GJontell JtnuiEraUy 
 
 Htljaca, Men) Jfnrlt 
 
 LIBRARY OF 
 
 THE SAGE SCHOOL 
 
 PHILOSOPHY 
 
Date Due 
 
 MftRg" 
 
 ^2 
 
 ^ 
 
 -Nt 
 
 -PES- 
 
 DW— 5" 
 
 warn 
 
 ^ 
 
 i^E 
 
 rr^ 
 
 fi^UggW" 
 
 4S&3 ,J Y 
 
 9 80 Klf ' 
 
 5 raSeM 
 
 • .'-'i 
 
 »PH^963 3 
 
 ^jsosiea 
 
 tt 
 
 OECT^" 
 
 ^a^nrr 
 
 *=F 
 
 +4-( ; 
 
 3 19^4 U»£^1 2™ 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31 924092291 1 1 5 
 
KANT'S CRITICAL PHILOSOPHY 
 
KANT'S 
 CRITICAL PHILOSOPHY 
 
 FOR ENGLISH READERS 
 
 JOHN P. MAHAFFY, D.D. 
 
 FELLOW AND TUTOR OF TRINITY COLLEGE, DUBLIN 
 PROFESSOR OF ANCIENT HISTORY IN THE UNIVERSITY OF DUBLIN 
 
 JOHN H. BERNARD, B.D. 
 
 FELLOW OF TRINITY COLLEGE, DUBLIN 
 ARCHBISHOP KING'S LECTURER IN DIVINITY IN THE UNIVERSITY OF DUBLIN 
 
 A NEW AND COMPLETED EDITION 
 
 VOL. I. 
 
 THE KRITIK OF THE PURE REASON EXPLAINED 
 AND DEFENDED 
 
 iLon&Ott 
 
 MACMILLAN AND CO. 
 
 AND NEW YORK 
 I889 
 
PREFACE TO THE NEW EDITION 
 
 The able co-operation of one of my colleagues, to- 
 gether with class-room criticisms of others, has en- 
 abled this work to appear in a new edition, not only 
 purged of many defects, but completed by the addition 
 of several chapters on the Dialectic and Methodology 
 of the Pure Reason. It is also offered to the public 
 in a cheaper form, one volume containing the whole 
 Kritik, while the second gives a revised translation of 
 the Prolegomena. Without the help of Mr. Bernard, 
 who has written the whole commentary from p. 237 
 onward, and who corrected and improved all the rest, 
 it is not likely that this, my third declaration upon 
 Kant, would ever have seen the light ; and indeed I 
 will here add, that I am not responsible for certain 
 Hegelianisms which appear in his notes on the 
 Dialectic. But haec hactcnns. It is my desire to 
 state candidly and unreservedly what my younger 
 colleague has done to help me. For to make such 
 
PREFACE 
 
 acknowledgment is a duty not always recognised 
 among literary men of established standing. 
 
 The plan of this commentary is to adhere much 
 more faithfully than Kuno Fischer, or indeed any 
 other Commentator we have met, to Kant's text, 
 which we have followed paragraph by paragraph, 
 shortening and simplifying, but shirking no diffi- 
 culties. We have also marked all our own reflec- 
 tions with a 1, which refers only to the paragraph 
 or chapter to which it is prefixed. It seems more 
 advisable to do this than to encumber the book with 
 foot-notes, which interrupt the reader's train of 
 reading. 
 
 But though this Commentary claims to be clear 
 and consistent, it is not, and cannot be, either easy 
 or short. It is not easy, because the subject is not 
 easy, and deals with notions exceedingly abstract, 
 and only familiar to those who have made mental 
 science a subject of special study. It is not short, 
 because Kant's book, if worth reading at all, is worth 
 reading and knowing accurately, and no pains are 
 misplaced if they result in a full and comprehensive 
 grasp of the greatest metaphysical system the world 
 has yet seen. 
 
 We are too much accustomed to general histories 
 in which a few pages are devoted to each thinker, so 
 
PREFACE 
 
 that the impressions left on the reader's mind, even 
 if true (which is seldom the case) are at all events 
 vague and misty. The greater lights in the philo- 
 sophical firmament can only be understood by special 
 study, and should therefore be made the subject of 
 separate monographs. We have endeavoured to do 
 this for Kant, being convinced that of all meta- 
 physicians he is certainly the greatest, and perhaps 
 the most imperfectly understood. 
 
 In the preface to the former edition I spoke of 
 the fancy for philosophical novelties in England, and 
 endeavoured to call attention to Kant, as of all 
 modern Germans the greatest philosopher, and 
 certainly the best adapted for practical minds. At 
 all events it is absurd to begin the study of Schelling 
 or Hegel without a prior intimacy with Kant, and 
 how many men are there now in England who 
 thoroughly understand the Critical Philosophy ? It 
 is also a remarkable fact that within the last few 
 years, philosophy even in Germany has reverted, as 
 I ventured to predict, from modern extravagance to 
 the soberness of Kant. His works are being reprinted, 
 illustrated, and attacked, on all sides. The sensual 
 school have discovered that the refutation of Kant 
 alone will give them a lawful victory, and to this task 
 they are applying all their energies. 
 
PREFACE 
 
 The influence of Grote and J. S. Mill, and the con- 
 stant appointment of Mr. Bain as a State Examiner in 
 Philosophy, brought this way of thinking into undue 
 prominence. All the youth of the country have been 
 crammed with Mr. Bain's handbooks, and have neither 
 time nor inducement to read an antidote. We must 
 therefore look to the Universities for a fair hearing, 
 and trust that there at least enlightened teachers will 
 not accept as true what the State has made fashion- 
 able. A polemical chapter on the Association 
 School, which appeared in the former edition of 
 this work, has been omitted, as the controversy 
 may now be regarded as obsolete. The most 
 important contribution to the better understanding 
 of Kant, in my edition of Kuno Fischer's Comment- 
 ary, was the true explanation of Kant's refutation 
 of idealism. A patent absurdity had been universally 
 attributed to him, and I showed that his attitude 
 had been totally misconceived. My argument was 
 candidly accepted by competent critics, and mio-ht 
 now be called a commonplace, did not Prof. Kuno 
 Fischer's recent declaration (a Critique of Kant, 1882) 
 show that it has not yet penetrated into his mind. 
 
 As the work is intended for English readers, we 
 have referred uniformly to the translation of Kant's 
 Kritik in Bonn's Library, still the most serviceable 
 
PREFACE 
 
 and to my own edition of Kuno Fischer's Comment- 
 ary. I may add that my other references are to the 
 following editions of the respective books, viz. the 
 ninth edition of Mr. Mill's Logic, and the third of his 
 Examination of Hamilton, and the fourth edition of 
 Mr. Lewes' History of Philosophy. The references to 
 the Prolegomena of Kant are to the pages of my 
 former edition, which will be indicated in the new 
 edition appended as a second volume to the present. 
 I conclude this Preface as I did the last, seventeen 
 years ago, with the earnest hope that the many 
 readers of my former books on Kant will find in this 
 a maturer and clearer exposition of the same views. 
 
 J. P. MAHAFFY. 
 
 Trinity College, Dublin, 
 March \6th, 1889. 
 
CONTEXTS 
 
 CHAPTER I 
 
 THE TWO PREFACES TO THE KRITIK 
 
 PAGE 
 
 I."! On Kant's Style . . I 
 
 2. Analysis of the First Preface . i 
 
 3.5 Contrast of the two Prefaces . 6 
 
 4. The Second Preface . .8 
 
 $.% Consistency of the Two Editions of the Kritik . 19 
 
 CHAPTER II 
 
 THE INTRODUCTION 
 
 Preliminary Considerations — The Distinction in Kind between 
 
 the Faculties . ... 24 
 
 1. Pure and Empirical Cognition . . 27 
 
 2. We possess a priori Cognitions, even in Ordinary Ex- 
 
 perience . . .28 
 
 3. We require a Science to determine these Systematically 3 1 
 
 4. Analytical and Synthetical Judgments . . . .32 
 
 5. All Theoretical Sciences contain Synthetical Judgments a 
 
 priori . . • • • 37 
 
 6. The General Problem of the Pure Reason . . . 39 
 
 7. General Conception of a Science called Kritik of the Pure 
 
 Reason .... .41 
 
 8.H History of Kant's Discoveries . . 42 
 
CONTENTS 
 
 QHAPTER III 
 THE TRANSCENDENTAL AESTHETIC 
 
 PAGE 
 
 I. Definitions ... 47 
 
 2 & 4. Metaphysical Exposition of Space and Time . 48 
 
 3 & 5. Transcendental Exposition of Space and Time 51 
 
 6. Deductions • 54 
 
 7. Objections Answered 57 
 S. General Remarks on the Aesthetic . 60 
 
 CHAPTER IV 
 
 INTRODUCTION TO THE TRANSCENDENTAL ANALYTIC 
 
 I. Of Logic Generally 7° 
 
 II. Of Transcendental Logic . 73 
 
 III. Division of General Logic into Analytic and Dialectic . 74 
 
 IV. Division of Transcendental Logic similarly . 77 
 
 CHAPTER V 
 
 TRANSCENDENTAL LOGIC, PART I 
 
 The Transcendental Analytic ... 78 
 
 Of the clue to discover the Categories .... 79 
 
 Of the Logical Use of the Understanding ... 81 
 
 On the term Function . . . . 83 
 
 9. Of the Logical Function of the Understanding in Judgments 85 
 
 10. Of the Categories. ....... 88 
 
 H Note on Mr. Mill's Derivation of the External World 91 
 
 Table of the Categories . 94 
 
 11. Remarks on the Table . 96 
 
 12. Quodlibet ens est unum verum bonum . 98 
 1i Kant and the Critics of his Categories . 100 
 
Contents 
 
 CHAPTER VI 
 
 INTRODUCTION TO THE DEDUCTION OF THE PURE CONCEPTS 
 OF THE UNDERSTANDING, SECTION I 
 
 l'AUE 
 
 13. Of the Principle of Transcendental Deduction . . . 105 
 
 Transition to the Deduction . 109 
 
 CHAPTER VII 
 
 THE DEDUCTION OF THE CATEGORIES. THE FIRST EDITION 
 AND THE PROLEGOMENA 
 
 'i Various forms of the Deduction . 112 
 
 The Prolegomena . . . 114 
 
 The First Edition . 117 
 
 CHAPTER VIII 
 
 THE DEDUCTION OF THE CATEGORIES. THE SECOND EDITION 
 OF THE KR1TIK 
 
 The Deduction in the Second Edition 123 
 
 Nature of Synthesis . . . . 125 
 
 The Transcendental Apperception . 127 
 
 Production of Objects . . . 129 
 
 Application of the Categories .... 130 
 
 Office of the Imagination . . . 132 
 
 Results . 135 
 
 Supplement to the Aesthetic . 137 
 
 CHAPTER IX 
 
 THE TRANSCENDENTAL ANALYTIC, BOOK II. THE ANALYTIC 
 OF PRINCIPLES 
 
 The Transcendental Faculty of Judgment . . 143 
 
 Of the Schematism of the Pure Understanding 144 
 
 If Schopenhauer Criticised . 15 J 
 
 15- 
 
 
 16, 
 
 1/ 
 
 IS, 
 
 19. 
 
 20, 
 
 23' 
 
 24. 
 
 
 27. 
 
 
 24-; 
 
 
CONTENTS 
 
 CHAPTER X 
 
 SECOND CHAPTER OF THE ANALYTIC OF PRINCIPLES 
 
 PAGE 
 
 1. The system of all the Principles of the Understanding . 155 
 The Highest Principle of all Analytical Judgments 156 
 
 2. The Highest Principle of all Synthetical Judgments 157 
 
 3. Systematic Exposition of the Principles . 160 
 Note, Analysis of Combination 162 
 
 CHAPTER XI 
 THE MATHEMATICAL PRINCIPLES 
 
 1. The Axioms of Intuition ... 1 63 
 
 2. The Anticipations of Perception . . 165 
 Proof of Continuity . 168 
 
 CHAPTER XII 
 THE DYNAMICAL PRINCIPLES 
 
 I . Analogies of Experience . 171 
 
 A. First Analogy — of Substance 174 
 Its bearing on Kant's Idealism 177 
 
 B. Second Analogy — of Causality . 180 
 The Empirical Criterion of Substance . 185 
 
 C. Third Analogy — of Co-existence 1S9 
 
 CHAPTER XIII 
 THE POSTULATES OF EMPIRICAL THINKING GENERALLY 
 
 1. The Postulates of Empirical Thinking . 193 
 
 2. The Postulate of Possibility . 194 
 
 3. The Postulate of Actuality . . . . . 195 
 
CONTENTS 
 
 I'AGE 
 
 4. The Postulate of Necessity . . . .196 
 
 5. Relation of Possible and Actual 197 
 
 6. General Remarks . . 199 
 
 1 CHAPTER XIV 
 
 KANT'S ATTITUDE TOWARDS IDEALISM 
 
 1. Kant's Attitude on Idealism . . 202 
 
 2. Collation of Passages . 203 
 
 3. His View of Things per se , 206 
 
 4. His Critique of Berkeley . 209 
 
 5. His Critique of Descartes . . . 214 
 
 6. Refutation of Kant's Critics 219 
 
 CHAPTER XV 
 
 THIRD CHAPTER OF THE ANALYTIC OF PRINCIPLES 
 
 1. The Contrast of Phenomena and Noumena . 222 
 
 2. The Categories only of Empirical Use 223 
 
 3. Cause of the Belief in Noumena . 225 
 
 CHAPTER XVI 
 
 APPENDIX ON THE AMBIGUITY OF THE CONCEPTS OF 
 REFLECTION 
 
 1. The Nature of Transcendental Reflection . 229 
 
 2. Identity and Diversity . . 230 
 Agreement and Opposition . 231 
 Internal and External 231 
 Matter and Form . . . . 232 
 
 3. The Transcendental Topic . . 234 
 
 4. Analysis of Something and Nothing . . . . 235 
 
CONTENTS 
 
 CHAPTER XVII 
 
 TRANSCENDENTAL LOGIC, PART II 
 
 PAGE 
 
 I. Of Transcendental Illusion . 2 37 
 
 II. A. Of Reason in General . .240 
 
 B. Of the Logical Use of Reason 2 4' 
 
 C. Of the Pure Use of Reason • 242 
 
 CHAPTER XVIII 
 THE CONCEPTS OF PURE REASON 
 
 1. Of Ideas in General . . 245 
 
 2. Of Transcendental Ideas ... . 246 
 
 3. System of Transcendental Ideas . . • 251 
 
 CHAPTER XIX 
 
 THE DIALECTICAL SYLLOGISMS OF PURE REASON 
 
 1. The Transcendental Paralogism • ■ 254 
 
 2. The Transcendental Antinomy . . 254 
 
 3. The Transcendental Ideal . • 254 
 11 The Categorical Syllogism and the Absolute Subject . 255 
 
 CHAPTER XX 
 
 SYNTHETICAL TREATMENT OF THE PARALOGISMS 
 
 First Paralogism — of Substantiality . 260 
 
 Second Paralogism — of Simplicity . . 262 
 
 Third Paralogism — of Personality . 265 
 
 Fourth Paralogism — of Ideality (of external relation) . 267 
 
 Discussion in Second Edition . . . 269 
 
CONTENTS 
 
 CHAPTER XXI 
 
 ANALYTICAL TREATMENT OF THE PARALOGISMS 
 
 PAGE 
 
 Dicta of Rational Psychology , . . . . .271 
 
 Critique thereon ...... 272 
 
 Mendelssohn's Argument for Immortality . 274 
 
 Transition from Rational Psychology to Cosmology . 277 
 
 CHAPTER XXII 
 THE ANTINOMY OF PURE REASON 
 
 Cosmical Concepts . . .281 
 
 1. System of Cosmological Ideas . . 282 
 
 2. Antithetic of Pure Reason . 286 
 
 CHAPTER XXIII 
 
 THE ANTINOMIES 
 
 Of Quantity . 289 
 
 Of Quality 291 
 
 Of Relation . 294 
 
 Of Modality . 297 
 
 U Remarks on Fourth Antinumy 300 
 
 CHAPTER XXIV 
 
 THE INTEREST OF REASON IN THESE CONFLICTS 
 
 3. Interest on Side of Dogmatism . 3°3 
 
 4. Necessity of a Solution . . 3°6 
 
 5. Sceptical Account of the Cosmological Questions . 308 
 
 6. Transcendental Idealism supplies the Key 3 10 
 
 b 
 
CONTENTS 
 
 CHAPTER XXV 
 
 THE CRITICAL SOLUTION 
 
 I'AOli 
 
 7. Critical Decision of the Cosmological Conflict . 312 
 
 8. Regulative Principle .... 3>5 
 
 9. Empirical Use of the Regulative Principle . . • 3 J 7 
 Transition from Mathematical to Dynamical Antinomies 319 
 Possibility of Causality through Freedom ... 321 
 
 CHAPTER XXVI 
 
 THE IDEAL OF PURE REASON 
 
 1. The Ideal in General . . . 330 
 
 2. Prototypon Transcendentale 331 
 
 CHAPTER XXVII 
 
 PROOFS OF THE EXISTENCE OF GOD 
 
 3. Three Possible Proofs ... . 336 
 
 4. The Ontological Proof 338 
 
 5. The Cosmological Proof . . . 341 
 
 6. The Physico-Theological Proof . 346 
 
 7. Kritik of all Speculative Theology . 348 
 
 CHAPTER XXVIII 
 
 THE REGULATIVE USE OF THE IDEAS 
 
 Negative Conclusion of the Analytic . . . . • 3SI 
 
 1. The Principle of Homogeneity .... 3c, 
 
 2. The Principle of Specification . . . . . • 3 S4 
 
 3. The Principle of Continuity of Forms . . . • 3S4 
 Ultimate Aim of the Natural Dialectic of Human Reason . 3 c 7 
 
CONTENTS 
 
 CHAPTER XXIX 
 
 TRANSCENDENTAL DOCTRINE OF METHOD 
 
 l'AGE 
 
 The Discipline of Pure Reason . . 36 1 
 
 1. In the Sphere of Dogmatism . 362 
 
 2. In Polemics . 367 
 
 3. In Hypothesis . . 371 
 
 4. In Proofs . • 373 
 
 CHAPTER XXX 
 
 CANON, ARCHITECTONIC, AND HISTORY 
 
 The Canon of Pure Reason . . . -377 
 
 1. The Ultimate End of the Pure Use of our Reason . 378 
 
 2. The Ideal of the Supreme Good . . 379 
 
 3. Opinion, Knowledge, and Belief . 381 
 The Architectonic of Pure Reason 384 
 The History of Pure Reason .... . 3^8 
 
CHAPTER I 
 
 THE TWO PREFACES 
 
 § i. IT The philosophical student who has been discouraged 
 from opening the Kritik of the Pure Reason, by its reputa- 
 tion for obscurity and difficulty, will be agreeably surprised 
 by the clearness and the elegance of Kant's first Preface. 
 So easy is the flow of thought, so felicitous the choice of 
 expression, that we can only find stray hints of the arduous 
 task that awaits us. The explanations of a commentator 
 are almost needless, and his analysis cannot do better than 
 adhere as closely as possible to the rich and suggestive lan- 
 guage of the great philosopher himself. 1 
 
 § 2. Among the various branches of human knowledge, 
 says Kant, there is one in regard to which our reason is 
 condemned to a very strange lot, being troubled with 
 questions which we cannot decline, seeing that they are 
 forced upon us by our very nature, but which, nevertheless, 
 we cannot answer, since they transcend all our faculties. 
 
 Our reason falls into these perplexities unawares. It 
 
 1 Mr. Lewes's remarks [History of Philosophy, vol. ii. p. 458) are, in 
 the first place, too severe, and in the second place unjust, as they omit 
 to mention the genuine and even sublime eloquence of some of Kant's 
 writings. We may notice that Kuno Fischer's Commentary is silent on 
 these Prefaces, the second of which utterly destroys his theory as to Kant's 
 idealism. 
 
 I E 
 
2 THE FORTUNES OF METAPHYSIC chap. 
 
 commences from principles whose use in the field of ex- 
 perience is inevitable. From these it proceeds, according 
 to its nature, to ascend higher, and to approach more 
 remote conditions. But we soon discover that such a pur- 
 suit will never end, because fresh questions are ever starting 
 up. Under these circumstances, nothing remains but to 
 take refuge in first principles, which transcend all experience, 
 and which, nevertheless, excite so little suspicion that even 
 the common sense of mankind does not quarrel with them. 
 These, however, lead the reason into such obscurities and 
 contradictions that we cannot but infer the presence of 
 some hidden errors. Yet the discovery of these errors is 
 impossible ; because the principles adopted by the reason, 
 as they transcend completely the bounds 1 of experience, 
 will neither acknowledge nor submit to any test which 
 originates within it. This arena of endless dispute is called 
 Metaphysic. There was a time when her claims to be 
 called Queen of the Sciences were admitted by all. It is 
 now the fashion to despise her, and, like the mourning 
 Hecuba, she sits forgotten and forlorn ; in the words of 
 Ovid : Modo maxima rerum, tot generis natisque potens — 
 nunc trahor exid inops ! 
 
 At first her rule, under the administration of the Dog- 
 matists, was despotic. But this barbarous form of govern- 
 ment degenerated through internal dissensions into complete 
 anarchy ; and the Sceptics, a sort of nomads who hate all 
 settled conditions, periodically scattered the community. 
 But they were too few in number to prevent mankind from 
 continually attempting a reconstruction, though without 
 any fixed or consistent principle. It seemed indeed once 
 
 1 We uniformly translate Grenu, bound, reserving the term limit 
 for Schranke. Cf. Kant's Proleg. p. 154, for the distinction— an im- 
 portant one, though confused in all the translations. 
 
INDIFFERENTISM NO SOLUTION 
 
 in later times, as if the celebrated Locke's Physiology of 
 the Human Understanding would put an end to the dis- 
 putes, and settle for ever the lofty claims of Metaphysic. 
 But no sooner had the descent of the pretended queen 
 been traced to the low origin of common experience, and 
 her assumptions accordingly questioned, than this genealogy 
 was found out to be fictitious, and accordingly she persisted 
 in her claims. And so things returned to the antiquated 
 and rotten dogmatism, and to the consequent contempt in 
 which the science was held. Now that men think every 
 path has been tried in vain, disgust prevails, and total 
 indifference, the mother of chaos and of night in the 
 sciences, but the prelude of a better day. 
 
 For it is vain to assume an artificial indifference on 
 subjects which cannot be indifferent to human nature ; and 
 the pretended indifferentists, though they may endeavour to 
 disguise it by assuming a popular garb, are ever falling back 
 into the metaphysical assertions which they profess to despise. 
 Nevertheless, this indifferentism is a phenomenon deserving 
 our deepest attention. It is the result of the ripe judgment 
 of our age, which will no longer tolerate insecurity or false 
 pretence. We hear, 1 indeed, complaints of shallowness, and 
 of the decay of sound science. But well-founded sciences, 
 such as Mathematic and Physic, refute this calumny, not 
 merely by holding their ground, but even by making great 
 onward strides. So would other branches of knowledge 
 progress also, were their principles placed on a firm basis. 
 ' In default of this indispensable requisite, doubt, indiffer- 
 ence, and severe criticism are rather evidences of a thorough- 
 going spirit. Our age is the proper age of Kritik (criticism), 
 to which everything must submit. Religion desires to 
 escape by its sanctity, Legislation by its majesty.' They con- 
 1 Cf. the first note to this Preface. 
 
PROVINCE OF THE KRITIK chap. 
 
 sequently excite just suspicion, when contrasted with those 
 sciences which have freely and fairly met the test. 
 
 Reason is therefore challenged to begin afresh that most 
 difficult task, the knowledge of itself, and establish its claims, 
 not by oracular dicta, but according to fixed and unchange- 
 able laws. The court which must decide this issue is the 
 Kritik of the Pure Reason itself. 
 
 This does not mean a Kritik of books and systems, but 
 of the faculty of reason, as regards what knowledge it may 
 attain, apart from all experience. It is in other words the 
 deciding of the possibility or impossibility of metaphysic in 
 general, and the determining of its sources, extent, and 
 bounds, exclusively from principles. 
 
 This path, says Kant, I have pursued, and flatter myself 
 I have discovered the source of all the errors which divided 
 reason against itself. I have not evaded these questions by 
 falling back upon the impotence of the human reason, but 
 have determined them completely from principles, and, after 
 discovering the point at which reason fell into a misunder- 
 standing with itself, have completely and satisfactorily solved 
 them. I have aimed at completeness above all things, and 
 I venture to say there is not a single metaphysical problem 
 which is not either solved or provided with a key for its 
 solution in this book. Pure reason being a unity complete in 
 itself, any analysis which fails to solve a single question fairly 
 suggested concerning it must be cast aside as perfectly idle. 
 
 The reader need not wonder at these pretensions as 
 boastful and impertinent. They are infinitely more modest 
 than the programme of any ordinary system that pretends 
 to prove the simple nature of the soul, or the necessity of 
 a first origin of the world. For these things far transcend 
 all the bounds of experience, whereas the present work 
 merely analyses the reason and its pure thinking. It is in 
 
i ITS' CERTAINTY AND CLEARNESS 5 
 
 fact to be compared to common Logic, which analyses 
 completely all the simple operations of thought, 'except 
 that the question here proposed is, How much can we attain 
 through these operations if deprived of all the materials and 
 the assistance of experience ? ' 
 
 If, then, the matter of the book must be complete and 
 explicit, it may also fairly be demanded that the form of its 
 demonstrations should be certain and dear. Of course all 
 mere opinion is worthless, when we seek to establish a priori 
 knowledge. But it is for the reader to judge whether the 
 grounds advanced by the author are certain, and equal to 
 his pretensions. He has already asserted that he considers 
 this quality absolutely indispensable. Some few points, 
 however, which are not essential to the work might possibly 
 excite suspicion, and may, therefore, be pointed out before- 
 hand. 1 
 
 There is no investigation more important, or that has 
 given the author more trouble, than the so-called Deduction 
 of the Pure Categories. This investigation has two sides. 
 The first concerns the objects of the pure understanding, and 
 is intended to explain the objective validity of its a priori 
 concepts concerning them. This is an essential part of 
 Kant's plan, since it shows what the understanding and the 
 reason can know, apart from all experience. It stands upon 
 a basis perfectly independent of the second side of the 
 deduction, which considers the understanding subjectively, 
 and endeavours to analyse its faculties. ' As the question, 
 How is the faculty of thought itself possible ? can only be 
 answered by inferring a cause from its effects, my solution,' 
 says Kant, ' may seem (though this is not really the case) 
 
 1 As the reader can hardly understand them, before he has mastered 
 the passages in question, he will be reminded of these explanations at 
 the proper places. 
 
 c 
 
CLEARNESS 
 
 to be a mere hypothesis, and the reader may think himself 
 at liberty to differ from me.' But this will not invalidate 
 the former side of the argument. 1 
 
 As to clearness, the reader has a right to expect logical 
 and discursive clearness in the arguments, which has been 
 carefully attended to, and sesthetical and intuitive clearness, 
 by means of sufficient examples and illustrations. The 
 length and intricacy of the discussion have compelled the 
 author to dispense with these to an extent he did not 
 originally intend. A general survey of a system is often 
 impeded by these illustrations, and ' many a book would 
 have been far dearer if the author had not endeavoured to 
 make it so clear? 
 
 Metaphysic is the only science, as we shall show, which 
 is capable of absolute completion within a short time, by 
 means of united efforts : for it is nothing but the systematic 
 inventory of what we possess by pure reason. Nothing can 
 here escape us ; nor can any experience increase our know- 
 ledge. By means of the present Kritik the ground has been 
 cleared and prepared, and here the reader must perform the 
 part of an impartial judge. When we proceed to build up 
 the system of pure reason, under the title, ' Metaphysic of 
 Nature,' he should join us as a zealous co-operator, especially 
 as the investigation of these details is easy, and more a 
 recreation than a difficult task. 
 
 §3.H Such in substance, and to a great extent in words, 
 was the remarkable Preface with which Kant introduced 
 his great treatise to the philosophic world. It bears a strong 
 family likeness to the utterances of other intellectual reformers, 
 and suggests the mental attitude of Bacon and of Descartes, 
 of Locke and of Hume. There is the same boldness in 
 
 1 The reader should bear in mind that this passage refers to the First 
 Edition of the Kritik. 
 
THE SECOND PREFACE 
 
 asserting the discovery, and the same modesty in attributing 
 it not to genius, but to method. There are the same hopes 
 of a speedy termination of error, the same conviction that 
 even ordinary minds, when armed with proper weapons, can 
 help in the victory. 
 
 Kant's analogy to Bacon was indeed so striking, that he 
 prefixed to his second edition a memorable motto from the 
 Preface to the Instauratio Magna- — ' De nobis ipsis silemus. 
 De re autem, quae agitur, petimus, ut homines earn non 
 opinionem sed opus esse cogitent : ac pro certo habeant, 
 non sectae nos alicuius, aut placiti, sed utilitatis et ampli- 
 tudinis humanae fundamenta moliri. Deinde ut suis com- 
 modis aequi — in commune consulant, et ipsi in partem 
 veniant. Praeterea ut bene sperent, neque instaurationem 
 nostram ut quiddam infinitum et ultra mortale fingant et 
 animo concipiant : quum revera sit infiniti erroris finis et 
 terminus legitimus.' 
 
 In his first Preface, therefore, Kant explained the his- 
 torical position of the Kritik, and announced that he was 
 about to revolutionise philosophy by a new method. And 
 yet similar claims had often before been made, and had 
 hitherto failed. What was the peculiar novelty of Kant's 
 method which inspired him with such extraordinary con- 
 fidence ? How did he discover it ? In the Preface to the 
 Second Edition, published in 1787, six years after the First, 
 these questions are fully answered ; and, furthermore, the 
 positive practical results of his apparently negative and 
 speculative system are brought before the reader. This 
 second Preface is therefore a material improvement to the 
 work, and is of peculiar interest, as disclosing to us the line 
 of thought that led Kant to his discoveries. 1 
 
 1 Thus, at the very outset, those are in error who affirm that this 
 Second Edition is no improvement on the first. 
 
HISTORY OF THE SCIENCES 
 
 § 4. ' We can easily determine,' says Kant, ' whether we 
 are pursuing any branch of knowledge scientifically by our 
 success in attaining results. If after much preparation, we 
 are constantly coming to a standstill, and are obliged to 
 recommence, and if our fellow-labourers cannot agree with 
 us how our common object is to be pursued, our proceeding 
 is no science, but a mere groping in the dark, which we 
 shall do well to abandon, even with the sacrifice of many 
 lofty aspirations.' 
 
 The history of any recognised science will prove this 
 position. Consider Logic. Since Aristotle it has never lost 
 one inch of ground. But what is far stranger, it has never 
 gained anything either, and appears to be complete and 
 settled. If modern writers have attempted to enlarge it by 
 psychological chapters on the human faculties, or metaphysi- 
 cal on the origin of knowedge or the nature of certainty, 
 or anthropological on human prejudices, they have merely 
 shown their ignorance of its nature. ' We do not enlarge, 
 but disfigure the sciences, by confusing their boundaries.' 
 And the boundaries of Logic are strictly determined as the 
 science which expounds the formal laws of all thought, and 
 nothing but thought. The success of Logic is, however, 
 owing to its narrow scope ; for Logic is forbidden to con- 
 sider any of the objects of knowledge, and must confine 
 itself to the understanding and its form. It is accordingly 
 but the outer court of those proper sciences which add to 
 our real knowledge. With far more difficulty did reason 
 enter upon the strict path of science, when objects also were 
 concerned. 
 
 If such sciences are to be rational, they must contain a 
 priori knowledge, 1 and this knowledge of the reason may 
 be related to the object in two ways — either as merely deter- 
 
 1 So far as this means power of prediction, no one will dispute it. 
 
THE TURNING-POINT 
 
 mining it and the concept of it (when given from elsewhere) 
 or as also making it actual. This latter is the practical 
 cognition of the reason [as when it produces from itself, for 
 example, the idea of duty.] The former is theoretical cogni- 
 tion. In either case the direst confusion must result if the 
 pure part, in which the reason determines its object alto- 
 gether a priori, be not separately treated. 1 
 
 Mathematic and Physic are the two theoretical branches 
 of rational knowledge which aim at determining their objects 
 a priori — the former quite purely, the later partly so. 
 Mathematic attained the safe position of a science long 
 since among the Greeks. Yet this does not prove that its 
 safe highway was constructed easily, as in Logic. Kant 
 thinks that for centuries, especially among the Egyptians, it 
 consisted in mere groping, and that the great change was 
 owing to a Revolution, produced by a happy thought of some 
 forgotten genius. He notices that even in later times the 
 discoverers of special principles were remembered though 
 these principles were unimportant ; and this he ascribes to 
 the indelible effect produced by the first great Revolution. 
 
 In what did this Revolution consist ? ' Whoever it was,' 
 says Kant, 'that first demonstrated the equality of the 
 [base] angles of an isosceles triangle found 2 new light dawn- 
 ing upon him ; for he found that he could not trace out and 
 learn the properties of the figure from what he saw in it, or from 
 mere thinking about it, but rather from what he had added to 
 
 1 Kant in no way implies that these different parts of cognition are 
 given separately, but says that if we confuse them in our treatment, we 
 are like men spending money without keeping accounts, and then un- 
 able to ascertain what part of their outlay can be diminished when 
 economy becomes necessary. 
 
 2 ' Who first demonstrated the right-angled triangle ' is Mr. Lewes's 
 translation, which makes no sense. He was probably misled by n 
 clerical error (equilateral) noticed by Kant himself in one of his letters. 
 
io THE HISTORY OF EXPERIMENTAL SCIENCE chap. 
 
 the figure according to concepts a priori, and had then repre- 
 sented by a construction. He also found that all the safe a 
 priori knowledge he could attain about it was merely the 
 necessary consequence of what he had himself introduced into it, 
 according to his own concepts." 1 This statement may well be 
 regarded as the very corner-stone of all Kant's discoveries. 
 In natural philosophy progress was far slower, and it is 
 only since the days of Bacon that it has attained the high- 
 way of science. But even from an empirical point of view 
 there is a close analogy in its history to Mathematic. 
 When Galileo and Torricelli and Stahl began to make their 
 well-known experiments, the same light dawned upon them 
 also. They comprehended that reason discovers what it 
 produces according to its own plans ; that it must, so to 
 speak, take the initiative, according to fixed principles sup- 
 plied by itself, and compel nature to reply, instead of wait- 
 ing upon her for instruction, since chance observations are 
 not connected by such necessary laws as the reason seeks 
 and requires. Those phenomena alone that agree with some 
 fixed principle of the reason can be regarded as the laws of 
 nature. ' Armed with such principles, then, in one hand, 
 and with experiments framed according to them in the 
 
 1 Mr. Lewes's translation of this passage (Hist, of Phil. vol. ii. p. 467) 
 appears, through some oversight, not to be even grammatical, and, more- 
 over, obscures the point of the argument. Mr. Meiklejohn's translation 
 would here have afforded him a fair version. His interpretation of the 
 passage is equally erroneous ; for he thinks Kant is insisting on the 
 metaphysical method as opposed to the experimental, and is highly in- 
 dignant at the proposal to study nature through our ideas. But Kant is 
 really showing the vast superiority of the experimental method over that 
 of mere observation. In the latter case the mind can only note down 
 occurrences ; in the former it approaches the facts with a theory of its 
 own construction, and compels nature to say whether the facts conform 
 to it or not. Surely this just difference is acknowledged by every scien- 
 tific inquirer. The next paragraph should have made the point plain to 
 any careful critic. 
 
METAPHYSIC BEHINDHAND 
 
 other, our reason must approach nature to learn from her 
 not in the capacity of a scholar, but in that of a judge, who 
 compels his witnesses to give their evidence. And so even 
 Physic owes its happy revolution to the idea of seeking from 
 (not inventing for) nature that information, of which reason 
 coidd know nothing of itself, according to what reason had 
 itself introduced into nature.' 1 Up to this discovery physical 
 philosophers were merely groping in the dark. 2 
 
 Metaphysic has not hitherto been so fortunate. It is a 
 peculiar and isolated branch of knowledge, aiming at what 
 lies beyond experience, and that through mere concepts, 3 
 without representing them in figures, like Mathematic. And 
 here men have been constantly meeting with checks, and 
 endeavouring to begin afresh, so that the science may be 
 compared to the arena of a tournament, in which none can 
 establish any lasting possession. This is evidently mere 
 groping in the dark, and among pure concepts too, where 
 verification is not easy. 
 
 Why has the high road of science not been discovered 
 in this branch of our knowledge ? Is it imaginary ? Then 
 why has our nature been visited with such unavoidable and 
 restless longing ? and can we trust it in other things, if it 
 here proves a delusion and a snare ? Or is it our fault, 
 and have we hitherto failed to discover the right way, for 
 want of the proper method ? 
 
 Surely the examples of Mathematic and Physic, re- 
 modelled by a sudden revolution, are sufficiently remark- 
 able to induce us to make a similar attempt in Metaphysic, 
 
 1 I have marked the italics by way of answer to Mr. Lewes's 
 criticism. 
 
 - Cf. Bacon, "mera palpatio." De Aui^. Sci. v. 2. 
 
 3 Cr. Kritik, p. 435, for the difference between the mathematical 
 and philosophical methods. We refer throughout to Mr. Meiklejohn's 
 Translation. 
 
THE ANALOGY OF ASTRONOMY chap. 
 
 for they are obviously analogous to it as rational cognitions. 
 Hitherto it was assumed that our knowledge of objects 
 must conform to them, and all attempts to extend it a priori, 
 by means of concepts, have failed. Let us attempt the 
 problem of Metaphysic under the assumption that the ob- 
 jects must conform to our faculty of knowing them, — an 
 assumption which at first sight agrees better with the re- 
 quired possibility of knowing objects a priori — that is, of 
 determining something concerning them before they are 
 given to us. 
 
 ' This idea resembles that of Copernicus, who, when he 
 found that the motions of the stars could not be explained 
 by assuming them to revolve round the spectator, tried the 
 effect of making the spectator revolve, and the stars remain 
 at rest.' In Metaphysic a similar attempt can be made as 
 regards the intuition of objects. And Kant made this 
 attempt successfully some years before he discovered how 
 to make it in the other parts of the science. If our intui- 
 tion must conform to the nature of its objects, how can we 
 know anything a priori about these objects ? If the object 
 must conform to the peculiar nature of our intuiting faculty, 
 we may easily do so. 
 
