLASGOW : PRINTED AT THE UNIVERSITY PRESS 
 BY ROBERT MACLEHOSE AND CO. 
 
i 
 
 PREFATORY NOTE. 
 
 My reason for presenting to the public these transla- 
 tions from the philosophical writings of Kant will be best 
 understood if I state how they came to be made. The 
 teacher of philosophy soon finds that a very powerful 
 irritant is needed to awaken his pupils from their " dog- 
 matic slumber." I do not doubt that it is possible to 
 secure the desired end by a systematic criticism of the 
 preconceptions that stand in the way of genuine philo- 
 sophical comprehension. But my experience is that it 
 is almost impossible, by this method, to prevent the 
 average student from accepting what he is told without 
 mastering it and making it his own. Thus he passes 
 from one form of dogmatism to another, and with the 
 new dogmatism comes the great enemy of all education, 
 a conceit of knowledge without its reality. The study of 
 philosophy is of little value if it does not teach a man to 
 think for himself. The process of self-education is 
 necessarily a severe one, and, therefore, distasteful to 
 the natural man. Yet any attempt to evade it by some 
 " short and easy method " defeats the end. What is 
 required is a process by which the student who is really 
 in earnest may pass, gradually and surely, from a lower 
 to a higher plane of thought. The philosophical writings 
 
 /. ^' o ^ > ^ n 
 
VI PREFATORY NOTE. 
 
 of Kant, which exhibit in brief the transition from the 
 old to the new, I beheve to be a potent instrument for 
 this end. But the struggle upwards must be made by 
 the student himself. A man may hear, and seem to 
 appreciate, a course of lectures on the Critical philo- 
 sophy, containing a clear, and even a full statement of 
 it, and may yet fail to enter into its spirit. To obviate 
 this danger as far as possible, I tried some years ago 
 what could be done by throwing the student more upon 
 himself. My plan was to set a class of more advanced 
 pupils at work upon extracts from the philosophy of 
 Kant, to watch them as they forced their way through its 
 perplexities, and to put forth a helping hand only when 
 it seemed to be needful. The experiment justified 
 itself No method that I have tried — and I have tried 
 several — has been so fruitful in results. 
 
 The limited edition of Extracts, originally printed 
 for the use of my own students, but also used in other 
 American Universities, is now out of print. I have, 
 therefore, gone carefully over the writings of Kant again, 
 selecting and re-translating all the passages that seem 
 to be essential to the understanding of his philosophy. 
 The Extracts have been taken from four treatises — the 
 Critique of Pure Reason, the Metaphysic of Morality, the 
 Critique of Practical Reason, and the Critique of Judgment. 
 
 In the translations I have sought to express Kant's 
 meaning as clearly and simply as I could; and in no 
 case, so far as I am aware, have I been biassed by a 
 pre-conceived theory of what he ought to say. To 
 render Kant into intelligible English I have not found 
 an easy task, but it has been made much lighter for me 
 by the labours of my predecessors, Mr. Meiklejohn, Mr. 
 
PREFATORY NOTE. VU 
 
 Max Miiller, Dr. Hutchison Stirling, Mr. Mahaffy, and 
 Mr. Abbott, to whom I beg to express my obligations. 
 My very special thanks are also due to Professor Edward 
 Caird, of Glasgow University, for his great kindness in 
 reading the whole of the manuscript, and making a 
 number of valuable suggestions. 
 
 The pages of the first and second editions of the 
 Kritik der reinen Vermmft are given on the margin — the 
 former to the right, the latter to the left. The references 
 in other cases are to the complete edition of Kant's 
 works published by Hartenstein in 1867. The Index at 
 the end of the volume, which I have tried to make as 
 complete as possible, will, I hope, be found useful. 
 
 What I call the Preface to the Critique of Pure Reason 
 really consists of two prefaces thrown into one ; but the 
 extracts have been taken mainly from the preface to the 
 second edition, though a few passages from that to the 
 first edition have been inserted. Here, and in one or 
 two other places, I have made a slight change from the 
 order of the original ; but the transpositions are few, and 
 are sufficiently indicated by the references on the margin. 
 As a rule, my editorial privilege has been exercised only 
 in the way of omission. 
 
 I am well aware that objection may be taken to the 
 whole principle of these Extracts. The work of a 
 great author, it may be said, should be represented *' all 
 in all, or not at all." The objection is not without 
 force, but it seems to me to apply mainly to the selec- 
 tion of disconnected passages, and to the mutilation of a 
 faultless work of art like the Republic of Plato. The 
 writings of Kant, which are full of confusing repetitions 
 that really mar their perfection of form, hardly deserve 
 
Vlll PREFATORY NOTE. 
 
 the same tenderness of treatment. This is a case in 
 which it may be doubted if the less does not contain the 
 greater and even more. At least it is safe to say that 
 most students are more likely to turn to the full text of 
 Kant after a study of the more important passages in his 
 works, than if they had to make their way against greater 
 obstacles. No doubt there are suggestive points which 
 the plan of this work has compelled me to omit, but I 
 have tried to reduce these to a minimum. I believe that 
 what is here given contains all the main ideas of Kant in 
 their systematic connection. It is to be hoped, how- 
 ever, that the student who has mastered these Extracts 
 will not be satisfied until he has read all that Kant has to 
 say. 
 
 University of Queen's College, 
 Kingston, Canada, 
 May, 1888. 
 
CONTENTS. 
 
 Critique of Pure Reason 
 
 
 
 
 PAGE 
 
 Preface, 
 
 I 
 
 Introduction, .... 
 
 
 
 7 
 
 Transcendental iEsthetic, . 
 
 
 
 22 
 
 Section I. — Space, 
 
 
 
 23 
 
 Section II. — Time, . 
 
 
 
 29 
 
 General Remarks, 
 
 
 
 36 
 
 Transcendental Logic, 
 
 
 
 40 
 
 Transcendental Analytic, 
 
 
 
 46 
 
 Book I. — Conceptions, 
 
 
 
 46 
 
 Chapter I. — Guiding-thread for Discovery ol 
 
 F 
 
 Categories, . . . . 
 
 46 
 
 Chapter II. — Deduction of Categories, . 
 
 53 
 
 Book II. — Judgments, 
 
 83 
 
 Chapter I. — Schematism of Categories, . 
 
 84 
 
 Chapter II. — Principles of Pure Understandinj 
 
 ?, 92 
 
 I. Axioms of Perception, 
 
 92 
 
 2. Anticipations of Observation, . 
 
 96 
 
 3. Analogies of Experience, . 
 
 lOI 
 
 
 106 
 
 B. Causality, .... 
 
 no 
 
 C. Community, 
 
 118 
 
 4. Postulates of Empirical Thought, 
 
 122 
 
 General Remark, .... 
 
 126 
 
 Chapter III. — Phenomena and Noumena, 
 
 129 
 
 Transcendental Dialectic, 
 
 135 
 
 Book I. — Ideas, 
 
 140 
 
 Book II. — Dialectical Conclusions, 
 
 143 
 
 Chapter I. — Paralogisms, 
 
 145 
 
 Chapter II.— Ant 
 
 inomi 
 
 les, 
 
 155 
 
CONTENTS. 
 
 Chapter III. — The Ideal, .... 
 Ontological Proof, ..... 
 Cosmological Proof, 
 Physico-theological Proof, 
 
 Metaphysic of Morality. 
 
 Section I. — Transition from Ordinary Moral Conceptions to 
 the Philosophical Conception of Morality, . 
 
 Section II. — Transition from Popular Moral Philosophy to 
 the Metaphysic of Morality, ..... 
 
 Section III. — Transition from the Metaphysic of Morality 
 to the Critique of Practical Reason, .... 
 
 PAGE 
 
 195 
 
 204 
 
 210 
 218 
 
 225 
 232 
 250 
 
 Critique of Practical Reason. 
 
 Book I. — Analytic, 
 
 Chapter I. — Principles, 
 
 Chapter II. — Object, . 
 
 Chapter III. — Motives, 
 Book II. — Dialectic, . 
 
 Chapter I. — General Consideration, 
 
 Chapter II. — The Summum Bonum, 
 
 Antinomy, 
 
 Critical Solution of Antinomy, 
 
 Immortality of the Soul, . 
 
 Existence of God, 
 
 Postulates, 
 
 Extension of Practical Reason, 
 
 I. 
 
 2. 
 4- 
 
 5. 
 6. 
 
 7- 
 
 8. Faith, 
 
 261 
 261 
 280 
 284 
 289 
 289 
 291 
 292 
 
 293 
 294 
 
 296 
 
 298 
 300 
 302 
 
 Critique of Judgment. 
 
 Introduction, .... 
 Critique of Teleological Judgment, 
 
 Section I. — Analytic, . 
 
 Section II. — Dialectic, 
 
 Appendix on Method, 
 
 Index, 
 
 307 
 323 
 323 
 331 
 343 
 
 351 
 
THE CRITIQUE OF PURE REASON. 
 
THE PHILOSOPHY OF KANT. 
 
 THE CRITIQUE OF PURE REASON. 
 
 PREFACE. 
 
 This may well be called the age of criticism, a criti- yi u 
 cism from which nothing need hope to escape. When 
 religion seeks to shelter itself behind its sanctity, and law 
 behind its majesty, they justly awaken suspicion against 
 themselves, and lose all claim to the sincere respect which 
 reason yields only to that which has been able to bear 
 the test of its free and open scrutiny. 
 
 Metaphysic has been the battlefield of endless ii 
 conflicts. Dogmatism at first held despotic sway ; but iii 
 . . . from time to time scepticism destroyed all 
 settled order of society ; . . . and now a widespread iv 
 
 xiv indifferentism prevails. Never has metaphysic been so 
 fortunate as to strike into the sure path of science, but 
 has kept groping about, and groping, too, among mere 
 
 XV ideas. What can be the reason of this failure ? Is a 
 science of metaphysic impossible ? Then, why should 
 nature disquiet us with a restless longing after it, as if it 
 were one of our most important concerns ? Nay more, 
 how can we put any faith in human reason, if in one of 
 the very things that we most desire to know, it not merely 
 
♦ I 
 
 c n 
 
 2 tftE. PHILOSOPHY OF KANT. 
 ..•.•'..* • * 
 
 * forsakes ii*3,-b*i>t'riires us on by false hopes only to cheat 
 us in the end ? Or are there any indications that the 
 true path has hitherto been missed, and that by starting 
 afresh we may yet succeed where others have failed? 
 
 xvi It seems to me that the intellectual revolution, by 
 which at a bound mathematics and physics became what 
 they now are, is so remarkable, that we are called upon 
 to ask what was the essential feature of the change that 
 proved so advantageous to them, and to try at least to 
 apply to metaphysic as far as possible a method that has 
 xi been successful in other sciences of reason. In mathe- 
 matics I believe that, after a long period of groping, the 
 true path was disclosed in the happy inspiration of a single 
 man. If that man was Thales, things must suddenly 
 
 xii have appeared to him in a new light, the moment 
 he saw how the properties of the isosceles triangle could 
 be demonstrated. The true method, as he found, was 
 not to inspect the visible figure of the triangle, or to 
 analyze the bare conception of it, and from this, as it 
 were, to read off its properties, but to bring out what was 
 necessarily implied in the conception that he had himself 
 formed a priori, and put into the figure, in the construc- 
 tion by which he presented it to himself 
 
 Physics took a much longer time than mathematics to 
 enter on the highway of science, but here, too, a sudden 
 revolution in the way of looking at things took place. 
 When Galileo caused balls which he had carefully weighed 
 to roll down an inclined plane, or Torricelli made the air 
 bear up a weight which he knew beforehand to be equal 
 
 xiii to a standard column of water, a new light broke on 
 the mind of the scientific discoverer. It was seen that 
 reason has insight only into that which it produces after a 
 
THE CRITIQUE OF PURE REASON. 3 
 
 plan of its own, and that it must itself lead the way 
 with principles of judgment based upon fixed laws, and 
 force nature to answer its questions. Even experimental " 
 physics, therefore, owes the beneficial revolution in 
 
 xiv its point of view entirely to the idea, that, while reason 
 can know nothing purely of itself, yet that which it has 
 itself put into nature must be its guide to the discovery 
 of all that it can learn from nature. 
 
 xvi In metaphysical speculations it has always been v 
 assumed that all our knowledge must conform to objects ; 
 but every attempt from this point of view to extend our ' ^,jjJ^ 
 knowledge of objects a priori by means of conceptions 
 has ended in failure. The time has now come to ask, 
 whether better progress may not be made by supposing 
 that objects must conform to our knowledge. Plainly ^ 
 this would better agree with the avowed aim of meta- 
 physic, to determine the nature of objects a priori, or 
 before they are actually presented. Our suggestion is 
 similar to that of Copernicus in astronomy, who, finding 
 it impossible to explain the movements of the heavenly 
 bodies on the supposition that they turned round the 
 spectator, tried whether he might not succeed better by 
 supposing the spectator to revolve and the stars to re- 
 
 xvii main at rest. Let us make a similar experiment in 
 
 metaphysic with /(?rr^//<?;2. If it were really necessary for ^ aV*''^ 
 our perception to conform to the nature of objects, I do u^'^'^^ 
 not see how we could know anything of it a priori ; but 
 if the sensible object must conform to the constitution of 
 our faculty of perception, I see no difficulty in the matter. ^ 
 Perception, however, can become knowledge only if it ~ 
 is related in some way to the object which it determines. 
 Now here again I may suppose, either that the conceptions 
 
4 THE PHILOSOPHY OF KANT. 
 
 through which I effect that determination conform to the 
 objects, or that the objects, in other words the experience 
 in which alone the objects are known, conform to concep- 
 tions. In the former case, I fall into the same perplexity 
 as before, and fail to explain how such conceptions can be 
 known a priori. In the latter case, the outlook is more 
 hopeful. For, experience is itself a mode of knowledge 
 which implies intelligence, and intelligence has a rule of 
 its own, which must be an a priori condition of all 
 knowledge of objects presented to it. To this rule, as 
 xviii expressed in a priori conceptions, all objects of 
 experience must necessarily conform, and with it they 
 must agree. 
 
 Our experiment succeeds as well as we could wish, and 
 xix gives promise that metaphysic may enter upon the sure 
 course of a science, at least in its first part, where it is 
 occupied with those a priori conceptions to which the 
 corresponding objects can be given. The new point of 
 view enables us to explain how there can be a prio7'i 
 knowledge, and what is more, to furnish satisfactory 
 proofs of the laws that lie at the basis of nature as a 
 totality of objects of experience. But the consequences 
 that flow from this deduction of our faculty of a priori 
 knowledge, which constitutes the first part of our 
 inquiry, are unexpected, and at first sight seem to be 
 fatal to the aims of metaphysic, with which we have to 
 deal in the second part of it. For we are brou_ght to the 
 conclusion that we never can transcend the limits of 
 possible experience, and therefore never can realize the 
 object with which metaphysic is primarily concerned. 
 XX In truth, however, no better indirect proof could be given 
 that we were correct in holding, as the result of our 
 
THE CRITIQUE OF PURE REASON. 5 
 
 first estimate of the a priori knowledge of reason, that 
 such knowledge relates not at all to the thing as it exists 
 in itself, but only to phenomena. For that which 
 necessarily forces us to go beyond the limits of experience 
 and of all phenomena is the nncondiiioned^ which reason 
 demands ot things in themselves, and by right and 
 necessity seeks in the complete series of conditions for 
 everything conditioned. If, then, we find that we cannot 
 think the unconditioned without contradiction, on the 
 supposition of our experience conforming to objects as 
 things in themselves ; while, on the contrary, the contra- 
 diction disappears, on the supposition that our knowledge 
 does not conform to things in themselves, but that objects 
 as they are given to us as phenomena conform to our 
 knowledge ; we are entitled to conclude that what we at 
 xxi first assumed as an hypothesis is now estabhshed as a 
 truth. 
 
 It may seem from this that the result of our critical 
 investigation is purely negative, and merely warns us not 
 to venture with speculative reason beyond the limits of 
 experience. And no doubt this is its first use ; but a 
 positive result is obtained when it is seen that the prin- 
 ciples with which speculative reason ventures beyond its 
 proper limits, in reality do not extend the province of 
 reason, but inevitably narrow it. For in seeking to go 
 xxv altogether beyond its true limits, the limits of sensibility, 
 those principles threaten to supplant pure reason in its 
 xxvii practical aspect. Let us suppose that the necessary 
 distinction which our criticism shows to exist between 
 things as objects of experience and the same things as 
 they are in themselves, had not been made. Then 
 the principle of causality, and with it the mechanical 
 
6 THE PHILOSOPHY OF KANT. 
 
 conception of nature as determined by it, would apply to 
 all things in general as efficient causes. Hence I could 
 not, without palpable contradiction, say of the same being, 
 for instance the human soul, that its will is free, and yet 
 is subject to the necessity of nature, that is, is not free. 
 But, if our criticism is sound and the object may be 
 taken in two distinct senses, on the one hand as a 
 phenomenon, and on the other hand as a thing in itself; 
 xxviii there is no contradiction in supposing that the very same 
 will, in its visible acts as a phenomenon, is not free ^ but 
 necessarily subject to the law of nature, while yet, as 
 belonging to a thing in itself, it is not subject to that law, but 
 xxix \sfree. Now, morality requires us only to be able to think 
 freedom without self-contradiction, not to understand it ; 
 it is enough that our conception of the act as free puts no 
 obstacle in the way of the conception of it as mechanically 
 necessary, for the act stands in quite a different relation 
 to freedom from that in which it stands to the mechanism 
 of nature. From the critical point of view, therefore, 
 the doctrine of morality and the doctrine of nature may 
 each be true in its own sphere ; which could never have 
 been shown had not criticism previously established our 
 unavoidable ignorance of things in themselves, and 
 
 XXX limited all that we can krww to mere phenomena. I 
 have, therefore, found it necessary to deny knowledge of 
 God^ freedo??i, and im7nortality^ in order to find a place 
 iox faith. 
 
 It is dogmatism, or the preconception that progress in 
 metaphysic may be made without a previous criticism of 
 pure reason, that is responsible for that dogmatic unbelief 
 
 xxxi which is so hostile to morality. The first and most 
 important task of philosophy is to deprive metaphysic 
 
THE CRITIQUE OF PURE REASON. 7 
 
 once for all of its pernicious influence by closing up the 
 XXXV sources of its errors. Our critique is not opposed to the 
 dogmatic procedure of reason as a science of pure know- 
 ledge, which must be strictly proved a priori from well- 
 founded principles, but only to dogmatism, that is, to 
 the presumption that we may follow the time-honoured 
 method of constructing a system of pure metaphysic out 
 of principles that rest upon mere conceptions, without 
 first asking in what way reason has come into possession 
 of them, and by what right it employs them. Dogmatism, 
 in a word, is the dogmatic procedure of reason without 
 any previous criticism of its own powers. 
 
 The critique of pure reason is not a criticism of books vi ^ , \ 
 and systems, but of the faculty of reason in general, in so ^' LmM 
 far as reason seeks for knowledge that is independent ^^ J 
 of all experience. I have evaded none of its questions, ^ 
 on the plea of the imbecility of human reason. In fact, vii 
 reason is so perfect a unity that, if it were in principle 
 inadequate to the solution of even a single one of the 
 questions which by its very nature it raises, we might at 
 once with perfect certainty set it aside as incapable of 
 
 xxxvii answering any of the others. For as it is a true organic 
 unity, in which the whole exists for the sake of each of 
 the parts, and each part for the sake of the whole, 
 
 xxxviii the slightest imperfection, whether it is due to a flaw 
 or to a defect, will inevitably betray itself in use. 
 
 INTRODUCTION. 
 
 r. Distinction of Pure and Empirical Knowledge. 
 
 There can be no doubt whatever that all our knowledge " 
 begins with experience. By what means should the 
 
8 THE PHILOSOPHY OF KANT. 
 
 faculty of knowledge be aroused to activity but by 
 objects, which, acting upon our senses, partly of them- 
 selves produce ideas in us, and partly set our understand- 
 ing at work to compare these ideas with one another, 
 and, by combining or separating them, to convert the 
 raw^ material of our sensible impressions into that know- 
 
 ,' ledge of objects which is called experience? In the 
 order of time, therefore, w^e have no knowledge prior to 
 experience, and with experience all our knowledge 
 begins. 
 
 -- But, although all our knowledge begins with experience, 
 it by no means follows that it all originates fro77i experi- 
 ence. For it may well be that experience is itself made 
 up of two elements, one received through impressions ot 
 sense, and the other supplied from itself by our faculty 
 of knowledge on occasion of those impressions. If that 
 
 2 be so, it may take long practice before our attention is 
 drawn to the element added by the mind, and we learn 
 to distinguish and separate it from the material to which 
 it is applied. 
 
 It is, therefore, a question which cannot be lightly put 
 aside, but can be answered only after careful investiga- 
 tion, whether there is any knowledge that is independent 
 of experience, and even of all impressions of sense. 
 Such knowledge is said to be a priori, to distinguish it 
 from empirical knowledge, which has its sources a 
 posteriori, or in experience. 
 
 The term a prioi'i must, however, be defined more 
 precisely, in order that the full meaning of our question 
 may be understood. We say of a man who undermines 
 the foundations of his house, that he might have known 
 a priori that it would fall ; by which we mean, that he 
 
THE CRITIQUE OF PURE REASON. 9 
 
 might have known it would fall, without waiting for the 
 event to take place in his experience. But he could 
 not know it completely a priori ; for it is only from ex- 
 perience that he could learn that bodies are heavy, and 
 must fall by their own weight when there is nothing to 
 support them. 
 
 3 By <2 priori knowledge we shall, therefore, in what 
 follows understand, not such knowledge as is inde- 
 pendent of this or that experience, but such as is 
 absolutely independent of all experience. Opposed to 
 it is empirical knowledge, or that which is possible 
 only a posteriori^ that is^ by experience. A priori 
 knowledge is pure^ when it is unmixed with anything 
 empirical. The proposition, for instance, that each 
 change has its own cause is a priori^ but it is not pure, 
 because change is an idea that can be derived only 
 
 ^ from experience. 
 
 2. Science and Common Sense contain a priori 
 
 Knowledge. 
 
 Evidently what we need is a criterion by which to 
 distinguish with certainty between pure and empirical 
 knowledge. Now, experience can tell us that a thing 
 is so and so, but not that it cannot be otherwise. 
 Firstly, then, if we find a proposition that, in being 
 thought, is thought as necessary, it is an a priori 
 judgment ; and if, further, it is not derived from any pro- 
 position except which is itself necessary, it is absolutely 
 a priori. Secondly, experience never bestows on its judg- 
 ments true or strict universality, but only the assumed 
 or comparative universality of induction; so that, properly 
 
 4 speaking, it merely says, that so far as our observation 
 
lO THE PHILOSOPHY OF KANT. 
 
 has gone, there is no exception to this or that rule. 
 If, therefore, a judgment is thought with strict univer- 
 sahty, so that there can be no possible exception to 
 it, it is not derived from experience, but is absolutely 
 a priori. Necessity and strict universality are, there- 
 fore, sure criteria of a priori knowledge, and are 
 also inseparably connected \nth each other. 
 
 Now, it is easy to show that in human knowledge there 
 actually are judgments, that in the strictest sense are 
 universal, and therefore pure a priori. If an example 
 from the sciences is desired, we have but to think of any 
 5 proposition in mathematics ; if an instance from common 
 sense is preferred, it is enough to cite the proposition, 
 that there can be no change without a cause. To take 
 the latter case, the very idea of cause so manifestly implies 
 the idea of necessary connection with an effect, that it 
 would be completely lost, were we to derive it, with 
 Hume, from the repeated association of one event with 
 another that precedes it, and were we to reduce it to the 
 subjective necessity arising from the habit of passing 
 from one idea to another. Even without appealing to 
 such examples to show that as a matter of fact there are 
 in our knowledge pure a priori principles, we might prove 
 a priori that without such principles there could be no 
 experience whatever. For, whence could experience 
 derive the certainty it has, if all the rules that it follows 
 were merely empirical and therefore contingent ? Surely 
 such rules could not be dignified with the name of first 
 principles. 
 
THE CRITIQUE OF PURE REASON. II 
 
 3. A Science is needed to deter 7?iine the possibility, the 
 principles, and the extent of all a priori Knowledge. 
 
 A far more important consideration remains than any- 
 thing that has yet been stated. There is a sort of 3 
 knowledge that even quits the field of all possible experi- 
 ence, and claims to extend the range of our judgments 
 beyond its limits, by means of conceptions to which no 
 corresponding object can be presented in experience. 
 Now, it is just in the province of this sort of knowledge, 
 where experience can neither show us the true path nor 
 put us right when we go astray, that reason carries on 
 7 those high investigations, the results of which we regard 
 as more important than all that understanding can dis- 
 cover within the domain of phenomena. Nay, we are 
 even willing to stake our all, and to run the risk of being 
 completely deluded, rather than consent to forego in- 
 quiries of such moment, either from uncertainty or from 
 carelessness and indifference. These unavoidable pro- 
 blems, set by pure reason itself, are God, freedom, and 
 i7nmortality, and the science which brings all its resources 
 to bear on the one single task of solving them is 
 metaphysic. 
 
 Now, one might think that men would hesitate to leave 
 the solid ground of experience, and to build an edifice of 
 truth upon knowledge that has come to them they know 
 not how, and in blind dependence upon principles of 
 which they cannot tell the origin, without taking the 
 greatest pains to see that the foundation was secure. One 
 might think it only natural, that they would long ago 
 have raised the question, how we have come into 
 possession of all this a priori knowledge, and what may 
 
12 THE PHILOSOPHY OF KANT. 
 
 8 be its extent, its import and its value. But the fact is, 4 
 that a part of this knowledge — mathematical knowledge, 
 for instance — has so long been established as certain, 
 that we are less ready to suspect the evidence for other 
 parts, although these may be of a totally different nature. 
 Besides, when we are once outside the circle of experi- 
 ence, we are sure not to be contradicted by experience ; 
 and so strong is the impulse to enlarge our knowledge, 
 that nothing short of a clear contradiction will avail to 
 arrest our footsteps. Now, such contradiction may easily 
 be avoided, even where we are dealing with objects that 
 are merely imaginary, if we are only careful in putting 
 our fictions together. Mathematics shows us by a 
 splendid instance, how far a science may advance a priori 
 without the aid of experience. It is true that by it 
 objects and conceptions are considered only in so far as 
 they can be presented in perception ; but it is easy to 
 overlook the limitation, because the perception in this 
 case can itself be given a priori, and is therefore hard to 
 
 ~ distinguish from a mere idea. Deceived by this proof of 5 
 the power of reason, we can see no limits to the extension 
 
 9 of knowledge. So Plato forsook the world of sense, 
 chafing at the narrow limits it set to our knowledge, and, 
 on the wings of pure ideas, launched out into the empty 
 space of the pure understanding. He did not see that 
 with all his efforts he was making no real progress. But 
 it is no unusual thing for human reason to complete its 
 speculative edifice in such haste, that it forgets to look 
 to the stability of the foundation. The reason why we 
 have no fear or anxiety while the work of construction is 
 going on, but take it for granted that the foundation 
 stands firm, is, that much of the work of reason, perhaps 
 
THE CRITIQUE OF PURE REASON. 1 3 
 
 the greater part, consists in the analysis of conceptions 
 10 which we already possess. This analysis really gives us 6 
 a kind of a priori knowledge that is safe and useful. 
 But, misled by this success, reason interpolates proposi- 
 tions of quite a different character, which but superficially 
 resemble the others. I shall therefore at the very outset 
 point out the distinction between these two kinds of 
 knowledge. 
 
 4. The distinctiofi between Analytic and Synthetic 
 
 Judg?ne7its. 
 
 There are two ways in which the predicate of an affirm- 
 ative judgment may be related to the subject. Either 
 the predicate B is already tacitly contained in the subject 
 A, or B lies entirely outside of A, although it is in some 
 way connected with it. In the one case I call the judg- 
 j J ment a?ialytic, in the other case synthetic. Analytic y 
 judgments are those in which the predicate is related to 
 the subject in the way of identity, while in synthetic 
 judgments the predicate is not thought as identical with 
 the subject. The former class might also be called 
 explicative, because the predicate adds nothing to the 
 subject, but merely breaks it up into its logical elements, 
 and brings to clear consciousness what was already 
 obscurely thought in it. The latter class we may call 
 ampliative, as adding in the predicate something that was 
 in no sense thought in the subject, and that no amount 
 of analysis could possibly extract from it. " Body is 
 extended," for instance, is an analytic judgment. For, 
 to be conscious that extension is involved in the con- 
 ception signified by the term body, it is not necessary to 
 go outside that conception, but merely to analyze it into 
 
14 THE PHILOSOPHY OF KANT. 
 
 the various logical elements that are always thought in 
 it. But in the proposition " Body has weight/' the predi- 
 cate is not implied in the very conception of body, but 
 is a perfectly new idea. The addition of such a predicate, 
 therefore, yields a synthetic judgment. 
 
 Judgments of experience are all by their very nature 
 synthetic. To say that I must have recourse to exper- 
 ience for an analytic judgment is absurd, because I can 
 frame the judgment without going beyond the conception 
 
 1 2 I already possess. I have, for instance, the conception 
 of body, and by mere analysis I become aware of the 
 attributes extension, impenetrabiHty, figure, etc., which 
 the thought of it involves. To enlarge my conception, 
 I turn again to experience, from which the conception 
 was originally derived, and, finding weight to be invari- 
 ably connected with those attributes, I attach it to them 
 by synthesis as a new attribute. The possibility of this 
 synthesis of the attribute weight with the conception body 
 therefore rests upon experience. The two ideas are 
 quite distinct, but they yet are parts of the same experi- 
 ence, and experience is itself a whole in which 
 a number of perceptions are synthetically though only 
 contingently combined. 
 
 In a priori synthetic judgments^ on the other hand, I n 
 
 1 2 can get no aid whatever from experience. But, if it is 
 here vain to look to experience for aid, on what other 
 support am I to rely, when I seek to go beyond a certain 
 conception A, and to connect B synthetically with it? 
 Take the proposition, that every event must have its 
 cause. No doubt I cannot have the conception of an 
 event without thinking of something as having a moment 
 of time before it, and from this certain analytic judg- 
 
THE CRITIQUE OF PURE REASON. 1 5 
 
 . ments may be derived. But the conception of a cause 
 lies entirely outside the conception of an event, and 
 introduces an idea not contained in it. By what right, 
 then, do I pass from the conception of an event to the 
 totally different conception of a cause ? How do I 
 know that there is a necessary connection between the 
 two conceptions, when I can perfectly well think the one 
 without the other ? What is here the unknown x, which 
 gives support to the understanding, when it seems to have 
 discovered an entirely new predicate B to belong neces- 
 sarily to the subject A? Experience it cannot be, 
 because the principle has a degree of universality that 
 experience can never supply, as it is supposed to connect 
 the new conception with the old in the way of necessity, 
 and must do so entirely a priori^ and on the basis of 
 mere conceptions. And yet our speculative a priori 
 knowledge must rest upon such synthetic or ampliative lo 
 propositions. 
 
 14 5. The principles of all Theoretical Sciences of reason 
 are a priori Synthetic Judgmejits. 
 
 (i) All mathematical judgments, without exception, 
 are synthetic. No doubt the mathematician, in his 
 demonstrations, proceeds on the principle of contradic- 
 tion, but it is a mistake to suppose that the propositions 
 on which his demonstrations rest can be known to be 
 true by that principle. The mistake arises from not 
 observing that, while a synthetic proposition may certainly 
 be seen to be true by the principle of contradiction, its 
 truth is in that case evident, not from itself, but only 
 because it is seen to follow from another proposition that 
 has been previously obtained by synthesis. 
 
1 6 THE PHILOSOPHY OF KANT. 
 
 The first thing to notice is, that no truly mathematical 
 judgments are empirical, but always are a priori. They 
 carry necessity on their very face, and therefore cannot 
 
 15 be derived from experience. Should any one demur to 
 this, I am willing to limit my assertion to the proposi- 
 tions of pure matheffiaiics, which, as everybody will 
 admit, are not empirical judgments, but perfectly pure a 
 prio7'i knowledge. 
 
 At first sight it may seem that the proposition 
 7 + 5 = 12 is purely analytic, and follows, by the principle 
 of contradiction, from the conception of a sum of 7 and 
 5. But, when we look more closely we see that the con- 
 ception of the sum of 7 and 5 is merely the idea of the 
 union of the two numbers, and in no way enables us to 
 tell what may be the single number that forms their sum. 
 To think that 7 and 5 are to be united is not to have the 
 conception 12, and I may analyze the idea of the possible 
 sum as long as I please, without finding the 12 in it. To 
 get beyond the separate ideas of 7 and 5, I must call in 
 the aid of perception, referring to my five fingers, or to 
 five points, and, starting with the conception 7, go on to 
 add to it, unit by unit, the 5 so presented to me in 
 
 16 perception. The propositions of arithmetic are therefore 
 all synthetic. This is even more manifest if I take larger 
 numbers, when it becomes at once obvious that without 
 fhe aid of perception no mere analysis of my concep- 
 tions, turn and twist them as I may, could ever yield the 
 sum. 
 
 Nor is any proposition of pure geometry analytic. 
 That the straight line between any two points is the 
 shortest, is a synthetic proposition. My idea of straight 
 is purely an idea of quality, not of quantity. From no 
 
THE CRITIQUE OF PURE REASON. 1 7 
 
 analysis of the conception of a straight line can the 
 knowledge that it is the shortest be derived. Per- 
 ception has to be called in to enable me to make the 
 synthesis. 
 
 '7 (2) The principles on yfhich physics rests are a priori 
 synthetic judgments. I shall content myself with citing 
 two such judgments : first, that in all changes of the 
 material world the quantity of matter remains the same ; 
 and, secondly, that in the communication of motion, 
 
 1 8 action and reaction are always equal. Both propositions, 
 it is plain, are not only necessary, and therefore in their 
 origin a priori^ but they are also synthetic. The con- 
 ception of matter does not include the idea of perman- 
 ence, but merely signifies its presence in the space 
 which it occupies. When, therefore, I say that matter is 
 permanent in quantity, I add /<? the conception of matter 
 an attribute which was not at first thought in it. Accord- 
 ingly, the proposition is not analytic, but at once a priori 
 and synthetic ; and so with the other propositions of 
 pure physics. 
 
 (3) Unsuccessful as metaphysic may hitherto have been 
 in solving the unavoidable problems set to it by human 
 reason, its aim undoubtedly is to acquire a priori syn- 
 thetic knowledge. That aim it certainly will never attain 
 by merely dissecting the conceptions of things which we 
 have in our mind a priori, and expressing them in 
 analytic propositions. For it seeks to enlarge our a 
 priori knowledge, and therefore it must try to show that 
 there are judgments that add to a conception something 
 not already contained in it, even if it should be led to 
 venture into a region where experience cannot follow, as 
 for instance in the proposition that the world must have 
 
 B 
 
1 8 THE PHILOSOPHY OF KANT. 
 
 had an absolute beginning. In its aim at least meta- 
 physic therefore consists entirely of a priori synthetic 
 propositions. 
 
 19 6. The Problem of Pure Reason. 
 
 It is of very great advantage, to others as well as to one- 
 self, to be able to bring together various topics of 
 investigation in a single problem. Now, the true problem 
 *" of pure reason may be put in this way — How are a 
 p7'iori synthetic judgments possible ? 
 
 20 Should this question be answered in a satisfactory way, 
 we shall at the same time learn what part reason plays 
 in the foundation and completion of those sciences which 
 contain a theoretical a priori knowledge of objects. 
 Thus we shall be able to answer the questions — How is 
 pure maiheinatics possible 2 How is pure physics possible '^ 
 As these sciences actually exist, we may fairly ask how 
 they are possible ; for that they must be possible is 
 
 21 proved by the fact that they exist. But as no real 
 progress has as yet been made in the construction of a 
 system that realizes the essential aim of metaphysic^ it 
 cannot be said that metaphysic exists, and there is, 
 therefore, reason to doubt whether it is possible 
 at all. 
 
 Yet in one sense metaphysic may certainly be said to 
 exist, namely, in the sense that there is in man a natural 
 
 22 disposition to seek for this kind of knowledge. But as 
 all attempts to answer the questions which human reason 
 is naturally impelled to ask, as, for instance, whether the 
 world had a beginning, or has existed from all eternity, 
 have always and unavoidably ended in self-contradiction ; 
 we cannot be satisfied with asserting the mere natural 
 
THE CRITIQUE OF PURE REASON. 1 9 
 
 disposition to metaphysical speculation, or, in other 
 words, with the bare ability of pure reason to construct 
 some sort of metaphysic. It must be possible for reason ^ 
 to attain to certainty one way or the other : we must be 
 able to ascertain whether reason can know the objects it 
 seeks, or whether it cannot know them ; we must find a 
 conclusive answer to the question whether pure reason is 
 capable or incapable of determining the nature of those 
 objects, and whether, therefore, its domain may with 
 confidence be enlarged beyond the limits of experience, 
 or must be restricted within them. Accordingly, the 
 third and last question, which flows from the general 
 problem of pure reason, may be correctly put in this way : 
 How is a science of metaphysic possible ? Thus a criticism 
 of reason in the end necessarily leads to science, whereas 
 the dogmatic employment of reason without previous 
 
 23 criticism can lead only to groundless assertions, to which 
 other assertions equally specious may always be opposed, 
 the inevitable result being scepticism. 
 
 24 7. Idea and Division of the Critique of Pttre Reason. 
 
 From all that has been said we get the idea of a unique 
 science, which may be called the Critique of Pure Reason. 
 
 25 It is not a doctrine, but a criticism of pure reason, and its 
 speculative value is entirely negative, because it does not 
 enlarge our knowledge, but only casts light upon the 
 nature of our reason and enables us to keep it free from 
 error. By transcendental knowledge I mean all know- 
 ledge that is occupied, not with objects, but with the way 
 in which a knowledge of objects may be gained, so far as 
 
 26 that is possible a priori. What we propose is not a 12 
 
20 THE PHILOSOPHY OF KANT. 
 
 doctrine of pure reason, but a transcendental criticism, 
 the purpose of which is not to extend knowledge, but to 
 rectify it, and to supply a touchstone of the value of all 
 
 V. a priori knowledge. 
 
 2 7 This transcendental criticism will afford a complete archi- ^ 3 
 tectonic plan of transcendental philosophy, as exhibited 
 in its principles, and will therefore give a perfect guar- 
 antee of the completeness and stability of the edifice in 
 all its parts. 
 
 28 The Critique of Pure Reason therefore contains all 14 
 that is essential to the idea of transcendental philosophy, 
 and if we distinguish it from that philosophy, the reason is 
 that it does not carry its analysis beyond what is required 
 
 in a complete estimate of a priori synthetic knowledge. 
 
 The main thing to be kept in view in the division of 
 such a science is that no ideas be allowed to enter that 
 are in any way of empirical origin, or, in other words, 
 that it consist only of perfectly pure a priori knowledge. 
 Hence, although the principles and fundamental concep- 
 tions of morality are a priori^ they form no part of a 15 
 
 29 transcendental philosophy, because they are necessarily 
 relative to the conceptions of pleasure and pain, desire, 
 and inclination, etc., which in their origin are empirical. 
 
 In a systematic division of this science we must have, 
 firstly, a doctrine of the elements^ secondly, a doctrine of 
 the 7nethod of pure reason. As to the subdivisions, it 
 seems enough to say at present that there are two stems 
 of human knowledge — Sensibility and Understanding, 
 which may perhaps spring from a common root, unknown 
 to us, and that by the one obj ects are given, by the other 
 
 30 they are thought. Now, if Sensibility is found to contain 
 an ^ /r/(?r/ element, without which objects could not be 
 
THE CRITIQUE OF PURE REASON. 21 
 
 given to us, an investigation into the nature of that 
 element will be one of the tasks of transcendental 
 philosophy. The doctrine of this transcendental element i6 
 of sensible perception will form the first part of the 
 science of elements, because we must consider the 
 conditions under which objects of human knowledge are 
 given, before we go on to inquire into the conditions 
 under which they are thought. 
 
22 THE PHILOSOPHY OF KANT. 
 
 TRANSCENDENTAL ESTHETIC. 
 
 I. 
 
 34 Sensation is the actual affection of our sensibility, or 20 
 capacity of receiving impressions, by an object. The 
 perception which refers itself to an object through 
 sensation, is empirical perception. The undetermined 
 object of such a perception is 2i phenomenon (Erscheinung). 
 
 That element in the phenomenon which corresponds 
 to sensation I call the matter^ while that element which 
 makes it possible that the various determinations of the 
 phenomenon should be arranged in certain ways relatively 
 to one another is its form. Now, that without which 
 sensations can have no order or form, cannot itself be 
 
 " sensation. The matter of a phenomenon is given to us 
 entirely a posteriori^ but its form must lie a priori in the 
 mind, and hence it must be capable of being considered 
 
 ^ by itself apart from sensation. 
 
 This pure form of sensibility is also called pure per- 
 
 35 ception. Thus, if from the consciousness of a body, I 
 separate all that the understanding has thought into it, as 
 substance, force, divisibiHty, etc., and all that is due to 
 sensation, as impenetrability, hardness, colour, etc. ; 2 1 
 what is left over are extension and figure. These, therefore, 
 belong to pure perception, which exists in the mind 
 
 a priori^ as a mere form of sensibility, even when no 
 sensation or object of sense is actually present. 
 
TRANSCENDENTAL ^ESTHETIC. 23 
 
 The science o f all the a tiriori principles of sensibility 
 ■^6 1 call Tran scejide7ital Esthetic ^ in contradistinction from 
 the s cier^q^ of the ^ ^inci^les^ of j)ure thought, whicK.I 
 ^ call Tran scendental Los^tc. 
 
 In Transcendental Esthetic we shall first of all isolate 22 
 sensibility, abstracting from all that the understanding 
 contributes through its conceptions, so that we may 
 have nothing before us but empirical perception. In the 
 next place, we shall separate from empirical perception 
 all that belongs to sensation ; when there will remain 
 only pure perception, or the mere form of phenomena, 
 ^ the sole element that sensibility can yield a priori. If 
 this is done, it will be found that there are two pure forms 
 of sensible perception, which constitute principles of 
 a p7'iori knowledge, namely. Space and Time. With 
 these it will now be our business to deal. 
 
 37 Section I. — Space. 
 
 2. Metaphysical Exposition of Space. 
 
 In external sense we are conscious of objects as out- 
 side of ourselves, and as all without exception in space. 
 In space their shape, size, and relative position are 
 marked out, or are capable of being marked out. 
 Inner sense, in which we are conscious of ourselves, 
 or rather of our own state, gives us, it is true, 
 no direct perception of the soul itself as an object ; but 
 it nevertheless is the one single form in which our own 23 
 state comes before us as a definite object of perception ; 
 and hence all inner determinations appear to us as 
 related to one another in time. We cannot be conscious 
 
24 THE PHILOSOPHY OF KANT. 
 
 of time as external, any more than we can be conscious 
 of space as something within us. What, then, are space 
 and time? Are they in themselves real things? Are 
 they only determinations, or perhaps merely relations of 
 things, which yet would belong to things in themselves 
 even if those things were not perceived by us ? Or, 
 ;^S finally, have space and time no meaning except as forms 
 of perception, belonging to the subjective constitution of 
 our own mind, apart from which they cannot be 
 predicated of anything whatever ? To answer these 
 questions I shall begin with a metaphysical exposition of 
 space. An exposition I call it, because it gives a distinct 
 although not a detailed, statement of what is implied in 
 the idea of space ; and the exposition is metaphysical^ 
 because it brings forward the reasons we have for 
 regarding space as given a priori. 
 
 (i) Space is not an empirical conception, which has 
 been derived from external experiences. For I could 
 not be conscious that certain of my sensations are 
 relative to something outside of me, that is, to something 
 in a different part of space from that in which I myself 
 am ; nor could I be conscious of them as outside of and 
 beside one another, were I not at the same time con- 
 scious that they not only are different in content, but are 
 ' in different places. The consciousness of space is, 
 therefore, necessarily presupposed in external perception. 
 No experience of the external relations of sensible things 
 could yield the idea of space, because without the 
 consciousness of space there would be no external 
 experience whatever. 
 
 (2) Space is a YitcQssd.xy a priori idea, which is pre- 
 supposed in all external perceptions. By no effort can 24 
 
TRANSCENDENTAL ESTHETIC. 25 
 
 39 we think space to be away, although we can quite 
 ^ readily think of space as empty of objects. Space we 
 
 therefore regard as a condition of the possibility of 
 phenomena, and not as a determination dependent on 
 ^.phenomena. It is thus a priori, and is necessarily pre- 
 supposed in external phenomena. 0^^\^~''^ ^ 
 
 (3) Space is not a discursive or general conception of 25 V>%, ^ 
 the relations of things, but a pure perception. For we ,%^^ . 
 can be conscious only of a single space. It is true that '^'" - ^ ^ * 
 we speak as if there were many spaces, but we really ^ .^-r" 
 mean only parts of one and the same identical space, w^^ 
 Nor can we say that these parts exist l?e/ore the one 
 all-embracing space, and are put together to form a whole; 
 
 but we can think of them only as in it. Space is essen- 
 tially single ; by the plurality of spaces, we merely mean 
 that because space can be limited in many ways, the 
 general conception of spaces presupposes such limitations 
 as its foundation. From this it follows, that an a priori 
 perception, and not an empirical perception, underlies all 
 conceptions of pure space. Accordingly, no geometrical 
 proposition, as, for instance, that any two sides of a 
 triangle are greater than the third side, can ever be 
 derived from the general conceptions of line and triangle* 
 but only from perception. From the perception, however, 
 it can be derived a priori, and with demonstrative 
 certainty. , ^ .-^ 
 
 (4) Space is presented before our consciousness aS'W»-V' 
 
 40 an infinite magnitude. Now, in every conception we 
 certainly think of a certain attribute as common to an 
 infinite number of possible objects, which are subsumed 
 under the conception ; but, from its very nature, no 
 conception can possibly be supposed to contain an 
 
26 THE PHILOSOPHY OF KANT. 
 
 infinite number of determinations within it. But it is 
 just in this way that space is thought of, all its parts being 
 conceived to co-exist ad injinitii7?i. Hence the original 
 consciousness of space is an a priori perception, not a 
 conception. 
 
 3. Transcendental Exposition of Space. 
 
 A transcendental exposition seeks to show how, from 
 a certain principle, the possibility of other a priori 
 synthetic knowledge may be explained. To be suc- 
 cessful, it must prove (i) that there really are 
 synthetic propositions which can be derived from the 
 principle in question, (2) that they can be so 
 derived only if a certain explanation of that principle 
 is adopted. 
 
 Now, geometry is a science that determines the 
 properties of space synthetically, and yet a priori. What, 
 then, must be the nature of space, in order that such 
 knowledge of it may be possible? Our original con- 
 . J sciousness of it must be perception, for no new truth, 
 such as we have in the propositions of geometry, can be 
 obtained from the mere analysis of a given conception 
 (Introduction, 5). And this perception must be a priori^ 
 or, in other words, must be found in us before we actually 
 observe an object, and hence it must be pure, not 
 empirical perception. For all geometrical propositions, 
 as, for instance, that space has but three dimensions, are 
 of demonstrative certainty, or present themselves in 
 consciousness as necessary ; and such propositions 
 cannot be empirical, nor can they be derived from 
 judgments of experience (Introduction, 2). 
 
 How, then, can there be in the mind an external 
 
TRANSCENDENTAL ^ESTHETIC. 27 
 
 perception, which is antecedent to objects themselves, 
 and in which the conception of those objects may be. 
 determined a prior il Manifestly, only if that perception 
 has its seat in the subject, that is, if it belongs to the 
 formal constitution of the subject, in virtue of which it is 
 so affected by objects as to have a direct consciousness 
 or perception of them ; therefore, only if perception is the 
 universal ^r;;z of outer sense. 
 
 Our explanation is, therefore, the only one that makes 
 the possibility of geometry intelligible, as a mode of 
 a priori synthetic knowledge. All other explanations fail 
 to do so, and, although they may have an external 
 resemblance to ours, may readily be distinguished from it 
 by this criterion. 
 
 42 Inferefices. 26 
 
 ^ (a) Space is in no sense a property of things in them- 
 selves, nor is it a relation of things in themselves to one 
 , another. It is not a determination that still belongs to 
 objects even when abstraction has been made from all 
 the subjective conditions of perception. For we never 
 could perceive a priori any determination of things, 
 whether belonging to them individually or in relation to 
 one another, antecedently to our perception of those 
 things themselves. 
 
 {p) Space is nothing but the form of all the phenomena 
 of outer sense. It is the subjective condition without 
 which no external perception is possible for us. The 
 receptivity of the subject, or its capability of being 
 affected by objects, necessarily exists before there is any 
 perception of objects. Hence it is easy to understand, 
 how the form of all phenomena may exist in the mind 
 
28 THE PHILOSOPHY OF KANT. 
 
 a priori^ antecedently to actual observation, and how, as 
 a pure perception in which all objects must be determined, 
 it may contain the principles that determine beforehand 
 the relations of objects when they are met with in 
 experience. 
 
 It is, therefore, purely from our human point of view 
 that we can speak of space, of extended things, etc. 
 Suppose the subjective conditions to be taken away, 
 without which we cannot have any external perception, 
 43 or be affected by objects, and the idea of space ceases 
 to have any meaning. We cannot predicate spatial 
 dimensions of things, except in so far as they appear 27 
 in our consciousness. The unalterable form of this 
 receptivity, which we call sensibility, is a necessary 
 condition of all the relations in which objects are per- 
 ceived as outside of us, and this form, when it is viewed 
 in abstraction from objects, is the pure perception that 
 is known by the name of space. We are not entitled 
 to regard the conditions that are proper to our sensibility 
 as conditions of the possibility of things, but only of things 
 *" as they appear to us. Hence, while it is correct to say, 
 that space embraces all things that are capable of 
 appearing to us as external, we cannot say, that it 
 embraces all things as they are in themselves, no matter 
 what subject may perceive them, and, indeed, whether 
 they are perceived or not. For we have no means of 
 judging whether other thinking beings are in their 
 perceptions bound down by the same conditions as 
 ourselves, and which for us hold universally. If we state 
 the limitations under which a judgment holds of a given 
 subject, the judgment is then unconditionally true. The 
 proposition, that all things are side by side in space, is 
 
TRANSCENDENTAL ^ESTHETIC. 29 
 
 true only under the limitation that we are speaking of our 
 own sensible perception. But, if we more exactly define 
 the subject of the proposition by saying, that all things 
 as external phenomena are side by side in space, it will 
 be true universally and without any exception. Our 
 44 exposition, therefore, establishes the reality, or objective 
 truth of space, as a determination of every object that can 28 
 possibly come before us as external \ but, at the same 
 time, it proves the ideality of space, when space is con- 
 sidered by reason relatively to things in themselves, that 
 is, without regard to the constitution of our sensibility. 
 We, therefore, affirm the empirical reality of space, as 
 regards all possible external experience ; but we also 
 maintain its transce?idental ideality, or, in other words, we 
 hold that space is nothing at all, if its limitation to 
 possible experience is ignored, and it is treated as a 
 ^ necessary condition of things in themselves. 
 
 46 Section II. — Time. 3^ 
 
 4. Metaphysical Exposition of Time. 
 
 (i) Time is not an empirical conception, which has 
 been derived from any experience. For we should not 
 observe things to co-exist or to follow one another, did 
 we not possess the idea of time a priori. It is, therefore, 
 only under the presupposition of time, that we can be 
 conscious of certain things as existing at the same time 
 (simultaneously), or at different times (successively). 31 
 
 (2) Time is a necessary idea, which is presupposed in 
 all perceptions. We cannot be conscious of phenomena 
 if time is taken away, although we can quite readily 
 
30 THE PHILOSOPHY OF KANT. 
 
 suppose phenomena to be absent from time. Time is, 
 therefore, given a priori. No phenomenon can exist at all 
 that is not in time. While, therefore, phenomena may 
 be supposed to vanish completely out of time, time itself, 
 as the universal condition of their possibility, cannot be 
 supposed away. 
 
 (3) Time is not a discursive, or general conception, 
 but a pure form of sensible perception. Different times 
 are but parts of the very same time. Now, the conscious- 
 ness of that which is presented as one single object, is 32 
 perception. Moreover, the proposition, that no two 
 moments of time can co-exist, cannot be derived from a 
 general conception. The proposition is synthetic, and 
 cannot originate in mere conceptions. It therefore rests 
 upon the direct perception and idea of time. 
 
 (4) The infinity of time simply means, that every 
 48 definite quantity of time is possible only as a Hmitation of 
 
 one single time. There must, therefore; be originally 
 a consciousness of time as unlimited. Now, if an object 
 presents itself as a whole, so that its parts and every 
 quantity of it can be represented only by limiting that 
 whole, such an object cannot be given in conception, for 
 conceptions contain only partial determinations of a 
 thing. A direct perception must therefore be the founda- 
 tion of the idea of time. 
 
 5. Transce7idental Exposition of Time. 
 
 47 Apodictic principles which determine relations in time, 31 
 
 *■ or axioms of time in general, are possible only because 
 
 time is the necessary a priori condition of all phenomena. 
 
 Time has but one dimension; different times do not 
 
 co-exist but follow one another, just as different spaces do 
 
TRANSCENDENTAL ^ESTHETIC. 3 1 
 
 not follow one another but co-exist. Such propositions 
 cannot be derived from experience, which never yields 
 strict universality or demonstrative certainty. If they 
 were based upon experience, we could say only, that it 
 has ordinarily been observed to be so, not that it must be 
 so. Principles like these have the force of rules, that 
 lay down the conditions without which no experience 
 whatever is possible : they are not learned from experi- 
 ence, but anticipate what experience must be. 
 
 Let me add here that change, including motion or 
 change of place, is conceivable only in and through the 
 idea of time. Were time not an inner a priori percep- 
 tion, we could not form the least idea how there should 
 be any such thing as change. Take away time, and 
 change combines in itself absolutely contradictory predi- 
 cates. Motion, or change of place, for instance, must 
 then be thought of as at once the existence and the 
 non-existence of one and the same thing in the same 
 49 place. The contradiction disappears, only when it is 
 seen that the thing has those opposite determinations one 
 after the other. Our conception of time as an a priori 
 form of perception, therefore explains the possibility of 
 the whole body of a priori synthetic propositions in 
 regard to motion that are contained in the pure part of 
 physics, and hence it is not a little fruitful in results. 
 
 6. Inferences. 32 
 
 {a) Time is not an independent substance nor an 
 objective determination of things, and hence it does not 
 survive when abstraction has been made from all the 
 subjective conditions of perception. Were it an indepen- 
 dent thing, it would be real without being a real object of 
 
32 THE PHILOSOPHY OF KANT. 
 
 consciousness. Were it a determination or order of 33 
 things as they are in themselves, it could not precede our 
 perception of those things as its necessary condition, nor 
 could it be known by means of synthetic judgments. 
 But the possibility of such judgments becomes at 
 once intelligible if time is nothing but the subjective 
 condition, without which we can have no perception 
 whatever. For in that case we may be conscious of this 
 form of inner perception before we are conscious of 
 objects, and therefore a priori. 
 
 {b) Time is nothing but the form of inner sense, that is, 
 of the perception of ourselves and our own inner state. 
 As it has no influence on the shape or position of an 
 CO object, time cannot be a determination of outer pheno- 
 mena as such ; what it does determine is the relation 
 of ideas in our own inner state. And just because this 
 inner perception has no shape of its own, we seek to 
 make up for this want by analogies drawn from space. 
 Thus, we figure the series of time as a line that proceeds 
 to infinity, the parts of which form a series; and we 
 reason from the properties of this line to all the properties 
 of time, taking care to allow for the one point of differ- 
 ence, that the parts of the spatial line all exist at once, 
 while the parts of the temporal line all follow one after 
 the other. Even from this fact alone, that all the 
 relations of time may thus be presented in an external 
 perception, it would be evident that time is itself a 
 perception. 
 
 {c) Time is the formal a priori condition of all pheno- 
 mena without exception. Space, as the pure form of all 34 
 external phenomena, is the a priori condition only of 
 external phenomena. But all objects of perception. 
 
TRANSCENDENTAL ESTHETIC 33 
 
 external as well as internal, are determinations of the 
 mind, and, from that point of view, belong to our inner 
 state. And as this inner state comes under time, which 
 is the formal condition of inner perception, time is an a 
 priori condition of all phenomena : it is the immediate 
 condition of inner phenomena, and so the mediate con- 
 
 51 dition of outer phenomena. Just as I can say, a priori^ 
 that all external phenomena are in space, and are de- 
 termined a priori in conformity with the relations of 
 space, so, from the principle of the inner sense, I can say 
 quite generally that all phenomena are in time, and stand 
 necessarily in relations of time. 
 
 If we abstract from the manner in which we immedi- 
 ately perceive our own inner state, and mediately all 
 external phenomena, and think of objects in themselves, 
 we find that in relation to them time is nothing at all. 
 It is objectively true in relation to phenomena, because 
 we are conscious of phenomena as objects of our senses ; 
 but it is no longer objective, if we abstract from our ^5 
 sensibility, and therefore from the form proper to our 
 perceptive consciousness, and speak of tJmtgs as such. 
 Time is therefore a purely subjective condition of human 
 perception, and in itself, or apart from the subject, it is 
 nothing at all. Nevertheless, it is necessarily objective 
 in relation to all phenomena, and therefore also to every- 
 thing that can possibly enter into our experience. We 
 
 52 cannot say that all things are in time, because when we 
 speak of things in this unqualified way, we are thinking of 
 things in abstraction from the manner in which we per- 
 ceive them, and therefore in abstraction from the con- 
 dition under which alone we can say that they are in 
 time. But, if we qualify our assertion by adding that 
 
34 THE PHILOSOPHY OF KANT. 
 
 condition, and say that all things as phenomena, or 
 objects of sensible perception, are in time, the proposi- 
 tion is, in the strictest sense of the word, objective, and is 
 universally true a priori. 
 
 We see, then, that time is empirically real, or is 
 objectively true in relation to all objects that are capable 
 of being presented to our senses. And as our perception 
 always is sensuous, no object can ever be presented to 
 us in experience, which does not conform to time as its 
 condition. On the other hand, we deny to time all claim 
 to absolute reality, because such a claim, in paying no 3^ 
 heed to the form of sensible perception, assumes time to 
 be an absolute condition or property of things. Such 
 properties, as supposed to belong to things in themselves, 
 can never be presented to us in sense. From this we 
 infer the transcendental ideality of time ; by which we 
 mean that, in abstraction from the subjective conditions 
 of sensible perception, time is simply nothing, and cannot 
 be said either to subsist by itself, or to inhere in things 
 that do so subsist. 
 
 53 7. Explanatory Remarks. 
 
 To this doctrine, which admits the empirical reality of 
 time, but denies its absolute or transcendental reality, 
 there is one objection so commonly made, that I must 
 suppose it to occur spontaneously to everybody who is 
 new to the present line of thought. It runs thus : No 
 one can doubt that there are real changes, for, even if it 37 
 is denied that we perceive the external world, together 
 with the changes in it, we are at least conscious of a 
 change in our own ideas. Now, changes can take place 
 only in time. Therefore time is real. 
 
TRANSCENDENTAL ESTHETIC. 35 
 
 There is no difficulty in meeting this objection. I 
 admit all that is said. Certainly time is real : it is the 
 real form of inner perception. It has reality for me 
 relatively to my inner experience; in other words, I 
 actually am conscious of time, and of my own determina- 
 
 54 tions as in it. Time is therefore real, not as an object 
 beyond consciousness, but as the manner in which I 
 exist for myself as an object of consciousness. But, if I 
 could be perceived by myself or by any other being without 
 the condition of sensibility, the very same determinations, 
 which now appear as changes, would not be known as in 
 time, and therefore would not be known as changes. 
 The empirical reality of time thus remains, on our theory, 
 the condition of all our experience. It is only its 
 absolute reality that we refuse to admit. Time is there- 
 fore nothing but the form of our inner perception. If 
 we take away from it the peculiar condition of our 
 sensibility, the idea of time also vanishes ; for time does 
 not belong to objects as they are in themselves, but only 38 
 to the subject that perceives them. 
 
 55 Time and space are two sources of knowledge from 
 which a variety of a priori synthetic judgments may be 39 
 derived. Mathematics, especially, supplies a splendid 
 instance of such judgments, in the science of space and 
 
 56 the relations of space. Time and space are the two pure 
 forms of all sensible perception, and as such they make 
 a priori synthetic propositions possible. And just because 
 they are mere conditions of sensibility, they mark out 
 their own limits as sources of a priori knowledge. 
 Applying only to objects regarded as phenomena, they 
 do not present things as they are in themselves. Beyond 
 the phenomenal world, which is their legitimate domain. 
 
36 THE PHILOSOPHY OF KANT. 
 
 they cannot be employed in determination of objects. 
 But this Hmitation in no way lessens the stability of our 
 empirical knowledge ; for, such knowledge, as depending 
 upon necessary forms of the perception of things, is just 
 as certain as if it rested upon necessary forms of things in 
 themselves. 
 28 Transcendental Esthetic cannot contain more than 41 
 these two elements. This is plain, if we reflect that all 
 other conceptions belonging to sensibility presuppose 
 something empirical. Even the idea of motion, in which 
 both elements are united, presupposes the observation of 
 something that moves. Now, there is nothing movable 
 in space considered purely by itself; hence that which is 
 movable can be found in space only by experience, and 
 is therefore an empirical datum. Similarly the idea of 
 change cannot be put among the a priori data of 
 transcendental aesthetic. Time itself does not change, 
 but only something that is in time ; hence the idea of 
 change must be derived from the observation of some 
 actual object with its successive determinations — that is, 
 from experience. 
 
 eg 8. Ge7ieral remarks o?i the Tra7isce7ide7iial yEstheiic. 
 
 . (i) A distinction is commonly drawn between what *- 
 belongs essentially to an object, and is perceived by every 
 one to belong to it, and what is accidental, being per- 
 ceived only from a certain position, or when a special organ 
 is affected in a particular way. In the one case, we are 
 said to know the object as it is in itself; in the other 
 case, to know it only as it appears to us. This, however, 
 is merely an empirical distinction. For, it must be re- 
 
TRANSCENDENTAL iESTHETIC. 37 
 
 membered, that the empirical object which is here called 
 the thing, is itself but an appearance. If this were all, 
 our transcendental distinction would be altogether lost 
 sight of, and we might imagine ourselves to know things 
 in themselves when we knew only phenomena. For the 
 truth is, that, however far we may carry our investigations 
 into the world of sense, we never can come into contact 
 
 63 with aught but appearances. For instance, we call the 
 rainbow in a sun-shower a mere appearance, and the rain 
 the thing itself Nor is there any objection to this, if we 
 mean to state merely the physical truth, that from what- 
 ever position it is viewed the rain will appear to our 
 senses as a real object of experience. But, if we go 
 beyond the fact, that the sensible object is here the same 46 
 for every one, and ask whether the object is known as it 
 is in itself, we pass to the transcendental point of view, and 
 the question now is in regard to the relation of our 
 consciousness of the object to the object as it exists 
 apart from our consciousness. In this point of view, not 
 merely the rain-drops, but their round shape, and even 
 the space in which they fall, must be regarded as mere 
 appearances, not as things in themselves. Every aspect 
 of the phenomenon, in short, is but a modification or a 
 permanent form of our sensible perception, while the 
 transcendental object remains to us unknown. 
 
 71 (2) It is recognized in natural theology, not only that 
 God cannot be an object of perception to us, but that 
 He can never be an object of sensuous perception to 
 Himself At the same time. His knowledge must be 
 perception, and not thought, for thought always involves 
 limitations. Now, the natural theologian is very careful 
 to say, that God, in His perception, is free from the 
 
38 THE PHILOSOPHY OF KANT. 
 
 limits of space and time. But, how can this possibly be 
 maintained, if it has previously been assumed, that space 
 and time are forms of things in themselves? It must 
 then be held that, even if those things were annihilated, 
 space and time would continue to be a priori conditions 
 of their existence. And if they are conditions of all 
 existence, they must be conditions of the existence even 
 72 of God. We can avoid this conclusion only by saying 
 LH,.i,x,v^ that space and time are not objective^ forms of all things, 
 ^ but subjective forms of our outer as well as of our inner 
 
 perceptions. In fact our perception is sensuous, just 
 because it is not original. Were it original, the very 
 existence of the object would be given in the perception, 
 and such a perception, so far as we can see, can belong 
 only to the Original Being. Our perception is dependent 
 upon the existence of the object, and therefore it is 
 possible only if our perceptive consciousness is affected 
 by the presence of the object. 
 
 Nor is it necessary to say, that man is the only being 
 who perceives objects under the forms of space and 
 time ; it may be that all finite thinking beings agree with 
 man in that respect, although of this we cannot be 
 certain. But, however universal this mode of perception 
 may be, it cannot be other than sensuous, simply because 
 it is derivative {inticitiis derivativus) and not original 
 {intuitus origi?iarius), and therefore is not an intellectual 
 perception. An intellectual perception, as we have 
 already seen reason to believe, is the prerogative of the 
 Original Being, and never can belong to a being which is 
 dependent in its existence as well as in its perception, 
 and in fact is conscious of its own existence only in re- 
 lation to given objects. 
 
TRANSCENDENTAL ^ESTHETIC. 39 
 
 73 Conclusio7i of the Tra7iscende?ital Esthetic. 
 
 We have, then, in the Transcendental Esthetic, one 
 of the elements required in the solution of the general 
 problem of transcendental philosophy : How are a priori 
 synthetic propositions possible 1 Such propositions rest 
 upon space and time, which are pure a priori perceptions. 
 To enable us to go beyond a given conception, in an a 
 priori judgment, we have found that something is needed, 
 which is not contained in the conception, but in the 
 perception corresponding to it, something therefore that 
 may be connected with that conception synthetically. 
 But such judgments, as based upon perception, can 
 never extend beyond objects of sense, and therefore hold 
 true only for objects of possible experience. 
 
40 THE PHILOSOPHY OF KANT. 
 
 74 TRANSCENDENTAL LOGIC. 50 
 
 I. Gene7-al Logic. 
 
 There are two ultimate sources from which knowledge 
 comes to us : either we receive ideas in the form of 
 impressions, or, by our spontaneous faculty of conception, 
 we know an object by means of those ideas. In the 
 former case, the object is giveji to us ; in the latter case, 
 it is thought in relation to the impressions that arise in 
 our consciousness. Perception and conception, there- 
 fore, are the two elements that enter into all our 
 knowledge. To every conception some form of percep- 
 tion corresponds, and no perception yields knowledge 
 without conception. Both may be either pure or 
 empirical ; empirical^ if sensation, which occurs only in 
 the actual presence of an object, is implied; pure^ if 
 there is no intermixture of sensation. We may call 
 
 75 sensation the matter of sensuous knowledge. Hence 
 pure perception contains only the form under which 51 
 a something is perceived, and pure conception the 
 form in which an object in general is thought. Pure 
 perceptions or pure conceptions alone are possible a 
 priori^ while empirical perceptions or empirical concep- 
 tions are possible only a posteriori. 
 
 Insensibility is the receptivity of the mind in the actual 
 apprehension of some impression, understanding is the 
 spQjitandty of knowledge, or the faculty that of itself pro- 
 
TRANSCENDENTAL LOGIC. 4 1 
 
 duces idea s. We are so constituted that our perceptmi 
 always is sensuous ; or it shows merely the manner in 
 which we are affected by objects. But, we have also 
 imderstanding, or the faculty of thinking the object of 
 sensuous perception. Neither of these is to be regarded 
 as superior to the other. Without sensibility no object 
 would be given to us, without understanding none would 
 be thought. Thoughts without content are empty, per- 
 ceptions without conceptions are blind. It is therefore 
 just as necessary to make our conceptions sensuous, 
 that is, to add the object to them in perception, as it 
 is to make our perceptions intelligible, that is, to bring 
 them under conceptions. Neither of these faculties 
 or capacities can do the work of the other. Under- 
 standing can perceive nothing, the senses can think 
 
 76 nothing. Knowledge arises only from their united action. 
 But this is no reason for confusing the function of either 
 with that of the other ; it is rather a strong reason for 5 2 
 carefully separating and distinguishing the one from the 
 other. Hence it is, that we di stinguish Esthetic, as the 
 science of the universal rules of sensibility, from Logic, 
 which is the science of the universal rules of under- 
 standing. 
 