 But we cannot stop at intuitions, and are compelled to 
 consider them as representations of some object which we 
 endeavour to determine through them. This object then 
 must be conceived. I may either assume that the concepts 
 by which I determine the object conform to it, and then 
 arises the old difficulty of obtaining any a priori knowledge ; 
 or I may assume that the objects conform to these con- 
 cepts. If I change the expression, and say that experience 
 conforms to my concepts, the result is the same. For in 
 and through experience alone do the objects become known 
 to us. This assumption then seems to promise good results, 
 
KANTS METHOD 
 
 13 
 
 for experience is a species of knowledge, and knowledge pre- 
 supposes the Understanding that knows. The Understand- 
 ing again presupposes certain rules, by which it acts, and 
 these rules must be considered logically prior to the objects 
 given through them. If we wish to express these rules, we 
 can only do so by a priori concepts, to which, accordingly, the 
 objects of experience must necessarily conform. Such is the 
 result attained by Kant's Analytic, and these a priori rules, 
 by which the Understanding proceeds, when it applies itself 
 to experience, are the Kantian Categories. There are other 
 objectors of which the Reason alone forms Ideas, and indeed 
 is bound to do so, but which cannot be given in experience, 
 or at least given as they are thought by the Reason. In 
 our attempts to think these objects we shall find an excel- 
 lent test of our new way of regarding the problem, which is 
 founded on the principle that we can only know that a priori 
 of objects which we have ourselves introduced into them. 1 
 This method, Kant adds, in a note, is borrowed from 
 physicaTscience, and consists in seeking the elements of the 
 pttreTeason in such a way that our results can be confirmed 
 or refuted by an experiment. We cannot indeed make ex- 
 periments with the objects of the pure reason (as in natural 
 philosophy), for they are ever beyond the bounds of all ex- 
 perience. But we can do it with the con cepts and principles 
 which we assume a priori, by so arranging them that the 
 same objects can be regarded from two different aspects — 
 first, as objects of Sense and Understanding suited to our 
 experience; secondly, as objects that are only thought, and 
 suited to the isolated Reason, transcending all experience. 
 If we find that, when things are regarded from this double 
 point of view, our reason is at harmony with itself, but if 
 
 1 Mr. Mill, in citing this passage {Exam, of Hamilton, pp. 31, 32), 
 omits the important words a priori. 
 
14 PLACE OF THE UNCONDITIONED chap. 
 
 from a single point, unavoidable contradictions arise, then 
 the experiment has proved the justice of our distinction. 1 
 
 Our attempt succeeds perfectly, and promises Metaphysic 
 the sure path of science in its first part, which is concerned 
 about such pure concepts as can. have corresponding ob- 
 jects given in experience. The possibility of a priori 
 knowledge can now be perfectly explained, and the laws 
 which lie a priori at the basis of nature (in the sense above 
 explained) can now for the first time be satisfactorily proved. 
 But there results a conclusion very adverse to the second 
 part of Metaphysic, which is that we can never advance 
 beyond the bounds of experience, and yet this was the chief 
 object of the science. ' Nevertheless, this very result tests 
 the truth of our first estimate of rational knowledge a priori, 
 which was that it concerns phenomena, and abandons the 
 thing per se as actual indeed in itself, but unknown to us.' 
 The Reason, indeed, necessarily requires the Unconditioned 
 to complete the series of conditions we find in phenomena. 
 [We cannot comprehend our mental phenomena without 
 presupposing necessarily a substance called Mind, beyond 
 and beneath all its various manifestations. This illustration 
 will explain what Kant means by the necessary belief in the 
 Unconditioned.] 
 
 Supposing that as long as we assume our empirical know- 
 ledge to conform to objects per se, the Unconditioned cannot 
 possibly be thought without contradictions, but that assuming 
 objects, as mere phenomena, to conform to our manner of 
 representing them to ourselves, the contradiction vanishes 
 [by confining the Unconditioned to things per se], then the 
 
 1 Kant's second Preface was written in answer to the criticism and 
 controversies excited by the First Edition, and therefore implies a gene- 
 ral knowledge of the book, without which the following observations 
 are necessarily obscure. 
 
I AIM AND VALUE OF THE KRITIK 15 
 
 latter assumption is established. We may, however, be able, 
 on practical grounds, to re-establish a priori what has been 
 shown to be unattainable by the speculative reason. In 
 fact, the ground may have been only cleared for such a 
 result. (This test, as Kant's note says, is similar to that 
 employed in chemistry, where the elements which have 
 been separated by analysis are again combined to reproduce 
 the original substance. Our Analytic separates pure a priori 
 knowledge into two heterogeneous elements, viz. things as 
 phenomena, and things per se. Our Dialectic combines 
 them again into harmony with the necessary idea of the 
 Unconditioned, and finds that this can only be done by 
 adopting the distinction.) Kant expressly tells us he has 
 adopted this hypothetical way of stating his conclusions, in 
 order to show the train of thought by which he arrived at 
 them. He insists that they have been perfectly established 
 in his work, by the nature of Space and Time, and of the 
 Categories. Copernicus's theories were at first mere hypo- 
 theses ; they ended not only by being demonstrated, but by 
 leading to the establishment of Newton's Law of Gravitation, 
 which could never have been discovered had they not been 
 assumed. 
 
 In this attempt, then, made after the model of geometers 
 and physical inquirers, consists the Kritik of the pure 
 reason. It is a treatise on the method, and not a com- 
 pleted system, of the science known as Metaphysic. But 
 owing to the unity and completeness of reason within itself, 
 the outline of the science is indicated by our method ; and 
 for the same reason Metaphysic, like Logic, is capable of 
 completion once for all, when the faculties of the human 
 mind have been surveyed in their nature and their use. 
 
 But what, it may be asked, is the value of this Kritik ? 
 It appears to be at first sight only negative, prohibiting our 
 
16 KNOWING AXD THIXKIXG chap. 
 
 speculative reason from transgressing the bounds of ex- 
 perience. This is indeed its first use. But this use 
 becomes f-csifizc, when we consider that previous attempts 
 have, as a necessary consequence, so extended the bounds 
 of mere sense, as to interfere seriously with the proper 
 practical use of the reason. For if such a practical use be 
 necessary to morality, though it cannot be assisted by 
 speculation, it must be secured against interference. You 
 might as well say the police were of no positive use, because 
 it is their negative duty to prevent peaceable citizens from 
 being molested in the pursuit of their business. The 
 analytical part of the Kritik shows that we cannot possiblv 
 have any speculative knowledge beyond the bounds of 
 experience 
 
 But it must be carefully observed, that though we cannot 
 kncu- objects as things per se, we are able to think them. 
 (In order to know a thing, I must be able to prove its 
 possibility either from its actuality in experience or a priori 
 through reason. But I can think what I like, provided I 
 do not contradict myself Such a thought is only logically 
 possible, and requires something additional to make it 
 really possible : however, this addition need not be sought 
 only in theoretical sources, but in practical also.) ' If we 
 ^ere unab'.e to think things per se there would follow the 
 absurdity of an appearance, without anything to appear.' 
 Xow let us suppose our distinction of things per se and 
 phenomena had not been made. If so, the law of causality 
 must bmd all beings absolutely. It would follow that the 
 
 kLT™ T "r^ P ° int ° f ™ W ' COUld not * "^ed 
 
 a thing - r wd as „ pheno^rir t^ti 
 po.nt of vtew ,ts v 1S ible actions cannot i ndee d bei^ bm 
 
i THE USE OF IDEAS 17 
 
 the result is different when we regard it from the former. 
 Without, therefore, being able to know my soul as free, I 
 liberate my attempt to think it such from an apparent 
 contradiction. 1 Suppose, now, that the freedom of the 
 will were one of the conditions, without which practical 
 morality were impossible, but that the speculative reason 
 had proved such an idea contradictory to itself, then freedom 
 must give way, and with it morality, to make place for the 
 necessity of natural causes. From such a result we are 
 saved by the Kritik. 
 
 The same great positive uses can be shown in our ideas 
 of God and of immortality. These cannot even be assumed 
 without checking the impertinences of the speculative 
 reason, which, by applying its empirical principles where 
 they are inapplicable, asserts all practical extension of the 
 reason to be impossible. Such knowledge is the real source 
 of immoral unbelief, and must be ordered to make way for 
 faith. We aim then at improving the reason of our pos- 
 terity by setting them to study a sober science, and saving 
 them from wasting time and trouble on idle groping and 
 pretentious dogmatism. Above all, we hope to dispose of 
 all objections against morality and religion after the manner 
 of Socrates, by proving clearly the ignorance of the 
 objectors. 
 
 And if there be any supposed loss in the surrender of 
 the claims made by the reason, the loss affects the monopoly 
 of the schools, not the interests of humanity. Were the 
 doctrines of the immortality of the soul, and of the existence 
 
 1 It must be carefully remembered that in this remarkable discussion 
 (Antinomy, sec. 9) Kant professed to prove, not the existence of 
 freedom, nor its probability, nor even its possibility, but simply that it 
 was not necessarily contradictory to causality. (See Kritik, p. 345.) 
 This guarded attitude has not been transferred to Mr. Lewes's exposition 
 (Hist, of Phil. ii. p. 519). 
 
 I C 
 
CLAIMS OF THE KRITIK 
 
 of God, ever really established by the subtile arguments 
 and distinctions of the schools ? Has it not been confessed 
 that the public are unable to grasp such refinements, and 
 that the former was rather proved by the profound inade- 
 quacy of the present life to satisfy our aspirations, and the 
 latter by the harmony, beauty, and mercy of nature ? 
 These proofs rather gain than lose, for the schools are 
 taught to pretend to no deeper knowledge, and to confine 
 themselves to the arguments that are accepted by the many, 
 instead of arrogating to themselves the sole possession of 
 such truths. But the schools remain in the possession of 
 a science most useful to the public, though the fact is not 
 recognised ; a science which cannot and need not ever be 
 popular, which refutes the arguments of the philosophers 
 who mystify the public, by objections equally subtile, but 
 saves them from drifting unconsciously into the assumptions 
 and the quarrels that have hitherto disgraced metaphysic. 
 ' Only by means of this science — the Kritik of the Pure 
 Reason — can Materialism, Atheism, Fatalism, Enthusiasm, 
 and Superstition be disarmed. We also cut away the very 
 roots of idealism and scepticism, but this rather affects the 
 schools than the mass of mankind. It would be far more 
 rational of governments to support such a science, than to 
 countenance the ridiculous dogmatism of the schools, which 
 raise an alarm about the public safety, when their cobwebs 
 are torn in shreds, though the public neither notice nor miss 
 them.' The reader will not fail to remember how very 
 similar, even in expression, was the design of Bishop 
 Berkeley. 1 
 
 Our Kritik is not opposed to the dogmatic procedure of 
 the reason in its pure cognition as a science, for every pure 
 
 1 Cf. Kritik, p. 265, for a like distinction between scepticism and 
 the sceptical method. 
 
CHANGES IN THE SECOND EDITION 19 
 
 science must demonstrate dogmatically — that is to say, from 
 sure principles strictly a priori. It is opposed to dogmatism, 
 which is the dogmatic procedure of the reason without pre- 
 vious criticism of its faculties} But we do not therefore 
 support shallow talking, which pretends to be popular, or 
 scepticism, which abolishes all Metaphysic. We rather 
 establish this science on a sound and systematic, not on a 
 popular, basis ; and in its future development we must fol- 
 low the steps of the great dogmatic philosopher, Wolff, who 
 may be regarded the originator of thoroughgoing and sys- 
 tematic inquiry in Germany. Had his ground been critically 
 prepared, he might indeed have established metaphysic as a 
 science. 
 
 § 5. IT These reflections, which agree closely with the 
 analytical and popular account Kant has given of his dis- 
 coveries in the Prolegomena (published in 1783), conclude 
 the exposition of his method and its results. There remains 
 an explanation of the changes introduced into his Second 
 Edition, a subject of less importance, had not Schopenhauer 
 made his pretended discovery that these two Editions 
 differed very materially, not only in exposition, but in 
 doctrine. It was said that Kant had become afraid of the 
 idealistic conclusions drawn from his principles, and had 
 suppressed the passages which resolve the whole external 
 object into our own sensations, and their form (imposed by 
 the mind also). More particularly, there was one paragraph 
 inserted into the Deduction of the Categories which dis- 
 tinctly states that the matter of our intuitions is given by a 
 source apart from, and independent of, the understanding; 1 
 and a refutation of Idealism was introduced into the Prin- 
 ciples of the pure understanding, in which Kant attempted 
 to prove that the objective existence of things in space is 
 
 1 Cf. JCritik, p. 89. 
 
20 THE PREFACES COMPARED chap. 
 
 the condition of our internal experience. Above all, in the 
 First Edition the distinction between soul and body was 
 explained to be a difference, not of substance (of which we 
 know nothing), but of representation ; and from this point 
 of view the community or relation of both was discussed 
 (cf. Appendix C, vol. ii. of this work). This was supposed 
 to be contradicted or extenuated 1 in the Second and follow- 
 ing Editions, for the purpose, Kuno Fischer thinks, of 
 gaining adherents. The question therefore assumes con- 
 siderable importance ; for it must determine, in the first 
 place, the degree of Kant's own conviction as to the truth 
 of his doctrine ; and, secondly, the real import of his 
 system. 
 
 Let us then first of all consult the author himself, and 
 consider what he says in his second and more elaborate 
 Preface 2 : — 'As regards this Second Edition, I naturally did 
 not wish to let the opportunity escape of remedying, as far 
 as possible, the difficulties and the obscurity from which 
 may have arisen the sundry misapprehensions that have 
 occurred to many acute men (perhaps without my fault) in 
 their estimate of this work. In the positions themselves, and 
 the grounds of proof , as well as in the form and completeness 
 of the plan, I have found nothing to alter : a fact which must 
 be ascribed partly to the long consideration to which I sub- 
 mitted my work previous to its publication, partly to the 
 nature of the subject itself, — I mean the constitution of a 
 purely speculative reason, which contains a veritable system 
 of members, where everything is organic — that is, where 
 
 1 In this particular case we have an exactly parallel statement in the 
 Second Edition. Compare Proleg. p. 234, with Kritik, p. 252. This 
 passage in the Second Edition does not seem to have been noticed by 
 critics. 
 
 2 The silence of Kuno Fischer as to these two Prefaces is very re- 
 markable. 
 
CHANGES OF DETAIL 
 
 the whole is for the sake of each individual part, and each 
 individual part for the sake of the whole; so that any 
 defect, however trifling, whether it be a positive error, or a 
 mere deficiency, is certain to betray itself in use. 1 . . . But 
 in the exposition much remains to be done, and in this 
 respect I have attempted to improve this Second Edition, 
 with the intention of clearing away partly the misappre- 
 hensions of the Aesthetic, especially of the concept of Time; 2 
 partly the obscurity in the Deduction of the Categories ; 3 
 partly to supply the supposed want of sufficient evidence in 
 the demonstrations of the Principles of the Pure Under- 
 standing ; 4 partly, in fine, to remove misapprehension as to 
 the Paralogisms laid to the charge of Rational Psychology. 5 
 . . . But the necessary consequence of this improvement, 
 except we made the work altogether too long, is a slight 
 loss to the reader, since a good deal (that did not indeed 
 belong substantially to the completeness of the whole) must 
 be omitted, or put into a shorter form, 6 which, nevertheless, 
 many readers might not wish to lose. This was done to 
 
 1 Cf. Kritik, p. xxxix. 
 
 2 Kant added Section i. § 6, on Time, and the General Remarks, ii.- 
 iv. (pp. 41-43). In his Introduction, Sections i. and ii. were greatly 
 expanded, and v. and vi. added. 
 
 3 From Section ii. § n, of the Transcendental Logic to the end of 
 the Deduction it was completely rewritten. 
 
 1 Under each of the Definitions of the Principles (with the exception 
 of the Postulates) the first paragraph, headed 'proof,' was added; as 
 well as two Appendices, entitled, ' General Remarks on the System of 
 Principles,' and the Refutation of Idealism, on which he also adds a 
 note in the second Preface. 
 
 6 From the words, ' but we shall for brevity's sake' (p. 241), the 
 whole discussion was rewritten. 
 
 The third chapter of the Analytic (on Phenomena and Noiimena), 
 and the Refutation of Rational Psychology, were considerably shortened, 
 part of the latter reappearing in the Refutation of Idealism. The De- 
 duction of the Categories is likewise abbreviated. 
 
22 NO REAL CHANGES chap. 
 
 make room for my present, and I venture to hope now 
 intelligible exposition, which in substance, as regards the 
 propositions, and even in their method of proof, changes 
 absolutely nothing ; but still varies [from the former] 
 here and there in the method of the exposition in such a 
 manner as could not be managed by interpolation. This 
 slight loss, which, by the way, can be supplied, if any one 
 chooses, by a compariso7i with the First Edition, is, I hope, 
 more than counterbalanced by the greater clearness ' [of 
 the present Edition]. 1 
 
 In the face of this declaration, which explicitly asserts 
 that nothing whatever has been altered in the system, and 
 which invites the reader to compare the two Editions, we 
 are told that the Second Edition is a mutilated, distorted, 
 and depraved work, caused by the weakness of old age, and 
 the fear of public opinion in Kant ! It can be proved by 
 the theological attitude of this, and of his later works, that 
 these charges are perfectly absurd. It will be also shown in 
 the course of this work that the supposed evidence of the 
 theory is derived from a series of blunders and oversights 
 (if not actual suppressions) in the interpretation of Kant's 
 very clear, though not dogmatic declarations. 
 
 As Schopenhauer's opinion is fashionable in Germany, we 
 do not wish to open the discussion without giving the 
 reader the means of judging for himself, by comparing the 
 two Editions ; he will find, accordingly, in the footnotes 
 and in the appendices to the second volume of this work, all 
 the passages of any importance which appear in the First 
 Edition only. The results of such a comparison are 
 simply these : that we may safely defy the advocates of the 
 First Edition to find any doctrine there stated to which 
 there is not a corresponding assertion in the Second ; or to 
 1 Cf. A'ritik, p. xli. 
 
NO REAL CHANGES 
 
 point out a supposed alteration in the Second Edition which 
 we cannot prove to be supported by quotations from the 
 original work. 1 The assertion of the honest author is most 
 decidedly true ; in the propositions themselves, and even in 
 their proof, absolutely nothing has been changed. 
 
 1 In Appendix C are added short footnotes, showing the special 
 points of agreement ignored by the critics, and explaining the sup- 
 posed points of difference ; and these will save us in this place from 
 quotations, as well as from the discussion of them. The vacillating 
 attitude of Mr. Mill and Mr. Lewes on this question has been noted in 
 the Preface : they seem unable to resist the force of the argument, but 
 at the same time they will not accept thoroughly the conclusions which 
 these arguments justify. The long note in the second Preface, which 
 refers to the Refutation of Idealism, and endeavours to improve the form 
 of his proof, will best be discussed in connection with the passage to 
 which it refers. The reader may therefore pass it by for the present. 
 
 Since the first appearance of this work a great deal has been written 
 on the question of the divergence between Kant's First and Second 
 Editions ; in particular, Mr. Stirling has given a valuable history of the 
 controversy in Mind, vol. viii. pp. 525 sqq. His conclusion is, that 
 Kant, ' being scandalised by certain misrepresentations of his doctrine 
 as mere Berkeleianism,' 'allowed the idealistic standpoint somewhat to 
 recede.' 'The passages most loudly idealistic in sound have been 
 omitted from the Second Edition.' But this latter statement cannot be 
 accepted without caution, in the face of the very passages quoted by 
 Mr. Stirling himself. It may be that Kant was somewhat frightened at 
 the charge of Berkeleianism ; but it still appears true, that though he 
 altered his language here and there, he did not at all alter his opinions. 
 This is the view of Ueberweg in his short Latin tract on the subject, and 
 is substantially the view originally put forward in the text. 
 
CHAPTER II 
 
 For the benefit of the reader we here give a general analysis of the 
 Kritik. Its main divisions are called respectively Elementology (pp. 
 1-430) and Methodology (pp. 431-517). The former is subdivided thus : — 
 
 sthetic (Sensibility) 
 
 I. 
 External intuition 
 
 Internal intuition 
 
 Logic (Understanding) 
 
 I I I 
 
 Analytic (of Concepts) Analytic (of Principles) Dialectic 
 
 (Understanding) (Judgment) (Reason) 
 
 I • „ I I 
 
 a. Categories a. Schematism a. Psychology 
 
 b. their Deduction b. Principles b. Cosmology 
 
 1. Theology 
 
 THE INTRODUCTION — -PRELIMINARY CONSIDERATIONS 
 
 The Distinction in Kind between the Cognitive Faculties 
 
 IT It is evident from the considerations urged in the Pre- 
 face, that a precise theory of the human faculties lies at the 
 basis of Kant's inquiry. He investigates the object through 
 the subject and its conditions. What then are the faculties 
 which teach us all we know of things ? The general 
 answer was given long ago, according to which they have 
 been unanimously divided by philosophers into two classes : 
 on the one hand, sense, sensation, external intuition or 
 impressions from without ; on the other, intellect, reflection, 
 
chap, ii THE FIRST DISTINCTION 25 
 
 internal intuitio n. though t. 1 But while this distinction was 
 generally admitted, there had ever been an a prioi'i fallacy 
 afloat (as Mr. Mill would say) that only one kind of know- 
 ledge could be true, and that all the information given us 
 from other sources than this must be either illusory or 
 defective. Hence Kant, who discussed the subject in 
 several shorter treatises prior to his Kritik, notices, that 
 under this assumption sensibility and understanding had 
 been in turn exalted into the sole source of true knowledge. 2 
 The idealists, from Plato to Leibnitz, had mistrusted the 
 senses, and considered sensibility a mere vague and con- 
 fused copy of the reality attained by thought. The sensual- 
 ists, from Epicurus to Condillac, had considered the senses 
 as the primary source from which the understanding com- 
 pounded or abstracted a faint reproduction of the external 
 reality. While, therefore, idealists and realists 3 are opposed 
 diametrically as to the true cognitive faculty, they agree in 
 one point, that Sensibility and Understanding differ in degree 
 only. 
 
 If Kant had joined either party it would certainly have 
 been that of the idealists. But he perceived that the 
 clearest of all our knowledge, that of geometrical figures, 
 was given by Sense, and that many concepts of the Under- 
 standing, such as the notion of right, could only be appre- 
 hended with great difficulty. He concluded that there 
 
 1 It will be seen, however, that Kant distinguishes between internal 
 intuition and thought, as, indeed, Locke had done before him. Internal 
 intuition belongs to sensibility, not to understanding. 
 
 2 Cf. Kritik, p. 516. 
 
 3 It has been customary in English Philosophy to confine the term 
 Realist to those among the schoolmen who asserted the separate exist- 
 ence of general ideas, as opposed to Nominalists. There is no reason 
 for this restriction. The term is here used to mean the philosophers 
 who assert the separate existence of our external percepts, as opposed 
 to the idealists. 
 
26 THE THREE FACULTIES chap. 
 
 wer^.tw^^ognitiyeTaculties^JiatalLy distinct,. ancL -differing 
 in kind — Sensibility and Understanding. This^ position is 
 the basis of the whole critical philosophy. Kant himself 
 has explicitly argued this point, and asserted this generic 
 distinction, and any commentator who endeavours to refine 
 it away, by speaking of these faculties as mere laws of 
 development, 1 or by insisting that it is after all the same 
 mind that knows in either case, 2 must be rejected as a 
 misleading guide. It is even questionable whether we 
 should designate them both under the same name of 
 faculty; for we shall see that Sensibility is. a _ passive recep- 
 tivity, whilst Understanding is an active spontaneity. We 
 find, indeed, at a later stage of the Kritik, a_ third faculty 
 introduced, and called the Reason, as distinguished from 
 the Understanding. But if we attend to Kant's own 
 explanation, 3 he tells us ' that the Reason does not give 
 birth to any [new] concept, but only frees the concepts of 
 the understanding from the limits of experience.' In fact, 
 it differs from the understanding not in its essential nature, 
 but in its aim, which is the unconditioned. This second 
 distinction, then, though very important, is by no means so 
 fundamental or so trenchant as that which has just been 
 explained. In accordance with the latter, Kant divides his 
 Kritik of the reason into the Transcendental Aesthetic, or 
 Kritik of the Sensibility, and the Transcendental Logic, or 
 Kritik of the Understanding. The former had been 
 sketched, completely enough, in his earlier treatises, and 
 seems to have cost him far less labour to discover (as it 
 also costs us far less to understand) than the latter. 4 
 
 1 Dr. Webb, Intellectual: sm of Locke, p. 168, note. 
 This question 1ms been fully discussed by Dr. Ingleby and Pro- 
 fessor Sylvester. See Sylvester's Laws of Verse, Appendix. 
 
 3 Kritik, p. 256. 
 
 4 In pointing out the relation between Reason and Understanding in 
 
THE SECOND DISTINCTION 27 
 
 We can now follow Kant's Introduction, observing closely 
 the order of his exposition. He claims particularly to have 
 been the first who understood, and who therefore stated 
 correctly, the Problem of Mctaphysic. We are about to 
 investigate the human faculties. We can only do this by 
 analysing the effects which they produce, and these effects 
 are know ledge or cognition. 1 The q n flstion, th erefore, of 
 t he Krit ik isthlsT How is the fact of cognition possible ? 
 and more particularly, ho w is syntheti cal a priori cognition 
 possible? 2 
 
 Kant's Introduction 
 
 § 1 . Of the Distinction of Pure and Empirical Cognition. 
 — There can be no doubt that our knowledge begins chrono- 
 logically with experience. For our senses are first affected 
 by external objects, and our understanding is first occupied 
 in 3 comparing or arranging the materials so obtained. But 
 though all knowledge begins with experience, this does not 
 imply that it originates * from experience. For it might, 
 even when obtained from experience, be a composite thing, 
 consisting partly of impressions, partly of additions made 
 (in the act of receiving them) by our understanding. And 
 
 the Kritik, it must be observed that Kant uses the latter term in two 
 distinct meanings — ( 1 ) as the general faculty of thought, as opposed to 
 Sensibility ; (2) as the special faculty of concepts, as distinguished from 
 Judgment and Reason. The word concept also is used ambiguously, 
 generally being opposed to intuition, but is sometimes used in a wide 
 sense, to signify any mental state [Cf. infra, p. 31, note]. Cf. the 
 analytical table on p. 24. 
 
 1 These terms are in the following work used synonymously. 
 
 - As Kuno Fischer puts it, there are three questions to be answered 
 in the sequel — (1) What is cognition ? (2) Does it exist? (3) How is 
 its existence to be explained, or how is it possible ? 
 
 3 Cf. Locke, Essay, book ii. chap. 12, § 1. 
 
 4 Cf. the distinction between origo and exordium. 
 
28 A PRIORI COGNITIONS chap. 
 
 it may be very difficult to separate these additions, and 
 recognise them as such. The question, whether there be 
 cognitions independent of all the impressions of the senses, 
 is not therefore to be lightly decided. [It may be objected, 
 that elements added by the cognitive faculties cannot pro- 
 perly be called cognitions, for they do not teach us to know 
 things different from the mind, but rather interfere with 
 such knowledge. We may admit this objection so far as to 
 allow that if our understanding fuses its own conditions 
 with the impressions received from without, these things, as 
 they are apart from us, cannot be known. But surely, in 
 things as they appear to us, these elements must be of the 
 last importance.] 
 
 The cognition of these elements is called a priori, as 
 distinguished from that which is derived from experience, 
 which is a posterioi'i. The popular meaning of a priori is 
 simply that our knowledge is derived, as opposed to special 
 experience, from a general rule, which may have been itself 
 originally derived from experience. You say that a man 
 who undermines the foundations of his house might have 
 known a priori that it would fall. Yet he must have 
 learned from experience that bodies are heavy, before he 
 could make this inference. We intend, then, to use the 
 phrase a priori cognition of such as is absolutely independent 
 of all experience. A priori cognitions are pure, if they 
 have no empirical elements mixed with them ; if they have, 
 they are mixed. So the assertion : every change has a 
 cause, is a mixed cognition, because change is a notion 
 that can only be obtained from experience. 
 
 § 2. We possess certain a priori cognitions, and even the 
 ordinary undej-standifig always contains them. — By what 
 mark, says Kant, can we surely know that we possess any 
 pure, as opposed to empirical, knowledge ? Firstly, Should 
 
THEIR CRITERIA 29 
 
 there be a proposition which is thought together with its 
 necessity, then it is an a priori judgment, and if underived 
 from any other x not itself necessary, absolutely a priori. 
 Secondly, Experience never gives us strict and absolute, but 
 only comparative universality, gained by induction, and 
 which asserts merely that so far we have found no exception. 
 Should any judgment be thought in strict universality, i.e. 
 so that no exception is regarded possible, it cannot be 
 derived from experience, but is valid absolutely a priori. 
 Empirical universality is then but an arbitrary or contingent 
 exaggeration from the cases we and others know, to all 
 cases ; whereas strict universality is essential to the judgment 
 in which it is found, and points to a peculiar source of 
 knowledge, which we have designated a priori. Necessity 
 and strict universality are certain marks of an a priori 
 cognition, and are inseparable. But as empirical limitation 
 is at times more easily shown than contingency, and it is 
 often more convincing to show the unlimited universality of 
 a judgment than its necessity, we may use these two criteria 
 separately, each of them being in itself infallible. 
 
 ft This view of the criteria of a priori knowledge has 
 not met with general acceptance. Kant says indeed very 
 justly, that exceptions in experience are more easily shown 
 than abstract contingency in judgments, but many philo- 
 sophers would demur to having necessity proved by uni- 
 versality. Sir W. Hamilton, indeed, distinctly deduces the 
 latter from the former. But the school of Mr. Mill, while 
 admitting the importance of universality, hold that it can 
 prove only a subjective necessity, or conviction, stronger in 
 degree than empirical conviction, but not differing from it in 
 kind. We cannot enter now into the details of this con- 
 
 1 Mr. Stirling, in his translation, omits the important words, 'not 
 itself necessary. ' — Text Book to Kant, p. 117. 
 
30 OBJECTIVE NECESSITY CHAr. 
 
 troversy ; but the important point to be noticed is that the 
 necessary judgment is thought as necessary and exception- 
 less ; in thinking that A is B in such a case, we simul- 
 taneously think that it could not be otherwise. 
 
 That there are strictly universal and necessary, and there- 
 fore a priori judgments in human knowledege, says Kant, is 
 easily shown in science, by mathematical judgments, in 
 ordinary life, by the assertion that every change must have 
 a cause ; so plainly indeed does the latter concept contain 
 these criteria, that it would be altogether lost were we to 
 deduce it, as Hume did, from mere frequent association, 
 and so allow it only a subjective necessity. But Kant thinks 
 that without any examples, pure a priori principles can be 
 shown indispensable to experience. For experience must 
 deduce its certainty from some fixed principles, and not 
 from rules, which are themselves all empirical and con- 
 sequently contingent. These, he adds, could hardly count 
 as first principles. [Had Kant expanded this proof, it 
 would have been an instance of what he calls his transcen- 
 dental proof, which, from the existence of a fact in our 
 cognition, proves the existence of the necessary conditions 
 from which alone the fact can result. 1 ] He is ' content 
 however, in this place, to note the existence of a pure use 
 of our cognitive faculties, and its attributes.' We must 
 remember that they belong, not only to judgments, but some- 
 times to notions. So the space occupied by a body, or 
 what we consider its substance, cannot be abstracted from it. 
 
 IT It is, I think, much to be regretted, that Kant did not 
 give more weight to the force of custom, or subjective 
 necessity, as he calls it, and show clearly that it may in all 
 cases be distinguished from real or objective necessity. And 
 this omission in the Kritik is the more remarkable, as he 
 1 Cf. Kritik, p. 478. 
 
MATHEMATIC HERE MISLEADING 
 
 had before him the writings of Hume, in which the effacing 
 of this distinction was a capital feature. 
 
 § 3. Philosophy requires a (special) science to determine the 
 possibility, the principles, and the extent of all a priori cog- 
 nitions. — We have seen in the Preface how certain cognitions 
 attempt to transcend all experience, and to enlarge our 
 knowledge independently of it. Nay, this very knowledge 
 is generally regarded as the most noble and important. 
 Such are the problems that concern God, Freedom, and 
 Immortality. But it might naturally have been expected 
 that we should have determined accurately the origin, validity, 
 and value of the principles we have applied in these re- 
 searches. If we mean by naturally, what ought to be, this 
 remark is just ; but if we mean what usually happens, there 
 are solid reasons for expecting this investigation to be long 
 delayed. 1 For the recognised security of mathematical 
 knowledge leads us to expect the same from other a priori 
 cognitions, though they are quite different in nature. And 
 these we pursue with such ardour, that only clear contradic- 
 tions will check us. Unfortunately the facts of experience, 
 which in other sciences test idle theories strictly, have here 
 no application. We ignore the fact that Mathematic, which 
 has made brilliant advances in a priori knowledge, is strictly 
 confined (as we shall see) to intuition. But the intuitions 
 with which Mathematic deals are given a priori, and are 
 therefore hardly distinguishable from pure concepts. 2 This 
 example then excites us with the hope of great results. The 
 fleet dove, that cuts the resisting air in her flight, might 
 think to increase her speed if space were a vacuum. So 
 
 1 Cf. Kritik, p. 434. 
 
 2 This remark shows why Kant vacillates in his language about space 
 and time, calling them, in an earlier treatise, even conceplus spatii et 
 temporis. He also speaks of the ' ' Begriff des Raumes, " at the beginning 
 of the Metaphysical Exposition of Space. Cf. note, p. 24, supra. 
 
32 THE THIRD DISTINCTION chap. 
 
 Plato left the world of Sense, and ventured on the wings 
 (as it were) of Ideas, into the vacuum of the pure Under- 
 standing. He failed to perceive that he could make no 
 way, for want of a resisting medium, in which to apply his 
 powers. 
 
 Speculation is ever hastening to complete her structure, 
 and only then begins to consider the soundness of the 
 foundation. Our suspicions are generally lulled during the 
 construction by this fact, that perhaps the greater part of 
 the work of our reason consists in the mere analysis, in 
 formal explication of the concepts we already (though 
 perhaps confusedly) possess. This sober and useful process 
 seduces the reason to make unwittingly quite a different 
 sort of assertion about given concepts, in which new matter 
 is joined to them a priori, without questioning our right to 
 do so. This distinction must be forthwith explained at 
 greater length. 
 
 § 4. The distinction between analytical and synthetical 
 judgments. — Though this distinction has become a house- 
 hold truth in philosphy, Kant's analysis has never been 
 accurately expounded. The reader must pay particular 
 attention to it, if he wishes to understand clearly the objective 
 necessity of mathematical judgments. 
 
 If I assert, says Kant, of a body, that it is extended, I 
 only assert an attribute necessarily contained in the notion. 
 It is by an analysis of the notion that I form the judgment, 
 and it is hence called analytical or explicative, as enumerat- 
 ing clearly elements contained obscurely or confusedly in 
 the concept. But if I assert of a body, that it has weight, 
 I assert what cannot be discovered by any analysis of my 
 notion of a body. This judgment is therefore synthetical 
 or ampliative, enriching our notion by the addition of a 
 new attribute. 
 
n LOCKE AND KANT 33 
 
 H It was pointed out in a former work x that Locke had 
 completely anticipated this celebrated distinction. Mr. 
 Lewes 2 thinks that a glance at the Prolegomena (p. 24) would 
 have shown any one that Kant fully recognised Locke's 
 priority. We do not know what a glance at the passage might 1 
 have done, but a careful perusal of it has convinced us that • 
 Kant (who was not 'fully alive to Locke's priority ') did not 
 know the really decisive passage. It is not that cited by 
 Mr. Lewes (Locke, Essay iv. 8, 8) and Kant (iv. 3, 9), but 
 that cited by Dr. Webb in the first chapter of the 4th book 
 of the Essay, where Locke enumerates the four kinds of 
 agreement and disagreement between our ideas: (1) Identity 
 and diversity, viz. 'blue is not yellow;' (2) Relation, viz. 
 ' the three angles of a triangle are equal to two right ; ' (3) 
 Co-existence : ' gold is soluble in aqua regia ; ' (4) Real 
 existence, viz. ' God is.' Here are Kant's analytical, his 
 synthetical a priori, and synthetical a posteriori judgments 
 accurately distinguished, and his very examples almost 
 anticipated ; and in the fourth the distinctness of existential 
 judgments is asserted, which, as we shall see, Kant proved 
 to be synthetical, but sttbjectively so, by the addition, not of 
 an attribute, but of a relation to ourselves, and therefore he 
 also distinguished them from other synthetical judgments. 
 It may be noted, however, that Locke nowhere gives an ex- 
 ample of a proposition which would correspond to Kant's 
 synthetical a priori in Physic, e.g. every event has a cause. 
 That a proposition such as this might be synthetical and yet 
 a priori was perhaps Kant's greatest discovery. 3 
 
 1 Fischer's Comm. p. 28, note. ' Op. cit. vol. ii. p. 475, note. 
 
 3 Indeed, though the distinction between analytical and synthetical 
 may have occurred to Locke, he does not consistently uphold it — e.g. 
 in iv. 3, 18, when discussing the possibility of an a priori science of 
 ethics, the propositions that he suggests as its basis are only analytical. — 
 Cf. Essay i. 2, 9-20, and iv. 8, 13, on this subject. 
 I D 
 
34 DESCARTES AND KANT chap. 
 
 U I agree with Kant that mere hints are not anticipations, 
 and do not, therefore, claim any exaggerated importance for 
 a curious passage in Descartes's 14th Regie pour la direction 
 de F esprit, though in it he lays down the Kantian distinction 
 of analytical and synthetical as plainly as it can well be ex- 
 pressed, and shows how previous philosophers had confused 
 these judgments, and consequently fallen into errors. Here 
 is the passage : — 
 
 ' Passons maintenant a ces paroles : un corps a de 
 V etendue; bien que nous comprenions que dans cette phrase 
 etendue signifie. autre chose que corps, cependant nous ne 
 formons pas dans notre imagination deux idees distinctes, 
 l'une d'un corps et l'autre de l'etendue, mais une seule, celle 
 d'un corps qui a de l'etendue. Au fond c'est comme si je 
 disais: un corps a de F etendue; ou plutot ce qui a de Vetendue 
 a de l'etendue; cela est particulier a tout etre qui n'existe que 
 dans un autre et qui ne peut etre compris sans un sujet; il en 
 est autrement pour les etres qui se distinguent rdellement 
 des sujets. Si je dis, par exemple, Pierre a des richesses, 
 l'idee de Pierre est entitlement differente de celle de 
 richesses; de meme si je dis : Paul est riche, je m'imagine 
 tout autre chose que si je disais, le riche est riche. Faute 
 d'apercevoir cette difference, la plupart pensent a tort que 
 l'etendue contient quelque chose de distinct de ce qui a 
 de l'etendue, comme les richesses de Paul sont autre chose 
 que Paul.' Descartes, however, in trying to evolve the 
 existence of God out of our idea of Him, asserts, at least in 
 this particular case, that existential judgments are analytical. 
 The truth is, though other philosophers may have hinted at 
 the distinction, Kant was the first to grasp it thoroughly. It 
 is also approached by Hume, Essays, vol. ii. p. 165 : — ' That 
 the square of the hypotenuse is equal to the squares of the other 
 two sides, cannot be known, let the terms be ever so exactly 
 
II PRINCIPLE OF SYNTHETICAL JUDGMENTS 35 
 
 defined, without a train of reasoning and inquiry. But to 
 convince us of this proposition, that where there is no pro- 
 perty there is no injustice, it is only necessary to define the 
 terms and explain injustice to be a violation of property.' 
 But to return. 
 
 All analytical judgments depend upon the Laws of 
 Identity and Contradiction. 1 You cannot deny to a con- 
 cept any of its parts, without at once contradicting your own 
 act of conception. All analytical judgments are also a priori, 
 however empirical the concept concerned may be, for they 
 require no additional experience, but a mere dissection of 
 given notions. But here a priori is used in the popular 
 sense explained above. Synthetical judgments must of 
 course conform to the logical law of Contradiction also, but 
 still they can never be obtained from it alone, and require 
 some distinct principle in addition. What can this prin- 
 ciple be ? 
 
 All empirical judgments are synthetical. For it were 
 idle to apply to experience for any information that could 
 be obtained by analysis of our concepts. But if these 
 judgments join new elements to our previous concepts, what 
 guarantee have we that these elements ought to be so joined? 
 We must know that the predicate belongs to the subject, or 
 we have no cognition. Kant has replied to this difficulty 
 very fully, especially in his First Edition. Let us first take 
 the case of a posteriori judgments, such as ' all bodies are 
 heavy.' How do we know this ? From experience. The 
 concepts body and heavy are distinct from each other, and 
 can only be brought into connection by comparing them 
 both with some third thing (x, as Kant calls it) with which 
 they will both agree. For all synthetical a posteriori judg- 
 ments this x is our past experience. ' Both concepts (al- 
 1 Cf. Proleg. p. 17- 
 
36 PRINCIPLE OF SYNTHETICAL JUDGMENTS chai\ 
 
 though the one is not contained in the other) still belong 
 to one another (though only contingently) as parts of the 
 whole experience, which is itself a synthetic combination of 
 intuitions.' 1 
 
 A more difficult problem remains. All synthetical judg- 
 ments are not a posteriori, for many of them are universal 
 and necessary. If they be synthetical, and also a priori, 
 what is the x, which affords us the real synthesis ? 2 When 
 we assert of an effect, that it must have some cause, this 
 never could be obtained from the analysis of the concept of 
 effect ; where then did we find the combination a priori of 
 effect and cause ? for as this judgment is absolutely univer- 
 sal and necessary, we could not as before have recourse to 
 our complete experience, of which it is indeed one of the 
 very conditions. The answer to this question was one of 
 Kant's greatest discoveries. But he made it first in the 
 field of Mathematic. 
 
 1 The question whether propositions can ever pass from the class of 
 synthetical to that of analytical has been much debated. The truth 
 seems to be this : & judgment must be one or the other ; either the pre- 
 dicate is thought in the subject, or it is thought out of it: there is no 
 alternative. At the same time, a proposition (i.e. a judgment expressed 
 in words) may be analytical for one man, and synthetical for another ; 
 but then it does not represent the same judgment for both individuals, 
 for they are using terms in different senses. And even for the same in- 
 dividual a proposition may at one time be synthetical, and at a later 
 period, when larger experience has added to his concepts, analytical. 
 For omniscience all true propositions are analytical, but for a finite in- 
 telligence the distinction is fundamental. Cf. Monck, Logic, pp. 122 
 sgq., for a discussion of this question. 
 
 2 Kant's answer to this question is, we shall see, briefly this : In a 
 priori synthetical judgments in Mathematic, the x is the a priori in- 
 tuition ; in those in Physic it is the possibility of experience ; in Meta- 
 physic such judgments are impossible, for no x can be found. 
 
II SYNTHETICAL JUDGMENTS AS PRINCIPLES 
 
 37 
 
 § 5. All theoretical Sciences contain Synthetical Judgments 
 a priori as Principles 
 
 1. While Philosophy 1 is satisfied with discursive judg- 
 ments about concepts, Mathematic insists on proving each 
 step by intuition, and this observation gives us the clue to 
 its first condition. This condition must be intuition, and it 
 must be a priori. For though mathematical, like all other 
 true judgments, must conform to the Law of Contradiction, 
 such mere analysis does not explain their real nature. 
 Consider the judgment 7+5 = 12. All previous philo- 
 sophers considered this a mere analytical inference. But 
 Kant denies that the concepts of 7, of 5, and of their 
 addition, actually contain 12 as a necessary element. 'We 
 must go beyond these concepts, and obtain the assistance 
 of the intuition corresponding to either of them — suppose 
 the fingers of a hand, or five points in a space — and add 
 the units of the five given by intuition successively to the 
 concept of 7.' 2 When this operation is completed, and 
 then only, do we see the result to be 1 2. All such arith- 
 metical judgments are therefore synthetical, as may be 
 easily proved by considering the addition of large numbers. 
 We there find that no analysis of our concepts will give us 
 the required result. 3 Geometrical judgments are equally 
 
 1 Cf. Proleg. p. 36. 
 
 3 The reader will observe that Kant proposes to add the intuition of 
 5 to the concept of 7, the very expression repeated verbatim in the Pro- 
 legomena. He appears to mean that the 7 is a made up group, whereas 
 the 5 units are added seriatim. 
 
 3 It is curious to find, in a book published in 1886, a reiteration of 
 the opinion that mathematical judgments are analytical ; and yet Pro- 
 fessor Knight in his Hume falls into this blunder (p. 164), and remarks 
 that the difference between mathematical judgments and the law of 
 causality is practically that the first are analytical, the second synthe- 
 tical. This, he says, ' leads straight to Kant's reply to Hume ! ' As to 
 
38 THE PRINCIPLES OF MATHEMATIC chap. 
 
 synthetical. If I say that a right line is the shortest pos- 
 sible between two points, I cannot elicit anything about its 
 shortness, which is quantity, from the mere concept of its 
 straightness, which is quality. 1 
 
 (There are indeed, 2 Kant parenthetically observes, in 
 geometry and also in arithmetic analytical judgments de- 
 pending on the Law of Contradiction, such as a = a and 
 a + b> a, but neither are these the principles on which the 
 demonstration is based, nor would they be admissible in 
 mathematics were they not capable of being expressed in 
 intuition.) What misleads us about the synthetical judg- 
 ments of mathematic, and makes us regard them as analy- 
 tical, is an ambiguity of expression. To have a synthetical 
 a priori judgment we must necessarily join the predicate 
 in thought to the subject, the given concept. But the 
 question is not, what we must join in thought to the given 
 concept, but what we actually think in it, though only 
 obscurely ; and then it becomes manifest, that the predicate 
 in synthetical a priori judgments pertains to the concept, 
 necessarily indeed, yet not as thought in the concept itself, 
 but by virtue of an intuition, which must be added to it. 3 
 
 the nature of the judgment, 7 + 5 = 12, which Kant regards as synthetical, 
 the curious reader may find a long controversy in Mind, vol. viii. — Mr. 
 Sidgwick, in opposition to Mr. Monck, upholding the opinion of Ueber- 
 weg, that such a judgment may be regarded as analytical if we go back 
 to the definitions of 7, 5, &c 
 
 1 In the Prolegomena Kant adds a still clearer example. All the proofs 
 of equal triangles resolve themselves ultimately into super-position, which 
 is no logical analysis, but a direct appeal to intuition. 
 
 2 Professor Max Midler, in his Translation, puts this sentence in~a 
 parenthesis, ; and [remarks (Preface, p. lviii.) that Dr. Vaihinger first 
 gave this explanation of the passage. But that these words are 
 parenthetical was pointed out long ago by Mr. Monck, and they were 
 placed in brackets in the first edition of this book. 
 
 3 The ambiguity is in the words, ' we must join the predicate to the 
 subject.' This, in synthetical judgments, says Kant, does not mean 
 
THE PRINCIPLES OF METAPHVSIC 39 
 
 2. Physical science contains a priori synthetical judg- 
 ments as its principles. The examples which Kant gives 
 are not the principle of causality [as Kuno Fischer alleges], 1 
 but the assertions that the quantity of matter in nature is 
 constant, and that action and reactio?i are always equal. 
 Permanence is not part of our concept of matter ; Reaction 
 is not so either, and so these propositions are synthetical. 
 
 3. Metaphysic, whether we grant its scientific value or 
 not, at all events pretends to occupy itself not about analys- 
 ing concepts, but about extending our a priori knowledge, 
 and it employs such a priori synthetical principles as our 
 experience cannot even grasp. We can take as an example, 
 the world must have had a beginning. Metaphysic then 
 aims, at all events, at consisting of nothing but synthetical 
 a priori judgments. [When Kuno Fischer gives as an 
 example judgments asserting existence, he forgets that the 
 synthetical nature of such judgments is only established in 
 the latter part of the Kritik by a long and difficult discussion, 
 and could therefore not be here quoted as a commonly re- 
 ceived truth.] 
 
 § 6. The general Problem of the Pure Reason. — It is 
 very useful to comprehend a number of investigations 
 under a single formula. Both the proof and the refutation 
 are thereby simplified. A single question expresses the 
 problem of the pure reason, how are synthetical a priori 
 judgments possible ? By the solution of this problem or 
 the proof that such judgments are not possible, Meta- 
 physic must stand or fall. David Hume, of all previous 
 philosophers, approached nearest to this problem, but 
 
 that the predicate is a part of the subject, but that it must be necessarily 
 added to it. 
 
 1 Ueberweg goes wrong in giving the Law of Causality as an ex- 
 ample of a synthetical a priori judgment in Metaphysic (Hist, of Phil. 
 vol. ii. p. 156). 
 
40 HUME AND KANT 
 
 did not state it to himself either distinctly or universally 
 enough. He confined his attention to the Principle of 
 Causality, 1 and exploded Metaphysic as in truth borrowed 
 from experience, though decked out with an apparent 
 necessity engendered by habit. A larger consideration of 
 the question would have shown him that his conclusion, 
 viz. that no synthetical proposition could be a priori? 
 disproved the possibility of Mathematic, a result at which 
 his good sense must have revolted. 3 Kant tells us, in his 
 Prolegomena, that this scepticism of Hume was the exciting 
 cause that prompted his first critical doubt. 
 
 The solution of the above problem explains the possi- 
 bility of all sciences which contain a theoretical cognition 
 of objects a p7-iori, and therefore answers the questions : 
 How is pure Mathematic and how is pure Physic possible ? 
 That they are possible, their actual existence proves. As 
 to Metaphysic, its want of success excites reasonable doubts 
 as to its possibility. Yet as a fact in human nature, a 
 certain spontaneous Metaphysic cannot be denied. The 
 reason is irresistibly impelled to discuss those questions 
 which transcend the bounds of experience, and in this 
 sense there has been since the dawn of speculation, and 
 there will ever be, Metaphysic. The question therefore 
 
 ' Kant did not know Hume's Treatise, in which substance is dis- 
 cussed at length, or he would not have said this. His knowledge of 
 Hume was confined to the Enquiry. 
 
 - But, of course, as Mr. Stirling points out {Text Book, p. 359), the 
 notion of a priori in this reference never occurred to Hume. 
 
 3 Kant says that Hume regarded the propositions of mathematics as 
 analytical {Proleg. p. 28 ; Theory of Ethics, Abbott's Translation, p. 
 142). This, however, is not so certain. The passage Kant seems to 
 have had in his mind is at p. 22, vol. ii. of the Essays, viz. ' Mathe- 
 matical propositions are discoverable by the mere operation of thought, 
 without dependence on what is anywhere existent in the universe.' 
 (See p. 34, supra.) 
 
TRANSCENDENTAL PHILOSOPHY 
 
 remains : How can this impulse be explained from the 
 nature of the reason, or how is spontaneous mctaphysic 
 possible 2 
 
 But as it has confessedly led to perpetual contradictions, 
 we must insist upon the further and last issue : how is 
 Metaphysic as a science possible ? J These are the strictly 
 scientific, and closely defined limits of the Kritik, which is 
 concerned, not with objects, but with reason itself, and its 
 problems. All previous dogmatic attempts at Metaphysic 
 may be completely ignored, as either affording a mere 
 analysis of concepts, which though useful is not Metaphysic, 
 or as consisting of assumptions which have long since 
 become suspicious, owing to the contradictions which they 
 originated. 
 
 § 7. The general Conception and Subdivisions of a special 
 Science, called Kritik of the Pure Reason. — The reason is 
 the faculty which gives us the principles of a priori know- 
 ledge. An organon of the pure reason would then be a 
 summary (Inbegriff) of these principles, and its detailed 
 application would be the system of the pure reason. The 
 present work is a mere preliminary (or propadetitic 2 ) to this 
 system, of negative use, and devoted to clearing and puri- 
 fying our reason from errors on the subject, by means of 
 searching criticism. 
 
 ' I call all knowledge transcendental which is not directly 
 concerned with objects, but with the way in which we cog- 
 nise them, so far as it is possible to do so a priori! A 
 system of such knowledge is properly called transcendental 
 philosophy. Yet even this exceeds our design, as it should 
 contain a complete account of our analytical, as well as our 
 synthetical knowledge; whereas we shall only carry our 
 analysis as far as is absolutely necessary to the understand- 
 
 1 Considered in the Methodology. ' TpoiraiSevru:-/i. 
 
42 SUBDIVISIONS OF THE KRITIK CHAr. 
 
 ing of the principles of the a priori synthesis. This work 
 is then merely a transcendental criticism, or Kritik of the 
 Pure Reason. The main point in subdividing such a science 
 is to admit no concepts that have the smallest empirical 
 element. Thus the principles of morality, though they are 
 by no means based on pleasure and pain, or on desires and 
 inclinations, all of which are empirical in origin, yet imply 
 them necessarily either as obstacles to duty, or incitements 
 to action. They must therefore be excluded. 
 
 Our science must of course contain first Stoicheiology, 1 
 and next Methodology. Each of these will be subdivided 
 according to principles explained in the sequel. One point 
 must here be mentioned : that there are two stems or trunks 
 of human knowledge which perhaps spring from a common, 
 but to us unknown, root, and these are Sensibility 'and 
 Understanding ; through the former of which objects are 
 given to us, through the latter they are thought. So far as the 
 sensibility may contain a priori representations that are the 
 conditions of objects being given to us, so far does it enter 
 into transcendental philosophy. And as objects must be 
 given to us, before they can be thought, this transcendental 
 doctrine of Sensibility, or Aesthetic, must be our first con- 
 sideration. 
 
 If Concluding Reflections on the Introduction 
 
 § 8. The History of Kanfs Discoveries, and his peculiar 
 Method of Proof. — Kuno Fischer has given, in his Com- 
 mentary on the Kritik, 2 a very interesting sketch of the 
 chronological development in Kant's earlier writings. This 
 sketch is particularly valuable, because it shows that the 
 
 1 Or Doctrine of Elements, ffroix«ti. 
 2 Pp. 28-33. Cf, also Kant's Proleg. (Introduction.) 
 
n KANT S PRELIMINARY STEPS 43 
 
 critical philosophy was not adopted by Kant till he had 
 actually supported some of the most popular solutions 
 adopted in the present day. They were tried by him, and 
 found wanting. As early as the year 1762, Kant declared 
 that all logical judgments were analytical and a priori. 
 The following year he contrasted with them the connection 
 of cause and effect, which he declared to be synthetical. 
 He had then discovered that real cognitive judgments, as 
 opposed to logical, were synthetical. But a few years after- 
 wards he declared with Hume, that the concept of cause 
 was obtained empirically. He did not yet perceive how 
 synthetical judgments could be a priori. This is in fact 
 the attitude of Mr. Mill, and his school, who explain the 
 apparent necessity of judgments by association. It cannot 
 be said therefore of Kant, as has been said of Sir William 
 Hamilton and of Dean Mansel, that he ' ignored ' insepar- 
 able association, and did not give that theory his serious 
 consideration. 
 
 But a deeper reflection on mathematical judgments 
 altered his views. Surely these cannot be empirical, and 
 yet they most certainly give us real knowledge. As early 
 as 1764, Kant saw that they depended upon intuition, 
 and he declared space to be that intuition. But still he 
 ascribed to it ' a reality proper to itself,' which lay at the 
 basis of all matter. This was the view of Hamilton and 
 his school. But if space were thus given from without, 
 how could its judgments be anything but empirical, and 
 hence how could they be universal and necessary ? If they 
 are such, space must be an intuition not given with objects 
 from without, but a priori. This step he made between 
 1768 and 1770. 
 
 By maintaining the a priori and yet synthetical character 
 of mathematical judgments Kant parted company with 
 
44 A PRIORI PHYSIC ESTABLISHED chap. 
 
 Hume, and entered upon his critical path. It was obvious 
 when the existence of such judgments was ascertained in 
 one science, that the same problem must be solved in other 
 sciences. How about Metaphysic ? If it means the 
 science of things in themselves, all judgments whatever 
 about such things are rendered impossible by our late dis- 
 covery. For if space and time are necessarily imposed by 
 the mind upon all the objects it can know, how can things 
 apart from these conditions ever be brought before the 
 mind ? How can we speak of things as they are in them- 
 selves, when we only know them under these all-important 
 modifications ? 
 
 There is only one other sense in which a metaphysic of 
 things is possible — in the sense of phenomena. Is there 
 any universal and necessary knowledge of phenomena pos- 
 sible ? Is there such a thing as a priori Physic ? This 
 was the last and by far the most subtle of all Kant's dis- 
 coveries. He would not publish his Kritik, or consider his 
 system complete, till he had ascertained that as we intuite 
 phenomena under a priori conditions, so we also think and 
 connect them under a priori conditions. As the a priori 
 conditions of intuition give us synthetical a priori judgments 
 in Mathematic, so the a priori conditions of thought give us 
 similar judgments in Physic. But we only mean thought 
 about phenomena — thought applied to experience. The cog- 
 nition of the things of sense need not itself be sensuous cognition} 
 Ten years of thought brought him to this conclusion. The 
 critical philosophy therefore, like most great discoveries, was 
 
 1 See Kritik, p. 238. There is a curious analogy in Descartes's 
 great discovery — the application of mathematics to nature, which con- 
 sisted of two steps : first, the solution of figure-problems by algebraic 
 symbols ; secondly, the principle that all nature is reducible to figures, 
 i.e. modes of extension. Cf. 'Descartes' in Blackwood' 's Philosophical 
 Series. 
 
n KANT S TRANSCENDENTAL PROOFS 45 
 
 not the offspring of a happy guess, or a sudden inspiration, 
 but the slow and gradual result of a long life of labour. 
 
 We are left in no darkness as to all these points. Not 
 only the chronological sequence of Kant's works but the 
 general account of his discoveries given in the second Pre- 
 face, and in his Prolegomena, are explicit. In this latter 
 work, published for the use of teachers in expounding the 
 Critical Philosophy, he gives the analytical or regressive 
 view of the system synthetically constructed in the Kritik. 
 We have endeavoured to combine both in the preceding 
 Commentary. One point, however, deserves special atten- 
 tion, before we enter upon our task. The nature of Kant's 
 demonstration throughout the Kritik may appear at first 
 sight illogical, inasmuch as he argues from the position of 
 the consequent to the position of the antecedent, and this 
 he calls his transcendental proofs (p. 478). But this argu- 
 ment is only illogical on account of the plurality of causes. 
 Given a cause, its effect will follow ; but given the effect, 
 we cannot infer the particular cause, except we are certain 
 that no other cause could have produced the effect. We 
 may safely argue from the effect to its only possible cause. 
 And such is Kant's investigation, which infers from the fact 
 of cognition the only possible conditions under which it 
 could exist. When these conditions are established, they 
 show not the existence of the fact, by which they were 
 themselves proved, but its legitimacy. Thus the legitimacy 
 of Mathematic and Physic, and the illegitimacy of Meta- 
 physic, as a science of things per se, are demonstrated from 
 the conditions they involve. It may be objected that if 
 Metaphysic be a fact, its conditions must be as real as those 
 of any other science : how then can it be rejected ? In 
 answer it may be observed, firstly, that the conditions of 
 
 1 Cf. Kuno Fischer, pp. 24-28. 
 
46 SUBDIVISIONS OF THE KRITIK chap, ii 
 
 Metaphysic are absolutely inconsistent with those of mathe- 
 matic. This raises a strong presumption against the more 
 doubtful science. Secondly, if it be found that from the 
 conditions of Mathematic and Physic the possibility of an 
 illusory science of Metaphysic can be explained, whereas 
 from those of Metaphysic the very existence of Mathematic 
 and Physic can be shown impossible — in such a sense, we 
 cannot hesitate as to our decision. The Aesthetic shows 
 the legitimacy of Mathematic, the Analytic that of Physic; 
 the Dialectic proves the illegitimacy, as well as the apparent 
 existence of Metaphysic. These are the main divisions of 
 the Kritik. 
 
CHAPTER III 
 
 THE TRANSCENDENTAL AESTHETIC 
 
 § I. [Definitions.] — The immediate knowledge we have of 
 objects is intuition. This only occurs if an object is given 
 us — that is to say, if it produces an affection or modification 
 of our minds. The faculty of obtaining representations 
 through this affection produced by objects, is sensibility. 
 Kant notices that this so-called faculty is properly a recep- 
 tivity, as opposed to the spontaneity of thought. Sensibility 
 alone gives us intuitions : when these are thought by the 
 understanding, we obtain concepts. All thinking must 
 refer either mediately or immediately to sensibility. In no 
 other way can an object be given us. The effect of an 
 object on our sensibility, so far as we are affected by it, is 
 sensation. So far as an intuition is of this sort, it is empirical. 
 ' The undetermined object of an empirical intuition is called 
 an appearance, or phenomenon.' The element in it that 
 corresponds to sensation is the matter; the element that 
 renders its variety reducible to fixed relations is the form 
 of the phenomenon. The form differs completely from 
 the matter in this, that while the matter is given a posteriori, 
 the form exists a priori, as it were, ready in the mind, and 
 can therefore be considered quite separately. 1 Representa- ■ 
 1 Kant says expressly (Kritik, p. 98, note) that there are elements 
 
THE AESTHETIC 
 
 tions are therefore pure, in the transcendental sense, if 
 nothing belonging to sensation is to be found in them, 
 and the pure form of intuition, that reduces the variety 
 of phenomena to order, is in the mind, and may be 
 called pure intuition?- Abstract from a body what the 
 understanding thinks about it, such as substance, force, 
 and divisibility ; abstract also what belongs to sensation, 
 such as incompressibility, hardness, colour, etc., and there, 
 still remain extension and figure. These belong to the 
 pure a priori intuition which exists in the mind, even 
 without an actual object of sense, as a mere form of 
 sensibility. The science of all a priori principles of 
 sensibility may be called transcendental aesthetic, using this 
 word not in the sense of the philosophy of taste, but in 
 the Greek sense of cucr^o-is, as opposed to vo?;o-is. Our 
 investigation shows that there are two such forms, space 
 and time, which we proceed to consider. 
 
 §§ 2 and 4. Metaphysical Exposition of Space and Time. — 
 Although Kant considers the two forms separately, we may 
 combine them, so far as the same observations apply strictly 
 to both. This is the case with the metaphysical exposition, 2 
 defined by Kant, as ' containing the description of a notion, 
 as given a priori. ' 
 
 By means of our external sense, which is a property of 
 
 not given by sensibility included in his Aesthetic, so difficult is it to 
 separate in treatment what is one in nature. But many of his critics 
 have assumed, in direct opposition to this statement, that Kant made a 
 sharp separation between the two faculties in their actual use, and have 
 proceeded to charge him with want of psychological insight. Cf. also 
 Kritik, p. 84, note. 
 
 1 It is very misleading to find Mr. Stirling telling us (p. 352, Joe. cit.) 
 that Anschauung by itself means pure intuition. Surely Kant's lan- 
 guage is plain enough ; and Mr. Stirling himself admits on p. 358 that 
 Anschauung is also used of empirical intuition. 
 
 - Cf. Kritik, p. 444. 
 
SPACE AND TIME 49 
 
 our minds, we represent objects as without us, and their 
 form and relations are determined in what we call Space. 
 The internal sense, by means of which the mind intuites itself 
 or its own internal states, gives us frideed no intuition of 
 the soul itself, as an object, but has nevertheless a form, viz. 
 Time, without which no such internal experience is possible. 
 What then are Space and Time ? Are they actual beings ? 
 [This is the popular belief.] Or are they mere relations, 
 that belong however to things in themselves, whether we 
 intuite them or not ? [This was the view of many previous 
 philosophers, e.g. Leibnitz and Wolf.] Or do they belong 
 merely to the subjective nature of our mind, as forms of its 
 intuition, through which alone they can be added to things ? 
 U Kant does not consider the possibility of their coming 
 under more than one of these three heads. For in the first 
 place, if any one of these suppositions satisfies all the 
 phenomena, the philosophical law of Parcimony forbids us 
 to assume an additional one without a?iy reason at all. 
 Secondly, if the subjective origin of space and time be 
 established, it is specially absurd to assume that the peculiar 
 element added to objects by the mind, which constitutes in 
 fact the essential difference between the phenomenon and 
 the thing per se — that this element is added by the mind to 
 objects which have it already in themselves. Recent contro- 
 versies will make it necessary to revert to this subject, when 
 we have considered Kant's express utterances on the point. 
 
 (1) Space and time are not empirical concepts, deduced 
 from our experience. For we cannot refer our sensations to 
 anything without us (that is, in space), nor can we assert 
 them to be simultaneous or successive, unless the representa- 
 tions of space and time were already in the mind. 
 
 (2) Space and time are necessary a priori representations, 
 lying at the basis of external and internal intuitions respect- 
 
 1 E 
 
50 SPACE AND TIME 
 
 ively. It is impossible to conceive either of them annihi- 
 lated, though we can easily conceive all objects in them 
 removed. They are therefore the necessary conditions of 
 this very possibility of phenomena. 1 
 
 (3) Space and time are not general concepts of the 
 relations of things, but pure intuitions. For there is but 
 one space and one time, of which all separate spaces and 
 times are parts. 2 And these parts are not considered as 
 constituent elements, composing space and time, and there- 
 fore prior to them, but rather as limitations of space and 
 time, and existing in them. Hence an a priori intuition 
 lies at the base of all our notions of space and time. 
 
 (4) Space and time are represented as unlimited quan- 
 tities. 3 [And the infinite parts which they contain are 
 neither constituent elements chemically fused to produce 
 them, nor logical parts contained under them. J Every 
 concept indeed is represented as comprising a possibly 
 
 1 (1) shows that Space and Time are a priori and not empirical, i.e. 
 not products of special sense ; (2), shows that they are not only a priori, 
 but that they are the necessary conditions of all intuitions. 
 
 - Cf. note, Kritik, p. 98 and p. 274. We now go on to show that 
 they are intuitions, not concepts, from the peculiar nature of their unity 
 in (3), of their infinity in (4) ; they being integral not universal wholes. 
 
 3 Kant says of space that it is represented as an infinite given quantity, 
 a statement justly attacked by his German critics (especially Montgomery), 
 who deny that on his own principles infinity can be given to sense, or 
 indeed given at all (cf. Kritik, p. 270). Possibly he meant to say 
 unlimited, for he is far more cautious in the parallel remark on time, 
 where he says the ' infinity of time means but this, that all definite 
 quantities of time are only possible by limiting the single (total) time 
 lying at their basis ' (cf. also Kritik, pp. 257, 262). Hence time is ori- 
 ginally given as unlimited. A knowledge of the absence of limits may 
 be given, though positive infinity cannot. Kant's opponents should 
 have given him the benefit of this reasonable explanation. Even as to 
 space, he spoke of its infinity, in the first Edition, as the absence of 
 bounds in the extension of intuition (die Grenzenlosigkeit im Fortgange 
 der Anschauung). 
 
in THE TRANSCENDENTAL EXPOSITION 51 
 
 infinite number of individuals under itself, but it cannot be 
 conceived as containing them within itself. The parts of 
 space and time are in infinite space and time, not contained 
 under our concepts of them. Our original representation 
 of space and time is therefore not concept, but intuition. 
 [Kant had added, in his first Edition, that no general con- 
 cept of relations (in space) could of itself imply the endless- 
 ness of these relations, as our notion of space certainly 
 does.] 
 
 §§ 3 and 5. Transcendental Exposition of Space and 
 Time. — Kant defines this to be ' the explanation of a 
 concept as a principle from which the possibility of other 
 synthetical a priori cognitions can be understood.' It is 
 necessary to show (a) that such cognitions actually flow 
 from the given concept ; (b) that these cognitions are only 
 possible by presupposing this particular explanation of the 
 concept. 
 
 Geometry is a science that determines the properties of 
 space synthetically, and yet a priori. What then must our 
 representation of space be, to produce such cognitions ? It 
 must obviously be intuitive, for from mere concepts we can- 
 not obtain synthetical propositions ; and also a priori, for 
 these propositions are apodeictic, and carry with them 
 necessity, such as the statement that space has but three 
 dimensions. 1 So also there are axioms concerning the 
 relation of time equally demonstrable and necessary : e.g. 
 Time has but one dimension, different times cannot be 
 simultaneous (as different spaces are), but successive. So 
 also the concept of Change, and with it of Motion (change 
 
 1 Briefly, the propositions in geometry are synthetical — therefore 
 space is an intuition ; they are necessary — therefore space is a priori. 
 Mr. Watson puts the argument thus : ' As specifications of the forms of 
 intuition mathematical judgments are a priori, and as specifications of 
 those forms they are synthetical.' {Kant and liis English Critics, p. 41.) 
 
52 THE BASIS OF ARITHMETIC chap. 
 
 of place), are only possible through our notion of time ; and 
 through this latter only as an a priori intuition, for no con- 
 cept could possibly make us understand the possibility of 
 change, which is, in fact, the combination of contradictory 
 predicates — the existence and non-existence of the same 
 thing in the same place. It is only in time — that is to say 
 successively — that this is possible. All the synthetical pro- 
 positions, therefore, derived from our idea of motion in 
 general (and they are not a few) are wholly dependent on 
 our idea of time. 
 
 11 The expositors of Kant 1 have uniformly derived the 
 science of Arithmetic from the intuition of Time — a deri- 
 vation so important, if true, that he could not possibly 
 have omitted to mention it. If Arithmetic be based on 
 Time, it would be natural to make use of the propositions 
 of Arithmetic to prove that Time is an intuition, in the 
 Transcendental Exposition. But although in his Prole- 
 gomena? he has (perhaps in deference to his critics, and 
 seduced by his passion for symmetry) conceded en passant, 
 that this view is possible, he has left us in no doubt, from 
 several passages in his Kritik, that the units of the science 
 of arithmetic, being essentially simultaneous, and not 
 successive units, are given us primarily in space, and not 
 in time. The original intuition of 5, for example, is not a 
 group of five successive thoughts or intuitions, but the 
 immediate perception, through sight, and perhaps through 
 touch also, of five simultaneous, adjacent units. The 
 fact that we can, if we choose, apprehend them successively 
 in five separate acts of attention, makes us apply arithmet- 
 ical laws to sensations in time also ; but we do not see 
 how a summation, or subsumption of several units under a 
 
 1 E.g. K. Fischer, Dean Mansel, Sir Wm. Hamilton, Ueberwe? 
 and many others. 2 p , _ 
 
m THE BASIS OF ARITHMETIC 53 
 
 higher number, regarded as a unit itself — how this would 
 ever have been accomplished, were we not aware intuitively 
 of the simultaneous presence of the units within a small 
 definite portion of space. There is no other practical way 
 of teaching arithmetic to a child or savage, than by appeal- 
 ing to space intuitions. Let us add that the subdivision 
 of units into fractions is equally unattainable, originally, 
 through intuition of time, but is easily obtained through 
 space, where all the units assumed are intuitively divisible. 
 It has been pointed out in another work 1 how the opinion 
 appears to have arisen. Our only way of exemplifying 
 quantity in time is by the act of adding (mentally) units to 
 
 1 Cf. Fischer's Commentary on Kant' s Kritik, p. 95, note. Dr. Tarle- 
 ton has an able paper in Hermathena No. I., in which he develops a 
 positive theory of the growth of arithmetic from subdivisions of Time. 
 As he observes, we can conceive a mind ignorant of space distinguishing 
 units and their addition by the striking of a clock. And if any given 
 number of strokes were gathered, he thinks, by means of some longer 
 division of time under a larger unity, an acute reasoner might even 
 under such conditions construct an arithmetic. This is, we believe, the 
 only attempt yet made ;to meet the difficulty in a full and scientific 
 manner. We are not disposed to deny the possibility of arithmetic being 
 so obtained in the absence of spatial data, as, for instance, in blindness 
 or perpetual darkness ; but we are certain that, as a matter of fact, 
 ordinary human beings do not learn it so, and that Kant felt this, 
 and intended to derive our ordinary notions of number from space, 
 though he admitted the other. The apprehension of units may no 
 doubt be successive, and so involve time, but we deny that in the early 
 counting of small numbers, which the eye can take in simultaneously, 
 this succession is a conscious element, and except it be explicitly present 
 in consciousness, it has of course no claim to be regarded as the basis 
 of arithmetic. We must add that Dr. Tarleton has supplied solutions to 
 some of the difficulties which have been raised, but surely his derivations 
 are too subtle for ordinary use. In any case it is doubtful whether 
 similar representations repeated in time would have given the same 
 notion of different units, which we derive from co-ordinated units in 
 space. The units in time may be the same thing reproduced; this 
 cannot be the case with separate units in space. 
 
54 DEDUCTIONS 
 
 one another, in other words by number. But this schema 
 of quantity, which will be discussed in its proper place, is 
 expressly contrasted by Kant^ with any representation which 
 can be reduced to an image. The derivation of arithmetic 
 from the intuition of space has never yet, so far as we know, 
 been refuted, and though not definitely stated hitherto, 
 was distinctly implied as far back as Descartes. 2 
 
 § 6. Deductions from the preceding Notions. — {a and b) 
 Neither time nor space represents any properties of things 
 in themselves, nor do they express any relations of such 
 things to one another which still exist, if we abstract from 
 things the subjective conditions of intuition. For in no 
 case can properties, absolute or relative, be intuited a 
 priori, prior to the existence of the things which they deter- 
 mine. To assert then the absolute reality of space and 
 time is to assert that we know a priori the properties or 
 relations of things which we do not know. Such an asser- 
 tion has no meaning whatever, unless we assume that 
 space and time belong to the conditions imposed on the 
 mind, and are, therefore, logically prior to the cognition of 
 objects. 3 Upon this supposition the necessary and universal 
 judgments obtained from the intuitions of space and time 
 follow as a matter of course. Also our external image of 
 time is an endless right line, which expresses all its relations 
 in an intuition, and, therefore, proves it to be such. 
 
 It is then only from the human point of view that we 
 can speak of extended beings, or of events in time. If we 
 abandon the subjective conditions, under which alone we 
 
 1 Kritih, p. no. 
 
 2 See his Highs pour la Direction de P 'Esprit, xiv. and xv. 
 
 3 Space is nothing more than the ' formal capacity of the subject to 
 be affected by objects' {Kritik, p. 25). Time is ' the mode in which 
 the mind is affected by its own activity' (p. 41). 
 
RESULTS 55 
 
 can be affected by objects, then both space and time have 
 absolutely no meaning. These forms are a necessary part 
 of every intuition which we can have, because we intuite 
 through them. But we may not impose the conditions 
 which limit our cognition of things on things themselves. 
 Space and time then include all things which appear to us, 
 but beyond this we can give neither of them any reality. 
 For as they are not things, but our particular way of look- 
 ing at things, it is absurd to imagine that they can belong 
 to things per se as qualities. Again, as we cannot possibly 
 investigate the conditions which limit the intuition of other 
 beings than ourselves, we cannot say whether any of them 
 are obliged to look at things as we do. We are, therefore, 
 incompetent to affirm the existence of space and time, even 
 in this sense, which is the only possible sense in which they 
 can be conceived beyond our own experience without 
 absurdity. But when thus limited, nothing can be more 
 objectively certain. There is no possible intuition which 
 we can have apart from space and time. We hold, there- 
 fore, as strongly as possible, their empirical reality, but 
 assert their transcendental ideality, — that is, that they are 
 nothing, if we omit the conditions of experience, and regard 
 them as belonging to things in themselves. [We also deny 
 their absolute reality, which can never be revealed to us, 
 and which is in some senses absurd. 1 ] 
 
 1 When Trendelenburg says that Kant forgot to inquire whether 
 space and time might not be both subjective and objective, as a third 
 possibility, he shows by his very statement of the question his ignorance 
 of Kant's system. To say, as Trendelenburg does {Beitra^e, pp. 215 
 sq.), that Kant made them purely subjective is equally false. Kant 
 would never have conceded such a statement of his views. He denied 
 the subjectivity in the usual sense of space and time. He asserted 
 them expressly, in the sense which Trendelenburg desires, to be both 
 subjective and objective. He never denied their objectivity except in 
 an absurd sense. 
 
56 SPACE AND TIME OBJECTIVE chap. 
 
 Space is the only one of our subjective representations 
 relating to what is without us which can be called objective 
 a priori. For synthetical a priori propositions cannot be 
 deduced from any other of them, so that such sensations 
 as heat, colour, and sound can claim no ideality at all, 
 accurately speaking. 1 They agree, indeed, with space in 
 belonging to the subjective nature of sensibility, but being 
 sensations and not intuitions, they in themselves give no 
 object, not to say an object a priori. 
 
 (<r) Time is the formal condition of all phenomena 
 generally. Space is confined to those which are external. 2 
 But all representations, whatever be their object, must be- 
 long, as affections of the mind, to our inner states ; and as 
 these inner states are subject to the formal condition of 
 time immediately, even external phenomena are mediately 
 subject to the same condition. There is therefore no part 
 of our experience free from the condition of time. 
 
 Kant cautions the reader repeatedly not to attempt to 
 illustrate his transcendental doctrine of space and time by 
 comparing them to the subjective affections which objects 
 excite in us, and which are known as secondary qualities. 
 It might be thought that as heat, colour, and taste are not 
 attributes of objects, but affections excited in us, and wholly 
 subjective, so space and time are also subjective affections 
 added by us to objects to which they really do not belong. 
 
 1 He means that we cannot conceive a mere sensation as having even 
 an ideal existence separate from our thinking. To afford us this possi- 
 bility, we must conceive objects as distinct from mere affections of the 
 senses. 
 