 77 General logic, as distinguished from the special logic 53 
 or organon of a particular science, is either pure or 
 
 78 applied; but only the former is in the strict sense a 
 science. There are two rules that must ever be kept in 54 
 mind in pure general logic, (i) As general logic, it 
 abstracts from all content of thought, and from all dis- 
 tinction of objects, and deals only with the pure form of 
 thought. (2) As pure logic, it has no empirical prin- 
 ciples. Psychology has no influence on the canon of the 
 
42 THE PHILOSOPHY OF KANT. 
 
 understanding, and therefore it does not, as has some- 
 times been supposed, contribute anything to pure logic. 
 Logic is a demonstrative science, and whatever it 
 contains must be certain entirely a priori. 
 
 79 
 
 2. Transce?idental Logic. ^^ 
 
 Pure general logic, then, abstracts from all the content 
 of knowledge, or what is the same thing, from all relation 
 of knowledge to its objects, and considers merely the 
 logical form implied in the relation of one element of 
 knowledge to another, or the universal form of thought. 
 Now, we have learned from the Transcendental Esthetic 
 that there are pure as well as empirical perceptions, and 
 it may well be, that a similar distinction obtains between 
 the pure and the empirical thought of objects. In that 
 
 ^° case, there will be a logic that does not abstract from all 
 the content of knowledge. Containing merely the rules 
 of the pure thought of an object, it will exclude all 
 knowledge, the content of which is empirical. It will 
 also refer our knowledge of objects to its origin, in 
 so far as that origin cannot be ascribed to objects 56 
 themselves. 
 
 8t Let us suppose, then, that there are conceptions which 57 
 relate to objects a priori, but which, as mere functions 
 of pure thought, stand to objects in quite a different 
 relation from that in which perceptions stand to them, 
 whether these are pure or sensuous. As these concep- 
 tions will be of neither empirical nor aesthetic origin, we 
 get the idea of a science of pure understanding and pure 
 reason, the aim of which is to examine into the know- 
 ledge which we obtain by thinking objects completely a 
 
TRANSCENDENTAL LOGIC. 43 
 
 priori. Such a science, as setting forth the origin, the 
 limits, and the objective validity of pure conceptions, we 
 must call Transcendefital Logic. 
 
 g2 3. Division of General Logic into Analytic and 
 
 Dialectic. 
 
 84 General logic analyzes the whole formal procedure of 60 
 understanding and reason into its elements, and presents 
 these as principles by which the logical validity of know- 
 ledge may be estimated. This part of logic, which is 
 well called Analytic, supplies a negative touchstone of 
 truth . . . but it does not enable us to determine 
 
 85 positively anything in regard to objects. At the same 
 time, there is something so seductive in an art that 
 enables us to reduce all our knowledge to the form of 
 understanding, however empty and poor in content it may 61 
 be, that general logic, although it is merely a canon of 
 judgment, is apt to be used as an organon by means of 
 which new truth, or rather the specious appearance of 
 new truth, may be obtained. When it is thus misused 
 
 as a supposed organon, logic is called Dialectic. 
 
 87 4. Division of Transce?tdental Logic into Analytic and 62 
 
 Dialectic. 
 
 Just as in Transcendental Esthetic we isolated the 
 sensibility, so in Transcendental Logic we shall isolate 
 the understanding, and throw into relief that element in 
 our knowledge which has its origin in the understanding 
 alone. This pure element can be employed in actual 
 knowledge, only on condition that objects are presented 
 
44 THE PHILOSOPHY OF KANT. 
 
 in perception to which it may be applied. For, without 
 perception, the pure element of knowledge has no object, 
 and therefore remains perfectly empty. Tha^partjof 
 Transcendental Logic which sets forth the pure element 
 in_Jcnmvledge that belongs to understanding, and the 
 principles without which no object whatever can be 
 thought, is Transcendental Analytic. It is a logic of 
 truth, because no knoAvledge can contradict it without 
 losing all content, that is, all relation to an object, and 6^ 
 therefore all truth. But there is a very seductive and 
 deceptive tendency to employ that pure knowledge of 
 understanding and those principles by themselves, and to 
 apply them even beyond the limits of experience. Only 
 in experience, however, can any matter or object be 
 88 found to which the pure conceptions of understand- 
 ing may be applied. There is thus a danger that 
 understanding, with a mere show of rationality, may make 
 a material use of its purely formal principles, and pass 
 judgments upon all objects without distinction, whether 
 they are given to us or not, and perhaps even although 
 they cannot be given to us at all. That which is merely 
 a canon for the criticism of understanding in its empirical 
 use, is misused; when it is supposed to be an organon 
 that may be employed universally and without restriction, 
 and when it permits understanding to venture upon 
 synthetic judgments about objects in general, and to pro- 
 nounce and decide upon them. Pure understanding is 
 then employed dialectically. The second part of Tran- 
 scendental Logic must therefore consist of a criticism of 
 dialectical illusion. It is called Dialectic, not because it 
 is an art of producing illusion dogmatically — a favourite 
 art of too many metaphysical jugglers — but because it is 
 
TRANSCENDENTAL LOGIC 45 
 
 a criticism of understanding and reason in their hyper- 
 physical use j a criticism, the aim of which is to expose 
 their specious and groundless pretensions to the dis- 64 
 covery and extension of knowledge through purely 
 transcendental principles, and to preserve understanding 
 from all sophistical illusion. 
 
 
46 THE PHILOSOPHY OF KANT. 
 
 89 TRANSCENDENTAL ANALYTIC. 
 
 BOOK I.— ANALYTIC OF CONCEPTIONS. 
 
 Chapter L — Guiding-thread for the Discovery 
 
 OF the Categories. 
 
 90 The first part of Transcendental Analytic deals with the 65 
 conceptions^ the second part with the judgments of pure 
 understanding. 
 
 92 It is the privilege as well as the duty of transcendental 67 
 philosophy, to proceed in the search for its conceptions 
 upon a definite principle ; for these conceptions spring 
 from the understanding pure and unmixed, and must 
 therefore be connected together in the unity of a single 
 conception or idea. This one fundamental conception is 
 
 a systematic principle, by the application of which we 
 may be certain a priori that we have found out all the 
 pure conceptions of understanding, and have assigned to 
 each its proper place in the whole system. , / "/ 
 
 Section I. — The Logical Use of Understanding. ., , % ^ ,^ / 
 
 Understanding has already been defined, negatively, 
 as a non-sensuous faculty of knowledge. Now, as 
 without sensibility we can have no perception, under- 68 
 
 93 standing cannot be a faculty of perception. But, apart 
 from perception, the only other mode of obtaining 
 knowledge is by means of conceptions. Therefore the 
 knowledge that is due to understanding, or at least to 
 
 ^.^v, ■ v^ ^tJ'-^^-^- ^ 
 
TRANSCENDENTAL ANALYTIC. 47 
 
 human understanding, is a knowledge by means of con- 
 ceptions ; it is not perceptive, but discursive. All per- 
 ceptions, as sensuous, rest upon affections, whereas 
 conceptions rest upon functions. By function I mean 
 the unity of act, in which various ideas are brought under 
 a common idea. Conceptions are based on the spon- 
 taneity of thought, sensuous perceptions on the receptivity 
 of impressions. Now the only use that understanding 
 can make of these conceptions is to judge by means of 
 them. And, as without perception there is no direct 
 consciousness of an object, a conception is never related 
 directly to an object, but always indirectly, through a 
 perception or through another conception. Judgment is 
 therefore the indirect knowledge of an object, or the 
 knowledge of knowledge. In every judgment there is a 
 conception which holds true of various ideas, and, 
 among others, of one which is directly referred to an 
 object. Thus, in the judgment that all bodies are 
 divisible, the conception of divisibility applies to various 
 other conceptions, but it is in an especial way related to 
 the conception of body, as this again is related to certain 
 94 objects that we directly perceive. Of these objects we 69 
 are therefore conscious only indirectly in the conception 
 of divisibility. Accordingly, all judgments are functions 
 of unity, because they do not consist in the direct know- 
 ledge of an object, but bring that and other knowledge 
 under the unity of a higher and more comprehensive con- 
 ception. And as we can reduce all acts of understanding 
 to judgments, understanding itself may be said to be a 
 faculty of judg?ne7it. For, as we have seen above, under- 
 standing is the faculty of thought. To think is to know 
 by means of conceptions. But conceptions, as predi- 
 
i8 
 
 THE PHILOSOPHY OF KANT. 
 
 cates of possible judgments, are relative to the idea of an 
 object not yet determined. By the conception of body 
 is meant something — metal, for instance — which may be 
 known by means of that conception. Body is a 
 conception^ just because it contains under it other deter- 
 minations by means of which it may be referred to actual 
 objects. It is thus the predicate of a possible judgment, 
 such as, that every metal is a body. We may, 
 therefore, find out all the possible functions of judgment 
 if we can but tell what are all the possible functions of 
 unity in judgment. And this, as we shall see in the next 
 section, can quite readily be done. 
 
 / 
 
 95 Section II. — 9. The Logical Function of U'?idcrsta?iding 7^ 
 
 i?i Judgment. 
 
 If we abstract from all the content of a judgment, and 
 only pay heed to the mere form of understanding, we 
 find that the functions of thought in judgment may be 
 brought under four heads, each of which contains three 
 subdivisions. Thus we get the following table : — 
 
 I. Quantity of Judgments. 
 Universal. 
 Particular. 
 Singular. 
 
 ■- ( i 
 
 ^JL 
 
 2. Quality. 
 
 3. Relation. 
 
 Affirmative. 
 
 Categorical. 
 
 Negative. 
 
 Hypothetical. 
 
 Infinite. 
 
 Disjunctive. 
 
 
 4. Modality. r • f / 
 
 •*-^ U^. 
 
 Problematic. "' /'' 
 
 -^< /-yu^^^^ 
 
 Assertoric. Jyi^^ W kj. ^^ 
 
 fy/ »l4f^ 
 
 Apodictic. /' y 
 
 VA^ 
 
 
^^^ ^ TRANSCENDENTAL ANALYTIC. (/^§ V 
 
 1 02 6'<?^//^« ///. — 10. The Pure Conceptions of Under sta?2di?ig 76 
 JtM^'trr-^^^ ^^1" / or Categories.^' ^ <^-^ a-.v^-f^^'^^-v^ ■ . 
 
 General Logic, as has been said, abstracts from all the 
 content of knowledge, and looks to some other source, 
 whatever that may be, for the content that it is to trans- 
 form by analysis into conceptions. Transcendental 
 Logic, on the other hand, has lying before it a complex 7 7 
 oi a priori sensibility, which it receives from Transcen- 
 dental ^Esthetic; without this complex, as a material 
 upon which to operate, the conceptions of pure under- 
 standing would be without content or perfectly empty. 
 Now, space and time have not only themselves^ as pure 
 a priori perceptions, a complexity of content ; but, as 
 they are the conditions without which the mind could 
 not be receptive of impressions, and therefore could not 
 be conscious of objects, they must always affect our con- 
 ception of objects. Conception, however, is due to 
 the spontaneous activity of thought, and hence the com- 
 plex content of pure perception must first be surveyed, 
 taken up into thought and combined, before there can be 
 any knowledge. This act I call synthesis. 
 
 103 By synthesis, in its most general sense, is meant the 
 act of putting various ideas together, and grasping their 
 multiplicity in one consciousness. Such synthesis ispure^ 
 if the multiplicity is given, not empirically but a priori^ as 
 in the case of space and time. Now, before we can 
 analyze any idea, we must first have the idea, and hence 
 the content of a conception cannot originally come into 
 consciousness by analysis. It is by synthesis of various 
 elements, whether those elements are given empirically or 
 a priori^ that we first get knowledge. No doubt the 
 
50 THE PHILOSOPHY OF KANT. 
 
 synthesis may at first be crude and confused, and it may 
 stand in need of analysis, but yet it is by synthesis 
 that the various elements are gathered together and 
 united in the knowledge of a certain concrete object. 
 It is to synthesis, therefore, that we must first direct ^g 
 our attention, if we would learn the true origin of our 
 knowledge. 
 
 Synthesis in general, as we shall afterwards see, is due 
 solely to the operation of imagination, a blind but indis- 
 pensable function of the soul, without which we should 
 have no knowledge whatever, but of which we are 
 seldom even conscious. To bring this synthesis to con- 
 ceptions is the function of understanding, and it is only 
 by this operation of understanding that we obtain what 
 can properly be called knowledge. 
 :o4 Piwe synthesis, viewed in its most general aspect, is t he 
 pure conception of understanding. By this pure syn- 
 thesis I understand that which rests upon a basis of a 
 priori synthetic unity. Thus in arithmetical addition, as 
 is readily seen in the case of larger numbers, the synthesis 
 conforms to a conception, because it proceeds on a 
 common basis of unity, as, for instance, the decade. By 
 this conception the unity in the synthesis of a complex is 
 made necessary. 
 
 By analysis various ideas are brought under a single 
 conception, as is shown in general logic. But it belongs 
 to transcendental logic to tell us how thQ pure synthesis of 
 ideas is brought to conceptions. The first element that 
 enters into the knowledge of all objects a priori is the 
 complex content of pure perception. The second 79 
 element is the synthesis of this content by imagination. 
 But as even this is not enough to constitute knowledge, 
 
TRANSCENDENTAL ANALYTIC. 5 1 
 
 a third element is supplied by understanding, in the 
 conceptions which give tcnity to this pure synthesis, and 
 which consist solely in the consciousness of this necessary 
 synthetic unity. 
 
 The same function which gives unity to various ideas 
 
 105 in 2l judgment also gives unity to the mere synthesis of 
 various ideas in 2, perception ; and this synthesis, in its 
 most general expression, is the pure conception of under- 
 standing. Unde rstanding^ at once gives analytic unity to 
 conceptions, and synthetic unity to the complex content 
 of perception j and indeed the logical form of judgment 
 presupposes and rests upon the very same acts of thought 
 as those by which a transcendental content is given to 
 the various determinations of our consciousness. Hence 
 it is that the pure conceptions of understanding, as they 
 are fitly called, apply to objects a priori^ and therefore 
 do not fall within the view of general logic. 
 
 In this way there arises exactly the same number of 
 pure conceptions of understanding, applying a priori to 
 all objects of perception, as there are logical functions of 
 judgments in the preceding table ; for those functions 
 completely specify understanding, and give a perfect 
 measure of its powers. We shall call the pure concep- 
 tions categories, after Aristotle, because our object is the ^o 
 same as his, although our method and results are widely 
 different. 
 
 106 Table of Categories. 
 
 I. Quantity, 
 Unity. 
 Plurality. 
 Totality. 
 
52 THE PHILOSOPHY OF KANT. 
 
 2. Duality. 
 
 3. Relatio7t. 
 
 Reality. 
 
 Inherence and Subsistence 
 {substantia et accidens). 
 
 Negation. 
 
 Causality and Dependence 
 (cause and effect). 
 
 Limitation. 
 
 Community (reciprocity 
 between the active and 
 the passive). 
 
 4. 
 
 Modality. 
 
 Possibility 
 
 Impossibility. 
 
 Existence 
 
 Non-existence. 
 
 Necessity 
 
 Contingency. 
 
 This, then, is a list of all the primary pure conceptions 
 of synthesis that understanding contains within itself a 
 priori. Because it contains these pure conceptions, it is 
 called pure understanding, and only by them can it 
 understand anything in the complex content of perception, 
 that is, think an object. The table has not been left to 
 the uncertain suggestions of empirical induction, but has 
 been drawn up systematically, on the basis of a single 
 principle, namely, the faculty of judgment, or, what is the 81 
 same thing, the faculty of thought. 
 
 109 II. 
 
 The table of categories suggests some nice points, which, 
 perhaps, might be found to have an important bearing 
 on the scientific form of all knowledge of reason, (i) 
 no The four classes of categories naturally fall into two 
 groups ; those in the first group being concerned with 
 objects of perception, pure as well as empirical, while 
 those in the second group are concerned with the exist- 
 ence of those objects, as related either to one another 
 
TRANSCENDENTAL ANALYTIC. 53 
 
 or to understanding. The first may be called the 
 mathematical^ the second the dynamical categories. The 
 former, as is obvious, have no correlates, the latter have 
 correlates. This distinction must have some ground 
 in the nature of understanding. (2) It is also suggestive 
 that the number of categories in each class is three, 
 because usually all a /r/(?r/ division must be by dichotomy. 
 To this it must be added that the third category in each 
 class arises from the union of the second category with 
 1 1 1 the first. Thus totality or allness is '^wsX plurality regarded 
 as unity^ limitation is reality combined with negation^ 
 community is causality in which two substances mutually 
 determine one another, and lastly, necessity is \\}i's>\. existence 
 given by mere possibility Itself. 
 
 jt6 Chapter II. — Deduction of the Categories. 84 
 
 13. Principles of a Transcendefital Deduction. 
 
 There is a distinction in law between the question of 
 right {quid Juris) and the question of fact {quid facti). 
 Both must be proved, but proof of a right or claim is 
 1 1 7 called its deduction. Now, among the variety of con- 85 
 ceptions that make up the very mixed web of human 
 knowledge, there are certain conceptions that put in a 
 claim for use entirely a priori, and this claim of course 
 stands in need of deduction. It is useless to refer to the 
 fact of experience in justification of such a claim, but at 
 the same time we must know how conceptions can 
 possibly refer to objects which yet they do not derive 
 from experience. An explanation of the manner in 86 
 which conceptions can relate a priori to objec-ts, I call a 
 transcendental deduction; and from it I distinguish an 
 
54 THE PHILOSOPHY OF KANT. 
 
 empirical deduction, which simply tells us how a con- 
 ception has been acquired by experience and reflection 
 on experience. The former proves our right to the 
 use of a certain conception, the latter merely points 
 out that as a matter of fact it has come into our 
 possession in a certain way. 
 
 We had no difficulty in explaining how space and 89 
 time, although they are themselves known a p7'iori, are 
 yet necessarily related to objects, and make possible a 
 synthetic knowledge of objects which is independent of 
 all experience. For, as it is only by means of these pure 
 forms of sense that we can be conscious of an object in 
 122 empirical perception, space and time are pure percep- 
 tions, which contain a priori the condition of the pos- 
 sibility of objects as phenomena, and therefore synthesis /y^^- 
 
 /\vv>C^>- 
 
 >£^' 
 
 in them has objective validity. 
 
 The categories of understanding, on the other hand, 
 are not conditions under which objects are given in ^^~\, ^ 
 perception ; hence objects might certainly be presented ^^T^ 
 to us, even if they were not necessarily related to ^ vA''*^^'^ 
 functions of understanding, as their a priori condition. 
 Here, therefore, a difficulty arises that we did not meet 
 with in the field of sensibility. The difficulty is, how sub- \ , 
 jective conditions of thought should have objective validity, 1 ' 
 or, in other words, how they should be conditions ^^dth- 
 out which no knowledge of objects would be possible. 
 Take, for instance, the conception of cause. Here we 90 
 have a peculiar sort of synthesis, in which something 
 B is conceived as following upon something else quite 
 different A, in conformity with a rule. It is hard to see 
 why phenomena should be subject to such an a priori 
 conception. Why should not the conception be perfectly 
 
 1 
 
TRANSCENDENTAL ANALYTIC. 55 
 
 empty, and without any phenomenal object corresponding 
 to it? 
 
 123 We cannot avoid the toil of such investigations by 91 
 saying that experience is perpetually giving us examples 
 of such conformity to law on the part of phenomena, and 
 that we are thus enabled to form an abstract conception 
 of cause, and to be certain of its objective validity. The 
 conception of cause cannot possibly originate in that way ; 
 and hence we must either show that it rests completely 
 a priori upon understanding, or we must discard it 
 
 124 altogether as a mere fiction of the brain. For the con- 
 ception demands that something A should be of such a 
 nature that something else B follows from it necessarily^ 
 and in conformity with an absolutely universal rule. No 
 pure conception of understanding can be the product of 
 empirical induction without a complete reversal of its 
 nature and use. 
 
 126 The transcendental deduction of all a priori concep- 94 
 tions must therefore be guided by the principle, that 
 these conceptions must be the a priori conditions of all 
 possible experience. Conceptions which make experience 
 possible are for that very reason necessary. An analysis 
 of the experience in which they occur would not furnish 
 a deduction of them, but merely an illustration of their 
 
 127 use. Were they not the primary conditions of all the 
 experience in which objects are known as phenomena, 
 their relation to even a single object would be utterly 
 incomprehensible. 
 
 / 
 
56 THE PHILOSOPHY OF KANT. 
 
 Section II. — A Priori Conditions of Experience* 
 
 It would be quite a sufficient deduction of the 9^ 
 categories, and justification of their objective applica- 
 tion, to show that, apart from them, no object whatever 97 
 is capable of being thought. But there are two reasons 
 why a fuller deduction is advisable : firstly, because, in 
 thinking an object, other faculties besides understanding, 
 or the faculty of thought proper, come into play ; and, 
 secondly, because it has to be explained how under- 
 standing can possibly be a condition of the knowledge of 
 real objects. We must, therefore, begin with a considera- 
 tion of the primary activities of the subject that are 
 essential in the constitution of experience ; and these we 
 must view, not in their empirical, but in their transcen- 
 dental character. 
 
 If consciousness were broken up into a number of 
 mutually repellent states, each isolated and separated 
 from the rest, knowledge would never arise in us at all, 
 for knowledge is a whole of related and connected 
 elements. When, therefore, I call sensible perception a 
 synopsis, in order to mark the complexity of its content, 
 it must be remembered that in this synopsis a certain 
 synthesis is implied, and that knowledge is possible only 
 if spontaneity is combined with receptivity. This is the 
 reason why we must say that in all knowledge there is a 
 three-fold synthesis : firstly, the apprehe?ision in perception 
 of various ideas, or modifications of the mind ; secondly, 
 their reproductio7i in imagination ; and, thirdly, their 
 recognition in conception. These three forms of synthesis 
 
 * AW that comes under this heading is taken from the first edition of 
 the " Critique of Pure Reason," and forms what is called in the preface 
 (p. X.) the " subjective deduction." 
 
TRANSCENDENTAL ANALYTIC. 57 
 
 point to three sources of knowledge, which make under- 
 standing itself possible, and through it all experience 98 
 as an empirical product of understanding. 
 
 I. Synthesis of Apprehension in Perception. 
 
 Whatever may be the origin of our ideas, whether they 
 are due to the influence of external things or are produced 
 by internal causes, whether as objects they have their 
 source a priori or in experience, as modifications of the 99 
 mind they must all belong to the inner sense. All 
 knowledge is, therefore, at bottom subject to time as the 
 formal condition of inner Sc^nse, and in time every part of 
 it without exception must be ordered, connected, and 
 brought into relation with every other part. This is a 
 general remark, which must be kept in mind in the whole 
 of our subsequent inquiry. 
 
 We should not be conscious of the various determina- 
 tions that every perception contains within itself were we 
 not, in the succession of our impressions, conscious of 
 time. If each feeling were limited to a single moment, 
 it would be an absolutely individual unit. In order that 
 the various determinations of a perception, as, for instance, 
 the parts of a line, should form a unity, it is necessary 
 that they should be run over and held together by the 
 mind. This act I call the synthesis of apprehension. It 
 is apprehension^ because it goes straight to perception ; it 
 is synthesis, because only by synthesis can the various 
 elements of perception be united in one object of con- 
 sciousness. 
 
 Now, this synthesis of apprehension must be employed 
 a priori also, or in relation to determinations not given in 
 sensible experience. Otherwise we should have no 
 
58 THE PHILOSOPHY OF KANT. 
 
 consciousness of space and time a priori^ for these can loo 
 be produced only by a synthesis of the various determi- 
 nations that are presented by sensibility in its original 
 receptivity. There is therefore a pure synthesis of 
 apprehension. 
 
 2. Synthesis of Reproductio7i in Imagination. 
 
 There is an empirical law of the association of ideas. 
 When any two ideas have often followed, or accompanied 
 each other, an association between them is at last formed, 
 and they are so connected that, even when an object is 
 not present, the mind passes from the one to the other in 
 conformity with a fixed rule. But this law of reproduc- 
 tion presupposes that phenomena are themselves actually 
 subject to such a rule, and that the various elements in 
 these phenomena of which we are conscious should accom- 
 pany or follow one another in accordance with certain 
 rules. On any other supposition our empirical imagina- 
 tion would have nothing to reproduce in any way 
 conforming to its own nature, and would therefore lie 
 hidden in the depths of the mind as a dead, and to us 
 unknown faculty. Were cinnabar, for instance, some- 
 times red and sometimes black, sometimes light and loi 
 sometimes heavy ; or were the same name given at one 
 time to this object, and at another time to that, without 
 the least regard to any rule implied in the nature of the 
 phenomena themselves, there could be no empirical 
 synthesis of reproduction. 
 
 There must, therefore, be something which makes the 
 reproduction of phenomena possible at all, something 
 which is the a priori ground of a necessary synthetic 
 unity. That this is so, we may at once see, if we reflect 
 
TRANSCENDENTAL ANALYTIC. 59 
 
 that phenomena are not things in themselves, but are 
 merely the play of our own ideas, and therefore at bottom 
 determinations of the inner sense. Now, if we can show 
 that even our purest a priori perceptions can yield 
 knowledge, only in so far as they involve such a com- 
 bination as makes a thoroughgoing synthesis of reproduc- 
 tion possible, we may conclude that this synthesis of 
 imagination, being prior to all experience, rests upon a 
 priori principles. We must then assume a pure tran- 
 scendental synthesis as the necessary condition of all 102 
 experience, for experience is impossible unless phenomena 
 are capable of being reproduced. Now, if I draw a line 
 in thought, or think of the time from one day to another, 
 or even think of a certain number, it is plain that I must 
 be conscious of the various determinations one after the 
 other. But if the earlier determinations — the prior parts 
 of the line, the antecedent moments of time, the units as 
 they arise one after the other — were to drop out of my 
 consciousness, and could not be reproduced when I 
 passed on to the later determinations, I should never be 
 conscious of a whole ; and hence not even the simplest 
 and most elementary idea of space or time could arise in 
 my consciousness. 
 
 The synthesis of reproduction is therefore inseparably 
 bound up with the synthesis of apprehension. And as 
 the synthesis of apprehension is the transcendental ground 
 of the possibility of all knowledge — of pure a priori as 
 well as empirical knowledge — the reproductive synthesis 
 of imagination belongs to the transcendental functions of 
 the mind, and may therefore be called the transcendental 
 faculty of imagination. 
 
6o THE PHILOSOPHY OF KANT. 
 
 3. Synthesis of Recognition i?i Conceptions. 103 
 
 Were I not conscious that what I think now is identical 
 with what I thought a moment ago, all reproduction in 
 the series of ideas would be useless. The idea reproduced 
 at a given moment would be for me a perfectly new idea. 
 There would be no identical consciousness bound up 
 with the act of producing one idea after another; and as 
 without such consciousness there could be for me no 
 unity, I should never be conscious of the various 
 members of the series as forming one whole. If, in 
 counting, I should forget that the units lying before my 
 mind had been added by me one after the other, I should 
 not be aware that a sum was being produced or generated 
 in the successive addition of unit to unit; and as the 
 conception of the sum is simply the consciousness of this 
 unity of synthesis, I should have no knowledge of the 
 number. 
 
 At this point it is necessary to have a clear idea of 104 
 what we mean by an object of consciousness. We have 
 seen that a phenomenon is just a sensation of which we 
 are conscious, and that no sensation can be said to exist 
 by itself as an object outside of consciousness. What, 
 then, do we mean when we speak of an object as corre- 
 sponding to our knowledge, and therefore as distinct from 
 it? It is easy to see that this object can be thought of 
 only as something = x^ for there is nothing beyond know- 
 ledge that we can set up as contrasted with knowledge, 
 and yet as corresponding to it. 
 
 It is plain that in knowledge we have to do with 105 
 nothing but the various determinations of our own 
 consciousness; hence the object-^, which corresponds 
 
TRANSCENDENTAL ANALYTIC. 6 1 
 
 to these determinations, if it is supposed to be distinct 
 from every object of consciousness, is for us nothing at 
 all. The unity which the object demands can be only 
 the formal unity of consciousness in the synthesis of its 
 various determinations. In saying that we know the 
 object, we mean that we have introduced synthetic unity 
 into the various determinations of perception. But this 
 is impossible, if the perception could not be produced by 
 a function of synthesis, which, in conforming to a rule, 
 makes the reproduction of those determinations a priori 
 necessary, and renders possible a conception that unites 
 them. 
 
 There can be no knowledge without a conception, io6 
 however indefinite or obscure it may be, and a conception 
 is in form always a universal that serves as a rule. The 
 conception of body, for instance, as a unity of the various 
 determinations thought in it, serves as a rule in our 
 knowledge of external phenomena. Now, it is always a 
 transcendental condition that lies at the foundation of 
 that which is necessary. There must, therefore, be a 
 transcendental ground of the unity of consciousness in 
 the synthesis of the various determinations implied in 
 every perception ; and this ground must be necessary to 
 the conception of any object whatever, and therefore to 
 the conception of every object of experience. In no 
 other way can there be any object for our perceptions; 
 for the object is nothing but that something = ^, the 
 conception of which involves necessity of synthesis. 
 
 This original and transcendental condition is just 
 transcendental apperception. The consciousness, in internal 107 
 perception, of oneself as determined to certain states, is 
 merely empirical, and is always changing. In the flux of 
 
62 THE PHILOSOPHY OF KANT. 
 
 inner phenomena there can be no unchanging or per- 
 manent self. This form of self-consciousness is usually 
 called in7ter sense or empirical apperceptio7i. Now, from 
 empirical data it is impossible to derive the conception 
 of that which must necessarily be numerically identical. 
 What we require, in explanation of such a transcendental 
 presupposition, is a condition that precedes all experience, 
 and makes it possible. 
 
 No knowledge whatever, no unity and connection of 
 objects, is possible for us, apart from that unity of 
 consciousness which is prior to all data of perception, 
 and without relation to which no consciousness of objects 
 is possible. This pure, original, unchangeable conscious- 
 ness I call transcendental apperceptioji. That this is the 
 proper name for it is evident, were it only that even the 
 purest objective unity, that of the a priori conceptions of 
 space and time, is possible only in so far as perceptions 
 are related to it The numerical unity of this appercep- 
 tion is, therefore, just as much the ^ /;V^r/ foundation of 
 all conceptions as the various determinations of space 
 and time are the a priori foundation of the perceptions of 
 sense. 
 
 It is this transcendental unity of apperception which io8 
 connects all the possible phenomena that can be gathered 
 together in one experience, and subjects them to laws. 
 There could be no such unity of consciousness were the 
 mind not able to be conscious of the identity of function, 
 by which it unites various phenomena in one knowledge. 
 The original and necessary consciousness of the identity 
 of oneself is at the same time the consciousness of 
 a necessary unity in the synthesis of all phenomena 
 according to conceptions. These conceptions are 
 
TRANSCENDENTAL ANALYTIC. 63 
 
 necessary rules, which not only make phenomena capable 
 of reproduction, but determine perception as perception 
 of an object, that is, bring it under a conception of some- 
 thing in which various determinations are necessarily 
 connected together. It would be impossible for the 
 mind to think itself as identical in its various determina- 
 tions, and indeed to think that identity a priori, if it did 
 not hold the identity of its own act before its eyes, and 
 if it did not, by subjecting to a transcendental unity all 
 the synthesis of empirical apprehension, make the con- 
 nection of the various determinations implied in that 
 synthesis possible in accordance with a priori rules. 
 
 129 15. Possibility of any Co7nbination whatever j^ 
 
 Though a perception is merely sensuous or receptive, 
 the various determinations of consciousness may be given, 
 while the form, as simply the way in which the subject is 
 affected, may lie a priori in the mind. But the combina- 
 tion {conjuncfio) of those determinations can never come 
 to us through the medium of sense, and therefore cannot 
 be contained even in the pure form of sensible perception. 
 
 130 Combination is a spontaneous act of consciousness, and, 
 as such, it is the especial characteristic of understanding, 
 as distinguished from sense. All combination, therefore, 
 whether we are aware of it or not, whether it is a combina- 
 tion of the various determinations of perception or of 
 several conceptions, and whether the determinations of 
 perception are empirical or pure, is an act of understand- 
 ing. This act we call by the general name of synthesis, 
 to draw attention to the fact that we can be conscious of 
 
 *What follows (15-27) constitutes the "objective deduction" of 
 the categories, as it appears in the second edition of the *' Critique." 
 
64 THE PHILOSOPHY OF KANT. 
 
 nothing as combined in the object, which we have not 
 ourselves previously combined. And as it proceeds 
 entirely from the self-activity of the subject, combination 
 is the element, and the only element, that cannot be 
 given by the object. It is easy to see that this act must 
 in its origin always be of one and the same nature, no 
 matter what may be the form of combination ; and that 
 the resolution or analysis^ which seems to be its opposite, 
 in point of fact always presupposes it. If understanding 
 has previously combined nothing, there is nothing for 
 it to resolve ; for without the combining activity of 
 understanding there can be no consciousness of an object 
 at all. 
 
 By combination, however, must be understood not 
 merely the synthesis of the various determinations of 
 sense, but also their unity. Combination is con- 
 131 sciousness of the syntheticunity of various determinations. 
 The consciousness of this unity cannot be the result of 
 the combination, for were we not, in being conscious of 
 various determinations, also conscious of their unity, we 
 should have no conception of combination at all. Nor 
 must this unity, which precedes any conception of 
 combination, be confused with the category of unity 
 (10); for all categories rest upon logical functions of 
 judgment, and, in these, combination, or the unity of 
 given conceptions, is already implied. For an explana- 
 tion of the unity in question, which is qualitative (12), 
 we must go further back, and seek it in that which, as the 
 gi-ound of the unity of various conceptions in judgment, 
 is implied in the possibility even of the logical use of 
 understanding. 
 
TRANSCENDENTAL ANALYTIC. 65 
 
 16. The origi?ial Synthetic iinity of Apperception. 
 
 The " / think " must be capable of accompanying all 
 132 my ideas; for, otherwise, I should be conscious of some- 
 thing that could not be thought ; which is the same as 
 saying, that I should not be conscious at all, or at least 
 should be conscious only of that which for me was 
 nothing. Now, that form of consciousness which is prior 
 to all thought, is perceptio7i. Hence, all the manifold 
 determinations of perception have a necessary relation to 
 the '■^ I think^^ in the subject that is conscious of them. 
 The "/ think" however, is an act of spontaneity, which 
 cannot possibly be due to sense. I call \\. pier e appercep- 
 tion^ to distinguish it from empirical apperceptio?i. I call 
 it also the original apperceptio?i, because it is the self- 
 consciousness which produces the "/ thinks Now, 
 the " / think " must be capable of accompanying all 
 other ideas, and it is one and the same in all con- 
 sciousness ; but there is no other idea beyond the 
 *' / think^^ to which self-consciousness is bound in a 
 similar way. The unity of apperception I call also the 
 transcendental tmity of self-consciousness, to indicate that 
 upon it depends the possibility of a priori knowledge. 
 For, the various determinations given in a certain per- 
 ception would not all be in my consciousness, if they did 
 not all belong to one self-consciousness. True, I may 
 not be aware of this, but yet as they are determinations 
 of my consciousness, they must necessarily conform to the 
 condition, without which they are not capable of standing 
 133 together in one universal self-consciousness. In no other 
 way would they all without exception be mine. From this 
 original combination important consequences follow. 
 
66 THE PHILOSOPHY OF KANT. 
 
 The absolute identity of apperception in relation to all 
 the determinations given in perception, involves a 
 synthesis of those determinations, and is possible only 
 through consciousness of the synthesis. For, the 
 empirical consciousness, which accompanies each deter- 
 mination as it arises, is in itself broken up into units, and 
 is unrelated to the one identical subject. Relation to 
 a single subject does not take place when I accompany 
 each determination with consciousness, but only when I 
 add one determination to the other, and am conscious 
 of this act of synthesis. It is only because I am capable 
 of combining in one consciousftess the various determina- 
 tions presented to me, that I can become aware that in 
 every one of them the consciousness is the same. The 
 analytic unity of apperception is, therefore, possible only 
 134 under presupposition of a certain synthetic unity. The 
 thought, that the determinations given in a perception all 
 belong to me, is the same as the thought, that I unite 
 them, or at least that I am capable of uniting them in one 
 self-consciousness. This does not of itself involve a 
 consciousness of the synthesis of determinations, but it pre- 
 supposes the possibility of that consciousness. It is 
 only because I am capable of grasping the various deter- 
 minations in one consciousness, that I can call them all 
 mine ; were it not so, I should have a self as many- 
 coloured and various as the separate determinations of 
 which I am conscious. Synthetic unity of the various 
 detenninations of perception as given a priori^ is 
 therefore the ground of that identity of apperception 
 itself, which precedes a priori every definite act of 
 thought. Now, objects cannot combine themselves, nor 
 can understanding learn that they are combined by 
 
TRANSCENDENTAL ANALYTIC. 67 
 
 135 observing their combination. All combination is the 
 work of understanding, and in fact understanding is itself 
 nothing but the faculty of combining a priori^ and 
 bringing under the unity of apperception, the various 
 determinations given in perception. The unity of 
 apperception is, therefore, the supreme principle of all 
 our knowledge. 
 
 This principle of the necessary unity of apperception, 
 is no doubt in itself an identical and therefore an analytic 
 proposition; but it also reveals the necessity for a 
 synthesis of the various determinations given in percep- 
 tion, because without such synthesis the thoroughgoing 
 identity of self-consciousness is inconceivable. In the 
 simple consciousness of self, no variety of determination 
 is given ; such variety of determination can be given 
 only in the perception which is distinguished from the 
 consciousness of self, and can be thought only by being 
 combined in one consciousness. An understanding in 
 which the consciousness of self should at the same time 
 be a consciousness of all the complex determinations of 
 objects, would be perceptive) but our understanding can 
 only think, and must go to sense for perception. I am 
 conscious of my self as identical in the various determina- 
 tions presented to me in a perception, because all 
 determinations that constitute one perception I call mine. 
 But this is the same as saying, that I am conscious of a 
 necessary synthesis of them a priori^ or that they rest 
 upon the original synthetic unity of apperception, under 
 
 136 which all the determinations given to me must stand, but 
 under which they can be brought only by means of a 
 synthesis. 
 
68 THE PHILOSOPHY OF KANT. 
 
 17. The syjithetic unity of Apperception is the sup7'eme 
 principle of U?idersta?iding. 
 
 In the Transcendental Esthetic, we have seen that the 
 supreme principle, without which perception in its 
 relation to sensibility is impossible, is, that all the 
 determinations of perception should stand under the 
 formal conditions of space and time. Now, the supreme 
 principle, without which perception, in its relation to 
 understanding is impossible, is, that all determinations of 
 perception should stand under conditions of the original 
 synthetic unity of apperception. Under the former stand 
 all determinations of perception, in so far as they are 
 given to us ; under the latter, in so far as they must be 
 137 capable of being co?nbi?ied in one consciousness. Apart 
 from the synthetic unity of apperception, nothing can be 
 thought or known, because the determinations given in 
 perception, not having the act of apperception, " / ////;^/('," 
 m common, would not be comprehended in one self- 
 consciousness. 
 
 Speaking quite generally, understanding is the faculty 
 of knowledge. Knowledge consists in the consciousness 
 of certain given determinations as related to an object. 
 An object, again, is that, in the conception of which the 
 various determinations of a given perception are united. 
 Now, all unification of determinations requires unity of 
 consciousness in the synthesis of the determinations. 
 Hence, the unity of consciousness is absolutely necessary, 
 to constitute the relation of determinations to an object, 
 give them objective validity, and make them objects of 
 knowledge ; and on that unity therefore rests the very 
 possibility of understanding. 
 
TRANSCENDENTAL ANALYTIC. 69 
 
 The principle of the original synthetic unity of 
 apperception, as being completely independent of all 
 conditions of sensuous perception, is the first pure cog- 
 nition of the understanding, upon which all its further 
 use depends. Space, as the mere form of external 
 sensuous perception, does not of itself yield any 
 knowledge : it but supplies the various elements of 
 a priori perception that are capable of becoming 
 138 knowledge. To know anything spatial, as, for instance a 
 line, I must draw it, and so produce by synthesis a 
 definite combination of the given elements. Thus, the 
 unity of the act of combination is at the same time 
 the unity of the consciousness in which the line is 
 thought, and only in this unity of consciousness is a 
 determinate space known as an object. The synthetic 
 unity of consciousness is, therefore, an objective con- 
 dition of all knowledge. It is not merely a condition 
 which I must observe in knowing an object, but it is a 
 condition under which every perception must stand, 
 before it can become an object for me at all. Without this 
 synthesis, the various determinations would not be united 
 in one consciousness. 
 
 Although it is thus proved, that the synthetic unity of 
 consciousness is the condition of all thought, the unity of 
 consciousness, as has been already said, is in itself an 
 analytic proposition. For, it says only, that all the 
 determinations of which /am conscious in a given per- 
 ception must stand under the condition, which enables 
 me to regard them as mine^ or as related to my identical 
 self, and so to comprehend them as synthetically com- 
 bined in one apperception, through the "/ tJwik^^ 
 expressed in all alike. 
 
70 THE PHILOSOPHY OF KANT. 
 
 But this is not the principle of every possible under- 
 standing, but only of an understanding, through the pure 
 apperception of which, in the consciousness "/«ot," no 
 
 139 determinations are given. If we had an understanding, 
 which, by its mere self-consciousness, presented to itself 
 the manifold determinations of perception ; an under- 
 standing, which, by its very consciousness of objects, 
 should give rise to the existence of these objects ; such an 
 understanding would not require, for the unity of con- 
 sciousness, a special act of synthesis of manifold deter- 
 minations. But this act of synthesis is essential to 
 human understanding, which thinks, but does not 
 perceive. It is, indeed, the supreme principle of human 
 understanding. Nor can we form the least conception 
 of any other possible understanding, whether of one that 
 itself perceives, or of one that is dependent upon 
 sensibility for its perception, but not upon a sensibility 
 that stands under the conditions of space and time. 
 
 18. Objective unity of Self-consciousness. 
 
 The transcendental unity of apperception is that unity 
 through which all the determinations given in a percep- 
 tion are united in a conception of the object. It is, 
 accordingly, called objective^ and must be distinguished 
 from the subjective unity of consciousness, which is a 
 deter mitiation of the inner sense, through which the 
 complex of perception is given empirically to be com- 
 bined into an object. Whether I shall be empirically 
 conscious of certain determinations as simultaneous, or 
 
 140 as successive, depends upon circumstances, or empirical 
 conditions. Hence, the empirical unity of consciousness, 
 
TRANSCENDENTAL ANALYTIC. 7 I 
 
 through the association of the elements of perception, is 
 itself a phenomenon, and is perfectly contingent. But 
 the pure form of perception in time, as merely perception 
 in general, stands under the original unity of conscious- 
 ness just because the various determinations given in it 
 are necessarily related to an "/ thinks It therefore 
 stands under that original unity by means of the pure 
 synthesis of understanding, which is the a priori ground 
 of the empirical synthesis. Only the original unity of 
 apperception is objective; the empirical unity, with 
 which we are not here concerned, and which besides is 
 only derived from the other, under given conditions in 
 conc7'efo, is merely subjective. To one man, for instance, 
 a certain word suggests one thing, to another a different 
 thing. In what is empirical, the unity of consciousness 
 does not hold necessarily and universally of that which 
 is given. 
 
 19. The Logical Form of all Judgments consists in the 
 objective unity of the Apperception of the Co?iceptio?is 
 they cont<iin. 
 
 141 A judgment is simply the way in which given ideas are 
 brought to the objective unity of apperception. This is 
 
 142 the force of the copula "/j-," which just marks the dis- 
 tinction between the objective unity and the subjective 
 unity of given ideas. It indicates their relation to the 
 original apperception, and their necessary unity. This 
 holds good even if the judgment is itself empirical and 
 therefore contingent. I do not mean, that, in the pro- 
 position, " Bodies are heavy," the idea of heavy is 
 necessarily connected with the idea of body in empirical 
 perception, but that they are connected with each other 
 in the synthesis of perceptions through the necessary 
 
72 THE PHILOSOPHY OF KANT. 
 
 u?iity of apperception. That is to say, the two ideas are 
 connected with each other in conformity with the prin- 
 ciples by which ideas are objectively determined and 
 become knowledge. Now, those principles are all 
 derived from the supreme principle of the transcendental 
 unity of apperception. Through this principle alone, 
 ideas are related in the way of judgment, and become 
 objectively valid. Thus we get a sufficient test of the 
 distinction between the relation of ideas in a judgment, 
 and a relation of the same ideas that is only of subjective 
 validity, as, for instance, a relation depending upon the 
 laws of association. In the latter case, all that I could 
 say would be, that if I lift a body, I have a sensation of 
 weight, but not, that the body is heavy. To say that the 
 body is heavy, means, that the two ideas of heavy and 
 body are connected together in the object, whatever the 
 state of the subject may be, and not merely that they are 
 contiguous in my observation, repeat it as often as I 
 please. 
 
 143 20. All sensuous Perceptions sta?id under the Categories as 
 conditions ufider which alo?ie their various deternmi- 
 ations ca?t co??ie together in one Consciousness. 
 
 The various determinations given in a sensuous per- 
 ception stand under the original synthetic unity of 
 apperception, because in no other way could there 
 possibly be any u?iity of perception (17). But that act 
 of understanding, by which the determinations given in 
 consciousness, w^hether these are perceptions or concep- 
 tions, are brought under a single apperception, is the 
 logical function of the judgment (19). Hence, all the 
 elements given in an empirical perception are determined 
 
TRANSCENDENTAL ANALYTIC. 73 
 
 by one of the logical functions of judgment, and thus 
 brought into one consciousness. But the categories are 
 just the functions of judgment, in so far as these are 
 applied in determination of the various elements of a 
 given perception (13). Therefore, the various determina- 
 tions in a given perception necessarily stand under the 
 categories. 
 
 146 22. The Category has no other application in Know- 
 
 ledge than to Objects of Experience. 
 
 To think an object is not the same thing as to kno7u 
 it. Knowledge involves two elements : firstly, the con- 
 ception or category, by which an object in general is 
 thought ; secondly, the perception by which it is given. 
 If no perception could be given, corresponding to the 
 conception, I should no doubt be able to think an object 
 so far as its form was concerned, but as there would be 
 no object in which that form was realized, I could not 
 possibly have knowledge of any actual thing. So far as 
 I could know, there would be nothing, and could be 
 nothing, to which my thought might be applied. Now, 
 the Esthetic has shown to us that all the perception that 
 we can have is sensuous ; hence the thought of an object 
 in general, by means of a pure conception of understand- 
 ing, can become knowledge, only by being brought into re- 
 
 147 lation with objects of sense. Sensuous perception is either 
 the pure perception of space and time, or the empirical 
 perception of that which is directly presented through 
 sensation as actually in space and time. By the deter- 
 mination of space and time themselves, we can obtain 
 that a priori knowledge of objects which mathematics 
 supplies. But this knowledge is only of the form of 
 
74 THE PHILOSOPHY OF KANT. 
 
 phenomena, and it is still doubtful if actual things must 
 be perceived in this form. Mathematical conceptions, 
 therefore, can be called knowledge, only if it is presup- 
 posed that there are actual things which cannot be 
 presented to us except under the form of that pure 
 sensuous perception. Now, things in space and time are 
 given to us only through empirical observation, that is, in 
 perceptions that are accompanied by sensation. Hence, 
 the pure conceptions of understanding, even if they are 
 applied to a priori perceptions, as in mathematics, do 
 not yield a knowledge of things. Before there can be 
 any knowledge, the pure perceptions, and the concep- 
 tions of understanding through the medium of pure 
 perceptions, must be applied to empirical perceptions. 
 The categories, therefore, give us no knowledge of actual 
 things, even with the aid of perception, except in so far 
 as they are capable of being applied to etnpirical percep- 
 tioji. In other words, they are merely conditions of the 
 possibility oi ernpirical knowledge. Now, such knowledge 
 is called experience. Hence the categories have a share 
 148 in the knowledge of those things only that are objects of 
 possible experience. 
 
 23. 
 
 The above proposition is of the greatest importance, 
 for it marks out the limits of the pure conceptions of 
 understanding in their application to objects, just as 
 Transcendental -Esthetic marked out the limits of the 
 pure forms of sensuous perception. Space and time are 
 but the conditions under which objects that are relative 
 to our senses are capable of being presented to us, and 
 therefore they apply only within the limits of experience. 
 
TRANSCENDENTAL ANALYTIC. 75 
 
 Beyond those limits they have no meaning whatever, for 
 they are only in the senses, and have no reality apart 
 from them. The pure conceptions of understanding are 
 free from this limitation, and extend to objects of percep- 
 tion of any kind, whether that perception is like or unlike 
 ours, if only it is sensuous, and not intellectual. But 
 this extension of conception beyond our sensuous per- 
 ception does not help us in the least. For, the concep- 
 tions are in that case quite empty, and we are therefore 
 unable even to say that there are any objects correspond- 
 ing to them. They are mere forms of thought without 
 objective reality, for we have no perception at hand, and 
 therefore no object, to which the synthetic unity of 
 apperception, which is the sole content of those forms of 
 149 thought, could be applied. Only our sensuous and 
 empirical perception can give to them meaning and 
 reality. 
 
 If I suppose an object of a non-sensiious perception to 
 be given, I can, no doubt, think of it as having all the 
 predicates implied in my presupposition. I can say that 
 the object has none of the determinations proper to 
 sensuous perception : that it is not extended or in space, 
 that its duration is not time, that there is in it no change 
 or succession of states in time, etc. But no real know- 
 ledge of an object is gained by merely indicating how it 
 is not perceived, so long as I cannot tell what is the con- 
 tent of its perception. I cannot in that way understand 
 even the possibility of an object to which my pure 
 conception could apply, for I am unable to bring forward 
 a perception corresponding to such an object, and can 
 say only that my perception can never bring me into 
 contact with it. But what most concerns us here, is, 
 
76 THE PHILOSOPHY OF KANT. 
 
 that to a thing of that nature, not even a single category 
 could be applied. I could not say, for instance, that 
 such a thing is a substance, that is, a thing that can exist 
 as subject, but never as mere predicate. For, how could 
 I apply the conception of substance, when, in the absence 
 of all empirical perception, I should not even know that 
 anything corresponding to my idea could exist at all ? 
 
 150 24. The application of the Categories to objects 
 
 of se7tse. 
 
 Understanding is capable of applying its pure concep- 
 tions to any object of perception, whether the perception 
 is the same as ours or not, if only it is sensuous. But 
 what this shows is that those conceptions are but mere 
 forms of thought^ which in themselves yield no knowledge 
 of a determinate object. As we have seen, the synthesis, 
 or combination of various elements implied in these forms 
 of thought, is relative merely to the unity of apperception, 
 and only m relation to that unity does it make possible 
 any a priori knowledge, or rather that knowledge which 
 rests upon understanding. It is, therefore, not only 
 transcendental, but also purely intellectual. But there 
 lies in us a certain form of a priori sensuous perception, 
 which is bound up with our sensibility, or the receptive 
 side of our consciousness. Hence understanding, by its 
 spontaneity, is capable of determining the inner sense, by 
 bringing the various elements given in pure perception 
 into conformity with the synthetic unity of apperception. 
 Thus it can think synthetic unity of the apperception of 
 the elements implied in a priori sensuous perception as 
 the condition under which all objects of human percep- 
 
TRANSCENDENTAL ANALYTIC. 77 
 
 tion must necessarily stand. In this way the categories, 
 which in themselves are mere forms of thought, obtain 
 
 151 objective reality, or application to objects that can be 
 given to us in perception. These objects, however, are 
 merely, phenomena, for only to phenomena do the a 
 priori forms of perception apply. 
 
 This synthesis of the units of sensuous perception, 
 which is possible and necessary a priori^ may be called 
 figural synthesis {synthesis speciosa), to distinguish it from 
 that intellectual synthesis {synthesis inteilectimlis), which 
 is thought in the mere category as applicable to all the 
 determinations of a perception. Both are transcende?ita\ 
 not merely because they precede a priori other know- 
 ledge, but because they make other a priori knowledge 
 possible. 
 
 But the figural synthesis, when it is considered merely 
 in relation to the original synthetic unity of apperception, 
 that is, to the transcendental unity which is thought in 
 the categories, must be called, in distinction from the 
 purely intellectual combination, the transcendental syn- 
 thesis of imagifiatiofi. Lnaginatio?! is the faculty of 
 setting before the mind in perception an object that is 
 7iot itself present. Now, all our perception is sensuous, 
 and hence imagination can give a perception correspond- 
 ing to the conceptions of understanding, only under the 
 subjective condition of time. Imagination therefore 
 pertains to sensibility. At the same time its synthesis is 
 
 152 the expression of spontaneous activity; for, unlike sense, 
 imagination is not simply capable of being determined, 
 but it is itself determining ; and hence it can a priori 
 determine sense in its form, in accordance with the unity 
 of apperception. Imagination, then, is in one point of 
 
78 THE PHILOSOPHY OF KANT. 
 
 view the faculty of determining the sensibiUty a priori ; 
 and its synthesis of the elements of pure perception, con- 
 forming as it does to the categories, must be called the 
 transcendental synthesis of imagination. This synthesis 
 is the result of an action of understanding on the sensi- 
 bility, or it is the first application, and so the condition 
 of all other applications, of understanding to objects that 
 we are capable of perceiving. The figural synthesis is 
 distinguished from the intellectual synthesis simply in 
 this, that the latter is due purely to understanding in 
 isolation from imagination. In so far as imagination is a 
 spontaneous activity, I sometimes call \\.productive\vi\2i<g\n- 
 ation, to distinguish it from reproductive imagination, the 
 synthesis of which is entirely dependent upon empirical 
 laws of association. As this latter synthesis in no way 
 helps to explain how a p?'iori knowledge is possible, it 
 belongs to psychology, not to transcendental philosophy. 
 
 ICQ 26. Transce?idejital Deduction of the Categories as 
 
 employed in Experience. 
 
 In the metaphysical deduction it has been proved that the 
 categories have their origin a priori, because they per- 
 fectly agree with the universal logical functions of 
 thought. In the t7'anscendental dediiction (20, 21), we 
 have seen how the categories make possible the a priori 
 knowledge of objects of perception in general. We 
 have now to explain how, by means of the categories, we 
 are capable of knowing ^/r/m objects of which we are 
 conscious only when our senses are actually affected. 
 What we propose to explain is not how there can be an 
 a priori knowledge of sensible objects as regards the form 
 of perception, but how there can be an a priori know- 
 
TRANSCENDENTAL ANALYTIC. 79 
 
 ledge of the laws by which the combination of objects 
 
 1 60 is effected, or, as we may say, what are the laws imposed 
 upon nature, without which there would be no nature 
 at all. 
 
 The first thing to be observed is that by synthesis of 
 apprehension is meant the putting together of various 
 determinations in an empirical perception, an act without 
 which there could be no observation or empirical con- 
 sciousness of a phenomenal object. 
 
 In space and time we have a priofi forms of outer as 
 well as inner perception ; and to these the synthesis of 
 apprehension must always conform, because in no other 
 way can apprehension take place at all. But space and 
 time are more than mere forms of sensuous perception : 
 they are themselves perceptions that contain a complex 
 of elements, and these elements are conceived a priori to 
 be determined to unity (see Transcendental Esthetic). 
 Along with these perceptions (not in them) there is pre- 
 supposed a priori^ as condition of all synthesis of 
 
 161 apprehension, a unity of synthesis of the various determin- 
 ations of inner and outer perception ; and this, again, 
 implies that whatever can be perceived as in space and 
 time must submit to combination. This synthetic unity 
 can only be the combination, in conformity with the 
 categories, of the various elements of any given perception 
 in an original consciousness, in so far as the combination 
 is applied to our sensuous perception. Hence, all synthesis, 
 including even that through which sensible observation is 
 possible, stands under the categories. And, as experience 
 is knowledge by means of connected observations, the 
 categories are conditions of the possibility of experience, 
 and therefore hold a priori of all objects of experience. 
 
8o THE PHILOSOPHY OF KANT. 
 
 162 I observe a house, for instance, by the apprehension of 
 various determinations given in empirical perception. 
 The necessary unity of space, and of external sensuous 
 perception in general, is presupposed, and I draw as it 
 were an outline of the house, in conformity with this 
 synthetic unity of its determinations in space. But, if I 
 abstract from the form of space, the very same synthetic 
 unity has its seat in understanding, and is the category of 
 qiiafitity, or the category of the synthesis of the homo- 
 geneous in any perception whatever. To this category, 
 therefore, the synthesis of apprehension — that is, the 
 observation — must completely conform. 
 
 163 Categories are conceptions which a priori prescribe 
 laws to phenomena, and therefore to nature as the sum 
 total of all phenomena {natura viaterialiter spectata). 
 Now, the categories are not derived from nature, nor do 
 they adapt themselves to nature as their model, for in 
 that case they would be merely empirical. How, then, 
 one asks, can it be shown that nature must adapt itself to 
 them ? How can the categories determine a priori the 
 combination of the complex phenomena of nature, instead 
 of going to nature to find out how phenomena are com- 
 bined ? Here is the solution of the problem. 
 
 164 It is no more wonderful that the laws of phenomena in 
 nature must agree with understanding and its a priori 
 form, or faculty of combining any complex given to it, 
 than that phenomena themselves must agree with the 
 form of ^/nV/ sensuous perception. Just as phenomena 
 have no existence at all, apart from a subject that has 
 senses, so there exist no laws in phenomena, apart from 
 a subject that has understanding. Things in themselves 
 \A'Ould of course have laws of their own, even if they did 
 
TRANSCENDENTAL ANALYTIC. 8 1 
 
 not come within the knowledge of the subject through 
 his understanding. But phenomena are merely the 
 manner in which things appear in consciousness, and give 
 no knowledge of what things may be in themselves. As 
 mere appearances they are subject to no law of con- 
 nection but that which is imposed by the connective 
 faculty. Now, it is imagination that connects the various 
 units of sensuous perception, and imagination is de- 
 pendent upon understanding for the unity of its intellec- 
 tual synthesis, and upon sensibility for the complexity of 
 apprehension. But nothing can come under observation 
 without a synthesis of apprehension, and this empirical 
 synthesis is dependent upon the transcendental synthesis, 
 and therefore upon the categories. Hence, all that can 
 165 be observed, or can come to empirical consciousness, 
 that is, all phenomena of nature, must depend for com- 
 bination upon the categories. In the categories, there- 
 fore, nature as a system of necessary laws {natura 
 formaliter spectaid) has its ground and origin. Pure 
 understanding, however, cannot by mere categories pre- 
 scribe <7 //'/<?/'/ any laws to phenomena other than those 
 universal laws of nature that apply to all objects in space 
 and time. Special laws, as relating only to what is 
 empirically determined, cannot be completely derived from 
 the categories, although they must all, without exception, 
 stand under the categories. To learn what are the special 
 laws of nature, we must go to experience ; but it is none 
 the less true that only the a priori laws imposed by 
 understanding tell us what is necessary for any experience 
 whatever, and what is capable of being known as an 
 object of experience. 
 
82 THE PHILOSOPHY OF KANT. 
 
 27. Result of the Deduction of the Categories. 
 
 We cannot think an object without categories; we 
 
 ^x/^ cannot know an object so thought without perceptions 
 
 \j^JY'^ that correspond to categories. Now, all our perceptions 
 
 are sensuous, and therefore all our knowledge of objects 
 
 that are presented in perception is empirical. But 
 
 t66 empirical knowledge is experience. Hence there can be 
 
 no a priori knowledge, except of objects that are capable 
 
 of entering into experience. 
 
 But although such knowledge is limited to objects of 
 experience, it is not therefore altogether derived from 
 experience. For pure perceptions as well as pure con- 
 ceptions are elements in knowledge, and both are found 
 in us a priori. There are only two ways in which we can 
 account for a necessary coincidence of the data of 
 experience with the conceptions which we form of its 
 objects : either that experience must make the concep- 
 tions possible, or the conceptions must make experience 
 
 167 possible. The former supposition is inconsistent with 
 the nature of the categories, not to speak of pure sensuous 
 perception ; for the categories, as a priori conceptions, 
 are independent of experience, and to derive them from 
 
 > ~ experience would be a sort of generatio aeqiiivoca. The 
 alternative supposition, which involves what may be 
 called an epigenesis of pure reason, must therefore be 
 adopted, and we must hold that the categories, as pro- 
 ceeding from understanding, contain the grounds of the 
 possibility of any experience whatever. 
 
 168 Short Statement of the Deduction. 
 
 What has been shown in the deduction of the cate- 
 gories is that the pure conceptions of understanding, on 
 
TRANSCENDENTAL ANALYTIC. 83 
 
 which all theoretical a priori knowledge is based, are 
 principles that make experience possible. In other 
 169 words, they are principles for the general determination of 
 phenomena in space and time, a determination that 
 ultimately flows from the principle of the origifial syn- 
 thetic unity of apperception as the form of understanding 
 in relation to space and time, the original forms of 
 sensibility. 
 
 \ 
 
 BOOK II.— THE ANALYTIC OF JUDGMENTS 
 
 171 Transcendental Judgment. 132 
 
 If understanding is called the faculty of rules, judgment 
 will be the faculty of suhsujnption under rules, that is, the 
 faculty of deciding whether something stands under a 
 given rule or not {casus datae legis). Now pure general 
 logic does not, and indeed cannot, lay down rules for the 
 application of judgment. For, as it abstracts from all the 
 cofitent of knowledge, its sole business is to analyze the 
 
 172 pure form of knowledge, as expressed in conceptions, ^ZZ 
 judgments, and inferences, and from this analysis to 
 derive formal rules for the general use of understanding. 
 
 174 The business of transcendental logic, on the other hand, 135 
 is to lay down definite rules which may enable judgment 
 
 to make a correct and certain use of the conceptions of 
 understanding. For transcendental philosophy has the 
 peculiarity that it not only brings to light the rules, or 
 rather the universal condition of rules, implied in the 
 
 175 pure conceptions of understanding, but it is able also to 
 indicate a priori the case to which each rule should be 
 
84 THE PHILOSOPHY OF KANT. 
 
 applied. The reason of its superiority in this respect 
 over all other theoretical sciences, except mathematics, 
 is that the conceptions with which it deals relate to 
 objects a priori. 
 
 The transcendental doctrine of judgment consists of 136 
 two chapters. The first treats of the sensuous condition 
 without which no pure conceptions of understanding can 
 be used. This is called the schematism of understanding. 
 The second deals with the synthetic judgments, which 
 arise a priori \N\].en the pure conceptions of understanding 
 are brought into use under that condition, and which 
 underlie all other a priori knowledge. It treats, in other 
 words, of the principles of pure understanding. 
 
 176 Chapter I. — The Schematism of the Categories. 137 
 
 In all subsumption, the object of which we are conscious 
 must be homogeneous with the conception under which it 
 is brought ; in other words, the conception must contain 
 some determination that is also present in the object 
 subsumed under it. This in fact is what we mean when 
 we say that an object is contained under a conception. 
 The empirical conception of a piaie, for instance, is 
 homogeneous with the pure geometrical conception of a 
 circle, because the roundness which is thought as a deter- 
 mination of the plate is presented as a perception in the 
 circle. 
 
 Now, a pure conception, or category, is quite hetero- 
 geneous from an empirical perception, or indeed from 
 any sensuous perception, and hence no pure concep- 
 tion can ever be found realized in a perception. No 
 
 177 one will say that the category of cause can be made 
 
TRANSCENDENTAL ANALYTIC. 85 
 
 visible to sense, or can be presented in a particular 
 perception as a property of it. How then can a percep- 
 tion be subsumed under a pure conception? How can 
 a category be applied in determination of an object of 
 sense? It is because this very natural and very important 138 
 question demands an answer that a transcendental doc- 
 trine of judgment is necessary. It must be shown how 
 pure conceptions of understanding can possibly be applied 
 to phenomena. In other sciences it is not necessary to 
 show that conceptions are applicable to objects, because 
 the general conception of the object is not in the same 
 way distinct and heterogeneous from the object as 
 presented in concrefo. 
 
 Manifestly there must be some third thing, which is 
 homogeneous on the one hand with the category, and on 
 the other hand with the object of sense, and which thus 
 makes the application of the one to the other possible. 
 This mediating idea must be pure, or free from any 
 empirical element, and yet it must be at once intellectual 
 and sensuous. Such an idea is the transcendental schema. 
 
 The category contains the pure synthetic unity of any 
 elements of which we can be conscious as different. 
 Time, as the formal condition of the various determina- 
 tions of inner sense, and therefore of the connections of 
 all our ideas, contains a priori m pure perception a variety 
 of differences. Now, a transcendental determination of 
 time is so far homogeneous with the category which 
 178 gives unity to it, that it is universal, and rests upon an a 
 priori rule. But, on its other side, that determination is 
 to a certain extent homogeneous with the object of sense, 
 since time is present in every object of which we can be 
 empirically conscious. By means of the transcendental 
 
86 THE PHILOSOPHY OF KANT. 
 
 determination of time or schema, the category may there- 
 fore be applied to phenomena, or, what is the same thing, 
 the phenomenon may be subsumed under the category. 
 
 179 In itself a schema is merely a product of imagination ; 140 
 but, as in producing it imagination does not seek to set 
 before itself an individual object of perception, but to 
 produce unity in the general determination of sensibility, 
 
 we must distinguish between the schema and the image. 
 If I set down five points one after the other, thus 
 
 I have before me an image of the number 
 
 five. But if I think simply of number — of any number at 
 all, be it five or a hundred — my thought is rather of the 
 method by which a certain sum, say a thousand, may be 
 presented in an image, in conformity with a certain con- 
 ception, than itself an image. It would, in fact, be very 
 hard to compare the image of so large a number as a 
 thousand with the conception of it. Now, the conscious- 
 
 180 riess of a universal process of imagination, by which an 
 image is provided for a conception, is what I call the 
 schema of a conception. 
 
 In point of fact, schemata, and not images, lie at the 
 foundation of our pure sensuous conceptions. No image 141 
 of a triangle can ever be adequate to the general 
 conception of triangle. The conception includes all 
 triangles — right-angled, obtuse-angled, etc. ; and, hence, 
 the image which I can set before myself can never reach 
 to the universality of the conception, but occupies only a 
 part of its sphere. The schema of the triangle can exist 
 nowhere but in thought : it is simply a rule for the 
 synthesis of imagination, in the determination of pure 
 figures in space. Much less can a single object of 
 experience, or an image of a single object, reach to the 
 
TRANSCENDENTAL ANALYTIC. 87 
 
 universality of an empirical conception. The direct 
 relation of an empirical conception is to the schema of 
 imagination, or the rule by which a perception is deter- 
 mined in conformity with that conception. The concep- 
 tion of a dog, for instance, is a rule for the guidance of 
 imagination in tracing out the figure of a certain four- 
 footed animal ; but it cannot be restricted to any single 
 determinate figure that experience can supply, nor can it 
 even be presented in concreto in any possible image that I 
 am capable of imagining. This schematism of our under- 
 standing, in its application to phenomena and to their 
 pure form, is an art hidden away in the depths of the 
 jgj human soul, the secret of which we need not hope to 
 drag forth to the light of day. This much may be said : 
 that the i??iage is a product of the empirical faculty of 
 productive imagination; while the schema of sensuous 
 conceptions, as, for instance, of figures in space, is a 142 
 product, and as it were a monogram, of pure a priori 
 imagination, which makes the consciousness of an image 
 possible at all. An image is necessarily connected with 
 a conception through the schema, and is in no case quite 
 congruent with the conception. But what distinguishes 
 the schema of a pure conception of understanding as 
 such, is that it cannot be presented in an image at all, 
 but is simply the pure synthesis, which conforms to a rule 
 of unity expressed in the category. Such a schema is a 
 transcendental product of imagination. It is a determina- 
 tion of the inner sense according to conditions of its form 
 of time in view of all ideas, a determination which is 
 necessary, if ideas are to be brought together a priori 
 in one conception, in conformity with the unity of 
 apperception. 
 