 2 Mr. Hodgson shows {Space and Time, p. n6) that Kant need not 
 have denied to space its universality even as a condition of internal ex- 
 perience. For as external phenomena come secondarily (as thoughts) 
 under the form of Time, so internal states come indirectly under space, 
 being always localised within our bodies. It would appear indeed that 
 Kant felt this in his Refutation of Idealism. 
 
HI SPACE AND TIME OBJECTIVE 57 
 
 Such an illustration would mislead the student, for it is 
 based on the contrast between the reality of the object, and 
 the subjectivity of the affection, which may vary in different 
 minds. The rose is more real than its colour or fragrance : 
 the wine is more real than its flavour. These qualities are 
 neither universal nor necessary. But space and time are 
 as universal and necessary as anything in our experience 
 can be ; they are as real and as objective as any part of our 
 experience. To contrast them with the objects to which 
 they belong, would be to lapse into the old error of regard- 
 ing these objects as things per se. Concerning these latter, 
 experience knows and asks nothing ; they are perfectly un- 
 known correlatives of the phenomena which appear to us 
 in space and time only. These latter are in no wise sen- 
 sations, produced by an affection of our sensibility, but pure 
 forms of intuition, perceived by us as figures and relations. 
 § 7. Further Explanations. — It was objected almost 
 unanimously, by intelligent critics of Kant's First Edition, 
 that changes were actual, as is proved by our internal ex- 
 perience, even abandoning the evidence of our external 
 experience. But changes being only possible in time, time 
 must be actual. To this Kant replies by admitting the 
 whole argument, which is perfectly true, but also perfectly 
 irrelevant. Kant never denied the actuality of Time, as a 
 part of oicr experience. But suppose that some other being 
 could behold our minds with an intuition free from the 
 limitations by which ours are bound, what we call changes 
 would produce a cognition in him in which the representa- 
 tion of time, and therefore of change, would form no part. 
 In other words the empirical reality of time is asserted by 
 both Kant and his opponents ; but they have sought to 
 infer from it absolute reality, which cannot be conceded on 
 his principles. 
 
58 OBJECTIONS RETORTED chap. 
 
 The objectors did not press the same argument in the 
 case of space, [because idealism had long since proved 
 that the actuality of external objects was not demonstrable, 
 whereas the objects of the internal sense are proved actual 
 by consciousness. They did not consider, that without dis- 
 puting for a moment the actuality of either as representa- 
 tions, both only belong to [the genus] phenomenon, which 
 has always two sides : (i) in regard to which the object is 
 viewed as in itself, apart from the manner of its intuition 
 (in which case its nature is always problematical), and (2) 
 where the form of the intuition is considered. This, though 
 actually and necessarily belonging to the phenomenal object, 
 must be sought in the subject. 
 
 The objection is in fact retorted with great force by 
 Kant. For while this theory supports the reality of all our 
 empirical cognition, those who hold the absolute reality of 
 space and time are at variance with the principles of experi- 
 ence. They must either regard space and time as subsist- 
 ing independently of things (the prevalent theory of mathe- 
 matical physicists), and so assume two eternal and infinite 
 self-subsisting nonentities [Undinge] 1 which exist merely for 
 the purpose of containing all that is actual ; or they must re- 
 gard them, with some metaphysical philosophers, as relations 
 of phenomena, abstracted from experience, and confused in 
 the process. In this latter case they must deny the validity, 
 or at least the apodeictic certainty of the mathematical a 
 priori deductions which concern actual things in space, for 
 such certainty cannot be obtained a posteriori ; and space 
 and time are, on this hypothesis, the creatures of the im- 
 agination, abstracted from experience, generalising its rela- 
 tions, but for that very reason depending for application on 
 
 1 They are Undinge — that is, unlike sensations, they are not referred 
 to objects other than themselves. — Cf. Stirling, loc. cit. p. 377. 
 
in FALSE EXTENSIONS OF THE THEORY 59 
 
 the restrictions which nature has imposed. The former 
 school save their mathematical conclusions, but create for 
 themselves endless difficulties when they leave this (pheno- 
 menal) field. The latter have the advantage of not being 
 hampered by space and time, when they desire to consider 
 objects not as phenomena, but in relation to thought ; they 
 cannot, however, explain the possibility of a prioii mathe- 
 matical cognitions, or bring the laws of experience into 
 necessary harmony with them. Kant's theory solves all 
 these difficulties. 
 
 Finally, the transcendental Aesthetic cannot contain more 
 than these two elements, space and time. For every other 
 notion pertaining to our sensibility, even that of motion, which 
 combines them both, presupposes something empirical. 
 Space itself does not move, but something {empirically given) in 
 space. Time changes not, but something perceived in time. 1 
 
 IT This remark refutes the attempts which have been 
 made to construct an a priori Mechanic on the presenta- 
 tion of Force.- We cannot perceive or intuite dynamical 
 force, without motion, and for this empirical data are re- 
 quired, which will destroy the a priori purity of the deduc- 
 tions from such intuition. When the late Dean Mansel in 
 another work (his well-known Bampton Lectures) endeavours 
 to deduce moral laws from the datum of ^Personality, he is 
 guilty (we conceive) of a different error, for the notion of 
 Personality, though a priori, is not an intuition at all, but, 
 as will be seen in the latter part of this Commentary, a 
 logical supposition of thought, giving us no definite know- 
 ledge or basis for further deductions such as are found in 
 pure Mathematic. 
 
 § 8. General Remarks on the Transcendental Aesthetic— 
 
 1 Cf. Kritik, p. 95 note, and Proleg. p. 45- 
 2 See Mansel's Prolegomena Logica, Appendix A. 
 
6o RECAPITULATION 
 
 I. Kant here again sums up his doctrine, insisting that 
 space and time are not necessary even to [rational] beings 
 as such, but to the subjective sensibility of all human beings. 
 They are its pure form, and therefore cognoscible apart from 
 the sensation, which is the matter given a posteriori, and 
 infinitely various. The form being a necessary part of all 
 our intuitions, no possible analysis or increased clearness in 
 them can bring us in the least nearer to things per se. To 
 say then that our sensibility is only a confused representation 
 of things, containing what really belongs to them as such, 
 but under a congeries of attributes and partial representa- 
 tions, which we do not consciously explicate — to say this is 
 to falsify the whole notion of sensibility and of phenomenon, 
 and make it idle and void. The contrast between distinct 
 and indistinct representations is merely logical, and does 
 not concern their content No doubt such a concept as 
 that of right is the same in sound common sense, and when 
 analysed by the subtlest speculation ; the latter merely de- 
 velops what is unconsciously felt in the former. ' But the 
 common notion of right is not therefore sensuous, or a 
 mere phenomenon, for in any case it is a concept of the 
 understanding, and represents a property of actions which 
 belongs to them as such. The representation of a body in 
 intuition, on the* contrary, contains nothing at all which 
 could belong to an object per se, but only the phenomenon 
 of something ; and the way in which we are affected by it, 
 and this receptivity of our cognitive faculty, which is called 
 sensibility, must ever differ toto mlo from a cognition of 
 the objector se, however thoroughly we may penetrate the 
 phenomenon to its deepest elements.' The logical distinc- 
 tion therefore- drawn by the school of Wolf and Leibnitz 
 between the two divisions of our knowledge 1 is false, the 
 1 That is, between the Sensibility and the Intellect, which latter they 
 
in ONLY PHENOMENA PERCEPTIBLE 61 
 
 distinction is transcendental, and concerns their origin and 
 content. Abstract from objects our subjective constitution, 
 and they must disappear with the qualities which this very 
 constitution gave them, thereby determining their form as 
 phenomena. 
 
 We distinguish indeed in phenomena that which affects 
 the senses of all men in the same way, and that which 
 depends on the peculiar organisation of isolated individuals. 
 We say commonly that the former is the object in itself, 
 the latter only appears to be such. But this is merely an 
 empirical distinction. We must go further and regard even 
 the object as it appears to all mankind, as a mere pheno- 
 menon, in which no property at all of a thing per se is 
 to be found ; otherwise our trandescendal distinction (just 
 now made) is lost. 
 
 In the case, for example, of a rainbow, we say physically 
 that it is a mere appearance, while the rain is the reality J 
 itself (die Sache an sich selbst). But if we inquire further 
 into the question whether this thing which is real to the 
 senses of all men, also represents a thing per se beyond 
 these senses, 'then the question of the relation of the 
 representation to its object becomes transcendental, and 
 not merely the drops of rain, but their round form, and 
 even the space through which they fall, are mere modifica- 
 tions or conditions of our sensibility; the transcendental 
 object remains totally unknown.' 
 
 ' The second point of importance in the transcendental 
 Aesthetic is this, that it should not find favour merely as 
 a plausible hypothesis, but claim to be as certain and un- 
 regarded as capable of knowing things per se more distinctly. Kant does 
 not dispute this latter point yet, as it here suffices for his argument to 
 establish the true character of sensibility. 
 
 1 Not, as Mr. Stirling renders it, the thing in itself (Ding-an-sich). 
 
62 THE ARGUMENT FROM MATHEMATIC chap. 
 
 doubted as can possibly be demanded from a theory, which 
 is to serve as organon,' or basis of a scientific system. Let 
 us illustrate this by an example. Suppose that space and 
 time were objective/!?;' se, and conditions of the possibility 
 of things in themselves. It is a fact that from them 
 (especially from space) a number of propositions demon- 
 strable a priori, and synthetical, are derived. Whence can 
 these geometrical truths, with this twofold character, be 
 obtained ? There is no way possible, except either through 
 intuitions, or through concepts, and these either a priori or 
 a posteriori. All a posteriori knowledge of either kind is at 
 once excluded, because mathematical judgments are strictly 
 necessary, a feature which no experience can supply. All 
 concepts are also excluded, for from them we can only 
 obtain analytical, and not synthetical judgments. Torture 
 the concepts of right lines, and of the number two as you 
 please, you will never obtain from them the proposition : 
 Two right lines cannot enclose a space. We must therefore 
 have recourse to intuition, and intuition a priori, as just 
 established. 
 
 But if space and time were given to us as properties of 
 things apart from our way of looking at things (our sensi- 
 bility), how could we possibly know anything about these 
 things a priori 1 Or granting, as we must do, that this a 
 priori knowledge of them must come from our subjective 
 conditions, how could we assert it to belong to a triangular 
 object as such ? Both these alternatives are impossible. 
 From the fact that these qualities belong to all objects 
 a priori, they must come from our sensibility ; from the 
 fact that propositions concerning them can be uncondition- 
 ally asserted — that we can assert for example that the 
 mathematical properties of a triangle belong to any triangular 
 object in nature — from this we must infer that it can be 
 
Ill RELATIVITY OF TIME 63 
 
 nothing beyond our sensibility. It is therefore not prob- 
 able, but indubitably certain, that space and time are 
 nothing but the necessary, though merely subjective condi- 
 tions of our intuitive experience — for this reason universal 
 and necessary, but relating only to experience, not at all to 
 things in themselves. 
 
 II. In his Second Edition Kant adds the following 
 supplementary reflections. 
 
 To corroborate this theory of the ideality of both ex- 
 ternal and internal senses, and therefore of all the objects 
 of our senses, as mere phenomena, the following remark 
 may be of service. Whatever in our cognition belongs to 
 intuition (the emotions of pleasure and pain, and the will, 
 are not cognitions) is nothing but relations, whether they 
 be of extension, motion, or moving forces — all of which are 
 relations or changes of place, or else the laws that determine 
 them. Now relations tell us nothing of a thing absolutely, 
 it might therefore be fairly inferred that our external sense 
 merely gives us the relation of an object to the subject. 
 The same may be said of our internal intuition. For, in 
 the first place, the representations of our external senses 
 are the matter which supplies our minds. Secondly, time, 
 which is the formal condition of all our experience, and is 
 logically prior to it, existing in the mind as a form of in- 
 tuition — time contains relations of sequence, simultaneity, 
 and permanence. Being prior to any thought of an object, 
 it must be our intuiting ; and if it contains nothing but rela- 
 tions, it must be the form of this intuiting, or the way in 
 which the mind, when acting, is affected by its own activity. 
 Whatever is represented to us through a sense must be a 
 phenomenon ; there must therefore either be no internal 
 sense, or else our mind or subject, when it is the object of 
 this sense, must be given to us merely as a phenomenon 
 
64 FALSE EXTENSIONS chap. 
 
 and not as it would judge itself, were its intuition self- 
 acting, and therefore purely intellectual. The great diffi- 
 culty is to explain how a subject can have internal intuition 
 of itself, but this difficulty is common to every theory. Our 
 self-consciousness (apperception) is the simple representa- 
 tion of the Ego, and cannot by a pure act of spontaneity 
 cognise all the variety which exists in the subject. It must 
 apprehend this variety by permitting its internal intuiting 
 faculty to be affected, and then this faculty must receive 
 the variety in time, ordering it according to the laws and 
 conditions under which alone the mind can act. Our 
 faculty of internal intuition is not a spontaneity but a 
 receptivity, and does not therefore present to us the Ego in 
 any way more distinctly than our external intuition presents 
 to us foreign bodies ; we know both only as phenomena. 
 
 IT The late Dean Mansel, failing to apprehend the force 
 of this remark, was led to distinguish between the data of 
 our external and our internal sense, and to ascribe to the 
 latter the first origin of our ideas of substance. Descartes 
 long ago set up our internal experience as more trustworthy 
 than our external, but he ascribed our knowledge of self 
 not to intuition, but to thought. This theory is more 
 reasonable than that of Mansel, for unless our intuition be 
 made intellectual, like Schelling's, it must act under the 
 condition of time (as well as of the categories, as we shall 
 see), and therefore adulterates the object, so as to prevent 
 us from knowing it per se. It will be seen in the sequel that 
 this is not the only conclusive objection to the theory. 
 
 III. This critical view of space and time does not by 
 any means reduce the objects given in them to mere illu- 
 sion — the charge, of all others, most warmly denied by 
 Kant. He calls it, in his Prolegomena, 'an objection 
 arising from an unpardonable, and, he would almost say, 
 
WHAT IS ILLUSION I 65 
 
 intentional misconception.' 1 In phenomena both the ob- 
 jects and even the qualities we add to them are regarded as 
 really given ; but as these latter qualities depend upon a 
 particular relation, we contrast things per se with such 
 phenomena, without in the least denying the reality of the 
 latter. Kant never said that bodies only seemed to exist 
 without us, or that our minds seemed to be present in con- 
 sciousness, when he asserted that space and time, the 
 conditions (to us) of knowing them, are in the subject 
 only. Phenomenon is not illusion. It is not asserting an 
 illusion to say that a rose appears to be red, or scented. 
 But if we think that Saturn has handles, as was once 
 believed, we are subject to an illusion, which consists in 
 attributing to the object per se what belongs to it merely 
 in relation to our senses. If I said that a rose was red 
 per se, and not merely that it appears to us red, then I 
 am subject to an illusion like that concerning Saturn. Just 
 so if I assert that space is a property of objects per se, I 
 attach to them what they possess only in relation to me. 
 The common theory therefore is actually that which turns 
 reality into illusion. For if we regard space and time as 
 conditions of things in themselves, and consider the ab- 
 surdities that follow : how two infinite things, that are 
 neither substances nor attributes, exist and are the necessary 
 conditions of all things, and remain even when things are 
 removed— considering all this, we cannot blame Berkeley 
 for degrading bodies to mere illusion ; nay even our own 
 existence, if depending on the self-subsistent reality of a 
 nonentity, time, might come to be regarded an illusion — an 
 
 1 Cf. vol. ii. p. 67. It is melancholy to see a man of Trendelenburg's 
 position clinging to these foolish objections, and followed by second- 
 rate English writers. As Stewart thought the Berkeleian idealism a 
 test of metaphysical acumen among the Scotchmen of his day, so 
 Kantian idealism might serve us at present. 
 I F 
 
66 TWO KINDS OF INTUITION chap. 
 
 absurdity with which no one has as yet ventured to identify 
 himself. 
 
 IV. What notion do we form of the cognitive faculties 
 of the Deity in our natural theology? His knowledge 
 cannot be a process of thinking, which always must imply 
 limits ; it must be intuiting. And as we cannot regard the 
 Divine intuition as sensuous, we are careful not to attribute 
 to it the limitations of space and time. But how could we 
 avoid doing this, if they were the conditions of the existence 
 of things a priori, and existing independently of them ? 
 And in such case they must be the conditions of God's 
 existence also. It remains for us to make them the sub- 
 jective forms of our intuition, called sensuous, because it is 
 not primitive, but a faculty of the subject depending on the 
 presence of the object, and affected by it. Primitive in- 
 tuition, 1 which belongs, as far as we can see, to the Prime 
 Being only, is that which gives of itself the existence 
 (Dasein) of the object of intuition, without depending upon 
 such affection. 2 
 
 It is possible, that not only men, but all finite intelli- 
 gences, have their intuition thus limited. Even if this be 
 the case, it is not the less sensibility, because it is not a 
 primitive (intuitu* origi nanus), not therefore an intellectual 
 intuition which belongs perhaps to the First Being only, 
 but a derived (derivativus) intuition, belonging to a being 
 dependent both in its existence and in its intuition. 
 
 This latter remark is to be considered an illustration of 
 the Aesthetic, not an argument in support of it. 
 
 1 Cf. Theory of Ethics, p. 235. 
 
 2 In other words, a primitive intuition would perceive the object as 
 existing, whereas we only perceive it as affecting us. Our intuition is 
 therefore dependent on the present existence and action of the object ; 
 that of the Deity is conceived as not so dependent, but as perceiving 
 the object directly, and without waiting to be affected by it. 
 
in AN INDIRECT PROOF 67 
 
 Conclusion of the Transcendental Aesthetic. — One part of 
 the problem of the transcendental philosophy is now settled, 
 which problem is this : how are synthetical a priori judg- 
 ments possible ? We have discovered pure a priori intuitions, 
 space and time, in which, when we wish to enlarge a given 
 concept a priori, we discover a priori what is not given in 
 the concept, and join these additional features to it syn- 
 thetically. But owing to this origin, such judgments only 
 concern objects of the senses, and are only valid for objects 
 of possible experience. 
 
 IT Those who are not satisfied with these proofs and 
 illustrations are invited by Kant to consider the absurdities 1 
 which result from regarding the world of phenomena as a 
 real aggregate of things {per se) in real space and time. 
 This he does in the seventh section of his Antinomies of the 
 Pure Reason,' at which we shall arrive in due time. If the 
 world be a thing per se, it must be either finite or infinite. 
 But both these suppositions can be proved false, for we 
 cannot conceive space and time either as having bounds, or 
 as having none absolutely. It follows that our hypothesis 
 was false, and that the world is not an aggregate of real 
 things in space and time, but of intuitions necessarily sub- 
 ject to the forms of our faculties, which must therefore 
 always accompany them, and appear infinite, though they 
 are only indefinitely extensible. He also notices (p. 98, 
 note) that mathematical figures presuppose not merely in- 
 tuition, which gives the parts, but the gathering of these 
 parts into a unity, which is an act of the understanding. 3 
 
 1 Cf. Theory of Ethics, p. 196. 
 
 2 Kritik, p. 316, or else Kuno Fischer's Commentary, p. 230. 
 
 3 This remark, though contained in a footnote, and not brought 
 prominently forward by Kant, is of the greatest importance owing to 
 recent objections (such as Mr. Caird's), which assert that Kant unphilo- 
 
6S THE FACULTIES LOGICALLY SEPARATED chap, hi 
 
 This action of the understanding is treated fully in the next 
 chapter of the Kritik. 
 
 sophically isolated the mental faculties, and regarded them as acting 
 separately. He found it necessary to treat them logically as if they 
 were separate, but was not so stupid an observer as to mistake plain 
 facts. As far as the Aesthetic is concerned, he regards individual objects 
 as known by the special senses without the aid of understanding; but 
 this is only a provisional statement, which is corrected in the Analytic. 
 
CHAPTER IV 
 
 INTRODUCTION TO THE TRANSCENDENTAL ANALYTIC 
 
 If Before entering on a new division of Kant's Kritik, a few 
 general remarks will not be out of place. The Analytic 
 affords the reader a far longer and more weary task than 
 the Aesthetic. The latter is perhaps too compressed, owing 
 (probably) to Kant's earlier discussions having to a great 
 extent forestalled it At all events, there is hardly any 
 repetition, or enforcing of the same truth in slightly varied 
 language, when Kant discusses the basis of Mathematic. 
 The Analytic, on the contrary, just thought out by the 
 great philosopher, is born, if we may so say, with the 
 pangs of labour. We see Kant wrestling with his utterance 
 to put it clearly before the world. As might have been 
 expected, such a discussion defeated its own end. Repeti- 
 tions and explanations weary and confuse us, when they 
 are carried beyond reasonable limits. And so Kant labours 
 again and again at the Deduction or justification of his 
 Categories, in the first Edition, then in the Prolegomena, 
 then in his second Edition ; and yet his first exposi- 
 tion, though not the most complete, is by far the 
 clearest he has given. The difficulties, however, of any of 
 them seem quite sufficient for most English philosophers. 
 A few have made bold to discuss and comment on the 
 Aesthetic, or its doctrines. But as soon as we approach 
 
70 ■ SOURCES OF KNOWLEDGE chap. 
 
 the Analytic, we find little to help us, but either servile 
 repetition or silence. In truth, the duty of a commentator 
 on this part of Kant's Kritik is not merely to paraphrase, 
 or to expand. If he would have his author receive full 
 justice, he must, above all things, abbreviate. He must 
 bring together Kant's varied reassertions of the same fact, 
 and reproduce them in a single but complete form. This 
 will be the plan of the Commentary now offered on the 
 Deduction of the Categories, and, therefore, the practice of 
 adhering to the paragraphs of the original will be sometimes 
 advisedly abandoned. 
 
 The division on which we now enter is the second part 
 of the Transcendental Stoicheiology, viz. Transcendental 
 Logic. Kant's Introduction is naturally devoted to the 
 accurate definition of this expression, which is not in itself 
 obvious. 
 
 I. Of Logic Generally — 'Our knowledge springs from 
 two sources within us. The first is the faculty of receiving 
 impressions {sensibility), the second, that of knowing objects 
 by means of these impressions — or the faculty of producing 
 concepts {understanding). By means of the former an 
 object is given to us ; by the latter the object is thought, in 
 relation to this given representation, as a mere determina- 
 tion of our minds.' The distinction of pure {a priori) and 
 empirical {a posteriori), already applied 1 to intuitions, holds 
 good of concepts also. As the pure intuition contains the 
 mere form, in which something is intuited, so the pure 
 concept contains the mere form in which we think an 
 object generally. This receptivity of impressions and this 
 spontaneity of our understanding are both essential to 
 knowledge. Though perfectly distinct in their nature, they 
 combine in their action, and there is no ground for pre- 
 1 Above, pp. 27, 28. 
 
IV SENSIBILITY AND UNDERSTANDING 71 
 
 ferring one to the other. No object can be given without 
 sensibility, no object can be thought without understanding. 
 Taken by themselves, concepts without intuitions are vacua, 
 intuitions without concepts cceca. ' It is, therefore, as 
 necessary to make our con'cepts sensuous, as to make our 
 intuitions understandible, or bring them under concepts. 
 For the understanding can intuite nothing, and the senses 
 can think nothing — they cannot exchange duties ; we can 
 only have knowledge, or cognition, by combining them. 
 But this affords no reason for confusing their respective con- 
 tributions, which we should rather carefully distinguish and 
 separate. We have done so under the titles Aesthetic and 
 Logic' 
 
 11 It has been a common objection to Kant, urged 
 among others by Mr. Lewes and by Edmund Montgomery, 
 that Kant separated faculties which are never separate 
 in nature. The previous paragraph shows, we think, that 
 they have not done justice to Kant. He knew, as well as 
 they, that sensibility and understanding cannot be severed 
 in use. He even shows, what he claims as an original 
 psychological observation, that, in our ordinary perceptions 
 the understanding is necessary, acting through the imagina- 
 tion. 1 But if scientific analysis means anything, it means 
 the separate consideration of the fused elements given in 
 our ordinary experience. We cannot forbear to add, that 
 whatever mistakes in theory Kant may have made, his 
 psychological observation was far too subtile to allow any 
 obvious fact to escape his notice ; any objection founded 
 on such grounds can generally be straightway refuted from 
 his own express statements. 
 
 We may approach Logic with two different objects: 
 either to ascertain the rules of our understanding, which 
 1 Cf. vol. ii. p. 211, note. 
 
72 LOGIC, PURE AND APPLIED chap. 
 
 are absolutely necessary for using it in any way whatever, 
 or to ascertain those which guide its use when applied to a 
 particular class of objects. The latter is called the organon 
 (or logic) of this or that science. In the schools it is 
 generally supposed to be a preliminary to the study of a 
 science; in reality it is the very last question settled in 
 each science. We must know the objects very well before 
 we can state the rules which make a science of them 
 possible. Thus, for example, while mathematical studies 
 are in a very advanced condition, what is called the Logic 
 of Mathematic is still in obscurity. The first notions 
 necessary for a sound method in Geometry, in the Differ- 
 ential Calculus, or in Mechanics, still divide the minds of 
 competent inquirers. 
 
 Turning back then to the former — general Logic, which 
 makes no distinction as to objects — even this Logic may be 
 applied Logic, as well as pure. For pure general Logic 
 must abstract from all empirical conditions, not only 
 external but internal. It must exclude the influences of 
 the senses, the play of imagination, the laws of memory, 
 and all those sources of particular prejudices, which only 
 affect us in particular cases. It merely contains a canon of 
 the understanding and reason, as regards the formal elements 
 in their use. So far the matter is plain enough. But how 
 can a general Logic be applied, for this seems to imply par- 
 ticular objects ? There is one set of conditions, which, 
 though empirical, yet apply to all mankind. These are 
 the psychological conditions that may hinder or advance 
 reasoning. Logic can be general, and yet include a con- 
 sideration of these, viz. attention and its consequences, 
 the causes of error, the conditions of doubt, of conviction, 
 and such like. For these, though individually contingent 
 are, as a whole, necessary for any concrete use of the 
 
TRANSCENDENTAL LOGIC 73 
 
 understanding. This is the only sense in which Kant 
 admits the term applied Logic, as a kathartikon, or purifier, 
 of common sense. It bears the same relation to pure 
 general Logic that moral teaching, or practical Ethic bears 
 to pure Morals, a science which contains nothing but the 
 necessary moral laws of a free will, while practical Ethic 
 discusses the difficulties and hindrances with which men 
 must contend in carrying out these laws. 
 
 It is necessary for logicians to sever most carefully the 
 pure from the applied (though still general) side of Logic. 
 The former alone is strict science. They should keep 
 before them two rules; (1), as general pure Logic abstracts 
 from all content whatever in our knowledge ; (2), as being 
 pure it possesses no empirical principles, more especially 
 it must borrow nothing from empirical Psychology. In a 
 demonstrative science everything must be certain a priori. 
 
 II. Of Transcendental Logic. — We spoke above of the 
 Logic of a particular use of the understanding. When this 
 particular use is determined by a special class of objects 
 (or special science), its Logic becomes the organon of that 
 science. But the particular use may be determined not by 
 the object, but by the special procedure of the understanding. 
 As there are not only empirical but pure intuitions, so there 
 may be a similar distinction between empirical and pure 
 thinking. We might then have a Logic which did not 
 abstract from all the content or matter of knowledge. For 
 though the analysis of the pure thinking of objects must (1) 
 exclude all cognitions of empirical content, in order to be 
 general, it may (2) discuss the origin of our knowledge, 
 when pure — a subject quite beyond the range of general 
 pure Logic, which must take the facts as it finds them, and 
 in no way concerns itself about their origin, but only about 
 the way in which the understanding connects them. 
 
74 WHAT IS TRANSCENDENTAL ? chap. 
 
 Kant here inserts a remark of great importance. The 
 term transcendental is not applied in the Kritik to every a 
 priori cognition, but only to such as inform us, that certain 
 representations (intuitions or concepts) are applied altogether 
 a priori, how they are so applied, or that they are possible 
 a priori, and how so. For example, space or any of its 
 geometrical figures is no transcendental representation, but 
 the knowledge that it is a priori, and the possibility of its 
 applying, though a priori, to the objects of the senses, this 
 may be called transcendental. The contrast of empirical 
 and transcendental concerns a distinction in our cognitions, 
 and not their relations to their objects. 
 
 Let us assume that there are concepts referring a priori 
 to objects, as we found intuitions so doing. These con- 
 cepts being mere acts of pure thinking, neither empirical 
 nor aesthetical in origin, suggest to us the notion of a 
 peculiar science, the science of that cognition by which we 
 think objects completely a priori. Such a science deter- 
 mining the origin, sphere, and objective validity of such 
 cognitions, may be called Transcendental Logic, being con- 
 cerned with the laws of understanding and reason only, 
 and with these only so far as they apply to objects a 
 priori. 
 
 III. The Division of General Logic into Analytic and 
 Dialectic. — The celebrated old question by which men 
 thought either to puzzle logicians, or make them confess 
 their ignorance, was this : What is truth 1 The definition 
 of the term — the agreement of cognition with its object — 
 is here pre-supposed ; what we demand is the universal and 
 safe criterion of the truth of each single cognition. Prudens 
 interrogate dimidium scientiae. A silly question may not 
 only disgrace the interrogator, but mislead the incautious 
 respondent into silly answers, and so (according to an old 
 
iv FORMAL CRITERION OF TRUTH 75 
 
 proverb) while one milks the he-goat, the other holds the 
 sieve for him. 
 
 If truth means the agreement of a cognition with its 
 object, it is implied that this object is distinguished from 
 all others, for the cognition, whatever other agreements it 
 may contain, is only true if it correspond to its particular 
 object. Now an universal criterion of truth must be valid 
 for all cognitions, without distinguishing their objects. As, 
 therefore, such a criterion must abstract from all content of 
 knowledge, and is at the same time to concern this very 
 content, to demand it is absurd and self-contradictory. An 
 universal as well as sufficient mark of such truth cannot be 
 found. As the content is called the matter of knowledge, 
 our result may be stated thus : ' No universal criterion of 
 the truth of our cognition as to its matter can be required, 
 because such a criterion is self-contradictory.' 1 
 
 As regards mere form, the science of Logic, in expound- 
 ing the universal and necessary rules of the understanding, 
 must evidently present us in these rules criteria of truth. 
 Whatever contradicts them must be false, or the understand- 
 ing would be in conflict with its own general laws. But 
 these criteria are insufficient, as they only affect the form of 
 truth. For a cognition might fully satisfy the logical form, 
 — that is, not be self-contradictory, — and yet contradict the 
 object. The mere logical condition^ of truth, or agreement 
 of a cognition with the universal and formal laws of under- 
 standing and reason, is then the sine qua non, or negative con- 
 dition of all truth : but logic can do no more, and is power- 
 less to detect the errors which affect not form, but content. 
 
 General logic analyses all the formal operations of un- 
 derstanding and reason, and determines the principles of 
 all logical estimating of our knowledge. ' This part then 
 1 Cf. Mansel, Pro/eg. Logica, p. 189, 190. 
 
76 ANALYTIC AND DIALECTIC chap. 
 
 may be called Analytic, and is the negative touchstone of 
 truth, whose rules must be thoroughly satisfied before we 
 examine our knowledge as to matter, to see whether it con- 
 tains positive truth as regards the object.' 1 But as the 
 mere form of knowledge, however perfect, is quite insuffi- 
 cient to guarantee the material (objective) truth of know- 
 ledge, no man can venture to assert from Logic anything 
 about objects, without obtaining farther information. He 
 may then attempt to use or combine his materials accord- 
 ing to logical laws, or better still, merely to test them in 
 this way. But the act of giving all our cognitions the form 
 of understanding, however indigent we may be as regards 
 their matter, is so seductive that general logic, which is a 
 mere canon for testing, has been used (and accordingly 
 abused) as an organon for producing, or apparently produc- 
 ing, objective assertions. General logic, when it claims to 
 be an organon, is called Dialectic. 
 
 However the ancients may have varied in their accepta- 
 tion of this term, they practically used it as the Logic of 
 illusion. They applied the accurate method of logic so- 
 phistically to give their ignorance, or even their deliberate 
 fallacies the appearance of truth. But we may lay it to heart 
 as a safe and useful caution : that universal Logic, considered 
 as an organon, is always dialectical, or a logic of illusion. 
 The pretence, therefore, of using it as an instrument 
 (organon) to extend, even apparently, our knowledge, turns 
 out mere idle talk. Such a proceeding is totally unworthy 
 of philosophy. Dialectic has therefore been included in 
 Logic as the Kritik of dialectical illusion, and as such we 
 shall here use it. 
 
 1 Mr. Meiklejohn has seen the true sense of this sentence, in which 
 the pronouns are very confused. We have followed his version, which 
 is not the obvious one. 
 
iv DIVISIONS OF TRANSCENDENTAL LOGIC 77 
 
 IV. The Division of Transcendental Logic into Transcend- 
 ental Analytic and Transcendental Dialectic. In transcend- 
 ental Logic we isolate the understanding (as we isolated 
 the sensibility in the transcendental Aesthetic) and select 
 from our knowledge that part of thinking which has its 
 origin in the understanding. But we cannot use this pure 
 cognition except objects be given us in intuition — a neces- 
 sary condition, without which cognition is void. The part 
 of transcendental Logic, then, which expounds the elements 
 of pure rational cognition, and the principles without which 
 nothing can be thought, is called Transcendental Analytic, 
 and is a Logic of Truth. But though experience alone 
 provides us with the matter to which these pure concepts of 
 the understanding can be applied, there is the strongest 
 temptation to use them by themselves, and beyond the 
 limits of experience. Hence the understanding is in danger 
 of making a material use of its formal principles, and of 
 judging about objects which are not, and perhaps even 
 cannot be, given. Thus a mere canon for controlling the 
 empirical use of the understanding, is misapplied into an 
 organon of universal and unlimited use, and the understand- 
 ing ventures without further aid to assert and decide syn- 
 thetically about objects in general. The second part of our 
 transcendental Logic must therefore criticise this dialectical 
 illusion, and is called transcendental Dialectic — not in the 
 sense of the art of producing such illusion, but — as the 
 Kritik of the Understanding and Reason in their hyper- 
 physical employments, which exposes the illusion of their 
 false pretensions, and reduces their claim of discovering 
 and extending knowledge by purely transcendental principles 
 to a mere protecting the understanding from sophistical 
 illusions 
 
CHAPTER V 
 
 TRANSCENDENTAL LOGIC. PART I 
 
 The Transcendental Analytic. — This Analytic is the re- 
 solution of our whole a priori cognition into its component 
 elements. It requires the following conditions: — (i) the 
 concepts must be pure ; (2) they must belong, not to 
 sensibility, but to thought and understanding ; (3) they 
 must be elementary concepts, well distinguished from those 
 deduced from them, or composite ; (4) the list must be 
 complete. This last quality cannot be obtained by a mere 
 aggregate of observations. We must start from a notion of 
 a priori cognition as a whole, and subdivide it into the 
 concepts belonging to it, which are then connected systematic-, 
 ally. The pure understanding, apart from all that is em- 
 pirical, and even from all sensibility, is an independent 
 self-contained unity, not to be enlarged by additions from 
 without. Hence the sum total of its cognition forms a 
 system falling under one idea, and the completeness and 
 perfect articulation of this system will be a touchstone to 
 test the claims of all cognitions that belong to it. The 
 Analytic is divided, on the model of ordinary Logic, into 
 two books, of which the first contains the concepts, the 
 second the fundamental Judgments [or Principles, Grundsdtze'] 
 of the pure understanding. 
 
ANALYTIC OF CONCEPTS 79 
 
 Transcendental Analytic, Book I. 
 
 The Analytic of Concepts. — Kant does not mean by this 
 term the ordinary analysis of concepts, in order to make 
 them clear and distinct, but rather the analysis of the 
 faculty of Understanding itself — a task seldom attempted. 
 For we shall endeavour to ascertain the possibility of a priori 
 concepts by seeking them in the understanding alone as their 
 place of birth, and analysing its pure use ; this is the proper 
 business of transcendental philosophy. These pure con- 
 cepts lie prepared in their forms and dispositions in the 
 human understanding, until they develop by occasion of 
 experience, and may then be found in the same understand- 
 ing, freed of the empirical conditions that attach to them. 
 
 Analytic of Concepts [Chapter I.] 
 
 Of the Clue to discover all pure Concepts of the Under- 
 standi?ig. — If we bring any of our cognitive faculties into 
 action, in due time sundry notions present themselves, which 
 make this faculty known to us, and which can be collected 
 more or less completely, according to the number or acute- 
 *3e*ss of our observations. But we can never make sure 
 that our task is completed ; and moreover the concepts 
 attained in this accidental way come in no order or sys- 
 tematic unity, nor can they attain these qualities by being 
 ranked according to their greater or less comprehension, 
 however methodical this arrangement may be. 
 
 Transcendental philosophy has the advantage, and is 
 also under the obligation, of seeking its concepts on a fixed 
 principle, as they spring pure and unmixed from the under- 
 standing as an absolute unity, and must therefore be con- 
 nected according to some one notion. This connection 
 must afford a rule, according to which every pure concept 
 
So HOW WERE THE CATEGORIES OBTAINED ? chat. 
 
 may be ranked, and the completeness of the list fully deter- 
 mined a priori 
 
 IT It appears manifest from these introductory remarks 
 that the common charge made against Kant, of having 
 picked up his Categories empirically, and without proper 
 reduction, can only be true in one sense. He may have 
 blundered in the carrying out of his idea — a question 
 which cannot be discussed as yet — but he certainly did not 
 intend to proceed empirically, nay, he even professedly 
 repudiated the error attributed to him by his critics. His 
 mistake then, if he really made it, was not a mistake in 
 principle, but a mistake in psychological insight — an 
 inability to see how several of his separate heads might 
 really be reduced to one. We must repeat that such a 
 fault is not likely to occur in a great master thinker, and 
 that if his starting point be right, as his critics must confess 
 (according to their own statements, which coincide with 
 his), he is less likely than they to have committed an error 
 from want of acuteness. But of this anon. 
 
 IT We now come to the first really difficult discussion in 
 the Kritik. The Aesthetic, however the commentators may 
 have blundered in details, has been, as a whole, compre- 
 hended even by English philosophers, and has had its full 
 and fair effect on philosophical thought, ever since Kant's 
 days. But with the exception of Hegel, whose system 
 implies a profound knowledge of Kant's transcendental 
 Logic, it would be difficult to find any philosopher who had 
 fully utilised Kant's teaching in this part of his work. The 
 later commentaries are indeed progressing towards a fuller 
 knowledge. That of Kuno Fischer, for example, following 
 the lead of Schopenhauer, is tolerably clear and precise on 
 the difficult Deduction of the Categories, so far as it was ex- 
 pounded in the First Edition, though he shirks all com- 
 
MEANING OF FUNCTION' 
 
 parison with the changes in the later Editions. But even 
 Kuno Fischer settles the hard question before us, the 
 passage from the ordinary table of logical judgments 
 to that of the Categories, in a single short paragraph, 
 omitting all mention of Kant's close analysis of the 
 function of thought, and the functions of unity produced 
 among phenomena of the understanding. We can here 
 obtain no help but from the author himself, and must seek 
 from a careful analysis of his argument and from the short 
 explanations in his Prolegomena to develop his meaning. 
 