88 THE PHILOSOPHY OF KANT. 
 
 182 The pure image of all magnitudes {quanta) that are 
 presented in outer sense is space ; the pure image of all 
 objects of sense, inner as well as outer, is time. But 
 quantity {qua7ititas\ as a conception of understanding, 
 has as its schema nu7nbej'^ or the idea of the successive 
 addition of homogeneous unit to homogeneous unit. 
 Number is, therefore, the unity of synthesis implied in i43 
 putting together any homogeneous units of perception 
 whatever, a unity which results from the generation of 
 time itself in the apprehension of the perception. 
 
 The category of reality is the conception of that which 
 corresponds to any sensation whatever, and therefore of 
 that, the very idea of which is that it has being in time ; 
 the category of negation is the conception of that, the 
 very idea of which is that it has no being in time. The 
 opposition of reality and negation therefore rests upon 
 the distinction between a time as filled and the same 
 time as empty. And, as time is merely the form of per- 
 ception, that which in the phenomenon corresponds to 
 sensation is the transcendental matter, or reality, of all 
 objects as actual things. Now, every sensation has a 
 degree or magnitude by which it is capable of filling the 
 same time more or less, or, in other words, of occupying 
 the inner sense, with more or less comj^leteness, down to 
 the vanishing point ( = o = negatio). Hence, there is 
 
 183 a relation and connection, or rather a transition from 
 reality to negation, which makes us capable of setting 
 every reality before ourselves as a quantum. The schema 
 of reality, as the quantity of something so far as it fills a 
 time, is just this continuous and uniform generation of 
 reality in time, by the gradual descent from a sensation 
 that has a certain degree in time to its disappearance, or, 
 
TRANSCENDENTAL ANALYTIC. 89 
 
 what is the same thing, the gradual ascent from the 
 negation of sensation to its definite degree. 
 
 The schema of substance is the permanence of the i44 
 real in time, or the idea of the real as presupposed in the 
 empirical determination of time, and as persisting while 
 all else changes. Time does not itself pass away, but the 
 changeable in time passes away in its particular being. 
 What corresponds in the phenomenon to time, which is in 
 itself unchangeable and permanent, is the unchangeable 
 in existence, or substance; and only in reference to 
 substance can the succession and the co-existence of 
 phenomena be determined in time. 
 
 The schema of cause, and of the causality of a thing 
 in general, is the real which is supposed never to exist 
 without being followed by something else. It consists, 
 therefore, in the succession of various determinations, in 
 so far as that succession is subject to a rule. 
 
 The schema of community (reciprocal action), or of 
 the reciprocal causality of substances as regards their 
 184 accidents, is the co-existence in conformity with a uni- 
 versal rule of the determinations of one substance with 
 those of another. 
 
 The schema of possibility is the harmony of the 
 synthesis of different ideas with conditions of time in 
 general. Opposites, for instance, cannot exist in the 
 same thing at the same time, but only the one after the 
 other. The schema of possibility therefore determines 
 how a thing is capable of being known at any time. 
 
 The schema of actuality is existence in a determinate 145 
 time. 
 
 The schema of necessity is the existence of an object 
 at all times. 
 
9© THE PHILOSOPHY OF KANT. 
 
 From all this it is plain that the schema of every one 
 of the categories is in some way relative to time. The 
 schema of quantity is the generation or synthesis of time 
 itself in the successive apprehension of an object; the 
 schema of quality, the synthesis of sensation, as implied 
 in observation, with the consciousness of time, or, in 
 other words, it is the filling up of time ; the schema of 
 relation, the relation of different perceptions to one 
 another at all times, or in conformity with a rule for the 
 determination of time ; lastly, the schema of modality, 
 in its three forms, time itself as the correlative of the 
 determination whether and how an object belongs to 
 time. The schemata are, therefore, just ^/r/«??'/ determi- 
 nations of time in conformity with rules. Following the 
 order of the categories, we find that these rules, which 
 185 apply to all possible objects of experience, relate to the 
 series of ti7?ie, the conie?it of time^ the order of time, and the 
 coniprehe7isio7i of time. 
 
 We thus see that the schematism of understanding, 
 through the transcendental synthesis of imagination, is 
 neither more nor less than the way in which the various 
 determinations of perception are reduced to unity in the 
 inner sense, and so indirectly to the unity of apperception, 
 the function that corresponds to the receptivity of inner 
 sense. The schemata are, therefore, the true and only 146 
 conditions under which the categories obtain significance, 
 by being brought into relation with objects. In the end, 
 therefore, the categories have no other application than 
 to objects of a possible experience. They merely serve 
 to bind phenomena together under universal rules of 
 synthesis, by means of a necessary a priori unity that has 
 its source in the necessary combination of all conscious- 
 
TRANSCENDENTAL ANALYTIC. 9I 
 
 ness in the original unity of apperception. Thus it is 
 that the categories make phenomena fit for a thorough- 
 going connection in one experience. 
 
 Within this whole of possible experience all our know- 
 ledge lies, and in the universal relation to possible 
 experience consists that transcendental truth which 
 precedes empirical truth and makes it possible. 
 
 But no one can fail to see that, although only the 
 
 186 schemata of sensibility can realize the categories, they 
 none the less restrict them. For the schemata limit the 
 categories by conditions that lie outside of understanding 
 and in sensibility. The schema is in harmony with the 
 category, but it is properly merely the sensuous appear- 
 ance or sensuous conception of an object. Now, it is 
 naturally supposed that the sphere of a conception 
 previously restricted is enlarged when the restriction is 
 taken away. Hence it may be thought that the categories 147 
 in their purity, or apart from all conditions of sensibility, 
 hold true of things as they really are ; while the schemata 
 present them only as they appear. On this view the 
 categories will have a much wider meaning than the 
 schemata, and will be quite independent of them. And 
 this is so far true that, even apart from all sensuous con- 
 ditions the categories are not meaningless, for they still 
 have the logical meaning of the unity of our ideas of ob- 
 jects. But no conception has in itself objective meaning, 
 because, apart from the conditions of sensibility, there is 
 
 no object to which it can be applied. Substance, for 
 instance, viewed apart from the sensuous determination 
 of permanence, simply means, that which can be thought 
 only as subject, never as the predicate of anything else. 
 
 187 But such an idea has no meaning for us, because it tells 
 
92 THE PHILOSOPHY OF KANT. 
 
 US nothing whatever about the actual nature of the thing 
 that is thought to be an ultimate subject. Without 
 schemata, therefore, the categories are only functions of 
 understanding for conceptions, and give no knowledge 
 of objects. Meaning comes to them from sensibility, and 
 sensibility realizes understanding only by restricting it. 
 
 Chapter II. — Principles of Pure Understanding. 148 
 
 In the preceding chapter, we have considered the 
 transcendental faculty of judgment with reference only 
 to the universal conditions, under which it is justified in 
 employing the categories for the production of synthetic 
 judgments. We have now to set forth, in systematic 
 order, the judgments which understanding, under that 
 critical provision, actually produces a priori. The 
 table of categories will no doubt be a safe and natural 
 200 guide. Accordingly we find that all the principles of 161 
 pure understanding are — 
 
 I . Axioms of Perception. 
 2. Anticipations of 3. Analogies of 
 
 Observation. Experience. 
 
 4. Postulates of Empirical Thought. 
 
 I. Axioms of Perception. 
 
 202 -^ ^ 
 
 The principle of these is : All perceptions are extensive 
 magnitudes. 
 
 Pf'oof. 
 
 203 By an extensive magnitude, I mean a magnitude in 162 
 which the idea of the parts necessarily precedes and 
 makes possible the idea of the whole. I cannot have 
 
TRANSCENDENTAL ANALYTIC. 93 
 
 the idea of a line, however small it may be, with- 
 out drawing it in thought ; only by producing its parts 
 one after the other, beginning from a certain point, do 163 
 I mark out the line as a perception. Similarly with 
 every portion of time, even the smallest. I am con- 
 scious of time only in the successive advance from one 
 moment to another, and it is by the addition of all the 
 parts that a definite quantity of time is at last generated. 
 Now, either space or time is present in every 
 204 phenomenon as its pure element ; and as this pure 
 element can be known in apprehension only by a 
 successive synthesis of part with part, every perception 
 is an extensive magnitude. No phenomenon, therefore, 
 can be perceived at all without being perceived as an 
 aggregate or collection of previously given parts, a 
 characteristic which does not hold good of every sort of 
 magnitude, but only of those magnitudes, which, from 
 their very nature, are apprehended and presented in 
 consciousness as extensive. 
 
 On this successive synthesis of productive imagina- 
 tion in the generation of figures. Geometry, as the 
 mathematics of extension, is based. The axioms of 
 geometry express the conditions of sensuous perception 
 a priori^ without which no schema of any pure concep- 
 tion of an external object is possible ; as, for instance, 
 that between any two points only one straight line can be 
 drawn ; that two straight lines cannot enclose a space, 
 etc. Such axioms as these properly apply only to 
 magnitudes (^quanta) as such. 
 
 As to quantity {quantiias), that is, the answer to the 
 question how large a thing is, there are, strictly speaking, 
 no axioms, although several of the propositions referring 164 
 
94 THE PHILOSOPHY OF KANT. 
 
 to it are synthetic and immediately certain {indemon- 
 strabilia). The propositions, that if equals be added to 
 equals the wholes are equal, and that if equals be taken 
 from equals the remainders are equal, are analytic pro- 
 
 205 positions ; for I am directly conscious that the quantity 
 generated in the one case is identical with the quantity 
 generated in the other ; these propositions, therefore, 
 have no title to be called axioms, which must needs be 
 a priori synthetic propositions. There are, indeed, 
 simple numerical propositions which are synthetic ; but, 
 unlike the synthetic propositions of geometry, they are 
 not universal, and therefore even they cannot be called 
 axioms, but only numerical formulae. That 7 + 5 = 1 2 is 
 not an analytic proposition. For neither the idea of 7, 
 nor that of 5, nor the idea of the combination of the two, 
 yields the number 12. But, while it is synthetic, 
 the proposition 7 + 5 = 12 is merely individual. The 
 synthesis of the homogeneous can here take place only 
 in one way, although no doubt the numbers may afterwards 
 be employed universally. If I say that a triangle may be 
 constructed out of three lines, any two of which are 
 together greater than the third, I have before my mind 
 the mere function of productive imagination, which may 
 draw the lines greater or smaller, and bring them 165 
 together in all sorts of angles at will. On the other 
 hand, the number 7 is possible only in one way, and the 
 number 12 can be produced only by the synthesis of 5 
 
 206 with it. If mere numerical formulae like this are to be 
 called axioms, the number of axioms will be infinite. 
 
 This transcendental principle of the mathematics of 
 nature greatly enlarges our api'iori knowledge. It shows, 
 as nothing else can show, that mathematics in all its 
 
TRANSCENDENTAL ANALYTIC. 95 
 
 precision is applicable to objects of experience ; and this, 
 so far from being self-evident, has been the occasion of 
 much controversy. Phenomena are not things in them- 
 selves. Empirical perception is possible only through 
 the pure perception of space and of time ; and, therefore, 
 whatever geometry says of pure perception is beyond 
 dispute true also of empirical perception. All attempts 
 to evade this conclusion, by showing that objects of 
 sense need not conform to the rules of geometrical 
 construction — for instance, the rule of the infinite divisi- 
 bility of lines and angles — must be at once set aside. 
 Were such a contention true, the objective truth of 
 geometry, and therefore of all mathematics, would be 
 overthrown, and it would be impossible to say why and 
 how far mathematics should be applied to phenomena at 
 all. The synthesis of spaces and times, as the essential 
 forms of all perception, is that which makes the appre- 
 hension of a phenomenon even possible, and hence it is 166 
 the condition of all external experience, and so of all 
 knowledge of external objects. Whatever pure mathe- 
 matics proves to be true of space and time must necessarily 
 hold good of all external objects. All objections to the 
 207 truth of applied mathematics are but the chicanery of an 
 ill-advised reason, which wrongly seeks to free objects 
 of sense from the formal condition of our sensibility, 
 and to treat them as if they were things in themselves 
 apprehended by understanding. If phenomena were 
 really things in themselves, we could know nothing what- 
 ever of them a priori ; and as no synthetic judgments 
 can be based upon pure conceptions of space, geometry, 
 as the science of the properties of space, would itself be 
 impossible. 
 
g6 THE PHILOSOPHY OF KANT. 
 
 2. Anticipations of Observation. 
 
 The principle of these is : In all phenomena the real, 
 which is an object of sensation, has intensive magnitude 
 or degree. 
 
 Proof. 
 
 209 If it should turn out that in all sensations as such, how- 167 
 ever they may differ from one another, there is something 
 that can be known a priori ; this would, in a very special 
 sense, deserve to be called an anticipation. For by this 
 name we should call attention to the remarkable fact that 
 
 it is possible to say something a priori about the nature 
 of empirical objects, that is, about that very element in 
 them which is due to experience. 
 
 If no heed is paid to the succession of different 
 sensations, apprehension by means of mere sensation 
 is found to occupy only a moment. Here there is no 
 successive synthesis, advancing from the consciousness of 
 the parts to the consciousness of the whole, and therefore 
 that in the phenomenon which is called sensation has no 
 extensive magnitude. The absence of sensation from the 
 moment that it fills would therefore carry with it the 
 
 210 consciousness of that m.oment as empty = 0. Now that 168 
 which in empirical perception corresponds to sensation is 
 reality {realitas phaenome?ion) ; that which corresponds to 
 the absence of sensation is negation = o. But every 
 sensation is capable of diminution, so that it can decrease 
 and gradually disappear. Between reality in the pheno- 
 menon and negation, there is, therefore, a continuous 
 series of many possible intermediate sensations, the 
 difference between any two of which is always less than 
 the difference between the given sensation and zero or 
 
TRANSCENDENTAL ANALYTIC. 97 
 
 complete negation. That is to say, the real in the 
 phenomenon always has a quantity, but of this quantity 
 there is no consciousness in apprehension, because 
 apprehension, so far as it is due to the inner sensation, 
 takes place in one moment, and does not consist in 
 a successive synthesis of different sensations, and there- 
 fore does not advance from the parts to the whole. 
 Hence the real has magnitude, but not extensive 
 magnitude. 
 
 Now, a magnitude that is apprehended only as unity, 
 plurality being conceived in it as simply approximation 
 to negation = o, I call an intensive inagjtitude. Every 
 reality in a phenomenon has therefore intensive magni- 
 tude or degree. This reality may be regarded as a cause, 
 either of sensation or of some other reaHty in the pheno- 
 menon, for instance, a change. The degree of reality is 
 then called momentum^ as when we speak of the momen- 
 tum of gravity, to indicate a quantity, the apprehension of i6^ 
 which is not successive but instantaneous. I make this 
 remark merely in passing, for this is not the place to treat 
 of causality. 
 211 Every sensation, then, and consequently every reality 
 in a phenomenon, however small it may be, has an 
 intensive magnitude or degree that can always become 
 less, and between reality and negation there is a con- 
 tinuous series of possible realities, and of possible smaller 
 perceptions. The colour red, for instance, has a degree 
 which, however small it may be, is never the smallest 
 possible j and so with heat, the momentum of gravity, etc. 
 
 The property of magnitudes, by which no part in them 
 is the smallest possible, or no part is simple, is called their 
 continuity. Space and time are quanta continua^ because 
 
98 THE PHILOSOPHY OF KANT. 
 
 no part of them can be presented that is not enclosed 
 between limits (points or moments), and therefore each 
 part of space is itself a space, each part of time is itself a 
 time. Space consists only of spaces, time of times. 
 Points and moments are but limits, that is, mere places 
 of limitation in space and time, and as such always 
 presuppose the perceptions which they are to limit or 
 determine. Mere places are not constituent parts, which 
 can be given prior to space or time, and out of which 170 
 space and time can be made up. Such magnitudes may 
 also be called y?z^^;//, because the synthesis of productive 
 imagination, by which they are generated, is a progression 
 
 2 1 2 in time, the continuity of which is usually designated by 
 the \.^x\xiflux oxfloivmg. 
 
 All phenomena are continuous magnitudes, and that 
 in two ways : as pure perceptions, they are continuous 
 extensive magnitudes, and as perceptions of sense contain- 
 ing sensation, and therefore reality, they are continuous 
 intensive magnitudes. When the synthesis of determina- 
 tions is interrupted, we have an aggregate of various 
 objects of sense, not a single phenomenon as a quantum. 
 Such an aggregate is produced, not by continuing without 
 break the productive synthesis with which we begin, but 
 by continually renewing a synthesis that is continually 
 coming to an end. 
 
 As all phenomena, whether they are viewed extensively 171 
 
 213 or intensively, are continuous magnitudes, the continuity 
 of all change, or transition of a thing from one state into 
 another, might readily be proved here, and indeed proved 
 mathematically. But the causality of a change, as pre- 
 supposing empirical principles, does not come within the 
 province of transcendental philosophy. Understanding 
 
TRANSCENDENTAL ANALYTIC. 99 
 
 can give us no hint a priori that there can be a cause, 
 which is capable of changing the state of things, that is, 
 determining them to the opposite of a given state. It is 
 not simply that we cannot understand a priori how this 
 can take place — for there are many other instances of a 
 similar failure in a priori knowledge, — but that only certain 
 determinations of objects are capable of change at all, 
 and what these determinations are we can learn only 
 from experience, although no doubt the cause must lie in 
 that which is unchangeable. The only data that we have 
 here before us, are the pure conceptions implied in all 
 possible experience, which contain nothing empirical ; 
 nor can we avail ourselves of the primary facts of experi- 
 ence which lie at the foundation of pure physics without 172 
 destroying the unity of our system. 
 
 At the same time, there is no difficulty in showing that 
 the principle of understanding now under consideration 
 is of great value in enabling us to anticipate perceptions of 
 sense, and even to some extent to supply their place, by 
 guarding us against all false inferences that might be 
 drawn from their absence. 
 214 If all reality in perception has a degree, between which 
 and negation there is an infinite series of ever smaller 
 degrees, and if each sense must have a definite degree of 
 receptivity for sensations, it is evident that no perception, 
 and therefore no experience, can prove, directly or indi- 
 rectly, by any possible ingenuity of reasoning, that a 
 phenomenon is absolutely destitute of reality. That is to 
 say, there is no way of proving from experience that there 
 is empty space or empty time. For, in the first place, 
 the complete absence of reality from a perception of 
 sense can never be observed ; and, in the second place, 
 
lOO THE PHILOSOPHY OF KANT. 
 
 the absence of all reality can never be inferred from any 
 variation in the degree of reality of a phenomenon, nor 
 ought it ever to be brought forward in explanation of that 
 variation. For, although the whole perception of a 
 certain definite space or time is real through and through, 
 so that no part of it is empty ; yet, as every reality has a 
 degree, which may diminish by infinite degrees down to 
 nothing (the void), while the extensive magnitude of the 173 
 phenomenon remains unchanged, there must be an infinity 
 of degrees with w^hich space or time may be filled ; hence 
 the intensive magnitude may be greater or less in different 
 phenomena, although the extensive magnitude of the 
 perception remains the same. 
 
 217 The quality of sensation — colour, taste, etc., — is always 175 
 merely empirical, and cannot be known a priori. But 
 the real that corresponds to sensations in general, and is 
 opposed to negation = o, stands merely for that the very 
 conception of which implies being, and it has, therefore, 
 
 no other meaning than the synthesis in empirical con- 
 sciousness generally. In the inner sense, that empirical [^5 
 consciousness can be raised from o to any higher degree, 
 so that the extensive magnitude of a perception may be 
 greater or less, even when the intensive magnitude 
 remains the same. Thus, the degree of sensation excited 
 by an illuminated surface, may be as great as that pro- 
 duced by a number of less illuminated surfaces, the 
 aggregate extent of which is twice as large. In consider- 
 
 2 18 ing the intensive magnitude of a phenomenon, we may, 
 therefore, abstract entirely from its extensive magnitude, 
 and think only of the sensation, filling a single moment, 
 as generated by a synthesis that advances uniformly from 
 o to the given empirical consciousnes. Thus, while all 
 
TRANSCENDENTAL ANALl'TlC. lOl" ' 
 
 sensations as such are given a posteriori, it can be known 
 a priori that to all belongs the property of having a 
 degree. It is remarkable that of quantity in general only 
 a single quality — the quality of continuity, can be known 
 a priori, but that of quality, or the reality of phenomena, 
 nothing more than the intensive quantity, or the posses- 
 sion of degree, can be known a priori, while all else has 
 to be learned from experience. 
 
 3. Analogies of Experience. 
 
 The principle of these is : Experience is possible only, 
 through the consciousness of a necessary connection of 
 perceptions of sense. 
 
 Proof. 
 
 219 The three w^/// of time zxq permanefice, successioji, and 177 
 co-existence. All the relations of phenomena in time will 
 therefore be expressed in three rules, which precede all 
 experience and make it possible at all. These rules state 
 
 all the conditions under which phenomena can possibly 
 exist, in conformity with their unity in time. 
 
 220 The principle of all three analogies rests upon the neces- 
 sary unity of apperception in all empirical consciousness, 
 or perceptions of sense, «/ every moment of time. And 
 as the unity of apperception is the a priori condition of 
 all perception, that principle is based upon the synthetic 
 unity of all phenomena as regards their relation in time. 
 The original apperception is related to the inner sense, 
 which contains all possible objects of consciousness, or, 
 more exactly, it is related a priori to the form of inner 
 sense, as the manner in which the manifold determinations 
 of empirical consciousness are ordered in time. Now, in 
 
I02 THE PHILOSOPHY OF KANT. 
 
 original apperception all those determinations are to be 
 united, as regards their relations in time ; for nothing can 
 enter into my knowledge, or be mine, nothing can be for 
 me an object, that does not stand under the a priori 
 transcendental unity of apperception. This synthetic 
 unity in the temporal relation of all perceptions is, there- 
 fore, determined a priori, and is expressed in the law, 
 that all empirical determinations in time must stand 
 under universal rules of determination in time. The 178 
 Analogies of Experience must therefore be rules of this 
 kind. 
 
 These Analogies have the peculiarity, that they are not 
 concerned with the synthesis of empirical perception, 
 implied in the consciousness of objects of sense, but only 
 with the existence of such objects, and the relations to one 
 another by which their existence is determined. Now, a 
 221 phenomenon may be so determined a priori,\h?i\. the rule 
 of its synthesis yields at once the perception which is 
 presented to us in every empirical instance of it ; or, in 
 other words, the rule may not only tell us the character 
 of the synthesis, but may set the object before us as a 
 perception. But the existence of phenomena cannot thus 
 be known a priori. We may indeed in this way come to 
 know that something exists, but we cannot definitely 
 know what it is, nor can we anticipate how it will differ 
 from other objects, when it is empirically perceived. 
 
 The two principles already discussed, which I called 
 mathematical, to indicate that they justify the application 
 of mathematics to objects of sense, showed merely how 
 phenomena were possible, and how their perceptive 
 form, as well as the real of sense perception, could 
 be generated in conformity with rules of a mathe- 
 
TRANSCENDENTAL ANALYTIC. I03 
 
 matical synthesis. Both principles, therefore, entitle us 
 to estimate phenomena numerically and quantitatively. 
 The degree of sensation of sunlight, for instance, may be 179 
 determined a priori^ or constructed, by putting together, 
 say, 200,000 illuminations of the moon. Those princi- 
 ples may therefore be called constitutive. 
 
 It is quite different with the principles that show how 
 the existence of phenomena comes under a priori rules. 
 222 Existence cannot be constructed ; all that can be done is 
 to state the rules that determine the relations of existence, 
 and these rules yield only regulative principles. Here, 
 therefore, there can be neither axioms nor anticipations. 
 If in observation something is presented as related in 
 time to something else, as yet unknown, it is impossible 
 to tell what that something else may be, or what may be 
 its magnitude ; all that we can tell is how the two per- 
 ceptions, to exist at all, must be connected with each 
 
 other An analogy of experience is, there- 180 
 
 fore, merely a rule which states the conditions under 
 which observations of sense may be reduced to the unity 
 of experience. Incompetent to tell us the conditions of 
 observation, so far as its empirical element is concerned, 
 it is not a principle constitutive of objects of sense or 
 phenomena, but is merely regulative. In like manner, 
 the postulates of empirical thought are regulative prin- 
 ciples. The certitude is as great for the regulative as 
 for the mathematical or constitutive principles, for both 
 are a priori^ but the kind of evidence is different. 
 
 762 In regard to the general method of proof in philosophy, 7^4 
 it must be observed that a demonstratio?i is an apodictic 
 proof which rests upon direct perception. But, in the 
 
104 THE PHILOSOPHY OF KANT. 
 
 case of discursive knowledge, even those judgments 
 which are based upon ^/n'm conceptions, and are there- 
 fore apodictic, cannot be proved by a direct appeal to 
 perception. It is only mathematics that admits of 
 demonstrative evidence, for mathematics alone derives 
 its knowledge, not from conceptions, but from the con- 
 struction of conceptions — that is, from the perception 
 which corresponds to certain conceptions and can be 
 presented a priori. Even the solution of an algebraic 
 equation is a process of construction, though not of 
 geometrical construction ; for, it consists in presenting 
 conceptions in perception by means of symbols, and 
 especially conceptions of the relation of quantities. 
 Although, therefore, in its method algebra is not heur- 
 istic, it is able to guard against error in its results by 
 placing all the conceptions that it employs directly before 
 the eyes. But, while mathematics views the universal in 
 coficreto — that is, in pure perception, where every error be- 
 comes immediately visible — philosophical knowledge has 
 to dispense with this advantage, and to consider the univer- 
 
 763 sal in abstractor or through the medium of conceptions. 
 
 It is therefore contrary to the true spirit of philosophy, 735 
 and especially of pure philosophy, to boast of its dogmatic 
 procedure, and to bedeck itself with the orders and the 
 titles of mathematics. Such empty boasts can only 
 retard the progress of philosophy, and prevent it from 
 detecting the illusion into which reason falls when it is 
 unaware of its true limits. 
 
 764 Apodictic propositions may be distinguished as either 736 
 dogmata or niathcmata. By a dogma is meant a synthetic 
 proposition which is directly derived from conceptions ; 
 
 by a mathetna^ one that is obtained by the construction 
 
TRANSCENDENTAL ANALYTIC I05 
 
 of a conception. Of these two classes of a priori 
 synthetic propositions, popular language permits us to 
 apply the term dogma only to philosophical knowledge, 
 for we should hardly call a proposition in arithmetic or 
 geometry a dogma. The ordinary use of words thus 
 ^ confirms the distinction we have drawn between judg- 
 ments that are derived from conceptions, and judgments 
 that rest upon the construction of conceptions. 
 
 Now, it is impossible to find in the whole domain of 
 pure speculative reason a single synthetic judgment that 
 is directly derived from conceptions. For, reason is 
 765 unable to obtain from its pure ideas any synthetic judg- 
 ment which holds true objectively. It is true that, by 
 means of the conceptions of understanding, reason is able 737 
 to show that there are certain principles which rest upon 
 a solid foundation ; but these principles it does not 
 directly derive from conceptions, but only indirectly, by 
 showing the relation of the conceptions in question to 
 something that is perfectly contingent — namely, possible 
 experience. If something is presupposed as an object of 
 possible experience, no doubt those principles are 
 apodictically certain ; but, in themselves, or directly, they 
 can never be known a priori. Thus, no one, simply 
 from the conceptions contained in it, can see what is the 
 foundation of the proposition, that whatever happens has 
 its cause. Such a proposition can certainly be shown 
 readily enough to be apodictic, if it is applied only 
 within the field of experience ; but it cannot be a dogn)a. 
 It must be called d. principle^ and not a theorem^ because 
 it has the peculiar property, that it is the condition of 
 that by which it is proved — namely, experience, and 
 must always be presupposed as essential to experience 
 
I06 THE PHILOSOPHY OF KANT. 
 
 Now, if in the speculative use of pure reason there are 
 no dogmata, all dogmatic methods, whether they are 
 borrowed from the mathematician or are peculiar to the 
 individual thinker, are self-condemned. For tliey only 
 serve to conceal defects and errors, and to give rise to 
 philosophical illusion, instead of securing the true aim of 
 philosophy, which is to exhibit every step of reason in 
 the clearest possible light. Yet the method of philosophy, 
 though it is not dogmatic, may always be systematic. For 
 766 our reason is itself subjectively a system, though, if we 738 
 regard it merely as a source of pure conceptions, it is not 
 a system of knowledge, but only a system by which our 
 investigations may be carried on : in other words, it 
 supplies the principles of unity for knowledge, and must 
 look to experie?ice to supply the materials to be determined 
 in accordance with these principles. 
 
 224 -<4« First Analogy. 182 
 
 Principle of the Permanence of Substance. 
 
 In all the changes of phenomena substance is perma- 
 nent, and its quantum in nature neither increases nor 
 diminishes. 
 
 Proof. 
 
 225 Our apprehension of the various determinations of a 
 phenomenon is always successive, and therefore is always 
 changing. Hence there is nothing in apprehension, 
 taken by itself, that enables us to say whether those 
 determinations are, as an object of experience, co-existent 
 or successive. An object of experience is possible only 
 
 226 if there is something that ahuays is, somtthmg per7?ia?i- 
 ent and persistent, all change and co-existence being 
 
TRANSCENDENTAL ANALYTIC. I07 
 
 nothing but so many modes of time in which that 
 permanent something exists. Only in the permanent can 
 there possibly be the relations of simultaneity and succes- 
 sion, which are the sole relations in time. The permanent 183 
 is therefore the sicbstratiim of the empirical consciousness 
 of time itself, and only in it is any determination of time 
 possible at all. Permanence is time considered quite 
 generally, as the constant correlate of all change and all 
 concomitance of actual objects of experience. For, 
 change does not affect time itself but only phenomena in 
 time; just as co-existence is plainly not a mode of time 
 itself, the parts of time not being together, but following 
 one another. If it is said, that time itself comes into 
 being part by part, we must suppose that there is another 
 time in which it successively comes to be. Only through 
 the permanent does existence in a number of successive 
 moments acquire a magnitude, which we call duration. 
 In mere succession, taken by itself, existence is always 
 vanishing and appearing, and never has even the smallest 
 magnitude. Apart from the permanent, there is therefore 
 no relation of time. Now, time cannot be perceived by 
 itself; hence the permanent is the substratum of all the 
 determinations of phenomena in time, and therefore the 
 227 condition without which there could be no synthetic unity 
 in our perceptions, that is, in experience. Thus we learn 
 that all existence and all change in time must be regarded 
 as simply a modus of the existence of that which does 
 not change but persists. In all phenomena the per- 
 manent is therefore the object itself, that is, the substance 
 (phaenofHc7ioH), while all that changes, or can change, per- 184 
 tains merely to the manner in which substance or substances 
 exist, and therefore to the determinations of substance. 
 
Io8 . THE PHILOSOPHY OF KANT. 
 
 229 The determinations of a substance are called accidents. 186 
 They are always real, because they are just the manner 
 
 in which the substance exists, whereas negations are 
 merely determinations which affirm that a substance does 
 
 230 not exist in a certain manner. If we wish to say, that 
 what is real in a substance has a special sort of existence, 
 as, for instance, that motion is the manner in which 
 matter exists, we are wont to speak of this mode of ex- 
 istence as inherence^ to distinguish it from the existence 
 of the substance, which is called subsistence. But this is 
 apt to lead to much misapprehension, and it is more precise 187 
 and more correct to say, that an accident is simply the 
 manner in which the existence of a substance is positively 
 determined. At the same time, the conditions under 
 which understanding in its logical use operates, gives a 
 kind of independence to that in the existence of a sub- 
 stance which can change while the substance remains 
 unchanged, and this changing element we are led to view 
 
 as standing in relation to the really permanent and 
 radical element. It is for this reason that the category 
 of substance is put among the categories of relation ; for, 
 although strictly speaking it does not itself contain a 
 relation, it yet is the condition of relations. 
 
 The conception of change can be properly understood 
 only by reference to the idea of permanence. Coming 
 to be and ceasing to be are not changes of that which 
 comes to be or ceases to be. Change is a mode of 
 existence that follows upon another mode of existence of 
 the very same object. All that changes is pernianejit^ 
 and only its state alters. As this alteration affects only 
 the determinations, which can cease to be or begin to be, 
 
 231 we may say, in words that sound somewhat paradoxical, 
 
TRANSCENDENTAL ANALYTIC. IO9 
 
 that only the permanent changes, while the changeable 
 is subject to no change, but only to an alternation, in 
 which certain determinations cease to be as others begin 
 to be. 
 
 Change, then, can be observed only in substances. i88 
 An absolute beginning or cessation can by no possibility 
 be observed, but only a determination of that which is 
 permanent j because only by reference to that which is 
 permanent can there be any consciousness of the transi- 
 tion from one state into another, and from not being to 
 being. And these states can be known empirically only 
 as alternating determinations of that which is permanent. 
 If we suppose something absolutely to begin to be, we 
 must also suppose that there was a point of time in which 
 that something was not. But with what are we to con- 
 nect this point of time, if not with something that already 
 is ? For, an empty time, if we suppose such to precede 
 the point of time in question, is not a possible object of 
 perception ; and if we connect what is supposed 
 absolutely to begin to be with things that existed before 
 it, and continue to exist up to the moment of its origina- 
 tion, that which is supposed absolutely to begin to be 
 must be really a determination of the permanent that 
 existed before it. So, also, that which absolutely ceases 
 to be requires us to presuppose the empirical conscious- 
 ness of a time in which there was nothing to observe. 
 
 Substances, then, are the substrates of all determina- 
 tions of phenomena in time. If some substances could 
 come into being, and others cease to be, even the sole 
 condition, under which the empirical unity of time is 
 possible, would be taken away. We should in that case 
 232 be compelled to suppose, that phenomena were in two 
 
no THE PHILOSOPHY OF KANT. 
 
 distinct times, and that existence flovved away in two 
 parallel streams. But this is absurd, for there is only one 
 time, and different times are not side by side, but follow 189 
 one another. 
 
 Permanence is therefore a necessary condition, without 
 which phenomena cannot be determined as things or 
 225 objects in a possible experience. The permanent is the 
 substance, or the real, in a phenomenon, which, as the 
 substratum of all change, always remains the same. And 
 as substance can be subject to no change in existence, its 
 quantum in nature can neither increase nor diminish. 
 
 232 B. — Second A nalogy. 
 
 Principle of Causal Succession. 
 
 All changes take place in conformity with the law of the 
 connection of cause and effect. 
 
 Proof. 
 
 234 The apprehension of the various determinations of a 
 phenomenon is always successive. The ideas of the parts 
 follow one another in consciousness. Whether the parts 
 follow one another in the object also, is a different 
 thing. Now, anything whatever of which we are con- 
 scious, anything of which we have an idea, we may 
 certainly call an object ; but it is not so easy to say what 
 
 235 is meant when the term object is applied to a phenome- 190 
 non. In this case by an object must be understood, not 
 
 a mere idea, but only that in the idea which stands for 
 an object. But in so far as by an object we mean merely 
 our own ideas as objects of consciousness, there is no 
 distinction between actual objects of sense and the 
 
TRANSCENDENTAL ANALYTIC. Ill 
 
 apprehension or reception of them in the synthesis of 
 imagination. So far we must therefore say that the 
 various determinations of phenomena are always pro- 
 duced in the mind successively. Were phenomena 
 things in themselves, no man could tell how the various 
 determinations, as they arise one after the other in con- 
 sciousness, might be connected in the object. As we 
 cannot go outside of our own consciousness, there is no 
 possible way of knowing how things may be in them- 
 selves, apart from the ideas through which we are 
 affected by them. But, although phenomena are not 
 things in themselves, and yet are the only things that 
 can be presented to us as knowledge, it is necessary to 
 explain what there is in phenomena themselves that can 
 connect their various determinations in time, while yet 
 the consciousness of those determinations is in apprehen- 
 sion always successive. Thus, for instance, the appre- 
 hension of the various determinations contained in the 
 perception of a house is successive. But no one would 
 think of saying that the determinations of the house itself 
 236 are successive. Now, when I ask how an object is to be 
 conceived from the transcendental point of view, I find 
 that the house is not a thing in itself, but only a phe- 
 nomenon, that is, it is the consciousness of something, ^9^ 
 the transcendental object of which is unknown. The 
 question therefore is, what is meant by the connec- 
 tion of various determinations in the phenomenon itself, 
 that phenomenon being yet no thing in itself. Here that 
 which Hes in the successive apprehension is considered as 
 mere modes of my consciousness, while the phenomenon 
 which is given to me, although it is nothing but a complex 
 of these modes, is yet regarded as their object, and the con- 
 
1 12 THE PHILOSOPHY OF KANT. 
 
 ception which I derive from them is held to harmonize 
 necessarily with that object. It soon becomes evident 
 that, as truth consists in the agreement of knowledge 
 with its object, the only question here must be in regard 
 to the formal conditions of empirical truth. The phe- 
 nomenon as an object can be opposed to apprehension 
 as a series of states of consciousness, only on the ground 
 that it is a unique mode of apprehension, which stands 
 under a rule that necessitates the connection of its various 
 determinations in a certain way. That in the phenome- 
 non which contains the condition of this necessary rule 
 of apprehension, is the object. 
 
 Let us now go on to our special problem. There can 
 be no empirical observation that something has occurred, 
 that is, that something, or some state, has come to be 
 2 -> 7 which before was not, unless there has previously been 
 observed something that does not contain this state in 
 itself. For, an actual thing following upon an empty 192 
 time, an absolute beginning preceded by nothing, can 
 no more be apprehended than empty time itself. Every 
 apprehension of an event is therefore a perception that 
 follows upon another perception. But, as this is true in 
 all synthesis of apprehension, even in such a synthesis as 
 that of the determinations of a house already instanced, 
 there is nothing in the mere succession of perceptions to 
 distinguish the apprehension of an event from any other 
 apprehension. But I note further, that when I am con- 
 scious of a phenomenon as containing an event, the 
 perception of the antecedent state A cannot follow the 
 perception of the consequent state B, but, on the con- 
 trary, B, in my apprehension, always follows A, while A 
 never follows B but can only precede it. I see, for 
 
TRANSCENDENTAL ANALYTIC. 1 13 
 
 instance, a ship moving down stream. I first observe 
 it higher up the stream, and then lower down, and it is 
 impossible that in the apprehension of the phenomenon 
 I should first observe the ship lower down the stream 
 and then higher up. The order in which the perceptions 
 follow one another in my apprehension is here deter- 
 mined, and to that order my apprehension is tied down. 
 238 In the former example of the house, my apprehension 
 might begin with a perception of the roof and end with 
 the basement, but it might just as well begin from below^ 
 and end above, or again the units of my empirical 
 observation might be apprehended from right to left or 
 from left to right. In that series of observations there 193 
 was therefore no fixed order that made it necessary for 
 my apprehension to begin at a certain moment in the 
 empirical combination of the various elements of percep- 
 tion. But, in the observation of any event, there always 
 is a rule that makes the order in which the elements of 
 perception follow one another in my apprehension a 
 necessary order. 
 
 In this case, therefore, the subjective succession of 
 apprehension must be derived from the objective succes- 
 sion of the phenomena. Were it not so, there would be 
 nothing whatever to determine the order of succession 
 in my apprehension, and to distinguish one sort of phe- 
 nomenon from another. Viewed by itself a mere 
 succession of apprehension is quite arbitrary, and tells us 
 nothing about the connection of the elements of per- 
 ception in the object. The objective connection must 
 therefore consist in the order in which the elements of 
 perception follow each other, the order being this, that 
 the apprehension of one event follows the apprehension 
 
 H 
 
114 THE PHILOSOPHY OF KANT. 
 
 of another event in conformity with a rule. Thus only am 
 I justified in saying, that there is succession in the phe- 
 nomenon, and not merely in my apprehension, or^ in other 
 words, that I cannot possibly have the apprehension in 
 any other order. 
 
 In conformity with this rule, there must lie in that 
 
 239 which precedes any event, the condition for a rule by 
 which the event always and necessarily follows ; but 
 
 I cannot say, conversely, that I can go back from 194 
 the event and apprehend what precedes it. No pheno- 
 menon goes back from a given point of time to an 
 antecedent point of time, but it yet is related to so7ne 
 antecedent poifit of tifne; on the other hand, the progression 
 from a given time to the precise time that follows is 
 necessary. Now something certainly follows, and this I 
 must necessarily refer to something else, which precedes 
 it and upon which it follows necessarily or in conformity 
 with a rule. Accordingly, the event, as that which is 
 conditioned, points back with certainty to some 
 condition, and this condition is what determines the 
 event. 
 
 240 When therefore we have experience of any event, we 
 always presuppose that something has gone before, 
 on which the event follows according to a rule. Other- 
 wise I should not say that the object follows, for I am 
 not justified in saying that there is succession in an 
 object merely because there is a succession in my 195 
 apprehension, but only because there is a rule that 
 determines the succession of my apprehension by relation 
 
 to what precedes. It is therefore always by reference to 
 such a rule that I make my subjective synthesis or 
 synthesis of apprehension objective, and under this 
 
TRANSCENDENTAL ANALYTIC. II5 
 
 presupposition, and this presupposition only, is even 
 the experience of an event possible at all. 
 
 No doubt this seems to contradict the whole view of 
 the course of thought that the facts have always been 
 held to warrant. The accepted doctrine is, that, from 
 the repeated observation and comparison of many cases 
 in which certain events follow certain antecedents, we are 
 
 241 first led to the discovery of a rule according to which the 
 events invariably follow those antecedents, and then by 
 reflection on the rule, to the general conception of cause. 
 But on that showing, the conception of cause would be 
 merely empirical, and the rule based upon it, that every 196 
 event has a cause, would be just as contingent as the 
 experience from which it was derived. Having no 
 a priori foundation, but resting merely on induction, 
 it would have no genuine universality, but only a purely 
 suppositious universality and necessity. The truth is, 
 that here the same principle applies as in the case of 
 other pure a priori elements, for instance, space and 
 time : the principle that we can derive a clear conception 
 from experience, only because we have ourselves put it 
 into experience, and, indeed, have thereby made ex- 
 perience possible at all. No doubt we cannot have a 
 logically clear idea of cause, as a rule that determines 
 the series of events, until we have made use of it in 
 experience, but it is none the less true, that a tacit 
 reference to that rule, as a condition of the synthetic 
 unity of phenomena in time, was the foundation of 
 experience itself, and therefore preceded it a priori. 
 
 244 No experience whatever is even possible without 199 
 understanding, and the first thing that understanding 
 does, is not to make the conception of special objects 
 
Il6 THE PHILOSOPHY OF KANT. 
 
 clear, but to make the very consciousness of an object 
 
 245 possible. Now, this it effects by conferring upon phen- 
 omena and their existence order in time, assigning a 
 priori to each of them as consequent a determinate posi- 
 tion in time relatively to what precedes. Were such 200 
 position in time not assigned to phenomena, they would 
 not harmonize with time itself, all the parts of which have 
 their position determined a priori. Now, the determinate 
 position of phenomena cannot be learned from the 
 relation of phenomena to absolute time, for absolute 
 time cannot be observed ; on the contrary, it is the 
 phenomena that must determine for one another their 
 position in time, making the order in time in which each 
 occurs necessary. That which follows or occurs, must 
 follow in conformity with a universal rule, on that which 
 was contained in a preceding state. Thus arises a series 
 
 of phenomena, which, by the action of understanding, 
 necessarily assumes in the series of possible perceptions 
 the very same order and unbroken connection which are 
 found a priori in time itself, as the form of inner 
 perception in which all perceptions must have their 
 position. 
 
 The perception of an event is therefore a possible 
 experience, which becomes an actual experience, when I 
 regard a phenomenon as determined to its position in 
 time, and therefore as an object that can always be found 
 in the connection of perceptions in accordance with a rule. 
 
 246 This rule, by which a thing is determined conformably 
 with the succession of time, is, that in what precedes is to 
 be found the condition under which an event always 
 or necessarily follows. 
 
 The proof of this proposition rests entirely upon the 201 
 
TRANSCENDENTAL ANALYTIC. 117 
 
 following grounds. All empirical knowledge implies the 
 synthesis by imagination of various determinations. 
 This synthesis is always successive, or, in other words, 
 the various determinations always follow one another in 
 consciousness. In this synthesis of imagination, how- 
 ever, there is no fixed order of succession, for the series 
 of ideas may be taken just as well backwards as forwards. 
 But, if this synthesis is a synthesis of apprehension, 
 in which there is a consciousness of the various deter- 
 minations contained in a given phenomenon, the order is 
 determined in the object, or, more exactly, there is in our 
 apprehension an order of successive synthesis that 
 determines an object, and in conformity with that order 
 something must necessarily precede, and if it exists, 
 something else must necessarily follow. If, therefore, in 
 my observation I am to obtain the knowledge of an event, 
 that is, of something that actually takes place, my 
 observation must carry with it an empirical judgment, in 
 which the succession is thought as so determined that the 
 event in question is preceded by something else, which it 
 247 follows necessarily or according to a rule. Were this not 
 so, were I to determine the antecedent as existing, with- 
 out being forced to recognize the event as following, I 
 should be compelled to regard the succession as a mere 
 subjective play of my imagination, or, if I still supposed 
 it to be objective, I must call it a mere dream. Hence 
 that relation of phenomena, that is of possible perceptions, 202 
 in which the consequent is necessarily determined in 
 its existence in time by some antecedent in accordance 
 with a rule — the relation, in a word, of cause and effect 
 — is the condition of the objective validity of our 
 empirical judgments with regard to the series of percep- 
 
Il8 THE PHILOSOPHY OF KANI. 
 
 tions, and therefore the condition of experience. The 
 principle of causaHty thus appHes to all objects of 
 experience that stand under the conditions of succession, 
 just because it is itself the ground of the very possibility 
 of such experience. 
 
 256 C. — Third Analogy. 211 
 
 Principle of Community, 
 
 All substances, in so far as they can be observed to 
 CO- exist in space, are in thoroughgoing reciprocity. 
 
 Proof. 
 
 258 Things are co-existent which exist at one and the same 
 time. But how do we know that they exist at one and 
 the same time ? Only if in the synthesis of apprehension 
 the order in which the various determinations arise in 
 consciousness is indifferent, or can go either from A 
 through B, C, D, to E, or conversely from E to A. 
 Were the determinations actually to follow one another 
 in time, that is, in an order that began with A and ended 
 with E, it would be impossible for apprehension to 
 start from E and go backwards to A ; for A would 
 in that case belong to a time that was past, and therefore 
 could no longer be an object of apprehension. 212 
 
 Now, suppose that a number of substances could 
 be observed, each of which was so completely isolated 
 from the rest that none acted upon any other or was 
 
 259 itself acted upon: then I say, that those objects could 
 not possibly be observed to co-exist^ and that there is no 
 way in which by empirical synthesis we could pass from 
 the existence of one to the existence of another. If the 
 objects are assumed to be separated by a space that 
 
TRANSCENDENTAL ANALYTIC. II9 
 
 is quite empty, no doubt the existence of each might be 
 presented in turn in a series of observations; but this 
 would not enable us to say, whether the different 
 phenomena themselves followed one another or existed 
 at the same time. 
 
 If, therefore, our supposed substances are to be known 
 as empirically co-existent, it must be by something more 
 than its mere existence that A determines the position in 
 time of B, and, conversely, B the position in time of A. 
 Now, only that which is the cause of a thing or of its 
 determinations, can determine the position of that thing 
 in time. And, as a substance does not itself begin to be 
 in time, but cily its determinations, every substance must 
 contain in itself at once the causality of certain deter- 
 minations in another substance, and the effects of the 
 causality of that other substance. In other words, if 
 substances are to to be known in experience as co-exist- 
 ing, they must, directly or indirectly, stand with one 123 
 another in a relation of dynamical community. Now, we 
 must regard as necessary to the objects of experience, 
 260 that without which the experience of these objects would 
 itself be impossible. It is, therefore, necessary that all 
 substances, in so far as they are co-existent phenomenal 
 should stand with one another in thoroughgoing com- 
 munity of reciprocity. 
 
 The word community is here used in the sense of 
 dynamical community (commercium), without which even 
 local community {communio spaiii) could never be empiri- 
 cally known. Any one may easily gather from his own 
 experience, that only continuous influences in all points 
 of space can lead our senses from one object to another. 
 The light which plays between our eye and the heavenly 
 
I20 THE PHILOSOPHY OF KANT. 
 
 bodies produces a mediate community between us and 
 them, and shows us that they co-exist. Nor could we 
 change our position empirically, that is, observe the 
 change in our position, if matter were not everywhere, 
 and if the parts of a material object did not manifest 
 their simultaneous existence by means of their influence 
 on one another. It is in this indirect way that we learn 
 the co-existence of all material objects, even the most 
 distant. Without community there could be only a 214 
 number of detached observations ; the chain of empirical 
 261 ideas constituting experience would be continually begin- 
 ning with a new object, having absolutely no connection 
 with that which preceded it, and standing with it in no 
 relation of time. This does not prove that there is no 
 empty space ; empty space there may perhaps be, to 
 which perception cannot reach, and where there is, there- 
 fore, no empirical knowledge of co-existent objects ; but 
 such a space is certainly not a possible object of 
 experience. 
 
 A word by way of further explanation may be useful. 
 All phenomena, of which we can possibly be conscious 
 in experience, must in our mind stand in the community 
 of apperception. And so far as we can be conscious of 
 objects as co-existing, we must be conscious that, by 
 reciprocally determining their position in time, they con- 
 stitute a whole. If this subjective community is to rest 
 upon an objective ground, or to hold of phenomena as 
 substances, it must be because the observation of one 
 object is the necessary condition of the observation of 
 another, and vice versa. Otherwise we must say, that 
 the succession which belongs to all observation viewed as 
 apprehension holds also of objects, and that objects 
 
TRANSCENDENTAL ANALYTIC. 121 
 
 cannot be known as co- existing. But if objects can be 
 known in experience as co-existing, there must be a 
 reciprocal influence, or real community {commercium) of 
 substances. Through this commercium phenomena, in so 215 
 
 262 far as they are external to one another and yet stand in 
 connection, are members of a systematic whole {compositum 
 reale) and are related in many ways within that whole. 
 The three dynamical relations, from which all others flow, 
 are therefore the relations of inherence^ consecution^ and 
 composition. 
 
 These are the three Analogies of Experience. They 
 are simply the principles by which the existence of 
 phenomena in tim.e is determined, in conformity with 
 the three possible modes of time. There is, firstly, the 
 relation to time itself as a magnitude, the magnitude of 
 existence, that is, duration ; secondly, the relation in 
 time as a series, the parts of which follow one another ; 
 lastly, the relation likewise in time as a sum of all exis- 
 tence, the members of which are co-existent. This unity 
 of determination in time is dynamical through and 
 through, that is, time is not regarded as that in which 
 experience directly determines to each existence its own 
 place ; this indeed is impossible, because it is not possi- 
 ble to observe an absolute time, in which phenomena 
 might be held together. The unity is due to a rule of 
 understanding, through which only the existence of 
 phenomena can obtain synthetic unity in conformity 
 with relations of time, and which determines to each 
 object its place in time, and that too a priori^ and as 
 holding for all and every time. 
 
 263 Nature, in the empirical sense of the word, is the 216 
 
122 THE PHILOSOPHY OF KANT 
 
 connected system of phenomena as conforming in their 
 existence to necessary rules or laws. There are therefore 
 certain laws, and these a priori^ that make nature possible 
 at all. Empirical laws can be found out and established 
 only by means of experience, and even they are subject 
 to those primary laws that make experience possible. 
 Our analogies, therefore, properly exhibit the unity of 
 nature in the connected system of phenomena under 
 certain exponents, and these exponents express nothing 
 but the relation of time, as embracing all existence within 
 itself, to the unity of apperception, a unity that is possible 
 only in the synthesis conformed to rules. Taken together 
 they affirm, that all phenomena belong, and must belong, 
 to one system of nature, inasmuch as, apart from this 
 a priori \m\\.y ^ there could be no unity of experience, and 
 therefore no determination of objects in experience. 
 
 r 4. Postulates of all Empirical Thought. « 
 
 I. That which agrees with the formal conditions of 
 experience, or conforms to pure perception and pure 
 conception, is possible. 
 266 2. That which is bound up with the material con- 
 ditions of experience, or with sensation, is actual. 
 
 3. That which, in its connection with the actual, is 
 determined in accordance with the universal conditions 
 of experience, is necessary^ or necessarily exists. 
 
 Explanation. 219 
 
 The categories of modality have this peculiarity, that 
 they do not in the least enlarge the conception to which 
 they are attached as predicates, but merely express its 
 
TRANSCENDENTAL ANALYTIC. 1 23 
 
 relation to the faculty of knowledge. Granting the con- 
 ception of a thing to be quite complete, I may yet ask 
 whether the object is possible or actual, and if actual, 
 whether it is also necessary. Such determinations are 
 not conceived to belong to the object itself; the only 
 point is how the object, together with its determinations, 
 is related to understanding in its empirical use, to 
 empirical judgment and to reason as applied to experience. 
 
 267 (i) The first postulate demands that the conception of 220 
 things should agree with the formal conditions of any 
 experience whatever. Now this objective form of experi- 
 ence includes all synthesis that is essential to the know- 
 ledge of objects. A conception may imply synthesis, but 
 
 if the synthesis does not belong to experience, either as 
 being derived from it, or as forming its a priori con- 
 dition, the conception must be held as empty, and as 
 
 268 not related to any object. There is, for instance, no 
 contradiction in the conception of a figure that is enclosed 
 by two straight lines, for the conception of two straight 
 lines, and the conception of two such lines meeting, do 
 not involve the negation of a figure : the impossibility 221 
 arises, not from the conception in itself, but from the 
 conditions of space and of the determination of space, 
 which prevent the construction of such a figure. But 
 these conditions have an objective reality of their own, 
 
 or apply only to possible things, because they contain in 
 themselves the a priori form of experience in general. 
 
 270 Setting aside everything the possibility of which can be 223 
 learned only from actual experience, let us limit ourselves 
 
 to the question whether anything is possible through a 
 
 271 priori conceptions. Now, I maintain that nothing can be 
 determined as possible through such conceptions alone, 
 
124 THE PHILOSOPHY OF KANT. 
 
 but only in so far as they are merely formal and objective 
 conditions of experience in general. 
 
 At first sight it no doubt seems as if the possibility of 
 a triangle could be known from the mere conception of 
 it ; the conception certainly is independent of experi- 
 ence, and we can as a matter of fact give to it an object, 
 that is, we can construct the triangle completely a priori. 
 But as the triangle so constructed is merely the form of 
 an object, it would remain a mere product of imagination, 224 
 and we could not tell whether any object corresponding 
 to it was possible, if we could not show that such a figure 
 is thought under no conditions but those on which all 
 
 272 objects of experience rest. It is true that we are able to 
 know and to characterize the possibility of things even 
 prior to experience ; but this we can do, only because we 
 are able to determine completely a priori the formal 
 relations under which any object whatever can be known ; 
 and even then we can determine the possibility of things 
 only relatively to experience and within the limits of 
 experience. 
 
 (2) The postulate which relates to the knowledge of 225 
 things as actual dQvcid,V).diS perceptio/is of sense, and therefore 
 sensations of which we are conscious. True, it is not 
 necessary that we should be directly conscious through 
 sense of the object that is to be known, but we must be 
 avvare of its connection with some actual perception, in 
 accordance with those analogies of experience which 
 exhibit the conditions of all real connection in experi- 
 
 273 ence. That which is characteristic of actuality is found 
 solely in the perception of sense that gives to a concep- 
 tion its matter. At the same time, we may know even 
 before perception that a thing actually exists, and there- 
 
TRANSCENDENTAL ANALYTIC. I25 
 
 fore is in a sense a priori, if we can but show that it is 
 inseparably related to certain perceptions, in accordance 226 
 with the principles or analogies that determine the 
 empirical connection of all perceptions. Thus from 
 observation of the attraction of iron filings, we know that 
 a magnetic matter pervades all bodies, although our 
 organs of sense are so constituted that we cannot directly 
 perceive it. For, by the laws of sensibility and the con- 
 text of our perceptions, we should have a direct percep- 
 tion of that matter in experience were our senses only 
 fine enough. Our knowledge of the existence of things, 
 therefore, extends as far as perception, or valid inferences 
 from perception, will carry us. But if we do not start 
 
 274 from experience, and proceed in accordance with the 
 laws of the empirical connection of phenomena, we shall 
 in vain try to guess or to discover the existence of any- 
 thing whatever. 
 
 279 (3) The third postulate refers to material necessity 
 or necessity in existence, not to merely formal and 
 logical necessity in the connection of conceptions. Now, 
 the existence of an object of sense cannot be known 
 completely a priori, but only comparatively a priori, or 
 relatively to something else the existence of which is 227 
 already known ; hence necessity of existence can never 
 be derived from conceptions, but only from the connec- 
 tion of an object through general laws of experience with 
 what has been perceived. There is no existence, how- 
 ever, that can be known to be necessary, on condition 
 that other phenomena have been presented, except the 
 existence of effects as following from given causes in 
 conformity with laws of causality. It is, therefore, not 
 the existence of things or substances that we can know to 
 
126 THE PHILOSOPHY OF KANT. 
 
 be necessary, but only the existence of their state ; and 
 280 the existence of their state we can know to be necessary 
 only from its connection, in accordance with empirical 
 laws of causality, with other states given in perception. 
 From this it follows that the criterion of necessity lies 
 entirely in the law of possible experience, the law that 
 whatever happens is determined a priori in the object 
 through its cause. We cannot know any effects in nature 
 to be necessary except those effects the causes of which 
 are given to us, and hence the criterion of necessity in 
 existence does not apply beyond the field of possible 
 experience. Nor does it apply even within experience 
 to the existence of things as substances, because sub- 
 stances can never be regarded as empirical effects, or 
 something that happens and begins to be. Necessity 228 
 applies only to the relations of phenomena as standing 
 under the dynamical law of causality, and to the possi- 
 bility that is based upon it of concluding a prioj'i from a 
 given mode of existence (the cause) to another mode of 
 existence (the effect). 
 
 General Remark ofi the Principles of Judgment. 
 
 288 It is very remarkable that there is nothing in a cate- 
 gory, taken by itself, that enables us to say whether a 
 real thing corresponding to it is possible, and that a pure 
 conception of understanding can be shown to have 
 objective reality only if a perception is brought forward 
 to which it can apply. 
 
 291 But what is still more remarkable is that the categories 
 cannot be shown to be conditions of the possibiHty of 
 things, and therefore to have objective reality, without the 
 
TRANSCENDENTAL ANALYTIC. 1 27 
 
 aid not only of perceptions, but even of external percep- 
 tions. Take, for instance, the pure conceptions of 
 relation. Here we find (i) that, in order to show that 
 there is something permanent, which corresponds to the 
 conception of substance, and thus to prove the objective 
 reality of the conception, we must have the perception of 
 that which is in space, in other words, the perception of 
 matter; for only space has in it anything permanent, 
 whereas time, and therefore all that exists in the inner 
 sense, is in perpetual flux. (2) The perception which 
 corresponds to the conception of causality is change. 
 Now, to have a real consciousness of change, we must 
 have the perception of motion, or change of place, and 
 indeed it is only by reference to motion that we can 
 realize what change means. No pure understanding can 
 comprehend how change is possible, for in itself change 
 combines determinations that contradict one another 
 292 when they are predicated of the same thing. How, in 
 the very same thing, there should follow from a given 
 state another state that is its opposite, is not only beyond 
 the power of reason to comprehend without a special 
 instance, but without perception it cannot be made 
 intelligible to it at all. The only perception which 
 fulfils this requirement is that of the motion of a point in 
 space, for, by its presence in different places, the point 
 presents us with a series of reciprocally exclusive deter- 
 minations, and thus enables us to realize the meaning of 
 change. Even in the case of inner changes, we have to 
 figure time, the form of inner sense, as a line, and the 
 inner changes themselves as the generation of that line, 
 that is, as motion. Thus it is by means of external per- 
 ception that we make intelligible to ourselves the various 
 
128 THE PHILOSOPHY OF KANT. 
 
 successive states in which we ourselves exist. The true 
 explanation of this is that no change can possibly be an 
 object of experience apart from the consciousness of 
 something that is permanent, and that in inner sense 
 nothing that is permanent can be found. (3) Nor can 
 the possibility of the category of conwiunity be conceived 
 by reason alone, and hence its objective reality can be 
 seen to be possible only by reference to perception, and 
 indeed only by reference to external perception in space. 
 How can we think it to be possible that there should be 
 anything in the existence of one substance to affect 
 reciprocally the existence of other substances, and that, 
 
 293 therefore, because there is something in the former, there 
 must be something also in the latter which could not be 
 understood from the existence of the latter when it is 
 considered merely by itself? This is what community 
 demands, and yet it is inconceivable, if things subsist by 
 themselves, or are completely isolated from one another. 
 The answer is that we can make the possibility of the 
 community of substances, that is, of objects of experience, 
 intelligible to ourselves only by representing them in 
 space, and, therefore, in external perception. For space 
 by its very nature contains in itself a priori formal 
 external relations, and these are conditions of the possi- 
 bility of the real relations of action and reaction, and 
 therefore of community. Similarly it might readily be 
 shown that the possibility of things as qtiania^ and there- 
 fore the objective reality of the category of quantity, 
 requires external perception, and that only by means of 
 external perception can we have experience even of any- 
 thing in inner sense as a quantum. 
 
 294 The net result of this whole section is this : — All 
 
TRANSCENDENTAL ANALYTIC. I 29 
 
 principles of pure understanding are nothing but a priori 
 principles of the possibility of experience; and all a 
 priori synthetic propositions relate only to experience, 
 and indeed from that relation they derive their possibility. 
 
 Chapter III. — Distinction of Phenomena 235 
 
 AND NOUMENA. 
 
 295 We have seen that, whatever understanding produces 236 
 from itself, it holds in trust solely in the interest of 
 
 296 experience. The principles of pure understanding, 
 whether as mathematical they are a priori constitutive 
 principles, or as dynamical merely regulative principles, 
 contain nothing but what may be called the pure schema 
 
 for a possible experience. For experience derives its 237 
 unity entirely from the synthetic unity which understanding 
 imparts, originally and spontaneously, to the synthesis of 
 imagination in relation to apperception ; and phenomena, 
 as the data for a possible knowledge, must therefore 
 stand a priori in relation to that synthetic unity and in 
 harmony with it. 
 
 2Q7 Now the proposition that understanding can never 238 
 make a transcendental use, but only an empirical use, of 
 any of its a p?'iori principles, is seen to have very 
 
 298 important consequences, so soon as it is thorougly under- 
 stood. A conception is employed transcendentally 
 when it occurs in a proposition regarding things as such 
 or in the??iselves ; it is employed empirically when the 
 proposition relates merely \.o phenoinena^ or objects of a 
 possible experience. Only the empirical use is admis- 239 
 sible. Every conception requires, firstly, the logical 
 
 form of conception or thought, and, secondly, the possi- 
 
 I 
 
130 THE PHILOSOPHY OF KANT. 
 
 bility of an object being empirically given to which it may 
 be applied. ^Vhere no such object can be given, the 
 conception is empty and meaningless, containing nothing 
 but the logical function which is necessary in order to 
 form a conception out of any data that may be given. 
 Now, the only way in which an object can be presented 
 is in perception. And this perception must be empirical ; 
 for, although pure perception is possible <a:/r/<?r/ before 
 the presentation of an object, yet, as it is a mere form, it 
 can by itself have no object to which it may apply, and 
 therefore it can have no objective value ascribed to it. 
 Hence all conceptions, and with them all principles, even 
 when they are possible a priori^ are none the less relative 
 to empirical perceptions as the data for a possible experi- 
 ence. Apart from this relation they have no objective 
 validity, but are a mere play of imagination or of under- 
 standing. 
 
 300 That this limitation applies to all the categories, and to 240 
 all the principles derived from them, is evident, if only 
 from this, that we cannot give a real definition of even a 
 single one of them, or in other words, make the possi- 
 bility of their object intelligible, without directly referring 
 to the conditions of sensibility, and therefore to the form 
 of phenomena. The categories are thus necessarily 241 
 limited to phenomena as their sole object, and, if this 
 limitation is taken away, all meaning or objective relation 
 vanishes from them, and no possible instance of an 
 object can be adduced to make the conception compre- 
 hensible. 
 
 303 There is therefore no way of avoiding the conclusion 246 
 that the pure conceptions of understanding can never be 
 employed transcendentally, but only empirically, and that 
 
TRANSCENDENTAL ANALYTIC. I3I 
 
 the principles of pure understanding can apply only to 
 objects of sense, as conforming to the universal conditions 
 of a possible experience, and never to things as such, or 
 apart from the manner in which we are capable of 
 perceiving them. 
 
 The Transcendental Analytic has brought us to this 
 important conclusion, that understanding can never do 
 more than supply by anticipation the form for a possible 
 experience; and, as nothing but a phenomenon can be 
 an object of experience, it has taught us that under- 
 standing cannot possibly transcend the limits of sensi- 
 bility, beyond which no objects aie presented to us. The 
 principles of pure understanding are merely exponents of 247 
 phenomena, and for the proud name of Ontology, as a 
 science that claims to supply in a systematic doctrine 
 a priori synthetic knowledge of things as such, must be 
 substituted the more modest claims of an Analytic of 
 Pure Understanding. 
 309 If from empirical knowledge is taken away all that 253 
 thought contributes in its categories, there is no longer 
 any knowledge of an object. By mere perception nothing 
 whatever is thought, and the mere fact that I am con- 
 scious of an affection of my sensibility does not entide 
 me to say that I am conscious of my affection as 
 related to any object. On the other hand, even if all 
 perception is taken away, there still remains the form of 254 
 thought, or the manner in which the various elements of 
 a possible perception are capable of being combined in 
 relation to an object. The categories have therefore in 
 this sense a wider reach than perceptions of sense, that 
 they think objects in general, without looking to the 
 particular manner in which they may be presented. But 
 
132 THE PHILOSOPHY OF KANT. 
 
 although they are so far independent of sensibility, they 
 do not determine a larger sphere of objects ; for we are 
 not entitled to say that non-sensuous objects can be 
 presented, unless we can show that a sort of perception 
 is possible that is not sensuous. Now this we cannot 
 possibly do. 
 
 310 A conception which cannot be known in any way to 
 have objective reahty may be called problematic, if it is 
 not self-contradictory, and if it is bound up with the 
 knowledge gained through certain conceptions the range 
 of which it serves to limit. Now the conception of a 
 noumeno7i, that is, of a thing that cannot be an object of 
 sense, but is thought, by pure understanding alone, as a 
 thing in itself, is certainly not self-contradictory ; for we 
 cannot know with certainty that sensibility is the only 
 possible mode of perception. Moreover, the conception 
 of a noumenon is necessary to prevent sensuous percep- 
 tion from claiming to extend to things in themselves, and 
 to set a limit to the objective validity of sensuous know- 
 ledge. In the end, however, we are unable to understand 255 
 how such noumena are possible at all, and the realm 
 beyond the sphere of phenomena is for us empty^ 
 We have indeed an understanding that problematically 
 stretches beyond the sphere of phenomena, but we have 
 
 no perception in which objects beyond the field of sensi- 
 bility can be presented, nor can we conceive how such 
 a perception is even possible. Hence understanding 
 cannot be employed assertorically beyond the world of 
 phenomena. The conception of a noumenon is, there- 
 
 311 fore, merely the conception of a limit, a conception which 
 is only of negative use, and but serves to check the 
 presumption of sensibility. But although it is unable to 
 
TRANSCENDENTAL ANALYTIC. 1 33 
 
 establish anything positive beyond the sphere of pheno- 
 mena, the idea of a noumenon is not a mere arbitrary 
 fiction, but is connected in the closest way with the 
 limitation of the sensibility to phenomena. 
 