 The Transcendental Clue to discover all the Pure 
 Concepts of the Understanding [Section I.] 
 
 Of the logical Use of tlie Understanding generally. — 
 We know already that our understanding has no faculty of 
 intuition. As there is no other way of cognising except 
 through concepts, the understanding must cognise through 
 them, but discursively, as opposed to intuitive cognition. 
 All intuitions are affections of sense, all concepts functions 
 of the understanding. By Function Kant means ' the unity 
 of action in ranging various representations under a higher 
 one.' So the understanding is spontaneous in forming 
 its concepts, not receptive of impressions, like the sensibility. 
 What we here desire to ascertain is the exact number 
 of the primitive functions of the understanding — that is, 
 of its various a priori unities of action, or ways of ordering 
 its representation under higher notions. How does Kant 
 set about doing this ? In the first place, suppose we have 
 the concept ready, how can we make use of it ? It is con- 
 fessed, that no concept refers directly to objects. Intuitions 
 alone are immediate representations of objects, for we still 
 hold [in spite of Sir Wm. Hamilton and other Scotchmen] 
 that external objects (in the common sense) are not pre- 
 
82 THINKING IS JUDGING chap. 
 
 sented to the mind, but represented by intuitions, or 
 modifications of our sensibility. Therefore we regard both 
 intuitions and concepts as representations, but the former 
 as immediate, the latter as mediate ; and when the concept 
 refers to objects through the intuition, it produces in us the 
 representation of an (immediate) representation. But as 
 we cannot use concepts to intuite objects through them, 
 there is no use left for them except that we should judge 
 objects by means of them. For in a judgment we can 
 bring an intuition (as subject) under a concept (as predicate), 
 and so we indirectly cognise an object through our concept. 
 We may repeat the process, making a lesser concept the 
 subject, and bringing it under a greater, as predicate. For 
 example, in the judgment all bodies are divisible, both terms 
 are concepts, but we have previously brought many intuitions 
 under the former, by judging : this is a body, and that is a 
 body. But then we bring the concept body under a higher 
 concept of divisibility, which applies to other concepts also. 
 Judging, therefore, consists in bringing many intuitions, or- 
 many lower concepts, under one higher concept, which 
 embraces the many. Hence judgments are called by Kant 
 functions of unity — that is to say, functions producing unity, 
 in our cognitions. But, if concepts can only be used in 
 judgments, then the understanding has no action apart from 
 judging; it may, in fact, be called the faculty of judging. 
 For what is thinking (as opposed to intuition) but cognising 
 through concepts? Whenever we do this, we regard the 
 concept as the predicate of a judgment in which we assert 
 that another representation comes under it. Every concept, 
 therefore, as it must contain many representations, so it 
 must be also regarded as the predicate of many possible 
 judgments. If therefore all the acts of the understanding 
 can be classed as judgments, and if concepts can only be 
 
v FUNCTIONS OF UNITY 83 
 
 used in judgments, it is obvious that an analysis of the 
 various kinds of judgments will be the clue to discover 
 the various kinds of concepts. In Kant's language, the 
 functions of unity in judgments, if they can be reduced to 
 fixed classes, will discover to us all the functions of the 
 understanding. 
 
 H We think we have left no difficulty unresolved in the 
 above paragraph, except the definition of the word Function, 
 which Kant says is ' The unity of the act (die Einheit der 
 Handlung) of ranging diverse representations under a 
 common one.' The reader must not confuse this with 
 what Kant calls functions of unity, which are evidently 
 functions producing unity, as is clear from his exposition of 
 judgments, which he describes in these words. We had 
 once thought of suggesting a slight emendation (der Einheit 
 die Handlung), so that the passage would run thus : ' I 
 mean by function of unity, the act of ranging, etc' This 
 would evade the difficulty, and give a true and easy sense. 
 But the recurrence of the words, ' the unity of the act of 
 combining,' etc., in at least three other places in the 
 Critick, is decisive against the solution of the difficulty. 
 
 These are two of the passages (p. 85) : 'I cannot know a 
 line, or anything else in space, without drawing it, and so 
 producing a certain combination of various elements syn- 
 thetically, so that the unity of this act is at the same time 
 the unity of consciousness (in the notion of a line), and so 
 alone can an object in space be cognised.' He accordingly 
 says, in a very decisive passage (p. 80), when talking of 
 synthesis, as a spontaneous combination of variety by 
 the understanding : ' You here perceive easily, that this 
 action is originally one [einig] and equally valid for all 
 combination.' 1 
 
 1 This sentence is completely ruined in Mr. Meiklejohn's translation, 
 
THE UNITY OF OBJECTS 
 
 The third is a passage expounding the very same sub- 
 ject as we are now treating, and is a sort of appendix to 
 the Aesthetic thrust into the Deduction of the Categories 
 in the Second Edition. 1 He is speaking of the action of 
 the understanding on the materials supplied by intuition, 
 and says, ' the synthesis [of the understanding], considered 
 separately, is nothing but the unity of the [act or] action, of 
 which, as such, the understanding is conscious, even without 
 sensibility, but by which it determines even the sensibility,' 
 etc. He proceeds in the next paragraph to illustrate his 
 meaning by the very example adduced in the previous 
 passage, adding several other similar ones. 
 
 11 It will thus appear that Kant declares no object can 
 be considered by the mind as such, till its parts or qualities 
 have been brought together by a spontaneous act of the 
 mind, and that though the various parts must be gathered 
 successively, there is an unity in this act in every case, and 
 this is really the unity attributed to the object. A slight 
 variation in expression will be found in the corresponding 
 passage of the First Edition to that last quoted. 2 He there 
 speaks of the identity of the function, and even of the 
 identity of the act, which subjects intuited impressions to an 
 a priori unity. There is good reason to speak of the 
 identity of the function, for it is in every case a synthesis of 
 variety ; but we think Kant wisely changed his expression in 
 the Second Edition, and spoke of the unity of the act, for 
 though the various Categories are indeed phases of the same 
 identical function of synthesis, yet the acts of unifying 
 
 who strangely substitutes the following : ' The reader will easily enough 
 perceive that the possibility of the conjunction must be grounded in the 
 very nature of this act,' etc. ! 
 
 1 Latter part of §§ 24 and 25, or §§ 20 and 21. 
 
 2 Cf. vol. ii. p. 201. 
 
v THE CATEGORIES 85 
 
 multiplicity in each can hardly be called identical. And 
 so, in the Second Edition, in the close of the passage we 
 have been expounding, he says that an analysis of judg- 
 ments will disclose to us the Functions of the understanding ; 
 that is to say, the phases of that higher function, which is 
 identical in all consciousness, and consists simply in making 
 all representations our oivn, or modifications of an identical 
 self. The action of the understanding in the case of each ' 
 Category is identical, for each Category is one way of 
 producing unity in our representations. The action of the 
 understanding in all the Categories together is one kind of 
 action, and so far one, though it may vary in the detail of 
 each Category. 
 
 The Clue to discover all Pure Concepts of the 
 Understanding [Section II.] 
 
 § 9. 1 Of the Logical Fiindioji of the Understanding in 
 Judgments. — If we abstract from all content, and apply our- 
 selves to the mere form of judgments, we find that the 
 function of thought in them can be brought under four 
 heads, each of which contains three phases [Momente]. 
 Here is the table : 
 
 1 This numbering of paragraphs continued (after a long suspension), 
 from p. 60 above, is used by Kant down to the end of the Deduction, 
 merely for the convenience of reference, and has nothing to say to the 
 divisions and subdivisions of chapters and sections in which he luxuriates. 
 We have preserved it for the same reason. Mr. Meiklejohn has un- 
 warrantably changed it, and so confused some of Kant's references. 
 He has inserted §§ 1-4 in the earlier part of the Analytic, where Kant 
 does not use them, 'and so finds himself at § 5, where Kant resumes 
 with § 9. The reader who refers to his translation will note this, as we 
 shall refer to Kant's paragraphing, even when we give the pages 
 according to the English translation. 
 
LOGICAL TABLE OF JUDGMENTS 
 
 I. Quantity of Judgments. 2. Quality. 
 
 Universal. Affirmative. 
 
 Particular. 1 Negative. 
 
 Singular. Infinite. 
 
 3. Relation. 4- Modality. 
 
 Categorical. Problematical. 
 
 Hypothetical. Assertative. 
 
 Disjunctive. Apodeictical. 
 
 Some remarks on this table are necessary. 
 
 1. Logicians say justly, that singular judgments can be 
 treated as universals in syllogisms, for the predicate is 
 affirmed of the whole subject. But in their quantity, as 
 cognitions, they differ as unity does from infinity, and are, 
 therefore, essentially distinct. Hence the judicium singulare, 
 as a mere cognition, compared in quantity with other cog- 
 nitions, must have a separate place in a complete table of 
 the phases of thinking, though indeed this would not be 
 necessary in a logic limited merely to the consideration of 
 the use of judgments in reference to each other. 
 
 2. Similarly in transcendental Logic, infinite judgments 
 must be distinguished from affirmative, which is contrary to 
 the practice of General Logic. The latter abstracts wholly 
 from the content of the predicate, and merely examines 
 whether it is affirmed or denied of the subject. But the 
 former is concerned about the value of this affirmation by 
 means of a negative predicate, and what our cognition can 
 gain by it. If I say the soul is not mortal, this judgment, 
 regarded as negative, at least guards against an error. Re- 
 garded as affirmative (not-mortal), it divides all the universe 
 into mortal and immortal things, and excluding the mortal, 
 
 1 Kant calls these Plurative judgments. 
 
v ANALYSIS OF DISJUNCTIVE JUDGMENTS 87 
 
 includes the soul in the infinite number of non-mortal 
 things. Many other parts might in like manner be ab- 
 stracted, without in the least increasing or determining our 
 notion of the soul. These infinite judgments then, as re- 
 gards logical extension, are really limitative of the general 
 content of a cognition, and must not be passed over in our 
 transcendental table, as this function of the understanding 
 may be important in the field of pure a priori cognition. 
 
 3. All the relations of thinking in judgments are either 
 (a) that of predicate to subject, (/3) that of antecedent 
 to consequent, (7) that of a divided cognition to its com- 
 bined members. In the first kind two concepts, in the 
 second two judgments, in the third several judgments, are 
 considered in relation to each other. This is obvious in 
 the first and second cases. The relation of the various 
 parts in a disjunctive judgment is not only one of opposition, 
 in that they are mutually exclusive, but also of community, 
 in that they jointly make up the sphere of the cognition 
 concerned. Each then completes the rest. For example, 
 ' the world exists either through blind chance, or through 
 internal necessity, or through an external cause.' Each of 
 these judgments embraces a portion of the sphere of our 
 possible knowledge about the world's existence : taken 
 together they express the whole of this sphere. Remove 
 one and you posit the rest, and vice versa. There is then 
 this community in these judgments, that though mutually 
 exclusive, they make up, when combined, the whole con- 
 tent of one given cognition. 
 
 4. The Modality of judgments is a peculiar function, in 
 that it adds nothing to their content, but only affects the 
 
 ■ value of the copula in relation to our thinking. Problem- 
 atical judgments are those in which the affirmation or 
 negation is only considered possible, or optional. In assert- 
 
THE CATEGORIES 
 
 ative, it is considered actual, or true ; in apodeictical, neces- 
 sary. Thus in a hypothetical judgment, both antecedent 
 and consequent are optional, and only the consequence 
 is assertative. Whether you assume them true or not, the 
 consequence is actual. So the various judgments contained 
 in a disjunctive judgment are problematical, and may be 
 merely assumed for a moment, to lead the mind towards 
 the true solution. The assertative judgment asserts logical 
 actuality or truth, as, for example, in a hypothetical syllogism 
 the antecedent which has occurred problematically in the 
 major premiss, is repeated assertatively in the minor, and 
 shows that the proposition is no longer optional, but con- 
 nected with our understanding according to its laws. The 
 apodeictical judgment considers the assertion as determined 
 by these laws of the understanding, and, therefore, as mak- 
 ing its assertion a priori, or necessarily. As these various 
 stages are reached gradually, the mind proceeding from an 
 assertion first to its truth, and then to its necessity, the 
 three functions of modality may be counted as distinct 
 phases of thought. 
 
 The Clue to discover all the Pure Concepts of 
 the Understanding [Section III.] 
 
 § 10. On the Pure Concepts of tlie Understanding, or 
 Categories. — 51 We now come to the second serious difficulty 
 in the Analytic, which has not yet been adequately explained. 
 Even Kuno Fischer's professed Commentary quietly ignores 
 all the argument in this paragraph, and contents itself with 
 stating the result. 1 As Kant's own exposition has not 
 penetrated the minds of his readers, it seems advisable to 
 abandon paraphrasing, and to state the argument in an 
 independent form. 
 
 1 Cf. Kuno Fischer's Commentary, pp. 71 sqq. 
 
FORMATION OF JUDGMENTS 
 
 1T We saw above (p. 82) that the function of the under- 
 standing was simply to gather up objects, or lesser concepts, 
 under higher notions — in fact to produce unity in our 
 knowledge ; and we saw that this was done by acts of 
 judging, which compared the lesser notion with the greater, 
 and either reduced it under this notion, or referred it to 
 some notion, to be otherwise determined. But an analysis 
 of judgments is a ■ task long since accomplished by logi- 
 cians, and from this analysis it appears that our judgments, 
 or acts of producing unity among our representations, 
 must assume, and can only assume, one in each group of 
 the forms laid down in the logical table of judgments. 
 This fact points to an original feature in the constitution of 
 our understanding, and indicates that our power of com- 
 bining variety is confined to these forms, or to something 
 corresponding to these forms. 
 
 Now let us consider that all the subjects of our earliest 
 judgments are not concepts, but intuitions of objects. We 
 cannot in fact have a general concept till we have formed it 
 by means of a number of such judgments. Nevertheless, 
 from the very outset our judgments must have taken the 
 various forms laid down in the table of judgments — that is 
 to say, we must have considered the objects about which 
 we judged as one, as many, or as both ; as substances, 
 as causes, etc., etc. How did we come to range our 
 intuitions under such classes ? For we received through 
 our sensibility nothing but various impressions or repre- 
 sentations, which our sense placed in space and in time, 
 but which were in sense not otherwise combined, and were 
 nothing but a confused congeries of single impressions. 
 How, in fact, have we come to consider them as separate 
 objects at all, and judge about them as single, as plural, as 
 substances, or as causes ? 
 
90 TRANSCENDENTAL SYNTHESIS chap. 
 
 IT Kant says there is from the very beginning another 
 faculty besides sense at work, unconsciously indeed and 
 darkly working, but occupied in binding up the various 
 data of sense into groups or unities. This faculty is the 
 imagination, and its function he calls synthesis. We can 
 never obtain a notion of any combined set of impressions, 
 or of such a complex as distinct from surrounding im- 
 pressions, without the action of this faculty. It is not, 
 however, left to combine impressions at random, upon no 
 principle, for being closely allied with the understanding, 
 the spontaneous action of the understanding supplies it 
 with its rules, and so it proceeds to combine and form 
 objects according to the laws prescribed by the under- 
 standing. 
 
 But we saw above that the laws of combination under 
 which the understanding acts are expressed by the forms 
 of the various judgments. Hence these forms are a clue to 
 the laws which the understanding prescribes to the imag- 
 ination in its synthesis of representations, in order to form 
 objects. It is the same action which produces unity in the 
 judgment, and unity in the intuition — taking unity in a 
 large sense, to signify systematic connection among the 
 parts. Hence the various phases of this synthesis of 
 intuitions correspond to the various forms of judgment, 
 and both are called by Kant Categories. They must of 
 course be general phases, general ways of combining — in 
 Kant's words, synthesis, represented generally?- They must 
 affect the pure synthesis of the imagination, and must 
 involve no data save those given by the transcendental 
 
 1 So general, as he observes of the science of nature in his Prolegomena 
 (p. 65), as to reduce nature generally, including both the object of the 
 external and that of the internal sense (the object both of physic and of 
 psychology) to universal laws. 
 
THE FORMATION OF OBJECTS 
 
 Aesthetic, viz. multiplicity generally, presented in space j 
 and time. 
 
 Let us endeavour to make the whole matter plain by 
 an illustration. When you sit in a room, the senses supply 
 you with a number of impressions all separate in space or 
 in time, various patches of colour, various sensations of 
 texture, of sunlight, of cold, of odours, and many others. 
 The sensibility may give these partly simultaneously, partly 
 in succession, but no more than these. How does the 
 mind come to sever them all into distinct groups, and call 
 them separate objects, or the qualities of separate objects ? 
 Surely in this way, that the imagination, by reproducing 
 some of the sensations just past, binds them up with those 
 really present, and so produces a certain grouping of re- 
 presentations. This grouping, however, is not carried on 
 at random, but according to laws imposed on the imagina- 
 tion by the understanding. What are these laws ? They 
 are not special laws binding up any particular kind of repre- 
 sentations, but general laws, applicable to any sort of 
 representation, and they are not difficult to indicate. 1 We 
 
 1 It will be observed that Kant's theory of our empirical knowledge 
 is just as much a ' psychological theory ' as Mr. Mill's, inasmuch as he 
 holds the external world, as we know it, to be made up of our percep- 
 tions and their relations, and of these only. But Kant differs from the 
 Association School in attributing the notions of Externality, and also 
 of Permanence and Substantiality, to original laws of our minds, and 
 not to the gradual teaching of associations. Mr. O'Hanlon's difficulties, 
 even as they appear in Mr. Mill's short extracts {Exam, of Hamilton, 
 chap, xi.), are not answered, and there are many others in reserve. 
 
 For instance, Mr. Mill's object (pp. 222, 223) consists of an indefinitely 
 large group of sensations, of which a few only are present, but suggest 
 all the rest by inseparable association. This implies such a set of dis- 
 tinct inseparable associations of each of the present sensations with each 
 of the remainder, that our completed notions of substances should be 
 very slowly obtained. 
 
 He next (p. 224) advances the statement that the constant ante- 
 
92 MILL'S THEORY REFUTED chap. 
 
 see a single chair close to us, we see several others through 
 the room, we speak of the whole furniture of the room. 
 This means that the understanding has directed our imagin- 
 ation to consider a certain quantity of sensations as making 
 up a single chair, and has therefore imposed the Category 
 of Unity on these sensations. "When we speak of several 
 chairs, then a larger group of sensations, in which certain 
 features recur, has been arranged by the imagination as a 
 plurality of units — here is the Category of Plurality. The 
 whole mass of representations produced by the objects 
 around us (as we call them) is again considered as a 
 
 cedences and consequences which we observe in Nature, and which 
 form the most important part of our knowledge of external things 
 — that ' almost all of these do not obtain between [present] sensa- 
 tions, but between the [absent] permanent possibilities of sensation ! ' 
 We confess we have not yet thought ourselves into the attitude 
 of Mr. Mill's school, so as to comprehend how a series of all-import- 
 ant indissoluble associations can arise between a set of absent possi- 
 bilities, which are only suggested by certain present sensations not so 
 associated. 
 
 Here is a third difficulty. He thinks (p. 230) that with every 
 sensation we have, we associate the idea of something different from 
 it. This, of course, must be some different sensation. He argues 
 that, in like manner, we come to associate with the sum total of our 
 sensations this notion of something different from it. This inference 
 appears to violate all the conditions of inseparable association. We 
 seldom think of this sum total at all, and when we do, it comprises all 
 the differing sensations ; so that, of all our ideas, it is that which does 
 not imply any idea of difference, except that of individual sensations, as 
 opposed to a whole of sensations. For how could an association 
 between two differing sensations suggest an association between a 
 sensation and a thing not a sensation ? He goes on to state (p. 230) 
 that this something, regarded as different (we suppose in kind) from a 
 sensation, or a possibility of sensation, is identified with permanent 
 possibilities of sensation, because these latter are extremely unlike 
 actual sensations [only]. Are they unlike possibilities of sensation also? 
 Surely this part of Mr. Mill's argument falls to pieces whenever we 
 try to grasp it. 
 
ILLUSTRATIONS 93 
 
 plurality of units, making up one great unity — this is the 
 Category of Totality in application. It appears, then, that 
 we could not obtain our ordinary knowledge of objects, if 
 the understanding did not order our receptive faculties to 
 group the representations they receive, as to quantity, into 
 units, pluralities, and totalities. In like manner, the sensa- 
 tions which affect us are classed as Real, such as those of 
 light and heat in the room, as the Negation of real, such as 
 darkness and cold, or as Limitations of the real, as when 
 we feel transitions from the one to the other. We must 
 also consider our groups of representations in relation to 
 each other, and so we judge that the hardness, colour, and 
 texture of the chair are Accidents of a Substance, that the 
 heat of the room is the Effect of a Cause (the fire), and 
 that the various articles of furniture in the room are not re- 
 lated in either of these ways, but are substances simultane- 
 ously existing, and determining each other in place, as, 
 for example, we indicate the place of a stool by its proximity 
 to a sofa or a chair. Finally, we consider certain sensations 
 possible, as when we see the end of the poker bluish, 1 or 
 brick colour, and judge that it will possibly burn us if we 
 touch it. The judgment becomes actual, if we make the 
 experiment, and burn our fingers. A person sitting in the 
 room before we came in, who had seen it thrust between 
 the bars of the grate, and removed just when it was 
 becoming red, would judge its burning heat to be necessary. 
 Thus, then, in our ordinary experience, we are perpetually 
 using the Categories, not only as the frames, or pure forms 
 of our judgment, but as the frames, or pure notions, of the 
 objects of intuition around us. 2 Such is Kant's famous 
 
 1 We do not say red hot, in which case the judgment is stronger 
 than a possible one. 
 
 2 It was noticed by Mr J. C. Malet, that this binding up of in- 
 
94 TRANSCENDENTAL TABLE OF CATEGORIES chap. 
 
 derivation of his Categories from the forms of judgments 
 a derivation imperfectly comprehended for want of holding 
 fast to Kant's express statement, that these pure a priori 
 concepts were the rules not merely for judgments, but for 
 all objects of our intuition. 
 
 The following is accordingly the table of these pure 
 concepts of the understanding : — 
 
 1. Quantity. 3- Relation. 
 Unity. 1 Of Inherence & Subsistence. 2 
 Plurality. Of Causality & Dependence. 
 Totality. Of Community (reciprocal 
 
 action between agent and 
 patient). 
 
 2. Quality. 4. Modality. 
 Reality. Possibility — Impossibility. 
 Negation. Existence — Non-existence. 
 Limitation. Necessity — Contingency. 
 
 IF Kant has called them Categories, because his original 
 design is identical with Aristotle's, though the execution of 
 it differs widely. On this observation English philosophers, 
 especially Hamilton and Mansel, have made such comments 
 as disclose a complete ignorance of Kant's derivation of the 
 Categories. They say that he brought a new, nay an 
 
 tuitions produces integral wholes, whereas the binding up of concepts 
 in judgments produces universal wholes — a remark worthy of notice, as 
 suggesting a transcendental affinity between logical and real unity. 
 Cf. Prolegomena, pp. 73, 7&- Kant there holds that the Categories 
 determine our intuitions, in order to fit them for judgments. 
 
 1 Cf. vol. ii. p. 75, note, where we see, as is otherwise plain, that 
 the Category of Unity corresponds to singular judgments. 
 
 2 In the Categorical judgment A is B, B is regarded as inherent in 
 A, which is the relation of accident and substance. The Category of 
 Reality signifies presence to sensation ; that of Actuality existence at 
 some definite time, not necessarily the present time. 
 
v ARISTOTLE AND KANT 95 
 
 opposed meaning into Aristotle's time-honoured nomencla- 
 ture, because the investigations were totally diverse. Ham- 
 ilton, in his usual antithetical style, which palms oft" rugged 
 clearness upon us as good sense, says that Aristotle was 
 investigating the laws under which the object is known ; 
 Kant, on the contrary, the laws under which the subject 
 thinks. 1 If the reader has followed the exposition in the 
 previous pages, he will have seen that Kant's Categories are 
 just as much as Aristotle's, ' the highest classes to which 
 objects of our knowledge could be generalised,' nor was 
 Kant likely to make so silly a blunder as to assert his 
 intention to be the same as Aristotle's, if it were only to be 
 contrasted with it. This is not the only place in which 
 commentators have had the audacity to contradict the plain 
 statements of the Kritik as to Kant's intentions. 
 
 But as Kant says, the execution of the plan differs 
 widely from Aristotle's. For the above is the list of all 
 the pure notions of combination, or synthesis, which the 
 understanding contains, and through which it is an under- 
 standing, for by them alone can it tmderstand the variety of 
 intuition, by grouping it into objects. This table then and 
 its divisions are deduced from one general principle (that 
 of judgments) instead of being picked up at random by 
 mere empirical induction, which can neither certify the 
 completeness of the list, nor give any reason why these and 
 no other should be in the pure understanding. Hence 
 Aristotle's table, even with the post -predicaments, 2 was 
 never complete ; it includes modes of sensibility (quando, 
 ubi, situs, etc.), an empirical notion {motus), and even 
 deduced concepts {actio, passid)? 
 
 1 Cf. Note A to Reid's Works, p. 762. 
 
 2 Viz. oppositum, simul, prius, motus, habere 
 
 3 Mansel considers these criticisms to be wrong, for (1) the post- 
 
96 DEFINITION OF THE CATEGORIES chap. 
 
 Apropos of the last-named notions, Kant observes that 
 the Categories, as generic concepts of the pure understand- 
 ing, have their pure deduced concepts under them, which he 
 proposes to call predicables of the pure understanding, e.g. 
 subordinate to the category of cause we should have force, 
 action, passion ; under community, presence, resistance ; under 
 modality, origination, extinction, change, etc. To draw out 
 a table of these would be here to turn from his object, and 
 he therefore passes the matter by with this notice. He 
 also declines to define the Categories, as he wishes to avoid 
 collateral controversies, 1 and is only here concerned with 
 analysing them for his special purpose, not building up a 
 system of pure reason, in which such definitions would be 
 fairly demanded. He considers that with the basis here 
 supplied the task will not be difficult. 
 
 § ii. In the Second Edition, two sections of reflections 
 on the Categories were appended, which we shall pass over 
 as briefly as possible, as not belonging to his main argu- 
 ment. But we must call attention to the very first statement 
 in these sections, which contains a vindication of the 
 common charge brought against Kant, that he was enslaved 
 by the love of symmetry, and forced such Ideas as those of 
 
 predicaments were not so called by Aristotle — the term is of scholastic 
 origin, and was employed to denote the five subjects treated of by 
 Aristotle after the Categories proper. (2) The Categories, with Aris- 
 totle, belonged to the matter of thought, and were of course generalised 
 from experience. — Proleg. Logica, App. 
 
 1 In a passage of his First Edition, which will be found in vol. ii. p. 
 220 sqq., he goes more fully into the reasons why the mere Categories 
 cannot be defined. Cf. note, chap. xv. § 2. These passages were 
 modified in the latter Editions. Notwithstanding there are, in the 
 discussion of the Schemata of the various Categories (Krilik, pp. 110- 
 112, 181) various indications of the Categories, which amount almost to 
 Definitions, by means of the schemata. Cf. Kritik, p. 79, ' They are 
 concepts of an object in general, by means of which its intuition is con- 
 templated in relation to one of the logical functions of judgment." 
 
v THE CLASSIFICATION NATURAL 97 
 
 the world and God into the scheme supplied by the Cate- 
 gories and logical syllogisms. Not only is the table, says 
 Kant, necessary in the theoretical part of philosophy, both 
 in giving us a plan for a complete system of our science, as 
 well as its division on fixed principles ; but as it contains all 
 the elementary notions of the understanding, and even the 
 form of their systematic combination, so it must suggest for 
 any speculative science both the heads, and the arrangetnent 
 of them. Kant himself applies them to natural science 
 in another treatise. If the critics had objected to Kant 
 that he was violating the natural relations of things by 
 forcing them into his classification, he would answer that 
 not he, but nature, had prescribed it. All our thinking is 
 only comparison, all comparison is judging. Analyse there- 
 fore the possible forms of judgment, and you have all the 
 possible kinds of thinking. If this be so, the arrangement 
 according to the Categories (and of course according to the 
 syllogisms derived from them) is not artificial, but the most 
 natural and necessary possible, and moreover one which 
 will apply equally to all theoretical sciences. 
 
 Remark 1. The table naturally falls into two subdivisions, 
 the one referring to objects of intuition, the other to the 
 existence of these objects, either in relation to one another 
 or to the understanding. These he distinguishes as mathe- 
 matical and dynamical Categories. The second, as they 
 consist of relations of objects either to other objects, or to 
 us, have correlates. 1 
 
 Remark 2. Instead of a division by dichotomy, each 
 
 1 Mathematical Categories are concerned with the construction of 
 objects ; Dynamical with their connection either with each other {Rela- 
 tion) or with experience as a whole {Modality). Cf. chap, xi., infra, 
 sub fin. Kant's view, to which we adhere, differs from that of Mr. 
 Caird, cf. Phil, of Kant, p. 438. 
 
 I H 
 
EXPLANATIONS 
 
 class has three members, the third arising from a com- 
 bination of the first and second. So totality is nothing but 
 plurality considered as unity ; community the causation of 
 different substances in determining one another, and so of the 
 rest. Nevertheless, this third Category is no mere predicable, 
 or deduced concept. For it requires a distinct act of the 
 understanding. So our notion of a number (which is a 
 totality of several units) is not always possible, even when 
 we have the concepts of unity and plurality before us, as 
 for example in the case of infinity. Neither do substance 
 and cause explain, when merely combined, how one substance 
 can influence, or be the cause of something in another. 
 
 Remark 3. The agreement of the Category of Com- 
 munity with the corresponding form of the disjunctive 
 judgment is not at first sight obvious. But we have al- 
 ready seen (above, p. 87) that the sum of the predicates 
 in a disjunctive judgment makes up a total sphere of know- 
 ledge, in which the parts are not subordinate, but co-ordinate, 
 as mutually exclusive, being the parts of an aggregate. A 
 similar combination is conceived in the totality of things, 
 which are not subordinated to one another as effects are to 
 causes, but thought as co-ordinate, and affecting one another; 
 as, for example, in a body the various parts are co-ordinate, 
 and mutually attract and repel each other. The under- 
 standing pursues the same proceeding, when representing 
 to itself the sphere of a divided concept, and the parts of a 
 divided thing ; both consist of a whole, containing mutually 
 exclusive parts. 
 
 § 12. The celebrated position, Quodlibet ens est unum 
 verum bonum, though not included by scholastic philosophers 
 in the list of Categories, nevertheless holds among them a 
 position which ought to give it this status. It is worth in- 
 quiring whether some real though misunderstood rule of the 
 
v SCHOLASTIC ADDITIONS 99 
 
 understanding may not be at the basis of this principle, 
 though it is now antiquated and in desuetude in our 
 philosophical books. These pretended transcendental attri- 
 butes of things are nothing but logical requisites and 
 criteria of our concepts of things generally, and place the 
 Categories of Quantity at the basis of this cognition. But 
 though the schoolmen were using these Categories only 
 logically, in a formal sense, they incautiously raised them to 
 the position of the properties of things per se. In every 
 cognition of objects there is a unity of the concept, which 
 may be called qualitative unity, as when we speak of the 
 unity of a play, a speech, or a story. Secondly, there is 
 truth as regards its consequences. The more true con- 
 sequences follow from a given concept, the more evidences 
 have we of its objective reality. This may be called the 
 qualitative plurality. Finally, Completeness, in that this 
 plurality can be reduced to the unity of the concept, and to 
 it alone — this we may call its qualitative completeness or 
 totality. It appears, then, that the Categories of Quantity 
 in which the units that produce the quantum should be 
 thoroughly homogeneous, are here used to combine 
 heterogeneous parts of our cognition, by using the quality 
 of cognition as our principle. So the criterion of a 
 hypothesis (or other concept) is its unity, i.e. its not re- 
 quiring auxiliary hypotheses ; the truth, or agreement with 
 itself, and with experience, of the consequences deduced 
 from it ; and, finally, its completeness, in that the consequences 
 point back to this, and this alone, so that what we conceived 
 synthetically can be shown analytically. The list of Cate- 
 gories is not then to be increased by these pretended attri- 
 butes of things, which only come into existence by leaving 
 out all relation to things, and reducing our mere treatment 
 of concepts to general logical rules. 
 
KANT'S LIST CRITICISED 
 
 IT Kanfs Table of Categories, and his Critics. — It is 
 curious that the very ground upon which Kant attacks the 
 Categories of Aristotle has been urged as the particular 
 objection to Kant's own list. ' It was a design,' he says 
 (p. 65), 'worthy of an acute thinker like Aristotle, to search 
 for these fundamental concepts. Destitute, however, of 
 any guiding principle, he picked them up just as they 
 occurred to him. 1 Now, let us hear Schwegler ■} 'The 
 method of Fichte, just like that of Hegel afterwards, is a 
 combination of the analytical and synthetical methods, by 
 which Fichte earned the credit of having first deduced the 
 Categories of philosophy from one single point, and of 
 having brought them into connection, instead of taking them 
 merely empirically, and co-ordinating them, as had been 
 done, even by Kant.' The same view is taken by Mansel : 2 
 ' The Kantian Categories are not deduced from an analysis 
 of the act of thought, but generalised from the forms of the 
 proposition, which latter are assumed without examination, 
 as they are given in the ordinary logic. A psychological 
 deduction, or preliminary criticism, of the forms themselves, 
 might have considerably reduced the number.' And so 
 both Fichte and Mansel have given further analyses, which 
 the curious reader may find in the treatise just quoted of 
 Mansel's, and in Fichte's Wissenchaftslehre? These ana- 
 lyses are in substance the same, and consist in identifying 
 quality with quantity, and discarding relation and modality, 
 on the principle that substance and cause are implied in 
 them, and that these notions exclude them from the first 
 rank. We suspect that, upon a careful perusal of Mansel's 
 discussion, the reader will be glad to fall back upon Kant's 
 plainer classification, and will agree with him in not taking 
 
 1 Ed. Stirling, p. 262. 
 2 Metaphysics, p. 193, note. 3 Works, vol. i. p. 166 sqq. 
 
DEFENCE OF KANT 
 
 any interest in the subtleties of modern philosophers on 
 the subject. 1 There is, however, one charge from which 
 Kant must be cleared, and that is, that he did not go upon 
 a fixed principle in his table. His introduction to the sub- 
 ject is quite explicit. ' Transcendental philosophy,' he 
 says (Kritik, p. 56), 'has the advantage, and moreover the 
 responsibility, of searching for its concepts upon a principle, 
 because they originate pure and unmixed from the under- 
 standing, as an absolute unity, and must hence be con- 
 nected according to one concept or notion. Such a 
 connection gives a rule,' etc. 
 
 What is the principle according to which we must 
 proceed ? He shows that the Understanding has no power 
 of intuition, and hence can only regulate and bring into 
 classes and unities the intuitions given by our sensibility. 
 This spontaneous faculty he calls the function of the Un- 
 derstanding. And what is the only use we can make of 
 these unities ? To judge by means of them. And how do 
 we judge by means of them ? We repeat the process by 
 which they have been already formed, and bring an ad- 
 ditional representation under them. The understanding 
 has no other duty at all ; hence it may be simply called 
 our judging faculty. This is the a priori argument and 
 principle upon which he bases his Table of the Categories ; 
 so that, in this sense, his list is neither purely empirical, nor 
 picked up at random. 
 
 The number of the classes of judgments he did take for 
 
 1 The great diversity of philosophers as to the reduction of Kant's 
 Categories is remarkable, and is an argument against such reduction. 
 Cousin reduces them to substance and cause ; Fischer and Schopenhauer, 
 to cause only ; Hamilton to Condition, which appears to be the Cate- 
 gory of cause without the schema, or of relation generally. When 
 philosophers difter so widely, it may be well to inquire whether any 
 remedy is really required. 
 
IS A REDUCTION FEASIBLE ? 
 
 granted, from the existing treatises on logic (which, we 
 suppose, discovered them empirically) ; but this because 
 there could not be a class of judgments without a corre- 
 sponding expression for them in human language, and the 
 grammatical analysis of language is long since completed ; 
 and because he saw distinctly that psychologically they 
 depended upon different acts of the mind. 1 That it was 
 possible to reduce them in number, was a point which came 
 distinctly before him, and which he combats in his obser- 
 vations on the Table of Judgment ; and in further remarks 
 (p. 67), he even insists on some judgments, which are 
 logically reducible under one head, being kept apart as 
 psychologically distinct. It is not fair then to charge Kant 
 with having evaded or overlooked a more complete psycho- 
 logical deduction ; but we must rather place his authority 
 (and his psychological acumen) over against those of the 
 critics, and supposed improvers, of his system. 
 
 It is obvious that two sorts of reduction are possible : 
 we may either reduce the number of the Categories under 
 each head, or we may reduce the various heads or classes 
 to a lesser number. The first description of reduction has 
 been (as was observed already) noticed and rejected by 
 Kant. The second has been attempted by Mansel. Now, 
 that there exists an analogy between the classes of Categories 
 would be naturally suggested, and probable, from the unity 
 of the pure reason, upon which Kant insists frequently, and 
 this would also suggest the same number of judgments 
 under each head. But the question remains — Is this 
 similarity Identity, or merely Analogy t Kant could only 
 regard them as identical, if the quantity and quality of 
 judgments were proved identical. Take, for example, the 
 
 1 Cf. for another defence of Kant, Stirling in Schwegler's Hist, of 
 Phil. p. 423. 
 
v MANSEL CRITICISED 103 
 
 supposed identical Categories of unity and reality. Because 
 affirmation asserts unity between two representations, can 
 we jump at the conclusion that affirmation is identical with 
 unity ? Certainly not ; an asserted unity between repre- 
 sentations has nothing to do with the Category of unity, 
 derived from singular judgments. Of what does a judgment 
 consist ? Of a subject, a predicate, and a copula. What 
 can we say about the subject ? It may be either one, or 
 many, or a totality (the many regarded as a unity). What 
 about the predicate ? We may assert it to be identical 
 with this one, or many, or whole, or the reverse. How can 
 this act of mind be declared the same as the former ? If 
 the predicate of a judgment were singular, and we affirmed 
 it of any sort of subject, we should be much nearer the 
 Category of unity. 
 