 The positive division of objects into phenomena and 
 noumena, and of the world into a sensible and intelligible 
 world, is therefore quite inadmissible. Certainly, the 
 distinction of conceptions as sensuous and intellectual is 
 legitimate. But, as intellectual conceptions do not 
 determine any object for themselves, they can have no 
 objective validity. If abstraction is made from sense, 
 how shall it be made intelligible, that the categories, 256 
 which are then the only means of determining noumena, 
 have any meaning whatever ? The mere unity of thought 
 is not the same thing as the determination of an object ; 
 for knowledge also requires that the object to which 
 that unity can be applied, should be capable of being 
 presented in a perception. At the same time, if the 
 conception of a noumenon is interpreted in a problematic 
 sense, it is not only admissible but indispensable, serving 
 as it does to define the limits of sensibility. In that 
 sense, however, a noumenon is not a special kind of 
 object for our understanding, namely, an intelligible 
 object \ on the contrary it is problematic whether there is 
 any understanding that could have such an object actually 
 312 before it. Such an understanding would not know its 
 object discursively by means of categories, but intuitively in 
 a non-sensuous perception ; and how this is possible we 
 cannot form even the faintest conception. Still, in the con- 
 ception of a noumenon our understanding gets a sort of 
 negative extension ; for in calling things in themselves 
 noumena, and viewing them as not objects of sense 
 
134 THE PHILOSOPHY OF KANT. 
 
 it rather limits the sensibility than is limited by sensi- 
 bility. At the same time, understanding cannot limit 
 sensibility without also setting limits to itself, for it has 
 instantly to add, that things in themselves cannot be 
 known by means of categories, and all that remains is 
 to think them under a name that indicates something 
 unknown. 
 315 There are, therefore, no principles through which the 259 
 conception of pure, merely intelligible objects could 
 ever be applied, for we cannot imagine any way in 
 which such objects could be presented to us. The 
 problematic thought, which leaves a place open for 
 intelligible objects, serves only, as a sort of empty space, 
 to limit the empirical principles, without containing 260 
 within it or indicating any object of knowledge that 
 lies outside the sphere of those principles. 
 
TRANSCENDENTAL DIALECTIC. 1 35 
 
 349 TRANSCENDENTAL DIALECTIC. 293 
 
 INTRODUCTION. 
 
 I. Transcendental Illusion. 
 
 351 We have here nothing to do with optical illusion, 295 
 
 352 or with empirical illusion of any kind, which occurs in 
 the empirical use of correct rules of understanding, and 
 arises from the misleading influence of imagination upon 
 judgment. What we propose to consider is transcen- 
 dental illusion, which is due to the use of principles 
 that have no bearing upon experience, and therefore 
 cannot be tested by experience. Contrary to all the 
 warnings of criticism, this illusion tempts us to apply 
 the categories beyond experience, and cheats us with 
 the dream of an extension of pure understanding be- 
 yond the limits of experience. Principles which are 
 applied entirely within the limits of possible experience 296 
 we shall call immanent, those which seek to transcend 
 these limits we shall call transcendent. In calling a 
 principle transcendent, I do not mean to indicate simply 
 the transcendental use, or, as we should rather call it, 
 misuse of the categories. This is merely a defect in 
 judgment, when it has not been chastened by criticism, 
 and therefore does not pay due heed to the limits 
 within which alone pure understanding has full sway. 
 
 A principle is said to be transcendent, when it positively 
 
136 THE PHILOSOPHY OF KANT. 
 
 asserts our right to break down all such barriers, 
 and to take possession of an entirely new realm, into 
 which it can enter only if all limits to knowledge have 
 been taken away. Transcendental and transce?ident are 
 therefore not the same thing. The principles of pure 
 understanding, as we have seen above, can be employed 
 353 only empirically, never transcendentally, inasmuch as 
 they do not extend beyond the limits of experience. 
 But a principle which denies those limits, or even com- 
 mands us to transcend them, must be called transceti- 
 dent. Now, if our Critique succeeds in exposing the 
 illusion of these pretended principles, the principles 
 which are employed only empirically may be called, in 
 contrast to the former, immanent principles of pure 
 understanding. 
 
 The logical illusion of a sophistical syllogism con- 
 sists in an imitation of the form of reason, and arises 
 solely from a want of attention to the rules of logic. 
 It therefore vanishes the moment our attention is 297 
 aroused. Transcendental illusion, on the other hand, 
 does not disappear, even when it has been brought 
 under the light of transcendental criticism, and when 
 its fallacy has been clearly detected ; as is the case, 
 for instance, in the proposition, that the world must 
 have a beginning in time. The explanation of this is, 
 that in our reason, considered as simply a faculty of 
 human knowledge, there lie fundamental rules and 
 maxims of its use, which have all the appearance of 
 objective principles. Hence we commonly mistake the 
 subjective necessity, which is essential to the connection 
 by understanding of our conceptions, for an objective 
 necessity in the determination of things in themselves. 
 
TRANSCENDENTAL DIALECTIC. 137 
 
 354 Transcendental Dialectic must therefore be satisfied 
 with bringing to light the illusion in transcendent judg- 
 ments, and guarding us against its deceptive influence ; 
 but it can never put an end to the illusion and cause 298 
 it to disappear, as in the case of logical illusion. For, 
 the illusion is here natural and unavoidable, resting as 
 
 it does upon subjective principles which we cannot 
 help supposing to be objective. So closely is this illus- 
 ion interwoven with the operations of human reason 
 that even after it has been detected in its work of 
 
 355 deception, it never fails to fascinate the reason and 
 to lead to momentary errors, which need to be corrected 
 again and again. 
 
 2. Pure Reason as the Seat of Transcendental 
 
 Illusion. 
 
 356 In the Analytic it has been shown that understanding 299 
 is the faculty of rules ; and now we distinguish reason 
 from understanding by calling it Xhe faculty of pri?iciples. ^oo 
 
 By a principle is usually meant any sort of knowledge 
 that can be employed as a principle, even if in itself, and 
 from the point of view of its origin, it is not a principle 
 at all. Every general proposition, which can stand in a 
 syllogism as its major premise, is in this sense called a 
 principle, even if it has been derived by induction from 
 experience. 
 
 357 Strictly speaking, however, only that knowledge is a 
 principle in which I know the particular in the universal 
 
 by means of conceptions. Now, if we consider the 301 
 universal a priori propositions of pure understanding in 
 themselves and according to their origin, they are very 
 far from yielding knowledge by means of conceptions. 
 
138 THE PHILOSOPHY OF KANT. 
 
 For they would not be even possible a priori^ if we could ;■ 
 
 not refer to pure perception, or to the universal conditions j 
 
 of a possible experience. Understanding cannot possibly '^\ 
 
 358 derive synthetic knowledge from conceptions, and such 
 knowledge is what I mean when I speak of principles in 
 the strict sense of the term, 
 
 359 If, then, understanding is the faculty of reducing 302 
 phenomena to the unity of rules, reason is the faculty of 
 bringing the rules of understanding under principles. 
 Reason never goes directly to experience or to any 
 object, but seeks by means of conceptions to give a 
 priori unity to understanding and its various knowledge. 
 This unity, which may be called the unity of reason, is 
 quite different in kind from that which understanding is 
 capable of producing. 
 
 363 The question arises, then, whether pure reason in 306 
 itself contains a priori synthetic principles and rules, 
 and, if so, what those principles are. 
 
 From the formal and logical procedure of reason in 
 syllogisms we may readily learn the ground upon which 
 the transcendental principle of pure reason in its synthetic 
 knowledge must rest. 
 
 Firstly, in the process of inference reason does not 
 bring perceptions directly under rules, as understanding 
 does with its categories, but deals with conceptions and 
 judgments. No doubt pure reason also relates to objects 
 of perception, but its direct relation is not to perceptions, 
 but only to understanding and its judgments. It is 307 
 understanding only that applies directly to perceptions of 
 sense, and by its judgments determines them as objects. 
 The unity of reason therefore differs essentially from the 
 unity of a possible experience as due to understanding. 
 
TRANSCENDENTAL DIALECTIC. 1 39 
 
 The proposition that whatever happens has a cause, is not 
 a principle known and prescribed by reason. It makes 
 364 the unity of experience possible, and borrows nothing 
 from reason, which could never have prescribed such a 
 synthetic unity from mere conceptions, that is, indepen- 
 dently of all relation to possible experience. 
 
 Secondly, reason in its logical use seeks to reach a 
 premise which contains the universal condition of the 
 judgment that constitutes the conclusion of the syllogism, 
 and the syllogism is itself simply a judgment in which 
 that condition is subsumed under a universal rule con- 
 tained in the major premise. Now, as reason may again 
 seek for a universal condition of that rule, or, in other 
 words, may go as far as it can in search of the condition 
 of a condition, by means of a pro-syllogism, it is plain 
 that the peculiar principle of reason in its logical use is 
 to find for every conditioned knowledge of understanding 
 the unconditioned, and so to complete the unity of 
 knowledge. 
 
 This logical maxim, however, can be regarded as a 
 principle of pure reason only if we assume that when the 
 conditioned is given the whole series of conditions in 
 subordination to one another, and therefore the uncon- 308 
 ditioned, is actually realized, the object being seen in itself 
 and in the whole of its relations. 
 
 Now, such a principle of pure reason is manifestly 
 synthetic, for while it is no doubt true that the con- 
 ditioned, is related analytically to some condition, it is 
 not possible to derive the unconditioned from it by 
 36; analysis. From that principle must also proceed various 
 other synthetic propositions, of which pure understanding 
 knows nothing. These propositions will be transcendent^ 
 
140 THE PHILOSOPHY OF KANT. 
 
 SO far as phenomena are concerned ; that is to say, it will 
 be impossible ever to make an adequate use in experience 
 of the supreme principle of pure reason. That principle 
 will therefore be essentially different from all the principles 
 of understanding, which are always immanent and have 
 no object but to make experience possible. We must 
 therefore inquire, whether the principle that the series of 
 conditions extends as far as the unconditioned has any 
 objective truth, and how the answer to this question 
 affects our view of the empirical use of understanding ; or 309 
 whether it has no objective truth, but is simply a logical 
 rule, telling us to get as near to completeness as we 
 possibly can in the ascent to ever higher conditions, and 
 so to bring our knowledge to the highest unity of which 
 our reason is capable. 
 
 366 BOOK I. 310 
 
 377 Section II. — TraJisce7idental Ideas. 321 
 
 In the Transcendental Analytic we have seen how, from 
 the mere logical form of our knowledge, there arise pure 
 a priori conceptions, which yield the consciousness of 
 
 378 objects antecedently to all experience, or rather point to 
 the synthetic unity that alone makes an empirical know- 
 ledge of objects possible. By conceiving the form of 
 judgments as supplying conceptions for the synthesis of 
 perceptions, we were led to the discovery of the cate- 
 gories, which we found to be the guide of understanding 
 in the whole of its empirical use. We may therefore 
 expect that from the form of syllogisms, as applied to the 
 synthetic unity of perceptions in conformity with the 
 
TRANSCENDENTAL DIALECTIC. I4I 
 
 categories, will arise a special class of a priori conceptions, 
 which may be called pure conceptions of reason or tran- 
 scendental ideas, and which will determine how under- 
 standing is to be employed within the realm of experience 
 as a whole in conformity with principles. 
 
 379 A transcendental conception of reason is, therefore, 322 
 just the conception of the totality of conditions of any- 
 thing that is given as conditioned. Now, the uncon- 
 ditioned alone makes a totality of conditions possible, 
 while conversely the totality of conditions is always itself 
 unconditioned ; hence a pure conception of reason may 
 be defined, generally, as a conception of the uncondi- 
 tioned, in so far as it contains a ground for the synthesis 
 of the conditioned. 
 
 There are as many pure conceptions of reason as 323 
 understanding has functions of relation in its categories. 
 Hence we have, firstly, the unconditioned of the cate- 
 gorical synthesis in a subject; secondly, the unconditioned 
 of the hypothetical synthesis of the members of a series ; 
 thirdly, the unconditioned of the disjunctive synthesis of 
 the parts in a system. 
 
 384 Transce7idental ideas, then, are problematic conceptions 327 
 of pure reason, which regard all empirical knowledge as 
 determined through an absolute totality of conditions. 
 They are not mere fictions, but spring from the very 
 nature of reason itself, and therefore stand in a necessary 
 relation to the whole use of understanding. And, lastly, 
 they are transcendent, inasmuch as they overleap the 
 limits of all experience, in which no object can be pre- 
 sented that is adequate to the transcendental idea. 
 
 It must not be supposed, however, that because tran- 
 scendental conceptions of reason are only ideas, they are 
 
142 THE PHILOSOPHY OF KANT. 
 
 therefore superfluous and useless. For, although ideas 
 cannot determine an object, they may lie at the basis of 
 understanding as an unseen canon, for its extended and 
 consistent use. Adding nothing to what we know of an 
 object by means of the conceptions of understanding, 
 they yet may guide understanding to clearer and wider 
 knowledge; not to mention that they may, perhaps, 
 make the transition possible from the sphere of nature to 
 the sphere of morality. 
 
 390 Section III. — System of Traiiscendental Ideas. 333 
 
 391 All transcendental ideas can be brought under three 334 
 heads : the first, containing the absolute or unconditioned 
 tmity of the thinki?io siibject; the second, the absolute 
 
 luiity of the series of conditio?is of phenomena ; the third, 
 the absolute unity of the condition of all objects of thought 
 whatever. 
 
 The thinking subject is the object oi psychology^ the 
 sum-total of all phenomena (the world) is the object of 
 cosmology, and the being that contains the supreme con- 
 dition of all that can be thought (the Being of all beings) 
 \s \hQ oh]Qci oi theology. Pure reason therefore supplies 
 the idea for a transcendental doctrine of the soul {psycho- 
 
 392 logia rationalis), the idea for a transcendental science of 
 the world {cosmologia ratio?ialis), and, lastly, the idea for 335 
 a transcendental knowledge of God {theologia tra?iscen- 
 dentalis.) 
 
 393 It is readily seen, that the sole aim of pure reason is 336 
 absolute totality of synthesis on the side of the conditions, 
 and that it has nothing to do with absolute completeness 
 
 OJi the side of the conditioned. For the presupposition that 
 
TRANSCENDENTAL DIALECTIC. I43 
 
 the series of conditions should be complete is satisfied, 
 if reason can only present to the understanding a priori 
 a condition that is itself complete and unconditioned, 
 leaving it to understanding to descend from the condition 
 to the conditioned. 
 
 394 It is obvious also, that those three ideas display a 337 
 certain connection and unity among themselves, by 
 means of which pure reason is able to reduce its know- 
 ledge to system. To advance from the knowledge of 
 oneself (the soul) to a knowledge of the world, and 
 through it to a knowledge of the Supreme Being, is a 
 progression so natural, that it suggests the logical 
 
 395 advance of reason from premises to conclusion. 
 
 396 BOOK II. 338 
 
 The Dialectical Conclusions of Pure Reason. 
 
 We may say that the object of a purely transcendental 
 idea is something of which we can form no conception, 
 although the idea itself has arisen with absolute necessity 
 from the primary laws of reason. In fact it is impossible 
 for understanding to have the conception of an object that 
 should be adequate to the demands of reason, for this 
 would mean that we should have a conception that could 
 be exhibited and brought to perception in a possible ex- 
 
 397 perience. But it is better, because less misleading, to say, 330 
 that we can have no knowledge of the object which 
 corresponds to an idea, although we may have a pro- 
 blematic conception of it. 
 
 Now, at least the transcendental or subjective reality 
 of the ideas of reason is reached by a necessary inference. 
 
144 THE PHILOSOPHY OF KANT. 
 
 There are, accordingly, inferences the premises of which 
 contain nothing empirical, and in these we reason from 
 something that we know to something else that we 
 cannot comprehend, but to which by an unavoidable 
 illusion we ascribe objective reality. In their actual 
 result those inferences are sophistical rather than 7'atioJial ; 
 at the same time they are not mere arbitrary fictions, but 
 spring from the very nature of reason, and in that sense 
 are well entitled to be called rational. They are sophis- 
 tications of pure reason itself, which even the wisest man 
 cannot shake off. After much effort he may avoid posi- 
 tive error, but he need not hope to be perfectly free from 
 an illusion that will never cease to mock and bewilder 
 him. 
 
 Corresponding to the three ideas, there are three 
 kinds of dialectical inference. In the first, I reason from 340 
 398 the transcendental conception of the subject, which is 
 perfectly simple, to the absolute unity of the subject itself, 
 of which I have no conception at all. This dialectical 
 illusion I shall call the transcendental /(^r^^^^/j-w. The 
 second kind of dialectical inference is to the tran- 
 scendental conception of an absolute totality in the series 
 of conditions to any given phenomenon. Here I reason, 
 that, as my conception of the unconditioned synthetic 
 unity of the series is always self-contradictory, the op- 
 posite unity, which is equally inconceivable, must be 
 regarded as true. The attitude of reason in this form of 
 dialectical inference I shall call the antinotny of reason. 
 In the third and last kind of sophistical inference of 
 reason, I conclude from the totality of conditions de- 
 manded by the thought of objects as a whole, in so far as 
 these can be given, to the absolute synthetic unity of all 
 
TRANSCENDENTAL DIALECTIC. 1 45 
 
 conditions of the possibility of things in general ; in other 
 words, I reason from things that I cannot know through 
 the mere transcendental conception of them to a Being of 
 all beings that I know still less, since the transcendental 
 conception throws no light upon its existence or its 
 unconditioned necessity. This sort of dialectical in- 
 ference I shall call the ideal of pure reason. 
 
 399 Chapter I. — The Paralogisms of Pure Reason. 34 i 
 
 A logical paralogism is an inference invalid in form, 
 the invalidity of which is quite independent of its content. 
 A transcendental paralogism is an inference also invalid 
 in form, but its formal invalidity has a transcendental 
 source. The wrong conclusion will here be due to the 
 very nature of human reason, and will carry with it an 
 unavoidable though not an inexplicable illusion. 
 
 There is one conception, that we must now put along 
 with the transcendental conceptions contained in the 
 table of categories, but without in any way changing or 
 adding to that table. This is the conception, or, if it is 
 preferred, the judgment, "■ I thmk.''^ It is easy to see, 
 that " / thtJtk " is the common vehicle of all conceptions, 
 and therefore of transcendental as well as empirical con- 
 ceptions. As the vehicle of transcendental conceptions 
 it is itself transcendental, but it cannot claim a special 
 place in the list of these transcendental conceptions, be- 
 
 400 cause it merely serves to indicate that all thought belongs 
 
 to consciousness. And although it is pure, or free from all 342 
 empirical elements or impressions of sense, it yet serves 
 to distinguish between two different kinds of objects, from 
 the different ways in which they are related to conscious- 
 
146 THE PHILOSOPHY OF KANT. 
 
 ness. /, as thinking, am an object of inner sense, and 
 am called soul, while that which is an object of outer 
 sense is called body. Accordingly, the very term / 
 designates the thinking being which is the object of 
 psychology. Psychology, therefore, may be called the 
 rational science of soul, if it seeks to know nothing about 
 the soul but what can be inferred, independently of 
 all experience, from the conception / as present in 
 all thought ; that is, if no attempt is made to determine 
 the I in concrete as a particular object of experience. 
 
 Now, the ratio7ial doctrine of the soul necessarily 
 attempts to do this \ for, if the smallest empirical element 
 of my thought, or any particular perception of my own 
 state, should mingle with the principles of the science, it 
 would no longer be a rational, but would be merely an 
 empirical doctrine of the soul. We have before us, then, 
 what claims to be a science, built upon the single 
 proposition, " I thi?ik,'' and this is the proper place to try 
 its solidity, or want of solidity, by the principles of a 
 transcendental philosophy. It is not a valid objection to 
 rational psychology to say, that, as the proposition 
 " /MzV/y^," expresses the perception of oneself as revealed 
 in experience, the doctrine built upon that perception 343 
 401 can never be purely rational, but must be founded in part 
 upon an empirical principle. For this inner perception 
 is but the mere apperception, " / thitik^^ which is the 
 condition of all transcendental conceptions, and means, I 
 think substance, cause, etc. The determination of the 
 constitution and possibility of inner experience in general, 
 or the general relation of one perception to another, 
 apart from the particular distinction and empirical deter- 
 mination of perception, cannot be regarded as empirical 
 
TRANSCENDENTAL DIALECTIC. I47 
 
 knowledge, but only as the knowledge of what any 
 empirical object must be. Now the investigation into 
 the possibility of experience in general is undoubtedly a 
 transcendental investigation, though the addition of even 
 the smallest ingredient of sense, were it only the feeling 
 of pleasure or pain, to the pure idea of self-consciousness, 
 would at once convert a rational into an empirical 
 psychology. 
 
 ^^ I think,'' is therefore the text of rational psychology, 
 and from this single proposition the whole system must be 
 derived. It is easy to see, that, if this thought is to be 
 used as determining the self as an object, it can bring to 
 it only transcendental predicates^ for any empirical predi- 
 cate whatever must destroy the purity of a rational science, 
 and make it dependent upon experience. 
 
 402 The categories will naturally be our guiding-thread, but 344 
 
 as there is here given to us a thing, the / as a thinking 
 
 being, we shall begin with the category of substance, 
 
 which is predicated of the / as a thing in itself. The 
 
 topic of the rational doctrine of soul, from which all else 
 
 that may be contained in it is derived, is therefore as 
 
 follows : — 
 
 I. The soul is substance. 
 
 2. As to quality, simple. 3. As to the various times 
 
 in which it exists, numeri- 
 cally identical, that is, 
 unity (not plurality). 
 4. In relation \o possible objects in space. 
 
 403 Corresponding to these elements, the transcendental 345 
 doctrine of the soul contains four paralogisms. That 
 doctrine is therefore wrongly held to be a science of pure 
 
 404 reason concerninsj the nature of our thinking self. It has 
 
148 THE PHILOSOPHY OF KANT. 
 
 no foundation but the simple idea /, which is so com- 346 
 pletely empty of all content, that it cannot be called even 
 a conception, but merely a consciousness that accompanies 
 all conceptions. This /, or he, or it, this thing that 
 thinks, is nothing but the idea of a transcendental subject 
 of thought = X, which is known only through the thoughts 
 that are its predicates, and which, apart from them, can- 
 not be conceived at all. We turn round and round it in 
 a perpetual circle, for we can make no judgment about it 
 without making use of the idea of it in our judgment. 
 Nor can this inconvenience be avoided, for consciousness 
 in itself is not so much the distinct idea of a particular 
 object, as a general form of all the ideas through which 
 knowledge of objects is to be obtained, and indeed the 
 only form of which I can say, that without it I can tJimk 
 nothing whatever. 
 406 Now, as the proposition, " / think" taken problem- 348 
 atically, contains the form of every judgment of under- 
 standing, and accompanies all categories as their vehicle, 
 it is clear that the inferences from it must rest upon a 
 purely transcendental use of understanding, in which all 
 aid from experience is rejected. After what has already 
 been shown, we cannot have much faith in the success of 
 such an undertaking. We shall therefore follow the pure 
 doctrine of soul with a critical eye through all its predica- 
 ments. 
 
 But, before doing so, a general remark may be made, 
 which will help to bring out the real character of those 
 inferences. I do not know an object by simply thinking 
 it, but only by determining a given perception relatively 
 to the unity of consciousness in which all thought consists. 
 To know myself as an object, it is therefore not enough 
 
TRANSCENDENTAL DIALECTIC. 1 49 
 
 simply to be conscious of myself as thinking, but I must 
 be conscious of the perception of myself as determined 
 relatively to the function of thought. Now, none of the 
 
 407 7nodi of self-consciousness in thinking are by themselves 
 conceptions of objects or categories : they are merely, 
 logical functions, which can give me no knowledge of 
 myself as an object, because they can give no knowledge 
 of an object at all. To know myself as an object of my 
 own inner perception, I must be conscious of the self as 
 object, and not simply as deterjnining subject; in other 
 words, I must be conscious of the various determinations 
 of myself, in so far as these can be brought together in 
 conformity with the unity of apperception, which is the 
 universal condition of all combination in thought. 
 
 (i) There is no doubt that in any judgment I am the 
 detennitiing subject of the relation in which the judgment 
 consists. The proposition that I, I that think, am the 
 subject in every act of thought, and cannot possibly be 
 regarded as a predicate pertaining to thought, is not only 
 a necessary but even an identical proposition. But this 
 does not mean that I am conscious of myself as an object 
 in such a way that I can determine myself as a self- 
 sichsistent being or substafice. The latter proposition goes 
 a very long way indeed beyond the former, and demands 
 for its proof data that will certainly never be found in the 
 / as simply thinking, and perhaps will never be found in 
 it at all, in so far as it is considered as thinking. 
 
 (2) That the / of apperception, and therefore the / in 
 each act of thought, is o?te, and cannot be resolved into a 
 plurality of subjects, or is a logically simple subject, is 
 implied in the very conception of thinking, and may be 
 
 408 derived from it by mere analysis. But this does not 
 
150 THE PHILOSOPHY OF KANT. 
 
 mean that the thinking / is a svnple substance^ which 
 would be a synthetic proposition. The conception of 
 substance is ahvays relative to perceptions, and as these 
 in us can only be sensuous, they lie quite beyond the 
 field of understanding and its thinking. But it is precisely 
 of thought, as distinguished from perception, that we are 
 speaking, when we say that the / in thinking is simple. 
 Now, in all other cases, it is a very difficult thing to tell 
 what in any given perception is substance, and it is still 
 more difficult to say whether the substance can be simple, 
 as, for instance, whether matter is made up of simple 
 parts. It would therefore be a very remarkable thing 
 indeed, if the poorest of all our ideas should by a sort of 
 revelation enable us to say at once that the / is a simple 
 substance. 
 
 (3) The proposition, that in the various determinations 
 of my consciousness I am identical with myself, is Hke- 
 wise implied in the conceptions themselves, and is there- 
 fore an analytical proposition. But this identity of the 
 subject in all the determinations of which I can be 
 conscious is not the same thing as a perception in which 
 the self is presented as an object which can be recognized 
 as self-identical. The mere consciousness of the identity 
 of the subject in all its determinations does not mean 
 the identity of the person, if by that is meant, the con- 
 sciousness of the identity of one's own substance as a 
 thinking being in all changes of its state. No mere 
 analysis of the proposition, " Ithi?ik,'' can prove identity 
 
 409 in this latter sense ; for that we should require synthetic 
 judgments derived from an actual perception of the 
 self. 
 
 (4) That I distinguish my own existence as a thinking 
 
TRANSCENDENTAL DIALECTIC. 151 
 
 being from things outside of me, one of which is my own 
 JT body, is also an analytic proposition ; for by other I mean 
 
 other tha?i ?ne, or distinct from me. But this does not 
 enable me to know whether I could be conscious of 
 myself at all, were things not given to me in perception 
 as outside of me, and whether I could exist merely as a 
 thinking being without being also a sensuous being. 
 
 The analysis, then, of my consciousness of self as the 
 subject that thinks, does not enable me to take a single 
 step towards the knowledge of myself as an object. To 
 suppose so is simply to confuse the logical analysis of 
 thinking in general with the metaphysical determination 
 of an object. 
 
 The truth is, that it would be a great stumbling-block, 
 and, indeed, the only thing that our Critique could have 
 reason to fear, if it could be shown a priori^ that all 
 thinking beings are in themselves simple substances; 
 that, as a necessary consequence, personality is insepar- 
 able from them ; and that they are conscious of their 
 own existence as separate and distinct from all matter. 
 Were it possible in this way to take a step beyond the 
 410 world of sense, and to enter the world of noumena, whc 
 should then deny to us the right to go forward in this 
 new region, to settle in it, and, if we were under a lucky 
 star, to take complete possession of it ? For, the 
 proposition, that every thinking being is by its very 
 nature a simple substance, is an a priori synthetic pro- 
 position ; firstly, because it goes beyond the conception 
 with which it starts, and adds to the act of thinking in 
 general the mode of existence ; and, secondly, because it 
 adds to that conception the new predicate of simplicity, 
 which cannot be given in any experience. Hence a priori 
 
152 THE PHILOSOPHY OF KANT. 
 
 synthetic propositions would be possible and admissible, 
 not simply, as we have contended, in relation to objects 
 of a possible experience, and indeed as principles of the 
 possibility of experience, but even as determinations of 
 things in general and of things in themselves. This 
 would make an end of our whole Critique, and bring 
 us back to the old dogmatism. The danger, however, is 
 not so great as it seems, as may be seen when we look at 
 the matter more closely. 
 
 The whole procedure of rational psychology is vitiated 
 by a paralogism, which may be exhibited in the following 
 syllogism : — 
 
 That which can be thought only as subject, must 
 exist as subject, and is therefore substance. 
 
 A thinking being from its very nature can be thought 
 only as subject. 
 
 Therefore, a thinking being can exist only as subject, 
 that is, as substance. 
 
 Now, in the major premise of this syllogism, by "that 
 which can be thought" is meant a being in every 
 relation in which it can be thought, and therefore in 
 relation to possible perception. But, in the minor 
 premise, the only being spoken of is a "thinking being," 
 or one that is conscious of itself as subject, simply from 
 its relation to thought and to the unity of consciousness, 
 but not at all from its relation to a perception by which it 
 is presented to thought as an object. The conclusion is, 
 therefore, reached /(?r sophLwia figurae didionis. 
 412 That we are perfectly right in resolving this famous 
 argument into a paralogism will be at once evident, if we 
 call to mind what has already been pointed out. The 
 conception of a thing that can exist by itself as a subject, 
 
. TRANSCENDENTAL DIALECTIC. 153 
 
 but cannot exist as a mere predicate, does not carry with 
 it objective reality. We cannot possibly know that there 
 is an object corresponding to the conception, because we 
 cannot understand how an object of that sort could 
 exist at all. If by the term " substance " is meant an 
 object that can be presented to us, we must say that the 
 indispensable condition of the objective reality of our 
 conception is, that it should be presented to us in a 
 
 413 permanent perception. Now, in inner perception there 
 is nothing permanent, for the / is merely the conscious- 
 ness of my thinking. So long, therefore, as we limit 
 ourselves to mere thinking, we are without the necessary 
 condition for the application of the conception of 
 substance to the self as a thinking being ; we are unable, 
 in other words, to say that the self is an independent 
 subject. And along with the objective reality of the 
 conception of substance completely disappears the sim- 
 pHcity of substance, leaving only the logical quahtative 
 unity of self-consciousness in thinking in general, a unity 
 which exists whether the subject is composite or simple. 
 
 A 2 1 Rational psychology is, therefore, not a doctrine which 
 enables us to add anything to our knowledge of self; it 
 is merely a discipline^ which sets impassable limits to 
 speculative reason in this field, and prevents us, on the 
 one hand, from throwing ourselves into the arms of a 
 soulless materialism, and on the other hand, from giving 
 ourselves up to a mystic spiritualism that has lost its 
 hold of actual Hfe. The refusal of reason to answer our 
 curious questions as to a life beyond the present, we 
 ought to interpret as a hint to apply our self-knowledge 
 to fruitful practical ends, and to turn away from fruitless 
 and transcendent speculations. 
 
154 THE PHILOSOPHY OF KANT. 
 
 The claim of rational psychology to take rank as a 
 science, rests upon a mere misunderstanding. The unity 
 of consciousness, which is the supreme condition of the 
 422 categories, is simply confused with a perception of the 
 subject as object, and hence we suppose that we may 
 apply to the subject the category of substance. But the 
 unity of consciousness is merely the unity implied in all 
 thinking, and by means of this unity no object is given, 
 nor can the category of substance, which always presup- 
 poses a given perception, be applied to it. There is 
 therefore no knowledge of the subject as an object. The 
 subject no doubt thinks the categories, but that is no 
 reason for saying that it can have a conception of itself 
 as an object of the categories. It cannot think the 
 categories without presupposing its own pure self-con- 
 sciousness, and therefore self-consciousness cannot be 
 brought under the categories. If the subject, in which 
 the consciousness of time has its origin, cannot determine 
 by means of that consciousness its own existence in time, 
 no more can it determine itself as a mere thinking being 
 by means of the categories. 
 
 426 The result of our investigation, then, is, that the dialecti- 
 cal illusion in rational psychology arises from the confu- 
 sion of an idea of reason — the idea of a pure intelligence 
 — with the perfectly undetermined conception of a 
 thinking being in general. Abstracting from all actual 
 experience, I first think of myself as the subject of a 
 possible experience, and then I infer that I can be 
 conscious of my own existence even apart from experi- 
 
 427 ence and the empirical conditions of experience. But 
 this is to confuse the possible abstraction of my own 
 existence as empirically determined, with the conscious- 
 
TRANSCENDENTAL DIALECTIC. 155 
 
 ness of a possible separate existence of my thinking self 
 Thus arises the belief, that I have an actual knowledge 
 of what is substantial in me as a transcendental subject, 
 when in truth I have in my thought merely the unity of 
 consciousness as the form of knowledge that is presup- 
 posed in all determination of objects. 
 
 432 Chapter II. — The Antinomy of Pure Reason. 405 
 
 433 The second class of dialectical arguments, in analogy 406 
 with the hypothetical syllogism, has for its content the 
 unconditioned unity of objective conditions in the phen 
 omenal world. The transcendental paralogism produced 
 merely a one-sided illusion, in regard to the idea of the 
 subject of our thought ; nor is there, in that connection, 
 anything whatever in the conceptions of reason to 
 suggest that the opposite may be true. It is quite other- 
 wise with the objective synthesis of phenomena, where 407 
 reason thinks to establish its principle of unconditioned 
 unity with the greatest ease, until it finds, as it soon does, 
 that in trying to do so it becomes involved in contradic- 
 tions, which force it to give up all pretensions to a 
 rational cosmology. This is a new experience for human 
 reason, for here it falls of itself into a perfectly natural and 
 
 434 unavoidable Antithetic, which is not due to artificial 
 refinements or logical tricks. 
 
 All those transcendental ideas which relate to absolute 
 totality in the synthesis of phenomena, I shall call 
 cosmical conceptions. I call them cosmical, partly because 408 
 the conception of the world as a whole, which is itself 
 only an idea, rests upon that unconditioned totality, and 
 partly because they are concerned only with the synthesis 
 
156 THE PHILOSOPHY OF KANT. 
 
 of phenomena, and therefore with objects of experience. 
 435 On the other hand, absolute totality in the synthesis of 
 the conditions of all possible things gives rise to an ideal 
 of pure reason, and this idea, although it is no doubt 
 related to the conception of the world as a whole, is yet 
 quite distinct from it. Just as the paralogisms of pure 
 reason were the source of a dialectical psychology, so the 
 antinomy of pure reason will set before our eyes the 
 transcendental principles on which a pure or rational 
 cosmology is supposed to rest. 
 
 Section I. — System of Cosmological Ideas. 
 
 It must be observed, firstly, that reason does not of 409 
 itself give rise to any conception, but simply seeks to 
 free a cojiception of understanding from the unavoidable 
 limitation of a possible experience and to extend it 
 beyond the limits of experience, though still without 
 436 losing its connection with experience. Demanding 
 absolute totality on the side of the conditions, it converts 
 the category into a transcendental idea, and tries to give 
 absolute completeness to the empirical synthesis, by 
 carrying it up to the unconditioned. The principle by 
 which reason is here guided, is, that if the condi- 
 tioned is given, the whole sum of conditio7is required to 
 acco7int for the conditioned, and therefore the absolutely 
 uncoftditioned, is likewise given. But, secondly, the only 
 categories which can be so employed, are those which in 
 their synthesis constitute a series of conditions subordin- 
 ated to one another, not those in which the conditions are 
 co-ordinate. This synthesis, as starting from the side of 
 the conditions, and going back step by step to the more 
 
TRANSCENDENTAL DIALECTIC. 157 
 
 remote conditions may be called 7'egressive, to distinguish 
 it from a progressive synthesis, which would start from 
 the nearest consequent on the side of the conditioned 
 and gradually advance to more remote consequents. 
 The former proceeds i?i antecedentia, the latter in con- 
 sequentia. 
 442 When we have rejected the categories which do not 415 
 conform to these requirements, we find that there are but 
 four cosmological ideas, corresponding to the four titles 
 of the categories, that necessarily imply a series in the 
 synthesis of phenomena. 
 
 I. Absolute completeness in 
 
 the composition 
 
 of the given whole of all phenomena. 
 
 2. Absolute completeness 3. Absolute completeness 
 in the divisio?i in the origination 
 
 of a given whole in the world of a phenomenon as such, 
 of phenomena. 
 
 4. Absolute completeness, 
 as regards dependence of existence, 
 of the changeable in the phenomenal world. 416 
 
 448 Section II. — Antithetic of Pure Reasoji. 420 
 
 By the term Antithetic we may denote, not the dog- 
 matic assertion of the opposite of a thesis, but the 
 conflict between two propositions, each of which seems 
 to be true, but neither of which has any more claim to 
 our assent than the other. When we are not content to 42 1 
 
 449 apply our reason to objects of experience, and in subor- 
 dination to the principles of understanding, but venture 
 
158 
 
 THE PHILOSOPHY OF KANT. 
 
 to go beyond the limits of experiencej there arise certain 
 pseudo-rational propositions, which experience can nei- 
 ther confirm nor overthrow. Each of these propositions 
 is not only in itself free from contradiction, but it can 
 appeal to the very nature of reason in support of its 
 truth, although, unfortunately, the opposite proposition 
 can make out just as good a claim to be regarded as 
 necessarily true. 
 453n The antinomies follow in the order of the transcenden- 425n 
 tal ideas as given above. 
 
 454-5 
 
 The Antinomy of Pure Reason. 
 First Conflict of the Transcendefital Ideas. 
 
 426-7 
 
 thesis. 
 
 The world has a beginning 
 in time, and is enclosed within 
 limits of space. 
 
 Proof. 
 
 Assume that the world has 
 no beginning in time. Then, 
 up to every given point of time 
 an eternity must have elapsed, 
 and hence an infinite series of 
 states of things must have 
 passed away one after the 
 other, and come to an end in 
 the world. Now, the infinity 
 of a series just consists in 
 this, that the series can never 
 be completed in a successive 
 
 antithesis. 
 
 The world has no begin- 
 ning in time, and no limits 
 in space, but is infinite as 
 regards both time and space. 
 
 Proof 
 
 Assume that the world has 
 a beginning. Then, as no- 
 thing can begin to be which 
 has not been preceded by a 
 time in which the thing 
 that begins was not, we must 
 hold that there was a time 
 antecedent to that in which 
 the world began to be, that 
 is, an empty time. But, no- 
 thing whatever can come in- 
 to being in an empty time, 
 
TRANSCENDENTAL DIALECTIC. 
 
 159 
 
 synthesis. Hence an infinite 
 series of states cannot have 
 passed away in the world, 
 and therefore a beginning of 
 the world is a necessary con- 
 dition of its existence. This 
 was the first thing to be 
 proved. 
 
 As to the second point, 
 again assume the opposite. 
 Then, the world must present 
 itself to us as an infinite 
 whole of coexistent things. 
 Now, if a magnitude is not 
 presented in a perception as 
 within certain limits, there is 
 no other way in which we can 
 think its dimensions, than by 
 the synthesis of its parts ; 
 and the magnitude as a whole 
 we can think only by the re- 
 peated addition of unity to 
 itself until the synthesis is 
 complete. Hence, in order 
 to think the world, which fills 
 all space, as a whole, we 
 must suppose the successive 
 synthesis of the parts of an 
 infinite world to have been 
 completed. But this is the 
 same as saying that an infin- 
 ite time must have elapsed 
 during the summation of 
 the totality of coexisting 
 things. Now this is im- 
 possible. Hence an infinite 
 
 for no part of an empty time 
 has in it any condition of ex- 
 istence rather than of non-ex- 
 istence, which distinguishes 
 it from any other part ; and 
 this is true, whether we sup- 
 pose things to originate of 
 themselves, or to be pro- 
 duced by some other cause. 
 Hence, although many series 
 of things may begin in the 
 world, the world itself can 
 have no beginning, and is 
 therefore infinite as regards 
 time. 
 
 As to the second point, 
 let us begin by assuming the 
 opposite, namely, that the 
 world is finite and limited as 
 to space. Then, the world 
 must exist in an empty space 
 which has no limits. Things 
 must therefore not only be 
 related 171 space, but they 
 must also be related to space. 
 But the world is an absolute 
 whole, outside of which no 
 object of perception, and, 
 therefore, no correlate of the 
 world, can be found. The re- 
 lation of the world to empty 
 space would therefore be 
 the relation of it to no ob- 
 ject. But such a relation, 
 and therefore the limitation 
 of the world by empty space. 
 
i6o 
 
 THE PHILOSOPHY OF KANT. 
 
 aggregate of actual things 
 cannot present itself to us 
 as a whole, and therefore not 
 as a whole all the parts of 
 which coexist. The world 
 is therefore not infinitely ex- 
 tended in space, but is en- 
 closed within spatial limits. 
 And this was the second 
 thing to be proved. 
 
 is nothing at all. Hence the 
 world cannot be limited as 
 regards space, or, the world 
 is infinite in its extension. 
 
 462-: 
 
 Second Co7ifiict of the Transcendental Ideas. 
 
 THESIS. 
 
 Every composite substance 
 in the world is made up of 
 simple parts, and nothing 
 whatever exists but the 
 simple, or that which is 
 composed out of the simple. 
 
 Proof. 
 
 Assume that composite 
 substances are not made 
 up of simple parts. Then, 
 if we think all composition 
 to be away, no composite 
 part will be left. And, by 
 hypothesis, there is no 
 simple part. Hence, no- 
 thing at all will remain, 
 and therefore no substance. 
 Either, then, it is impos- 
 sible to think all composi- 
 tion to be away, or even 
 after composition is thought 
 
 ANTITHESIS. 
 
 No composite thing in the 
 world is made up of simple 
 parts, nor does anything 
 simple exist anywhere in the 
 world. 
 
 Proof 
 
 Assume that a composite 
 thing or substance is made 
 up of simple parts. Then, 
 as no external relation, and 
 therefore no composition out 
 of substances, is possible ex- 
 cept in space, the composite 
 thing must be made up of 
 exactly the same number of 
 parts as the space which it 
 occupies. Now, space is 
 not made up from simple 
 parts, but consists of spaces. 
 Every part of the composite 
 
TRANSCENDENTAL DIALECTIC. 
 
 l6l 
 
 to be away, there must be 
 something left, which exists 
 without composition, that is, 
 the simple. In the former 
 case, the composite cannot 
 be made up of substances, 
 for composition is merely an 
 accidental relation of sub- 
 stances, which may be taken 
 away without at all affect- 
 ing their existence as per- 
 manent realities. But, by 
 hypothesis, substances do 
 exist, and hence we must 
 adopt the other supposition, 
 that the composite substan- 
 ces in the world consist of 
 simple parts. 
 
 It directly follows, that all 
 the things in the world are 
 simple ; that composition is 
 merely an external state of 
 those things ; and that, al- 
 though we can never take 
 elementary substances out of 
 their state of composition and 
 isolate them, reason must 
 think of them as the prim- 
 ary subjects, which exist as 
 simple beings antecedently 
 to all composition. 
 
 thing must therefore occupy 
 a space. But the absol- 
 utely primary parts of every 
 composite thing are simple. 
 Hence each of those simple 
 parts occupies a space. Now, 
 as every real thing, which 
 occupies a space, contains 
 within itself a number of 
 parts that are outside of one 
 another, and is therefore 
 composite ; and as this real 
 composite thing is not made 
 up of accidents, since these 
 could not, apart from sub- 
 stance, be outside of one 
 another ; we must conclude, 
 that simple substance is com- 
 posite, which is absurd. 
 
 The second proposition of 
 the antithesis, that nowhere 
 in the world does there exist 
 anything simple, is only in- 
 tended to mean, that the 
 existence of the absolutely 
 simple cannot be shown from 
 any experience or perception, 
 external or internal ; and that, 
 as the absolutely simple is 
 therefore a mere idea, the 
 objective reality of which 
 can never be presented in 
 experience, it is without all 
 application and object in the 
 explanation of phenomena. 
 Let it be even admitted that 
 
l62 THE PHILOSOPHY OF KANT. 
 
 an object might be found in 
 experience corresponding to 
 that idea, or, in other words, 
 that we might have an em- 
 pirical perception of an ob- 
 ject which contained no parts 
 that are outside of one an- 
 other and combined to a 
 unity. Yet we could not 
 legitimately infer the impos- 
 sibility of finding any differ- 
 ence of parts in the object 
 from the fact that we are not 
 conscious of such difference. 
 But nothing less than this 
 will establish absolute sim- 
 plicity, and hence absolute 
 simplicity cannot be inferred 
 from any perception, no mat- 
 ter what its nature may be. 
 As, therefore, an absolutely 
 simple object can never be 
 presented in any possible ex- 
 perience, and as the world of 
 sense must be regarded as 
 the sum total of all possible 
 experience, it follows that 
 there is nothing in the world 
 that is absolutely simple. 
 
 472-3 Third Conflict of the TranscendeJital Ideas. 444-5 
 
 THESIS. ANTITHESIS. 
 
 Causality in conformity There is no freedom, but 
 
 with laws of nature is not all that com.es to be in the 
 the only causality, from which world takes place entirely 
 
 1 
 
TRANSCENDENTAL DIALECTIC. 
 
 163 
 
 all the phenomena of the 
 world can be derived. To 
 explain those phenomena it 
 is necessary to suppose that 
 there is also a free causality. 
 Proof. 
 Assume that the only caus- 
 ality is that in conformity 
 with laws of nature. Then, 
 all that comes to be presup- 
 poses an antecedent state up- 
 on which it follows according 
 to an inviolable rule. Now, 
 that antecedent state must 
 itself be something that 
 comes to be, or arises in 
 a time in which it pre- 
 viously was not ; for if it 
 had always existed, its effect 
 also must always have ex- 
 isted, and would not have 
 just come to be. The caus- 
 ality of the cause through 
 which something comes to 
 be must therefore itself be an 
 event, which again, accord- 
 ing to the law of nature, 
 presupposes an antecedent 
 state and its causality, and 
 this again a still earlier 
 state, and so on. If there- 
 fore all that comes to be 
 must conform to the law of 
 nature, there is never an 
 absolute beginning, but only 
 a relative beginning", and 
 
 in accordance with laws of 
 nature. 
 
 Proof. 
 Assume that there isfree- 
 do7n, in the transcendental 
 sense, as a special kind of 
 causality by which the se- 
 quence of events in the world 
 may be explained ; in other 
 words, that there is a faculty 
 of absolutely bringing into 
 existence a certain state, and 
 therefore a series of conse- 
 quents of that state. Then, 
 not only must this spontane- 
 ity originate the series, but it 
 must first determine itself to 
 originate it, and its act must 
 take place without any ante- 
 cedent to determine it in 
 accordance with fixed laws. 
 But every beginning of an 
 act presupposes a state in 
 which the cause has not yet 
 begun to act, and a dynami- 
 cally first beginning of the 
 act presupposes a state of 
 that cause which has no 
 causal connection with the 
 preceding state, and in no 
 way follows from it. Tran- 
 scendental freedom is there- 
 fore opposed to the law of 
 
164 
 
 THE PHILOSOPHY OF KANT. 
 
 hence there can be no com- 
 pleteness in the ascending 
 series of causes. Now, the 
 law of nature just consists 
 in this, that nothing can 
 come to be without a cause 
 sufficient to determine it a 
 priori. The proposition, that 
 all causality is possible merely 
 by laws of nature, is therefore 
 self-contradictory, if it is 
 taken in its unlimited uni- 
 versality ; and hence that 
 sort of causality cannot be 
 the only one. 
 
 We must, then, admit that 
 there is another sort of caus- 
 ality, a causality by means 
 of which something may 
 come to be, the cause of 
 which is not itself deter- 
 mined according to neces- 
 sary laws by another cause 
 antecedent to it. This will 
 be an absolutely spontajieous 
 causality, bringing into exist- 
 ence by itself 2i series of phe- 
 nomena which arise in con- 
 formity with laws of nature. 
 Hence, without transcenden- 
 tal freedom it is impossible 
 ever to have completeness, 
 on the side of causes, even 
 in the series of phenomena 
 which follow one another in 
 the course of nature. 
 
 causality, and demands such 
 a connection of the succes- 
 sive states of efficient causes 
 as makes the unity of ex- 
 perience impossible. As it 
 cannot be found in any ex- 
 perience, it is a mere idea 
 without any content. In 
 nature, therefore, and not 
 in freedom, we must seek 
 for the order and connection 
 of all events that occur in 
 the world. 
 
TRANSCENDENTAL DIALECTIC. 
 
 165 
 
 480-1 
 
 Fourth Conflict of the Transcendental Ideas. 
 
 452-3 
 
 THESIS. 
 
 There exists an absolutely- 
 necessary being, which be- 
 longs to the world either as 
 a part or as the cause of it. 
 
 Proof. 
 The world of sense is not 
 simply the sum total of all 
 phenomena, but it contains a 
 series of changes. Were 
 there no such changes, we 
 should have no conscious- 
 ness even of a series of time 
 as a condition of the possi- 
 bility of the world of sense. 
 But every change stands un- 
 der a condition, which pre- 
 cedes it in time, and makes 
 it necessary. Now, every- 
 thing that is presented as con- 
 ditioned, presupposes for its 
 existence a complete series of 
 conditions, ending in the per- 
 fectly unconditioned, which 
 alone is absolutely necessary. 
 Something absolutely neces- 
 sary must . therefore exist, if 
 there exist a change as its con- 
 sequence. And this neces- 
 sary existence must itself 
 belong to the world of sense. 
 For if it were outside that 
 world, we should have to say, 
 
 ANTITHESIS. 
 
 There nowhere exists an 
 absolutely necessary being, 
 either in the world, or out- 
 side of the world as its cause. 
 
 Proof 
 Assume that the world it- 
 self is a necessary being, or 
 that a necessary being exists 
 in it. Then, either there is 
 a beginning in the series of 
 its changes that is absolutely 
 necessary, and therefore with- 
 out a cause, or the series it- 
 self, having no beginning, is 
 as a whole absolutely ne- 
 cessary and unconditioned, 
 though it is contingent and 
 conditioned in all its parts. 
 Now, the former supposition 
 contradicts the dynamical 
 law of the determination of 
 all phenomena in time ; and 
 the latter supposition contra- 
 dicts itself, because the ex- 
 istence of a series cannot be 
 necessary as a whole, if no 
 single member of the series 
 is necessary. 
 
 Assume, on the other hand, 
 that there is an absolutely 
 necessary cause of the world, 
 which is outside of the world. 
 
i66 
 
 THE PHILOSOPHY OF KANT. 
 
 that the series of changes in 
 the world derived their be- 
 ginning from a necessary 
 cause which did not itself 
 belong to the world of sense. 
 Now, this is impossible. For, 
 as the beginning of a series 
 in time can be determined 
 by that only which is in a 
 time antecedent to the series, 
 the highest condition of the 
 beginning of a series of 
 changes must exist in a time 
 when the series as yet was 
 not. Hence the causality of 
 the necessary cause of the 
 changes, and therefore also 
 the cause itself, must belong 
 to time and to phenomena in 
 time, and cannot be thought 
 as separated from that sum 
 total of all phenomena which 
 constitutes the world of sense. 
 Something absolutely neces- 
 sary is therefore contained in 
 the world itself, whether that 
 something is the whole series 
 of changes in the world or a 
 part of that series. 
 
 5^4 Section IV. — Necessity of a Solutio7i of the Transcendental ^1^ 
 
 Problems of Pure Reason. 
 
 505 Transcendental philosophy cannot admit, that any 477 
 question which concerns an object presented to the pure 
 reason of man is unanswerable by the reason that 
 
 Then, to that cause, as the 
 highest member in the series 
 of the causes of changes in 
 the world, would originally 
 be due the beginning of the 
 existence of those changes 
 as a series. But the cause 
 must itself begin to act, and 
 its causality would therefore 
 belong to time, and so to 
 the sum total of phenomena ; 
 or, in other words, that cause, 
 as belonging to the world, 
 would not itself be outside 
 of the world. But this is 
 contrary to our hypothesis. 
 Hence, neither in the world, 
 nor as a cause outside of the 
 world, though in causal con- 
 nection with it, does there 
 exist any absolutely neces- 
 sary being. 
 
TRANSCENDENTAL DIALECTIC. 1 67 
 
 suggested it It is vain to allege our unavoidable 
 ignorance and the unfathomable depth of the problem, 
 as a reason for avoiding the obligation of giving a 
 thorough and complete answer. The very conception 
 which enables us to ask the question, must also give 
 us the means of answering it, because the object to 
 which it refers has no existence except in the conception. 
 
 506 It is, however, only in connection with the cosmo- 478 
 logical ideas that questions arise in transcendental 
 philosophy, which put upon us the obligation to answer 
 them. For here the object must be presented in experi- 
 ence, and the only question is whether it can conform 
 
 to the idea. If the problem, for instance, is whether 
 the soul, as that which presents itself in our conscious- 
 ness as thinking, is in its own nature a simple substance ; 
 or, whether there is an absolutely necessary cause of all 
 things ; the object is transcendental, and therefore itself 
 unknown ; and hence we have to inquire, whether there 
 is any object whatever, corresponding to our idea. In 
 this case, therefore, we may confess that the object is 
 
 507 unknown to us, without saying that it cannot possibly 479 
 exist. Only the cosmological ideas have the peculiarity, 
 that they can presuppose their object and the empirical 
 synthesis essential to the conception of it. The sole 
 question which they raise, is, whether the empirical 
 synthesis can be carried so far as to comprehend an 
 absolute totality of conditions. Now, as there is here 
 
 no question of a thing in itself, but only of a thing as an 
 object of possible experience, the answer to the trans- 
 cendental problem of cosmology cannot be found in 
 anything outside of the idea. We are not asking what is 
 the nature of any object in itself; we are not even asking 
 
1 68 THE PHILOSOPHY OF KANT. 
 
 what can be presented in cojicreto according to the con- 
 ditions of experience ; but our whole question is in regard 
 to what is contained in the idea itself, to which the 
 empirical synthesis will be found merely to approximate, 
 and the answer must be derived entirely from the idea. 
 Reason cannot evade a solution of the problem by putting 
 all the responsibility upon the unknown object, for the 
 idea is a pure creation of reason itself. 
 
 525 Section VII. — Critical Solution of the Cosniological 497 
 
 Proble7n. 
 
 The whole antinomy of pure reason rests upon this 
 dialectical argument : — 
 
 If the conditioned is given, the w^hole series of con- 
 ditions is given. 
 
 But objects of sense are given as conditioned. 
 
 Therefore, the whole series of conditions of objects of 
 sense is given. 
 
 The sophistical character of the argument will be more 
 
 526 readily seen, if we first correct and define some of the 
 conceptions contained in it. 
 
 Now, in the first place, it is plain and undeniable, 
 that, if the conditioned is given, a regress in the series of 498 
 all its conditions is demafided of us. The very conception 
 of the conditioned implies that something is referred to a 
 condition, and, if that condition is itself conditioned, to a 
 more remote condition, and so on through all the members 
 of the series. The proposition, therefore, that if the con- 
 ditioned is given we must seek for the whole series of 
 conditions, is analytical, and can have nothing to fear 
 from a transcendental criticism. 
 
TRANSCENDENTAL DIALECTIC. 1 69 
 
 In the second place, if the conditioned as well as its 
 condition are things in themselves, not only is the regress 
 to the condition demanded, if the conditioned is given, 
 but it is actually given along with the conditioned. And, 
 as this h9lds of all members of the series, the complete 
 series of conditions, and therefore the unconditioned, is 
 given at the same time, or, rather, it is presupposed in 
 virtue of the fact that the conditioned, which is possible 
 only through it, is given. The synthesis of the con- 
 ditioned with its condition is here a mere svnthesis of 
 ^ understanding, which assumes to present things as they 
 527 are, without first asking whether and how we can have 
 a knowledge of them. But, if I have to do with 
 phenomena, which, as existing only for consciousness, 
 are not given at all unless they are empirically known, I 499 
 cannot in the same sense say, that if the conditioned is 
 given, all its conditions are also given, and hence I can 
 in no way infer the absolute totality of the series of 
 conditions. For phenomena in our apprehension are 
 themselves nothing but an empirical synthesis in space 
 and time, and are therefore given only in that synthesis. 
 It does not follow, because the conditioned as a 
 phenomenon is given, that the synthesis which con- 
 stitutes its empirical condition, is given along with or 
 presupposed in it; for the synthesis exists only in the 
 regress and in no sense apart from it. What we can 
 say, in such a case, is, that a regress, or continuous 
 empirical synthesis, on the side of the conditions, is 
 enjoined and dtinanded, and that the conditions given 
 in that regress cannot be wanting. 
 
 It is evident, then, that in the major premise of the 
 cosmological argument, the conditioned is taken in the 
 
170 THE PHILOSOPHY OF KANT. 
 
 transcendental sense of a pure category ; while in the 
 minor premise, it is taken in the empirical sense of a 
 conception of understanding that is applied to mere 
 phenomena. Here therefore we have an instance of 
 
 528 the dialectical fallacy called sophisi7ia figurae dictionis. 500 
 The fallacy, however, is not artificial; it is by a per- 
 fectly natural illusion that reason in the major premise 
 assumes blindly, that if something is given as con- 
 ditioned, its conditions and their series must all be 
 present. In fact the assumption is just the logical 
 postulate, that every conclusion must have complete 
 premises. Moreover, the connection of the conditioned 
 with its condition is naturally thought to be inde- 
 pendent of any succession in time, and both are 
 assumed to be given together. Nor is it less natural, 
 
 in the minor premise, to regard phenomena as things 
 in themselves and as objects given to pure under- 
 standing, than to take the conditioned in the sense of a 
 pure conception in the major premise, where abstraction 
 has been made from all the conditions of perception 
 without which objects cannot be given at all. Yet this 
 overlooks an important distinction between these con- 
 ceptions. The synthesis of the conditioned with its con- 
 dition, and indeed with the whole series of its conditions, 
 as expressed in the major premise, carries with it no 
 limitation through time and no idea of succession. But 
 the empirical synthesis and the series of conditions in 
 phenomena, as subsumed in the minor premise, is 
 necessarily successive, the members of the series being 
 given as following one another in time. Here, therefore, 
 
 529 we cannot presuppose absolute totality of the synthesis 
 and of the series presented in it. In the former case all 
 
TRANSCENDENTAL DIALECTIC. lyi 
 
 the members of the series are given in themselves 
 irrespective of any condition of time ; but in the latter 
 case they are possible only by means of a successive 501 
 synthesis, and can be given as a whole only if that 
 synthesis can actually be completed. If, then, we are 
 to settle the dispute between the two parties to the 
 satisfaction of both, we must be able to show that they 
 are really quarrelling about nothing, and that a certain 
 53° transcendental illusion has mocked them with a reality 502 
 where none is to be found. 
 
 532 If we regard the two propositions {a) that the world is 504 
 infinite in extension, and (b) that the world is finite in 
 extension, as contradictory opposites, we assume that 
 the world, or the whole series of phenomena, is a thing 
 
 in itself. For, whether the regress in the series of its 
 phenomena is denied to be infinite, or denied to be finite, 
 in both cases the world is supposed to be absolutely real. 
 But if I challenge this supposition, or rather this trans- 
 cendental illusion, and deny that the world is a thing in 
 
 533 itself, the contradictory opposition of the two statements 505 
 is converted into a dialectical opposition. As the world 
 does not exist at all as a thing in itself, that is, indepen- 
 dently of the regressive series of my ideas, it cannot be 
 said to be in itself QiXh^r an infinite whole or 2i finite whole. 
 Apart from the empirical regress in the series of pheno- 
 mena, the world has no existence whatever. If, therefore, 
 that series is always conditioned, and as a consequence is 
 never given as complete, the world cannot be an uncon- 
 ditioned whole, and therefore cannot exist as an uncon- 
 ditioned whole that is either infinite in magnitude or 
 finite in mag;nitude. 
 
 What has been said of the first cosmological idea is 
 
172 THE PHILOSOPHY OF KANT. 
 
 equally true of the other three. The series of conditions 
 exists only in the regressive synthesis itself, not in an 
 object of sense given as an independent thing prior to all 
 regress. Hence I must say that the number of parts in 
 a given phenomenon is neither finite nor infinite. A 
 phenomenon has no existence in itself, and its parts are 
 only given in and through the regress of the decomposing 
 synthesis ; and this regress, being never absolutely com- 
 plete, cannot be said to be either finite or infinite. The 
 same thing holds of the series of causes that proceed in 
 an ascending series, and of the series that proceeds from 
 534 conditioned existence to unconditioned necessary exist- 506 
 ence. Neither series can be regarded as in itself either 
 finite or infinite in its totality ; for, as a series of subordin- 
 ated ideas consists only in the dynamical regress itself, 
 it cannot possibly exist in itself before that regress as a 
 self-subsistent series of things in themselves. 
 
 Thus the antinomy of pure reason in its cosmological 
 ideas disappears. It is purely dialectical, or a conflict 
 due to an illusion. The idea of absolute totality, which 
 has no proper meaning except as a condition of things in 
 themselves, is wrongly applied to phenomena, which exist 
 only in our consciousness, and, if they form a series, 
 only in a successive regress, but which have no other 
 existence whatever. From tiiis antinomy, however, we 
 may gain, not indeed a dogmatic, but a critical and 
 doctrinal advantage. It supplies an indirect proof 
 of the transcendental ideality of phenomena, which 
 ought to convince anyone who may not have been quite 
 satisfied with the direct proof in the Transcendental 
 -•Esthetic. The new proof would consist in the following 
 dilemma : — 
 
TRANSCENDENTAL DIALECTIC. 1 73 
 
 If the world is a self-existent whole, it is either finite 
 
 or infinite. 
 But it is neither finite nor infinite (as is shown in the 
 Antithesis and Thesis respectively). 
 535 Therefore the world (the sum total of all phenomena) 507 
 is not a self-existent whole. 
 Phenomena have, therefore, no existence apart from our 
 consciousness of them ; and this is what we mean when 
 we speak of their transcendental ideality. 
 
 C36 Section VIII. — Regulative Principle of Pui-e Reason t^oS 
 
 in the Cosmological Ideas. 
 
 By the cosmological principle of totality, a maximum 
 in the series of conditions of the objects of sense is not 
 given^ but is only demanded. Still, that principle is true 
 if it is taken in the proper sense. No doubt it is not an 
 axio?n, requiring us to think totality as actually present in 
 the object, but it is a problem for understanding, and 
 therefore for the subject of understanding, calling upon 
 him to begin and to follow out the regress in the series of 
 conditions for that which is given as conditioned, in con- 
 formity with the completeness contained in the idea. In 
 the presentation of sensible objects as in space and time, 
 every condition which we are capable of reaching is 
 found to be itself conditioned. If phenomena were 
 things in themselves, we might perhaps find in them 
 something unconditioned ; but, as a matter of fact, they 
 are merely empirical objects, and as such can appear only 
 under the forms of space and time, the condition of all 
 our perceptions. The principle of reason is therefore 
 537 merely a riile^ which demands a regress in the series of 509 
 
174 THE PHILOSOPHY OF KANT. 
 
 conditions of given phenomena, and will not permit us 
 to assume that we have reached anything absolutely uncon- 
 ditioned. It is nota principleofthepossibility of experience 
 and of the empirical knowledge of objects of sense, and 
 hence it cannot be ranked with the fundamental judgments 
 of understanding ; for every experience is of an object, 
 which, as conforming to the conditions of perception, is 
 enclosed within limits. Xor is it a constitutive principle 
 of reason, which enlarges our conception of the world 
 of sense beyond all possible experience, but only a 
 principle that tells us to continue and enlarge our 
 experience as far as we possibly can. Refusing to admit 
 that any given empirical limit is absolute, the principle 
 of reason serves as a rtde which postulates what must 
 take place, if we make the regress, but does not anticipate 
 what is present, before any regress is made, in the object 
 as it is in itself. I call it, therefore, a regulative principle 
 of reason, to indicate that it is not a constitutive 
 cosmological principle, that is, a principle that determines 
 objects of sense as things in themselves having an 
 absolute totality in the series of their conditions. That 
 there is no such constitutive principle, I indicate by 
 calling the principle of reason regulative, and in this way 
 trying to prevent what otherwise would be inevitable, the 
 transcendental subreption which attributes objective 
 reality to an idea that serves merely as a rule. 
 
 543 Section IX. — Empirical use of the Regulative Principle r i c 
 
 of Reason. 
 
 We have seen that no transcendental use can be made 
 of pure conceptions, whether these belong to under- 
 
TRANSCENDENTAL DIALECTIC. 1 75 
 
 Standing or to reason ; that absolute totality in the series 
 of conditions in the world of sense rests entirely upon a 
 transcendental use of reason, in which absolute complete- 
 
 544 ness is demanded from that which is presupposed as a 
 thing in itself; and that such completeness cannot be 
 found in the world of sense. It is therefore vain to ask 516 
 whether the series of conditions is in itself absolutely 
 limited or absolutely unlimited; the only question is, 
 how far we ought to go back in our empirical regress in 
 search of the conditions of experience, in order that, 
 guided by the rule of reason, we may find an answer 
 which is conformable to the nature of the object in 
 question. 
 
 Now, it has been clearly enough shown that the 
 principle of reason is not a constitutive principle of 
 objects in themselves, but is merely a rule for the co7i~ 
 tinuation and extension of a possible experience. If we 
 keep this steadily before our eyes, the conflict of reason 
 with itself is at an end. For our critical solution not 
 only does away with the illusion in which the contra- 
 diction has its origin, but it reveals the true sense in 
 which reason is in harmony with itself Thus the mis- 
 apprehension which was the sole cause of the conflict has 
 been removed, and a dialectical principle has been con- 
 verted into a doctrinal principle. 
 
 545 I. Solution of the First Antinomy. 517 
 
 546 For the solution of the first cosmological problem, we 518 
 have simply to determine, whether, in the regress to the 
 unconditioned extension of the world in time and space, 
 there is a regress to infinity^ or merely a regress that is 
 capable of being continued indefinitely {in indefnitum). 
 
176 THE PHILOSOPHY OF KANT. 
 
 The perfectly general idea of the series of all past 
 states of the world, as well as of all the things which co- 
 exist in space, is merely the thought of a possible 
 empirical regress, the extent of which has not yet been 
 determined. Only through this idea can there arise the 
 conception of such a series of conditions of a given 
 
 547 perception. Now, the world as a whole exists for me 
 only as a conception, never as a perception. Hence I 519 
 cannot reason from the quantity of the world to the 
 quantity of the regress, and determine the latter by the 
 former ; on the contrary, I can form a conception of the 
 quantity of the world only by finding out the quantity of 
 the empirical regress. Of the empirical regress, how- 
 ever, I can never say more than that I must always 
 advance empirically from every given member of the 
 series of conditions to a higher and more remote member. 
 But in this way the quantity of phenomena as a whole 
 cannot be absolutely determined, and hence I cannot say 
 that the regress proceeds to infinity. To say that it 
 proceeds to infinity would be to anticipate members of 
 the regress that have not yet been reached, and to 
 represent their number as so great that no empirical 
 regress could ever reach them ; it would in fact be to 
 determine the quantity of the world (although only 
 
 C48 negatively) prior to the regress, which is impossible. 520 
 The first or negative answer to the first cosmological 
 problem therefore is, that the world has no first 
 beginning in time, and no extreme limit in space. 
 
 540 The affirmative answer directly follows, that the regress 521 
 in the series of phenomena as a determination of the 
 quantity of the world proceeds in i7idefi?titum. This is 
 the same as saying that the world of sense has no 
 
 I 
 
TRANSCENDENTAL DIALECTIC 177 
 
 550 absolute quantity. Every beginning is in time, and every 522 
 limit of that which is extended is in space. But space 
 and time belong only to the world of sense. Hence, 
 while phenomena in the woi'ld are conditionally limited, 
 the world itself is neither conditionally nor uncon- 
 ditionally limited. Similarly, as the world cannot be 
 given as complete^ and as even the series of conditions for 
 that which is given as conditioned cannot be given as 
 complete^ the conception of the quantity of the world is 
 
 551 given only in the regress, and not in a collective percep- 5^3 
 tion prior to it. But that regress consists simply in the 
 act of determining the quantity, not in a deter?ninate con- 
 ception, and hence it does not yield the conception of a 
 quantity that could be called infinite when measured by a 
 certain standard. The regress, therefore, does not pro- 
 ceed to the infinite, as if the infinite could be presented, 
 but only to an indefinite distance, and it is only in the 
 regress that any quantity of experience is actually given. 
 
 2. Solution of the Second Antino?fiy. 
 
 If I divide a whole that is presented to me in a per- 
 ception, I proceed from something conditioned to the 
 conditions which make it possible. The division into 
 parts {subdivisio or decompositid) is a regress in the series 
 of those conditions. The series could be presented as 
 an absolute totality, only if the regress could finally reach 
 perfectly simple parts. But if all the parts in a con- 
 tinuously progressive decomposition are themselves again 
 divisible, the division or regress from the conditioned to 
 its conditions proceeds in infinitum; for all the parts or 
 conditions, being contained in the conditioned itself, 
 
 M 
 
178 THE PHILOSOPHY OF KANT. 
 
 552 which is completely presented in a perception that is 524 
 enclosed between its limits, are presented along with the 
 conditioned. The regress in this case, therefore, cannot 
 
 be called merely a regress i7i indefinittun^ whereas in the 
 first cosmological idea that was the only kind of regress 
 that could be allowed, inasmuch as it was necessary to 
 proceed from the conditioned to conditions outside of it, 
 which were not presented along with the conditioned, 
 but were added to it only in the empirical regress. At 
 the same time, it is not permissible to say of a whole 
 which is divisible to infinity, that it is made up of an 
 iirfifiite iiumber of parts. For, although all the parts are 
 contained in the perception of the whole, the whole 
 divisiofi is not so contained, but it consists only in the 
 continuous decomposition or regress itself, and the series 
 has no existence at all prior to the regress. As this 
 regress is infinite, the members or parts reached in it are 
 certainly all contained in the given whole viewed as an 
 aggregate. But the whole series of the stages in divisio7i 
 is not contained in the given whole ; for these constitute 
 a successive infinite, which is never complete, and there- 
 fore never reaches an infinite multitude of parts, nor can 
 its parts be combined into a whole. 
 
 This general statement may easily be applied to 
 space. Every space perceived within its limits is a 
 whole, the parts of which, as obtained by decomposition, 
 are always themselves spaces. A space is, therefore, 5^5 
 infinitely divisible. 
 
 553 From this a second application of the statement 
 follows quite naturally. The divisibility of an external 
 object or body, which is enclosed within its limits, 
 depends upon the divisibility of the space that is the 
 
TRANSCENDENTAL DIALECTIC. 179 
 
 condition of the perception of the body as an extended 
 whole. The body is, therefore, infinitely divisible, with- 
 out, however, being made up of an infinite number of 
 parts. 
 
 556 Transition from the Mathematical to the Dynatnical 528 
 
 Antinomies. 
 
 So far we have assumed that the conditions belonging 
 to the conditioned are themselves in space and time. 
 Now this assumption, which is always made by common 
 sense, was the real source of the apparent conflict of 
 reason with itself For it forced us to hold that all 
 dialectical conceptions in regard to totality in the series 
 of conditions for that which is given as conditioned were 
 of exactly the same character. The condition was, 
 therefore, in all cases connected with the conditioned as 
 a member of the same series, and was homogeneous with 
 it. Hence the regress in the series of conditions was 
 never thought as complete, or, if it was thought as 
 complete, a member of the series, which was really 
 conditioned, was falsely supposed to be the first member 
 of the series, and, therefore, to be unconditioned. If the 
 
 557 object, or conditioned, was not considered merely 529 
 according to its magnitude, at least the series of con- 
 ditions of that object was so considered. Thus arose a 
 difficulty, which could be got rid of in no other way 
 than by cutting the knot, that is, by recognizing that 
 reason made the series either too long or too short for 
 understanding, so that understanding could never be 
 coincident with the idea of reason. 
 
 But, in all this we have been overlooking an essential 
 distinction that obtains between the objects, that is, the 
 
rSo THE PHILOSOPHY OF KANT. 
 
 conceptions of understanding, which reason tries to raise 
 to ideas. Two of the classes of categories contained in 
 the table given above imply a ??iathe?fiatical, the other two 
 a dyimmical synthesis of phenomena. That distinction 
 becomes important, now that we have come to consider 
 how far the dynamical conceptions of understanding are 
 adequate to the idea of reason, and opens up an entirely 
 new mode of escape from the suit in which reason is 
 involved. In the former suit the case was dis?nissed, 
 because both parties raised a false issue. But in the 
 558 dynamical antinomies, it seems as if reason might be able 53° 
 to establish its claims, for the judge has himself supplied 
 the proofs which were wanting, and which had been 
 overlooked by both parties. It is, therefore, possible 
 that the suit may be adjusted to the satisfaction of both 
 sides, a thing that was impossible in the case of the 
 mathematical antinomies. 
 
 The conditions even in the dynamical ideas are no 
 doubt all homogeneous, in so far as we look merely at 
 the extension of the series, and ask whether it conforms 
 to the idea, or whether the idea is too large or too small 
 for it. But the conception of understanding, on which 
 the idea rests, may contain merely a synthesis of the homo- 
 geneous — which is certainly the case in the composition 
 or division of every magnitude — or it may contain also a 
 synthesis of the heterogeneous. This latter sort of syn- 
 thesis is at least conceivable in the case of the dynamical 
 synthesis, whether it takes the form of causal connection 
 or of the connection of the necessary with the con- 
 tingent. 
 
 As, in the riiathematical connection of the series of 
 phenomena, every condition is itself a part of the 
 
TRANSCENDENTAL DIALECTIC. l8l 
 
 series, no condition can be allowed to enter that is not 
 sensuous. But, in the dynamical series of sensuous con- 
 ditions, a heterogeneous condition, or one that is not a 
 part of the series, is not inadmissible. Such a condition 
 
 559 as being ^Mrtly vitelHgible, would lie outside of the series. 531 
 Supposing it to be possible, the claims of reason would 
 
 be sufficiently satisfied by the unconditioned being placed 
 above phenomena, while yet the series of phenomena 
 would not cease to be conditioned, nor would it be cut 
 short in defiance of the principles of understanding. 
 
 If the dynamical ideas admit of a condition of phe- 
 nomena lying outside of the series of sensuous conditions, 
 a condition which is not itself a phenomenon, we reach 
 quite a different conclusion from that to which we were 
 brought in the case of the mathematical antinomies. In 
 these we were forced to say, that both of the contradic- 
 tory dialectical assertions were false. But, while the 
 dynamical series is necessarily conditioned throughout in 
 so far as it is a series of phenomena, it yet is connected 
 with a condition, which, though it is empirically uncon- 
 ditioned, is non-seiisuous . Thus satisfaction is given, on 
 the one hand to understandings and on the other hand 
 to reason. We are rid of the dialectical arguments, which 
 in one way or the other sought unconditioned totality in 
 mere phenomena, and we see that the propositions of 
 
 560 reason may both he true when taken in their proper sense. 532 
 This we could not possibly show in the case of the cos- 
 mological ideas that refer only to a mathematically un- 
 conditioned unity, for in them no condition of the series 
 
 of phenomena could be found, which was not itself a 
 phenomenon and therefore one of the members of the 
 series. 
 
l82 THE PHILOSOPHY OF KANT. j 
 
 y 
 
 f 'r 
 
 i 
 
 3. Solutio?i of the Third Antinomy. I 
 
 There are only two ways m which we can conceive 
 events to be due to a cause : either the causahty is 
 7iatural^ or it springs from freedovi. By natural causality 
 is meant, that connection of one state with another that 
 precedes it in the world of sense, in which the second 
 state follows the first in conformity with a rule. Now, 
 the causality of phenomena rests upon conditions of time, 
 and the preceding state cannot always have existed, for, 
 if it had, the effect produced by it would not only now 
 have come into being. Hence the causality of the cause 
 of something that happens or comes into being must 
 itself have come into beings and by the principle of under- 
 standing requires another cause to account for it. 
 561 By freedom, again, in the cosmological sense, is meant, 533 
 the power of bringing a state into existence spoiitaneotisly. 
 The causality of this state will therefore not itself stand 
 under another cause, which determines it in time in con- 
 formity with the law of nature. Taken in this sense, 
 freedom is a transcendental idea \ for, in the first place, 
 it contains in it nothing borrowed from experience, and. 
 in the second place, its object cannot be presented as 
 determined in any experience. That whatever comes to 
 be must have a cause, is a universal law, without which 
 there can be no experience at all. As the causality of 
 this cause comes to be or originates, it must itself have a 
 cause. Thus the whole field of experience, however far 
 it may extend, contains nothing but what is natural. \ 
 
 But, as in this way no absolute totality of conditions in 
 the way of causality can be obtained, reason creates for 
 itself the idea of a spontaneity, which can begin to act 
 
TRANSCENDENTAL DIALECTIC. 1 83 
 
 purely of itself, without being determined to activity 
 by another cause, as the law of natural causality de- 
 mands. 
 
 563 Now, if phenomena were things in themselves, and 535 
 space and time forms of the existence of things in them- 
 selves, the conditions would always be members of exactly 
 the same series as the conditioned. Here, therefore, 
 
 as in the other transcendental ideas, the antinomy would 
 arise, that the series must inevitably be too large or too 
 small for understanding. But it is characteristic of the 
 dynamical conceptions of reason, that they do not 
 consider an object with regard to its magnitude, but 
 
 564 only with regard to its existence. In this case, therefore, 536 
 we may abstract from the magnitude of the series of 
 conditions and direct our attention solely to the dynami- 
 cal relations of condition and conditioned. Thus we at 
 once come upon the difficulty, whether freedom is pos- 
 sible at all, and if it is, whether it can exist along with 
 the universality of the natural law of causality. Can we 
 affirm, disjunctively, that every effect in the world must 
 arise either from nature or from freedom, or must we say, 
 that in different relations the same event is due both to 
 nature and to freedom ? That every event in the world 
 
 of sense is connected with a preceding event according 
 to an unchangeable law of nature, has, in the Transcen- 
 dental Analytic, been shown to be a fundamental principle 
 which admits of no exception. The only question now 
 is, whether, assuming that principle, the same effect may 
 not only be determined in accordance with nature, but 
 may also depend upon freedom, or whether freedom is 
 completely excluded by that inviolable rule. Here the 
 common but false presupposition of the absolute reality 
 
184 THE PHILOSOPHY OF KANT. 
 
 of phenomena shows its baleful influence and confuses 
 our reason. If phenomena are things in themselves, 
 freedom cannot be saved. For, nature will then be the 
 complete and adequate cause of every event, and the 
 condition of an event will be contained only in the series 
 of phenomena that with its effect is necessary according 
 
 565 to the law of nature. If, however, phenomena are not 537 
 taken to be more than they really are ; if they are re- 
 garded, not as things in themselves, but simply as objects 
 connected with one another in our consciousness in con- 
 formity with empirical laws ; then they must themselves 
 have their source in that which is not a phenomenon. 
 Such an intelligible cause is not determined in its caus- 
 ality by phenomena, although the effects of its causality 
 are presented to us as phenomena, and can therefore be 
 regarded as determined by other phenomena. The in- 
 telligible cause, together with its causality, is itself outside 
 
 of the series, while yet its effects are to be found in the 
 series of empirical conditions. The effect can therefore 
 be regarded as free in respect of its intelligible cause, and 
 may at the same time be viewed on its phenomenal side 
 as following from phenomena, according to the necessity 
 of nature. 
 
 566 Possibility of Causality through Freedom. 538 
 
 571 The only question here is this: granting that in the 543 
 whole series of events there is to be found nothing but 
 the necessity of nature, is it yet possible to regard the 
 very same event, which on one side is merely an effect of 
 nature, as on the other side an effect of freedom, or is 
 there between these two sorts of causality a direct 
 contradiction ? 
 
TRANSCENDENTAL DIALECTIC. 1 85 
 
 Among the causes in the phenomenal world there 
 certainly can be nothing that absolutely and from itself 
 could cause a series to begin to be. Every act that pro- 
 duces an event is, as a phenomenon, itself an event or 
 result, which presupposes another state to serve as cause. 
 Everything that comes to be is therefore merely a 
 continuation of the series, and nothing that begins of 
 itself can enter into the series. Hence all the modes in 
 
 572 which natural causes act in the succession of time are 544 
 themselves effects, for which there must again be causes 
 in the series of time. It is vain to seek in the causal 
 connection of phenomena for an original act by which 
 something may come to be that before was not. 
 
 But, granting that the cause of a phenomenal effect is 
 itself a phenomenon, is it necessary that the causality of its 
 cause should be entirely empirical ? May it not be that 
 while every phenomenal effect must be connected with its 
 cause in accordance with laws of empirical causality, this 
 empirical causality, without the least rupture of its con- 
 nection with natural causes, is itself an effect of a causality 
 that is not empirical but intelligible ? May the empirical 
 causality not be due to the activity of a cause, which 
 in its relation to phenomena is original, and which there- 
 fore, in so far as this faculty is concerned, is not phen- 
 omenal but intelligible ; although as a link in the chain of 
 nature it must be regarded as also belonging entirely to 
 the world of sense ? 
 
 574 Let us see how this would apply to experience. Man 546 
 is one of the phenomena of the world of sense, and in so 
 far one of the natural causes, the causality of which must 
 stand under empirical laws. Like all other things in 
 nature, he mjjst have an empirical character. This 
 
lB6 THE PHILOSOPHY OF KANT. 
 
 character we learn from an observation of the powers 
 and faculties which he exhibits in the production of 
 effects. In lifeless nature, or in the mere animal, we find 
 no reason for thinking that there is any faculty but that 
 which is sensuously conditioned. But man, who knows 
 all the rest of nature solely as an object of sense, is 
 aware of himself also by mere apperception, and that 
 in acts and inner determinations, which he is quite unable 
 to regard as due to impressions of sense. On the one 
 side, he is no doubt for himself a phenomenon, but, 
 inasmuch as his actions cannot be ascribed to the recep- 
 575 tivity of sense, he is, on the other side, a purely Intel- 547 
 ligible object with respect to certain of his faculties. 
 These faculties we call understanding and reason; and 
 reason in particular we distinguish in quite a peculiar and 
 especial way from all forces that are empirically con- 
 ditioned, because it contemplates its objects purely in the 
 light of ideas, and determines understanding in accord- 
 ance with them. 
 
 That our reason actually has causality, or that we at 
 least suppose it to have causality, is evident from the 
 imperatives which we impose upon ourselves as rules for 
 our own conduct. The ought expresses a kind of neces- 
 sity and connection with conditions which we shall look 
 for in vain in all the rest of nature. Understanding can 
 know only what is, has been, or will be. It is impossible 
 for anything to exist for understanding otherwise than as 
 a matter of fact it does exist in those three relations of 
 time ; nay, if we fix our eyes simply upon the course of 
 nature, the 02ight has no meaning whatever. It is as 
 absurd to ask what nature ought to be, as to ask what sort 
 of properties a circle ought to have. The only question 
 
TRANSCENDENTAL DIALECTIC. 1 87 
 
 we can properly ask is, what comes to pass in nature ? 
 just as we can only ask, what actually are the properties 
 of a triangle ? 
 
 Now, this ought expresses a possible activity, the 
 ground of which is a bare conception, whereas a mere 
 
 576 natural activity must always have a phenomenal ground. 54^ 
 No doubt an act that ought to be, must be possible 
 under conditions of nature ; but these have no influence 
 
 in determining the will itself, but only in determining the 
 effect and what follows from it in the phenomenon. No 
 matter how many natural influences, how many sensuous 
 impulses may be brought to bear upon my will, they can- 
 not give rise to the ought. The volition which is due 
 to such influences is always conditioned and by no 
 means necessary, and the otight of reason confronts such 
 a volition with a limit and ideal, nay, forbids or author- 
 izes it. Whether the object willed is sensuous pleasure, 
 or even the good which is the object of pure reason, 
 reason refuses to yield to the influence of that which 
 is given empirically, and to follow the order of things as 
 they present themselves in the phenomenal world. With 
 perfect spontaneity it makes for itself an order of its own 
 in accordance with ideas, into which it fits the empirical 
 conditions, and guided by the idea of this order it 
 declares actions which have not yet taken place, and 
 which perhaps never will take place, to be necessary. 
 Thus reason assumes that it has in itself the power of 
 originating actions; for otherwise it would not expect to 
 find in experience the influence of its ideas. 
 
 Now let us pause here for a moment, and assume that 
 
 577 it is at least possible for reason to have causality with 549 
 respect to phenomena. Reason though it be, it must yet 
 
1 88 THE PHILOSOPHY OF KANT. 
 
 manifest an empirical character. For, every cause pre- 
 supposes a rule in conformity with which certain pheno- 
 mena follow as effects, and every rule requires uniformity 
 in the way of effects. It is upon such uniformity, in fact, 
 that the conception of cause as the faculty of producing 
 an effect is based. This uniformity of effect, as learned 
 from simply observing phenomena, may be called the 
 empirical character of a cause, and this empirical charac- 
 ter is unchangeable, although the effects appear in change- 
 able forms according as the accompanying and partly 
 limiting conditions vary. 
 
 Thus the will of every man has an empirical character, 
 which is simply a certain causality of his reason, in so far 
 as that causality manifests, in its effects in the pheno- 
 menal world, a rule from which we may infer the kind 
 and degree of the motives from which his actions have 
 been done, and so estimate the subjective principle of 
 his will. This empirical character must itself be gathered 
 from our observation of the effects of his causality in the 
 phenomenal world, and from the rule with which experi- 
 ence supplies us. It is therefore solely from a considera- 
 tion of the man's empirical character and of the other 
 causes that co-operate with it in conformity with the 
 order of nature, that we are able to determine his 
 578 actions on their phenomenal side; and if we could 550 
 trace all the manifestations of his will to their source, 
 we could tell with certainty what his actions in every 
 case must be, and show that they necessarily followed 
 from the given condition. So long, therefore, as we 
 look only at a man's empirical character, we cannot 
 find any trace of freedom. Yet this is the only thing 
 that comes before us, if we simply observe man, and 
 
TRANSCENDENTAL DIALECTIC. 1 89 
 
 investigate the motives of his actions from the point 
 of view of anthropology. 
 
 But if we change our point of view, and consider the 
 very same actions in their relation to reason, — by which 
 I do not mean speculative reason, which merely explains 
 how the actions have come to be, but reason only in so 
 far as it is the cause thsit prodi/ces them — if, in a word, 
 we view a man's actions in their relation to reason as 
 practical^ we find that they come under an entirely 
 different rule and order from the order of nature. We 
 find, it may be, that nothing ought to have taken place^ 
 which as a matter of fact has taken place in conformity 
 with the course of nature, and could not but take place 
 under the given empirical conditions. But sometimes we 
 find, or at least believe that we find, that the ideas of 
 reason have actually proved their causaHty with reference 
 to the actions of man as phenomena, and that those 
 actions have taken place, not because they were deter- 
 mined by empirical causes, but because they were deter- 
 mined by grounds of reason. 
 579 Now, if we could say that reason has causality in regard 551 
 to phenomena, should we be entitled to say that reason 
 acts freely, although the action is determined with absolute 
 precision and necessity in its empirical character, or as a 
 mode of sense ? The empirical character, it must be 
 observed, will in that case be itself determined in the 
 intelligible character or manner of thinking. The in- 
 telligible character, however, we do not directly know, 
 and hence we have to indicate its nature by means of 
 phenomena, which properly give us a knowledge only of 
 objects of sense, and therefore only of the empirical 
 character. Now, the action, in so far as our manner of 
 
190 THE PHILOSOPHY OF KANT. 
 
 thinking: is to be called its cause, is not a result that 
 follows according to empirical laws from that manner of 
 thinking ; that is, it is not preceded by the conditions of 
 pure reason, but only by the effects of pure reason as 
 they appear in the inner sense. Pure reason, as a faculty 
 that is merely intelligible, is not subject to the form of 
 time, and therefore it is not subject to the conditions 
 belonging to the succession of time. The causality of 
 reason in the intelligible character does not arise or begin 
 580 to be at a certain time in order to produce an effect. If 552 
 it did, it would itself be subject to the natural law of 
 phenomena, in conformity with which causal series are 
 determined in time, and its causality would then be 
 natural and not free. What we must say is, that if reason 
 can have causality with regard to phenomena, it is a 
 faculty by means of which the sensuous condition of an 
 empirical series of effects first begins to be. For the 
 condition which lies in reason is not sensuous, and there- 
 fore does not itself begin to be. In that case we should 
 find what we missed in all empirical series, that the con- 
 dition of a successive series of events might itself be 
 empirically unconditioned. For the condition would lie 
 outside of the series in the intelligible character, and 
 would therefore be subject to no sensuous condition, and 
 to no determination of time through preceding causes. 
 
 585 It must be observed that we have had no intention of 557 
 
 586 proving that there actually is freedom, and that it is one 558 
 of the faculties which are the cause of the phenomena of 
 our world of sense. Freedom has here been viewed 
 simply as a transcendental idea, which leads reason to 
 think that it can absolutely bring into existence the series 
 
 of conditions in the phenomenal world by means of the 
 
TRANSCENDENTAL DIALECTIC. IQI 
 
 sensuously unconditioned, and thus involves it in an 
 antinomy with its own laws — the laws which it lays down 
 for the understanding in its empirical work. The only 
 thing that we have been able to show, or that we have 
 tried to show, is that this antinomy has its source in a 
 mere illusion, and that nature at least does not contradict 
 the causality of freedom. 
 
 5S7 4. Solution of the Fourth Ajititiomy, 559 
 
 In what immediately precedes we have considered the 
 changes of the world of sense in their dynamical series — 
 a series each member of which stands under another as 
 its cause. We shall now take this series of states as our 
 guide in the search for an existence that may serve as the 
 supreme condition of all that changes ; that is, in our 
 search for the necessary being. Here we have to deal, 
 not with an unconditioned causality, but with the un- 
 conditioned existence of substance itself What we have 
 before us is therefore really a series of conceptions, and 
 not a series of perceptions, in which one perception is 
 the condition of the other. 
 
 Now, it is easy to see that, as every object in the 
 totality of phenomena is changeable, and therefore is con- 
 ditioned in its existence, no member of the series of 
 dependent existence can possibly be unconditioned ; in 
 other words, we cannot regard the existence of any 
 member of the series as absolutely necessary. Hence, if 
 phenomena were things in themselves, and if as a con- 
 sequence their condition always belonged to one and the 
 same series of perceptions, there would be no possibility 
 588 of introducing a necessary being as condition of the 5^° 
 existence of phenomena in the world of sense. 
 
192 THE PHILOSOPHY OF KANT. 
 
 But there is a peculiar distinction between the dyna- 
 mical and the mathematical regress. The mathematical 
 regress has to do only with the composition of parts into 
 a whole, or the division of a whole into parts ; and as in 
 this case the conditions must always be regarded as parts 
 of the series, and therefore as homogeneous, they cannot 
 but be phenomena. But in the dynamical regress we are 
 concerned, not with the possibility of an unconditioned 
 whole formed out of given parts, or of an unconditioned 
 part for a given whole, but with the derivation of a state 
 from its cause, or of the contingent existence of substance 
 itself from its necessary existence. Here, therefore, there 
 is no reason why the condition should enter into the 
 same empirical series with that which is conditioned. 
 
 Thus a way of escape from the apparent antinomy now 
 under consideration is opened up to us. Both of the 
 conflicting propositions may be true if they are taken in 
 different senses. All things in the world of sense may be 
 contingent, and therefore have only an empirically con- 
 ditioned existence, while yet there may be a condition of 
 the whole series that is not empirical ; that is, there may 
 be an unconditionally necessary being. For, this necessary 
 being, as the intelligible condition of the series, could not 
 589 belong to it as a member, not even as the highest member 561 
 of it, nor would it make any member of the series 
 empirically unconditioned, or in any way interfere with 
 the empirically conditioned existence of all the members, 
 which form the world of sense as a whole. Thus the 
 manner in which an unconditioned existence is here con- 
 ceived as the condition of phenomena is different from 
 the manner in which, in the last chapter, we sought to 
 explain the empirically unconditioned causality of free- 
 
TRANSCENDENTAL DIALECTIC. 193 
 
 dom. For there the thing itself {substantia phaenomenoji) 
 was as a cause conceived to belong to the series of con- 
 ditions, and only its causality was regarded as intelligible. 
 Here, on the other hand, the necessary being must be 
 conceived as entirely outside of the series of the sensible 
 world (as ens extramundanum)^ and as purely intelligible. 
 In no other way, indeed, can we regard it as free from 
 the law of contingency and dependence to which all 
 phenomena are subject. 
 
 The regulative principle of reason in the present case 
 may therefore be stated in this way. Everything in the 
 world of sense has an empirically conditioned existence, 
 and no property of a sensible object has unconditioned 
 necessity. We are entitled to expect that in a possible 
 experience there will be found an empirical condition for 
 every member of the series of conditions, and the search 
 for sach conditions we ought always to follow up as far 
 as we can. Nothing can justify us in referring any 
 particular mode of existence to a condition outside of 
 the empirical series, or even in regarding a particular 
 mode of existence within the empirical series as absolutely 
 independent and self-subsistent. At the same time there 
 rno is no reason to deny that the whole series may be de- 562 
 pendent upon an intelligible being, which is free from 
 every empirical condition, and is itself the condition of 
 the possibility of all phenomena. 
 
 593 Concluding Re7nark on the whole Antinomy of Pure 565 
 
 Reason. 
 
 So long as reason in its conceptions is seeking simply 
 the totality of conditions in the world of sense, and 
 
 N 
 
194 THE PHILOSOPHY OF KANT. 
 
 trying to find satisfaction in that direction, our ideas are 
 cosj?iological, though at the same time transcendental. 
 But the moment the unconditioned, in which we are 
 mainly interested, is conceived as lying entirely outside 
 of the world of sense, and therefore beyond all possible 
 experience, our ideas become transcendent. For then they 
 are not merely ideas which reason employs in seeking to 
 complete experience — an end which must always be pur- 
 sued, though it can never be fully attained ; rather they are 
 ideas that entirely separate themselves from experience, 
 and create for themselves objects, for which experience sup- 
 plies no material, and which cannot rest their claim to ob- 
 jective reality upon the completion of the empirical series, 
 
 594 but only upon pure apriori conceptions. Nevertheless, the 5^*^ 
 cosmological idea which gave rise to the fourth antinomy 
 urges us to take this step. Finding that phenomena are 
 always conditioned modes of existence, and have no 
 support in themselves, we are driven to look about for 
 something different from all phenomena, and therefore 
 
 for an intelligible object which is entirely free from con- 
 
 595 tingency. Thus the very first step which we take beyond 567 
 the world of sense compels us to enter upon an inquiry 
 into the nature of the absolutely necessary being, and to 
 derive from our conceptions of it our conceptions of all 
 things in their purely intelligible nature. This inquiry is 
 the subject of the next chapter. 
 
TRANSCENDENTAL DIALECTIC. 1 95 
 
 BOOK II.— THE IDEAL OF PURE REASON. 
 
 Section I. — The Ideal in General. 
 
 We have seen above that no object can be known 
 by means of the pure conceptions of understanding, 
 if these are isolated from all conditions of sensibility ; 
 for the conditions of objective reality are then absent, 
 and nothing is left but the mere form of thought. On 
 the other hand pure conceptions can be presented 
 in concrete if they are brought into connection with 
 phenomena, for in phenomena they obtain the appro- 
 priate material by which they become conceptions of 
 experience. A conception of experience, in fact, is 
 simply a conception of understanding in concreto. Now, 
 ideas are even further removed from objective reality 
 than categories^ for, as no phenomenon can be found 
 to which they might apply, they cannot be presented 
 
 596 in conc7'eto at all. They demand a certain complete- 568 
 ness which is beyond the reach of all possible empi- 
 rical knowledge, and reason has in them merely a 
 systematic unity, to which it brings the unity that is 
 possible in experience as near as possible, though it can 
 never hope to bring experience into -complete harmony 
 with its ideas. 
 
 But what I call the ideal seems to be still further 
 removed from objective reality than even the idea. 
 By the ideal I mean the idea, not merely in concreto^ 
 but in individuo ; I mean, in other words, an individual 
 thing that is determinable or even determined simply 
 by the idea. 
 
 599 Reason, in its ideal, aims at absolutely complete deter- 571 
 
196 THE PHILOSOPHY OF KANT. 
 
 mination in accordance with a priori rules. Hence it 
 sets before itself an object which it conceives as capable 
 of thoroughgoing determination in conformity with its 
 own principles. The conditions, however, that are 
 required for such a determination cannot be found in 
 experience, and thus the conception is itself transcendent 
 
 Section II. — The T^-anscendental Ideal. 
 
 601 The proposition, that all existetice is completely deter- 573 
 ?niried, means, that to know a thing completely, it is 
 necessary to know all that can possibly exist, and to 
 determine the thing in question, either affirmatively or 
 negatively, by reference to our ideal. The absolutely 
 complete determination of a thing is therefore a mere idea, 
 which can never be presented in its totality in concreto. 
 This idea has its source entirely in reason, which pre- 
 scribes the rule by which understanding must be guided 
 
 in seeking completeness of knowledge. 
 
 The idea of the totality of all possible existence will be 
 found to exclude a number of predicates. It excludes, 
 to wit, all those predicates that are derived from other 
 
 602 predicates already given, as well as those that cannot 574 
 stand along with them; and thus it leaves us with 
 
 a conception that is determined absolutely a priori^ 
 that is, with the conception of an individual object which 
 is completely determined by the mere idea of it. This 
 is what is meant by an ideal of pure reason. 
 60 -» Now, a negation cannot be definitely thought, except in 575 
 contrast to the affirmation that is its opposite. A man 
 born blind has no idea of darkness, because he has no 
 idea of light. All conceptions of negations are therefore 
 
TRANSCENDENTAL DIALECTIC. 1 97 
 
 derivative, and positive realities contain all the data, and 
 so to speak, the matter or transcendental content, for the 
 possibility and the complete determination of all things. 
 
 This transcendental substratum for the complete deter- 
 mination of things, which is presupposed by our reason, 
 
 604 is simply the idea of a totality of reahty {onmitudo 57^ 
 realitatis). All true negations are therefore merely 
 limitations of that unlimited totality of reality which 
 reason presupposes. 
 
 It is by supposing a thing in itself to possess this 
 totality of reality that we conceive of it as completely 
 determined. Moreover, the conception of this thing in 
 itself as an ens realissimum is the conception of an 
 individual being, for, in determining it, we are forced to 
 assign to it one out of every possible pair of contra- 
 dictory predicates, namely, that predicate which expresses 
 positive being. Thus it is a transcendental ideal which 
 necessarily compels us to conceive of all that exists as 
 completely determined, and to this ideal, as constituting 
 the supreme and the complete material condition of 
 their possibility, all objects must be referred in so far as 
 their content is concerned. Nor is human reason capable 
 of having any other genuine ideal, for in no other way 
 can a conception, which in itself is general, be completely 
 determined from itself, and recognized to be the idea of 
 an individual thing. 
 
 605 Now, it is self-evident that reason can think of things 577 
 as necesarily completely determined, without presuppos- 
 
 606 ing the existence of a being conforming to its ideal. It 578 
 is enough that the idea of that being should be pre- 
 supposed. In its ideal reason finds the prototype of 
 which all things are but imperfect copies or ectypes 
 
198 THE PHILOSOPHY OF KANT. 
 
 and from which they derive the material of their possi- 
 bihty. To this ideal things approximate more or less, 
 but they must always remain at an infinite distance 
 from it. 
 
 All things, then, with the synthesis of various deter- 
 minations which form their content, are regarded as 
 deriving their possibility solely from that which contains 
 all reality within itself and alone is originally possible. 
 The predicates by which all other modes of being are 
 distinguished from the truly real being, are all negative, 
 and negations are merely limitations of a higher reality, 
 and ultimately of the highest reality of all, from which 
 only their content is derived. The manifold deter- 
 minations of things are therefore simply various ways 
 of limiting the conception of the highest reality, which 
 is their common substratum, just as all geometrical figures 
 are merely the various ways in which infinite space is 
 capable of being limited. Hence the object which 
 reason sets before itself as an ideal is also called the 
 origmal Bemg {ens origin ariutfi) ; as having no being 
 higher than itself, it is called the supreme Being {ens 
 607 swufnuni); and it is also named the Being of all beifigs 579 
 {ens entiuni), to indicate that all other beings are con- 
 ditioned and subject to it. But all this does not entitle 
 us to say that there is an actual object which is so 
 related objectively to other things, but only that there 
 is an idea which is so related to our conceptions 
 of things. Whether a Being of such transcendent 
 perfection actually exists we are left in complete 
 ignorance. 
 
 Again, we cannot say that an original being consists 
 of a number of derivative beings, for each of these 
 
TRANSCENDENTAL DIALECTIC. 1 99 
 
 presupposes the original being, and therefore cannot 
 constitute it. The ideal of the original being must 
 therefore be conceived as simple. 
 
 The derivation of all other possibility from the 
 original being cannot properly be said to be a limitation 
 of its supreme reality, and as it were a division of it 
 into parts ; for the original being would in that case 
 be a mere aggregate of derivative beings, and this we 
 have just seen to be impossible, although in our first 
 rough sketch we represented the matter in that way. 
 The supreme reality we must conceive, not as the suvi 
 of all things, but as the necessary condition of their 
 possibility. The manifold determination of things must 
 be regarded, not as a limitation of the original being 
 itself, but as its complete product, to which will belong 
 our whole sensibility, and all the reality in the phenom- 
 enal world, which cannot enter as an ingredient into the 
 idea of the Supreme Being. 
 608 If we follow out this idea and hypostatise it, we shall 58c 
 be able to determine the original being, simply from 
 our conception of the supreme reality, as one, simple, 
 all sufficient, eternal, etc. j in a word, we shall be able 
 to determine it in its unconditioned completeness through 
 all predicaments. Now this is the conception of God, 
 in its transcendental sense, and thus the ideal of pure 
 reason is the object of a transcendental theology. 
 
 By such a use of the transcendental idea, however, 
 theology oversteps limits set to it by its very nature. 
 Reason only demands the conception of all reality as 
 essential to the complete determination of things ; it does 
 not require us to suppose that all this reahty should be 
 given objectively, and should itself constitute a thing. 
 
200 THE PHILOSOPHY OF KANT. 
 
 It is by a mere fiction that we combine the manifold 
 content of our idea into an ideal, and reahze it in a 
 particular being. We have no right to assume without 
 question that such a substantiation of the ideal is even 
 possible ; nor can it be said that any of the consequences 
 that flow from such an ideal have the least bearing upon 
 the complete determination of things in general, although 
 it was only for the sake of that determination that the 
 idea was put forward. 
 
 609 How, then, does it come that reason derives the whole 581 
 possibility of things from one single possibility, namely, that 
 
 of the highest reality, and why does it assume that this 
 reality must be contained in a particular original being ? 
 The answer will readily present itself if we look back to 
 what has been shown in the Transcendental Analytic. 
 There we found that objects of sense are possible only in 
 relation to our thought, which supplies the a priori 
 element or empirical form that is implied in them. But 
 unless the matter were given, that is, the real element in 
 the phenomenal object which corresponds to sensation, 
 the pbject could not be thought at all, nor could we com- 
 prehend how it should be possible. Now an object of 
 sense can be completely determined, only if we are able 
 to compare it with all possible determinations of pheno- 
 mena, and predicate these of it either affirmatively or 
 negatively. But that which constitutes the thing itself, 
 or the real element in the phenomenon, must be given, 
 
 610 and unless it is given the object cannot be thought at all. 582 
 Now the real element of all phenomena is given in the 
 one all-embracing experience; and hence the matter 
 which makes all objects of sense possible must be pre- 
 supposed as given in one comprehensive whole, and only 
 
TRANSCENDENTAL DIALECTIC. 20I 
 
 by the limitation of this whole are empirical objects 
 possible, distinguishable from one another, and capable 
 of complete determination. As a matter of fact, no other 
 objects can be given to us but objects of sense, and 
 these nowhere but in the context of a possible experience. 
 Hence there is /or us no object which does not presup- 
 pose the comprehensive whole of all empirical reality as 
 the condition of its possibility. It is therefore a natural 
 illusion which leads us to suppose that a principle which 
 properly holds only of things that are presented as objects 
 of our senses, is applicable to all things without exception. 
 We simply drop the limitation to phenomena, and imagine 
 that the empirical principle of our conceptions of the 
 possibility of phenomenal objects is a transcendental 
 principle of the possibility of things as such. 
 
 And the reason why we afterwards hypostatise this 
 idea of a comprehensive whole of all reality is, that we 
 change dialectically the distributive unity, implied in the 
 empirical use of understanding, into the collective unity of 
 a whole of experience, and think of this whole of pheno- 
 mena as an individual thing, which contains all reality 
 6 1 1 within itself. Our next step is, by means of the transcen- 583 
 dental subreption already mentioned, to confuse this 
 individual thing, which includes in itself all empirical 
 reahty, with the conception of a thing that constitutes the 
 supreme condition of the possibility of all things, and 
 suppUes the real conditions for their complete determina- 
 tion. 
 
202 THE PHILOSOPHY OF KANT. 
 
 Section III. — Arguments of Speculative Reason for the 
 Existence of a Supreme Being. 
 
 614 The natural course of human reason in seeking to 586 
 prove the existence of a Supreme Being is as follows. 
 First of all reason persuades itself that some necessary 
 being must exist. This being it regards as having uncon- 
 ditioned existence. Then it looks out for that which 
 
 615 can be conceived as independent of all conditions, and 5^7 
 this it finds in that which is itself the sufficient condition 
 
 of all other things, that is, in that which contains all 
 reality. Now, as the unlimited All is absolute unity, and 
 carries with it the conception of a single, supreme being, 
 reason concludes that a Supreme Being must necessarily 
 exist as the original condition of all things. 
 
 Let us suppose that every step in this argument is 
 5i5 valid. Grant, in the first place, that from any given ^88 
 existence, were it only my own, I may legitimately infer 
 the existence of an unconditionally necessary being. 
 Grant, secondly, that I must regard a being which con- 
 tains all reality and therefore all conditions, as uncondi- 
 tionally necessary, and that the conception of this being 
 harmonizes with the idea of absolute necessity. Admit- 
 ting all this, we yet are not entitled to say that there is 
 anything contradictory of absolute necessity in the con- 
 ception of a limited being, which does not possess the 
 highest reality. For, while it is no doubt true that from 
 the conception of a limited being we cannot derive the 
 idea of the unconditioned, which by its very nature 
 implies a totality of conditions, yet it by no means follows 
 that a limited being must in its existence be conditioned. 
 On the contrary, there is nothing to hinder us from 
 
k 
 
 i 
 
 TRANSCENDENTAL DIALECTIC. 203 
 
 supposing that all limited beings may be unconditionally 
 necessary, although no doubt their necessity cannot be 
 inferred merely from the general conception which we 
 have of them. The argument given above cannot, there- 
 fore, help us in the least to determine the nature of a 
 necessary being, and in fact it leads to nothing at all. 
 
 617 But although that argument, resting as it does upon 5^9 
 the internal insufficiency of the contingent, is undoubtedly 
 transcendental, it yet is so simple and natural that it 
 never fails to commend itself even to the most ordinary 
 mind, the moment its bearing is understood. We see 
 things change, arise and perish ; hence they, or at least 
 their state, must have a cause. But for every cause that 
 
 6i8 can be presented in experience, we are forced to seek a 590 
 new cause. Now, where should we more naturally expect 
 to find the ^rsf cause than in the supreme cause, that is, 
 in the Being which originally contains within itself the 
 sufficient explanation of every possible effect, and which 
 besides is so easily conceived through the single mark of 
 all-comprehensive completeness ? This supreme cause 
 is then held to be absolutely necessary, because it is 
 absolutely necessary for us to ascend to it in thought, 
 while yet we see no reason for going beyond it. Hence, 
 even among nations that are in a state of the blindest 
 polytheism, some gleams of monotheism are visible, to 
 which they have been brought, not by reflection and deep 
 . speculation, but simply by following the path that gradu- 
 ally and naturally opened up before them. 
 
 There are only three ways in which the existence of 
 God may be sought to be proved on the basis of specula- 
 
 619 tive reason. The first is the J^/iyst'eo-^/ieo/ogi'ca/ proof, the 591 
 second is the cosmological, and the third the ontologicaL 
 
204 THE PHILOSOPHY OF KANT. 
 
 This is the order in which the three proofs come before 
 reason as it gradually widens its vision. We shall, how- 
 ever, examine them in the reverse order, for, as we shall 
 immediately see, while experience gives the first impulse 
 to reason, it is the transcendental conception only which 
 leads the way, and sets before reason the goal of all its 
 efforts. I shall therefore begin with an examination of 
 the transcendental proof, and then pass on to consider 
 how far that proof may be strengthened by the addition 
 of empirical elements. 
 
 620 Section rV. — The Onto logical Proof. 592 
 
 From what has been said it is obvious that the concep- 
 tion of an absolutely necessary being is a pure conception 
 of reason. It is a mere idea, the objective reality of 
 which is by no means proved by the fact that reason 
 requires it. All that we can say is that the idea of an 
 absolutely necessary being points to a certain ideal com- 
 pleteness, but as this completeness is unattainable, the 
 idea really limits the sphere of understanding instead of 
 extending its knowledge to new objects. 
 
 People have at all times spoken of an absolutely 
 necessary being, but they have begun by seeking to prove 
 its existence without first asking whether and how a thing 
 of that sort could even be conceived. It is certainly easy 
 enough to give a verbal definition of it, as something the 
 
 621 non-existence of which is impossible. But this throws no 593 
 light upon the conditions which force us to regard 
 the non-existence of a thing as absolutely unthinkable. 
 Now it is just these conditions that we really wish to 
 know. We wish to know whether under the conception 
 
TRANSCENDENTAL DIALECTIC. 205 
 
 of a necessary being we are thinking anything at all or 
 not. To speak of the " unconditioned," and thus to take 
 away all the conditions by means of which understanding 
 is able to regard anything as necessary, does not help us 
 to understand, whether in the conception of an uncon- 
 ditionally necessary being we are thinking of a real being, 
 or, as may perhaps be the case, of nothing at all. 
 
 It has commonly been supposed that this conception, 
 now so familiar to us, but originally hit upon by accident, 
 might be justified by bringing forward a number of 
 examples, and that thus all further inquiry into its 
 intelligibility was rendered superfluous. Every geometrical 
 proposition, it was said, as, for instance, that a triangle 
 has three angles, is absolutely necessary ; and people 
 talked as if such examples entitled them to say that they 
 had a perfectly clear conception of what they meant by 
 an object that lay entirely beyond the sphere of human 
 understanding. 
 
 The examples brought forward were, however, all with- 
 out exception taken ixom judgments, not from things and 
 their existence. But the unconditioned necessity of a 
 judgment is not the same thing as an absolute necessity 
 of a thing. The absolute necessity of a judgment is only 
 a conditioned necessity of the thing predicated, that is, 
 622 of the predicate in the judgment. The proposition just 594 
 cited does not say that three angles are absolutely 
 necessary, but only that, if a triangle exists, that is, is 
 presented in perception, it must contain three angles. 
 But this mere logical necessity has proved a fruitful 
 source of illusion. People have framed a priori the con- 
 ception of a thing that seems to include existence within 
 its content, and have then assumed that, because existence 
 
206 THE PHILOSOPHY OF KANT. 
 
 belongs necessarily to the object as conceived, it must 
 also belong necessarily to the thing itself. Thus it is 
 inferred that there is an absolutely necessary being, 
 because the existence of that being is thought in a con- 
 ception that has been arbitrarily assumed, and assumed 
 under the supposition that there is an actual object 
 corresponding to it. 
 
 If in an identical judgment I retain the subject after 
 rejecting the predicate, a contradiction arises, and hence 
 I say that the predicate belongs to the subject necessarily. 
 But if I reject the subject as well as the predicate, there 
 is no contradiction, for nothing is left to which a contra- 
 diction could apply. To assume that there is a triangle, 
 and yet to deny that it has three angles, is contradictory, 
 but there is no contradiction in denying both the triangle 
 and its three angles. It is exactly the same with the 
 623 conception of an absolutely necessary being. If the 595 
 existence of that being is denied, the thing itself with all 
 its predicates is at the same time denied. How can this 
 be shown to involve a contradiction ? The contradiction 
 cannot come from without, for the thing is not said to be 
 necessary because of its relation to anything external ; 
 nor can it come from within, for, in denying the reality 
 of the thing itself, the reality of all that it contains is at 
 the same time denied. " God is almighty " is a necessary 
 judgment. The predicate " almighty " cannot be denied, 
 so long as the subject "God" is affirmed, for the con- 
 ception of God, that is, of an infinite being, is identical 
 with the conception of a Being that is " almighty." But 
 if you say, " There is no God," neither the predicate 
 " almighty " nor any other predicate remains : in the 
 denial of the subject every possible predicate is denied, 
 
TRANSCENDENTAL DIALECTIC. 207 
 
 and there is therefore not the least contradiction in 
 saying that God does not exist. 
 
 624 At this point, however, I am told that there is one 59^ 
 conception, although only one, the object of which can- 
 not without contradiction be denied to exist. The con- 
 ception is that of an absolutely necessary Being. This 
 Being, it is said, possesses all reality, and such a Being, 
 
 as I am willing to admit, we are justified in assuming to 
 be possible. Now that which comprehends all reality, 
 the objector goes on, must also comprehend existence. 
 Hence existence is in this case involved in the con- 
 
 625 ception of a thing as possible. If, therefore, the thing is 597 
 denied to exist, even its internal possibility is denied, 
 and this is self-contradictory. 
 
 Now I simply ask, whether the proposition, that this or 
 that thing exists, is an analytic or a synthetic proposition. 
 If it is analytic, nothing is added to the thought of a 
 thing by predicating existence of it. Either the thought 
 in you must itself be the thing, or you have simply 
 assumed existence to be implied in mere possibility, and 
 then derived existence from internal possibility, which is 
 nothing but a wretched tautology. It does not mend 
 matters to use the word "reality" in speaking of the 
 conception of a thing, and the word "existence" in 
 speaking of the conception of the predicate. Call all 
 that is assumed " reality," and in the conception of the 
 subject the thing with all its predicates is already assumed 
 to be actual, and this assumption is simply repeated in 
 
 626 the predicate. Admit, on the other hand, as every 598 
 rational being must admit, that every proposition which 
 affirms existence is synthetic, and how can it be any 
 longer maintained that the predicate of existence cannot 
 
2o8 THE PHILOSOPHY OF KANT. 
 
 be denied without contradiction ? That is the privilege 
 of analytic propositions only, and is bound up with their 
 very nature. 
 
 The illusion which arises from confusing a logical 
 predicate with a real predicate, that is, with one that 
 determines an actual thing, stubbornly resists almost all 
 attempts to correct it. As logic abstracts from all con- 
 tent, anything at all may serve as a logical predicate ; 
 nay, the subject may even be predicated of itself But a 
 determination is a real predicate, which adds something 
 to the conception of the subject and enlarges it. Hence 
 it must not be assumed in the conception of the subject. 
 Bei?ig is evidently not a real predicate, that is, a con- 
 ception of something that is capable of being added to 
 the conception of a thing. It is merely the ungrounded 
 assertion of a thing or of certain determinations as an 
 object of thought. In logic being is simply the copula 
 of a judgment. The proposition, God is omnipotent^ 
 contains two conceptions, the objects of which are 
 respectively God and onmipotence ; and the word is adds 
 627 no new predicate, but is merely a sign that the predicate 599 
 omnipotent is asserted in relation to the subject God. 
 If, then, I take the term God, which is the subject, to 
 comprehend the whole of the predicates, including the 
 predicate onuiipotent, and say, God is, or Tliere is a God, 
 I do not enlarge the conception of God by a new predicate, 
 but I merely bring the subject in itself with all its 
 predicates, in other words, the object, into relation with 
 my conception. The content of the object and of my 
 conception must be exactly the same, and hence I add 
 nothing to my conception, which expresses merely the 
 possibility of the object, by simply placing its object 
 
TRANSCENDENTAL DIALECTIC. 209 
 
 before me in thought, and saying that it is. The real 
 contains no more than the possible. A hundred real 
 dollars do not contain a cent more than a hundred 
 possible dollars. The one signifies the conception, the 
 other the object as it is set over against the conception ; 
 but if the object contained more than the conception, the 
 conception would not express the whole object, and 
 would therefore be an inadequate conception. No doubt 
 there is in my purse a hundred dollars more if I actually 
 possess them, than if I have merely the conception, that 
 is, have merely the possibiUty of them. As real, the 
 object is not simply contained in my conception 
 analytically, but it is added to it synthetically, the con- 
 ception as such being merely a determination of my own 
 state. But the hundred dollars do not become more 
 than a hundred whether they exist outside of my con- 
 ception or not. 
 628 No matter therefore what or how many are the predi- 600 
 cates by which I think a thing, no matter if I should 
 think it even in the completeness of its determinations, I 
 add absolutely nothing to it by saying that it is. To 
 think of a Being of the highest reality, a Being in whom 
 no reality is wanting, in no way settles the question, 
 whether that Being does or does not exist. For, 
 although my conception of the possible real content of a 
 thing may want nothing, it may be only a conception, 
 and relatively to my whole state of thinking, this may 
 be awanting, that I have no knowledge whether the 
 object of my conception is also possible a posteriori. 
 And here we come upon the true source of the difficulty. 
 Were it an object of sense that was in question, we 
 
 should never think of identifying the existence of the 
 
 o 
 
2 TO THE PHILOSOPHY OF KANT. 
 
 thing with the mere conception of it. In that case we at 
 once see that the conception of a thing signifies merely 
 the agreement of the object with the universal conditions 
 of all possible empirical knowledge ; whereas, by the 
 existence of a thing we mean that the object is thought 
 as contained in the context of experience as a whole. 
 
 629 The conception of the object is not in the least enlarged 601 
 by its connection with the context of experience as a 
 whole, but our thought is enriched by the possibility 
 of another perception. It is therefore not surprising 
 that, if we try to think existence simply by means of 
 the pure category, we cannot mention a single mark 
 which distinguishes existence from mere possibility. 
 The conception of a Supreme Being is in many re- 
 spects a most valuable idea, but, just because it is 
 
 630 only an idea, it is quite incapable by itself of extend- 602 
 ing our knowledge of actual existence. It cannot even 
 enable us to say that something may possibly exist apart 
 from the idea. Leibnitz is therefore very far from having 
 shown a priori^ as he fondly supposed he had shown, 
 that so sublime an ideal Being is even possible. 
 
 The labour and energy spent upon the famous ontolog- 
 ical or Cartesian proof from mere conceptions of the exist- 
 ence of a Supreme Being are therefore thrown away, and 
 a man has no more chance of extending his knowledge 
 by means of mere ideas than a merchant can better his 
 position by adding a few noughts to his cash account. 
 
 631 Section V, — The Cos??iological Proof . 603 
 
 632 The cosmological proof, like the ontological, affirms 604 
 the connection of absolute necessity with the highest 
 
TRANSCENDENTAL DIALECTIC. 211 
 
 reality ; but, instead of reasoning from the highest reality 
 to necessity of existence, it reasons from the uncondi- 
 tioned necessity of some being or other as given to the 
 unlimited reaUty of that being. It thus enters upon a line of 
 reasoning which at least seems to be natural, whether it is 
 to be called rational or sophistical, and which has a certain 
 persuasive force with the speculative not less than with 
 the common intellect. This proof, which is called by 
 Leibnitz the argument a contingentia mujidt, we shall 
 now state and examine. 
 It runs thus : — 
 
 If 
 anything exists, an absolutely necessary Being exists. 
 
 Now, 
 at least, I myself exist. 
 
 Therefore, 
 an absolutely necessary Being exists. 
 
 The minor premise of this syllogism contains the 
 633 statement of a particular experience ; the major premise, 605 
 the inference from any experience at all to the existence 
 of something that is necessary. The proof therefore 
 properly starts from experience, and thus it does not 
 proceed completely a priori or ontologically. For that 
 reason, and also because the object of all possible 
 experience is called the world, it is known as the 
 cosmological proof. And as it makes abstraction from 
 all the special properties of objects of experience which 
 make our world different from any other possible world, 
 this argument is distinguished also from the physico- 
 theological method of proof, which reasons from the 
 peculiar nature of the world of sense as it is presented to 
 our observation. 
 
2 12 THE PHILOSOPHY OF KANT. 
 
 The proof then goes on as follows. There is only one 
 way in which the necessary being can be determined, or, 
 in other words, it must have one out of each possible 
 pair of opposite predicates. Hence the conception of 
 the necessary being must co77ipletely determine it. Now 
 there is only a single conception possible, which com- 
 pletely determines a thing a priori^ namely, the concep- 
 tion of the ens realissiviiini. Therefore, the conception 
 634 of the absolutely real being is the only one under which 606 
 a necessary being can be thought, that is, a Supreme 
 Being necessarily exists. 
 
 In this cosmological argument so many sophistical 
 propositions are brought together, that it seems as if 
 speculative reason had exhausted its dialectical skill in 
 producing the greatest possible transcendental illusion. 
 I shall at present simply mention in their order the 
 sophisms by which an old argument has been clothed in 
 a new form and an appeal made to the agreement of two 
 witnesses, experience and reason, when in reality the 
 only witness is pure reason, which assumes a different 
 dress and voice, and pretends to be a second witness. 
 To make quite sure of its stability this proof takes its 
 stand upon experience, and therefore affects to be differ- 
 ent from the ontological proof, which puts its entire trust 
 in pure a priori conceptions. But the only use the cos- 
 mological proof makes of experience is to enable it to 
 make the first step, and to reach the conclusion that 
 some sort of necessary being exists. Experience, how- 
 ever, cannot tell us what may be the nature of this 
 necessary being, and hence reason is forced to make 
 a perfectly new start, and to seek by an examination 
 of mere conceptions to find out what must be the attri- 
 
TRANSCENDENTAL DIALECTIC. 213 
 
 635 butes of such a being ; in other words, it asks which 607 
 among all possible things contains in itself the conditions 
 essential to absolute necessity. The required conditions, 
 
 as it comes to believe, are found simply and solely in 
 the conception of an absolutely real being, and accord- 
 ingly it infers that this is the absolutely necessary being. 
 It is plain, however, that in this reasoning it is assumed, 
 that the conception of a being of the highest reality 
 perfectly coincides with the conception of absolute neces- 
 sity of existence, and that we can therefore reason from 
 the one to the other. Now this was also the assumption 
 of the ontological argument; so that a principle is as- 
 sumed and made the basis of the cosmological proof 
 which it was the express object of that proof to avoid 
 using. Absolute necessity is existence that follows from 
 mere conceptions. If it is said, that the conception of 
 the ens reaHssimu?n is a conception, and indeed the only 
 one, which is appropriate and adequate to necessary 
 existence, it must also be admitted, that the one can 
 be inferred from the other. Plainly, therefore, it is the 
 ontological argument from mere conceptions which gives 
 to the so-called cosmological proof all its force. The 
 appeal to experience is quite idle, serving at the most to 
 suggest the conception of absolute necessity, but not to 
 connect that conception with any object in particular. 
 The moment we try to effect the connection, we are 
 forced to leave experience altogether, and to search among 
 
 636 pure conceptions for one which contains in itself the con- 608 
 ditions of the possibility of an absolutely necessary Being. 
 But if in this way we could be sure that such a Being 
 
 is possible at all, its existence would at the same time be 
 established ; for the argument amounts to this, that of all 
 
214 THE PHILOSOPHY OF KANT. 
 
 possible beings there is only one which carries with it 
 absolute necessity ; in other words, that there is only one 
 Being which exists with absolute necessity. 
 
 637 The second path, then, upon which speculative reason 609 
 enters in its effort to reach the existence of the Supreme 
 Being, is not only equally deceptive with the first, but it 
 has the additional fault of leading to an igjioratio eleiichi. 
 
 It undertakes to lead us by a new way, but after a short 
 circuit it brings us back to the very path that it had 
 induced us to abandon. 
 
 I have said that in this cosmological proof there lies 
 hidden a whole nest of dialectical assumptions, which 
 transcendental criticism has no difficulty in detecting and 
 exposing. I shall content myself with simply enumerating 
 them, leaving it to the reader who is familiar with our 
 method to find out for himself wherein their fallacy con- 
 sists. 
 
 We find in it (i) the transcendental principle of reason- 
 ing from the contingent to its cause. This principle is 
 no doubt applicable within the world of sense, but beyond 
 that world it has no meaning whatever. No synthetic 
 proposition like that of causality can be derived from the 
 purely intellectual conception of the contingent. The 
 principle of causality has no meaning, and no criterion 
 for its employment, except in relation to the world of 
 sense ; whereas, in the cosmological argument, it is used 
 for the very purpose of taking us beyond the world of 
 
 sense. 
 
 638 There is (2) the inference to a first cause from the 610 
 impossibiUty of an infinite series of causes being pre- 
 sented one after the other in the world of sense. This 
 
 is an inference which reason does not permit us to employ 
 
TRANSCENDENTAL DIALECTIC. 215 
 
 as a principle even within experience, much less to 
 extend it beyond experience, where there is no chain of 
 causes at all. 
 
 Also (3) the false self-satisfaction of reason that the 
 series is complete merely because all the conditions of 
 the series have finally been eliminated. It is overlooked 
 that there can in that case be no necessity, and it is 
 simply assumed that the conception is complete because 
 there is no longer anything to serve as an object of 
 conception. 
 
 And (4) the confusion of the logical possibility of the 
 conception of reality as a complete whole, a conception 
 which no doubt is self-consistent, with the transcendental 
 possibility of that reality. For the latter there is needed 
 a principle that shows the practicability of such a syn- 
 thesis, and a principle of that kind can apply only to the 
 field of possible experience. 
 
 642 Source of the Dialectical Illusion in all Transcendental 614 
 
 Proofs. 
 
 Both of these proofs are transcendental, or attempts to 
 prove the existence of God independently of empirical 
 
 643 principles. What, then, in these transcendental proofs is 6 1 5 
 the cause of the dialectical and yet natural illusion, by 
 which the conception of necessity is connected with the 
 conception of the highest reality, and by which that 
 which is only an idea is realized and hypostatized ? 
 
 644 If I am forced to think something to be necessary as a ^ ^ ^ 
 condition of the existence of things in general, and if yet 
 
 I am unable to think of a single thing which is in itself 
 necessary, it inevitably follows that necessity and con- 
 
2l6 THE PHILOSOPHY OF KANT. 
 
 tingency can have no meaning as applied to things them- 
 selves. Were it otherwise, a contradiction would arise. 
 Hence neither of these two principles can be objective. 
 But this does not hinder them from being subjective 
 principles of reason, one of which calls upon us to seek 
 for something that is necessary as a condition of all that 
 is presented as existing, and to be content with nothing 
 short of a complete a priori explanation ; while the other 
 warns us that we need never hope to obtain a complete 
 explanation, or, in other words, that we must not suppose 
 anything empirical to be unconditioned and to admit of 
 no further derivation. Taken in this sense, the two 
 principles, as merely heuristic and regulative^ concern 
 only the formal interest of reason, and are quite con- 
 sistent with each other. The one says that in our 
 speculations on nature we should proceed as if there 
 were a necessary first cause of all that belongs to exist- 
 ence; for here our object is simply to bring our know- 
 ledge to systematic unity, an object that may be attained 
 if we only keep before our minds the idea of a supreme 
 cause as a point towards which we should direct our 
 645 efforts. The other warns us not to regard any single 617 
 determination relating to the existence of things as an 
 ultimate cause, that is, as absolutely necessary, but always 
 to view it as conditioned, and therefore to keep the way 
 ever open for further derivation. But if all that is observed 
 to belong to things must be viewed as only conditionally 
 necessary, nothing that is presented to us empirically can 
 be regarded as absolutely necessary. 
 
 From this it follows that we must conceive that which 
 is absolutely necessary to be outside of the world. Serving 
 simply as a principle for producing the greatest possible 
 
i 
 
 TRANSCENDENTAL DIALECTIC. 217 
 
 unity in phenomena through the idea of a supreme cause, 
 it can never be realized in the worlds because the second 
 rule bids us look upon all empirical causes of the unity 
 of phenomena as derived. 
 
 647 The ideal of the Supreme Being is therefore nothmg 6^9 
 but a regulative principle of reason, telling us to view all 
 connection in the world as if it proceeded from an all- 
 sufficient necessary cause. We can base upon it a rule 
 
 for the systematic unity which is necessary m the explana- 
 tion of the world according to universal laws, but it does 
 not entitle us to assert that there is an existence necessary 
 in itself At the same time it is impossible to avoid the 
 transcendental subreption by which this formal principle 
 is imagined to be constitutive, and the unity of the world 
 hypostatized. It is the same here as with space. Space 
 is merely a principle of sensibility, but as it originally 
 makes possible all the figures which are merely different 
 Hmitations of itself, it is held to be something absolutely 
 necessary and self-subsistent, and to be an object given in 
 itself a prio7'i. Similarly, the systematic unity of nature 
 cannot be shown to be a principle of the empirical use of 
 our reason, except in so far as we presuppose the idea of 
 an absolutely real being as the supreme cause. Accord- 
 ingly, this idea is supposed to be an actual object, and 
 this object, because it is the supreme condition of exist- 
 
 648 ence, is regarded as necessary. Thus a regulative prin- 620 
 ciple is changed into a constitutive principle. That such 
 
 a substitution has actually been made is evident from 
 this, that if I consider that Supreme Being, which rela- 
 tively to the world is absolutely or unconditionally 
 necessary, as a thing existing by itself, I cannot conceive 
 what such necessity means. The conception of necessity, 
 
2l8 THE PHILOSOPHY OF KANT. 
 
 therefore, is one which lies in my reason merely as a 
 formal condition of thought, not as a material and 
 hypostatic condition of existence. 
 
 Section VI. — The Physico-theo logical Proof. 
 
 If, then, neither the conception of things in general, 
 nor the experience of any existence in general^ can yield 
 what we require, it only remains to try whether a 
 determinate experiejice of the things that are presented 
 to us in the world, and of their constitution and order, 
 may not enable us to establish the existence of a 
 Supreme Being. Such a proof we should call the 
 physico-theological. Should this also fail, no satisfactory 
 proof can be given, on the basis of speculative reason, 
 of the existence of a Being corresponding to our 
 transcendental idea. 
 
 649 After what has already been said, it is soon seen 621 
 that an easy and conclusive answer to this last problem 
 may be expected. For how can any experience ever 
 be adequate to an idea? It is just the peculiarity of an 
 idea of reason that no experience can coincide with it. 
 The transcendental idea of a necessary and all-sufficient 
 Original Being is so transcendently great, and is raised 
 so far above all that is empirical and conditioned, that we 
 can never find in experience material enough to realize it 
 in its completeness. We are therefore forced to grope 
 about among things conditioned, seeking in vain for an 
 unconditioned, of which no law of empirical synthesis 
 can give us an example or even the least indication. 
 
 65 1 The physico-theological proof must always be mentioned 623 
 with respect. It is the oldest and simplest proof of all, 
 and never fails to commend itself to the popular mind. 
 
TRANSCENDENTAL DIALECTIC. 219 
 
 It imparts life to the study of nature, as it was itself 
 suggested by that study, and receives new vigour from it. 
 
 652 But although this line of argument must be admitted 624 
 to be both reasonable and useful, it cannot be shown to 
 have any claim to demonstrative certainty. On the 
 contrary, it must be maintained that the physico-theo- 
 logical argument in proof of the existence of a Supreme 
 Being cannot stand alone, but has to fall back upon the 
 ontological argument, which it simply serves to introduce, 
 in order to make up for its own deficiency. The ontologi- 
 cal is therefore the only possible argument. 
 
 6:57 The main steps in the physico-theological argument 525 
 are these : (i) There are in the world everywhere distinct 
 marks of adaptation to a definite end, an adaptation 
 which has been carried out with great wisdom, and which 
 is traceable in a whole indescribably complex in content 
 as well as unlimited in extent. (2) This adaptation does 
 not at all belong to the nature of things that exist in the 
 world, but is extraneous and accidental. That is to say, 
 different things could not all conduce to a single end 
 through such an infinite variety of means, were they not 
 specially selected and adapted to that end by a rational 
 principle acting from certain preconceived ideas. (3) 
 There must therefore exist a single wise and sublime 
 cause, or more than one, and this cause cannot be identi- 
 fied with the blind, all-powerful productiveness of nature, 
 but must be an intelligent B.nd free cause. (4) The unity 
 
 654 exhibited in the mutual relation of the parts of the world 626 
 is that of a skilfully constructed edifice ; and hence the 
 unity of the cause of the world is certain so far as our 
 observation extends, and by ail the principles of analogy 
 is probable even beyond its range. 
 
2 20 THE PHILOSOPHY OF KANT. 
 
 According to this argument, the adaptation and harmony 627 
 
 655 of so many forms of nature proves contingency merely in 
 the form of the world, but not in its matter or substance. 
 To prove the latter, it would be necessary to show that 
 the things of the world would not be capable of such 
 order and harmony, if they were not in their substance the 
 product of supreme wisdom. But to prove this, we 
 should have to take a totally different line of argument 
 from that which appeals to the analogy of human art. 
 All that the argument from design can possibly prove is 
 an architect of the world, who is very much limited by the 
 adaptability of the material in which he works : it cannot 
 prove a creator of the world, to whose idea everything is 
 subject. The argument is therefore very far from being 
 sufficient to prove what it set out to prove, namely, the 
 existence of an all-sufficient Original Being. To establish 
 the contingency of matter itself, we would need to have 
 recourse to a transcendental argument, and this is the 
 very thing which, in the argument from design, we have 
 been trying to avoid. 
 
 The physico-theological argument therefore reasons 
 from the contingent character of the order and adaptation 
 everywhere observable in the w^orld to the existence of a 
 c^iMSt adequate to its productio?t. But as this cause must 
 be conceived as something perfectly defi?iite, it can only 
 be the conception of a Being who possesses all power, 
 wisdom, etc., in a word, all that perfection which is 
 characteristic of an all-sufficient Being. 
 
 656 Now, I think no one will be bold enough to say that he 628 
 can tell how the greatness of the world which is presented 
 
 for his observation is related, either in content or extent, 
 to omnipotence ; how the order of the world is related to 
 
TRANSCENDENTAL DIALECTIC. 22 1 
 
 supreme wisdom, and the unity of the world to the 
 absolute unity of its Author. Hence the physico-theo- 
 logical argument cannot give us a definite conception of 
 the Supreme Cause of the world, and is therefore insuffi- 
 cient as the principle of a theology, which is itself to serve 
 as the basis of religion. 
 