 So, again, in a negative judgment we regard one attribute 
 as not co-existing with another ; but here, if we take a 
 singular judgment, viz. ' Socrates is not foolish,' we do not 
 necessarily imply other subjects which have this attribute, 
 and hence we do not obtain plurality. But supposing a 
 class were here implied, it would surely be just as much 
 implied in the corresponding affirmative judgment, which 
 would accordingly suggest plurality as much as unity. 
 Possibly Mansel was misled by his own statement, that in 
 a judgment two concepts are considered ' in relation to a 
 common object of intuition.' Perhaps the correct expression 
 would be, ' in relation to common objects' viz. how far the 
 objects which rank under one of these concepts rank also 
 under the other. If so, to think the co-existence or non- 
 existence of attributes in one or more subjects is obviously 
 distinct from thinking the unity or plurality of these subjects 
 themselves. The former are, indeed, unifying and dividing 
 processes, but so are all functions of thought, as Kant has said. 
 
104 ANALOGY IS NOT IDENTITY chap, v 
 
 It would be tedious in this place to urge all the similiar 
 objections which could be made in detail to Mansel's 
 reduction. But in general, except we can reduce the 
 psychological acts expressed in the various classes of judg- 
 ments to the same act, we have only demonstrated analogy, 
 and not identity. The attempts, then, of Fichte and 
 Mansel corroborate Kant's view of the symmetry and 
 harmony between the various acts of the understanding as 
 one complete whole ; for these analogies are strong enough 
 to suggest to acute minds complete identity. 
 
CHAPTER VI 
 
 INTRODUCTION TO THE DEDUCTION OF THE PURE CONCEPTS 
 OF THE UNDERSTANDING. SECTION I. 1 
 
 § 13. Of the Principles of Transcendental Deduction 
 generally. — Jurists, in discussing claims, are wont to dis- 
 tinguish the qucestio facti from the qucestio juris, and used to 
 call the proof of the latter the Deduction — that is to say, the 
 deduction of the claim from acknowledged principles, or 
 documents. We use a number of empirical concepts 
 without any such justification being required by ourselves 
 or others, because experience is always at hand, to prove 
 their objective reality. Yet here, too, there are some con- 
 cepts, such as luck and fate, which, though commonly 
 recognised, cause us great difficulties when the question 
 quid juris is asked as regards them, seeing that neither 
 experience nor reason affords the grounds of answering it. 
 
 But among the various concepts in the complex tissue 
 of our knowledge there are some meant for pure a priori 
 use, and their claims always require a deduction, for proofs 
 from experience are here insufficient, and yet we must know 
 how these pure concepts can refer to objects. The ex- 
 planation of the way in which they do so Kant calls their 
 
 1 The heading is : 'The Transcendental Analytic,' book i., part 2. 
 
106 WHAT IS DEDUCTION ? chap. 
 
 transcendental deduction, as distinguished from the empir- 
 ical, which shows how a concept is acquired by experience 
 or reflection upon it — that is to say, the facts from which 
 our use of it arose. 
 
 There are two very diverse kinds of notions which agree 
 in referring a priori to objects — the notions of Space and 
 Time as Forms of Sensibility, and the Categories as con- 
 cepts of the understanding. If a deduction of these (or 
 proof of their possibility) is necessary, it must, of course, be 
 transcendental, as their distinguishing feature consists in 
 applying to objects, without drawing anything from ex- 
 perience in order to represent them. 
 
 Of course, we may inquire quite independently into the 
 occasions when both this part and the rest of our know- 
 ledge arose in experience — an experience which contains 
 two dissimilar elements, the Matter given by the senses, 
 and the Form for ordering them, springing, on the occasion 
 of the former, from pure intuition and from thought. 
 Locke deserves the credit of first opening this path. But a 
 deduction of pure a priori concepts can never be thus 
 obtained. For then we only explain from experience the 
 fact that we possess pure cognition, whereas these concepts 
 must exhibit a very different pedigree from that of empirical 
 concepts. No one, therefore, who understands their nature 
 can accept anything but a transcendental deduction of 
 them. 
 
 But granting all this, is such a deduction absolutely 
 necessary ? We have, indeed, pursued the notions of space 
 and time to their source, by means of such a deduction, 
 and so explained and fixed their objective value. Yet 
 Geometry follows its course through nothing but a priori 
 cognitions, without asking for any certificate of the 
 legitimacy of its fundamental notion — space — from philo- 
 
vi IS DEDUCTION NECESSARY ? 107 
 
 sophy. But the use of this notion here applies to the 
 external world of sense, where all geometrical cognition is 
 immediately evident, being based on a priori intuition, in 
 which the objects are given (as to form) by our cognition. 
 When we come to the pure concepts of the understanding, 
 the necessity becomes apparent, not merely of justifying 
 them by a deduction, but of giving a deduction even of the 
 notion of space. For these concepts refer to objects not 
 through intuitive predicates, but through pure thinking. 
 They refer to them generally, and are neither based on 
 experience, nor can they show any object in a priori 
 intuition, affording a basis for their synthesis. Hence they 
 not only excite general suspicion about their objective 
 validity, and the limits of their use, but even involve the 
 concept of space in this suspicion, 1 as we are disposed to 
 apply it beyond the conditions of sensuous intuition. This 
 is the reason why a transcendental deduction 2 (of space) 
 was given above, i.e. to show that space and time have no 
 application beyond experience. The reader who is not 
 convinced of the inevitable necessity of the deduction of 
 the Categories also will only grope in the dark, and end as 
 ignorant as he began. But he must also perceive the 
 obscurity and difficulty of the investigation, and not 
 complain of weariness in its solution. For upon this it 
 
 1 As Mr. Caird puts it : 'If we have general conceptions which are 
 independent of sensible experience, it is not at once obvious that the 
 knowledge we get by their use should be confined to the objects given 
 by sense. And not only so, but the mind which is possessed of such 
 conceptions is inevitably driven to consider whether the forms of time 
 and space also, inasmuch as they give rise to necessary and universal 
 perceptions, may not apply to objects in general, irrespective of sensible 
 experience.' — Phil, of Kant, p. 320. 
 
 2 He calls it a transcendental exposition in the Aesthetic above, 
 
 P- 51. 
 
THE DIFFICULTY STATED chap. 
 
 depends whether we are to give up all our claims to any 
 possessions in the favourite department of metaphysic, 
 beyond the bounds of all experience, or bring this critical 
 investigation to a successful issue. 
 
 There was no great difficulty in showing (in the 
 Aesthetic) how the notions of space and time, though a 
 priori cognitions, were yet related necessarily to objects, 
 and rendered a synthetical cognition of them possible, 
 independent of all experience. No object can appear in 
 empirical intuition except by means of these pure forms ; 
 hence they are pure intuitions, which render objects possible 
 as phenomena ; their synthesis is therefore objectively 
 necessary. 
 
 But as phenomena 1 can appear in intuition, without 
 being necessarily related to the functions of the under- 
 standing, a difficulty arises which did not occur in the 
 Aesthetic, viz. How are subjective conditions of thought to 
 have objective validity, or be the necessary conditions of all 
 cognition of objects ? Take, for example, the concept of 
 cause, which signifies a peculiar kind of synthesis, by which 
 we put after A some quite different B, according to a fixed 
 rule. It is not obvious a priori (empirical evidence being 
 inadmissible), why phenomena should contain anything of 
 the sort, and we may doubt whether such a concept is not 
 idle, and baseless in experience. Objects would not be 
 objects did they not conform to space and time, but it is 
 not so clear that they must conform to the conditions which 
 the understanding requires to produce unity in its thinking. 
 Phenomena could appear without any such conformity. It 
 has been suggested (by Hume and others) that experience 
 is perpetually offering us examples of these regularities in 
 phenomena, which are quite sufficient to suggest the notion 
 1 See p. 47 for the definition of Phenomenon. 
 
vi TWO CONDITIONS OF KNOWLEDGE 109 
 
 of cause, and establish its objective validity. Kant thinks 
 that such concepts must either show an a priori basis, or be 
 abandoned as mere chimeras. He points, as usual, to the 
 strict necessity of the sequence of cause and effect, and the 
 absolute universality of its application, as attributes which 
 cannot possibly be given by experience. In fact the 
 Categories cannot be derived from experience, for that very 
 experience implies them. 
 
 Transition to the Transcendental Deduction 
 of the Categories 
 
 That our synthetical representation should necessarily 
 correspond or coincide with its objects is only possible in two 
 cases ; either if the object makes the representation possible, 
 or if the representation, and it alone, makes the object 
 possible. In the former case, the relation is empirical, and 
 the representation could never be possible a priori. This is 
 the case with phenomena, so far as they are sensations. 
 In the second case, our representation does not, indeed, 
 cause the object to exist (if we except the causality of the 
 will, to be elsewhere discussed), but nevertheless determines 
 it a priori, if through the representation alone we can 
 know anything as an object. This knowledge requires two 
 conditions necessarily : intuition, by which the object is 
 given merely as an appearance ; and a concept, by which 
 an object is thought, corresponding to the intuition. We 
 have above explained how all phenomena must necessarily 
 agree with our intuition, as they only appear (and are 
 intuited) through it. Now comes the question, whether 
 there are not also concepts within us, as antecedent condi- 
 tions a priori, under which alone anything can be thought 
 as an object. If so, all empirical knowledge of objects 
 
LOCKE'S ATTEMPT 
 
 must correspond to these concepts necessarily, or else no 
 object of experience is possible. But as soon as we go 
 beyond the mere data of the intuition of the senses, all our 
 experience does contain concepts or notions of objects given 
 in intuition, or appearing in it ; so then concepts of objects 
 in general do lie at the basis of all our empirical cognition, 
 as a priori conditions. Consequently the objective validity 
 of the Categories, as a priori concepts, depends on this, 
 that through them alone experience (as far as the form of 
 thinking is concerned) is possible. Of course, they have a 
 necessary relation to objects of experience, if it is only 
 by means of them that such an object can be at all 
 thought. 
 
 This then is the principle on which the transcendental 
 deduction of all a priori concepts depends. To explain 
 the development of experience in which they occur, is not 
 a deduction, but an illustration of them, as it does not 
 prove them necessary. Without showing their primitive 
 relation to all the possible experience in which objects 
 occur, their relation to any single object could never be 
 comprehended. 
 
 In his First Edition, Kant closes this preface to the 
 Deduction with a mention of the three faculties of the mind 
 that afford the clue to the succeeding discussion, 1 viz., sense, 
 imagination, and apperception. In the Second Edition he 
 substituted the following observations. ' The celebrated 
 Locke,' he says, ' ignoring totally any such deduction, and 
 finding the pure concepts of the understanding in experi- 
 ence, deduced them from it, and was yet so illogical, that 
 he attempted in this way to reach cognitions far beyond the 
 
 1 This passage, for which see Fischer's Comm. p. 76, note, merely 
 anticipates the after discussion, and therefore requires no farther mention 
 here. Cf. also Appendix A, vol. ii. 
 
LOCKE AND HUME 
 
 limits of experience.' 1 David Hume saw that this latter 
 attempt necessarily required concepts with an a priori 
 origin. But as he could not see how the understanding 
 came to think certain concepts as necessarily joined in the 
 object, which were not in themselves joined in the under- 
 standing, and as it did not occur to him that the under- 
 standing might, through these very concepts, be the 
 originator of experience, he was compelled to derive them 
 from experience, and through constant association, which 
 produces custom or subjective necessity, falsely deemed 
 objective. He was logical enough to deny that with such 
 principles we could ever pass the bounds of experience. 
 But these empirical deductions are refuted by the fact, that 
 we actually possess scientific cognitions a priori, viz. pure 
 mathematic and general physic. 
 
 The former of these celebrated men opened the door to 
 enthusiasm, for if the reason has pretentions or claims on 
 its side, it will not be restrained by vague exhortations to 
 moderation ; the latter abandoned himself to scepticism, as 
 he thought he had discovered so universal a delusion 
 mistaken for sound reason in our faculty of knowledge. 
 We are now to essay whether the reason cannot be steered 
 safely between these rocks, and whether we cannot save for 
 it the whole sphere of its proper activity, while assigning to 
 it fixed and determinate boundaries. 
 
 1 This is the ordinary view of Locke's philosophy, which ignores the 
 intellectual side of his system. But this is not the place to correct it. 
 Cf. Prof. Webb's Intelkctualism of Locke, passim. Cf. also Schwegler, 
 Hist, of Phil. p. 182, who makes this criticism on Locke's proof of the 
 Existence of God. 
 
CHAPTER VII 
 
 THE DEDUCTION OF THE CATEGORIES. THE FIRST 
 EDITION, AND THE PROLEGOMENA. 
 
 H We have now arrived at the great crux to most readers 
 of the Kritik, the famous Deduction of the Categories. 
 The solution has been, to some extent, implied in the fore- 
 going discussion, and the intelligent student will, doubtless, 
 anticipate the gist of Kant's argument ; but we must not 
 be content with stating the theory ; we must also give 
 some comparative account of the various forms which the 
 discussion assumed in the First Edition of the Kritik, and 
 in the Prolegomena, published in 1783. Were this omitted, 
 the reader might ask why Kant had obscured by prolixity 
 and by repetition a discovery in itself plain and comprehen- 
 sible, and which can be stated in a brief compass. 
 
 H Perhaps the first and most important help the stu- 
 dent can obtain from a commentary on this part of the 
 Kritik is to have the repetitions in Kant's argument care- 
 fully pointed out. The discussion in the First Edition 
 goes over the same ground three times. As he tells us 
 himself, 1 ' I have thought it better, in the four following 
 paragraphs, rather to prepare than instruct the reader, and 
 
 1 Vol. ii. p. 194. 
 
CHAr. vn REPETITIONS IN KAXTS ARGUMENT 113 
 
 not to lay before him the systematic discussion till the 
 succeeding third section.' All the momenta of the proof 
 are, however, contained in this preparatory discussion ; x 
 and he accordingly, in opening the third section (p. 207), 
 says, ' The detached observations made in previous section 
 (containing the four paragraphs) we shall here unite and 
 present in a connected form.' The reader will, therefore, 
 find in this section a repetition of the observations in the 
 previous section, not, however, without some modifications. 
 For having hitherto pursued the analytical method, starting 
 from empirical perceptions, and passing up by analysis to the 
 a priori elements contained in them, he now tells us (p. 
 207) that he will 'begin from pure apperception' — in 
 other words, give us the same proof synthetically. He does 
 this briefly and completely (pp. 207-210) ; but having done 
 it, he reverts to his former analytical procedure, and says 
 (p. 210), ' We shall now expound the necessary connexion, 
 etc., beginning from below, from the empirical extremity.' 
 He then goes over the very same ground, and in the very 
 same order, as in the four preparatory paragraphs, amplify- 
 ing a little here and there, modifying a few expressions, but 
 adding little, except (as we shall see) that he shows the 
 close relation between the three faculties expounded, and 
 so brings into a connected form the observations before 
 detached. If the reader will attend to these hints, he will 
 considerably curtail his labour, and save himself the per- 
 plexity of endeavouring to find new arguments, where Kant 
 is merely repeating and enforcing old ones. 
 
 As the Second Edition differs mainly from the First in 
 developing at great length the synthetical proof, very briefly 
 given at first (pp. 207-2 10), 2 we shall consider that part of 
 the discussion by the light of the Second Edition, and 
 
 1 VoL ii. pp. 194-207. " Cf. pp. 120, 121, infra. 
 
 I I 
 
114 WHAT IS EXPERIENCE ? chap. 
 
 shall give an account, as brief as possible, of the First 
 Edition, omitting this short passage. 
 
 A few words will here show us the attitude of the Pro- 
 legomena on the same subject, in a passage following the 
 analytical method also, as Kant expressly tells us at the 
 outset of his shorter work. We already called attention to 
 the twofold character of the Categories, which are both 
 general concepts, or frames of objects of intuition, and also 
 pure general forms of judgments. These two functions are 
 closely related, for the frames into which the former bring 
 intuitions are necessary conditions of these intuitions be- 
 coming fit for judgments ; our intuitions are, as Kant says, 
 determined by these Categories, in relation to some one of 
 the pure forms of judgments. 1 The deduction of the 
 Categories need therefore only establish their objective 
 necessity in either of these relations, and the other will 
 necessarily follow. For when we speak of the Categories 
 being necessary for our experience, what do we mean by ex- 
 perience ? We mean a great complex, embracing a vast num- 
 ber of objects, and we also mean the legitimate and orderly 
 connexion of these objects into a great harmony, or unity. 2 
 This connexion of objects, which implies certain necessary 
 relations among them, can only be expressed or conceived 
 in judgments concerning objects. If the Categories are 
 necessary for the formation of the judgments of experience, 
 it is clear that they must also be necessary for the objects 
 of these judgments, since nothing can be for us an object 
 except it be either the subject or predicate of some judg- 
 ment. The necessary laws, therefore, of the connexion of 
 objects must hold good of these objects themselves. Such 
 an inquiry Kant calls a deduction of the possibility of [the 
 
 1 See especially his statement in the Prolegomena, vol. ii. p. 73. 
 2 Cf. Proleg. p. 65. 
 
vii ARGUMENT IN THE PROLEGOMENA 115 
 
 faculty of] experience, as contrasted with a deduction of 
 the possibility of the objects of experience. The latter 
 side of the deduction had been brought forward promi- 
 nently in the First Edition, and it is only : in the two 
 summaries of the discussion that he notices the power 
 of the understanding to make laws for nature — in fact, to 
 establish necessary connexions among the objects of our 
 experience. This latter is then the aspect of the Categories 
 which he takes up in his Prolegomena 2 in this way. 
 
 Starting from the statement that Nature means a neces- 
 sary synthesis of phenomena, not of things per se, he shows 
 (§15) that there is a pure science of such nature, which 
 possesses universal and necessary synthetical judgments 
 applicable to all nature, whether internal (psychology) or 
 external (physics). These judgments are the laws of 
 nature. But the word nature, he adds (§ 16), also means 
 the complex of all the objects of experience. He proceeds 
 to inquire (§17) which form of the problem (the same in 
 either case) is preferable, and he decides that, owing to the 
 ambiguity of the word object, the Kantian sense of which 
 was then new and strange, 3 it is better to take the formal 
 side, and discuss the possibility of [the faculty of] experi- 
 ence, upon which the possibility of the objects of experi- 
 ence necessarily depends. Following this course, he shows 
 (§ 18) that our ordinary judgments of perception cannot 
 
 1 In the fourth paragraph, in section 2, and at the end of section 3, 
 especially p. 203, vol. ii. 
 
 2 Vol. ii. pp. 63 sqq. 
 
 3 An object, in the proper Kantian sense, consists of a number of 
 sensations, bound up into a necessary unity. As this latter element 
 cannot be given by mere sense, an act of the understanding, or of the 
 imagination, is necessary in knowing any object. As the former ele- 
 ment can only reach us through our sensibility, objects only exist in our 
 experience, and things per se are improperly called by the same name. 
 This will appear more fully in the sequel. 
 
n6 JUDGMENTS OF EXPERIENCE 
 
 become judgments of experience without expressing a 
 necessary connexion, or a connexion in the object, as it is 
 called, when all men agree in the connexion. But this 
 objective validity or necessity is merely equivalent to uni- 
 versality. What, then, must be added (§ 20) to the per- 
 ceptive judgment, to make it a necessary judgment of ex- 
 perience ? Simply this : the perception in question must 
 be subsumed under such a concept as determines for it a 
 place among the general forms of judgment. 1 These 
 general forms of judgment will, therefore, suggest the table 
 of pure concepts which correspond to them. The whole 
 
 1 It is very difficult to reconcile Kant's language on p. 69 of the 
 Prolegomena with his general doctrine that all judgments have a 
 category at their basis. What he seems to mean is that judgments of 
 perception only involve the Mathematical Categories, judgments of ex- 
 perience the Dynamical as well. This is the common sense of the 
 matter, and it is quite consistent with the examples he gives (pp. 71-74). 
 As to these there is no difficulty except in the case of the proposition, 
 'the air is elastic' He arrives at this in a peculiar way, and himself 
 confesses that the illustration is obscure. If we compare the analogous 
 case of 'the sun warms the stone,' discussed in his note (p. 74), we 
 shall see that he considers the proposition ' when the sun shines on the 
 stone, it grows warm,' to express the attitude of the mind in mere per- 
 ception, whereas the categorical equivalent marks the classing of the 
 representations under an a priori concept. But the other case is not so 
 clear. Kant indicates that after we have established the air to be the 
 necessary condition, or cause of expansion, we then advance to the 
 judgment, ' the air is elastic,' in which we regard elasticity as a quality 
 necessarily belonging to the air. Mr. Monck has suggested that Kant 
 had the experiments of Torricelli in his mind, by which both the weight 
 and the elasticity of air were demonstrated. It was shown that if the 
 pressure on the atmosphere be diminished, its volume increases, as, for 
 example, if a partially filled bladder be placed in a vessel from which 
 the other air can be removed by a pump, the bladder will become fully 
 distended, or even burst. This expansion depends on the nature of the 
 air itself (requiring only the absence of counteracting causes) as its 
 positive cause. This explanation is probably the true one, and clears 
 up the difficulty as to Kant's illustration. 
 
vn ARGUMENT IN THE FIRST EDITION 117 
 
 argument is clearly and concisely summed up in § 23, 
 which the reader is recommended to study carefully. It is 
 briefly this : We have objectively valid judgments, as dis- 
 tinct from mere judgments of perception. These are only 
 possible by means of the Categories, which have therefore 
 objective validity. 
 
 We now revert to the exposition of the First Edition 
 (vol. ii., Appendix A), for the most part analytical, like that 
 of the Prolegomena, but differing, as we have said, in two 
 points : (a) there is a short synthetical exposition at the 
 opening of section 4 ; (/?) with the exception of the con- 
 cluding passages in the duplicate analytical exposition, the 
 Categories are rather considered as pure concepts of ob- 
 jects, than as pure forms of judgments combining objects. 
 If we omit these passages, and eliminate repetitions, we 
 may sum up the argument as briefly as possible in the 
 following way. 1 
 
 A concept (p. 191) is nothing but a combination of 
 attributes. If this combination is to mean anything more 
 than a mere empty frame, the attributes must be supplied 
 by intuition ; thus only can our concepts refer to the sole 
 objects we can know, the objects possible in our experience. 
 If then there be in us a priori concepts, even they must 
 refer to experience, not of course as its offspring, but as 
 rendering it possible ; on this ground only can they be 
 objectively valid. The possibility of experience, therefore, 
 is the real point to be investigated. Whatever attempts we 
 may make to grasp such notions as God or spirits by the 
 aid of the pure concepts or Categories at the basis of 
 experience, we must still conceive these things as objects, 
 and therefore start from the same basis as we do in our 
 
 1 We cite here the pages of the English version in vol. ii., Appendix 
 A, so that the reader may verify the commentary at every step. 
 
n8 SUBJECTIVE CONDITIONS FOR OBJECTS chap. 
 
 legitimate experience. The Categories, therefore, will be 
 sufficiently justified, and proved objectively valid, ' if we 
 prove that through them alone an object can be thought.' 
 (Here, then, he keeps out of sight the Category as a pure 
 form of judgment, and treats it as a pure frame of objects 
 of intuition.) But as other faculties are concerned, Kant 
 proceeds to expound the subjective sources which make 
 objects possible, and how far these sources are of tran- 
 scendental use, or at the basis of our experience of objects. 
 Let the reader hold fast to Kant's prefatory remark 
 (p. 194) that all our representations are bound together by 
 one bond at all events, that of Time, the formal condition 
 of our internal, and therefore indirectly of all our sense. 
 All our representations are brought into at least one mutual 
 relation, in Time. But (1) as all our knowledge of objects 
 requires a successive series of representations (both separate 
 in space and differing in kind), there must be in the mind 
 a power to grasp these separate details, and consider them 
 as one complex object of intuition. This act (which is a 
 function of our imagination applied to present objects of 
 sense) Kant calls the synthesis of apprehension. And since 
 even pure space and time, regarded as objects, cannot come 
 into the mind without their parts being grasped together in 
 this way, though they are a priori representations, it follows 
 that this synthesis is possible a priori, and there is in us a 
 pure as well as an empirical synthesis of apprehension. (2) 
 There is, moreover, another combinative faculty of mind, 
 which causes past representations to come back to us in 
 groups, and not singly. Our imagination reproduces them 
 according to a law of association, which implies some prior 
 affinity existing among them. This reproduction of past 
 phenomena is equally necessary to our experience, for we 
 could not think of a large number, or a long time, if we 
 
vn TRANSCENDENTAL APPERCEPTION 119 
 
 lost from our minds the earlier numbers or moments with- 
 out recovery. This faculty of the imagination may then be 
 called the Reproductive synthesis of the Imagination. 1 Even 
 this, however, is not enough. (3) When we have appre- 
 hended the present details, when we have reproduced the 
 past impressions, what guarantee have we that they are 
 identical with those formerly intuited ? They must be 
 recognised in the concept we form of them — a concept which 
 reduces this multiplicity to unity, and declares them to be 
 the phenomena of a single identical consciousness. When 
 we have produced this unity, we call it an object. What is 
 this object ? As we can know nothing beyond our repre- 
 sentations, an object can only mean a necessary combination 
 of them — a combination that cannot be made at random, or 
 capriciously. But such necessary combination can only be 
 produced by the understanding, for the senses merely give 
 us isolated representations. Thus an external object of 
 these representations, such as body, is conceived just as we 
 conceive (p. 199) a triangle to be an object, and yet this 
 consists in nothing but a certain necessary combination of 
 three right lines, under a concept, and may not exist in 
 external nature. 2 If the combination be necessary, it must, 
 according to Kant's fundamental principles, depend on a 
 transcendental condition. How can the necessary unities, 
 which produce in us the notion of objects, be produced a 
 priori by the understanding ? 
 
 The required condition Kant finds in the Transcendental 
 apperception. This does not mean the empirical conscious- 
 ness of self, given in what is called the internal sense, but 
 the transcendental condition, which renders this internal 
 sense possible ; that is to say, the mental unity, or identity, 
 
 1 Cf. p. 133, infra, for the further account given in the Second Edition 
 of the office of the imagination. 2 Cf. vol. ii. p. 215. 
 
OUR NOTION OF OBJECTS chap. 
 
 which we feel even when intuiting such a priori objects as 
 space and time. All representations must be mine, they 
 must come under the identity of my permanent conscious 
 self. Yet it is not the existence of this self, but its action or 
 function, of which we are conscious. 1 We mean no more 
 by this feeling of personal identity than that it is the same 
 identical function which combines phenomena into objects 
 in all our experience. We cannot, therefore, be conscious 
 of the identity of self except as the ' unity of the synthesis 
 of all phenomena,' of course according to some plan, or 
 some concepts yet to be determined. 
 
 Let us now turn back and consider what our notions are 
 of object in general, and what we mean by a concept having 
 objective validity. Our thoughts are said to refer to objects, 
 when they apply to our intuitions, which again are supposed 
 to refer immediately to (phenomenal) objects. But even 
 these objects are only phenomena, and are, therefore, re- 
 ferred further to a transcendent object, which is a mere 
 indefinite supposition of thought, not obtainable by any 
 intuition, and therefore the same in all cases, a mere un- 
 known quantity, or x. Our whole notion of a definite 
 object is a number of intuitions, necessarily combined. As 
 the determinate intuitions are ex hypothesi absent in a gene- 
 ral notion of object, what remains but the necessary com- 
 bination, or synthesis, the framework, so to speak, of ihe 
 intuitions ? If this be so, the transcendental apperception 
 just explained affords us the necessary bond of unity, or 
 synthesis, and supplies us with the general notion of object, 
 
 1 Observe that it is the business of internal intuition to inform us 
 how we exist. As Mr. Caird puts it : ' We are thus conscious of our- 
 selves as a transcendental subject {that we are) because we distinguish 
 our own activity from the matter on which it is exerted ; but we are 
 conscious of ourselves as a phenomenal subject (what we are) only 
 under the form of time as distinguished from space.' Cf. p. 139, infra. 
 
vii OBJECTIVITY OF THE CATEGORIES 121 
 
 which contains nothing else. This, then, is the form of 
 phenomena, as space and time are the form of intuitions. 
 
 Kant adds a paragraph explaining how the Categories 
 may be a priori cognitions, and enforcing this truth. When 
 I speak of my experience, I mean one great unity, one vast 
 combination of all the phenomena presented to me as my 
 phenomena. Except they belong to this unity, they are no 
 part of my experience. But whatever conditions are im- 
 posed on my (faculty of) experience must, of course, be 
 equally imposed on the objects of experience acquired 
 through that faculty. But the Categories have above been 
 shown the necessary conditions of judging, therefore of 
 thinking, in any experience ; therefore they afford to objects 
 also of experience their possibility. Thus they are a priori 
 cognitions, and yet objectively valid. The Categories are, 
 in fact, phases of the pure apperception, and all conscious- 
 ness must be subject to it. Through it, therefore, or under 
 it, they legislate for phenomena. 
 
 All empirical derivations of these Categories are idle, for 
 they do not account for their necessity ; and secondly, they 
 postulate an universal law of association among phenomena, 
 which suggests to us laws, and they give no explanation of 
 this association. How can phenomena be conceived as 
 related to one another by a thoroughgoing affinity, which 
 alone explains our associating them ? On Kantian prin- 
 ciples, this affinity is transcendental, and produced by their 
 all being subject to one condition — the pure apperception, 
 which binds them together by a pure synthesis. So it 
 comes that the understanding prescribes laws for nature. 
 Hence we can make a priori assertions about nature, and a 
 strict science of nature is possible. 
 
 But the reader who desires to avoid repetitions, and 
 obtain the shortest possible exposition of the analytical 
 
122 OFFICE OF THE IMAGINATION chap, vii 
 
 deduction, may pass by the four preparatory paragraphs, 
 and confine himself to the repetition of their substance 
 in pp. 210-217. For there, as Kant tells us, he has con- 
 nected and brought into relation the isolated facts of these 
 paragraphs. There it is that he declares the synthesis of 
 apprehension to be the work of the imagination, which is 
 necessary in knowing even ordinary objects (p. 211, note), 
 and so brings the synthesis of apprehension and reproduction 
 into relation. There it is more especially that he expounds 
 the relation of the imagination and pure apperception, 
 showing that association must be based on affinity ; that 
 this affinity is given by the union of all phenomena in one 
 consciousness, but that this synthetical unity of conscious- 
 ness, which is intellectual, can only act on sense by directing 
 the productive imagination to combine phenomena according 
 to fixed laws. (The Imagination is, in fact, the go-between, 
 which mediates between the pure understanding and the 
 phenomena ;| it is only reproductive as regards these pheno- 
 mena themselves, but productive as regards the order in 
 which they are reproduced. Kant here trenches upon the 
 ground to be hereafter occupied in the schematism. We 
 shall therefore say nothing more concerning it now. Finally, 
 the statement that the understanding prescribes laws for 
 nature is more fully developed in the end of the discussion 
 (pp. 215-217). He adds a caution, that we cannot suppose 
 empirical laws to be directly deducible from the pure 
 Categories, any more than all the varieties of intuition 
 could be conceived from pure space and time. But never- 
 theless, according to this very analogy of space and time, 
 all empirical laws must be built on the type of the Categories, 
 — in other words, all our experience, as to form, must be 
 conformable to them. More than this our deduction never 
 attempted to prove. 
 
CHAPTER VIII 
 
 THE DEDUCTION OF THE CATEGORIES." "* THE SECOND 
 EDITION OF THE KRITIK 
 
 11 We now approach the discussion in its ultimate form, as 
 it appears in all the later Editions of the Kritik. The first 
 point to be noted is that in his first Preface, he had himself 
 carefully distinguished two sides of his Deduction — one con- 
 sisting of a view of the subjective faculties of the mind, 
 the other a mere explanation how objects can become 
 possible. This latter is the proper Deduction ; the former, 
 which consists in assigning a particular set of causes for a 
 given effect, is not so. 1 It was, therefore, naturally to be 
 expected, that when he desired to compress some parts of 
 his original work, in order to make room for expansion in 
 others, this highly important but unnecessary exposition 
 should be curtailed. Moreover, the first four paragraphs 
 of the First Edition were fully reproduced in the sequel, 
 and besides the publication of the Prolegomena during the 
 interval supplied an independent analytical" exposition. 
 We may add that these paragraphs, and especially an intro- 
 ductory statement, 2 also suppressed, seemed to attribute to 
 
 1/1 Cf. above, p. 5. 
 
 2 Cf. K. Fischer's Commentary, p. 76, and Hartenstein's Kritik, p. 
 112, note. 
 
124 METHOD OF THE DISCUSSION chap. 
 
 sense a power of combination which Kant carefully qualified 
 on further consideration. He would not speak in the 
 Second Edition of a synopsis of sense, and he was more 
 careful to show that the synthesis of apprehension was really 
 the work of the imagination. 
 
 1T As therefore the pure apperception, or synthetical unity 
 of apperception, was the point most obscure, and most 
 difficult of comprehension, he determined to develop his 
 brief synthetical exposition in the First Edition, 1 in order 
 that he might ' begin from pure apperception,' and explain it 
 more fully to his critics. It was necessary to do this in two 
 directions — first, for those who could not understand him, 
 or see how the synthetical power of the understanding could 
 form a principle of unity in nature ; secondly, for those who 
 exaggerated our knowledge of this very pure apperception, 
 and held that the Ego was given to us directly, not as a 
 phenomenon, but as a noumenal reality. 2 The sections 
 16-19, therefore, contain his fuller exposition of the syn- 
 thetical unity of apperception. The second part of § 24 
 and § 25 contain his expanded refutation of the second 
 error. ( This latter passage, then, may be postponed by the 
 reader as not forming a necessary part of the deduction. ) 
 
 In the next place, the opening of § 1 5 3 is to some extent 
 a repetition of his remarks on synthesis, in § 10, adding a 
 few important points, but almost ignoring what he said of 
 
 1 Vol. ii. pp. 207-210. 
 
 2 The danger of being misunderstood in this direction affords another 
 reason for his striking out the passage (p. 213) in which he spoke of 'the 
 fixed and permanent Ego ' as the correlate of all our representations. 
 He insists here more strongly on what he had already stated in the First 
 Edition, that we are directly conscious only of a faculty of combining 
 (synthesis), not of a subject exercising this faculty. 
 
 3 Cf. the note on this numbering above. Mr. Meiklejohn numbers 
 , these §§ 6 and 1 1 respectively. 
 
SYNTHESIS 125 
 
 syniB^Bs in the former section. The same observation 
 appl^Hmore strictly to §§ 16 and 17, which merely amplify 
 and "Mnforce the same points in varied language. Let us 
 obser\^, lastly, that the first part of § 24 anticipates the 
 schematism of the Categories, and discusses the function of 
 the productive imagination, which must again come before 
 us at greater length in that chapter. In brief, then, we 
 recommend the reader the following course in this discussion. 
 Let him read § 15, comparing it with § 10, and let him 
 next read § 19. Let him then read § 16 or § 17, either of 
 which suffices for Kant's argument. Proceeding directly 
 through §§ 18, 20, 24, let him postpone the second part of 
 § 24 and § 25 till he has read the concluding sections. By 
 this means he will greatly curtail his labour, and be better 
 able to apprehend Kant's argument. We now proceed 
 to accompany him through these sections in the order 
 prescribed. 
 
 § 15. If we look back to the exposition of § 10, we there 
 see synthesis l put forth as the original and primary condition 
 of knowledge, at first rudimentary and almost instinctive in 
 its action, afterwards explicating itself into certain definite 
 phases, or ways of combining, which Kant calls Categories. 
 As his object at first is mainly to discover or determine the 
 pure Categories, he passes over the earlier or undeveloped 
 
 1 It may be asked why we have adhered to Kant's strange word 
 synthesis, and not used the English word combination, which may mean 
 the same thing. We have done so to avoid an important ambiguity, 
 which Kant saw when he avoided the word Verbindung. Combination 
 oftener means the result of an act (aivdrnxa.) of combining, than the act 
 itself (uiv6e<n%). This kind of ambiguity is common to imagination, 
 conception, and many other such words. But nothing is more vital in 
 this discussion than to hold fast that we are speaking of combination as 
 an act ox function of the understanding. This definite sense is exactly 
 expressed by synthesis. Adhering strictly to this sense, the reader may 
 substitute combination, or any other word he pleases. 
 
126 ANALYSIS PRESUPPOSES SYNTHESIS chap. 
 
 stage of synthesis, which is in existence and at work before 
 we can form a concept, as its necessary condition. It is to 
 this stage that Kant now turns his attention. It had been 
 mistaken by some of his critics, and it had been asked 
 whether the Category of unity was not sufficient to account 
 for the unity in objects of intuition. But, granting that our 
 intuition is sensuous, or purely receptive, it is plain that 
 representations are given us through it merely as such, i.e. 
 not as objects. Their form may also be a priori in us, and 
 yet be nothing but the way in which we are affected. 
 Neither of these can give us combinations of various repre- 
 sentations [objects], for this is not an act of receptivity, but 
 of spontaneity, and therefore the work of what we call the 
 understanding, as opposed to sense. Whether we are 
 conscious of it or not, whether the things combined be 
 concepts, or sensuous intuitions, or pure intuitions, the 
 combination is an act of the understanding, which Kant calls 
 synthesis, indicating that this alone cannot be given by 
 objects, but is the self-activity of the subject, and that we 
 can represent nothing as combined in objects which our 
 understanding has not previously itself combined. Kant 
 thinks it obvious that this act of the understanding is 
 originally one and equivalent in all combinations, 1 and also 
 the necessary condition of analysis. 2 It will be remembered 
 that he spoke in the First Edition of the identity of function 
 
 1 Cf. above, p. 84. 
 
 2 He shows this in an ingenious note to § 16. My notion of red is 
 obtained by analysing several red things, and abstracting the colour. 
 But I cannot do this without presupposing red as an attribute already 
 combined with others in this or previous representations. An attribute 
 common to many different representations, implies that they each con- 
 tain something different combined with it. These, then, are synthetical 
 unities, which must be presupposed before I can obtain an unity by 
 analysis. 
 
viii IMPERFECT DESCRIPTION OF JUDGMENTS 127 
 
 of the understanding, and also of the unity of the action, in 
 this synthesis. 
 
 But the very notion of such combination implies that 
 variety is brought into unity, and this unity is a prior con- 
 dition, not a result of the combination, which has no 
 meaning without presupposing unity. We cannot, therefore, 
 proceed to a judgment, or to a Category (such as that of 
 unity), which is based on the power of judging, without 
 presupposing this combination, or unity of given representa- 
 tions. We must therefore seek it at the very earliest stage 
 of cognition. Let it be observed (§§ 10 and 19) that this 
 unity is presupposed by all judgments, which are very 
 inadequately described as the assertion of a relation between 
 two concepts. In what does the relation consist ? In 
 nothing but reducing them to an unity, by means of this 
 synthesis of which we are speaking. I may do this myself 
 personally, in which case I call it a judgment of perception, 
 or may conceive it as being done by all mankind, in which 
 case I call it a judgment of experience, or a law of objects — 
 in either case every judgment implies unity produced by this 
 synthetical action of the understanding. 
 
 §§ 16, 1 7. Beyond the necessary reference of all intuitions 
 to space and time, they must be subject to another condition, 
 viz. they must be present in our consciousness. There must 
 be a conscious I to intuite them, or they are nothing. But 
 this conscious I is not a receptive, but a spontaneous faculty 
 — in short the faculty which combines them, as has just been 
 explained. Furthermore, it is nothing but this consciousness 
 of a combining faculty, the same in all acts of consciousness. 
 This faculty Kant calls pure, or original apperception, since 
 it is a priori and spontaneous. He also calls it the tran- 
 scendental unity of self-consciousness, because this oneness 
 of the conscious self is the source of a priori cogni- 
 
128 WHAT IS OUR PERSONAL IDENTITY ? chap. 
 
 tions. 1 If representations could not be brought under the 
 condition of being recognised as belonging to my single self- 
 consciousness, they could not be considered as mine at all. 
 The consciousness which merely accompanies different re- 
 presentations is fragmentary, and may be called empirical 
 consciousness. But when I combine these various acts, 
 and am conscious of this synthesis, then only do I become 
 aware that my consciousness in them all was one and 
 identical. When I say then that all these representations 
 belong to me, I mean this, and this only, that I am com- 
 bining, or am able to combine them by this mental synthesis 
 into one whole — the unity of self-consciousness, or, as Kant 
 calls it, the synthetical unity of apperception. To say, 
 therefore, that the understanding means the faculty of com- 
 bining a priori our various representations under the unity 
 of apperception (or consciousness), is to state an identical 
 proposition, but it explicates our notion of pure conscious- 
 ness, and shows that our identity of self cannot be thought 
 without a synthesis of the variety given in intuition. 2 
 These conditions are imposed on us because the faculty 
 (intuition) which gives us multiplicity, and the faculty (self- 
 consciousness) which gives us unity, are different in kind. 
 If our understanding could intuite, then the multiplicity of 
 intuitions or objects would be given directly in its representa- 
 tion, and it would not require the act of synthesis which 
 our understanding, which only thinks, must perform. To 
 us this condition is so necessary that we cannot even conceive 
 an intellect directly intuiting, or even intuiting through other 
 conditions than space and time ; but if these other conditions 
 made its intuition receptive, a synthesis similar to ours would 
 
 1 See his explanation of transcendental above, p. 74. 
 
 2 This is the transcendental parallel to the law of Identity. Cf. p. 
 '59> infra, for the parallel to the law of Contradiction. 
 
viii TRANSCENDENTAL UNITY OF APPERCEPTION 129 
 
 still be required for the understanding attached to such an 
 intuition. 
 
 When we speak of the understanding, or the faculty of 
 cognitions, we regard the latter as referring to objects. But 
 an object implies a group or combination of intuitions. 
 This combination, as we have seen, can only be made by 
 the synthesis of our consciousness. It follows that the 
 unity of our consciousness is the necessary condition of our 
 forming any notion of objects. Here is an illustration : 
 mere space gives us no object, but only the materials for an 
 object. In order to know something in space, as for example 
 a line, I must draiv it, and so produce a synthetical unity 
 of parts. The unity of this act, as comprising several suc- 
 cessive acts, is the unity of my consciousness, which gives 
 me the notion of a line, and so only can I obtain such an 
 object. The synthetical unity of consciousness is therefore 
 an absolutely necessary or objective condition of all cogni- 
 tion, for not only do I require it in order to know an object, 
 but every intuition must come under it, before it can even 
 become to me an object. 
 
 § 1 8. This transcendental unity of apperception is (as we 
 have said) the unity by means of which we combine the 
 variety of intuition into the notion of an object. For this 
 reason we call it objective, to contrast it with such determi?ia- 
 tions of our internal sense as are merely suggestive and 
 empirical, and therefore not necessary. Every man makes 
 some of these combinations for himself a posteriori. But 
 the pure form of intuition (given a priori), considered as 
 mere presented variety, must stand under the original and 
 primitive / think, which alone contains an objectively valid 
 unity, viz. valid for every understanding. It is under this 
 objective unity (§ 19) that intuitions are brought in the act 
 of judgment. When we say a body is heavy, we do not 
 
130 RELATION OF CATEGORIES TO APPERCEPTION CH. 
 
 merely assert what seems to us by association subjectively 
 combined, but we make an assertion which, whether true 
 or false, is only possible by understanding what necessary 
 unity of apperception is, and consequently bringing two 
 representations under it. We assert these notions to be 
 necessarily combined into unity, not in our empirical in- 
 tuition, but by the synthesis of our perceptions in our pure 
 consciousness. 
 
 Thus the first step in the Deduction has been reached. 
 It has been shown that objects of intuition can only be 
 obtained by a combination of multiplicity. This combina- 
 tion is not given in a sensuous intuition, which is pure 
 receptivity. It is therefore added by the understanding, 
 which is a faculty whose function is to combine. But all 
 the several acts of combining are recognised by us as belong- 
 ing to one and the same consciousness. The importance 
 therefore of the unity of apperception, and its objective 
 character, are manifest. 
 
 § 20. But what have the Categories to do with this argu- 
 ment ? What relation have they to the pure apperception ? 
 It is this. The intuitions can only be brought under it by 
 the logical function of judging. Whatever variety therefore 
 is given in intuition can only be brought under the pure 
 apperception by being brought under one of the functions of 
 judging (as exhibited in the table, p. 86). But the Cate- 
 gories are these very forms of judging, so far as they merely 
 combine the variety of intuition (§ 13). This variety 
 therefore stands under the Categories as various phases, 
 or ways, of reducing it under the unity of apperception. 
 
 § 21. We have now proved that the Categories, which 
 arise in the understanding, quite apart from sensibility, can 
 introduce unity into intuitions quite generally, for this might 
 still be the case, even were our faculty of intuition different 
 
vm NEGATIVE SIDE OF THE DEDUCTION 131 
 
 from what it now is, provided it were receptive. We have 
 not yet considered how empirical objects are actually given- 
 us, or whether we can identify the unities given in them 
 with the unity imposed by the Category. ) When this is 
 done, our deduction will be complete. But though we 
 have hitherto abstracted from the way in which intuitions 
 are given us, we could not abstract from the fact that they 
 are given to us — that is, given from some other (here unde- 
 termined) source than our understanding, and independent 
 of it. If our understanding possessed a power of intuiting, 
 the Categories, which are mere acts of combining variety 
 given to it, would be idle, for the objects would then be 
 given directly to it in the act of intuiting. This peculiarity 
 of our understanding, as opposed to an intuitive understand- 
 ing, is, of course, a primitive fact, and inexplicable. 
 
 § 22. Bqt before we consider how empirical intuitions 
 are given to us, as contrasted with other possible sensuous 
 faculties of intuition, it is important to limit the other side 
 of the process, and show that the Category is of no use in 
 cognising things, except when applied to objects of experience. 
 For thinking and knowi?ig (cognising) an object are not the 
 same. To know it, we want both a Category, or concept, 
 and also an intuition, without which the former is mere 
 form, or possibility of knowledge. But we can have none 
 but sensuous intuitions either of pure space and time, or of 
 sensations in space and time ; and, moreover, the objects 
 given by the former (mathematical figures) are mere forms, 
 which do not prove the existence of things corresponding 
 to them. Things in space and time must be representations 
 accompanied by sensations, or empirical perceptions. Hence 
 the Categories, even when applied to pure intuition, give us 
 no knowledge of things, till we appeal further to empirical 
 intuition, or experience. Our assertion is therefore proved. 
 
1 32 SYNTHESIS SPECIOSA chap. 
 
 §23. It was easy to perceive the corresponding limita- 
 tion in the case of space and time, for we cannot carry 
 them beyond our senses. The pure Categories are not so 
 restricted, and may apply to the objects of any sensuous [or 
 receptive] intuition, whether it be in other respects like ours 
 or not. But this extension proves vain. For beyond our 
 sensuous intuitions they are mere empty forms of objects, 
 since there is no actual intuition at hand, to which they can 
 apply their synthesis to produce an unity of apperception, 
 and this is the only function they can exercise. We can 
 only then describe an object of an intuition different from 
 ours by negative predicates — by judging, for example, that it 
 is not in space and time, or subject to change. But these 
 negations contain no positive cognition whatever. And even 
 if they did, we should still not have the least notion what 
 Category to apply to such an object, for empirical intuition 
 must determine this point also, as will appear when we con- 
 sider the schematism. 
 
 § 24. In this paragraph Kant comes to explain the office 
 of the imagination, as intermediate between the pure under- 
 standing and the sensuous intuitions. He anticipates to 
 some extent the schematism, but this is nevertheless requisite 
 to the full comprehending of the Deduction. The pure 
 Categories, referring to the combination of the data of 
 intuition generally, are mere forms of thought, and not only 
 transcendental, but purely intellectual. But as the form of 
 sensuous intuition lies a priori within us, the understanding 
 can act upon this, and through it upon sensuous intuitions. 
 By this means the purely intellectual synthesis of the naked 
 Category passes into an intuitible or figurative synthesis 
 {synthesis speciosd), though still a priori and transcendental. 1 
 
 1 The synthesis inteilectualis would then be the same for all rational 
 beings, but the character of the synthesis sfcciosa depends on the peculiar 
 
viii THIRD STAGE OF THE DEDUCTION 133 
 
 Kant regards this latter synthesis as the work of the imagi- 
 nation, which therefore performs a transcendental synthesis, 
 to be distinguished from that of the mere understanding. 
 As reproductive of intuitions, it is indeed a faculty belonging 
 to sensibility ; but as exercising a spontaneity which actively 
 determines intuitions in harmony with the Categories, it is 
 allied to the understanding, and may be called the product- 
 ive imagination, which performs a transcendental synthesis 
 under the direction of the understanding. 
 
 Omitting for the present the Appendix to the Aesthetic 
 here inserted by Kant, we proceed at once to the conclusion 
 and summary of the whole deduction. 
 
 § 26. In the metaphysical Deduction (or exposition, as he 
 calls it in the Aesthetic), the a priori origin of the Cate- 
 gories was proved generally by their perfect coincidence 
 with the general functions of thinking (§ 10). In the tran- 
 scendental (§§ 20, 21) their possibility was shown as a priori 
 cognitions of the objects of intuition generally — that is to 
 say, of any sensuous or receptive intuition. We now pro- 
 ceed to complete the deduction by showing the possibility 
 of cognising « priori, according to the laws of their com- 
 bination, whatever objects can be presented to our senses. 
 Our combination of variety in space and time, an act of the 
 imagination, called by Kant (above, p. 118) the synthesis of 
 apprehension, must obviously conform and correspond to 
 the forms of space and time. But space and time are not 
 mere forms of sensuous intuition, but themselves intuitions 
 — that is to say, their variety is represented a priori as com- 
 bined into unity?- It appears then that unity in the synthesis 
 
 nature of the sensibility and its forms. The former is the skeleton, as it 
 were, of the latter. 
 
 1 Cf. Kritilt, p. 98, n. "Space represented as an object contains 
 more than the mere form of the intuition : namely, a combination of the 
 
134 EXAMPLES CHAP - 
 
 of variety, both within us (in time), and without (in space), 
 is given as the first condition of sensuous apprehension 
 along with the very act. This can be no other unity than 
 the combination of sensuous intuitions in general, which 
 takes place in pure consciousness, according to the Cate- 
 gories, as above explained (p. 127); it is here applied to 
 our sensuous intuitions. As, therefore, experience is nothing 
 but a knowledge of connected perceptions, and these are 
 shown to stand under the condition of the Categories, the 
 Categories are fully proved to be the conditions of the pos- 
 sibility of experience. Here are some examples showing 
 that the conformity of the objective synthesis to the unity 
 (1) of space and (2) of time is equivalent to its conformity 
 to the Categories. 
 
 When I perceive a house, the necessary unity of space, 
 and of my external intuition generally, must be presupposed. 
 It is in accordance with this that I, as it were, draw its 
 figure, and separate it from surrounding perceptions. 1 This 
 is the synthesis of apprehension. But abstracting from 
 space, we find that the understanding exercises the same 
 spontaneity more generally by the synthesis of homogeneous 
 
 manifold given according to the form of sensibility into a representation 
 that can be intuited ; so the form of the intuition gives us merely the 
 manifold, but the formal intuition gives unity of representation . . . 
 and as by means of this unity alone, space and time are given as intui- 
 tions, it follows that the unity of this intuition a priori belongs to space 
 and time, and not to the concept of the understanding.' 
 
 1 That is to say, to perceive a house implies that we know it as a 
 single object, separated from the surrounding perceptions given at the 
 same time ; also that it exists in space, in which the surrounding percep- 
 tions also exist. Two unities, the larger one of space, the lesser one of 
 the house, are both implied in our knowledge of it as an object. Kant 
 further states that we separate the house from its surroundings by a spon- 
 taneous act, which he calls drawing the figure, or mentally marking it 
 out from the rest. 
 
EPIGENESIS OF THE PURE REASON 135 
 
 parts in any intuition, which is the Category of quantity. 
 The former synthesis must correspond with this latter. 
 
 When I perceive the freezing of water, I apprehend two 
 states of water standing in a time relation. But time is an 
 internal intuition (as well as a form) with a necessary syn- 
 thetical unity of parts, and the necessary condition of 
 perceiving this relation. This is the synthesis of apprehen- 
 sion. But apart from time, the unity under which the 
 understanding combines such varieties in intuition gener- 
 ally, is the Category of cause, which, when applied to my 
 sensibility, determines all events in time according to its 
 relation. Therefore the apprehension of the event, and 
 therefore the event itself, stands under the relation of cause 
 and effect. 
 
 The conclusion of the paragraph repeats the argument 
 already (ii. pp. 206, 215) developed, that as the Categories 
 prescribe laws for phenomena, or objects of nature (materi- 
 a/iter spectata), they must consequently legislate for the 
 legitimacy or order of nature (formaliter spectata). There 
 is no difficulty whatever in the argument, and as we have 
 explained it already, 1 we shall not weary the reader with 
 repetitions. 
 
 § 27. We have come to the strange conclusion that for 
 us no cognition a priori is possible, except of objects of possible 
 experience. Yet though thus limited, it is nevertheless not 
 borrowed from experience, but as regards both pure intui- 
 tions and pure Categories, found in us a priori. As, there- 
 fore, experience and the Categories are in harmony, and 
 experience is not the ground of possibility of the Categories, 
 the reverse must be the case. This Kant calls the Epi- 
 genesis of the pure reason, which begets the frame and order 
 of nature by means of its Categories. 
 1 Above, p. 115. 
 
136 PRE-ESTABLISHED HARMONY chap. 
 
 Another alternative 1 has been proposed : That we are so 
 organised as to have subjective dispositions implanted in us, 
 corresponding to the independent laws of nature. This is 
 a sort of pre-established harmony. In the first place, Kant 
 argues in reply, if we once begin to postulate such hypo- 
 theses, there is no limit to their further use in explaining 
 other difficulties. But it is still more decisive, that in such 
 case the Categories must lack that necessity, which belongs 
 to their very nature. He thinks that the law of Causality, 
 for example, which asserts the necessity of certain conse- 
 quences, would be false. For we should only be entitled 
 to say : I am so constituted that I cannot think the effect 
 and cause except thus conjoined. This is just what the 
 sceptic wants, for then all our supposed objective judgments 
 would be mere illusion ; and when men were found, as there 
 surely would be, who denied the necessity, though they must 
 feel it, we could, at all events, never dispute with them 
 about a matter depending on the peculiar constitution of 
 their thinking subject. 
 
 11 The reader will at once perceive the close analogy 
 between this reply and that of Locke to the idealist sceptics 
 of his day [Locke's Essay, iv. 2, § 14]. It is too, like that 
 passage in Locke, one of the weakest passages in the great 
 work of a great author. Surely if we are all agreed that the 
 laws of nature are a mental relation superadded to the bare 
 successive feelings given to our nerves of sense, then the 
 only question which remains is this : Did the mind impose 
 them originally, or abstract them from repeated sensations ? 
 That there should be an unknown order of nature, in addi- 
 tion to and corresponding with the order which our under- 
 standing is, on either theory, competent to impose on its 
 
 1 Cf. vol. ii. p. 101, note, where we see that he is alluding to the 
 opinions of Crusius. 
 
APPERCEPTION AND INTERNAL SENSE 137 
 
 sensations — to require this is so perfectly otiose and gratui- 
 tous, as to be wholly inadmissible in any reasonable theory 
 of human knowledge. We might as well assume a real 
 space and time, after all the phenomena have been perfectly 
 and adequately explained by the Kantian theory. 
 
 If the reader has been able to follow this long and 
 intricate discussion, he has mastered perhaps the greatest 
 difficulty in the Kritik. 
 
 We may conclude this chapter with an account of the 
 supplement to the Aesthetic, which is inserted in the middle 
 of the discussion on the Categories, in §§ 24, 25. 'This,' 
 Kant says, ' is the place to explain the paradox which must 
 have struck every one in the exposition of the internal sense ' 
 (§ 6), where it is said, ' that our internal sense represents 
 us to ourselves as phenomena, not as we exist per se,' — in 
 other words, that we only intuite our internal affections, not 
 our internal being (self). As this puts us into a passive 
 relation as regards ourselves, it has been usual to identify 
 the faculty of apperception with the internal sense, whereas 
 we distinguish them carefully. 
 
 In order to know ourselves, as in the knowledge of any 
 other object, our understanding must employ its primitive 
 faculty of combining the variety given in internal sense, and 
 bringing it under the unity of apperception. "We have seen 
 that our understanding is not a faculty of intuiting, and 
 must regard such a faculty, when acting in the sensibility, 
 as a faculty differing from itself, and the variety given in it 
 as a variety not obtained by its own direct action. If, then, 
 we turn our attention to the synthesis of the understanding, 
 regarded purely by itself, it is nothing but the unity of the 
 action, of which we are conscious even without sensibility, 
 and which binds up the variety of sense, given internally, 
 according to the form of internal intuition. So it is that 
 
i 3 S CONDITIONS OF FORMING OBJECTS 
 
 our understanding, by a transcendental synthesis of the ima- 
 gination, as it is called, being one of this subject's faculties, 
 acts upon the passive subject, and thus affects the internal 
 sense. Apperception and its unity, as the source of all 
 combination, act upon all intuitions in general, under the 
 title of Categories, before they act upon objects in sensuous 
 intuition ; in other words, the unity of apperception is neces- 
 sary to obtain the frame, or Category, which is logically 
 prior to our knowledge of any object (whether internal or 
 external) in this frame. The internal sense is the mere 
 form of intuition, which does not give us a definite intuition, 
 or object, till its variety has been combined by that tran- 
 scendental action of the imagination above called the 
 figurative synthesis. 
 
 We can easily observe this in ourselves. We cannot 
 conceive a line, or circle, or other figure, without drawing 
 it in thought, or conceive even time, without drawing its 
 external image, a right line. This means that we direct 
 our attention merely to the action of combining multiplicity, 
 by which we determine our internal sense successively, and 
 so observe the succession in that sense. This motion, as 
 an act of the subject 1 (not as a determination of an object), 
 if we attend to the mere action by which we determine the 
 internal sense according to its form, is what produces in us 
 the very notion of succession. The understanding does not 
 find the combination in sense ; but produces it by acting 
 upon sense. The difficulty as to how the thinking self can 
 
 1 Kant notices that the motion of an object in space belongs not to 
 pure science, and therefore not to Geometry, as it requires experience to 
 know that anything is movable. But motion, as the act of drawing 
 figures, which is presupposed by Geometry, is a pure act of successively 
 combining multiplicity in external intuition generally by our productive 
 imagination ; it therefore takes its place even in transcendental philo- 
 sophy. Cf. above, p. 59. 
 
OUR MINDS NO EXCEPTION 139 
 
 be regarded as different from the self-intuiting self, and yet 
 identical with it, cannot be avoided or diminished by any 
 other theory, if we regard ourselves (as we must) as objects 
 of our own internal perception. That this letter is an 
 intuition is plain, when we consider that the only image we 
 can form of time, in which we represent ourselves, is a line 
 in space, and that all measures of time are imaged by 
 changes in external things ; in fact, that the determinations 
 of the internal sense in time are strictly analogous to those 
 of the external in space. But we only intuite external 
 objects, when we are affected through the external sense ; 
 we only intuite internal when we are affected through the 
 internal sense ; in other words, we know ourselves as pheno- 
 mena in time, not directly, as to our real nature. Every 
 act of attention gives us an example of this internal relation. 
 Here anybody can perceive how his understanding, as an 
 active faculty, determines his internal sense, as a passive 
 state ; in other words, we actively choose that our minds 
 (here controlled as passive) shall attend to something dif- 
 ferent from the natural succession of ideas. § 25. But the 
 phenomenal self given in internal intuition by the synthetical 
 action of our understanding, is not the only datum we have. 
 This very transcendental synthesis implies a consciousness, 
 not of what we are, but that we exist. This representation 
 we reach by thought, not by intuition. Now, every human 
 cognition, or knowledge, requires (a) a combining action of 
 the understanding, which unites (/3) the multiplicity given 
 in some kind of intuition. It follows that this consciousness 
 that we exist, as it wants the second element, is not a cog- 
 nition of itself. This self is indeed no phenomenon, far 
 less an illusion, but can only become an object by an appeal 
 to internal sense. All the thinking in the world, all the 
 Categories, will not supply this element. I exist therefore 
 
i 4 o THE EGO APART FROM INTUITION 
 
 as an intelligence, conscious merely of its faculty of com- 
 bining, but subject to a limiting condition in the things 
 combined, viz. that they must be obtained by the internal 
 sense, and therefore in time. This time modifies all the 
 data we receive through it, and thus makes them pheno- 
 mena, that cannot inform us of things per se. To obtain 
 these latter we should possess an intellectual intuition. 
 
 The / think, gives us the act of determining our own 
 existence, but no determination of our existence. As I have 
 no self-intuiting faculty, to intuite the subject, prior to its 
 act of determining, like as I have a pure sensuous intuition 
 in time, prior to objects in time, it is impossible for me to 
 determine my existence, as a self-acting being. I represent 
 myself, therefore, as spontaneous in thought, but with an 
 existence determined only, as other phenomena are, sensu- 
 ously. It is the consciousness of spontaneity, however, 
 which enables me to call myself an intelligence. 
 
 H Probably this long and difficult parenthesis has been 
 one of the main reasons why the Deduction was not better 
 understood. It really does not bear on the argument of 
 the Deduction, but on the Aesthetic, and was inserted here, 
 because Kant could not treat it till he had explained pure 
 apperception and the transcendental synthesis of the ima- 
 gination. It is a direct refutation of the theory propounded 
 by Dean Mansel, that we are presented with ourselves 
 directly, or intuitively, as substances, in contrast to the 
 indirect presentation of external things through their 
 attributes. His theory shows how little Mansel had appre- 
 hended this part of the Kritik, as he seems to have fol- 
 lowed Kant pretty closely whenever he could understand 
 him. It is bad enough to say that we have an intuition of 
 self, when, as a matter of fact, we cannot make a single 
 assertion about the intuition, or explain it, but merely 
 
MANSEL S THEORY 
 
 reiterate the assertion — unmeaning in itself — for the sake of 
 a philosophical theory. But surely the further collocation 
 of words, ' intuiting ourselves as substance,' might have 
 made Mansel pause. How is it conceivable that we should 
 intuite substance, as distinguished from its attributes ? 
 Surely if such a thing were conceivable, the substance 
 which we postulate for external things would not be such 
 an utterly negative, inconceivable representation ? In 
 a private communication, as regards this criticism, he 
 defended himself by saying, that if we were conscious of 
 self as a cause, which Kant has explained just now, we 
 must necessarily be conscious of ourselves as substance, as 
 substance and cause are in this case identical. We hold, on 
 the contrary, that we may be conscious of causation, or 
 action, without knowing anything more of the substance 
 which is the subject of the action. We hold the present 
 case to be a very striking instance of this fact. The ulti- 
 mate appeal is to each man's consciousness, and in this 
 appeal the great majority of readers will probably agree 
 with the great majority of modern philosophers, who, when- 
 ever they have avoided amplifications of language, and 
 stated the facts clearly, have plainly denied the immediate 
 presentation of self as a substance. 
 
CHAPTER IX 
 
 THE TRANSCENDENTAL ANALYTIC, BOOK II. THE ANALYTIC 
 OF PRINCIPLES 
 
 General Logic is built on a basis agreeing perfectly with 
 the higher faculties of knowledge, which are understanding, 
 [the faculty of] judgment, and reason. We have accord- 
 ingly the doctrine of concepts, judgments, and syllogisms 
 arranged on this plan. As this formal Logic merely dis- 
 cusses the form of thinking, it can even comprise in its 
 Analytic the canon of the reason, for this faculty, apart 
 from the peculiar nature of the cognitions used, must have 
 its proceedings prescribed and fixed. 
 
 Transcendental Logic, which is limited to a definite 
 content, viz. pure a priori cognitions, cannot follow in its 
 wake. For it appears that the transcendental use of 
 Reason is not objectively valid, and so not the Logic of 
 Truth, or Analytic, but occupies, as a Logic of illusion, a 
 separate place, under the title of transcendental Dialectic. 
 It is then Understanding and Judging only that have a 
 canon of their objectively valid use in Transcendental Logic. 
 The Analytic of Principles is simply a canon for the proper 
 use of the faculty of judgment, and teaches it to apply to 
 phenomena the Categories, which contain the a priori con- 
 ditions for rules. Taking then the proper principles of the 
 
chap, ix THE FACULTY OF JUDGMENT 143 
 
 understanding for his subject, Kant indicates his scope by 
 the title Doctrine of the Faculty of Judging. 
 
 Introduction. — Of the transcendental Faculty of Judgment 
 generally. 1 
 
 If the understanding be the faculty of rules, the judging 
 faculty is the power of subsuming under rules, or distinguish- 
 ing whether a given case comes under the rule. General 
 Logic cannot possibly give any rules for this faculty. For 
 as Logic abstracts altogether from the content of knowledge, 
 and adheres to the pure form only, were we to attempt to 
 show generally what should come under its general rules, 
 this could only be done by another general rule, and the 
 application of this would raise the same difficulty. The 
 faculty of judging is then a special talent, which can be 
 practised, but not imparted. It is, in fact, that mother- 
 wit, which no schooling can replace, for even though we 
 cram our minds with any quantity of rules derived from 
 other sources, the faculty of using them must belong to our- 
 selves naturally, and no learning can cure stupidity. But 
 it is very useful to exercise this faculty by examples, especi- 
 ally as they seldom conform exactly to the rule, and so 
 teach us to apply it in a wider sense ; from this point of 
 view they are specially requisite to men whose natural 
 talent for judging is weak. 
 
 We have seen above that general Logic can prescribe no 
 
 1 The reader will observe that throughout this chapter Kant uses the 
 word "judgment" in a sort of practical every-day sense, not merely as 
 the general faculty of comparing representations. In this latter sense 
 he stated above (p. 82), " that judging and thinking were coextensive." 
 He now uses it as we do, when we speak of a " man of judgment," viz. 
 a man who knows how to apply his principles, or bring (subsume) 
 particular cases under the right principles. This ambiguity was first 
 noticed by Dr. Toleken. 
 
i 4 4 THE SCHEMATISM chap. 
 
 rules for the faculty of judgment. It is so very different a 
 case with transcendental Logic, that it appears to be its 
 special business to direct and secure the use of the pure 
 understanding by fixed rules. For in obtaining extension 
 for our field of knowledge a priori, or as Doctrine, to use 
 Kant's word, philosophy appears ill equipped, and has done 
 nothing ; but as Kritik, to prevent errors of judgment in 
 the few pure concepts that we possess — in this negative 
 duty it must exercise all its skill and acuteness. 
 
 But transcendental philosophy has this peculiarity, that 
 beyond the rule (or better, the general condition of rules) 
 given in the Categories, it can also show a priori the case 
 to which the rule should be applied. It shares this advan- 
 tage with Mathematic alone of other sciences, because it 
 treats of concepts which are to refer a priori to objects ; 
 consequently its objective validity, as well as the general 
 conditions under which objects can be given, in conformity 
 with these Categories, can only be shown a priori. Were 
 this not done, they would be mere logical forms, and not 
 Categories. Our transcendental theory of judging contains 
 two parts — first, the Schematism, treating of the sensuous 
 condition, under which Categories must be used ; secondly, 
 the Principles of the pure understanding, or the synthetical 
 judgments, which flow from the Categories under these 
 conditions, and lie a priori at the basis of all the rest of 
 our knowledge. 
 
 The Transcendental Theory of Judging, or 
 Analytic of Principles. Chap. I. 
 
 Of the Schematism of the Pure Understanding. — When- 
 ever we subsume an object under a concept, the two repre- 
 sentations must be homogeneous, as a matter of course. 
 
IX THE TRANSCENDENTAL SCHEMA 145 
 
 Thus the concept of a plate is homogeneous with the 
 purely Geometrical notion of a circle, for the roundness 
 thought in the former, can be intuited in the latter. But 
 the pure Categories are completely heterogeneous from all 
 sensuous intuitions. How then can the latter be subsumed 
 under the former, and how is, consequently, the application 
 of the Categories to objects of sense possible ? For surely 
 none will assert that any Category, such as Causality, can 
 be intuited in phenomena and contained in them. Here 
 then the necessity of the Theory of Judgment, or applica- 
 bility of the pure Categories to experience, becomes apparent. 
 In other sciences this divergence between the general con- 
 cepts and their concrete representation does not exist. 
 There must obviously be something intermediate, homo- 
 geneous on the one hand with the Category, on the other 
 with the phenomenon, and this must make the application 
 possible. This mediating representation must be pure, 
 and yet not only intellectual but sensuous. We shall call it 
 the transcendental schema. 
 
 We saw that the concept of the understanding produces 
 pure synthetical unity of various parts generally. Now 
 time, as the formal condition of the variety given in internal 
 sense, and so of the combination of all our representations, 
 also affords us an a priori multiplicity in pure intuition ; 
 that is to say, the (pure) times, in which a series of various 
 representations are given and combined by the mind, may 
 themselves be regarded as an a priori multiplicity, com- 
 bined, or combinable a priori into a pure unity. Therefore 
 a transcendental determination of time must have this in 
 common with the Category (which brings this time-deter- 
 mination into unity), that it is universal, and depends on 
 an a priori rule. But, on the other hand, it is also in con- 
 formity with the phenomenon, inasmuch as time is contained 
 
 1 1. 
 
146 THE SCHEMA CHAr - 
 
 in every empirical representation of variety. Here then 
 we have the schema we require. The deduction has 
 already taught us that the Categories are only applicable to 
 objects of experience, as distinguished from things per se ; 
 that they consequently must require modifications of our 
 sensibility, and this implies that formal conditions of sense 
 (especially internal sense) are also necessary, as constituting 
 the universal condition under which alone we can apply 
 the Category to an object. This last condition is the 
 schema of the Category, 1 and the proceeding of the under- 
 standing as regards these schemata we call the schematism 
 of the pure understanding. 
 
 The schema in itself is indeed, like the image, the pro- 
 duct of our imagination, but also differs in not being an 
 individual picture, as it merely aims at representing the 
 general way in which the unity of intuition is produced. 
 
 Thus I can place five points together thus , and they 
 
 produce a picture of one number 5. But when I think of 
 what number in general means, I have before me the pecu- 
 liar way in which the imagination proceeds to form such 
 an arrangement of points. "When we represent to our- 
 selves the general procedure of the imagination, in procuring 
 an image for a concept, I call this the schema belonging to 
 this concept." 
 
 IT The illustration used by Kant shows that he was a 
 Conceptualist, as regards the object of the mind in part at 
 least of its thinking concerning general ideas, as they were 
 called by the schoolmen. It also shows that, as usual, he 
 took a deeper and fuller view of the mental state which was 
 once the subject of such bitter controversy. Locke's 
 abstract idea of a triangle, which had roused the ire and 
 
 1 I.e. the schema is the way in which we determine time in onler tu 
 apply the Category. 
 
ix LOCKE S ABSTRACT IDEA 147 
 
 the laughter of so many critics, is here shown to be not 
 only free from absurdity, but even the truest account of the 
 matter contained in any previous philosophy. But Locke's 
 fault had been to attend to the unimportant part of the 
 process. It is the act of the mind in putting together the 
 image of a triangle, not the completed image, which affords 
 us the proper object in general thinking. 1 For the actual 
 images are in every case different, and even inconsistent, 
 but the act of making them in general is one and the same 
 in all cases. It is what Kant calls a 'unity of action.' 
 ' In fact,' he says, ' not images, but schemata, lie at the 
 basis of our pure sensuous concepts. No image could ever 
 be adequate to the general concept of a triangle, as it can- 
 not embrace right-angled, scalenon, etc. The schema of 
 a triangle can exist in thought only, and means a rule of 
 the synthesis of the imagination, when applied to pure 
 objects in space.' The same is the case with empirical 
 concepts. My notion (schema) of a dog means a rule 
 followed by my imagination in drawing the general features 
 of a certain quadruped, without confining myself to any 
 particular figure. 
 
 This schematism of the understanding, as regards the 
 pure form of objects, is a hidden craft in the secrets of the 
 human mind, which we can hardly expect ever plainly to 
 discover and to expound. But so much seems certain : the 
 image is the product of the empirical working of the pro- 
 ductive imagination ; the schema of sensuous concepts, such 
 as figures in space, is as it were a sketch (monogram) of the 
 pure imagination a priori, through which alone images can 
 be brought into agreement with the concept. The schema 
 of a pure concept of the understanding (Category), on the 
 contrary, can never be reduced to an image, but is only a 
 
 1 On this view of the schema as an act, cf. Kritik, pp. 1 10, 435. 
 
THE SCHEMATA OF QUANTITY 
 
 pure synthesis, according to a rule of unity supplied and 
 expressed by the Category. This schema is, of course, a 
 transcendental product of imagination, affecting the deter- 
 mining of our internal sense generally, as to its form — Time. 
 It produces that unity expressed by the concept, which is a 
 phase of the transcendental unity of apperception. Let us 
 proceed to details. 
 
 The pure image of all quantities {quanta) in external 
 sense is space ; for all objects of the senses generally, it is 
 time. But the pure schema of quantity [guantifos], as a 
 Category, is number, or the successive addition of homo- 
 geneous units. The act of numbering is nothing but the 
 unity of combining the variety of a homogeneous intuition ; 
 in fact, I generate time itself in the successive apprehending 
 of my intuition. [Kant here means that the Category of 
 Unity is exemplified or imitated by the mind considering 
 the perception in a single act, or moment of time, to indi- 
 cate unity. The Category of Plurality is exemplified by 
 the mind requiring several successive moments to apprehend 
 a perception, of which the parts are separate but homo- 
 geneous, and therefore the several acts appear as separate 
 units. Totality implies the adoption of a large unit, under 
 which many smaller are combined.] 
 
 Reality in the Category is that which corresponds to 
 sensation generally, that of which the concept indicates 
 being in time, as opposed to Negation, or non-being in 
 time. Their opposition is therefore that of the same time 
 full (of sensation) and empty. ' As time is only the form 
 of intuition, or of objects as phenomena, that which in 
 these objects corresponds to sensation is the transcend- 
 ental matter of all objects, as things per se, x in fact their 
 
 1 Not ' actual things,' as Mr. Watson renders {Selections from Kant, 
 p. 88). 
 
THE SCHEMA OF REALITY 149 
 
 reality.' 1 We must judge of this reality bythe amount of sensa- 
 tion produced upon us. But every sensation has a degree, or 
 quantity, by which it can affect our faculty of representation 
 during the same length of time, more or less, varying from 
 the maximum of sensation down to its complete absence, 
 or negation. Thus a continuous transition from reality to 
 negation is possible, which enables us to regard every 
 reality as a quantum (of sensation). The schema of a 
 reality, in this sense, is its continuous and uniform genera- 
 tion in time, when we pass from total absence of sensation 
 in time to some particular degree of sensation. [As we 
 before had several successive homogeneous perceptions in 
 successive moments, giving us number, so we must here 
 suppose several successive perceptions homogeneous in 
 their character, but differing in the increasing intensity of 
 the sensation they produce. Kant considers that we must 
 conceive the maximum of sensation as made up of all the 
 lesser degrees which we could apprehend successively in 
 time. But as the sum of them is given to us in an equally 
 short time as each of the lesser degrees, we come to know 
 the difference between the same subdivision of time as 
 either full or empty. So it is that an object which affects 
 even three different senses together has more reality than 
 an object which affects only one. The moment in which 
 it affects us is a full time, and the fulness may be measured 
 by three times of equal length, each filled by one of the 
 sensations. The same fact is implied when we speak 
 of one object being ten shades darker than another 
 
 1 This statement, occurring in the First Edition, as well as the suc- 
 ceeding ones, is a strange way of preaching the absolute idealism which 
 Schopenhauer and Fischer ascribe to Kant ! He regards sensation 
 here as directly suggesting something apart from our cognition, though 
 we have no means of studying it save through our sensation. 
 
ijo THE SCHEMATA OF RELATION chap. 
 
 of the same kind. Quality is, after all, a quantity of 
 reality.] x 
 
 ' The schema of substance is the permanence of the real 
 in time, or the representation of it, as the substratum of an 
 empirical time-determination generally, which remains, while 
 all else changes. Time does not elapse, but rather the 
 existence of changeable things elapses in it.' Consequently 
 substance, or the permanent in existence, is what corre- 
 sponds in phenomena to time, itself unchangeable and per- 
 manent. It is therefore by substance alone that we can 
 determine sequence and co-existence of phenomena in time. 2 
 The schema of Causality consists in the succession of 
 various phenomena, so far as it is subject to a rule [invari- 
 able sequence]. The schema of community or reciprocal 
 causation of substances as regards their accidents, is the 
 necessary simitltaiieity of the determinations of both. The 
 schema of possibility is the agreement of our synthesis of 
 various representations with the conditions of time gener- 
 ally. For example, contradictory attributes can only exist 
 in a thing successively ; consequently possibility determines 
 the existence of a thing at some time. The schema of 
 
 1 Cf. Prokg, p. 86, n. 
 
 2 As he tells us afterwards that impenetrability is the empirical 
 criterion of substance {Kritik, p. 169), we may take this as a specimen 
 of a permanent reality (of sensation), which remains the same, and so 
 enables us to determine changes in other qualities. So the chameleon 
 remains a solid body while its colours change, and we accordingly talk 
 of its substance remaining the same, while its accidents vary. But were 
 there not a permanent phenomenon of some kind, corresponding to the 
 general lapse of time, we should not know that other sensations occupied 
 shorter time, and changed while the cause of them is conceived un- 
 changeable. We may add, that these illustrations of the various schemata 
 are developed and explained by the succeeding chapters on the Prin- 
 ciples which embody them, and that it is impossible to make them 
 clear to the reader till he has studied the theory of the Principles. 
 