 657 The truth is that, when it has led us to the point of 629 
 admiring the greatness of the wisdom, power, etc., of the 
 Author of the world, the argument from experience 
 cannot take us any further. Accordingly, we abandon it 
 altogether, and go on to reason from the contingency 
 which we had inferred at the very beginning from the 
 order and design of the world. From this contingency 
 
 we advance, by means solely of transcendental conceptions, 
 to the existence of something absolutely necessary. 
 And, finally, from the conception of the absolute necessity 
 of the first cause, we proceed to the thoroughly deter- 
 mined or determining conception of that cause, that is, 
 to an all-comprehensive reality. Thus the physico-theo- 
 logical proof, foiled in its attempt to prove the existence 
 of a Supreme Being, suddenly falls back upon the cosmo- 
 logical proof; and as the cosmological is simply the 
 ontological proof in disguise, the argument from design 
 really fulfils its aim by means of pure reason only, 
 although it began by disclaiming all connection with pure 
 reason, and professed to make use of nothing that was 
 not clearly proved by experience. 
 
 658 The physico-theological proof of the existence of an 630 
 Original or Supreme Being, therefore, rests upon the 
 cosmological proof, and the cosmological upon the onto- 
 logical. And as no other path is open to speculative 
 reason but these three, the ontological proof from pure 
 
222 THE PHILOSOPHY OF KANT. 
 
 conceptions of reason is the only one possible, even if we 
 admit that any proof of a proposition, which goes so far 
 beyond understanding as employed in experience, is 
 possible at all. 
 
 659 Sectio7i VII. — Criticism of all Speculative Theology. 631 
 
 667 Reason in its purely speculative use is quite incapable 639 
 of proving the existence of a Supreme Being. At the 
 same time it is of very great value in this way, that it 
 is able to correct our knowledge of that Being, should it 640 
 be possible to obtain a knowledge of it in any other way, 
 to bring it into harmony with itself and with all the aims 
 of our intelligence, and to purify it of all that is incon- 
 sistent with the conception of an Original Being, and of 
 all admixture of empirical limitations. 
 
 669 The Supreme Being is for purely speculative reason a 641 
 mere ideal, but still a perfectly faultless ideal, which 
 completes and crowns the whole of human knowledge. 
 And if it should turn out that there is a moral theology, 
 which is able to supply what is deficient in speculative 
 theology, we should then find that transcendental 
 theology is no longer merely problematic, but is indis- 
 pensable in the determination of the conception of a 
 Supreme Being, and in the continual criticism of reason, 
 which is so often deluded by sense and is not always in 
 harmony even with its own ideas. Necessity, infinity, 
 unity, existence apart from the world (not as a soul of 
 the world), eternity as free from conditions of time, 
 
 670 omnipresence as unaffected by conditions of space, etc., 642 
 are purely transcendental predicates, the purified concep- 
 tion of which, essential as it is to every theology, can be 
 derived only from a transcendental theology. 
 
THE METAPHYSIC OF MORALITY. 
 
THE METAPHYSIC OF MORALITY. 225 
 
 THE METAPHYSIC OF MORALITY. 
 
 V. 241 Section I. — Transition from ordinary Moral Conceptions to 
 the Philosophical Conception of Morality. 
 
 Nothing in the whole world, or even outside of the 
 world, can possibly be regarded as good without limita- 
 tion except a good will. No doubt it is a good and 
 desirable thing to have intelligence, sagacity, judg- 
 ment, and other intellectual gifts, by whatever name they 
 may be called; it is also good and desirable in many 
 respects to possess by nature such qualities as courage, 
 resolution, and perseverance ; but all these gifts of 
 nature may be in the highest degree pernicious and 
 hurtful, if the will which directs them, or what is called 
 the character, is not itself good. The same thing applies 
 to gifts of fortune. Power, wealth, honour, even good 
 health, and that general well-being and contentment with 
 one's lot which we call happiness, give rise to pride and 
 not infrequently to insolence, if a man's will is not good ; 
 nor can a reflective and impartial spectator ever look 
 with satisfaction upon the unbroken prosperity of a man 
 who is destitute of the ornament of a pure and good will. 
 A good will would therefore seem to be the indispensable 
 condition without which no one is even worthy to be 
 happy. 
 242 A man's will is good, not because the consequences 
 which flow from it are good, nor because it is capable of 
 
2 26 THE PHILOSOPHY OF KANT. 
 
 attaining the end which it seeks, but it is good in itself, 
 or because it wills the good. By a good will is not 
 meant mere well-wishing; it consists in a resolute 
 employment of all the means within one's reach, and 
 its intrinsic value is in no way increased by success or 
 lessened by failure. 
 
 This idea of the absolute value of mere will seems so 
 extraordinary that, although it is endorsed even by the 
 popular judgment, we must subject it to careful scrutiny. 
 
 243 If nature had meant to provide simply for the mainten- 
 ance, the well-being, in a word the happiness, of beings 
 which have reason and will, it must be confessed that, in 
 making use of their reason, it has hit upon a very poor 
 way of attaining its end. As a matter of fact the very 
 worst way a man of refinement and culture can take to 
 secure enjoyment and happiness is to make use of his 
 reason for that purpose. Hence there is apt to arise in 
 his mind a certain degree of misology^ or hatred of reason. 
 Finding that the arts which minister to luxury, and even 
 the sciences, instead of bringing him happiness, only lay 
 a heavier yoke on his neck, he at length comes to envy, 
 rather than to despise, men of less refinement, who follow 
 more closely the promptings of their natural impulses, 
 
 244 and pay little heed to what reason tells them to do or to 
 leave undone. It must at least be admitted, that one 
 may deny reason to have much or indeed any value in 
 the production of happiness and contentment, without 
 taking a morose or ungrateful view of the goodness with 
 which the world is governed. Such a judgment really 
 means that life has another and a much nobler end than 
 happiness, and that the true vocation of reason is to 
 secure that end. 
 
THE METAPHYSIC OF MORALITY. 227 
 
 The true object of reason then, in so far as it is prac- 
 tical, or capable of influencing the will, must be to 
 produce a will which is good in itself^ and not merely 
 good as a means to something else. This will is not the 
 only or the whole good, but it is the highest good, and 
 the condition of all other good, even of the desire for 
 happiness itself It is therefore not inconsistent with the 
 wisdom of nature that the cultivation of reason which is 
 essential to the furtherance of its first and unconditioned 
 object, the production of a good will, should, in this life 
 at least, in many ways limit, or even make impossible, 
 the attainment of happiness, which is its second and 
 conditioned object. 
 
 To bring to clear consciousness the conception of a 
 
 245 will which is good in itself, a conception already familiar 
 
 to the popular mind, let us examine the conception of 
 
 duiy^ which involves the idea of a good will as manifested 
 
 under certain subjective limitations and hindrances. 
 
 I pass over actions which are admittedly violations of 
 duty, for these, however useful they may .be in the attain- 
 ment of this or that end, manifestly do not proceed /r^;^ 
 duty. I set aside also those actions which are not 
 actually inconsistent with duty, but which yet are done 
 under the impulse of some natural inclination, although 
 not a direct inclination to do these particular actions ; 
 for in these it is easy to determine whether the action 
 that is consistent with duty, is done from duty or with 
 some selfish object in view. It is more difficult to make 
 a clear distinction of motives when there is a direct in- 
 clination to do a certain action, which is itself in con- 
 formity with duty. The preservation of one's own life, 
 for instance, is a duty; but, as everyone has a natural 
 
2 28 THE PHILOSOPHY OF KANT. 
 
 inclination to preserve his life, the anxious care which 
 most men usually devote to this object, has no intrinsic 
 value, nor the maxim from which they act any moral 
 import. They preserve their life in accordance with duty, 
 but not because of dxXy. But, suppose adversity and hope- 
 
 246 less sorrow to have taken away all desire for life ; suppose 
 that the wretched man would welcome death as a release, 
 and yet takes means to prolong his life simply from a sense 
 of duty; then his maxim has a genuine moral import. 
 
 247 But, secondly, an action that is done from duty gets its 
 moral value, ?iot from the object which it is intended to 
 secure, but from the maxim by which it is determined. 
 Accordingly, the action has the same moral value 
 whether the object is attained or not, if only the 
 principle by which the will is determined to act is 
 
 248 independent of every object of sensuous desire. What 
 was said above makes it clear, that it is not the object 
 aimed at, or, in other words, the consequences which flow 
 from an action when these are made the end and motive 
 of the will, that can give to the action an unconditioned 
 and moral value. In what, then, can the moral value of 
 an action consist, if it does not lie in the will itself, as 
 directed to the attainment of a certain object ? It can 
 lie only in the principle of the will, no matter whether 
 the object sought can be attained by the action or not. 
 For the will stands as it were at the parting of the ways, 
 between its a priori principle, which is formal, and its 
 a posteriori material motive. As so standing it must be 
 determined by something, and, as no action which is 
 done from duty can be determined by a material prin- 
 ciple, it can be determined only by the formal principle of 
 all volition. 
 
THE METAPHYSIC OF MORALITY. 229 
 
 From the two propositions just set forth a third 
 directly follows, which may be thus stated : Duty is the 
 obligation to act fro?n reverence for law. Now, I may 
 have a natural inclinatiofi for the object that I expect to 
 follow from my action, but I can never have revere?ice for 
 that which is not a spontaneous activity of my will, but 
 merely an effect of it ; neither can I have reverence for 
 any natural inclination, whether it is my own or another's. 
 If it is my own, I can at most only approve of it ; if it is 
 manifested by another, I may regard it as conducive to 
 my own interest, and hence I may in certain cases even 
 be said to have a love for it. But the only thing which I 
 can reverence or which can lay me under an obligation 
 to act, is the law which is connected with my will, not as a 
 consequence, but as a principle ; a principle which is not 
 dependent upon natural inclination, but overmasters it, 
 or at least allows it to have no influence whatever in 
 determining my course of action. Now if an action which 
 is done out of regard for duty sets entirely aside the 
 influence of natural inclination and along with it 
 every object of the will, nothing else is left by which 
 the will can be determined but objectively the law 
 itself, and subjectively pure reverence for the law as a 
 principle of action. Thus there arises the maxim, to 
 obey the moral law even at the sacrifice of all my natural 
 inclinations. 
 249 The supreme good which we call moral can therefore 
 be nothing but the idea of the law in itself, in so far as it 
 is this idea which determines the will, and not any con- 
 sequences that are expected to follow. Only a rational 
 being can have such an idea, and hence a man who acts 
 from the idea of the law is already morally good, no 
 
230 THE PHILOSOPHY OF KANT. 
 
 matter whether the consequences which he expects from 
 his action follow or not. 
 
 Now what must be the nature of a law, the idea of 
 
 250 which is to determine the will, even apart from the effects 
 expected to follow, and which is therefore itself entitled 
 to be called good absolutely and without qualification ? 
 As the will must not be moved to act from any desire for 
 the results expected to follow from obedience to a certain 
 law, the only principle of the will which remains is that 
 of the conformity of actions to universal law. In all 
 cases I must act in such a way that I can at the same 
 time will that my maxim should become a universal law. 
 This is what is meant by conformity to law pure and 
 simple ; and this is the principle which serves, and must 
 serve, to determine the will, if the idea of duty is not to 
 be regarded as empty and chimerical. As a matter of 
 fact the judgments which we are wont to pass upon 
 conduct perfectly agree with this principle, and in making 
 them we always have it before our eyes. 
 
 May I, for instance, under the pressure of circum- 
 stances, make a promise which I have no intention of 
 keeping ? The question is not, whether it is prudent to 
 make a false promise, but whether it is morally right. 
 
 251 To enable me to answer this question shortly and con- 
 clusively, the best way is for me to ask myself whether 
 it would satisfy me that the maxim to extricate myself 
 from embarrassment by giving a false promise should 
 have the force of a universal law, applying to others as 
 well as to myself And I see at once, that, while I can 
 certainly will the lie, I cannot will that lying should be a 
 universal law. If lying were universal, there would, 
 properly speaking, be no promises whatever. I might 
 
THE METAPHYSIC OF MORALITY. 23 1 
 
 say that I intended to do a certain thing at some future 
 time, but nobody would believe me, or if he did at the 
 moment trust to my promise, he would afterwards pay 
 me back in my own coin. My maxim thus proves 
 itself to be self-destructive, so soon as it is taken as a 
 universal law. 
 253 Duty, then, consists in the obligation to act from pure 
 reverence for the moral law. To this motive all others 
 must give way, for it is the condition of a will which 
 is good in itself^ and which has a value with which 
 nothing else is comparable. 
 
 There is, however, in man a strong feeling of antagon- 
 ism to the commands of duty, although his reason tells 
 him that those commands are worthy of the highest 
 reverence. For man not only possesses reason, but he 
 has certain natural wants and inclinations, the complete 
 satisfaction of which he calls happiness. These natural 
 inclinations clamorously demand to have their seemingly 
 reasonable claims respected ; but reason issues its 
 commands inflexibly^ refusing to promise anything to the 
 natural desires, and treating their claims with a sort of 
 neglect and contempt. From this there arises a natin-al 
 dialectic, that is, a disposition to explain away the strict 
 laws of duty, to cast doubt upon their validity, or at 
 least, upon their purity and stringency, and in this way 
 to make them yield to the demands of the natural 
 inclinations. 
 
 Thus men are forced to go beyond the narrow circle 
 of ideas within which their reason ordinarily moves, and 
 to take a step into the field of moral philosophy, not 
 indeed from any perception of speculative difficulties, but 
 simply on practical grounds. The practical reason of 
 
232 THE PHILOSOPHY OF KANT. 
 
 men cannot be long exercised any more than the 
 theoretical, without falling insensibly into a dialectic, 
 which compels it to call in the aid of philosophy ; and in 
 the one case as in the other, rest can be found only in a 
 thorough criticism of human reason. 
 
 254 Section II. — Transition from Popular Moral Philosophy 
 
 to the Metaphysic of Morality. 
 
 So far, we have drawn our conception of duty from the 
 manner in which men employ it in the ordinary exer- 
 cise of their practical reason. The conception of duty, 
 however, we must not suppose to be therefore derived from 
 experience. On the contrary, we hear frequent com- 
 plaints, the justice of which we cannot but admit, that no 
 one can point to a single instance in which an action has 
 undoubtedly been done purely from a regard for duty ; 
 that there are certainly many actions which are not 
 opposed to duty, but none which are indisputably done 
 
 2^^fro?n duty and therefore have a moral value. Nothing 
 indeed can secure us against the complete loss of our 
 ideas of duty, and maintain in the soul a well-grounded 
 respect for the moral law, but the clear conviction, that 
 reason issues its commands on its own authority, without 
 
 256 caring in the least whether the actions of men have, as a 
 matter of fact, been done purely from ideas of duty. 
 For reason commands inflexibly that certain actions 
 should be done, which perhaps never have been done ; 
 actions, the very possibility of which may seem doubtful 
 to one who bases everything upon experience. Perfect 
 disinterestedness in friendship, for instance, is demanded 
 of every man, although there may never have been a 
 
THE METAPHYSIC OF MORALITY. 233 
 
 sincere friend ; for pure friendship is bound up with the 
 idea of duty as duty, and belongs to the very idea of 
 a reason which determines the will on a priori grounds, 
 prior to all experience. 
 
 It is, moreover, beyond dispute, that unless we are 
 to deny to morality all truth and all reference to a 
 possible object, the moral law has so wide an application 
 that it is binding, not merely upon man, but upon all 
 ratiofial beings, and not merely under certain contingent 
 conditions, and with certain limitations, but absolutely 
 and necessarily. And it is plain, that no experience 
 could ever lead us to suppose that laws of this apodictic 
 character are even possible. 
 
 257 There is, therefore, no genuine supreme principle of 
 morahty, which is not independent of all experience, and 
 based entirely upon pure reason. If, then, we are to 
 have a philosophy of morality at all, as distinguished from 
 a popular moral philosophy, we may take it for granted 
 without further investigation, that moral conceptions, 
 together with the principles which flow from them, are 
 given a priori and must be presented in their generality 
 {in abstracto), 
 
 258 Such a metaphysic of morality, which must be entirely 
 free from all admixture of empirical psychology, theology, 
 physics and hyperphysics, and above all from all occult 
 or, as we may call them, hypophysical qualities, is not 
 only indispensable as a foundation for a sound theory of 
 duties, but it is also of the highest importance in the 
 practical realization of moral precepts. For the pure 
 idea of duty, unmixed with any foreign ingredient of 
 sensuous desire, in a word, the idea of the moral law, 
 
 259 influences the heart of man much more powerfully 
 
234 THE PHILOSOPHY OF KANT. 
 
 through his reason, which in this way only becomes 
 conscious that it can of itself be practical, than do 
 all the motives which have their source in experience. 
 Conscious of its own dignity, the moral law treats all 
 sensuous desires with contempt, and is able to master 
 them one by one. 
 
 From what has been said it is evident, that all moral 
 conceptions have their seat and origin in reason entirely 
 a priori^ and are apprehended by the ordinary reason of 
 men as well as by reason in its purely speculative activity. 
 We have also seen that it is of the greatest importance, 
 not only in the construction by speculative reason of a 
 theory of morality, but also with a view to the practical 
 conduct of life, to derive the conceptions and laws of 
 morality from pure reason, to present them pure and 
 unmixed, and to mark out the sphere of this whole 
 practical or pure knowledge o^ reason. Nor is it per- 
 missible, in seeking to determine the whole faculty of 
 260 pure practical reason, to make its principles dependent 
 upon the peculiar nature of human reason, as we were 
 allowed to do, and sometimes were even forced to do, in 
 speculative philosophy ; for moral laws must apply to 
 every rational being, and must therefore be derived from 
 the very conception of a rational being as such. 
 
 To show the need of advancing not only from the 
 common moral judgments of men to the philosophical, 
 but from a popular philosophy, which merely gropes its 
 way by the help of examples, to a metaphysic of morality, 
 we must begin at the point where the practical faculty of 
 reason supplies general rules of action, and exhibit clearly 
 the steps by which it attains to the conception of duty. 
 
 Everything in nature acts in conformity with law. 
 
THE METAPHYSIC OF MORALITY. 235 
 
 Only a rational being has the faculty of acting in con- 
 formity with the idea of law, or from principles; only a 
 rational being, in other words, has a will. And as 
 without reason actions cannot proceed from laws, will is 
 simply practical reason. If the will is infallibly deter- 
 mined by reason, the actions of a rational being are 
 subjectively as well as objectively necessary; that is, will 
 must be regarded as a faculty of choosing that only which 
 reason, independently of natural inclination, declares to 
 be practically necessary or good. On the other hand, if 
 the will is not invariably determined by reason alone, but is 
 261 subject to certain subjective conditions or motives, which 
 are not always in harmony with the objective conditions; 
 if the will, as actually is the case with man, is not in 
 perfect conformity with reason ; actions which are recog- 
 nized to be objectively necessary, are subjectively con- 
 tingent. The determination of such a will according to 
 objective laws is therefore called obligatio?i. That is to 
 say, if the will of a rational being is not absolutely good, 
 we conceive of it as capable of being determined by 
 objective laws of reason, but not as by its very nature 
 necessarily obeying them. 
 
 The idea that a certain principle is objective, and 
 binding upon the will, is a command of reason, and the 
 statement of the command in a formula is an imperative. 
 All imperatives are expressed by the word ought, to 
 indicate that the will upon which they are binding is not 
 by its subjective constitution necessarily determined in 
 conformity with the objective law of reason. An inopera- 
 tive says, that the doing, or leaving undone of a certain 
 thing would be good, but it addresses a will which does 
 not always do a thing simply because it is good. Now, 
 
236 THE PHILOSOPHY OF KANT. 
 
 that is practically good which determines the will by ideas 
 of reason, in other words, that which determines it, not 
 by subjective influences, but by principles which are 
 objective, or apply to all rational beings as such. Good 
 and pleasure are quite distinct. Pleasure results from the 
 influence of purely subjective causes upon the will of the 
 subject, and these vary with the susceptibiUty of this or 
 that individual, while a principle of reason is vahd for all. 
 
 A perfectly good will would, like the will of man, stand 
 under objective laws, laws of the good, but it could not 
 262 be said to be under an obligation to act in conformity with 
 those laws. Such a will by its subjective constitution 
 could be determined only by the idea of the good. In 
 reference to the Divine will, or any other holy will, 
 imperatives have no meaning ; for here the will is by its 
 very nature necessarily in harmony with the law, and 
 therefore 02ight has no application to it. Imperatives 
 are formulae, which express merely the relation of objec- 
 tive laws of volition in general to the imperfect will of 
 this or that rational being, as for instance, the will of 
 man. 
 
 Now, all imperatives command either hypothetically or 
 categorically. A hypothetical imperative states that a 
 certain thing must be done, if something else whicli is 
 willed, or at least might be willed, is to be attained. 
 ? The categorical imperative declares that an act is in itself 
 or objectively necessary, without any reference to another 
 end. 
 
 Every practical law represents a possible action as 
 good, and therefore as obligatory for a subject that is 
 capable of being determined to act by reason. Hence 
 all imperatives are formulae for the determination of an 
 
THE METAPHYSIC OF MORALITY. 237 
 
 action which is obligatory according to the principle of a 
 will that is in some sense good. If the action is good 
 only because it is a means to so??iething else, the impera- 
 tive is hypothetical; if the action is conceived to be good 
 in itself, the imperative, as the necessary principle of a 
 will that in itself conforms to reason, is categorical. 
 
 An imperative, then, states what possible action of 
 mine would be good. It supphes the practical rule for a 
 will which does not at once do an act simply because it 
 is good, either because the subject does not know it to be 
 good, or because, knowing it to be good, he is influenced 
 by maxims which are opposed to the objective principles 
 of a practical reason. 
 
 The hypothetical imperative says only that an action 
 263 is good relatively to a CQxtsm possible end or to a certain 
 actual end. In the former case it is problematic, in the 
 latter case assertoric. The categorical imperative, which 
 affirms that an action is in itself or objectively necessary 
 without regard to an end, that is, without regard to any 
 other end than itself, is an apodictic practical principle. 
 
 Whatever is within the power of a rational being may 
 be conceived to be capable of being willed by some 
 rational being, and hence the principles which determine 
 what actions are necessary in the attainment of certain 
 possible ends, are infinite in number. 
 
 Yet there is one thing which we may assume that all 
 finite rational beings actually make their end, and there 
 is therefore one object which may safely be regarded, 
 not simply as something that they may seek, but as 
 something that by a necessity of their nature they actually 
 do seek. This object is happiness. The hypothetical 
 imperative, which affirms the practical necessity of an 
 
238 THE PHILOSOPHY OF KANT. 
 
 264 action as the means of attaining happiness, is assertoric. 
 We must not think of happiness as simply a possible and 
 problematic end, but as an end that we may with confi 
 dence presuppose a priori to be sought by everyone, 
 belonging as it does to the very nature of man. Now 
 skill in the choice of means to his own greatest well-being 
 may be called prudence^ taking the word in its more 
 restricted sense. An imperative, therefore, which relates 
 merely to the choice of means to one's own happiness, 
 that is, a maxim of prudence, must be hypothetical ; it 
 commands an action, not absolutely, but only as a means 
 to another end. 
 
 Lastly, there is an imperative which directly commands 
 an action, without presupposing as its condition that 
 some other end is to be attained by means of that action. 
 This imperative is categorical. It has to do, not with the 
 matter of an action and the result expected to follow from 
 it, but simply with the form and principle from which the 
 action itself proceeds. The action is essentially good if 
 the motive of the agent is good, let the consequences be 
 what they may. This imperative may be called the im- 
 perative of morality. 
 
 26c How are all these imperatives possible ? The question 
 is not. How is an action which an imperative commands 
 actually realized ? but, How can we think of the will as 
 placed under obligation by each of those imperatives? 
 Very little need be said to show how an imperative of skill 
 is possible. He who wills the end, wills also the means in 
 his power which are indispensable to the attainment of the 
 end. Looking simply at the act of will, we must say that 
 this proposition is analytic. If a certain object is to 
 follow as an effect from my volition, my causality must be 
 
THE METAPHYSIC OF MORALITY. 239 
 
 conceived as active in the production of the effect, or as 
 employing the means by which the effect will take place. 
 The imperative, therefore, simply states that in the con- 
 ception of the wilHng of this end there is directly 
 implied the conception of actions necessary to this 
 end. No doubt certain synthetic propositions are 
 required to determine the particular means by which 
 a given end may be attained, but these have noth- 
 ing to do with the principle or act of the will, but 
 merely state how the object may actually be realized. 
 Were it as easy to give a definite conception of happi- 
 
 266 ness as of a particular end, the imperatives of prudence 
 would be of exactly the same nature as the imperatives of 
 skill, and would therefore be analytic. For, we should be 
 able to say, that he who wills the end wills also the only 
 means in his power for the attainment of the end. But, 
 unfortunately, the conception of happiness is so indefinite, 
 that, although every man desires to obtain it, he is unable 
 to give a definite and self-consistent statement of what he 
 actually desires and wills. The truth is, that, strictly speak- 
 ing, the imperatives of prudence are not commands at all. 
 They do not say that actions are objective or necessary, 
 and hence they must be regarded as counsels iconsilia)^ not 
 
 267 as commands {praeceptd) of reason. Still, the impera- 
 tive of prudence would be an analytic proposition, if the 
 means to happiness could only be known with certainty. 
 For the only difference in the two cases is that in the im- 
 perative of skill the end is merely possible, in the impera- 
 tive of prudence it is actually given ; and as in both all 
 that is commanded is the means to an end which is 
 assumed to be willed, the imperative which commands 
 that he who wills the end should also will the means, is in 
 
24© THE PHILOSOPHY OF KANT. 
 
 both cases analytic. There is therefore no real difficulty 
 in seeing how an imperative of prudence is possible. 
 
 The only question which is difficult of solution, is, 
 how the imperative of morality is possible. Here the im- 
 perative is not hypothetical, and hence we cannot derive 
 its objective necessity from any presupposition. Nor 
 must it for a moment be forgotten, that an imperative of 
 this sort cannot be established by instances taken from 
 experience. We must therefore find out by careful in- 
 vestigation, whether imperatives which seem to be cate- 
 gorical may not be simply hypothetical imperatives in 
 disguise. 
 268 One thing is plain at the very outset, namely, that only 
 a categorical imperative can have the dignity of a practical 
 law, and that the other imperatives, while they may no 
 doubt be called principles of the will, cannot be called 
 laws. An action which is necessary merely as a means to 
 an arbitrary end, may be regarded as itself contingent, 
 and if the end is abandoned, the maxim which prescribes 
 the action has no longer any force. An unconditioned 
 command, on the other hand, does not permit the will to 
 choose the opposite, and therefore it carries with it the 
 necessity which is essential to a law. 
 
 It is, however, very hard to see how there can be a 
 categorical imperative or law of morality at all. Such 
 a law is an a priori synthetic proposition, and we cannot 
 expect that there will be less difficulty in showing how a 
 proposition of that sort is possible in the sphere of 
 morality than we have found it to be in the sphere of 
 knowledge. 
 
 In attempting to solve this problem, we shall first of 
 all inquire, whether the mere conception of a categorical 
 
THE METAPHYSIC OF MORALITY. 24 1 
 
 imperative may not perhaps supply us with a formula, 
 which contains the only proposition that can possibly be 
 a categorical imperative. The more difficult question, 
 how such an absolute command is possible at all, will 
 require a special investigation, which must be postponed 
 to the last section. 
 
 If I take the mere conception of a hypothetical imper- 
 ative, I cannot tell what it may contain until the condition 
 under which it applies is presented to me. But I can tell 
 at once from the very conception of a categorical imper- 
 269 ative what it must contain. Viewed apart from the law, 
 the imperative simply affirms that the maxim, or sub- 
 jective principle of action, must conform to the objective 
 principle or law. Now the law contains no condition to 
 which it is restricted, and hence nothing remains but the 
 statement, that the maxim ought to conform to the uni- 
 versality of the law as such. It is only this conformity to 
 law that the imperative can be said to represent as 
 necessary. 
 
 There is therefore but one categorical imperative, which 
 may be thus stated : Act in co?iforinity with that maxim, 
 and that maxi?n Ofily, which you can at the same time will 
 to be a universal law. 
 
 Now, if from this single imperative, as from their prin- 
 ciple, all imperatives of duty can be derived, we shall at 
 least be able to indicate what we mean by the categorical 
 imperative and what the conception of it implies, although 
 we shall not be able to say whether the conception of 
 duty may not itself be empty. 
 
 The universality of the law which governs the succession 
 
 of events, is what we mean by nature, in the most general 
 
 sense, that is, the existence of things, in so far as their 
 
 Q 
 
242 THE PHILOSOPHY OF KANT. 
 
 existence is determined in conformity with universal 
 laws. The universal imperative of duty might therefore 
 be put in this way : Act as if the maxirti fro?n which 
 you act were to become through your will a imiversal la^c 
 of fiature. 
 
 272 If we attend to what goes on in ourselves in every 
 transgression of a duty, we find, that we do not will that 
 our maxim should become a universal law. We find it 
 in fact impossible to do so, and we really will that the 
 opposite of our maxim should remain a universal law, 
 at the same time that we assume the liberty of making an 
 exception in favour of natural inclination in our own case, 
 or perhaps only for this particular occasion. Hence, if 
 we looked at all cases from the same point of view, that 
 is, from the point of view of reason, we should see that 
 there was here a contradiction in our will. The con- 
 tradiction is, that a certain principle is admitted to be 
 necessary objectively or as a universal law, and yet is 
 held not to be universal subjectively, but to admit of 
 exceptions. What we do is, to consider our action at 
 one time from the point of view of a will that is in perfect 
 conformity with reason, and at another time from the 
 point of view of a will that is under the influence of 
 natural inclination. There is, therefore, here no real 
 contradiction, but merely an antagonism of inclination 
 to the command of reason. The universality of the 
 principle is changed into a mere generality, in order that 
 the practical principle of reason may meet the maxim 
 half way. Not only is this limitation condemned by our 
 own impartial judgment, but it proves that we actually 
 recognize the validity of the categorical imperative, and 
 
 273 merely allow ourselves to make a few exceptions in our 
 
THE METAPHYSIC OF MORALITY. 243 
 
 own favour which we try to consider as of no importance, 
 or as a necessary concession to circumstances. 
 
 This much at least we have learned, that if the idea of 
 duty is to have any meaning and to lay down the laws of 
 our actions, it must be expressed in categorical and not 
 in hypothetical imperatives. We have also obtained a 
 clear and distinct conception (a very important thing), 
 of what is implied in a categorical imperative which con- 
 tains the principle of duty for all cases, granting such an 
 imperative to be possible at all. But we have not yet 
 been able to prove a p7'iori, that there actually is such an 
 imperative ; that there is a practical law which commands 
 absolutely on its own authority, and is independent of 
 all sensuous impulses; and that duty consists in obedience 
 to this law. 
 
 In seeking to reach this point, it is of the greatest im- 
 portance to observe, that the reality of this principle cannot 
 possibly be derived from the peculiar constitution of human 
 nature. For by duty is meant the practically unconditioned 
 necessity of an act, and hence we can show that duty is a 
 law for the will of all human beings, only by showing 
 that it is applicable to all rational beings, or rather to all 
 rational beings to whom an imperative applies at all. 
 274 The question, then, is this: Is it a necessary Ibm for 
 all rational beings, that they must always estimate the 
 value of their actions by asking whether they can will that 
 their maxims should serve as universal laws ? If there 
 is such a law, it must be possible to prove entirely a priori, 
 that it is bound up with the very idea of the will of a 
 rational being. To show that there is such a connection 
 we must, however reluctantly, take a step into the realm of 
 metaphysic ;. not, however, into the realm of speculative 
 
244 THE PHILOSOPHY OF KANT. 
 
 275 philosophy, but into the metaphysic of morality. For 
 we have here to deal with objective practical laws, 
 and therefore with the relation of the will to itself, in so 
 far as it is determined purely by reason. All relation of 
 the will to what is empirical is excluded as a matter of 
 course, for if reason determines the relation entirely by itself^ 
 it must necessarily do so a priori. 
 
 Will is conceived of as a faculty of determining itself 
 to action in accordance with the idea of certain laws. Such 
 a faculty can belong only to a rational being. Now that 
 which serves as an objective principle for the self-deter- 
 mination of the will is an end^ and if this end is given 
 purely by reason, it must hold for all rational beings. On 
 the other hand, that which is merely the condition of the 
 possibility of an action the effect of which is the end, is 
 called the means. The subjective ground of desire is natural 
 inclination, the objective ground of volition is a motive ; 
 hence there is a distinction between subjective ends, 
 which depend upon natural inclination, and objective 
 ends, which are connected with motives that hold for every 
 rational being. Practical principles that abstract from 
 all subjective ends are formal ; those that presuppose 
 subjective ends, and therefore natural inclinations, are 
 material. The ends which a rational being arbitrarily sets 
 before himself as material ends to be produced by his 
 actions, are all merely relative ; for that which gives to 
 
 276 them their value is simply their relation to the peculiar 
 susceptibility of the subject. They can therefore yield 
 no universal and necessary principles, or practical laws, 
 applicable to all rational beings, and binding upon every 
 will. Upon such relative ends, therefore, only hypothetical 
 imperatives can be based. 
 
THE METAPHYSIC OF MORALITY. 245 
 
 Suppose, however, that there is something the existence 
 of which has in itself an absolute value, something 
 which, as an end in itself^ can be a ground of definite 
 laws ; then, there would lie in that, and only in that, the 
 ground of a possible categorical imperative or practical 
 law. 
 
 Now, I say, that man, and indeed every rational being 
 as such, exists as an end in himself, not merely as a 
 means to be made use of by this or that will, and there- 
 fore man in all his actions, whether these are directed 
 towards himself or towards other rational beings, must 
 always be regarded as an end. No object of natural 
 desire has more than a conditioned value ; for if the 
 natural desires, and the wants to which they give rise, did 
 not exist, the object to which they are directed would have 
 no value at all. So far are the natural desires and wants 
 from having an absolute value, so far are they from being 
 sought simply for themselves, that every rational being 
 must wish to be entirely free from their influence. The 
 value of every object which human action is the means 
 of obtaining, is, therefore, always conditioned. And 
 even beings whose existence depends upon nature, not 
 upon our will, if they are without reason, have only the 
 relative value of means, and are therefore called things. 
 Rational beings, on the other hand, are Q.z)\t^ persons^ 
 because their very nature shows them to be ends in 
 themselves, that is, something which cannot be made use 
 of simply as a means. A person being thus an object of 
 respect, a certain limit is placed upon arbitrary will. 
 Persons are not purely subjective ends, whose existence 
 has a value /?r us as the eftect of our actions, but they 
 are objective ends^ or beings whose existence is an end 
 
246 THE PHILOSOPHY OF KANT. 
 
 in itself, for which no other end can be substituted. If 
 all value were conditioned, and therefore contingent, it 
 would be impossible to show that there is any supreme 
 practical principle whatever. 
 
 If, then, there is a supreme practical principle, a 
 277 principle which in relation to the human will is a 
 categorical imperative, it must be an objective principle of 
 the will, and must be able to serve as a universal 
 practical law. For, such a principle must be derived 
 from the idea of that which is necessarily an end for 
 every one because it is an end in itself. Its foundation 
 is this, that ratio?ial nature exists as a?i e?id in itself. 
 Man necessarily conceives of his own existence in this 
 way, and so far this is a subjective principle of human 
 action. But in this way also every other rational being 
 conceives of his own existence, and for the very same 
 reason ; hence the principle is also objective, and from 
 it, as the highest practical ground, all laws of the will 
 must be capable of being derived. The practical impera- 
 tive will therefore be this : Act so as to use hic7nanity^ 
 whether in your own person or in the person of another, 
 ahvays as a?i end, never as merely a ?neans. 
 
 The principle, that humanity and every rational nature 
 is an end in itself, is not borrowed from experience. For, 
 in the first place, because of its universality it applies to 
 all rational beings, and no experience can apply so widely. 
 In the second place, it does not regard humanity sub- 
 jectively, as an end of man, that is, as an object which 
 the subject of himself actually makes his end, but as an 
 objective end, which ought to be regarded as a law that 
 constitutes the supreme limiting condition of all subjective 
 ends, and which must therefore have its source in pure 
 
THE METAPHYSIC OF MORALITY. 247 
 
 reason. The objective ground of all practical laws 
 consists in the rule and the form of universality, which 
 makes them capable of serving as laws, but their 
 sicbjedive ground consists in the end to which they are 
 directed. Now, by the second principle, every rational 
 being, as an end in himself, is the subject of all ends. 
 From this follows the third practical principle of the will, 
 which is the supreme condition of its harmony with 
 universal practical reason, namely, the idea of the will of 
 every rational being as a will which lays down universal 
 laws of actiofi. 
 
 280 This formula implies, that a will which is itself the 
 supreme lawgiver cannot possibly act from interest of any 
 sort in the law, although no doubt a will may stand 
 under the law, and may yet be attached to it by the bond 
 of interest. 
 
 At the point we have now reached, it does not seem 
 surprising that all previous attempts to find out the 
 principle of morality should have ended in failure. It 
 was seen that man is bound under law by duty, but it did 
 not strike anyone, that the universal system of laws to 
 which he is subject are laws which he imposes upon 
 himself and that he is only under obligation to act 
 in conformity with his own will, a will which by the 
 
 281 purpose of nature prescribes universal laws. Now so long 
 as man is thought to be merely subject to law, no matter 
 what the law may be, he must be regarded as stimulated 
 or constrained to obey the law from interest of some 
 kind ; for as the law does not proceed from his own will, 
 there must be somethi7ig external to his will which 
 compels him to act in conformity with it. This perfectly 
 necessary conclusion frustrated every attempt to find a 
 
248 THE PHILOSOPHY OF KANT. 
 
 supreme principle of duty. Duty was never established, 
 but merely the necessity of acting from some form of 
 interest, private or public. The imperative was therefore 
 necessarily always conditioned, and could not possibly 
 have the force of a moral command. The supreme 
 principle of morality I shall therefore call the principle 
 of the auto7i07ny of the will, to distinguish it from all other 
 principles, which I call principles of heiero.nomy. 
 
 The conception that every rational being in all the 
 maxims of his will must regard himself as prescribing 
 universal laws, by reference to which himself and all his 
 actions are to be judged, leads to a cognate and very 
 fruitful conception, that of a ki72gdo?n of ends. 
 
 Sy kingdom, I mean the systematic combination of 
 different rational beings through the medium of common 
 laws. Now, laws determine certain ends as universal, 
 and hence, if abstraction is made from the individual 
 differences of rational beings, and from all that is peculiar 
 to their private ends, we get the idea of a complete 
 totality of ends combined in a system ; in other words, we 
 are able to conceive of a kingdom of ends, which 
 conforms to the principles formulated above. 
 
 All rational beings stand under the law, that each 
 should treat himself and others, never simply as 7neans, 
 but always as at the same time ends in themselves. Thus 
 there arises a systematic combination of rational beings 
 through the medium of common objective laws. This 
 may well be called a kingdom of ends, because the 
 object of those laws is just to relate all rational beings 
 to one another as ends and means. Of course this 
 kingdom of ends is merely an ideal. 
 282 Morality, then, consists in the relation of all action to 
 
THE METAPHYSIC OF MORALITY. 249 
 
 the system of laws which alone makes possible a kingdom 
 of ends. These laws must belong to the nature of every 
 rational being, and must proceed from his own will. The 
 principle of the will, therefore, is, that no action should 
 be done from any other maxim than one which is con- 
 sistent with a universal law. This may be expressed in 
 the formula : Act so that the will may regard itself as in 
 its maxifns laying down universal laws. Now, if the 
 maxims of rational beings are not by their very nature in 
 harmony with this objective principle, the principle of 
 a universal system of laws, the necessity of acting in 
 conformity with that principle is called practical obligation 
 or duty. No doubt duty does not apply to the sovereign 
 will in the kingdom of ends, but it applies to every 
 284 member of it, and to all in equal measure. Autonomy 
 is thus the foundation of the moral value of man and of 
 every other rational being. 
 
 The three ways in which the principle of morality has 
 been formulated are at bottom simply different statements 
 of the same law, and each implies the other two. 
 
 292 An absolutely good will, then, the principle of which 
 must be a categorical imperative, will be undetermined as 
 regards all objects, and will contain merely \htform of 
 volition in general, a form which rests upon the autonomy 
 
 293 of the will. The one law which the will of every rational 
 being imposes upon itself, and imposes without reference 
 to any natural impulse or any interest, is, that the maxims 
 of every good will must be capable of being made a 
 universal law. 
 
 How such an a priori synthetic practical proposition is 
 possible., and why it is necessary, is a problem which it is 
 not the task of a metaphysic of morality to solve. We 
 
250 THE PHILOSOPHY OF KANT. 
 
 have not even affirmed it to be true, much less have we 
 attempted to prove its truth. To prove that practical 
 reason is capable of being employed synthetically, and 
 that morality is not a mere fiction of the brain, requires 
 us to enter upon a criticism of the faculty of practical 
 reason itself. In the next section we shall state the main 
 points which must be proved in a Critique of Practical 
 Reason, so far as is necessary for our present purpose. 
 
 294 Section TIT. — Transition fro7n the Metaphysic of Morality 
 to the Critique of Practical Reason. 
 
 The Idea of Freedo)n as the Key to the Autonomy of 
 
 the Will. 
 
 The will is the causality of living beings in so far as 
 they are rational. Freedom is that causality in so far as 
 it can be regarded as efficient without being determined to 
 activity by any cause other than itself. Natural necessity 
 is the property of all non-rational beings to be determined 
 to activity by some cause external to themselves. 
 
 The definition of freedom just given is negative^ and 
 therefore it does not tell us what freedom is in itself; but 
 it prepares the way for a positive conception of a more 
 specific and more fruitful character. The conception of 
 causality carries with it the conception of determination 
 by law (Gesetz), for the effect is conceived as determined 
 (gesetzt) by the cause. Hence freedom must not be 
 regarded as lawless (gesetzlos), but simply as independent 
 of laws of nature. A free cause does conform to un- 
 changeable laws, but these laws are peculiar to itself; 
 and, indeed, apart from law a free will has no meaning 
 
THE METAPHYSIC OF MORALITY. 25 1 
 
 whatever. A necessary law of nature, as we have seen, 
 implies the heteronomy of efficient causes ; for no effect 
 is possible at all, unless its cause is itself determined to 
 activity by something else. What, therefore, can freedom 
 possibly be but autonomy, that is, the property of the 
 will to be a law to itself? Now, to say that the will in 
 all its actions is a law to itself, is simply to say that its 
 295 principle is, to act from no other maxim than that the 
 object of which is itself as a universal law. But this is 
 just the formula of the categorical imperative and the 
 principle of morality. Hence a free will is the same 
 thing as a will that conforms to moral laws. 
 
 If, then, we start from the presupposition of freedom 
 of the will, we can derive morality and the principle of 
 morality simply from an analysis of the conception of 
 freedom. Yet the principle of morality, namely, that an 
 absolutely good will is a will the maxim of which can 
 always be taken as itself a universal law, is a synthetic 
 proposition. For by no possibility can we derive this 
 property of the maxim from an analysis of the conception 
 of an absolutely good will. The transition from the con- 
 ception of freedom to the conception of morality can be 
 made only if there is a third proposition which connects 
 the other two in a synthetic unity. Th.Q positive conception 
 of freedom yields this third proposition, and not the con- 
 ception of nature, in which a thing is related causally 
 only to something else. What this third proposition is 
 to which freedom points, and of which we have an a 
 priori idea, can be made clear only after some preliminary 
 investigation. 
 
252 THE PHILOSOPHY OF KANT. 
 
 Freedom is a property of all Rational Beings. 
 
 It cannot in any way be proved that the will of man is 
 free, unless it can be shown that the will of all rational 
 beings is free. For morality is a law for us only in so 
 far as we are rational beings, and therefore it must apply to 
 all rational beings. But morality is possible only for a 
 free being, and hence it must be proved that freedom also 
 29^ belongs to the will of all rational beings. Now I say, 
 that a being who cannot act except U7ider the idea of 
 freedoi7t^ must for that very reason be regarded as free so 
 far as his actions are concerned. In other words, even 
 if it cannot be proved by speculative reason that his will 
 is free, all the laws that are inseparably bound up with 
 freedom must be viewed by him as laws of his will. And 
 I say, further, that we must necessarily attribute to every 
 rational being that has a will the idea of freedom, because 
 every such being always acts under that idea. A rational 
 being we must conceive as having a reason that is 
 practical, that is, a reason that has causality with regard 
 to its objects. Now, it is impossible to conceive of a 
 reason which should be consciously biassed in its judg- 
 ments by some influence from without, for the subject 
 would in that case regard its judgments as determined, not 
 by reason, but by a natural impulse. Reason must there- 
 fore regard itself as the author of its principles of action, and 
 as independent of all external influences. Hence, as prac- 
 tical reason, or as the will of a rational being, it must be 
 regarded by itself as free. The will of a rational being, in 
 other words, can be his own will only if he acts under the 
 idea of freedom, and therefore this idea must in the 
 practical sphere be ascribed to all rational beings. 
 
THE METAPHYSIC OF MORALITY. 253 
 
 The Interest connected with Moral Ideas. 
 
 We have at last succeeded in reducing the true con- 
 ception of morality to the idea of freedom. This, how- 
 ever, does not prove that man actually is free, but only 
 that, without presupposing freedom, we cannot conceive 
 of ourselves as rational beings, who are conscious of 
 causality with respect to our actions, that is, as endowed 
 
 297 with will. We have also found that on the same ground 
 all beings endowed with reason and will must determine 
 themselves to action under the idea of their freedom. 
 
 From the presupposition of the idea of freedom there 
 also followed the consciousness of a law of action, the 
 law that our subjective principles of action, or maxims, 
 must always be of such a character that they have the 
 validity of objective or universal principles, and can be 
 taken as universal laws imposed upon our will by our- 
 selves. But why, it may be asked, should I subject 
 myself to this principle simply as a rational being, and 
 why, therefore, should all other beings who are endowed 
 with reason come under the same principle ? Admitting 
 that I am not forced to do so by interest — which indeed 
 would make a categorical imperative impossible — yet I 
 must take an interest in that principle and see how I come 
 to subject myself to it. 
 
 It looks as if we had, strictly speaking, shown merely 
 that in the idea of freedom the moral law must be pre- 
 supposed in order to explain the principle of the auto- 
 nomy of the will, without being able to prove the reality 
 and objectivity of the moral law itself. 
 
 298 It must be frankly admitted, that there is here a sort 
 of circle from which it seems impossible to escape. We 
 
254 THE PHILOSOPHY OF KANT. 
 
 assume that as efficient causes we are free, in order to ex- 
 plain how in the kingdom of ends we can be under moral 
 laws; and then we think of ourselves as subject to moral 
 laws, because we have ascribed to ourselves freedom of 
 will. Freedom of will and self-legislation of will are both 
 autonomy, and, therefore, they are conceptions which im- 
 ply each other ; but, for that very reason, the one cannot 
 be employed to explain or to account for the other. 
 
 301 How is a Categorical Imperative possible 1 
 
 As an intelligence, a rational being views himself as a 
 member of the intelligible world, and it is only as an 
 efficient cause belonging to this world that he speaks of 
 his own causality as will. On the other hand, he is con- 
 scious of himself as also a part of the world of sense, and 
 in this connection his actions appear as mere phenomena 
 which that causality underlies. Yet he cannot trace back 
 his actions as phenomena to the causality of his will, be- 
 cause of that causality he has no knowledge ; and he is 
 thus forced to view them as if they were determined 
 merely by other phenomena, that is, by natural desires 
 and inclinations. Were a man a member only of the in- 
 telligible world, all his actions would be in perfect agree- 
 ment with the autonomy of the will ; were he merely a 
 part of the world of sense, they would have to be regarded 
 as completely subject to the natural law of desire and 
 inclination, and to the heteronomy of nature. The former 
 would rest upon the supreme principle of morality, the 
 latter upon that of happiness. But it must be ob- 
 served that the intelligible world is the conditioji of the 
 world of se?ise^ and, therefore, of the laws of that world. 
 
THE METAPHYSIC OF MORALITY. 255 
 
 And as the will belongs altogether to the intelligible 
 world, it is the intelligible world that prescribes the laws 
 which the will directly obeys. As an intelligence, I am 
 therefore subject to the law of the intelligible world, that 
 is, to reason, notwithstanding the fact that I belong on the 
 other side of my nature to the world of sense. Now, as 
 subject to reason, which in the idea of freedom contains 
 the law of the intelligible world, I am conscious of being 
 302 subject to the autonomy of the will. The laws of the 
 intelligible world I must therefore regard as imperatives, 
 and the actions conformable to this principle as duties. 
 
 The explanation of the possibility of categorical imper- 
 atives, then, is, that the idea of freedom makes me a 
 member of the intelligible world. Were I a member of 
 no other world, all my actions would as a matter of fact 
 always conform to the autonomy of the will. But as I 
 perceive myself to be also a member of the world of sense, 
 I can say only, that my actions ought to conform to the 
 autonomy of the will. The categorical ought is thus an 
 a priori synthetic proposition. To my will as affected by 
 sensuous desires, there is added synthetically the idea of 
 my wilLas belonging to the intelligible world, and there- 
 fore as pure and self-determining. The will as rational 
 is therefore the supreme condition of the will as sensuous. 
 The method of explanation here employed is similar to 
 that by which the categories were deduced. For the 
 a priori synthetic propositions, which make all knowledge 
 of nature possible, depend, as we have seen, upon the 
 addition to perceptions of sense of the pure conceptions 
 of understanding, which, in themselves, are nothing but 
 the form of law in general. 
 
256 THE PHILOSOPHY OF KANT. 
 
 303 Lif?nts of Practical Philosophy. 
 
 Freedom is only an idea of reason, and therefore its 
 objective reality is doubtful. Thus there arises a dialectic 
 of practical reason. The freedom ascribed to the will 
 seems to stand in contradiction with the necessity of 
 
 3°4 nature. It is, therefore, incumbent upon speculative 
 philosophy at least to show that we think of man in one 
 sense and relation when we call him free, and in another 
 sense and relation when we view him as a part of nature, 
 and as subject to its laws. But this duty is incumbent 
 upon speculative philosophy only in so far as it has to 
 clear the way for practical philosophy. 
 
 306 In thinking itself into the intelligible world, practical 
 reason does not transcend its proper limits, as it would do 
 if it tried to know itself directly by means of perception. 
 In so thinking itself, reason merely conceives of itself 
 negatively as not belonging to the world of sense, without 
 giving any laws to itself in determination of the will. 
 There is but a single point in which it is positive, namely, 
 in the thought that freedom, though it is a negative 
 determination, is yet bound up with a positive faculty, 
 and, indeed, with a causality of reason which is called 
 will. In other words, will is the faculty of so acting that 
 the principle of action should conform to the essential 
 nature of a rational motive, that is, to the condition that 
 the maxim of action should have the universal validity of 
 a law. Were reason, however, to derive an object of will, 
 that is, a motive, from the intelligible world, it would 
 transcend its proper limits, and would make a pretence 
 of knowing something of which it knew nothing. The 
 conception of an intelligible world is therefore merely a 
 
THE METAPHYSIC OF MORALITY. 257 
 
 point of view beyond the world of sense, at which reason 
 sees itself compelled to take its stand in order to think 
 itself as p7'acticaL This conception would not be possible 
 at all if the sensuous desires were sufficient to determine 
 the action of man. It is necessary, because otherwise man 
 would not be conscious of himself as an intelligence, and, 
 therefore, not as a rational cause acting through reason or 
 operating freely. This thought undoubtedly involves the 
 idea of an order and a system of laws other than the order 
 and laws of nature, which concern only the world of sense. 
 Hence it makes necessary the conception of an intellig- 
 ible world, a world which comprehends the totality of 
 rational beings as things in themselves. Yet it in no way 
 entitles us to think of that world otherwise than in its 
 formal condition, that is, to conceive of the maxims of 
 the will as conformable to universal laws. 
 
 Reason would, therefore, completely transcend its 
 proper limits, if it should undertake to explain how pure 
 reason can be practical, or, what is the same thing, to 
 explain how freedom is possible. 
 307 We can explain nothing but that which we can reduce 
 to laws, the object of which can be presented in a possible 
 experience. Freedom, however, is a mere idea, the 
 objective reality of which can in no way be presented in 
 accordance with laws of nature, and, therefore, not in any 
 possible experience. It has merely the necessity of a 
 presupposition of reason, made by a being who believes 
 himself to be conscious of a will, that is, of a faculty 
 distinct from mere desire. The most that we can do is 
 to defend freedom by removing the objections of those 
 who claim to have a deeper insight into the nature of 
 things than we can pretend to have, and who, therefore, 
 
 R 
 
258 THE PHILOSOPHY OF KANT, 
 
 declare that freedom is impossible. It would no doubt 
 be a contradiction to say that in its causality the will is 
 entirely separated from all the laws of the sensible world. 
 But the contradiction disappears, if we say, that behind 
 phenomena there are things in themselves, which, though 
 they are hidden from us, are the condition of phenomena ; 
 and that the laws of action of things in themselves 
 naturally are not the same as the laws under which their 
 phenomenal manifestations stand. 
 311 While, therefore, it is true that we cannot comprehend 
 the practical unconditioned necessity of the moral impera- 
 tive, it is also true that we can comprehend its incompre- 
 hensibility ; and this is all that can fairly be demanded 
 of a philosophy which seeks to reach the principles which 
 determine the limits of human reason. 
 
THE CRITIQUE OF PRACTICAL REASON. 
 
v., THE CRITIQUE OF PRACTICAL REASON, 
 
 19 BOOK I.— ANALYTIC OF PURE PRACTICAL 
 
 REASON 
 
 Chapter I. — The Principles of Pure Practical 
 
 Reason. 
 
 I. Defi7iition. 
 
 Practical principles are propositions, which contain a 
 general determination of the will, a determination that 
 has under it various practical rules. They are subjective 
 principles, or maxitjis^ if the condition is regarded by the 
 subject as holding only for his own will ; they are 
 objective principles, or practical laws^ if the condition is 
 seen to be objective, or to hold for the will of every 
 rational being. 
 
 Re?nark. 
 
 A man may make it his maxim to avenge every injury 
 that is done to him, while yet he sees that this is not a 
 2 1 practical law, but is merely a maxim of his own. But tell 
 him, that he ought never to make a deceitful promise, and 
 he recognizes that here the rule concerns only his will, and 
 holds whether the special ends he may have in view can 
 be attained by obeying the rule or not. And if this rule 
 is practically right, it is a categorical imperative, and 
 therefore a law. Hence practical laws concern the will 
 
262 THE PHILOSOPHY OF KANT. 
 
 only, and not at all the consequences which may follow 
 in the world of sense through its causality. 
 
 2. Theorem i. 
 
 All practical principles that presuppose an object or 
 matter of desire as motive of the will, are empirical, and 
 such principles cannot yield practical laws. 
 
 By the matter of desire I mean an object which it is 
 desired to realize. Now, if the desire for this object 
 precedes the practical rule, and is the condition of its 
 being made a principle, I affirm, in the first place, that 
 the principle must be empirical. For the motive of the will 
 is in that case the idea of an object, which is so related 
 to the subject, that the faculty of desire is determined 
 2 to activity by it. But this means, that the subject 
 expects to receive pleasure from the realization of the 
 object. This pleasure must therefore be presupposed as 
 the condition without which the will would not be 
 determined to activity. Now, it is impossible to know 
 «/r/i7r/ whether an object will bring pleasure or pain, or 
 neither. The motive must therefore be altogether em- 
 pirical, and so also must be the material principle Vv^hich 
 is based upon it. 
 
 In the second place, a principle that rests entirely upon 
 the subjective condition of a peculiar sensibility to 
 pleasure or pain, may indeed serve as a 7naxwi for the 
 sensitive subject, but it cannot be a law even for him. 
 Such a principle can therefore never yield a practical 
 law. 
 
THE CRITIQUE OF PRACTICAL REASON. 263 
 
 3. Theorem 2. 
 
 All material practical principles are, as such, of one 
 and the same kind, and are reducible to the general 
 principle of self-love or individual happiness. 
 
 Pleasure in the idea of the existence of a thing, in so 
 far as it is to determine the desire for that thing, rests 
 upon the sensibility of the subject, because it is dependent 
 upon the existence of the thing desired. Pleasure there- 
 fore belongs to sense or feehng, not to understanding ; 
 for understanding implies a relation of the idea to an 
 object through conceptions, not to the subject by means 
 of feelings. Hence pleasure is practical only in so far as 
 the feeling of pleasure, which the subject expects to ex- 
 perience in the realization of the thing, determines the 
 faculty of desire. Now the consciousness on the part of 
 a rational being of agreeable feeling as continuing un- 
 broken through the whole of his life, is happiness, and the 
 principle which makes happiness the liighest motive of 
 the will, is the principle of self-love. Therefore all 
 material principles, which assign as the motive of the will 
 23 the pleasure or pain expected from the realization of some 
 object, are all of the same kind, inasmuch as they all 
 belong to the principle of self-love or individual happiness. 
 
 Corollary. 
 
 All material practical rules assume that the lower faculty 
 of desire determines the will, and if there were no ptirely 
 formal laws sufficient to determine the will, there would 
 be no higher faculty of desire at all. 
 
264 THE PHILOSOPHY OF KANT. 
 
 Retnark i. 
 
 It is a matter for surprise that men of intelligence 
 should imagine that a real distinction may be drawn 
 between the lower and the higher faculty of desire, on the 
 ground that some ideas which are associated with the 
 feeling of pleasure have their source in sense and others in 
 understa7idi?ig. For however the ideas themselves may 
 differ from one another, and whether they proceed 
 from understanding or even from reason, as distinguished 
 from sense, the feeling of pleasure, which is the real 
 motive by which the will is determined to act, is always 
 the same in kind, not only because it can be known only 
 empirically, but because in every desire the same vital 
 energy is always expressed. The only difference between 
 25 pleasures is therefore one of degree. However under- 
 standing and reason may be employed in furthering 
 individual happiness, the principle itself contains no 
 other motives than those which act upon the will through 
 the lower faculty of desire. We are therefore forced to 
 say, either that there is no higher faculty of desire at 
 all, or \ki2X pure reason is of itself practical, that is, is able 
 to determine the will by the mere form of the practical 
 rule, independently of all feeling, and therefore of all 
 ideas of pleasure and pain. 
 
 Remark 2. 
 
 27 Even if all finite rational beings were perfectly agreed 
 in regard to the objects that are fitted to bring pleasure 
 or pain, and also in regard to the means of attaining those 
 objects, the principle of self-love could not claim to be a 
 practical law. For the motive would not even then cease 
 
THE CRITIQUE OF PRACTICAL REASON. 265 
 
 to be subjective and empirical, nor would it possess the 
 necessity which every law implies, namely, the objective 
 necessity based upon a priori grounds. 
 
 28 4. Theorem 3. 
 
 If a rational being is to think of his maxims as prac- 
 tical universal laws, he must think of them as determining 
 the will, not by their matter, but simply by their Torm. 
 
 The matter of a practical principle is the object of will. 
 This object either determines the will, or it does not. In 
 the former case, the rule of the will is subjected to an 
 empirical condition; in other words, the idea which 
 determines the will is dependent upon a certain feeling of 
 pleasure or pain; and hence there can be no practical 
 law. But, if all matter is removed from the law, that is, 
 every object that is capable of determining the will, 
 nothing is left but the mere form of a universal system of 
 law. Either, therefore, a rational being cannot think of 
 his subjective practical principles or maxims as universal 
 laws ; or it is the mere form of his maxims which makes 
 them practical laws, and enables them to belong to a 
 system of universal laws. 
 
 Remark, 
 
 Any man of common sense can at once see without 
 being told, what form of maxim is fitted to serve as a 
 universal law, and what is not. Suppose, for instance, 
 that my maxim is to make as much money as I can. A 
 man at his death has left in my hands property in trust 
 for others, but he has not left in writing anything to show 
 29 that I received the money. Can I interpret my maxim 
 
266 THE PHILOSOPHY OF KANT. 
 
 in this way, that every one may deny having received a 
 deposit, if there is nothing to show that he has received 
 it? It is at once obvious, that such a principle, the 
 moment it is stated in the form of a law, becomes self- 
 contradictory; for if it were a universal principle of action, 
 no one would ever leave his money in trust. What is re- 
 cognized as a practical law must be universally appHcable; 
 in fact, this is an identical, and therefore a self-evident 
 proposition. If my will is to stand under a practical law, 
 I cannot regard my natural inclination — in the present 
 case my avarice — as a motive that harmonizes with a 
 universal practical law. So far is such a principle from 
 being in harmony with a universal system of laws, that it 
 destroys itself when it is stated in the form of a universal 
 law. 
 
 30 5. Problem I. 
 
 Granting that the mere form of universal law is the only 
 form of a maxim that is sufficient to determine a will; the 
 problem is, to find out what must be the nature of a will 
 that is determined purely by that form. 
 
 The mere form of the law can be apprehended only by 
 reason, and hence it is not an object of sense, nor can it 
 belong to phenomena. The idea which is to determine 
 the will is essentially different from the principles by which 
 events follow one another in nature in accordance with 
 the law of causality, for each of these events is determined 
 by that which is itself a phenomenon. Now, if nothing 
 else can serve as a law to the will but the mere form of 
 universal law, the will must be entirely independent of 
 the law which governs phenomena in their relation to one 
 another, namely, the law of natural causality. But inde- 
 
THE CRITIQUE OF PRACTICAL REASON. 267 
 
 pendence of natural law is fi-eedom^ in the strictest or 
 transcendental sense of the word. Therefore, a will for 
 which only the mere form of universal law can serve as 
 the form of its maxim, must be a free ^ill. 
 
 6. Problem II. 
 
 Granting that a will is free ; the problem is to find the 
 law which alone is fitted to determine it necessarily. 
 
 The matter of the practical law, or the object of the 
 maxim, can be given only in experience. Now a free 
 will must be independent of all empirical or sensuous 
 conditions, and yet it must be capable of being deter- 
 mined to activity. Such a will must find its principle of 
 determination in the law itself, abstracted from all the 
 matter of the law. But apart from its matter, the law 
 contains nothing but the form of law in general. 
 31 Therefore, the form of law in general, in so far as it 
 is contained in a maxim, is the only thing capable 
 of determining a free will. 
 
 Remark. 
 
 Freedom and unconditioned practical law mutually 
 refer to each other. I do not here ask, whether in their 
 actual nature they are different, or whether, on the 
 contrary, an unconditioned law is merely pure practical 
 reason as conscious of itself, and therefore identical with 
 the positive conception of freedom. My question is, 
 whether our k?iowledge of that which is unconditionally 
 practical, starts from the idea of freedom or from the idea 
 of a practical law. Now the idea of freedom cannot be 
 primary. For, as our first conception of freedom is 
 
268 THE PHILOSOPHY OF KANT. 
 
 negative, we cannot be directly conscious of it ; nor 
 again, can it be derived from experience, for experience 
 gives us a knowledge only of the law of phenomena, or 
 the mechanism of jiature, and nature is just the opposite 
 of freedom. It is therefore of the ??ioral law that we are 
 primarily and directly conscious. This law we apprehend 
 by thinking of maxims of the will in their form. Thus 
 reason presents the moral law as a principle of action, 
 which no sensuous condition can outweigh, nay, as a 
 principle which is completely independent of all sensuous 
 conditions. The consciousness of the moral law, there- 
 4n fore, leads inevitably to the conception of freedom. To 
 prevent misunderstanding it must be observed, that 
 while freedom is the ratio essefidi of the moral law, the 
 moral law is the ratio cognoscendi of freedom. The idea 
 of freedom is certainly not self-contradictory ; but, if 
 reason did not first give us a distinct idea of the moral 
 law, we should not feel justified in supposing that there 
 was such a thing as freedom at all. 
 
 32 7. Supreme Law of Pure Practical Reason. 
 
 Act so that the maxims of your will may be in perfect 
 harmony with a universal system of laws. 
 
 Re77iark, 
 
 33 The consciousness of this law may be called a fact of 
 reason. For it is impossible to derive it from any datum 
 known by reason antecedently to it, as, for instance, the 
 consciousness of freedom. It forces itself upon us as an 
 a priori synthetic proposition, which is independent of 
 any perception, either pure or empirical. If we were 
 
THE CRITIQUE OF PRACTICAL REASON. 269 
 
 permitted to presuppose freedom of will, it would no 
 doubt be an analytic proposition ; but the presupposition 
 of freedom as a positive conception could only be 
 justified by an intellectual perception, and we have no 
 ground to assume such a perception. To apprehend 
 this law in its true nature, however, it must be carefully 
 observed, that it is not given in the sense that it can 
 be verified in experience, but only in the sense that it is 
 the one fact of pure reason. It is therefore in relation to 
 this fact that reason proclaims itself to be the source 
 of law {sic volo, sic jubed). 
 
 Corollary. 
 
 Pure reason is practical purely of itself, and gives to 
 man a universal law, which is called the moral law. 
 
 Remark, 
 
 4 The principle of morality prescribes a universal law, 
 which is independent of all subjective differences, and 
 which serves as the supreme formal ground for the 
 determination of the will. For this very reason, that 
 principle is a law for all rational beings which have a will. 
 Hence it is not restricted to man, but holds for all finite 
 beings who have reason and will, and includes even the 
 Infinite Being, as the Supreme Intelligence. In the case 
 of finite beings, the law takes the form of an imperative ; 
 for while we may presuppose in them a pure will, we 
 cannot presuppose that, with their sensuous wants and 
 desires, they are possessed of a holy will, that is, a will 
 which is incapable of maxims that contradict the moral 
 law. The Supreme Intelligence, on the other hand, is 
 
270 THE PHILOSOPHY OF KANT. 
 
 incapable of any maxim which is not also an objective 
 law, and the conception of holiness which must therefore 
 be attributed to that Being, places Him, not indeed above 
 all practical law, but above all laws that restrict His will, 
 and therefore above obligation and duty. Yet hohness of 
 will is for finite beings a practical conception, serving 
 35 as the necessary ideal to which they can approximate 
 indefinitely. This ideal the pure moral law, which is 
 therefore itself called holy, rightly insists upon holding 
 ever before their eyes. To be assured of the infinite 
 progress of one's maxims and of their unchangeability as 
 the cause of a continuous advance constitutes virtue ; and 
 this is the utmost that the practical reason of finite beings 
 can bring about. Virtue, at least as a faculty acquired 
 naturally, can never be complete, for we cannot have 
 demonstrative certainty of possessing it, and nothing can 
 be more hazardous than an appeal to one's private 
 conviction of his own virtue. 
 
 8. Theorem 4. 
 
 Autonomy of will is the sole principle of all moral laws, 
 and of the duties which are in conformity with them. 
 Heteronomy of will, on the other hand, not only supplies 
 no basis for obligation, but it is contradictory of the 
 principle of obligation and of the morality of the will. 
 The single principle of morality thus consists in independ- 
 ence of all matter of the law, that is, of every object 
 of desire, and in the determination of the will through 
 the mere universal form of law, of which a maxim must 
 be capable. This independence of all matter is freedoin 
 in the negative sense, just as the self-legislation of pure 
 practical reason is freedom in the positive sense. Hence 
 
THE CRITIQUE OF PRACTICAL REASON. 27 1 
 
 the moral law simply expresses the autonomy of pure 
 practical reason, that is, of freedom. Autonomy is 
 therefore the formal condition of all maxims, and apart 
 from this condition there can be no harmony of the will 
 with the supreme practical law. If the matter of volition, 
 which is just the object of desire as connected with the 
 law, should enter into the practical law as the condition 
 of its possibility, there will be a heteronomy of the will; 
 for the will must then follow some natural impulse or 
 desire, and must therefore be dependent upon the law of 
 nature. Plainly the will in that case does not give law to 
 itself, but merely prescribes the rational course to be taken 
 in following certain pathological laws. Our maxims cannot 
 contain in themselves the form of universal law, and 
 therefore they not only cannot be the basis of obligation, 
 but they contradict the principle of a pure practical 
 reason. Even, therefore, if the action which proceeds 
 36 from them should be in harmony with moral law, they 
 are opposed to a truly moral disposition. 
 
 Remark. 
 
 Suppose that the matter of my maxim is my own happi- 
 ness. This is an object which every finite being seeks, 
 37 but the rule which prescribes it can be an objective practical 
 law, only if in one's own happiness is included the happi- 
 ness of others. Hence the law, to further the happiness 
 of others, does not originate from the presupposition, that 
 everyone makes his own happiness the object of his choice, 
 but only from this, that the form of universality, which 
 reason demands as the condition under which a maxim 
 of self-love obtains the objective validity of a law, should 
 
272 THE PHILOSOPHY OF KANT. 
 
 determine the will. It is not the securing of the happi- 
 ness of others that really determines the pure will, but the 
 pure form of law, which limits the maxim as based upon 
 mere desire, by imparting to it the universality of a law 
 and bringing it into conformity with pure practical reason. 
 Only by this limitation, and not by the addition of an 
 external impulse, can there arise the conception of the 
 obligation to extend the maxim of self-love so as to include 
 the happiness of others. 
 
 45 / — Deduction of the Principles of Pure Practical 
 
 Reason. 
 
 As the result of this Analytic, we learn that pure reason 
 can be practical, or, in other words, is capable of deter- 
 mining the will independently of all that is empirical. 
 This, indeed, is established not by an inference, but by a 
 fact. For reason actually proves itself to be practical 
 by the fact of autonomy in the fundamental principle of 
 morality, by which it determines the will to activity. 
 Another thing that we have learned is, that this fact is 
 inseparably bound up with the consciousness of freedom of 
 will, and, indeed, is identical with it. For a rational 
 being is conscious that in his will, or as he is in himself, 
 he belongs in the sphere of action to an intelligible order 
 of things, although he is also aware that, in so far as he 
 belongs to the world of sense, his will, like other efficient 
 causes, is necessarily subject to the laws of causality. 
 
 There is, therefore, a remarkable contrast between the 
 analytical part of the Critique of pure practical reason and 
 the analytical part of the Critique of pure speculative 
 reason. In the latter, not fundamental principles, but 
 
THE CRITIQUE OF PRACTICAL REASON. 273 
 
 the pure perceptions of space and time, constitute the 
 primary data by reference to which a priori knowledge 
 was shown to be possible, and possible only for objects 
 46 of sense. Speculative reason with perfect right denied 
 that there could be any positive knowledge of objects 
 which lie beyond the sphere of experience, and, therefore, 
 it denied all knowledge of things as noumena. At the 
 same time it at least showed, that the conception of nou- 
 mena, as a conception, is not only possible but necessary. 
 It proved, for instance, that there is nothing inconsistent 
 with the principles and limitations of pure theoretical 
 reason in the idea of freedom, taken in its negative sense. 
 Speculative reason, however, did not extend our know- 
 ledge by presenting noumena to us as definite objects, 
 but on the contrary showed that we are shut out from all 
 knowledge of them. 
 
 Nor does the moral law present things to our conscious- 
 ness as noumena, but it puts us in possession of a fact 
 which nothing in the whole sensible world, nothing that 
 comes within the range of theoretical reason in its widest 
 use, can possibly explain. This fact points to a purely 
 intelligible world, and even so far determines its character 
 positively, that we know something of it, namely, a 
 law. 
 
 This law gives to the world of sense, or rather to the 
 sensuous nature oixdXxondX beings, the form of an intelligible 
 world or supersensible nature^ without in any way interfer- 
 ing with the mechanism of the world of sense. Now, 
 nature, in the most general sense of the word, means the 
 existence of tbings under laws. The sensuous nature of 
 rational beings, viewed generally, is the existence of such 
 beings under empirically conditioned laws. Relatively 
 
2 74 THE PHILOSOPHY OF KANT. 
 
 to reason^ this is heterono7?iy. But the supersensible 
 nature of rational beings is their existence under laws 
 which are independent of all empirical conditions. These 
 laws, therefore, belong to the autonomy of pure reason. 
 Now, laws which are of such a character that a knowledge 
 of them is presupposed as the condition of the existence of 
 things, are practical laws. The supersensible nature of 
 rational beings is, therefore, just their nature as under 
 the autonomy of pure practical reason. But the law of 
 this autonomy is the moral law, which is tlierefore the 
 fundamental law of a supersensible nature and of a purely 
 intelligible world. The counterpart of this intelligible 
 world ought to exist in the world of sense, but without 
 interfering with its laws. The intelligible world is known 
 only by reason, and might be called the archetypal world 
 {iiatura archetypd); the world of sense, in so far as the idea 
 47 of the inteUigible world is capable of determining the will 
 and producing an effect upon it, we might call the ectypal 
 world {natitra ectypa). For, in point of fact the moral 
 law transfers us in idea into a realm of nature in which 
 pure reason, if it were accompanied by adequate physical 
 power^ would produce the highest good, and determines 
 our will to give to the world of sense the form of a system 
 of rational beings. 
 
 Now, in nature as it actually presents itself to our ex- 
 perience, the will, free as it is in itself, is not determined 
 to maxims which by themselves could be the foundation 
 of a system of universal laws, or which are even in har- 
 mony with such a system. On the contrary, the maxims 
 of the will rest upon private inclinations, which no doubt 
 constitute a system of pathological or physical laws, but 
 not such a system as would be possible were our will 
 
THE CRITIQUE OF PRACTICAL REASON. 275 
 
 determined by pure practical laws. Yet our reason makes 
 us conscious of a law to which all our maxims are subject, 
 just as if an ordered system of nature must be produced 
 by our will. This law must therefore be the idea of a 
 system of nature, which is not presented in experience, 
 but which yet is possible through freedom; a supersensible 
 system of nature, to which we ascribe objective reality, at 
 least in relation to action, because we regard it as the 
 object which as pure rational beings we ought to will. 
 
 There is therefore a distinction between the laws of a 
 nature to which the will is subject^ and the laws of a nature 
 
 48 which is subject to the will. In the one case, the object 
 must be the cause of the idea which determines the will ; 
 in the other, the will must be regarded as the cause of the 
 object; or, in other words, the causality of the will must 
 be determined entirely by pure reason. In this latter 
 connection, therefore, reason must be called pure prac- 
 tical reason. 
 
 49 So much by way of exposition of the supreme principle 
 of practical reason. The dedziction of that principle, that 
 is, the justification of its objective and universal validity, 
 and the proof that such an a priori synthetic proposition 
 is possible, we cannot expect to find so easy as the deduc- 
 tion of the principles of pure theoretical understanding. 
 
 50 The objective reality of the moral law cannot be 
 established by any appeal of theoretical reason either to 
 speculation or to experience, and even if its claim to 
 demonstrative certainty were renounced, it could not be 
 proved a posterioi'i by means of experience. Yet it rests 
 upon a solid foundation of its own. 
 
 No deduction of the principle of morality is possible in 
 either of those ways, but it turns out that the true method 
 
276 THE PHILOSOPHY OF KANT. 
 
 of deduction is just the reverse of what we might have 
 expected. It is the moral law which serves as the 
 principle of deduction of a faculty which cannot be known 
 theoretically or proved by experience, but which specula- 
 tive reason is forced to admit as at least possible. For 
 the moral law does not itself stand in need of any deduc- 
 tion, yet it proves not simply the possibility but the 
 51 actuality of freedom in beings who recognize it to be 
 binding upon themselves. In fact the moral law is a law 
 of a free cause, and therefore a law which makes a super- 
 sensible system of nature possible; just as the meta- 
 physical law of events in the world of sense was a law of 
 the causality of a sensible system of nature. The moral 
 law therefore does what speculative philosophy fails to do: 
 it determines the law for a causality of which the lalter 
 could give only a negative conception, and this for the 
 first time gives objective reality to the conception of a 
 free cause. 
 
 The moral law proves its own reality even to the 
 satisfaction of the Critique of Speculative Reason, by 
 adding to the merely negative conception of a free cause, 
 the possibility of which had to be assumed without being 
 understood, the positive conception of a reason which 
 directly determines the will. Thus the moral law is able 
 to give objective, though only practical, reality to the 
 ideas of reason ; and therefore practical reason makes 
 wt7fianent the use of ideas, which for speculative reason 
 were transcendent 
 
 The determination of the causality of beings in the 
 world of sense, from the very nature of the case, can 
 never be unconditioned. Yet, for every series of condi- 
 tions there must necessarily be something that is uncon- 
 
THE CRITIQUE OF PRACTICAL REASON. 277 
 
 ditioned, and therefore there must be a causality which is 
 completely self-determined. The possibility of freedom, 
 as a faculty of absolute spontaneity, was not a postulate, 
 but an analytic proposition of pure speculative reason. 
 But it is utterly impossible to find in experience any 
 particular instance of an action that conforms to the idea 
 of freedom. Hence speculative reason could only defend 
 the thought of a free cause from attack, by showing that a 
 being who belongs on the one side to the world of sense, 
 may yet on the other side be considered as a noumenon. 
 C2 It therefore maintained that there is nothing contradictory 
 in the supposition, that all the actions of a free being may 
 be physically conditioned in so far as they are regarded 
 as phenomena, while yet in so far as in acting it belongs to 
 the intelligible world, its causality is physically uncondi- 
 tioned. The conception of freedom thus turned out to 
 be a regulative principle of reason. True, no knowledge 
 of the nature of the object, of which free causality was 
 affirmed, was thus obtained, but an obstacle which 
 hindered us from admitting its existence was removed. 
 For, on the one hand, it was possible, in the explanation 
 of events in the world, and therefore in the explanation 
 of the actions of rational beings, to allow that the 
 mechanism of natural necessity might be followed back 
 ad infinitum from conditioned to condition ; while, on the 
 other hand, the place which speculative reason leaves 
 empty was kept open, namely, the realm of the intelligible, 
 and to this realm the unconditioned was transferred. 
 But this thought could not be realized ; in other words, it 
 could not be converted into the knowledge of a being 
 acting freely, nor indeed could it be shown that the know- 
 ledge of such a being was even possible. Pure practical 
 
278 THE PHILOSOPHY OF KANT. 
 
 reason, however, fills up the place left empty by specula- 
 tive reason with a determinate law of causality in an 
 intelligible world, namely, the moral law. Speculative 
 reason in this way gains nothing in the way of insight, but 
 it acquires certavity in regard to the problematic concep- 
 tion of freedom, inasmuch as this conception obtains 
 undoubted objective reality, though no doubt only prac- 
 tical reality. We cannot even say, that the conception 
 of causality in this way obtains any extension beyond the 
 limits of the world of sense, for that conception has 
 meaning and application only in relation to phenomena, 
 and serves simply to connect them with one another. To 
 justify the application of the conception of causality beyond 
 phenomena, it would be necessary to show, how the 
 logical relation of reason and consequent may be em- 
 ployed synthetically in a mode of perception that is not 
 sensuous ; in other words, it would have to be explained 
 how a 7ioiwienal cause is possible. But this cannot be 
 done, nor has practical reason any motive for trying to do 
 it. It is sufficient for practical purposes to show that the 
 causality of man as a sensuous being can be determined 
 by pure reason^ and that pure reason is therefore practical. 
 
 53 // Extension of Practical as cotnpared with Speculative 
 
 Reason. 
 
 58 Besides the theoretical relation in which it stands to 
 objects, understanding has also a relation to the faculty 
 of desire, and is therefore called will. And as pure under- 
 standing, which in this connection is called reason, is 
 practical through the mere idea of law, in its relation to 
 desire it is rightly called pure will. The objective reality 
 
THE CRITIQUE OF PRACTICAL REASON. 279 
 
 of a pure will, or, what is the same thing, of a pure 
 practical reason, may be said to be presented a prio?'i in 
 the moral law as a fact; for we may call a determination 
 of the will which is bound up with its very nature a fact, 
 without meaning to imply that it rests upon empirical 
 principles. Now, the conception of a will carries with it 
 the conception of causality ; and therefore the conception 
 of a pure will implies the conception of a free causality. 
 By a free causality is meant, a causality which cannot be 
 determined by laws of nature, and which therefore cannot 
 59 be proved to be real by empirical perception. Its objec- 
 tive reality can however be justified a priori through the 
 pure practical law. Now, the conception of a being who 
 has free will is that of a noumenal cause. That this con- 
 ception is not self-contradictory is plain, if we consider, 
 that the conception of cause has its source entirely in 
 pure understanding, and that it has been proved to have 
 objective reality in the Deduction of the Categories. 
 Being in its origin independent of all sensuous conditions, 
 the conception of cause is not in itself limited to pheno- 
 mena, nor is there anything to hinder it from being 
 applied to things which are objects of pure understanding. 
 We have, however, only a sensuous perception to which 
 we can apply it ; and hence a noumenal cause, though it 
 can be thought^ remains for theoretical reason an empty 
 conception. But it is not necessary to seek for a theoreti- 
 cal knowledge of the nature of a being that has a pure will ; 
 it is enough to show that there is such a being, and that 
 I may therefore combine the conception of causality with 
 the conception of freedom. This combination I am 
 certainly entitled to make, for the conception of causality 
 is not of empirical origin, and here I do not claim the 
 
28o THE PHILOSOPHY OF KANT. 
 
 light to make any other than a practical use of it ; in 
 Other words, to employ it in relation to the moral law 
 by which its reality is determined. 
 60 Moreover, the objective reahty of a pure conception of 
 understanding in the sphere of the supersensible, when it 
 has once been introduced, imparts objective validity to 
 all the other categories, although only in so far as these 
 stand in necessary connection with the moral law, through 
 which the pure will is determined. 
 
 61 Chapter II. — The Object of Pure Practical 
 
 Reason. 
 
 To determine whether a thing is an object of pure practical 
 reason or not, it is by no means necessary to ask whether 
 we are physically able to produce it. The only question is, 
 whether we ought to will an act, if we had the power to 
 bring the object into existence to which the act is directed. 
 The moral possibility must therefore precede the act, for 
 it is the law of the will, and not the object, which is to 
 determine the act. 
 52 The only objects of a practical reason slyc good Sind evil. 
 The one is a necessary object of desire, the other ot 
 aversion, and both rest upon a principle of reason. 
 
 Now, as pleasure and pain cannot be connected a 
 priori with the idea of an object, those who make a 
 feeling of pleasure the basis of their moral judgments, 
 must call that good which is the means to what is 
 agreeable, and that einl which is the cause of what 
 is disagreeable and painful. The practical maxims 
 which follow from this conception of the good, can- 
 not contain as the object of the will anything that is 
 
THE CRITIQUE OF PRACTICAL REASON. 28 1 
 
 good in itself, but only that which is good relatively to 
 something else. 
 64 Weal and woe are terms which always designate merely 
 a relation to our own state of feeling, as agreeable or dis- 
 agreeable, pleasant or painful. We desire an object as 
 weal or avoid it as woe only in so far as it is related to our 
 sensibility, and to the feeling of pleasure or pain which 
 the object produces in us. Good and evil, on the other 
 hand, always imply a relation to the will, in so far as it is 
 determined by a law of reason to make something an 
 object for itself. In this connection the will is never 
 determined directly by the idea of the object, but is a 
 faculty of making a rule the motive of its action. In the 
 proper sense of the word, therefore, good and evil are not 
 related to the state of sensation of the person, but to his 
 action. If there is anything absolutely good or evil, or 
 anything that is regarded as such, it cannot be the object 
 of the action, but only the mode of action, the maxim 
 of the will, and therefore the agent himself. 
 66 Now, if there be a principle which is thought as in itself 
 capable of determining the will, independently of all rela- 
 tion to possible objects of desire, it is an « /r/<?n practical 
 law, and pure reason must then be regarded as of itself 
 practical. In that case the law directly determines the 
 will, and the act conforming to it is in itself good. Hence 
 a will, the maxim of which is always in harmony with law, 
 is absolutely or in every respect good, and the supreme 
 condition of all good. But if, on the contrary, there is 
 something which precedes the maxim of the will and 
 determines desire, something which presupposes an object 
 fitted to produce pleasure or pain; and if therefore the 
 maxim, to seek the pleasant and avoid the painful, deter- 
 
282 THE PHILOSOPHY OF KANT. 
 
 mines our actions; then our actions are good only 
 relatively, or as means to another end, and our maxims 
 can never be laws, although no doubt they are practical 
 precepts of reason. 
 
 68 Now, the conceptions of good and evil, as consequences 
 of the a priori determination of the will, presuppose a 
 pure practical principle, and therefore a causality of pure 
 
 69 reason. Hence they are in all cases modi of the single 
 category of causality, in so far as that category is deter- 
 mined through the conception of a law of freedom which 
 reason gives to itself Thus reason proves itself to be 
 practical. But, although actions are, on the one hand, 
 under the law of freedom, and therefore belong only to 
 intelligence ; they are, on the other hand, as events in the 
 world of sense, also under the law of phenomena. Prac- 
 tical reason can therefore determine itself only in relation 
 to phenomena. And as its determinations must conform 
 to the categories of understanding, they cannot be em- 
 ployed theoretically, with the object of bringing the 
 various elements of sensuous perceptio7i a priori under 
 one consciousness, but only for the purpose of subjecting 
 a priori the various desires to the unity of consciousness, 
 as implied in a practical reason, or pure will, which issues 
 its commands through the moral law. 
 
 7 1 The Type of Pure Practical Judgvient 
 
 Prior to the conceptions of good and evil the will has 
 no object. But these conceptions themselves stand under 
 a practical rule of reason, which, in the case of pure 
 reason, determines the will a piiorim. respect of its object. 
 Now, to decide whether an action, that stands under 
 
THE CRITIQUE OF PRACTICAL REASON. 283 
 
 a rule is one that is possible for us in the world of 
 sense or not, is the business of practical judgment, the 
 function of which is to apply to an action in co7icreto that 
 which in the rule is stated universally or in abstrado. 
 
 'J2i But there is no perception, and therefore no schema, 
 which can serve as the middle term by means of which 
 the law of freedom, and therefore the conception of the 
 unconditionally good, can be appUed in concrete. Hence 
 the moral law can be applied to objects of nature only 
 through understanding and not through imagination. 
 Understanding, however, can supply no schema of sensi- 
 bility for an idea of reason, but only a law. Yet this law 
 can be exhibited in concreto in objects of sense, and 
 may therefore take the form of a law of nature. It thus 
 serves as the instrument of practical judgment, and may 
 therefore be called the type of the moral law. 
 
 The rule which judgment appHes, in subordination to 
 the laws of pure practical reason, is this : Ask yourself 
 whether you could regard the act which you have in view 
 as possible by your own will, if it were to occur in con- 
 formity with a law of nature. As a matter of fact this is 
 the rule by which everyone decides whether an action is 
 morally good or bad. 
 
 74 It is therefore permissible to take the nature of the 
 world of sense as the type of an i?itelligible nature, so long 
 as I merely conceive of the latter as under the form of 
 law, and do not transfer to it my perceptions and what is 
 dependent upon them. For all laws are as laws the same 
 in kind, no matter what may be the source from which 
 they spring. 
 
284 THE PHILOSOPHY OF KANT. 
 
 76 Chapter III. — The Motives of Pure Practical 
 
 Reason. 
 
 What is essential to the moral value of every action is, 
 that the moral law should directly determine the will. It 
 is not enough that the will should be determined in 
 harmony with the moral law. If feeling of any sort has 
 to be presupposed before the will can be determined, the 
 will is not determined because of the law, and therefore 
 the action is not ??wral but simply legal. By the word 
 " motive " we must understand that which determines the 
 will of a being, whose reason is not by its very nature 
 necessarily in harmony with the objective law. Hence, 
 firstly, we cannot speak of a motive in connection with the 
 divine will ; and, secondly, the only motive of the human 
 will, and indeed of every finite rational being, must be the 
 moral law. The objective ground must therefore always 
 be at the same time the only and the sufficient subjective 
 ground of determination of an action. On any other 
 supposition, only the letter of the law, and not its spirit, 
 would be fulfilled. 
 
 77 How a law can of itself directly determine the will, is 
 for human reason an insoluble problem, for it is identical 
 with the problem, how a free will is possible. What we 
 are called upon to show a priori^ is, not how the moral 
 law of itself can supply a motive, but what influence it 
 has, or rather must have, upon the mind in so far as it 
 does supply a motive. 
 
 The essential thing in all determination of the will by 
 the moral law, is, that the will as free should not only be 
 determined without the co-operation of sensuous desires, 
 but that it should even oppose such desires, and restrain 
 
THE CRITIQUE OF PRACTICAL REASON. 285 
 
 all natural inclinations that might prevent the realization 
 of the law. So far the influence of the moral law is merely 
 negative, and its character as a motive can be known only 
 a priori. For every natural inclination and sensuous 
 desire is based upon feeling, and the negative influence 
 of the law in opposing the natural inclinations itself takes 
 the form of feeling. Hence we know a priori^ that the 
 moral law in determining the will by thwarting all our 
 inclinations, must produce in us a feehng that may be 
 called pain. This is the first instance we have found, and 
 perhaps it is the only instance, in which we can tell from 
 a priori conceptions, what is the relation of knowledge 
 to the feeling of pleasure or pain. All natural inclinations 
 without exception arise from self-regard, the two forms of 
 which are self-love and self-esteem. Self-love, which is 
 natural and belongs to us prior to the moral law, pure 
 practical reason simply restrains, by bringing it into con- 
 78 formity with the law. It is then called rational self-love. 
 But self-esteem it completely destroys, for no man can 
 show the least title to respect, except in so far as his acts 
 conform to the moral law. The moral law, however, is 
 in itself positive, or, in other words, it is the form of an 
 intellectual causality, that is, of a free causality. In so far 
 as it counteracts the subjective antagonism of the natural 
 inclinations, and weakens self-esteem, the moral law is an 
 object of reverence; and in so far as it completely destroys 
 self-esteem, it is an object of the highest reverence. Thus 
 it gives rise to a positive feeling, which is not of empirical 
 origin, but is known a priori. Reverence for the moral 
 law is therefore a feeling which has an intellectual source, 
 and it is the only feeHng which can be known completely 
 a priori^ and which can be perceived to be necessary. 
 
286 THE PHILOSOPHY OF KANT. 
 
 80 Not only, therefore, is the moral law the formal con- 
 dition of action through pure practical reason ; not only 
 is it a material, though merely objective, condition of the 
 determination of those objects of action which we call 
 good and evil ; but it is also the subjective condition or 
 motive of action, inasmuch as it has an influence upon 
 the morality of the subject, and produces in him a feeling 
 that conduces to the influence of the law upon his will. 
 It is true that the sensuous feeling which is implied in all 
 our inclinations is the condition of the feeling of reverence, 
 but the cause which determines it lies in pure practical 
 reason. The feeling of reverence, therefore, is in its 
 origin noX pathological hut p7'acfical. Nor is reverence for 
 law an external motive to morality, but it is morality 
 itself, regarded subjectively as a motive. As an effect 
 upon feeling, it presupposes that the subject of it has a 
 sensuous nature, and is therefore finite. Hence we cannot 
 say that a Supreme Being feels reverence for law, nor can 
 we say that even a free finite being, who was devoid of 
 sensibility, would have such a feeling; for in neither case 
 is there any natural impulse which stands in opposition 
 to practical reason. 
 
 93 Critical Exafuination of the Analytic of Pure 
 
 Practical Reason. 
 
 102 To get rid of the apparent contradiction between the 
 mechanism of nature and freedom in the same act, we 
 must bear in mind what has been said, or what is implied, 
 in the Critique of Pure Reason. The natural necessity 
 which is contradictory of freedom attaches only to the 
 determinations of a thing that stands under conditions of 
 time. Hence it applies only to the agent in his 
 
THE CRITIQUE OF PRACTICAL REASON. 287 
 
 phenomenal aspect. In regard to actions the ground of 
 which hes in what belongs to the past and is no longer 
 ivithin his power ^ the agent is certainly not free. But he 
 is conscious of himself as also a thing in himself, and 
 from this point of view he looks upon his own existence 
 as not standing under conditions of tiine^ but as capable of 
 being determined only by laws which have their source 
 in reason. Nothing in his existence can in this connec- 
 tion be said to be antecedent to the determination of his 
 will. Every action that he does, or, more generally, 
 every change in the determinations of inner sense, and 
 indeed the whole series of such changes, he regards, in 
 so far as he is conscious of his existence as an intelligence, 
 as the result of his noumenal causality, never as determin- 
 ing that causality. From this point of view a rational being 
 can say with truth, that every wrong act done by him he 
 could have left undone, although as a phenomenon the 
 act is sufficiently determined and must inevitably take 
 place. For the act, together with all in the past that 
 determines it, belongs to the one continuous phenomenal 
 manifestation of the character which he has made for him- 
 self Looking upon himself in this way as a cause that 
 is independent of sensibility, he ascribes to himself the 
 causality of the whole phenomenal manifestation of his 
 being. 
 
 104 There is another difficulty in regard to the combination 
 
 105 of freedom and the mechanism of nature in a being that 
 belongs to the world of sense. Let it be granted, it may 
 be said, that the subject as an intelligence is free in the 
 performance of a certain act, although as a subject 
 belonging to the world of sense, he is conditioned by the 
 mechanism of nature. Yet if God — the universal Original 
 
288 THE PHILOSOPHY OF KANT. 
 
 Being— is admitted to exist, He must be regarded as the 
 cause of the existence of every substance. The actions of 
 man must therefore be due to a power which Hes entirely 
 outside of himself. For his actions must be referred to 
 a Supreme Being, who is distinct from himself, and upon 
 this Being, his existence, as well as all that is referred to 
 his causality, must be absolutely dependent. 
 
 1 06 A short solution of this difficulty is not far to seek. 
 Existence in time is a purely sensuous mode of conscious- 
 ness, which belongs only to thinking beings as they are 
 in the world, but does not hold of them as they are 
 in themselves. By the creation of thinking beings 
 
 107 we must therefore mean the creation of things in 
 themselves. The idea of creation has no meaning, in 
 so far as we are speaking of the sensuous mode of 
 apprehending existence or causality, but can refer only to 
 noumena. To speak of beings in the world of sense as 
 being created, is to speak of them as if they were 
 noumena. Now, it would be a contradiction to say that 
 God is the creator of phenomenal objects. It is equally 
 a contradiction to say that, as creator, He is the cause of 
 actions which take place in the world of sense, and are 
 therefore phenomenal objects, though it be admitted that 
 He is the cause of the existence of acting beings in their 
 character as noumena. Let us assume, then, that existence 
 in time holds only of phenomena, not of things in them- 
 selves. Now, if freedom is not incompatible with the 
 natural mechanism of actions regarded as phenomena, it 
 cannot be incompatible with the fact that the beings who 
 perform the actions are creatures. For creation has to do 
 only with their existence as intelligences, not with their 
 sensible existence, and therefore it cannot be regarded as 
 
THE CRITIQUE OF PRACTICAL REASON. 289 
 
 the ground of phenomena. It would be altogether 
 different if beings in the world existed as things in 
 themselves in time, for then the creator of a substance 
 would at the same time be the author of the whole 
 mechanism of this substance. 
 
 112 BOOK II.— DIALECTIC OF PURE PRACTICAL 
 
 REASON. 
 
 Chapter I. — General Consideration. 
 
 Pure reason is always dialectical, whether it is viewed in 
 its speculative or in its practical use. In both cases it 
 seeks to comprehend the absolute totality of conditions 
 for that which is presented as conditioned, and such a 
 totality cannot possibly be found anywhere but in things 
 in themselves. But all our conceptions of things have to 
 be brought into relation with perceptions, which in man 
 are always sensuous, and hence objects cannot be known 
 as things in themselves, but only as phenomena. It is 
 impossible to find the unconditioned in the series of the 
 conditioning and the conditioned, and an unavoidable 
 illusion arises from the application to phenomena of 
 the rational idea of the totality of conditions. The 
 deceptive character of this illusion would not indeed 
 be observed, if it did not betray itself by the self- 
 contradiction into which reason falls, when it seeks to 
 apply the principle in question, namely, that the con- 
 ditioned presupposes the unconditioned. Thus necessity 
 is laid upon reason, to trace back this illusion to its 
 source, and this is a task which can be accomplished only 
 by a thorough criticism of the whole faculty of pure 
 
290 THE PHILOSOPHY OF KANT. 
 
 [ 13 reason. The antinomy of pure reason, which makes 
 itself apparent in its dialectic, thus turns out to be the 
 most beneficial error into which human reason could 
 possibly have fallen. For it forces us to look about for 
 the clue by which we may escape from the labyrinth into 
 which we have wandered, and this clue, when it has been 
 found, unexpectedly leads us to a point where a higher 
 and an unchangeable order of things lies before us. In 
 this higher realm we find that we already exist, and in it 
 we are called upon to continue our existence, guiding 
 ourselves by certain definite precepts which the highest 
 reason lays upon us. 
 
 How the natural dialectic of pure speculative reason can 
 be explained, and how the error arising from a perfectly 
 natural illusion may be guarded against, has been fully 
 shown in the Critique of Pure Reason. But reason in its 
 practical use tails into as great a difficulty. It seeks to 
 find the unconditioned for the practically conditioned, 
 which depends upon the natural wants and inclinations, 
 although the unconditioned is not to be conceived as 
 determining the will, but simply as the unconditioned 
 totality of the object of pure practical reason. This object 
 is the higJiest good. 
 
 r 14 In regard to the dialectic of pure practical reason, which 
 is connected with its determination of the idea of the 
 highest good, a preliminary remark has to be made. The 
 moral law must of itself be capable of determining a pure 
 will. But this law is merely formal, or prescribes only the 
 form of that maxim which can be a universal law, and 
 hence it abstracts from all matter, that is, from every object 
 of volition. Accordingly, while it is true that the highest 
 good is always the whole object of a pure practical reason, 
 
THE CRITIQUE OF PRACTICAL REASON. 29 1 
 
 or a pure will, it is not to be regarded as determining 
 the will. The moral law must alone determine the 
 pure will, and its sole object is to produce, or help to 
 produce, such a will. For, as we have seen in the 
 Analytic, the supposition that the will is determined, 
 prior to the moral law, by some object called a good, 
 and that from it the supreme principle of action may 
 115 be derived, invariably gives rise to heteronomy and 
 destroys the moral principle. 
 
 116 Chapter II. — The Summum Bonum. 
 
 The conception highest contains two distinct ideas, 
 which must be carefully distinguished, if we are to 
 avoid needless perplexities. The highest may mean 
 either the sicpreme {supremum), or the cotnplete 
 {consiimmatum). The supreme is a condition which is 
 itself unconditioned or is not subordinate to anything 
 else {originariwn) . The complete, again, is a whole which 
 is not part of a larger whole of the same kind {perfectissi- 
 mum). Now virtue, or the worthiness to be happy, as 
 we have seen in the Analytic, is the supreme condition of 
 all that we can regard as desirable, and therefore the 
 supreme condition of all our search for happiness. 
 Virtue is therefore the supreme good. But it is not the 
 whole or complete good which finite rational beings 
 desire to obtain. The complete good includes happiness, 
 and that not merely in the partial eyes of the person who 
 makes it his end, but even in the judgment of unbiassed 
 reason, which regards the production of happiness in the 
 world as an end in itself If we suppose, for the sake of 
 illustration, that there exists a rational Being who has all 
 
292 THE PHILOSOPHY OF KANT. 
 
 power, it cannot be in accordance with the whole will of 
 such a being, that his creatures should be unable to secure 
 the happiness which their nature demands and of which 
 117 their obedience to the moral law makes them worthy. 
 The highest good of a possible world must therefore 
 consist in the union of virtue and happiness in the 
 same person, that is, in happiness exactly proportioned 
 to morality. By the highest good is here meant, 
 therefore, the whole or complete good. In this 
 complete good virtue is always, as a condition, 
 the supreme good, having no condition higher than 
 itself; while happiness is no doubt always agreeable 
 to the person who possesses it, but it is not good simply 
 in itself, and in all respects : it is good only under the 
 condition that a man's conduct is in conformity with 
 the moral law. 
 
 119 /. The A7itinomy of Practical Reaso7i. 
 
 In the highest good which is practical for us, that is, 
 which is to be realized through our will, virtue and 
 happiness are conceived as necessarily united, so that the 
 one cannot be apprehended by practical reason as 
 separated from the other. Now the connection of virtue 
 and happiness must be known either by analysis or by 
 synthesis. But it has been shown not to be known 
 analytically, and hence it must be synthetic, and synthetic 
 in the way of cause and effect. For we have here to do 
 with a practical good, that is, with a good which is 
 possible only by means of action. Either, there- 
 fore, the desire for happiness must be the motive 
 to maxims of virtue, or the maxims of virtue must 
 be the efficient cause of happiness. The former 
 
THE CRITIQUE OF PRACTICAL REASON. 293 
 
 is absolutely impossible ; for, as the Analytic has shown, 
 maxims which make the desire for happiness the 
 motive of the will, are not moral at all, and cannot 
 possibly be the foundation of virtue. The latter is also 
 impossible; for the causal connection in the world of 
 events which follow from the determination of the 
 will, does not conform to the moral disposition of the 
 
 120 will, but depends upon a knowledge of the laws of nature 
 and upon the physical power to make use of them for 
 certain ends. Hence the most scrupulous adherence to 
 the laws of morality cannot be expected to bring happi- 
 ness into connection with virtue, and to lead to the 
 attainment of the highest good. 
 
 //. Critical Solution of the Antinomy. 
 
 The solution of this antinomy is of the same nature as 
 the solution of the antinomy of pure speculative reason. 
 The first of the two propositions, namely, that virtue is 
 the result of the search for happiness, is absolutely false. 
 The second proposition, however, is not absolutely false, 
 but is untrue only if virtue is regarded as a form of 
 causality in the world of sense. In that case it is assumed 
 that a rational being can exist only as a sensuous being, 
 
 121 and the proposition is therefore conditionally false. Not 
 only can I think of my existence as a noumenon in the 
 world of intelligence, but in the moral law I have a purely 
 
 ntellectual principle which is capable of determining my 
 causality as manifested in the v\-orld of sense. There is, 
 therefore, nothing impossible in the idea that a moral 
 disposition should necessarily be the cause of happiness, 
 not indeed directly, but indirectly, or through the medium 
 of an intelligent Author of nature. Yet, though happiness 
 
2 94 THE PHILOSOPHY OF KANT. 
 
 might thus be an effect of virtue in the world of sense, the 
 connection of virtue and happiness in a system of nature, 
 which is merely an object of the senses, cannot be other 
 than contingent, and therefore it cannot be established in 
 the way required in the conception of the highest good. 
 
 In spite of the apparent self-contradiction of practical 
 reason, the highest good is necessarily the ultimate end 
 and the true object of a moral will. For the highest 
 good is practically possible, and the maxims of the will, 
 which are related to it on the side of their matter, have 
 objective reality. This reality was at first brought into 
 doubt by an antinomy in regard to the connection of 
 morality and happiness in accordance with a universal 
 law; but the antinomy arose simply from the false 
 assumption that things in themselves are related to 
 phenomena in the same way as phenomena are related 
 to one another. 
 
 128 IV. The Iminortaliiy of the Soul. 
 
 The object of a will that is capable of being determined 
 by the moral law, is the production in the world of the 
 highest good. Now, the supreme condition of the highest 
 good is the perfect harmony of the disposition with the 
 moral law. Such a harmony must be possible, not less 
 than the object of the will, for it is implied in the 
 command to promote that object. Perfect harmony of 
 the will with the moral law is holiness^ a perfection of 
 which no rational being existing in the world of sense 
 is capable at any moment of his life. Yet holiness is 
 demanded as practically necessary, and it can be found 
 only in an infinite progress towards perfect harmony 
 with the moral law. Pure practical reason therefore 
 
THE CRITIQUE OF PRACTICAL REASON. 295 
 
 forces us to assume such a practical progress towards 
 perfection as the real object of our will. 
 
 Now, this infinite progress is possible only if we pre- 
 suppose that the existence of a rational being is prolonged 
 to infinity, and that he retains his personality for all time. 
 This is what we mean by the immortality of the soul. 
 The highest good is therefore practically possible, only if 
 we presuppose the immortality of the soul. Thus 
 immortality is inseparably bound up with the moral law. 
 It is a postulate of pure practical reason, that is, a pro- 
 position that cannot be proved theoretically^ but depends 
 upon an a priori practical law of unconditioned validity. 
 
 '29 A finite rational being is capable only of an infinite 
 progress from lower to higher stages of moral perfection. 
 The Infinite Being, who is free from the limits of time, 
 sees in this series, which for us has no end, a whole that 
 is in harmony with the moral law. Holiness He demands 
 inexorably as a duty in order to assign to everyone his 
 exact share in the highest good ; and this holiness lies 
 completely before Him in a single intellectual perception 
 of rational beings. Created beings can hope to share in 
 the highest good only in so far as they are conscious of 
 having stood the test of the moral law. If in the past 
 they have advanced from lower to higher degrees of 
 morality, and have thus proved the strength of their 
 resolution, they may hope to make unbroken progress in 
 the future as long as they live here, and even beyond the 
 present life. They can never hope in this life, or, indeed, 
 at any imaginable point of time in the future life, to be in 
 
 130 perfect harmony with the will of God, but they may hope 
 for this harmony in the infinite duration of their existence 
 as it is surveyed by God alone. 
 
296 XHE PHILOSOPHY OF KANT. 
 
 V. The Existence of God. 
 
 The moral law leads us to postulate not only the 
 immortality of the soul, but the existence of God. 
 For it shows us how happiness in proportion to morality, 
 which is the second element of the highest good, is 
 possible, and to postulate it for reasons as perfectly dis- 
 interested as in the former case. This second postulate 
 of the existence of God rests upon the necessity of 
 presupposing the existence of a cause adequate to the 
 effect which has to be explained. 
 
 Happiness is the state of a rational being existing in 
 the world who experiences through the whole of his life 
 whatever he desires and wills. It, therefore, presupposes 
 that nature is in harmony with his whole end, as well as 
 with the essential principles by which his will is 
 determined. Now, the moral law, being a law of free 
 beings, commands us to act from motives that are entirely 
 independent of nature and of the harmony of nature with 
 our desires. But a rational agent in the world is not 
 the cause of the world and of nature itself There is no 
 reason whatever, in the case of a being who is a part of 
 the world and is dependent upon it, why the moral law 
 should imply a necessary connection between happiness 
 and morality proportionate to happiness. For the will of 
 such a being is not the cause of nature, and therefore he 
 131 has no power to bring nature into complete harmony with 
 his principles of action. At the same time, in the practical 
 problem of pure reason, that is, in the necessary pursuit 
 of the highest good, such a connection is postulated as 
 necessary. He ought to seek to promote the highest 
 good, and therefore the highest good must be possible. 
 
THE CRITIQUE OF PRACTICAL REASON. 297 
 
 He must therefore postulate the existence of a cause of 
 nature as a whole, which is distinct from nature, and 
 which is able to connect happiness and morality in exact 
 harmony with each other. Now, this supreme cause must 
 be the ground of the harmony of nature, not simply with 
 a law of the will of a rational being, but also with the 
 consciousness of this law in so far as it is made the 
 supreme principle of the agent's will. That cause must 
 therefore be in harmony not merely with the form of 
 morality, but with morality as willed by a rational being, 
 that is, with his moral character. The highest good is 
 thus capable of being realized in the world, only if there 
 exists a supreme cause of nature whose causality is in 
 harmony with the moral character of the agent. Now, a 
 being that is capable of acting from the consciousness of 
 law is a rational being, an intelligence^ and the causality 
 of that being, proceeding as it does from the consciousness 
 of law, is a will There is therefore implied, in the idea 
 of the highest good, a being who is the supreme cause of 
 nature, and who is the cause or author of nature through 
 his intelligence and will, that is, God. If, therefore, we 
 are entitled to postulate the highest derivative good, or the 
 best world, we must also postulate the actual existence of 
 the highest original good, that is, the existence of God. 
 Now, it is our duty to promote the highest good, and 
 hence it is not only allowable, but it is necessarily bound 
 up with the very idea of duty, that we should presuppose 
 the possibihty of this highest good, And as this possi- 
 bility can be established only under condition that God 
 exists, the presupposition of the highest good is inseparably 
 connected with duty, or, in other words, it is morally 
 necessary to hold the existence of God. 
 
298 THE PHILOSOPHY OF KANT. 
 
 138 VI. The Postulates of Pure Practical Reason. 
 
 The postulates of pure practical reason are not theor- 
 etical dogmas, but presuppositions which are practically 
 necessary. They do not enlarge our speculative know- 
 ledge, but give objective reality to the ideas of speculative 
 reason in general, and justify it in the use of conceptions 
 which it could not otherwise venture to regard as even 
 possible. 
 
 These postulates are i??imortality, freedom (in the 
 positive sense, as the causality of a being Avho belongs 
 to the intelligible world), and the existence of God. The 
 first rests upon the practically necessary cordition, that 
 existence should continue long enough to permit of the 
 complete realization of the moral law. The sccofid arises 
 from the necessary presupposition of man's independence 
 of the world of sense, and his capability of determining 
 his will in conformity with the law of an intelligible 
 world, that is, the law of freedom. The third depends 
 upon the necessity of presupposing a supreme, self- 
 existent good, that is, the existence of God, as the 
 condition under which the highest good may be realized 
 in such an intelligible world. 
 
 Our reverence for the moral law necessarily compels 
 us to seek for the realization of the highest good, and 
 hence the reality of the highest good must be presup- 
 posed. By means of the postulates of practical reason, 
 we are brought to conceptions, which speculative reason 
 
 139 no doubt set up as problems to be solved, but which it 
 was itself unable to solve. The first conception is that 
 of immortality. This conception involved speculative 
 reason m paralogisrns ; for it could find no trace of the 
 
THE CRITIQUE OF PRACTICAL REASON. 299 
 
 permanence required for the conversion of the psycho- 
 logical conception of an ultimate subject into the real 
 consciousness of a substance. Practical reason supplies 
 what is required, by the postulate of a duration adequate 
 to the complete realization of the moral law in the highest 
 good. It leads, secondly, to the cos77iological idea of an 
 intelligible world and the consciousness of our existence 
 in that world. This idea involved speculative reason in 
 an anti?io?ny, for the solution of which it had to fall back 
 upon a problematic conception, the objective reality of 
 which it could not prove. But practical reason, by means 
 of the postulate of freedom, shows that idea to have 
 objective reality. Lastly, practical reason brings us to 
 the conception of a Supreme Being. This conception 
 speculative reason was able to think, but it could not 
 show it to be more than a transcendental ideal. Practi- 
 cal reason, on the other hand, gives meaning to this idea, 
 by showing that a Supreme Being is the supreme principle 
 of the highest good in an intelligible world, and is 
 endowed with the sovereign power of prescribing moral 
 laws in that world. 
 
 Is our knowledge, then, actually enlarged by practical 
 reason? Is that which for speculative reason is ff^a?i- 
 scende?it for practical reason immanent ? Undoubtedly it 
 is, but only in relation to action. Practical reason 
 cannot give us a theoretical kfwwledge of our own soul, of 
 the intelligible world, or of a Supreme Being, as these are 
 in themselves. All that it can do is to unite the concep- 
 tion of them in the practical conception of the highest 
 good, which is the object of our will, and to unite them 
 entirely a priori through pure reason. This union is 
 effected only through the medium of the moral law, and 
 
300 THE PHILOSOPHY OF KANT. 
 
 merely in relation to that which it commands with a view 
 to the highest good. For we cannot understand how 
 freedom is possible, or how a free cause would appear to 
 us if it were theoretically and positively known ; all that 
 we can say is, that a free cause is postulated by the moral 
 140 law and for the sake of the moral law. The same remark 
 applies to the other ideas. No human intelligence can 
 ever understand how immortality and the existence of 
 God are possible ; but, on the other hand, no sophistry 
 will ever destroy the faith of even the most unreflective 
 man in their reality. 
 
 VII. Possibility of an extension of Pure Practical Reason 
 without a corresponding extension of Pure Speculative 
 Reason. 
 
 141 It is true that the three ideas of freedom, immortality, 
 and God, are not knowledge, but at least they are 
 thoughts the objects of which are not impossible. They 
 are necessary conditions of the possibility of that which 
 an apodictic practical law commands us to make our 
 object, and in this sense they have objective reality. 
 They indicate that they have objects, although we cannot 
 learn from them how they are related to these objects. 
 We can make no synthetical judgments in regard to 
 them, nor can v/e determine theoretically how they are 
 to be applied, and hence we cannot be said to have any 
 knowledge of them. Reason cannot make a theoretical 
 use of them, which is the same as saying that they are 
 not known by speculative reason. But, while the ideas 
 of practical reason do not enlarge our theoretical know- 
 ledge, the sphere of reason itself is in this sense enlarged, 
 
THE CRITIQUE OF PRACTICAL REASON. 30I 
 
 that by means of practical postulates we learn that there 
 are objects corresponding to those ideas. Conceptions 
 which before were problematic thus obtain objective 
 reality. No extension of our knowledge of supersensible 
 objects has taken place, but there has been an extension 
 of theoretical reason and of our knowledge of the super- 
 sensible in general, in so far as reason has been forced 
 to admit that there are such objects, though nothing 
 definite is known in regard to them. Even for this 
 relative extension of its sphere reason is indebted entirely 
 to its pure practical faculty. 
 143 If these ideas of God, an intelligible world or kingdom 
 of God, and immortality, are further determined by 
 predicates borrowed from the nature of man, it does not 
 follow that we have fallen into an anthropomorphism, 
 which makes pure ideas of reason se^isuous^ or that in 
 claiming a knowledge of supersensible objects our idea 
 becomes transcendent. For the predicates of which we 
 make use are those of intelligence and will, and these we 
 conceive of as related to each other shnply in the way 
 that the idea of the moral law demands. Hence we 
 make only a pure practical use of them. Abstraction is 
 made from all the predicates which are connected with 
 these conceptions psychologically, and are learned from 
 the observation of our faculties of understanding and will. 
 Of a Supreme Being, for instance, we cannot say, as we 
 can say of man, that His understanding is discursive, 
 and therefore deals dii*ectly only with conceptions, not 
 with perceptions ; that His perceptions follow one 
 another in time ; that His will is always dependent for 
 satisfaction upon the existence of the object to which it 
 is directed, etc. Now, when abstraction has been made 
 
302 THE PHILOSOPHY OF KANT. 
 
 from such predicates as these, the only predicates that 
 are left are those which belong to the idea of a pure 
 intelligence, in other words, those that are implied in the 
 mere thought of a moral law. Thus we have indeed a 
 knowledge of God, but only in a practical relation. If 
 we try to extend our knowledge to a theoretical relation, 
 we get the idea of an intelligence which does not think 
 but perceives, and a will which is directed to objects 
 upon the existence of which its own satisfaction does not 
 in the least depend. But these are all attributes of which 
 we can form no conception that enables us to have a 
 knowledge oi a Supreme Intelligence; and from this we 
 learn, that they can never be made use of in a theory of 
 supersensible beings, but must be limited in their use to 
 the practice of the moral law. 
 
 148 VIII. Faith as a Need of Pure Reason. 
 
 A need of pure practical reason arises from the duty of 
 making the highest good the object of will, and seeking 
 to promote it with all one's power. The possibility of this 
 highest good has therefore to be presupposed, as well as 
 the conditions without which it would not be possible, 
 namely, God, freedom, and immortality. The duty of pro- 
 moting the highest good is in itself apodictically certain, 
 and is entirely independent of the other presuppositions. 
 The idea of duty thus stands in need of no support from 
 any theory of the inner nature of things, the hidden pur- 
 pose of the world's history, or the existence of a Supreme 
 
 149 Ruler, to show that it is binding upon us in the most 
 absolute sense, and that we ought to act in conformity 
 with it. But the influence upon the agent of the moral 
 
THE CRITIQUE OF PRACTICAL REASON. 303 
 
 law, that is, the disposi'-ion which it produces in him 
 to promote the highest good that can be practically 
 realized by us, presupposes at the very least that the 
 highest good is possible. If it were not possible, we 
 should be trying to realize practically what could not 
 be realized, and to give effect to an idea that was empty 
 ici and without any object. Thus the principle which deter- 
 mines a moral judgment is no doubt subjective in relation 
 to us ; but, inasmuch as it is also the means by which 
 an object that is practically necessary may be promoted, 
 it is also the foundation of all beliefs which possess 
 moral certitude. That principle, therefore, takes the form 
 of a faith or conviction of pure practical reason. 
 
THE CRITIQUE OF JUDGMENT. 
 
THE CRITIQUE OF JUDGMENT. 307 
 
 THE CRITIQUE OF JUDGMENT. 
 
 177 INTRODUCTION. 
 
 / Division of Philosophy. 
 
 The object of philosophy is to search for the principles 
 by which reason may obtain a true knowledge of things. 
 Now, we may conceive of objects either from the theor- 
 etical or from the practical point of view, and hence the 
 ordinary division of philosophy into theoretical dJidi practical 
 is perfectly correct. But, in making this division, we 
 must be sure that the conceptions upon which the distinc- 
 tion of principles rests are themselves distinct. 
 
 There are two, and only two, classes of conception by 
 reference to which a distinction may be made in the 
 principles of philosophy. These are conceptions of nature 
 and the conception of freedom. The former are the con- 
 dition of theoretical knowledge in conformity with a priori 
 principles ; the latter in itself supplies merely a negative 
 principle of theoretical knowledge, but it is the source of 
 principles which enlarge the sphere of the will, and which 
 are therefore called practical. Philosophy has thus two 
 main divisions, theoretical philosophy or the philosophy 
 of nature^ and practical or moral philosophy. But these 
 terms have hitherto been grossly misapplied, both in the 
 division of the principles of philosophy and in the 
 178 division of philosophy itself. For it has been as- 
 
3o8 Till'; PHILOSOPHY OK KANT. 
 
 sninrd lliat llicrc is no distinction l)clwccn wlmt is 
 rnllcd ** i)r;irtical " in the si)1umi' oI n.Unrc, and what 
 is *' j)racli('al " rclalivciy to the idea, ol" liccdoni. Now, 
 tin's ronfnsidn bclwcH-n two perfectly distinct conceptions 
 has made the division of philosophy into theoretical and 
 ])ra(lical unmeaninf:^, inasnuu h as the same principle is 
 assnined to npply to both s])heres. 
 
 'I'he will as a facnlty of desire is simply one of the 
 many causes in the world ol" nature, namely, that cause 
 which acts from conct-ptlons. All that is |)ossil)le or 
 necessary through will is said to \)v i)ractically [)ossil)lc 
 or practically necessary, and with this is contrasted that 
 which is ])hysically jjossihle or necessary, that is, whati-ver 
 is the effect ol a cause wliii h acts, not by means ol conce[)- 
 tions, but by the nu'cham'sm of lifeless matter, or by animal 
 instinct. r>ut this in noway settles the (juestion, whether 
 it is a conception ol nature, or an idea of freedom, 
 which gives the rule when the will acts as a cause. 
 
 'The distinction, however, is of the greatest conse- 
 quence. I'or, if a conception of nature determines the 
 will, the principles arc technically practical ; whereas, if 
 the will is determined by the idea of freedom, the 
 principles are morally jiractical. And as the divisions of 
 a science of reason are determined by the nature of the 
 princii)les on which each rests, the lornur will belong to 
 theoretical philoso[)hy or tlie science of nature, the latter 
 to practical pliilosophy or the science of morality. 
 
 All technically j)ractical rules of art and skill, or of 
 that ])ractical sagacity which gives us a command over 
 men and enables us to inlluence their wills, so far as 
 their princii)les rest on conceptions, must be regarded as 
 corollaries of theoretical philosophy Only as 
 
THE CRITIQUE OF JUDGMENT. ^ 309 
 
 179 Standing under the conception of freedom is the will free 
 from nature, and hence the laws of freedom together with 
 their consequences alone constitute practical philosophy. 
 The practical arts of surveying, housekeeping, farming, 
 statesmanship, dietetics, etc., and even the precepts by 
 which happiness may be attained, are merely technically 
 practical rules. Only those rules which rest on the idea 
 of freedom are morally practical. For such rules are laws 
 which do not, like those of nature, rest upon sensuous 
 conditions, but, on the contrary, upon a supersensible 
 principle ; and hence they form a separate branch of 
 philosophy, which is properly called practical philosophy. 
 
 180 //. The Realm of Philosophy. 
 
 The term field simply indicates the general relation of 
 an object to our faculty of knowledge, no matter whether 
 the conception of that object makes knowledge of it 
 possible or not. That part of a field in which knowledge 
 is possible, is a solid ground or territory {tcrritorium) for 
 conceptions and their appropriate faculty. That part of 
 the territory, again, for which laws are prescribed in con- 
 ceptions, is the domain or realm [ditto) of these concep- 
 tions and their correspondent faculty. Empirical concep- 
 tions have, therefore, nature, as the sum of sensible 
 objects, for their territory ; but that territory is for them 
 not a realm but merely a dzuelling-place {domidlium)^ for 
 although they are under law they are not themselves 
 the source of law, and hence the rules based upon them 
 are empirical or contingent. 
 I Si Although understanding and reason operate on the 
 same territory of experience, their laws are distinct, and 
 
3IO THE PHILOSOPHY OF KANT. 
 
 do not interfere with each other. The conceptions that 
 are appHcable to nature have as Httle influence on the 
 law of freedom as the latter on the former. It is true 
 that in the sensible world each realm is perpetually 
 limited by the other, but in their laws they are quite 
 independent. The reason why they do not constitute 
 one realm is that the conception of nature has a meaning 
 only in relation to objects of perception or phenomena, 
 not in relation to things in themselves ; while, on the 
 other hand, the object of freedom is intelligible as a 
 thing in itself, but cannot be given in a perception. 
 There can, therefore, be no theoretical knowledge of 
 either realm as a thing in itself, or supersensible object. 
 
 The whole unlimited field of the supersensible thus 
 lies entirely beyond our knowledge, and affords no solid 
 ground, and therefore no realm, either for understanding 
 or for reason. This field we must indeed occupy with 
 182 ideas in the interest of theoretical as well as of practical 
 reason, but we can produce no other warrant for our 
 occupation of it than a practical one, and so far as 
 theoretical knowledge is concerned the supersensible 
 therefore remains as far beyond our reach as ever. 
 
 Between the sensible realm of nature and the super- 
 sensible realm of freedom a gulf is fixed, which is as 
 impassable by theoretical reason as if they formed two 
 separate worlds. Yet it lies in the very idea of freedom 
 to realize in the world of sense the end presented in its 
 laws, and hence nature, in its formal aspect as conform- 
 able to law, must at least be capable of harmonizing with 
 that end. There must, then, be a principle which unites 
 the supersensible substrate of nature with the supersen- 
 sible, that is involved practically in the conception 
 
THE CRITIQUE OF JUDGMENT. 31I 
 
 of freedom. And although that prmciple does not lead 
 to a knowledge of the supersensible, and hence has no 
 realm peculiarly its own, it yet enables the mind to make 
 the transition from the theoretical to the practical point 
 of view. 
 
 ///. The Critique of Judgment as counectifig link between 
 the two divisions of Philosophy, 
 
 183 There are three absolutely irreducible faculties of the 
 mind, namely, knowledge, feeling, and desire. The 
 laws which govern the theoretical knowledge of nature 
 
 184 as a phenomenon, understanding supplies in its pure 
 
 185 « priori conceptions. The laws to which desire must 
 conform, are prescribed a priori by reason in the con 
 ception of freedom. Between knowledge and desire 
 stands the feeling of pleasure or pain, just as judgment 
 mediates between understanding and reason. We must, 
 therefore, suppose that judgment has an a priori principle 
 of its own, which is distinct from the principles of 
 understanding and reason. And as pleasure or pain is 
 necessarily associated with desire, either preceding it as 
 in the lower desires or following it when desire is deter- 
 mined by relation to the moral law, we must further 
 suppose that judgment makes possible the transition 
 from mere knowledge or the realm of nature to the realm 
 of freedom, as, in its logical use, it makes possible the 
 transition from understanding to reason. 
 
 IV. Judgment as a Faculty of a priori Laws. 
 
 Judgment in general is the faculty of thinking the par- 
 ticular as contained under the universal. If the universal, 
 
312 THE PHILOSOPHY OF KANT. 
 
 that is, the rule, principle, or law, is given, the judgment 
 which subsumes the particular under it is detenrmiafit. 
 But if only the particular is given, for which the universal 
 has to be found, the judgment is merely reflective. 
 1 86 The determinant judgment subsumes particulars under 
 the universal transcendental laws supplied by the under- 
 standing, and has no need to seek for a law of its own by 
 means of which the particulars of nature may be brought 
 under the universal. But nature has many forms, which 
 may be regarded as modifications of the universal trans- 
 cendental conceptions, and the former are unaffected 
 by the latter, which are but the general conditions, 
 without which nature as a sensible object would not 
 be possible at all. There must, therefore, be laws 
 for those forms also, and such laws, as being empirical, 
 may be contingent so far as our intelligence is concerned, 
 and may yet be regarded as following necessarily from a 
 principle, which is the condition of the unity of the 
 multifarious forms of nature, though it is unknown to us. 
 The reflective judgment, which is compelled to ascend 
 from the particular to the universal, therefore requires a 
 principle of its own ; and that principle it cannot borrow 
 from experience, because its function is just to unite all 
 empirical principles under higher ones, and so to make 
 their systematic connection possible. 
 
 The principle of judgment as reflective must therefore 
 be conceived as if it were a unity imposed on nature by 
 an intelligence diff'erent from ours, with a view to the 
 reduction of our knowledge of nature to a system of 
 particular laws. We cannot, however, assert that there 
 actually is an intelligence of this kind, for judgment does 
 not give a law to nature but only to itself. 
 
THE CRITIQUE OF JUDGMENT. 313 
 
 187 Now a conception which contains the ground of the 
 actuality of an object is an end^ and by the agreement of 
 a thing with a character which is only possible in accord- 
 ance with ends, we mean that its form implies purpose. 
 The principle of judgment, in its relation to the forms of 
 things which come under empirical laws in general, is 
 thus the idea that in all its manifold variety fiafiire is 
 purposive. That is to say, nature is conceived as if the 
 unity of its manifold empirical laws were due to an 
 intelligence. 
 
 V. The Principle that the Form of Nature implies Purpose 
 is a Transcendental Principle of Judgment. 
 
 A transcendental principle of judgment is one which 
 enables us to think a priori the universal condition 
 without which things could not be objects of our know- 
 ledge at all. A metaphysical principle, on the other 
 hand, is one through which we think a priori the condition 
 without which objects, the conception of which must be 
 given empirically, cannot be further determined a priori. 
 Thus the principle, that the changes of empirical 
 substances must have a cause, is transcendental ; but if 
 we say that their changes must have an external csLUse, the 
 principle is metaphysical. In the former case, such 
 merely ontological predicates, or pure conceptions, as 
 substance are employed ; in the latter case, the empirical 
 conception of a body as a movable thing in space is 
 
 188 required, although when this has once been obtained, 
 the predicate of motion under the influence of external 
 causes may be deduced quite a priori. Now, the 
 principle that nature is purposive, is a transcendental 
 principle. For the conception of objects, so far as 
 
314 THE PHILOSOPHY OF KANT. 
 
 they are thought as standing under this principle, is 
 merely the pure conception of oDJects of possible 
 experience in general, and therefore contains nothing 
 empirical. But the principle that actions are purposive, 
 which is implied in the idea of the determination of a 
 free will, is a metaphysical principle, because the concep- 
 tion of desire must be given empirically. At the same 
 time neither principle is empirical, but both are a priori, 
 for the predicate may be connected with the empirical 
 conception that forms the subject of the judgment 
 completely a priori, and without any new experience. 
 
 The conception that nature is purposive is a transcen- 
 dental principle. This is sufficiently obvious from the 
 a priori maxims of judgment which are employed in 
 scientific inquiries into the specific laws of nature Such 
 maxims are continually applied as occasion demands, in 
 the shape of axioms of metaphysical wisdom : " Nature 
 takes the shortest way {lex parsinioniae) "j "Nature makes 
 no leaps {lex co7itinui in naturd) " ; " Nature has many 
 laws, but few principles {principia praeter fiecessitafem non 
 sufit 7nultiplicanda)," etc. 
 
 To attempt an explanation of the origin of these 
 propositions psychologically, is to go straight against 
 their sense. For they do not tell us what happens, that 
 189 is, by what rule our faculties operate or how we actually 
 judge, but they prescribe how we should judge ; and a 
 logical necessity of this sort is inexplicable if those 
 principles are merely empirical. The idea that nature is 
 purposive, is therefore a transcendental principle and 
 requires a transcendental deduction. 
 
 That which is at once seen to be necessary by the 
 principles which make experience possible, are the 
 
THE CRITIQUE OF JUDGMENT. 315 
 
 universal laws, without which nature, as an object of 
 sense, is not conceivable at all ; and these laws rest on 
 the categories in their application to the formal a priori 
 conditions of all experience that we can possibly have. 
 
 In relation to these laws judgment is determinant, its 
 sole function beinsr to subsume Darticulars under the laws 
 given to it. Thus understanding says, that every change 
 has a cause, or it lays down a universal law of nature. 
 Transcendental judgment, on the other hand, merely 
 presents the a priori condition on which subsumption 
 under the conception placed before it by understanding 
 takes place, namely, succession in the determinations of 
 one and the same thing. The law of causality is therefore 
 known to be an absolutely necessary condition of nature 
 as an object of possible experience. But the objects of 
 empirical knowledge are determined in many other ways 
 than by the formal condition of time ; at any rate we 
 may say a priori that they are at least capable of being 
 determined in many other ways. Hence the specific 
 forms of nature may be causes, not only in virtue of their 
 common character as belonging to nature in general, but 
 in an infinite variety of ways ; and each species of cause 
 must have its own necessary rule or law, although the 
 nature and limits of our knowledge may prevent us from 
 comprehending the necessity of the rule. We must, 
 therefore, suppose the empirical laws of nature to be 
 possibly infinitely various, and to be for us contingent or 
 incapable of being known a priori. So far as these 
 empirical laws are concerned, nature, as a possible unity 
 of experience or a system of laws, must accordingly be 
 190 regarded as contingent. Yet we must presuppose and 
 assume such a unity, for otherwise the thoroughgoing 
 
3l6 THE PHILOSOPHY OF KANT. 
 
 connection of empirical knowledge in a whole of experi- 
 ence would be impossible. The universal laws of nature 
 no doubt enable us to connect things in a system, so far 
 as they are viewed as belonging to nature in the most 
 general sense of the term, but not to connect them 
 in their specific character as particular modes of nature. 
 Judgment must therefore assume a pj-iori^ as a principle 
 required for its own use, that what in the empirical laws 
 of nature is from our human point of view contingent, yet 
 involves a unity in the connection of the multifarious laws 
 of nature, that are capable of being experienced, a unity 
 which can certainly be thought although it cannot be 
 comprehended by us. Now, a unity which is demanded by 
 our intelligence, but which is known as in itself contingent, 
 necessarily presents itself to us as the idea that objects 
 are purposive. Hence judgment, in relation to things 
 that may stand under empirical laws not yet discovered, 
 is merely reflective, and is compelled to think of nature 
 as in its special laws purposive as regards our knowledge, 
 a principle which is expressed in such maxims of judg- 
 ment as those that were cited above. This transcen- 
 dental conception of purpose in nature is neither a 
 conception of nature nor of freedom, for it attributes 
 nothing to nature as an object, but merely represents 
 the way in which we must necessarily proceed 
 in reflecting on natural objects, with a view to a 
 thoroughly connected experience. It is, therefore, a 
 maxim or subjective principle of judgment. 
 
THE CRITIQUE OF JUDGMENT. 317 
 
 193 VI. The Feeling of Pleasure connected with the Conception 
 
 that Nature is Purposive. 
 
 The reduction of the special laws of nature to unity of 
 principle is an end which understanding necessarily seeks 
 to secure. With the attainment of that end there arises 
 a feeling of pleasure which is determined by a ground a 
 priori for everyone, and indeed from the mere adaptation 
 
 194 of the object to our faculty of knowledge. . . . The dis- 
 covery that two or more heterogeneous laws of nature 
 may be combined in a common principle gives rise to a 
 very marked pleasure, and often to a feeling of wonder 
 that even familiarity does not destroy. 
 
 195 VII. The Esthetic Consciousness of Purpose in Nature. 
 
 The aesthetic character of an idea is determined solely 
 by its relation to the subject; its logical validity has 
 reference to the object as capable of being known. In 
 the apprehension of a sensible object both relations are 
 implied. In the presentation of objects as outside of me, 
 their spacial quality is merely a subjective element of my 
 perception, and they are accordingly thought of simply as 
 phenomena. But space is also an integral element in the 
 knowledge of phenomena. Seiisation, again, while no 
 doubt it is a purely subjective element in the perception 
 of objects as without us, yet affords the matter {reale) of 
 that which is given as existing, and hence it is essential 
 to the knowledge of those objects. But the feeling 
 oi pleasure ox pain, which accompanies our knowledge of 
 sensible objects, does not enter as an ingredient into know- 
 ledge at all, for it contributes nothing to the knowledge of 
 
3l8 THE PHILOSOPHY OF KANT. 
 
 an object, though it may be the result of that knowledge. 
 
 196 That an object of perception should be purposive, is 
 therefore no property of the object. Such an object is 
 therefore said to display purpose only if a feeling of 
 pleasure is immediately connected with the idea of it. 
 Here therefore we have the aesthetic consciousness of 
 
 purpose in nature When imagination, as the faculty 
 
 of a priori perception, is found to be in harmony with 
 understanding, and a feeling of pleasure is awakened by its 
 exercise, the object must be regarded as adapted to the 
 
 reflective judgment The object is then said to be 
 
 beautiful^ and the faculty which judges it to be so is called 
 T'aste. 
 
 198 The sensibility to pleasure arising from reflection on 
 the forms of things, whether of nature or of art, indicates 
 not only an adaptation of objects to reflective judgment, 
 an adaptation which is in conformity with the conception 
 of nature in the subject, but it also implies, conversely, 
 an adaptation of the subject in virtue of the conception 
 of freedom to the form or even formlessness of objects. 
 Hence the aesthetic judgment is related to the emotion 
 of the sublijne as well as to the feeling of the beautiful. 
 The Critique of ^Esthetic Judgment has therefore two 
 main divisions. 
 
 VIII. Logical Idea of Purpose in Nature. 
 
 An object of experience may be viewed as purposive 
 only relatively to the subject that is conscious of it, in 
 other words, the idea that it is purposive may rest upon 
 the mere harmony of the form of the object with our 
 faculty of knowledge, a form which is directly apprehended 
 without the intermediation of any conception. But the 
 
THE CRITIQUE OF JUDGMENT. 319 
 
 object may also be regarded as in itself purposive, if 
 the form of the thing, as given in a conception which 
 precedes it and is its condition, agrees with the possi- 
 199 bility of the thing itself. The former rests upon the 
 pleasure immediately felt in mere reflection upon the 
 form of an object; the latter requires us to have a 
 definite knowledge of an object through a conception ; 
 and as this knowledge is quite independent of any feel- 
 ing of pleasure in the contemplation of the object, it pre- 
 supposes a judgment of understanding. If the conception 
 of an object is given, the work of judgment lies in the 
 presentation {exhibitio) of a perception corresponding to 
 it. And we may either, as in art, endeavour to realize in 
 perception a conception set up by our own imagination 
 as an end, or we may make use of our conception of an 
 end in judging of certain natural objects, as, for instance, 
 in judging of organized bodies. In the latter case, not 
 merely the form of the thing implies purpose, but the 
 thing itself as a product is regarded as a natural end. 
 Now, although the subjective consciousness of purpose 
 does not imply any conception of an object, we may still, 
 by analogy with the conception of an end, attribute to 
 nature as it were a regard for our faculty of knowledge ; 
 hence we may look upon natural beauty as the presentation 
 of the conception of a formal or subjective purpose, and 
 we may regard natural end as the prese?itation of the con- 
 ception of a real or objective purpose. The former is 
 the object of aesthetic judgment or Taste, the latter is the 
 object of certain logical judgments, which understanding 
 and reason make by means of conceptions. The Critique 
 of Judgment has accordingly two parts, dealing respec- 
 tively with cesthetic judgment and teleological judgment. 
 
320 THE PHILOSOPHY OF KANT. 
 
 201 IX. Co7i?iection of Undersfa?idt?ig a?id Reason through 
 
 Judgment. 
 
 Understanding prescribes the a priori laws which make 
 experience or a theoretical knowledge of nature as an 
 object of sense possible. Reason prescribes the a priori 
 laws of freedom, and being itself a supersensible cause in 
 the subject, it gives rise to an unconditionally practical 
 knowledge. The realm of nature, which is under the laws 
 of understanding, and the realm of freedom which con- 
 forms to the laws of reason, are entirely removed from all 
 mutual influence by the great gulf which sunders the super- 
 sensible from the phenomenal world. The idea of free- 
 dom adds nothing to the theoretical knowledge of nature, 
 nor does the conception of nature affect the practical laws 
 of freedom. So far, therefore, there is no possibility 
 of throwing a bridge from the one realm to the other. 
 But while it belongs to the very idea of a free cause to be 
 independent of nature, and while the sensible cannot 
 determine that which in the subject is supersensible; yet 
 the converse is not impossible in a certain sense, and in 
 fact is implied in the very conception of a free cause, the 
 
 202 effect of which ought to be an event in the world. The 
 word caiise^ when applied to the supersensible, signifies 
 merely the ground which determines the causality of 
 things to an effect in accordance with the laws of nature ; 
 and while the possibility of causality in this sense cannot 
 be understood, it can be conclusively shown that it is not 
 self-contradictory, as some have maintained it to be. 
 The effect of freedom is the ultimate end which ought to 
 exist as a phenomenon in the world of sense, and the 
 condition of its possible realization is presupposed as 
 
\ 
 
 THE CRITIQUE OF JUDGMENT. 32 1 
 
 existing in the nature of man as a sensible being. Judg- 
 ment, as presupposing this a priori condition, independ- 
 ently of the practical, supplies us with the conception of 
 purpose in nature, a conception which mediates between 
 nature and freedom, and makes possible the transition 
 from the conception of conformity to law to the concep- 
 tion of an ultimate end. 
 
 The fact that understanding prescribes a priori IsiVfs to 
 nature, shows that nature is known merely as a phenom- 
 enon, and at the same time points to a supersensible 
 substrate of nature. This, however, leaves the nature of 
 this substrate quite undetermined. But judgment, by 
 means of its a priori principle for estimating nature 
 203 according to possible particular laws, brings the super- 
 sensible substrate, both in us and without us, within 
 reach of determination by our intellect. Reason, again, 
 through its practical a priori law actually determines it; 
 and thus judgment enables us to make the transition from 
 the realm of nature to that of freedom. 
 
 As to the higher faculties of the mind, that is, those 
 which contain an autonomy, understanding contains the 
 constitutive principles of knowledge ; judgment those for 
 the feeling of pleasure or pain ; reason those relative 
 to desire. The conception supplied by judgment of 
 purpose in nature is one of the conceptions of 
 nature, but it is merely a regulative principle of know- 
 ledge. The aesthetic judgment, as concerned with 
 certain objects of nature or art, which are the occasion of 
 that principle being applied, is a constitutive principle in 
 relation to the feeling of pleasure or pain. The spon- 
 taneity of the faculties of knowledge, from the harmonious 
 operation of which that pleasure arises, by intensifying 
 
322 
 
 THE PHILOSOPHY OF KANT. 
 
 the susceptibility of the mind for the moral feeling, makes 
 the conception of purpose in nature the fit connecting 
 link between the conception of nature and the idea of 
 freedom as manifested in its effects, inasmuch as these 
 imply the sensibility of the mind to moral feeling. 
 
 The following table exhibits all the higher faculties in 
 their systematic connection : — 
 
 204 Faculties of the Mind. Faculties of Knowledge. 
 
 Knowledge. Understanding. 
 
 Feeling of Pleasure or Pain. Judgment. 
 
 Desire. Reason. 
 
 A priori Principles. 
 Subordination to Law. 
 Purpose. 
 Ultimate End. 
 
 Application to 
 Nature. 
 Art. 
 Freedom. 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 323 
 
 THE CRITIQUE OF TELEOLOGICAL 
 JUDGMENT. 
 
 ^•j^ Section I. — Analytic of Teleological Judg^nent. 
 
 62. For77ial Objective Purpose. 
 
 Geometrical figures drawn on a principle often show a 
 remarkable objective adaptation to the purpose for which 
 they are employed, namely, the solution of several 
 problems by a single method, or of one problem in an 
 infinite variety of ways. The adaptation is here evidently 
 objective and intellectual, not subjective and aesthetic. 
 But, although such figures are adapted to the end in 
 view, namely, the production of a variety of geometrical 
 forms, they are regarded as possible independently of the 
 particular use made of them, and hence their adaptation 
 to that end is not the condition of their very existence in 
 
 376 thought This intellectual adaptation to an end is 
 
 therefore no doubt objective, and not like aesthetic 
 adaptation subjective ; but it is not real, but merely 
 formal. It can be conceived as adaptation in general 
 without the conception of end being presupposed, and 
 hence it is not an instance of teleology. 
 
 It is quite different when a number of things are pre- 
 sented as without me and enclosed within well-defined 
 limits, as, for example, trees, flowers, and walks disposed 
 in regular order in a garden ; for these are actually existing 
 things which must be known empirically, and not merely 
 
324 THE PHILOSOPHY OF KANT. 
 
 an idea of my own which is determined a priori according 
 to a principle. The adaptation in this case is empirical 
 or real^ and presupposes the conception of an end. 
 
 378 63. Relative as contrasted with I)iternal Purpose. 
 
 Experience leads our judgment to the conception of an 
 objective material purpose, that is, to the conception of 
 
 379 an end in nature, only if we find ourselves compelled 
 to presuppose the activity of a cause that is determined 
 to action by conceptions as the necessary condition of the 
 existence of a given effect. This may occur either when 
 the effect is regarded as itself a product of art, or when 
 it is regarded merely as material for the art of other 
 possible natural beings ; in other words, it is either an 
 end, or a means for the ends of other causes. Purpose 
 in the latter case is called utility in relation to man, 
 advantage when we are speaking of other creatures, and 
 
 4s merely relative ; while purpose in the former case is an 
 internal purpose exhibited in a natural being. 
 
 A sandy soil is most advantageous for the growth of 
 pine trees. Now, when the sea withdrew from the land 
 on our northern shores^ it left behind it large tracts of 
 ^80 sand, on which pine forests have grown up. Shall we 
 then say that the original deposit of these tracts of sand 
 is evidence of an end of nature, simply because it is of 
 advantage to pine trees ? Manifestly if this is an end of 
 nature, the sand also must be regarded as a relative end, 
 for which the withdrawal of the sea was a means. So also 
 if cattle, sheep, horses, etc., are to exist, grass must cover 
 
 the earth The objective purpose here supposed 
 
 is therefore not exhibited by things themselves, but is 
 merely relative or contingent. 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 325 
 
 381 From this it is quite plain, that purpose can be 
 regarded as an external natural end, only on condition 
 that the existence of that for which something else is 
 immediately or remotely advantageous, is in itself an end 
 of nature. But this can never appear from a mgre con- 
 templation of nature, and therefore relative purpose, 
 although it points hypothetically to natural ends, does 
 not of itself justify an absolute teleological judgment. 
 
 382 64. The Properties of Things which are Natural Ends. 
 
 To see that a thing is really a natural end, or cannot 
 be explained in a mechanical way, its form must be 
 incapable of explanation by the ordinary laws of nature 
 that are known and applied by the understanding to 
 objects of sense; in other words, it must be of such a 
 nature that it cannot be known in experience even as an 
 effect, except on presupposition of conceptions of reason. 
 Simply to know what are the conditions required for the 
 production of such a natural object, reason must per- 
 ceive its form to be necessary. Now, the very fact 
 that in the present case the form of the object is not 
 necessary but accidental, so far as the ordinary laws of 
 nature are concerned, is itself a ground for regarding that 
 form as possible only through reason. And as reason or 
 will is the faculty of acting from ends, an object which 
 is regarded as possible only through reason must be 
 conceived as an end. 
 
 383 To know a thing, however, not only as an end, but as 
 a 7iatiiral end^ more than this is required. A thing exists 
 as a natural end only if it is, in a double sense, its 
 own cause and its own effect. This may be illustrated by 
 an example. In the first place, a tree produces another 
 
326 THE PHILOSOPHY OF KANT. 
 
 tree according to a well-known natural law. The tree so 
 produced is of the same species ; hence a tree, being 
 continually self-produced, is, on the one hand, its own 
 effect, and, on the other hand, its own cause, and by such 
 continual self-production it perpetuates itself as a species. 
 In the second place, a tree is self-productive, even as 
 an individual. This no doubt is what we call growth, 
 but it must be observed that growth is quite different 
 from any mere increase in size according to mechanical 
 laws. The matter which the tree incorporates, it pre- 
 viously works up into a specifically peculiar quality, 
 which is not due to any natural mechanism outside of it ; 
 ar4 thus it develops itself by means of a material, which, 
 as assimilated, is its own product. No doubt the tree, so 
 far as the constituents obtained from external nature are 
 concerned, must be regarded as an educt; but, on the 
 other hand, it displays a power of separating, recombining 
 and shaping this raw material, which is far beyond the 
 384 reach of human art. In the third place, each part of the 
 tree is self-productive, so that the preservation of one part 
 is dependent on the preservation of all the rest. A bud 
 inoculated on the twig of another tree produces a plant 
 of its own kind, and so also a scion engrafted on a foreign 
 stem. We may, therefore, regard each twig or leaf of the 
 same tree as engrafted or inoculated on it, or as an inde- 
 pendent tree, externally attached to another and parasiti- 
 cally nourished by it. And while the leaves are a 
 product of the tree, the tree is in turn dependent for 
 its growth upon their effect on the stem, for if it is 
 repeatedly denuded of its leaves it dies. 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 327 
 
 65. Things which are Natural Ends are Organized 
 
 Beings. 
 
 Causal connection as thought by the understanding 
 always constitutes a regressive series of causes and 
 
 385 effects This sort of causal connection we call 
 
 that of efficient causes {nexus effectivus). But another 
 kind of causal connection is conceivable, which rests 
 upon the conception of ends. Here the series, if it can 
 be called a series, may be taken either backwards or 
 forwards, and hence that which has been named effect is 
 with equal propriety termed the cause of that of which 
 
 it is the effect Such a causal connection we name 
 
 that of final cause {nexus finalis). 
 
 For a thing to be a natural end, in the first place, its 
 parts must be possible only in relation to the whole. As 
 an end the thing itself is comprehended under a 
 conception or idea, which must determine a priori all 
 that is to be contained in it. This, however, does not 
 distinguish a natural product from an artificial product, 
 in which the cause is an intelligent being, distinct from 
 the material parts that are brought together and combined 
 in accordance with the idea of a whole that is possible 
 only by means of them. 
 
 Hence, in the second place, a natural product must 
 in itself or in its inner possibility imply relation to an 
 end ; in other words, it must be possible as a natural end 
 irrespective of any intelligent cause external to it. 
 Accordingly, the parts of such a natural product, which 
 combine in the unity of a whole, must be reciprocally 
 cause and effect of each other's form. Only in this way 
 can the idea of the whole determine conversely the form 
 
328 THE PHILOSOPHY OF KANT. 
 
 and combination of all the parts, not indeed as cause — 
 for then we should have an artificial product — but as the 
 ground on which the thing is known, by the subject 
 judging of it, in the systematic unity of the form and the 
 combination of all its parts. 
 
 386 A body is therefore a natural end, only if all its 
 parts mutually depend upon each other both as to their 
 form and their combination, and are thus themselves the 
 cause of the whole ; while, conversely, the idea of the 
 whole may be regarded as the cause of the body in 
 accordance with a principle. In such a body, accordingly, 
 the conjunction of efficient causes is at the same time 
 regarded as an effect through final causes. 
 
 In a natural product, each part not only exists by means 
 of the other parts, but is conceived as existing for the 
 sake of th.Q others and of the whole, that is, as an instrument 
 or organ ; and not only so, ^at its parts are all organs 
 reciprocally producing one another, which is never the 
 case with artificial instruments. Only a product of this 
 kind is called a natural end, and it receives this name 
 just because it is an organized and self-organizing being. 
 
 388 Organized beings are the only things in nature which, 
 in themselves and apart altogether from their relation to 
 other things, can be conceived to exist only as ends. 
 The conception of an end of tiature^ as distinguished from 
 a practical end, first obtains objective reality from a 
 consideration of such beings ; a.nd apart from them, the 
 Ideological consideration of nature as a special principle 
 of judgment would have no justification whatever. 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 329 
 
 66. The Principle by which Organized Beings are judged 
 to be internally purposive. 
 
 The principle that is applied when a thing is judged 
 to be internally purposive, a principle which is at the 
 same time a definition, is this : An organized product of 
 nature is one in which all the parts are reciprocally end 
 and means. Nothing in an organized being is useless, or 
 without purpose, and nothing in it can be ascribed to 
 blind natural mechanism. 
 
 This principle finds its occasion in the methodical 
 observation of experience, but, as it affirms the idea of 
 purpose to be of universal necessity, it cannot be derived 
 from experience, but must be a priori. At the same time, 
 as ends exist only as an idea in the judging subject, not 
 
 389 in any eflEicient cause, it is merely a regulative principle, 
 or a maxim, for judging of the internal purpose exhibited 
 in organized beings. 
 
 390 67. The Teleological Judgment in regard to Nature as a 
 
 System of Ends. 
 
 As has been shown above, external purpose does not 
 justify us in saying that things can be known to exist only 
 as ends of nature, or in employing the principle of final 
 cause to account for the purpose which may seem to be 
 
 implied in their effects Now, if there is no reason 
 
 for regarding a thing as in itself end, the external relation 
 can be only hypothetically judged to imply purpose. 
 
 To regard a thing as a natural end on account of its 
 internal form, is a very diff"erent thing from holding the 
 existence of that thing to be an end of nature. The 
 latter assertion is justifiable only if it can be shown, not 
 
33° THE PHILOSOPHY OF KANT. 
 
 merely that we have the conception of a possible end, 
 but that we have a knowledge of the ultimate end {scopiis) 
 of nature. But this requires the relation of such know- 
 391 ledge to something which is supersensible and far 
 transcends all our teleological knowledge of nature, for 
 the end of nature must be sought beyond nature. The 
 internal form of a simple blade of grass is sufficient to 
 show that for our human faculty of judgment its origin is 
 possible only according to the rule of ends. But if we 
 change our point of view, and look merely at its external 
 adaptation for the use of other natural beings, we get no 
 categorical end, but, finding always a new condition of 
 such adaptation, we are led to the idea of the un- 
 conditioned existence of a thing as ultimate end, and so 
 entirely beyond the physico-teleological consideration of 
 the world. So conceived the thing is not even a natural 
 end, for it is no longer regarded as a natural product. 
 
 Only organized matter, as in its specific form a product 
 of nature, necessarily demands the application of the 
 concept'on of natural end. But this conception, when 
 once obtained, necessarily leads to the idea of the whole 
 of nature as a system of ends, and to this idea all natural 
 mechanism must be subordinated in accordance with 
 principles ot reason. 
 
 It is manifest that this is not a principle of the 
 determinant, but only of the reflective judgment; that it 
 is regulative and not constitutive : and that it supplies us 
 with a guiding conception, by means of which natural 
 objects already determined may be considered according 
 to a new law and order, and our knowledge of them 
 extended by means of the principle of final cause. But 
 this principle in no way interferes with the principle of 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 33 1 
 
 mechanical causality already applied to them, nor does 
 
 392 it entitle us to regard anything whatever as a purposive 
 end of nature. 
 
 After the teleological judgment by natural ends as 
 applied to organized beings has brought us to the idea^t 
 a great system of ends of nature, even the beauty of 
 nature, that is, the harmony of nature with the free play 
 of our faculties of knowledge in apprehending and 
 judging of its appearance, may be regarded as a sort of 
 
 393 objective purpose exhibited by nature in a systematic 
 whole of which man is a member. 
 
 397 Section II. — Dialectic of Teleological Judgment. 
 
 70. Antinomy of Judgment. 
 
 398 In dealing with nature as a totality of sensible objects, 
 reason may either take its stand upon the a priori \dc^% 
 prescribed to nature by understanding, or upon laws 
 which are capable of indefinite addition as experience is 
 gradually extended. In applying the former sort of 
 laws, that is, the u?iiversal laws of material nature, judg- 
 ment needs no special principle of reflection; for an 
 objective principle is given to it by understanding, and 
 it is, therefore, merely determinant. But so multifarious 
 and diverse are the particular laws which have to be 
 learned from experience, that judgment must here supply 
 its own principle, if it is to conduct its investigations into 
 the phenomena of nature in an orderly way. Without 
 such a guiding-thread there is not the least hope that 
 our empirical knowledge may form a thoroughly con- 
 nected and orderly system, and may reduce the empirical 
 laws of nature to unity. Now, in a contingent unity of 
 
33 2 THE PHILOSOPHY OF KANT. 
 
 this kind it may very well happen, that judgment in its 
 reflection proceeds from either of those principles. It 
 may simply apply the a priori laws of understanding, or 
 it may start from the special principle, by means of which 
 reason, on occasion of particular experiences, seeks to 
 399 form a judgment upon corporeal nature and its laws. 
 Hence it comes, that these two maxims seem to be 
 mutually exclusive, and that a dialectic arises which 
 leads judgment into error in its application of the 
 principle of reflection. 
 
 The first ??iaxi7?i of judgment is the position : All 
 production of material things and the forms of material 
 things must be judged to be possible according to purely 
 mechanical laws. 
 
 The second 77iaxi7n is the coiinterposition : Some products 
 of material nature cannot be judged to be possible 
 according to purely mechanical laws, but require quite a 
 different law of causaUty, namely, that of final cause. 
 
 Now, if these regulative principles in the investigation 
 of nature are converted into constitutive principles, which 
 are supposed to determine the possibility of objects them- 
 selves, they will run thus : 
 
 Position : All production of material things is possible 
 according to purely mechanical laws. 
 
 Coiinterposition : Some production of material things is 
 not possible according to purely mechanical laws. 
 
 If we take the last pair of propositions as objective 
 principles of the determinant judgment, each is contra- 
 dictory of the other, and hence one of them must be 
 false. We shall then no doubt have an antinomy, 
 but it will be an antinomy not of judgment but of 
 reason. Reason, however, can prove neither the one 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 333 
 
 proposition nor the other, for there can be no a priori 
 principle which determines the possibility of things so 
 far as the purely empirical laws of nature are concerned. 
 The first two propositions, on the other hand, if they 
 are regarded simply as maxims of reflective judgment, 
 are not really contradictory. For, to say that all events 
 in the material world, and, therefore, all the forms which 
 are products of nature, must hQ judged to be possible on 
 purely mechanical laws, is not to say that they are possible 
 in this way alone, or apart from any other sort of causality. 
 All that is implied is, that we ought in all cases reflectively 
 to judge them by the principle of natural mechanism, and 
 to make this principle the foundation of all our investiga- 
 tions, and apply it as far as we can, since without it there 
 can, properly speaking, be no knowledge of nature at all. 
 But this in no way prevents us, if occasion is given 
 400 for it, from following the guiding-thread of the second 
 principle in our reflection upon certain natural forms, and 
 even by instigation of these upon the whole of nature, 
 the principle, namely, of final cause, which is quite 
 distinct from that employed in the explanation of natural 
 mechanism. The value of reflection of the kind indicated 
 in the first maxim is not in any way denied, but on the 
 contrary we are told to follow it as far as we can. Nor is 
 it said, that those forms are not possible at all on the 
 principle of natural mechanism : all that is said is, that 
 by following this path human reason will never be able to 
 discover any ground of the specific character of natural 
 ends, although it will certainly gain increased knowledge 
 of natural laws. Thus it is left undetermined, whether 
 in the inner ground of nature, which to us is unknown, 
 conjunction by physical mechanism and conjunction by 
 
334 THE PHILOSOPHY OF KANT. 
 
 ends may not themselves be connected together in the 
 same thing by one principle. We must conclude, how- 
 ever, that our reason is not in a position to unite the two 
 principles, and that it is not the determinant but the 
 reflective judgment which compels us to explain the 
 possibility of certain forms of nature by means of a 
 different principle from that of natural mechanism. 
 
 413 76. Remark. 
 
 Without conceptions of understanding, for which an 
 objective reality must be presented, theoretical reason 
 can make no objective or synthetical judgments. In 
 
 414 itself it contains no constitutive principle whatever, but 
 
 merely regulative principles Now the very nature 
 
 of our intelligence compels us to distinguish between the 
 possible and the actual. Such a distinction would not 
 be made, did not our knowledge involve the exercise of 
 two heterogeneous faculties, — understanding for concep- 
 tions and sensible perception for objects corresponding 
 to conceptions. Were our intelligence perceptive, its 
 
 objects would always be actual The distinction 
 
 of things into possible and actual is therefore a subjective 
 distinction, which is valid for human reason merely 
 because we can ahvays think something that has no 
 
 415 reality, or suppose something of whicn we have no con- 
 ception to be given as an object. That possible things 
 may not be actual, and, as a consequence, that actuality 
 cannot be deduced from possibility, is certainly true 
 when we are speaking of human reason, though it does 
 not follow that such a distinction applies to things them- 
 selves. That it has no such application is indeed plain 
 from, the irrepressible tendency of reason to suppose some 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 335 
 
 unconditionally necessary existence, or original ground, 
 in which the distinction of possible and actual no longer 
 holds good. 
 
 The conception of an absolutely necessary being is 
 thus an indispensable idea of reason, but it is an idea 
 which remains for human inteUigence a problem that it 
 cannot solve. It arises from the peculiar nature of our 
 faculties of knowledge, and therefore it does not hold 
 true objectively but merely subjectively. We cannot say 
 that such an idea is essential to every intelligence, for we 
 have no right to assume that in all thinking beings there 
 are two conditions of knowledge, so different in their 
 nature c.s thought and perception, and therefore we have 
 no right to suppose that in all thinking beings the 
 conditions of possibility and actuality are different. An 
 intelligence for whom this distinction did not exist, might 
 say • All objects that I know are, that is, exist; and such 
 a being could never suppose some objects to be possible 
 that have no existence, and therefore to be contingent 
 when they do exist, nor could it, in contrast to those 
 objects, think of others as necessary. 
 416 Just as theoretical reason must assume as an idea the 
 unconditioned necessity of the original ground of nature, 
 so practical reason presupposes its own unconditioned 
 causality, or freedom, v. 'lich is implied in the conscious- 
 ness of its own moral commands. Here the objective 
 necessity of an act, as being a duty, is opposed to the 
 necessity which it would have as^ an event, if its ground 
 lay in nature and not in freedom, that is, in the causality 
 of reason. The morally necessary act is regarded as 
 physically quite contingent, since that which ot^g/if 
 necessarily to take place, often does not take place. It 
 
336 THE PHILOSOPHY OF KANT. 
 
 is evidently owing to the subjective constitution of our 
 practical faculty, that moral laws must be represented as 
 commands, and the acts conforming to them as duties ; 
 and that reason expresses this necessity not in the form 
 that something is or happens, but in the form that it 
 ought to be. This would not be the case were reason 
 considered as a cause which acts quite independently 
 of sensibility, that is, independently of the subjective 
 condition of its application to objects of nature, and 
 therefore as a cause in an intelligible world that is 
 completely in accordance with moral laws. For in such 
 a world there would be no distinction between being and 
 doing, between a practical law of that which is possible 
 through us, and the theoretical law of that which is 
 actual through us. A purely intelligible world, then, 
 would be one in which whatever is possible is at 
 the same time actual, just because it is good. But even 
 freedom, which is the formal condition of an intelligible 
 world, is for us a transcendent conception, and is therefore 
 incapable of serving as a constitutive principle for deter- 
 mining an object and its objective reality. Yet, although 
 our nature is partly sensuous, freedom, in so far as it 
 involves the idea of conformity to reason, is for us, and 
 417 all other rational beings that have a connection with the 
 world of sense, a universal regulative p7'inciple. This prin- 
 ciple does not objectively determine the nature of free- 
 dom, but it commands everyone to act in accordance 
 with the idea of freedom, and that as absolutely as if it 
 were a constitutive principle. 
 
 Let us see the bearing of these considerations on the 
 topic immediately in hand. Between natural mechanism 
 and the technic of nature, that is, its teleological connec- 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 337 
 
 tion, there would be for us no distinction, were it not that 
 our intelligence is compelled by its very nature to advance 
 from the universal to the particular. There can, there- 
 fore, be no knowledge of the adaptation of the particular 
 to an end, and consequently no determinant judgments, 
 unless judgment has a universal law under which it may 
 subsume the particular. Now, the particular, as such, has 
 a certain contingency with respect to the universal ; and 
 yet reason demands conformity with law in the reduction 
 of particular laws of nature to unity. Conformity with 
 law in the case of the contingent is called purpose, 
 and from such a universal, particular laws, so far as 
 they imply a contingent element, cannot be derived a 
 priori. Hence the conception that natural products are 
 purposive, necessary as it is for our judgment, does not 
 enable us to determine the objects themselves. It is a 
 subjective or regulative principle of reason, although for 
 human judgment it has the same validity as if it were an 
 objective or constitutive principle. 
 
 77. The conceptio7i of Natural End as due to the peculiar 
 character of our Intellige7ice. 
 
 There are certain peculiarities of even our higher 
 faculty of knowledge which it is very natural to transfer 
 as objective predicates to things. But they really belong 
 only to ideas, for no possible object of experience can be 
 presented which corresponds to them. This holds good 
 418 even of the conception of a natural end, which as a 
 predicate can exist nowhere but in the idea. But, 
 as the effect corresponding to this idea, that is, the 
 product itself, is a real object in nature, the conception 
 
^^S THE PHILOSOPHY OF KANT. 
 
 of nature as a being acting from an end seems to 
 make the idea of a natural end a constitutive principle. 
 In this respect the idea of a natural end is different from 
 all other ideas. 
 
 The difference, however, lies in the fact that this idea 
 is not a principle of reason for the understanding, but 
 only for the judgment, and is therefore merely the appli- 
 cation of an intelligence in general to possible objects of 
 experience. For here judgment is not determinant but 
 merely reflective, and hence although the object is given 
 in experience, judgment cannot determine it by the idea, 
 but can only reflect on it. 
 
 It is therefore a peculiarity of human intelligence that 
 in its judgment upon natural things it assumes the form 
 of reflection. Now this suggests the idea of an intelli- 
 gence different from ours and presupposed in it, just as 
 in the Critique of Pure Reason it was by supposing the 
 possibility of a perception different from ours, that we 
 were able to see that our perception is by its very nature 
 limited to phenomena. It is, then, by reference to this 
 supposed intelligence that we are able to say : Certain 
 natural products, from the very nature of our intelligence, 
 must be considered by us as if they could not exist at all 
 unless they had been produced purposely, or from 
 conceived ends. But we cannot venture to say that 
 there actually is a particular cause which acts from 
 such ends, or that an intelligence higher than ours 
 may not find in the mere mechanism of nature, as 
 a sort of causality conceivable apart from intelligence, 
 a sufficient explanation of the possibility of such natural 
 products. 
 
 We must therefore expect to find that there is a certain 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 339 
 
 contingency in the nature of our intelligence as related to 
 its faculty of judgment, and if we can show wherein this 
 contingency consists, we shall learn how our intelligence 
 differs from other possible intelligences. 
 
 It is not difficult to see that the contingency lies in the 
 particular, which it is the function of judgment to bring 
 under the universal that is supplied in the conceptions of 
 
 419 understanding. For the universal of ^wr understanding 
 does not determine the particular, and it is contingent in 
 how many ways different things which agree in a common 
 mark may present themselves to our observation. 
 Knowledge involves perception as well as conception. 
 Now, a perfectly spontaneous faculty of perception would 
 be a faculty of knowledge different from sensibility, and 
 quite independent of it ; in other words, an intelligence 
 in the most general sense of the term. Thus we are able to 
 conceive of a perceptive intelligence, but only negatively 
 or simply as not discursive ; in other words, we can think 
 of an intelligence which does not advance from the 
 universal to the individual through the particular. For 
 such an intelligence there would not be that contingency 
 in the adaptation of particular laws of nature to under- 
 standing, which makes it so difficult for us to reduce the 
 multifarious forms of nature to the unity of knowledge. 
 
 In order, then, to think at least the possibility of such 
 an adaptation of natural things to our faculty of judgment, 
 we must at the same time conceive of another intelligence, 
 
 420 by reference to which, and apart from any end attributed 
 to it, we may represent as necessary that harmony of 
 natural laws with our faculty of judgment, which for our 
 intelligence can be thought only through the medium of 
 ends. 
 
340 THE PHILOSOPHY OF KANT. 
 
 It is the nature of our intelligence to proceed in 
 knowledge from an analytical tmiversal^ or a conception, 
 to the particular as given in empirical perception. The 
 multiplicity of the latter thus remains undetermined, 
 until judgment has determined it by bringing the percep- 
 tion under the conception. We may, however, conceive 
 of an intelligence different in kind from ours, an intelli- 
 gence which is perceptive and not discursive, and which 
 therefore proceeds from a synthetic universal to the 
 particular, that is, from a perceived whole to the parts. 
 For such an intelligence, the connection of the parts 
 which form a determinate whole would not be, or appear, 
 
 contingent as it is for us But, from the peculiar 
 
 character of our intelligence, a real whole in nature is 
 regarded only as the effect of the combined motive forces 
 of the parts. We may, however, instead of viewing the 
 whole as dependent on the parts, after the manner of our 
 discursive intelligence, take a perceptive or archetypal 
 intelligence as our standard, and seek to comprehend the 
 dependence of the parts on the whole, both in their 
 specific nature and in their interconnection. And as it is 
 a contradiction in terms to say that for a discursive intelli- 
 gence the connection of the parts necessarily presupposes 
 the whole, it must be the idea of the whole that for such 
 an intelligence explains the form of the whole and the 
 connection of its parts. Now, such a whole is an effect 
 ox product, the idea of which is treated as the cause that 
 makes it possible, and such a product is called an end. 
 It therefore arises solely from the peculiar character of 
 our intelligence, that we regard certain natural products 
 as due to a different sort of causality from that of the 
 material laws of nature, namely, that of ends and final 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 34 1 
 
 421 causes. This principle, therefore, does not determine the 
 manner in which things themselves, even when they are 
 regarded as phenomena, are capable of being produced, 
 but merely the manner in which our intelligence can 
 alone judge them to be produced. And this is the 
 reason why in our scientific investigations we are 
 dissatisfied with any explanation of natural products by 
 final causes. In such investigations our sole object 
 is to judge of natural products, so far as we are 
 capable of doing so in consistency with the nature of 
 our judgment, that is, our reflective judgment, not to 
 determine them by judgment as things in themselves. 
 The correctness of the view here taken does not 
 require us to show that an ijitelledus archetypus may 
 possibly exist ; it is enough that the idea is not self- 
 contradictory, and that a perceptive or archetypal 
 intelligence is the natural counterpart of a discursive 
 intelHgence like ours {t?itellecius ectypus), which by 
 its very nature is contingent and dependent upon the 
 presentation of particulars. 
 
 If we think of a material whole as in its form a product 
 of the parts, with their forces and power of combining 
 themselves with one another, we get the conception of a 
 mechanical mode of production. But we do not in this 
 wav obtain any conception of a whole as end, such as we 
 are compelled to suppose an organized being to be, — a 
 whole, the inner possibility of which is utterly inconceiv- 
 able apart from the idea of it, and on which depends the 
 very nature and mode of operation of the parts. It does 
 not follow, as we have just seen, that the mechanical 
 production of such a body is impossible ; for to say so 
 would be to say, that no mtelltgence could possibly think 
 
342 THE PHILOSOPHY OF KANT. 
 
 the different parts as combined in a unity, unless the 
 idea of the unity was at the same time the cause of the 
 whole ; unless, in other words, the production was pur- 
 posive. For the unity which is the necessary ground of 
 the form of natural products would then be merely that 
 of space ; and space is not a real ground of products, but 
 simply their formal condition, although no doubt it has 
 this in common with the real ground, that no part of it 
 can be determined except in relation to the whole. Now, 
 4.22 it is at least possible to regard the material world as a 
 mere phenomenon, and to conceive of its substrate as a 
 thing in itself, to which an intellectual perception corre- 
 sponds. Thus we get the idea of a supersensible and 
 real ground of the world of nature to which we our- 
 selves belong, although that ground is not for us an 
 object of knowledge. Accordingly, we may apply 
 mechanical laws in explanation of that which in the 
 sensible world is necessary, but the harmony and unity 
 of the particular laws and forms of nature — which 
 relatively to the mechanism of nature must be regarded 
 as contingent — we must view as an object of reason to 
 which teleological laws are applicable. Nature thus 
 comes to be judged on two distinct principles, the 
 mechanical and the teleological, but these in no way 
 conflict with each other. 
 
 From this point of view we can see, what even in other 
 ways might readily be guessed, but in no other way could 
 be proved and maintained with certainty, that the prin- 
 ciple of a mechanical derivation of those natural products 
 which exhibit purpose is quite consistent with the 
 teleological principle, but by no means enables us to 
 dispense with it. In the investigation of a thing that we 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 343 
 
 are forced to regard as a natural end, that is, an organized 
 being, we may try all the known and yet to be discovered 
 laws of mechanical production, and may even hope to 
 make good progress in that direction, but we need never 
 hope to get rid, in our explanation of natural products, of 
 the quite different principle of causation by ends. No 
 human intelligence, and indeed no finite intelligence, 
 however it may surpass ours in degree, need expect to 
 comprehend the production of even a blade of grass by 
 purely mechanical causes. The teleological connection 
 of causes and effects is absolutely indispensable in judg- 
 ing of the possibility of such an object. There is indeed 
 no adequate reason for regarding external phenomena 
 as such from a teleological point of view ; the reason for 
 it must be sought in the supersensible substrate of 
 phenomena. But, as we are shut out from any possible 
 view of that substrate, it is impossible for us to find in 
 nature grounds for an explanation of nature, and we are 
 423 compelled by the constitution of our intellectual faculty 
 to seek for the supreme ground of teleological connec- 
 tions in an original Intelligence which is the cause of the 
 world. 
 
 APPENDIX ON METHOD. 
 
 461 87. The Moral Proof of the Existence of God. 
 
 Theoretical reflective judgment is quite justified in sup- 
 posing, on the ground oi 2i physical teleology^ that there is 
 an intelligent cause of the world. Now, in our own 
 moral consciousness, and still more in the general con- 
 ception of a rational being who is endowed with free 
 causality, there is implied a moral teleology. But as the 
 
344 THE PHILOSOPHY OF KANT. 
 
 relation to ends, together with the laws connected with 
 them, is determined a priori in ourselves, and therefore 
 is known to be necessary, this internal conformity to law 
 does not require for its explanation the supposition of an 
 intelligent cause outside of ourselves. At the same time 
 moral teleology has to do with man as a being in the 
 world, and therefore with man as connected with other 
 things in the world. For, in the conception of ourselves 
 as beings under moral law, we find the standard by refer- 
 ence to which those other things are judged either to be 
 ends, or to be objects subordinate to ourselves as the 
 ultimate end. Moral teleology, then, has to do with the 
 relation of our own causality to ends, and even to an 
 ultimate end necessarily set up by us as our goal in the 
 world, as well as with the possibility of realizing that end, 
 the external world being what it is. Hence the question 
 462 necessarily arises, whether reason compels us to seek, in 
 a supreme intelligence outside of the world, for a principle 
 which shall explain to us even the purpose in nature 
 relatively to the law of morality within us. There is 
 therefore a moral teleology, which is concerned, on the 
 one hand with the nomothetic of freedom, and on the 
 other hand with that of nature. 
 
 If we suppose certain things, or even certain forms of 
 things, to be contingent, and therefore to depend upon 
 something else which is their cause, we may seek for this 
 supreme cause, or unconditioned ground of the condi- 
 tioned, either in the physical or in the teleological order. 
 That is to say, we may either ask, what is the supreme 
 productive cause of those things, or what is their supreme 
 and absolutely unconditioned end, that is, the ultimate 
 end of that cause in its production of those things, or 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 345 
 
 even of all things. In the latter case it is plainly implied 
 that the cause in question is capable of setting an end 
 before itself, that is, is an intelligence, or at least must be 
 thought of as acting in accordance with the laws of an 
 intelligence. 
 
 From the teleological point of view, it is dc fundamental 
 proposition admitted by every one, that there can be no 
 ultitnafe e?id 2X all presupposed by reason a priori, unless 
 
 463 that end is man as under 77ioral laws. A world con- 
 sisting of mere lifeless things, or even containing living 
 but unintelligent beings, would have no meaning or 
 value, because there would be in it no intelligent being 
 to appreciate its value. Again, suppose that in the world 
 there are intelligent beings, whose reason enables them 
 to value existing things for the pleasure they bring, but 
 who have not themselves any power of imparting a value 
 to things by means of freedom ; then, there will indeed be 
 relative ends, but there will be no absolute or ultimate 
 end, for the existence in the world of such intelligent 
 beings can never have an end. Moral laws, however, 
 are of this peculiar character, that they prescribe for 
 reason something as an end apart from all conditions, 
 and therefore exactly as the conception of an ultimate 
 end requires. The existence of a reason which can be 
 for itself the supreme law in the relation of ends, in other 
 words the existence of rational beings under moral laws, 
 can alone be conceived as the ultimate end of the 
 existence of a world. On any other supposition its 
 existence does not imply a cause acting from any end, or 
 it implies ends but no ultimate end. 
 
 464 The moral law, as the formal condition in reason of 
 the use of our freedom, lays its commands on us 
 
346 THE PHILOSOPHY OF KANT. 
 
 entirely on its own authority, without appealing to any 
 material condition as an end ; but it nevertheless 
 determines for us, and indeed a priori^ an ultimate 
 end as the goal to which our efforts ought to be 
 directed ; and that end is the highest good possible in 
 the world through freedom. 
 
 The subjective condition which entitles man to set 
 before himself an ultimate end subordinate to the moral 
 law is happiness. Hence the highest physical good 
 possible in the world is happiness^ and this end we must 
 seek to advance, as far as in us lies, but always under the 
 objective condition of the harmony of man with the law 
 of morality as worthiness to be happy. 
 
 But it is impossible, in consistency with all the faculties 
 of our intelligence, to regard the two requisites of the 
 ultimate end presented to us through the moral law as 
 connected by merely natural causes, and yet as conforming 
 to the idea of that ultimate end. If, therefore, nature is 
 the only cause which is connected with freedom as a 
 means, the conception of the practical necessity of the 
 ultimate end through the application of our powers, does 
 not harmonize with the theoretical conception of the 
 physical possibility of the realization of that end. 
 
 Accordingly, we must suppose a moral cause or authoi 
 of the world, in order to set before ourselves an ultimate 
 end that is conformable with the moral law ; and in so 
 far as the latter is necessary, in the same degree, and on 
 the same ground, the former also must necessarily be 
 admitted ; it must, in other words, be admitted that there 
 is a God. 
 
THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 347 
 
 469 88. Limitation of the Moral Proof. 
 
 The ultimate end, as merely a conception of our prac- 
 tical reason, is not an inference from data of experience 
 for the theoretical explanation of nature, nor can it be 
 applied in the knowledge of nature. Its only possible 
 use is for practical reason in relation to moral laws ; and 
 the ultimate end of creation is that constitution of the 
 world which harmonizes with the only end which we can 
 definitely present to ourselves as conforming to law, 
 namely, the ultimate end of our pure practical reason, in 
 so far as it is the nature of reason to be practical. Now, 
 we have in the moral law, which enjoins on us practically 
 the application of our powers to the realization of the 
 ultimate end, a ground for supposing the possibiHty and 
 practicability of that end, and therefore also a ground for 
 supposing a nature of things harmonious with it. Hence 
 we have a moral ground for representing in the world an 
 ultimate end of creation. 
 
 So far we have not advanced from moral teleology to 
 theology, that is, to the existence of a moral author of the 
 world, but have merely concluded to an ultimate end of 
 creation determined in that way. But, in order to 
 account for this creation, that is, for the existence of 
 things that are adapted to an ultimate end, in the first 
 place an intelligent being, and in the second place not 
 only an intelligent but a moral being or author of the 
 world, that is, a God, must be admitted to exist. But 
 this conclusion is of the peculiar character, that it holds 
 good merely for the judgment according to conceptions 
 of practical reason, and as such for the reflective not the 
 determinant judgment. It is true that in us morally 
 
34^ THE PHILOSOPHY OF KANT. 
 
 practical reason is essentially different in its principles 
 from technically practical reason. But we cannot assume 
 that in the Supreme Cause of the world, conceived of as 
 an intelligence, the same contrast exists, and that a 
 peculiar kind of causality is required for the ultimate 
 end, different in its character from that which is required 
 merely for ends of nature. We cannot assume, there- 
 fore, that in an ultimate end we have a reason for 
 admitting not merely a fuoral groimd or ultimate end of 
 creation as an effect, but also a juoral being as the original 
 470 ground of creation. But we may certainly say, that, 
 according to the constitution of our reason^ we cannot 
 make intelligible to ourselves the possibility of an adapta- 
 tion relative to the moral law, and to its object as it is in 
 this ultimate end, apart from an author and ruler of the 
 woi*fd, who is also a moral lawgiver. 
 
 Physical teleology sufficiently proves for theoretical 
 reflective judgment an intelligent cause of the world ; 
 moral teleology proves it for the practical judgment, 
 through the conception of an ultimate end, which must 
 be attributed to creation when we view it in relation to 
 action. It is true that the objective reaHty of the idea of 
 God, as the moral author of the world, cannot be shown 
 from a consideration of physical ends alone. But, it is a 
 maxim of pure reason to secure unity of principles, so far 
 as that is possible ; and hence the knowledge of physical 
 ends, when it is brought into relation with the knowledge 
 of the moral end, greatly aids us in connecting the 
 practical reality of the idea of God with its theoretical 
 reality as already existing for judgment. 
 
 To prevent a very natural misunderstanding two re- 
 marks must be made, which should be carefully borne in 
 
I 
 
 THE CRITIQUE OF TELEOLOGICAL JUDGMENT. 349 
 
 mind. In the first place, we can think the attributes of 
 the Supreme Being only by analogy. How, indeed, 
 could we investigate directly the nature of a Being to 
 whom nothing similar is given in experience ? Secondly, 
 the attributes by which we think the Supreme Being do 
 not enable us to know Him as He is, nor can we theor- 
 etically predicate them of Him. To contemplate that 
 Being as he is in Himse// speculsitive reason must assume 
 the form of the determinant judgment, and this is con- 
 trary to its very nature. 
 
INDEX. 
 
 JEsthetic, transcendental, 22; distin- 
 guished from transcendental logic, 
 23 ; deals only with space and time, 
 36 ; implies contrast of phenomenal 
 reality and reality of thing in itself, 
 36; summary, 39. 
 
 Affirmation and negation, 196, 
 
 Analogies of Experience, loi ; first 
 analogy, 106 ; second, no ; third, 
 118; general remark, 121. 
 
 Analysis perhaps main work of reason, 
 12 ; presupposes synthesis, 49, 64. 
 
 Analytic judgments contrasted with 
 synthetic, 13 ; also called explica- 
 tive, 13. 
 
 Anal3^ic, transcendental, its object, 
 43 ; analytic of practical reason, 261. 
 
 Animal, the, not free, 186. 
 
 Antinomy of pure reason, 155 ; first 
 antinomy, 158 ; second, 160 ; third, 
 162 ; fourth, 165 ; necessity of their 
 solution, 166 ; critical solution, 168 ; 
 solution of first antinomy, 175 ; so- 
 lution of second antinomy, 177 ; 
 transition from the mathematical to 
 the dynamical, 179 ; solution of 
 third antinomy, 182 ; solution of 
 fourth antinomy, 191 ; contrast of 
 antinomy and ideal, 193 ; antinomy 
 of practical reason, 292 ; antinomy 
 of judgment, 331. 
 
 Apperception, transcendental unity of, 
 61, 65 ; in itself analytic but the 
 condition of synthesis, 67 ; the su- 
 preme principle of understanding, 
 67, 68 ; not the principle of a per- 
 ceptive understanding, 6^], 69 ; con- 
 dition of objective unity, 70. 
 
 A priori and a posteriori, 8, 22. 
 
 Archetypal and ectypal intelligence, 
 
 341- 
 
 Association of ideas not an explana- 
 tion of causahty, 10; presupposes 
 synthesis of reproduction, 58. 
 
 Autonomy of will, 248, 270. 
 
 Axioms of perception, 92. 
 
 Beautiful and sublime the object of 
 aesthetic judgment, 318. 
 
 Categories, guiding-thread to, 46; 
 imply synthesis, 49 ; correspond to 
 logical functions, 51 ; table, 51 ; 
 distinguished as mathematical and 
 dynamical, 52 ; third of each class 
 the union of the other two, 52 ; their 
 deduction, 53 f.; principle of the 
 deduction, 53 ; an empirical deduc- 
 tion irrelevant, 55 ; subjective de- 
 duction, 56 f ; objective deduction, 
 63 f.; possibihty of combination, 
 63; original synthetic unity of ap- 
 perception, 64 ; objective unity, 70; 
 perceptions stand imder them, 72; 
 apply only to experience, 73 ; their 
 application to objects of sense, 76, 
 78 ; result of their deduction, 82 ; 
 short statement of deduction, 82 ; 
 peculiarity of modality, 122 ; re- 
 stricted to external perception, 126 ; 
 their real definition implies sensibi- 
 lity, 130 ; cannot be used transcen- 
 dentally, 130. 
 
 Causality a pure a priori judgment, 
 10 ; Hume's view, 10 ; a synthetic 
 judgment, 14 ; its category, 52 ; its 
 schema, 89 ; proof of the principle, 
 no; not a generahsation from ex- 
 perience, 115 ; implies perception of 
 motion, 127 ; antinomy of natural 
 and free causation, 162 ; solution of 
 the antinomy, 182 ; first cause and 
 supreme cause, 203 ; final cause in 
 physico-theological argument, 219 ; 
 will the causality of a rational being, 
 244 ; final cause as the principle of 
 judgment, 313 ; formal purpose, 
 313 ; esthetic idea of purpose, 317 ; 
 logical idea, 318 ; formal objective 
 purpose, 323; relative and internal 
 purpose, 324 ; real things as natural 
 ends, 325 ; nature as a system of 
 ends, 329, 
 
352 
 
 INDEX. 
 
 Change conceivable only through time, 
 S^j 35 ; wrongly supposed to dis- 
 prove ideality of time, 34 ; an em- 
 pirical conception, 36 ; its contin- 
 uity, 98 ; implies permanence, 108 ; 
 also implies causality, no. 
 Community, category of, 52 ; its 
 schema, 89; proof of the principle, 
 118; either dynamical or local, 119; 
 phenomena stand in community of 
 apperception, 120 ; implies percep- 
 tion of things in space, 128. 
 Conception an element in all know- 
 ledge, 40 ; either pure or empirical, 
 40; the form in which an object is 
 thought, 40 ; a function of under- 
 standing, 47 ; when problematic, 
 132 ; see Categories and Ideas. 
 Constitutive principles, 103; impos- 
 sible for reason, 174. 
 Continuity of magnitudes, 97. 
 Copernicus the type of the critical 
 
 philosopher, 3. 
 Cosmology, rational, defined, 142 ; 
 system of ideas, 156 ; see Antinomy 
 and Ideas. 
 Critical philosophy demands scrutiny 
 of all beliefs, i ; denies knowledge 
 of supersensible, affirms faith, 5 ; 
 not opposed to science, 7 ; its object 
 pure reason, not philosophical sys- 
 tems, 7 ; determines the possibihty, 
 principles and hmits of knowledge, 
 11; leads to science, 19; its idea 
 and division, 19; draws a plan of 
 transcendental philosophy, 20 ; its 
 method of proof, 103. 
 
 Deduction of categories, 53 f. ; of prin- 
 ciples of practical reason, 272 ; see 
 Categories and Principles. 
 
 Degree the schema of quality, 88. 
 
 Demonstration possible only in mathe- 
 matics, 103. 
 
 Design, argument from, 218. 
 
 Desires, natural, not an object of reve- 
 rence, 229 ; contrary to duty, 231 ; 
 supply no moral motive, 284. 
 
 Dialectic, transcendental, exposes the 
 illusions of reason, 44, 137 ; dia- 
 lectic of practical reason, 289 ; dia- 
 lectic of teleological judgment, 331. 
 
 Dogmata and mathemata, 104. 
 
 Dogmatism a phase of philosophy, 
 i; an enemy of morality, 6; its 
 definition, 7 ; leads to scepticism, 
 19. 
 
 Duty a Dure idea, 186 ; analysis of the 
 idea, 227f. ; not derived from ex- 
 perience, 233 ; implies a will not 
 perfectly good, 235, 269 ; expressed 
 in a categorical imperative, 238 ; 
 opposed to self-interest, 284. 
 
 Dynamical categories, 52 ; principles 
 of judgment, 103 ; principle of 
 causality, 183 ; contrast of dyna- 
 mical and mathematical regress, 192. 
 
 Empirical reality of space, 29 ; of time, 
 34 ; deduction of categories, 54 ; 
 apperception, 66, 70; character of 
 man, 188. 
 
 Existence, category of, 52 ; its schema, 
 89. _ 
 
 Experience contains two elements, 7 ; 
 does not yield universal judgments, 
 9, 14; supplies synthetic judgments, 
 14 ; limits the application of the 
 categories, 73. 
 
 Exposition, metaphysical, ofspace,23; 
 of time, 29 ; transcendental, of 
 space, 26 ; of time, 30 ; of prin- 
 ciples of practical reason, 261. 
 
 Faith, objects of, 6 ; a need of pure 
 reason, 302. 
 
 Final cause, see Teleology. 
 
 Form of perception an element in 
 knov^^ledge, 22 ; space a form of 
 outer sense, 27 ; time a form of 
 inner sense, 31 ; time the form of 
 all phenomena, 32 ; time and space 
 the only pure forms of sense, 35; 
 time not a form of God's perception, 
 37 ; form of thought the object of 
 general logic, 41 ; form of judgment 
 indicates objective unity, 71. 
 
 Freedom an object of faith, 6 ; a pro- 
 blem of pure reason, 11 ; its rela- 
 tion to natural causahty, 182; the 
 key to autonomy of will, 250 ; its 
 negative definition, 250 ; its positive 
 dehuition, 251 ; peculiar to rational 
 beings, 252 ; its relation to morahty, 
 253 ; not incompatible with natural 
 mechanism, 254, 273, 286; com- 
 patible with the causality of God, 
 287 ; its relation to idea of final 
 cause, 310. 
 
 Galileo a discoverer of scientific 
 
 method, 2. 
 Geometry, its judgments synthetic, 16, 
 
 26 ; involves productive imagina- 
 
INDEX. 
 
 353 
 
 tion, 93 ; its figures display formal 
 objective adaptation, 323. 
 
 God an object of faith, 6 ; His exist- 
 ence a problem of reason, 11 ; not 
 conditioned by space and time, 37 ; 
 conceived as ens realissimutn, 197 ; 
 the object of transcendental theo- 
 logy, 199 ; speculative proofs of His 
 existence, 202 ; ontological proof, 
 204 ; cosmological, 210 ; physico- 
 theological, 218 ; not under impera- 
 tives of duty, 236, 269 ; His caus- 
 ality compatible with human free- 
 dom, 287; moral proof of His 
 existence, 296. 
 
 Good and pleasure, 236, 280; the 
 summu7n bo?ium, 291 ; supreme and 
 complete good, 291 ; see Morality 
 and Will. 
 
 Happiness the complete satisfaction of 
 natural desire, 231 ; an assertoric 
 hypothetical imperative, 237 ; inde- 
 finable, 239 ; comprehends all mate- 
 rial principles, 263 ; impUed in the 
 summum bonum, 291, 
 
 Heteronomy of will, 271. 
 
 Hume, his view of causality, 10. 
 
 Ideas, transcendental, 140 ; rest upon 
 the unconditioned, 141 ; their sys- 
 tem, 142 ; idea of the soul, 142 ; 
 idea of the world, 155 ; idea of God, 
 196. 
 
 Ideal, the, i95f. 
 
 Ideality, transcendental, of space, 29 ; 
 ot time, 33. 
 
 Illusion, material and transcendental, 
 135 ; logical and transcendental, 
 136 ; dialectical, in rational psy- 
 chology, 145 ; in rational cosmo- 
 logy, 168 ; in rational theology, 
 
 215. 
 Image and schema, 86; space the 
 
 image of external magnitudes, time 
 
 the image of all magnitudes, 88. 
 
 Imagination synthetic, 49, 51, 77; 
 productive or reproductive, 78 ; its 
 product the schema, 86, 87. 
 
 Immortality an object of faith, 6 ; a 
 problem of reason, 11 ; its proof, 
 294. 
 
 Imperatives as rules of conduct, 186 ; 
 either categorical or hypothetical, 
 236 : only the categorical imperative 
 expresses law, 240 ; three ways of 
 stating it, 241 ; its proof, 254. 
 
 Induction does not prove a judgment 
 universal, 9 ; not the foundation of 
 causality, 115, 
 
 Indifferentism a phase of philosophy, 
 I. 
 
 Intellectual perception peculiar to 
 original being, 38 ; apperception 
 not its principle, 67, 69; indefinable, 
 755 133 ; implies archetypal inteUi- 
 gence, 340. 
 
 Judgment, empirical, not universal, 9 ; 
 pure a priori, found in common 
 sense and science, 10 ; analytic or 
 synthetic, 13 ; its definition, 47 ; 
 transcendental, 83 ; its schematism, 
 84 ; its principles, 92 ; its type, 282 ; 
 connects theoretical and practical 
 philosophy, 311 ; determinant or re- 
 flective, 311; its antinomy, 331; 
 teleological, 336. 
 
 Knowledge begins with experience, 7 ; 
 a priori or a posteriori. 8 ; pure or 
 mixed, 9 ; criteria of a priori, uni- 
 versality and necessity, 9; tran- 
 scendental, 19. 
 
 Logic and aesthetic, 23, 41 ; its divi- 
 sions, 43. 
 
 Magnitude, external, space its image, 
 88 ; time the image of all magnitude, 
 88 ; perceptions extensive magni- 
 tudes, 92 ; intensive magnitude, 96; 
 continuity, 97 ; relation of extensive 
 to intensive magnitude, 100. 
 
 Man, his intelligible and empirical 
 character, 186. 
 
 Materialism unprovable, 153. 
 
 Mathemata and dogmata, 104. 
 
 Mathematical categories, 52 ; prin- 
 ciples of judgment, 102; antinomies, 
 192. 
 
 Mathematics, its method, 2, 104 ; its 
 judgments synthetic, 15 ; how pos- 
 sible, 18 ; applicable to all pheno- 
 mena, 94 ; its figures display formal 
 objective purpose, 323. 
 
 Matter, its definition, 17 ; permanence 
 of its quantity a synthetic judgment, 
 17 ; matter of sense an element in 
 knowledge, 22; matter of desire, 
 262, 
 
 Mechanism of nature compatible with 
 freedom, 6, 254, 273, 286 ; and with 
 teleology, 340. 
 
354 
 
 INDEX. 
 
 Metaphysic, its past failure due to 
 false method, i ; true method, 3 ; 
 gives no knowledge of supersensible, 
 4; seeks for a priori synthetic know- 
 ledge, 17 ; exists as a natural dis- 
 position, 18 ; how possible as a 
 science, 19; metaphysic of morality, 
 
 225f. 
 
 Misology, 226. 
 
 Modality of judgments, 48 ; its cate- 
 gories, 52 ; its peculiarity, 122. 
 
 Momentum implies degree, 97. 
 
 Morality compatible with the mechan- 
 ism of nature, 5 ; its principles not 
 included in transcendental philo- 
 sophy, 20 ; metaphysic of morality, 
 225f; popular and philosophical con- 
 ception of morahty, 226 ; rests upon 
 idea of duty, 227 ; possible only for 
 a rational being, 229 ; distinguished 
 from prudence, 238 ; a law for all 
 rational beings, 243 ; necessity of 
 metaphysic of morality, 243 ; moral 
 conceptions a priori^ 246 ; implies 
 a kingdom of ends, 248 ; its laws 
 determine action by their mere form, 
 249. 
 
 Motion made conceivable by time, 31 ; 
 an empirical conception, 36, 
 
 Motives, 284f.; only moral motive con- 
 formity to law, 284. 
 
 Natura materialiter spectata and for- 
 maliter spectata, 80. 
 
 Natural theology assumes subjectivity 
 of space and time, 37. 
 
 Necessary being as condition of phe- 
 nomena, 191. 
 
 Necessity a criterion of a priori know- 
 ledge, 9 ; its category, 52 ; its 
 schema, 89 ; a postulate of empirical 
 thought, 125, 
 
 Negation, category of, 51 ; involves 
 limitation, 197, 
 
 Non-existence, category of, 52 ; its 
 schema, 89. 
 
 Noumena and phenomena, 129 ; the 
 idea of a limit, 132 ; not a division 
 of things, 133. 
 
 Object, meaning of, 60. 
 
 Ontology not a science, 131 ; onto- 
 
 logical argument, 204, 
 Organism implies internal purpose, 
 
 327- 
 Paralogism of rational psychology, 145. 
 
 Perception pure or empirical, 22 ; an 
 element m all knowledge, 40 ; an 
 extensive magnitude, 92 ; see hiiel- 
 lectual Perception. 
 
 Perceptive understanding, see Intel- 
 lectiial Perception. 
 
 Phenomenon defined, 22; implies time, 
 33 ; its reality, 36 ; determinable by 
 mathematics, 94 ; has degree, 96 ; 
 a continuous magnitude, 98 ; rela- 
 tion to noumenon, 129 ; man as 
 phenomenon, 186 ; distinction of 
 phenomena and noumena reconciles 
 human freedom with God's caus- 
 ality, 287. 
 
 Philosophy, its division into theoretical 
 and practical, 307 ; see Critical 
 Philosophy. 
 
 Physico-theological argument, 218. 
 
 Physics, its method, 2 ; contains a 
 priori synthetic judgments, 17 ; 
 how is pure physics possible, 18. 
 
 Plato, his theory of ideas, 12. 
 
 Pleasure relative to desire, 262 ; does 
 not admit of qualitative differences, 
 264 ; distinct from good, 281. 
 
 Plurality, category of, 51. 
 
 Possibility, category of, 52 ; its schema, 
 89 ; a postulate of empirical thought, 
 123 ; has no meaning for a percep- 
 tive intelligence, 334. 
 
 Postulates of empirical thought, 92, 
 122 ; possibility, 123 ; actuality, 124; 
 necessity, 125 ; postulates of pure 
 practical reason, 298. 
 
 Principles of judgment, 92 ; mathe- 
 matical constitutive, dynamical re- 
 gulative, 102 ; immanent or trans- 
 cendent, transcendent or transcen- 
 dental, 135, 139 ; principles of rea- 
 son, 137 ; regulative principles of 
 reason, 173 ; empirical use of regu- 
 lative principles, 174; practical prin- 
 ciples are maxims or laws, 261 ; 
 material principles reducible to self- 
 love, 263 ; they rest on the lower 
 desires, 263. 
 
 Prudence, its precepts hypothetical, 
 238. 
 
 Psychology distinct from logic, 41 ; 
 rational psychology, 142 ; its para- 
 logisms, 145 ; not a doctrine but a 
 discipline, 153. 
 
 Purpose in nature, see Teleology. 
 
 Quality of judgments, 48; its cate- 
 gories, 52 ; its schemata, 88. 
 
INDEX. 
 
 355 
 
 Quantity of judgments, 48 ; its cate- 
 gories, 51 ; its schema, 88 ; no 
 axioms of quantity [quaniitas), 93 ; 
 quantum of substance unchange- 
 able, no; quanta imply external 
 perception, 128. 
 
 Reality, empirical, of space, 29; of 
 time, 34 ; category of reality, 52 ; its 
 schema, 88; totality of reality, 197. 
 
 Reason organic, 7 ; its problem, 18 ; 
 the faculty of principles, 137 ; its 
 principles derived from the form of 
 inference, 138 ; its dialectical con- 
 clusions, 143 ; its paralogisms, 145 ; 
 its antinomies, 155 ; critical solu- 
 tion of its antinomies, 168 ; specu- 
 lative and practical, 256 ; limits of 
 practical, 257 ; critique of practical 
 reason, 261?. ; the source of moral 
 law, 269 ; object of practical reason 
 the summum bofium, 280 ; its mo- 
 tives, 284 ; extension of pure reason, 
 278, 300. 
 
 Reciprocity, category of, 52 ; its 
 schema, 89 ; proof of the principle, 
 118. 
 
 Recognition, synthesis of, 56, 
 
 Regulative principles of judgment, 
 103 ; of reason, 173 ; in regard to 
 necessary being, 193 ; in regard to 
 idea of God, 216. 
 
 Relation of judgments, 48 ; its cate- 
 gories, 52 ; its schemata, 89. 
 
 Reproduction, synthesis of, 56, 58. 
 
 Reverence for moral law, 229. 
 
 Scepticism a phase of philosophy, i ; 
 the result of dogmatism, 19. 
 
 Schematism of categories, 84 ; tran- 
 scendental schema, 85 ; schema and 
 image, 86 ; schemata underlie ma- 
 thematics, 86 ; the various schemata, 
 88; schemata restrict the categories, 
 91 ; schema and type, 283. 
 
 Self-consciousness, see Apperception 
 and Paralogism, 
 
 Self-love, 263. 
 
 Sensibility and understanding, 20; de- 
 finition of sensation, 22 ; external 
 and internal sense, 23 ; sensibility 
 receptive, 40 ; all sensations have 
 degree, 97 ; inner sense a flux, 127 ; 
 quantum of inner sense implies ex- 
 ternal perception, 128. 
 
 Soul the object of rational psychology, 
 ^45' 
 
 Space an object of transcendental 
 aesthetic, 23 ; metaphysical exposi- 
 tion, 23 ; transcendental exposition, 
 26 ; not a thing or property, 27 ; a 
 form ot outer sense, 27, 32 ; its 
 empirical reality and transcendental 
 ideality, 29 ; the source of a priori 
 synthetic judgments, 35 ; not a form 
 of divine perception, 37 ; the image 
 of external magnitudes, 88 ; a quan- 
 tum continuum, 97 ; empty space 
 unknowable, 99 ; its transcendental 
 ideality confirmed by critical solu- 
 tion of antinomies, 172. 
 
 Subject, thinking, not substance, 149. 
 
 Sublime and beautiful the object of 
 aesthetic judgment, 318. 
 
 Substance, category of, 52 ; its schema, 
 89 ; proof of the principle, 106 ; all 
 substances in dynamical community, 
 118; implies perception of matter, 
 127 ; not applicable to soul, 147 ; 
 contrasted with thinking subject, 
 149. 
 
 Summum bonum includes morality and" 
 happiness, 291. 
 
 Synopsis implies synthesis, 56. 
 
 Synthesis, definition of, 49 ; an opera- 
 tion of imagination, 50; pure syn- 
 thesis the category, 50; synthesis 
 of apprehension, 57, 79 ; of repro- 
 duction, 58 ; of recognition, 60 ; the 
 condition of knowledge, 63 ; figural 
 and intellectual, 'jj. 
 
 Synthetic judgments distinguished 
 from analytic, 13 ; also called am- 
 pliative, 13 ; some rest on experi- 
 ence, others do not, 14; mathema- 
 tics contains them, 15 ; examples 
 from arithmetic and geometry, 16 ; 
 physics contains them, 17; meta- 
 physic seeks for them, 17 ; how pos- 
 sible, 18 ; time and space their 
 source, 35 ; principles of reason 
 synthetic, 139. 
 
 Taste, aesthetic, 318. 
 
 Teleology in physico-teleologica argu- 
 ment, 219 ; in nature, 313 ; a maxim 
 of judgment, 316 ; implied in or- 
 ganized beings, 327 ; as a system, 
 329 ; its relation to the mechanism 
 of nature, 331. 
 
 Thales perhaps the discoverer of ma- 
 thematical method, 2. 
 
 Theology, natural, its conception of 
 God, 37 ; rational, 142 ; criticism of 
 
35^ 
 
 INDEX. 
 
 speculative theolofjy, 222 ; moral 
 theolo^'y, 345. 
 
 Thing in itself, 27, 31, see Noumena. 
 
 Time an object of transcendental aes- 
 thetic, 23 ; metaphysical exposition, 
 46; transcendental exposition, 29; 
 not a thing or determination of a 
 ^'^'"g. 31 > ^^^ form of inner sense, 
 32 ; represented as a line, 32 ; a form 
 of all phenomena, 32; its empirical 
 reality and transcendental ideality, 
 33; objection to its transcendental 
 ideality answered, 34 ; a source of 
 a priori synthetic judgments, 35 ; 
 not a form of divine perception, 37 ; 
 homogeneous both with category 
 and with object of sense, 85 ; the 
 image of all magnitudes, 88 ; im- 
 plied in all schemata, 90 ; a quan- 
 tum rontinuu7n, 97 ; empty time 
 unknowable, 99 ; its modi arc per- 
 man(;nce. succession and co-exist- 
 ence, loi ; its transcendental ideal- 
 ity confirmed by critical solution of 
 aiitiuDmies, 172. 
 
 Torricelli a discoverer of scientific 
 method, 2. 
 
 Totality, category of, 51. 
 
 Transcendental knowledge, 19 ; criti- 
 
 cism, 20; philosophy, 20; aesthetic, 
 23 f.; logic, 23, 41, 42; analytic, 43 
 f.; deduction of categories, 53 f.; 
 doctrine of judgment, 83 ; determi- 
 nation of time, 85 ; illusion, 135 ; 
 dialectic, 137 f, ; principle of pure 
 reason, 139; idccos, 140. 
 Type of pure practical judgment, 282. 
 
 Unconditioned, not a self-contradic- 
 tory idea, 5 ; the object of reason, 
 
 139- 
 
 Understanding one of the two stems 
 of knowledge, 20 ; a spontaneous 
 faculty, 40; its logical use, 46; 
 its faculty is judgment, 47; its logi- 
 cal functions, 48 ; makes objects 
 possible, 115; contrasted with rea- 
 son, 138. 
 
 Unity, category of, 51 ; implied in all 
 categories, 64. 
 
 Will, a good, the only absolute good, 
 225 ; a good will the highest good, 
 229 ; possible only for a rational 
 being, 229 ; divine will not subject 
 to imperatives, 236 ; autonomy the 
 basis of morality, 270 ; heteronomy, 
 271 ; will as desire, 308. 
 
 END, 
 
 Vt» 
 
 OI-ASOOW ; I'llINTKn at the university PBKflS BV ROBKBT MACLEH082 AND CO. 
 
RETURN TO the circulation desk ot any 
 University of California Library 
 or to the 
 NORTHERN REGIONAL LIBRARY FACILITY 
 Bldg, 400, Richmond Field Station 
 University of California 
 Richmond, CA 94804-4698 
 
 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 
 
 • 2-month loans may be renewed by calling 
 (510)642-6753 
 
 • 1-year loans may be recharged by bringing 
 books to NRLF 
 
 • Renewals and recharges may be made 4 
 days prior to due date, 
 
 DUE AS STAMPED BELOW 
 
 '/\PR 2 2 2003 
 
 12,000(11/95) 
 
 iierkeley 
 
GT 
 
 1)^ 
 
 iT'T' ^JT^ 
 
 H