SUMMARY 
 
 151 
 
 actuality is its existence at a definite time ; that of necessity 
 its existence at all times. 
 
 We can now see the relation of all the schemata. That 
 of quantity is the putting together (synthesis) of portions of 
 time, in our successive apprehension of an object or objects. 
 That of quality is the putting together of (feebler) sensations 
 in time, or filling time with them. The schema of relation 
 is the attitude of perceptions to one another in all time 
 (either as transient and permanent, as necessarily conse- 
 quent, or as necessarily simultaneous). The schema of 
 modality represents to us time itself, regarded generally as 
 the correlate for determining an object, when we consider 
 how or whether it belongs to time. The schemata affect 
 therefore respectively a priori, and according to rules — 1, 
 the succession of time ; 2, the content of time ; 3, the 
 order in time ; and 4, the sum total (Inbegriff) of time, as 
 regards all possible phenomena. 
 
 Kant concludes the discussion by reiterating what he 
 had said above (§§ 22, 23), as to the use of the Categories 
 being only empirical, and that they have no meaning except 
 as applied to objects of sense. The discussion of the 
 schematism has shown this more plainly, in that it proves 
 that the transcendental synthesis of the imagination is 
 wholly employed in time and its determinations. The 
 schema is only the sensuous concept of an object in 
 accordance with the Category. If we lay aside this 
 restriction, the Category preserves nothing but a logical 
 meaning, and can determine no object, not to say a 
 thing per se. 
 
 IT It will tend to put this schematism as well as the 
 deduction of the Categories in a clearer light, if we consider 
 Schopenhauer's criticism, which at first sight appears some- 
 what plausible. Kant's plan (he tells us) was to find for 
 
152 SCHOPENHAUER'S OBJECTIONS chap. 
 
 every empirical l function of the understanding its tran- 
 scendental parallel. Now when we use a very abstract 
 empirical concept symbolically (as Leibnitz would say), we 
 often glance back towards the empirical intuitions from 
 which we have obtained the concept, and we call up in 
 imagination a sort of imperfect image momentarily, merely 
 to assure ourselves that our thinking is possible in intuition 
 — a psychological fact which any one will discover for him- 
 self easily by reflection. This fugitive phantasm, inter- 
 mediate between abstract concepts and clear intuitions, 
 Kant called a schema, and thence concluded that between 
 the pure intuitive faculty of sensibility and the pure faculty 
 of thought there are similar schemata of the pure Categories. 
 But what is the use of this schematism in empirical think- 
 ing ? Merely to secure that the content of the concept be 
 correct. The matter has been abstracted from empirical 
 intuition : we refer to it occasionally, to make sure that our 
 thinking is about reality. But Schopenhauer objects that 
 the pure a priori concepts come from within, and are not 
 derived from intuition ; hence, such concepts cannot be 
 referred to any intuition to guarantee their reality. It was, 
 then, upon the misapplication of this psychological fact 
 above mentioned, that Kant based his elaborate schematism 
 of the pure understanding. 
 
 IT Although Schopenhauer's criticism is unsound, it has 
 been here stated, as the refutation of it will bring the real 
 doctrine of Kant into a clearer light. Schopenhauer has 
 well described the ' abstract idea ' of Locke as a fugitive 
 phantasm, which gives reality to our symbolical concepts. 
 What is the exact office of this schema ? To insure to us 
 that our (empirical) concepts are applicable in experience ; 
 to show us that they are not merely logically possible, but 
 1 He should have said logical, when he refers to Kant. 
 
REFUTED 153 
 
 objectively real.. Now, in empirical concepts this require- 
 ment is satisfied, if the content of the concept answers to 
 the schema, as the law of contradiction secures its possi- 
 bility, or logical correctness. But all our objects of experi- 
 ence stand not only under representative concepts (genus, 
 species, etc.), but also under assertative concepts (substance, 
 cause, etc.). These are the Categories, which were already 
 proved to be part of the (transcendental) content of repre- 
 sentations. Hence, such concepts must be shown to be 
 applicable to objects of experience just as generic concepts 
 are. These latter establish their claim by means of the 
 schema just mentioned — how can the Categories do so ? 
 
 IT Let us look back to the deduction of the Categories. 
 All phenomena were found to agree in one point at all 
 events — they must be my phenomena. It is this unity 
 which makes us speak of Nature as a unity, and yet as 
 consisting of many lesser units, called objects. For there 
 is no unity in our experience except what is imposed by 
 our minds. Accordingly this highest and most general syn- 
 thetical unity of consciousness acts upon phenomena by 
 imposing upon them various phases of its unity, various 
 lesser unities, all dependent upon the highest synthetical 
 unity. These lesser unities are the Categories. They are 
 imposed by the mind upon phenomena, which thus become 
 objects. But how ? The sensations which are the com- 
 ponent elements of the object, being received into the 
 mind successively, are reproduced, but not simply ; the 
 imagination moulds them, and so produces, not only the 
 received phenomena, but also the form of a concept along 
 with them ; so that, owing to this addition (which is the 
 transcendental content of the representation), that faculty 
 is properly called productive. But what is the form added 
 to the received elements by the imagination and under- 
 
1 54 THE PSYCHOLOGICAL PROCESS chap, ix 
 
 standing in this its action ? Surely no additional sensation, 
 no heterogeneous intellectual something, called a Category. 
 The imagination can only arrange or regulate the relations 
 to time of all our sensations. This is the point upon which 
 the imagination fixes ; for all our thoughts whatsoever must 
 be in time. The Categories must be thought under this 
 condition. The Categories then are imitated (so to speak) 
 or exemplified in time-determinations, which are imposed 
 by the productive imagination upon phenomena. Thus 
 the pure Category of substance is that which can only be 
 subject — and not predicate. An image of such a concept 
 is impossible ; but the nearest sensuous representation we 
 can get is something which is absolutely permanent in 
 time. This, then, is the schema under which the imagina- 
 tion brings certain phenomena, which are accordingly 
 declared to be substance, and it is only by means of such 
 schemata that we can assure ourselves that our thought is 
 applicable to experience. In Mr. Monck's briefer words : — 
 Sense or experience has its formal as well as its material 
 conditions. The office of the schema of an empirical con- 
 cept is to show that the material part is right ; that of the 
 schema of a pure concept to show that the formal part is 
 so also. Such is, in brief, the general notion of the schema- 
 tism, which follows necessarily from the productive imagin- 
 ation, and which forms one of the most remarkable claims 
 of Kant for originality and acuteness. 
 
CHAPTER X 
 
 SECOND CHAPTER OF THE ANALYTIC OF PRINCIPLES 
 
 § i. Of the System of all the Principles of the Pure Under- 
 standing. — So far we have only considered the general con- 
 ditions, which alone justify the transcendental judgment in 
 using the Categories for synthetical 'judgments. We now 
 proceed to give a systematic sketch of the judgments thus 
 actually produced. Of course, our clue will still be the 
 table of the Categories, since it is their relation to experi- 
 ence which constitutes all pure rational cognition. 
 
 A priori Principles (lit. fiiTidamental principles J ) are not 
 so called, merely because they contain the foundation of 
 others, but also because they themselves are not based on 
 higher and more universal cognitions. Yet this property 
 does not free them from requiring to be proved. Such 
 proof cannot, indeed, be objective, being rather the founda- 
 tion of all knowledge of its object. But a proof from the 
 subjective sources, which make it possible to produce cog- 
 nition of objects in general, is not only itself possible, but 
 even necessary, for otherwise such assertions must run the 
 risk of being considered mere assumptions. 
 
 Furthermore, we here confine ourselves to the Principles 
 
 1 To avoid cumbrousness, we uniformly translate Gnmdsatz by Prin- 
 ciple. 
 
156 PRINCIPLE OF ANALYTICAL JUDGMENTS chap. 
 
 which refer to the Categories. We therefore exclude the 
 Principles of the transcendental Aesthetic, that space and 
 time are the conditions of the possibility of phenomena, 
 and of phenomena only. For the same reason, mathe- 
 matical Principles are not part of this system. Yet still 
 their necessity, since they are synthetical and a priori, must 
 find place in it, not to prove their accuracy or necessity, 
 but merely to explain and justify (deduce) the possibility of 
 such pure cognitions. We shall first discuss the Principle 
 of analytical judgments, in order to free the synthetical, 
 with which they are contrasted, from misapprehension, and 
 illustrate their real nature. 
 
 Of the highest Principle of all Analytical Judgments. — 
 The highest, though but negative, condition of all our 
 judgments, quite irrespective of their objects, is that they 
 shall not contradict themselves. It is negative, because 
 judgments may conform to it, and yet be either false, as 
 combining concepts otherwise than the objects direct, or 
 baseless, as having no foundation either a priori ox posteriori. 
 This proposition, then : no thing can have a predicate which 
 contradicts it, is the Law of Contradiction, is an universal 
 but negative criterion of all truth, and is merely logical, 
 referring to cognitions as such, apart from their content. 
 We may, however, make a positive use of it, not merely to 
 avoid error, but to know truth. In analytical judgments 
 whether affirmative or negative, the truth can always be 
 adequately known according to this principle. For that 
 which is already thought in the concept of an object, must 
 be rightly affirmed of that object, and its contradictory 
 denied, else the object would contradict our concept of it. 
 The law, is, therefore, the universal and adequate principle 
 of all analytical knoivledge, but here its use as a sufficient 
 criterion ends. It is, of course, a sine qua non of all judg- 
 
x THE LAW OF CONTRADICTION 157 
 
 ments, but cannot determine their truth generally. And as 
 we are concerned only with the synthetical part of our 
 knowledge, we must, of course, beware of violating it, but 
 cannot expect from it any light as to the truth of this kind 
 of knowledge. 
 
 But there is a formula for this celebrated, though purely 
 formal law, which does contain a synthesis, unnecessarily 
 imported into it through inadvertence. Here it is : It is 
 impossible for the same thing to be and not to be at the 
 same time. Here apodictic certainty, which is implied of 
 course, is superfluously added by the word impossible, and, 
 what is more important, the proposition is affected by the 
 condition of time. It affirms a thing A, which is something 
 = B, cannot at the same time be not-B ; but it can be both 
 B and not-B successively. For example : a man who is 
 young, cannot at the same time be old, but becomes so in 
 the lapse of years. It is quite wrong to make the purely 
 logical law of contradiction depend on time-relations, and 
 such a course obscures its real import. The mistake arises 
 from first separating the predicate of a thing from our con- 
 cept of it, and then connecting its contradictory with that 
 predicate. This produces no contradiction with the sub- 
 ject, but only with the former predicate, which we had 
 connected synthetically with the subject, and even then 
 only when the two predicates are posited simultaneously. 
 If I say a man who is unlearned, is not learned, I must 
 add, at the same time, or it may be false. But if I say, No 
 unlearned man is learned, the proposition is analytical, and 
 is evident without the addition of at the same time. For 
 this reason, then, Kant alters the formula, in order to 
 express clearly the analytical nature of the law. 
 
 § 2. Of the highest Principle of all Synthetical Judgments. 
 — To determine the possibility of synthetical judgments is, 
 
158 PRINCIPLE OF SYNTHETICAL JUDGMENTS chap. 
 
 as we have seen, the duty not of general but of transcend- 
 ental Logic, and is indeed its most important function, for 
 thus alone can the compass and limits of the pure under- 
 standing be determined. As, in synthetical judgments, I 
 must pass out of the concept {subject) to consider something 
 quite different in relation with it — this is never a relation 
 of identity or contradiction, and so in the judgment, per se, 
 we cannot see either truth or error. 
 
 Granted, then, that we must pass out of a given con- 
 cept, to compare something else synthetically with it, some 
 third thing, or medium, is necessary, to contain the syn- 
 thesis of two concepts. What is this medium ? There is 
 but one envelope (Inbegriff) which embraces all our repre- 
 sentations, viz. — the internal sense, and its form, Time. 
 The synthesis of these our representations depends on the 
 Imagination, their synthetical unity on the unity of apper- 
 ception. Here, then, we must seek the possibility of syn- 
 thetical judgments, — nay, more, of synthetical a priori 
 judgments, which will be shown necessary from these 
 sources, if a cognition of representations can be accom- 
 plished, resting exclusively on the synthesis of repre- 
 sentations. 
 
 If a cognition is to have objective reality — that is, to 
 refer to an object, and have in it significance, the object 
 must of course be" somehow given. To have an object 
 given immediately, by representing it in intuition, means 
 nothing but to refer its representation to experience, either 
 actual or possible. Even space and time, pure and cer- 
 tainly a priori as they are, would have no objective validity, 
 and no meaning, were not their necessary use exhibited in 
 objects of experience ; nay, our very representation of 
 them is a mere schema, ever referring to the reproductive 
 imagination for matter to fill it, without which they would 
 
x THE POSSIBILITY OF EXPERIENCE 159 
 
 bear no signification. So it is with every one of our 
 concepts. 
 
 The possibility of experience, then, or of being experienced, 
 is what gives all our a priori cognitions objective validity. 
 Experience again is based on a synthetical unity — a syn- 
 thesis of phenomenal objects according to the Categories, 
 or principles of its form, which lie a priori at the basis of 
 experience. These are indeed universal rules of the unity 
 of phenomena, but their objective reality, nay, even their 
 possibility, can only be shown in experience. This is the 
 medium in which synthetical a priori propositions can 
 exhibit their objective reality. 
 
 We do, indeed, in the case of space, and of the figures 
 which the productive imagination draws in it, discover a 
 great deal a priori by way of synthetical judgments, and 
 actually without requiring any experience ; but such know- 
 ledge would be occupation with what is a mere chimera, 
 were not space considered as the condition of phenomena, 
 which are the materials of external experience. Hence, 
 even the pure synthetical judgments of Mathematic refer to 
 possible experience, or rather to its very possibility, and on 
 this alone is based the objective validity of their synthesis. 
 
 ' As, therefore, experience as an empirical synthesis is, 
 in its possibility, the only sort of cognition which gives to 
 all other syntheses reality, so experience as a cognition 
 a priori can only have truth, or agreement with its object, 
 by containing nothing more than what is necessary for the 
 synthetical unity of experience genera]]) - . Here, then, is 
 the highest Principle of all synthetical judgments : even- 
 object comes under the necessary conditions of the syn- 
 thetical unity of the manifold of intuitions in a possible 
 experience ' Thus the possibility of (the general faculty of) 
 experience, and the possibility of (there being) objects of ex- 
 
160 UNDERSTANDING AS SOURCE OF PRINCIPLES chap. 
 
 perience, lie under the same conditions, and thus our syn- 
 thetical a priori judgments about the former obtain (through 
 the latter) objective reality. 1 
 
 § 3. Systematic Exposition of all the synthetical Principles 
 of the pure Understanding. — We must ascribe it purely to 
 the understanding that there exists such a thing as prin- 
 ciples. For the pure understanding is not only a faculty of 
 rules, but the very source of Principles, according to which 
 every possible object stands necessarily under rules, which 
 add to phenomena the cognition of an object corresponding 
 to them. Even laws of nature, regarded as principles of 
 our ordinary experiences, carry with them the expression of 
 necessity, or at least the suggestion of being determined by 
 grounds valid for all experiences. But such laws again 
 stand under the higher Principles of the understanding, 
 which they merely apply to particular cases. There is no 
 danger of confusing these two classes of Principles, for the 
 absence of that necessity according to concepts, which the 
 latter have, is easily perceived in the former, however uni- 
 versally valid. But there are pure a priori Principles, 
 which are not properly to be attributed to the pure under- 
 standing, because they are drawn (not from pure concepts, 
 but) from pure intuitions, by means of the understanding ; 
 such are those of Mathematic. Still, their objective reality 
 in experience, and the deduction of their possibility, must 
 rest on the pure understanding. 
 
 Hence, says Kant, I shall not enumerate among my 
 Principles those of Mathematic, but only such as Mathe- 
 matic requires for an a priori basis of its possibility and 
 objective validity. These are the principles on which 
 
 1 The reader should compare the corresponding discussion in the 
 Prolegomena (pp. 63 sqq.), entitled, Hoiv is a pure Science of A T atnre 
 (Physic) possible? 
 
x TABLE OF THE PRINCIPLES 161 
 
 Mathematical judgments are based, and proceed from con- 
 cepts to intuitions, not vice versa (as the ordinary Mathe- 
 matical axioms do). 
 
 In applying the Categories to possible experience we 
 may use them either mathematically or dynamically, for as 
 they may refer merely to the intuition, so they may also 
 refer to the existence of a phenomenon. 1 The a priori 
 conditions of intuition as regards possible experience are 
 absolutely necessary ; those of the existence of objects of 
 a possible empirical intuition are, as such, only contingent. 
 Hence the Principles of Mathematical use must be uncon- 
 ditionally necessary, and be apodictical in form ; those of 
 dynamical use, though also (of course) necessary a priori, 
 are so mediately, under the condition of empirical thinking 
 in some experience. The latter have not therefore such 
 immediate evidence, though equally certain, as generally 
 applied to experience. 
 
 The table of the Categories is of course our natural clue 
 to the Principles, inasmuch as these are merely the rules of 
 objectively applying them. 
 
 These Principles are therefore — 
 
 r. 
 Axioms of Intuition. 
 
 2 3- 
 
 Anticipations Analogies 
 
 of Perception. of Experience. 
 
 4- 
 
 Postulates of 
 
 Empirical thinking in general. 
 
 These terms are carefully chosen, to indicate distinctions of 
 evidence and use. It will soon appear that the determining 
 
 1 Cf. p. 97, supra. 
 I M 
 
1 62 ANALYSIS OF COMBINATION chap, x 
 
 of phenomena by the Categories of Quantity and Quality 
 (regarding merely the Form of the latter) differs from that 
 of the others, one being intuitively, the other discursively 
 certain. Hence they are distinguished as Mathematical and 
 Dynamical Principles.^ At the same time, they are not 
 the Principles of Mathematic, nor of general Dynamic 
 (Physic), but the Principles of the understanding in respect 
 of our internal sense, upon which depends the possibility 
 of both these more special classes of axioms. Kant's Prin- 
 ciples are therefore so called on account of their application, 
 not their content. 
 
 1 Kant adds a note in the Second Edition, giving a sketch of the various 
 classes of combination. — All combination {conjujictio) is either composi- 
 tion or nexus. The former is the synthesis of multiplicity not necessarily 
 inseparable, as, for example, the two triangles formed by the diagonal 
 of a square. This is the nature of all synthesis of homogeneotts parts 
 that can be mathematically estimated. It is aggregation, if extensive ; 
 coalition, if intensive. Nexus again is the synthesis of inseparable parts, 
 such as accident and substance, effect and cause, which are heterogeneous, 
 though combined a priori. This combination, as being arbitrary \_will- 
 kuhrlich, he means that no reason can be assigned why these particular 
 heterogeneous elements are combined], I call dynamical, because it 
 concerns the combination of the existence of multiplicity, and this again 
 is either physical, of phenomena among one another, or metaphysical, 
 of phenomena in the cognitive faculty a priori. These four kinds of 
 combination evidently answer to the four classes of Principles. 
 
CHAPTER XI 
 
 THE MATHEMATICAL PRINCIPLES 
 
 (i.) Axioms of Intuition. — Their Principle is: All intui- 
 tions are extensive Quantities} 
 
 Proof. 
 All phenomena contain as to form an intuition in space 
 and time, which is their a priori basis. They cannot there- 
 fore be empirically apprehended except by combining mul- 
 tiplicity, and so generating the representation of a definite 
 space or time, the parts of which are homogeneous, and 
 conceived as a synthetical unity. This is the notion of a 
 
 1 See what he says about the relation of this Principle to axioms 
 proper in his more special discussion in the Methodology, Kritik, p. 
 446. Kant, with that habit of repetition and variation of statement 
 common to the books of almost all great discoverers, has in his Second 
 Edition inserted after the Definition of each of the first three classes of 
 Principles an additional paragraph, entitled Proof or Demonstration. 
 With one exception (that of Community) these inserted paragraphs 
 merely repeat in varying language what follows them, and what had 
 stood in the original edition. The reader who compares our Comment- 
 ary with Kant's text will therefore find that we have curtailed the 
 arguments considerably, but only by leaving out repetitions. We trust 
 there is no distinct point, however small, omitted. The parallel dis- 
 cussion in the Prolegomena, §§ 24-32, will be found in vol. ii. pp. 81 
 sqq. 
 
1 64 EXTENSIVE QUANTITY 
 
 quantity (quanti). As therefore the perception of an object 
 as phenomenon is only possible through this synthetical 
 unity of multiplicity, all phenomena are extensive quantities, 
 because they must be represented by the same synthesis 
 which determines space and time generally as quantities. 
 
 Kant defines an extensive quantity as that in which the 
 notion of the part precedes and renders possible the notion 
 of the whole. So I cannot conceive a line without drawing 
 it in thought, and thus generating successively all the parts 
 which make up the whole intuition. It is so also with 
 every time, even the shortest. As every phenomenon must 
 be intuited through space and time, it must also be an 
 aggregate, or combination of given parts, and this is only 
 the case with those quantities which we apprehend as 
 extensive. 
 
 ' On the successive synthesis of the productive imagina- 
 tion in generating figures is based Geometry and its axioms, 
 which express the a priori conditions of intuition under 
 which alone the schema of a pure concept of an external 
 intuition can exist ; e.g. between two points only one 
 right line is possible ; two right lines cannot enclose 
 a space. These axioms properly concern only quanta as 
 such.' 
 
 But as to determining the quantitas of a thing, and 
 answering the question, How large is it ? — though there are 
 various synthetical and indemonstrable propositions about 
 it, yet they cannot be called axioms. For that equals 
 added to equals make equals is an analytical proposition, 
 whereas axioms must be synthetical. On the other hand, 
 the self-evident propositions about particular numbers are 
 synthetical, but not universal, like geometrical axioms. 
 They should therefore be called numerical formulae. 7 + 
 5 = 12, as has been already seen (above, p. 37), is synthet- 
 
xi REALITAS PHENOMENON 165 
 
 ical. 1 But it is also singular. For this synthesis of unities 
 can only be made in one way, though the use of the num- 
 bers is afterwards universal. The construction of a triangle, 
 as a pure function of the productive imagination, may be 
 produced in a thousand ways, but 7 and 1 2 only in one way. 
 Were such propositions, then, axioms, they would be infinite 
 in number. 
 
 This transcendental Principle of the Mathematic of pheno- 
 mena extends our a priori cognition widely. For now we 
 see why pure Mathematic in all its precision is applicable to 
 objects of experience. Empirical intuition is only possible 
 through pure intuition. What Geometry says of the latter 
 must therefore be true of the former. All evasions of the 
 laws of construction of space (such as the endless divisi- 
 bility of lines and angles) must vanish. For these theories, 
 which deny the objective validity of space and of the 
 Mathematic based on it, are only the devices of a mis- 
 guided reason, which endeavours to free the objects of the 
 senses from the formal conditions of our sensibility, in 
 which case nothing whatever could be asserted of them a 
 priori, and Geometry itself would become impossible. 
 
 (2.) Anticipations of Perception. — Their principle 
 is : In all phenomena the Real, which is an object of sensa- 
 tion, has intensive quantity, or degree. 2 
 
 1 Kant (and Mansel) seem to have overlooked such numerical axioms 
 as a (b + c) = ab + ac, and ab = ba, which are both synthetical and 
 general. 
 
 2 This Principle was worded in a slightly different way in the First 
 Edition. ' The Principle, which anticipates all perceptions as such, is 
 thus expressed : In all phenomena sensation and the real, which corre- 
 sponds to it in the object (realitas phenomenon) has a degree, or intensive 
 quantity.' It will be seen that the form of the First Edition is more 
 realistic here than that of the Second. At the same time, we should be 
 most cautious not to suppose that the real, which corresponds to sensa- 
 
1 66 THE PROLEPSIS OF EPICURUS chap. 
 
 Proof. 
 
 Perception is empirical consciousness, in which there is 
 sensation. Phenomenal objects of perception are more 
 than mere formal intuitions (of space and time). They con- 
 tain in addition the materials for some object, represented as 
 existing in space and time. This is the real (element) of 
 sensation, and accordingly a mere subjective representation, 
 which makes us conscious of being affected, and which we 
 refer to an object in general. Now from this empirical 
 consciousness down to pure consciousness a gradual change 
 is possible till we reach a mere formal intuition of space and 
 time. So we may also regard sensation as a gradual pro- 
 duction, beginning with o, in pure intuition, and rising to a 
 certain quantity. Thus, while it is not an objective re- 
 presentation containing space and time, it still has an 
 intensive quantity, or a degree of influence on our sensibility, 
 which must accordingly be attributed to all objects of per- 
 ception. 
 
 All cognition, which determines empirical cognition a 
 priori, may be called irp6Xr]t(/i<s, and this was doubtless 
 Epicurus's meaning for the term. 1 But sensation, or the 
 matter of perception, is the very thing which distinguishes 
 the empirical as such from a priori knowledge ; this there- 
 fore should be the element never anticipated. On the other 
 hand, we may call any pure determinations in space or time, 
 whether of figure or magnitude, anticipations of phenomena, 
 
 Hon, is a thing per se. It will appear from Kant's discussion on the 
 first Analogy that we are obliged to conceive a phenomenal substratum, 
 distinct from each particular representation, and this we call the sub- 
 stance, or the substratum of what is real in nature, as it is permanent 
 and neither increases nor decays. The fuller explanation must be post- 
 poned to its proper place. It was probably this obscurity of anticipation 
 which caused Kant to alter the form in the Second Edition. 
 1 Cf. Diogenes Laertius, x. i. 21. 
 
XI INTENSIVE QUANTITY 167 
 
 because they represent a priori that which may always be 
 given a posteriori in experience. But supposing we can 
 find an universal feature in sensation as such, apart from 
 particular sensations, this might with exceptional propriety 
 be called anticipation, as outrunning experience in the very 
 province which we owe to it alone. This is here really the case. 
 
 Apprehension, if regarded not as a succession of sensa- 
 tions, but as one, fills an instant only ; if it be absent, then 
 the empty instant = o. But on the other hand, the in- 
 stant of time when filled with sensation is regarded not as 
 an extensive quantity that can be dissected, but as an 
 unit = i. 1 
 
 Intensive Quantity may then be defined : that which is 
 apprehended as unity, and in which multiplicity can only be 
 represented by approximating to o. Kant observes paren- 
 thetically that when this reality of phenomena is regarded 
 as a cause, e.g. of sensation or of change, the degree of 
 reality as cause is called moment, as in the ' Moment of 
 Weight,' because its apprehension is momentary — a very 
 curious criticism. 
 
 Every sensation, then, and every reality in phenomena 
 have a degree, which can be diminished, and between 
 reality and negation there is always a series of possible 
 realities and possible lesser perceptions. Every colour, 
 every temperature, has a degree, which, however small, is 
 not the least possible. 
 
 This attribute of quantities, that no part of them is the 
 least possible, or simple, is called their continuity. So there 
 is no part of space or time which is not itself space or time. 
 Points and instants are only boundaries or limitations, 
 which themselves presuppose the intuitions which they 
 limit. Of such boundaries, then, as elements, time and 
 
 1 Vol. ii. p. 86, note. 
 
1 68 QUANTUM AND AGGREGATE 
 
 space could never be composed. As generated by the 
 productive imagination in the equable lapse of time, they 
 may also be regarded as equably flowing quanta. 
 
 If this synthesis be interrupted, we have an aggregate of 
 many phenomena, produced by the repetition of an ever- 
 finishing synthesis. If I call twenty shillings a sum of 
 money, I may rightly consider this as a continuous quantity, 
 no part of which is the least possible, but might be a coin, 
 containing under it further subdivisions. But if I speak of 
 twenty shillings as so many coins, it were improper to call 
 them a quantum of shillings ; they are an aggregate, or 
 number. But in this number each unit is a quantum, and 
 as such a continuum. 
 
 All quantities, whether extensive or intensive quanta, 
 being continuous, we could here easily demonstrate with 
 mathematical precision the proposition, that all change is 
 continuous, did not the causality of change lie quite beyond 
 the reach of transcendental philosophy, and presuppose 
 empirical principles. The understanding gives us no light 
 a friori how a cause should be possible, which alters one 
 condition of a thing into another, and besides in this par- 
 ticular case only a certain number of the features of pheno- 
 mena are affected, which experience alone can show us, 
 while their cause is to be found in the unchangeable. As 
 we are here using nothing but pure a priori fundamental 
 concepts, we must postpone such inquiries. 
 
 But we are in no want of evidence how important our 
 Principle is, in anticipating perceptions, and in precluding 
 false impressions drawn from their deficiency. It is ob- 
 vious if all reality in perception must have a definite 
 degree, separated from negation by an infinite gradation of 
 lesser degrees, that no perception or experience can possibly 
 prove a total absence of reality in phenomena. In other 
 
xi ASSUMPTIONS IN HIYSICS 169 
 
 words, we can never from experience demonstrate empty 
 space or empty time. For such a thing can never be per- 
 ceived, nor can it be inferred from, or brought to explain, 
 the differences in degree of reality in any phenomenon. 
 For though the whole intuition should be real in every part, 
 yet there are infinite degrees between this and negation of 
 reality in such intuition, regarded as an intensive quantity, 
 though the extension may remain the same. 
 
 Here is an example. Almost all physicists, perceiving a 
 great difference of weight in the same volume or extension 
 of different bodies, have inferred that this volume contained 
 vacuities, but in various measure. Who could have thought 
 that these chiefly mathematical and mechanical inquirers 
 would have based their conclusions on a metaphysical 
 hypothesis — a thing which they so studiously avoid ? They 
 assumed that the real in nature was homogeneous in every 
 case, and only variable in the number of its parts, or ex- 
 tensively. To this purely metaphysical assumption Kant 
 opposes a transcendental proof, which does not indeed pre- 
 tend to explain the differences in the filling of space, but 
 destroys the notion that it can be explained by assuming 
 empty spaces, and leaves the mind free to adopt some other 
 explanation. We now see that though equal spaces be 
 perfectly filled with matter, this matter has in quality a 
 degree (say of resistance or of weight) which may be in- 
 finitely various. Thus the heat of a room may diminish 
 indefinitely without leaving a single particle of its space 
 empty, [by filling them still completely, but with a lesser 
 degree of its quality. ' I am not asserting this,' says Kant, 
 ' as a fact, but as a probability which overthrows a very pre- 
 valent false assumption.' 
 
 Nevertheless, this Anticipation of Perception does seem 
 strange to an inquirer trained to transcendental reflection, 
 
170 THE QUANTITY OF QUALITY chap, xi 
 
 and so rendered cautious. How is it that the understanding 
 can presume an a priori principle of distinction in sensa- 
 tion, abstracting from its empirical quality ? For here we 
 are declaring synthetically a priori concerning that which is 
 purely and specifically empirical. This is the answer. 
 The particular quality is always empirical, and in no sense 
 a priori. But the Real, which corresponds to sensation 
 generally, as opposed to negation or voidness, is nothing 
 but a notion of being, and is a mere synthesis (of quantity 
 and of quality) in empirical sensations generally. For 
 without changing the quantity, and abstracting from it com- 
 pletely, we can conceive a full sensation as affecting us 
 equally with an aggregate of lesser sensations conveyed to 
 us separately, each in its own moment ; so we can antici- 
 pate a priori that quality must have intensive quantity, just 
 as we determined a priori that all quantity must have a 
 quality, Viz. continuity. 
 
CHAPTER XII 
 
 THE DYNAMICAL PRINCIPLES 
 
 (3.) Analogies of Experience. — Their Principle is : 
 Experience is only possible by representing a necessary com- 
 bination of perceptions? 
 
 Proof. 
 
 Experience is an empirical cognition, or one which 
 determines an object by perceptions ; and this of course by 
 a synthesis of perceptions. This synthesis is not in the 
 perceptions, but consists in their variety being combined in 
 one consciousness ; this is the essential feature in cognising 
 objects of sense. But as perceptions occur in accidental or 
 contingent order in time, we cannot conceive any necessity 
 in their combination derived from this source. Since, then, 
 
 1 The First Edition has this principle in a different form, and brings 
 into prominence the element of time. 'AH phenomena as to their 
 existence come a priori under rules of their mutual determination in a 
 [portion of] time.' It is a mistake of Kuno Fischer to assert (Comm. p. 
 107) that Kant in the Second Edition omitted the all important element 
 of time in his analogies. He does the reverse, for instead of merely 
 stating it, he adds a new paragraph under the title proof, in which he 
 argues it out. The reader should observe that the reasoning in all these 
 proofs is of a similar character. We see that the necessity in experience 
 does not come (a) from the matter nor {b) from the form of Sense, and 
 therefore we conclude that it is due to the Understanding. 
 
172 THE DYNAMICAL PRINCIPLES chap. 
 
 experience is a cognition of objects not as they are acci- 
 dentally represented in time, but as they are in time object- 
 ively, and this pure time itself cannot be perceived, the a 
 priori combination which determines objects can only be 
 that of the categories and schemata which connect them a 
 priori. These carry with them necessity, and so a necessary 
 combination of perceptions becomes essential to experience. 
 
 The three modes of time are — Permanence, succession, 
 and simultaneity , from hence come our three rules. They 
 are in fact phases of the one great form of all our empirical 
 consciousness, viz. the necessary unity of apperception in 
 all time, whereby all perceptions must be mine, and so 
 bound together in one great chain. This apperception 
 affects and orders the internal sense, of which it is the 
 form. Thus ' all empirical time-determinations must come 
 under rules of universal and a priori time-determination,' 
 and these latter are the rules now under consideration. 
 
 These Principles have the peculiarity that they concern, 
 not phenomena and the synthesis which we make in in- 
 tuiting them empirically, but only their existence and their 
 mutual relation as regards this existence. This was stated 
 above in our first sketch of the Principles, as we saw that 
 the way in which any phenomenon must be apprehended 
 can be so determined a priori as to give us a rule applicable 
 to every particular case. But the existence of any pheno- 
 menon cannot be known a priori, or known determinately, 
 so as to distinguish it from others, even if we could infer a 
 priori some existence or other. 
 
 For this reason Kant called the former two Principles 
 mathematical, as showing that phenomena are only possible, 
 in intuition as well as in perception, by reason of a mathe- 
 matical synthesis. Even in the second case we construct 
 the sun's light, for example, which means our degree of 
 
xn THE ANALOGIES REGULATIVE 173 
 
 sensation, by combining 200,000 moonlights. These prin- 
 ciples may therefore be called Constitutive (of phenomena). 
 The others must be called in contrast Regulative 1 (of the 
 existence of phenomena). We cannot dream of having 
 axioms and anticipations here. But although, when one 
 perception is given to us in a time-relation with other 
 (though undetermined) perceptions, we cannot say a priori 
 what other and liow much perception must be conjoined 
 with it, we can still declare how they are necessarily com- 
 bined in this mode, or particular description, of time. 
 Philosophical are very different from mathematical analogies. 
 These latter assert such an equality of quantitative ratios 
 that if the first three members be given, we can construct 
 the last (as in the ordinary ' Rule of Three'). A Philoso- 
 phical analogy 2 is an equation not of quantitative but of 
 qualitative ratios, where the first three terms give us, not 
 the fourth term itself, but the relation to a fourth term — in 
 fact only a rule for seeking it in experience, and an attribute 
 by which to recognise it there. An analogy of experience 
 is then a rule which gives us unity of experiences in various 
 phenomena, and is accordingly a Principle merely regulative 
 of phenomena. The same remark holds good of the 
 Postulates of empirical thinking generally, which regard the 
 form, the matter, and the relation of perception to our em- 
 pirical thinking generally. All these regulative Principles 
 differ from the constitutive, not in certainty, but in the 
 nature of their evidence. 
 
 It is of great importance to observe here that all these 
 synthetical principles have their meaning and validity not 
 
 1 They are regulative of intuitions, but nevertheless ' are constitu- 
 tive in relation to experience, inasmuch as they render the concepts 
 without which experience could not exist, possible a priori' (Kritik, 
 p. 407). 
 
 2 Cf. vol. ii. p. 162, note, for an example. 
 
174 PERMANENCE OF SUBSTANCE chap. 
 
 in the transcendental, but in the empirical use of the under- 
 standing ; we must hence subsume the phenomena not 
 directly under the Categories, but under their schemata, 
 which are the exemplifications of the Categories in Time. 
 These schemata are merely (cf. above, p. 146) sensuously 
 restricted syntheses of the functions of the pure Categories. 
 The pure Category has no such restriction. Hence we can 
 only combine phenomena by means of these Principles 
 according to an analogy with the logical and universal unity 
 of the Categories. While therefore we use the pure Cate- 
 gory in the statement of the Principle, we must employ in 
 the application of the Principle the schematised Category, 
 which is its restricted formula, and which will give us the 
 key to its use. 
 
 A. First Analogy. 
 
 Principle of the Permanence of Substance. 
 
 In all succession of phenomena substance remains per- 
 manent, and its Quantum in nature neither increases nor 
 diminishes. This statement varies slightly from that of the 
 First Edition, which asserts that all phenomena contain the 
 permanent (substance) as the object itself, and the change- 
 able as its mere determination, or way in which the object 
 exists. 
 
 Proof. 
 
 Phenomena are in time, in which alone, as the unchange- 
 able substratum, or permanent form of internal intuition, 
 simultaneity and sequence are represented. But time in 
 itself cannot be apprehended. We must therefore find in 
 the objects of perception the substratum which represents 
 time, and by relation to which we apprehend change or 
 simultaneity. If this were not so, the very notion of the 
 unity of time and its consequences just stated were im- 
 
CHANGE POSTULATES PERMANENCE 175 
 
 possible. But if we turn to phenomena, that which is the 
 substratum of all that is real is known as substance. Hence 
 the permanent element, by which all time-relations of pheno- 
 mena can alone be determined, is the substance in the pheno- 
 mena, or the real element, which always remains the same, 
 as the substratum of modifications. As this (ex hypothesi) 
 can never alternate, its quantum can never be increased or 
 diminished in nature. 
 
 This proof requires fuller development, being at first 
 sight both obscure and inconclusive. Kant accordingly 
 adds : our apprehension of multiplicity is ever successive 
 and changing. We can therefore never know whether as 
 objects of experience its parts are simultaneous or successive, 
 if it have not something permanent and constant as its basis. 
 Change and simultaneity are only so many modes of the 
 existence of this permanent. In it alone, therefore, are 
 time-relations (viz. the two just mentioned) possible, and 
 thus the permanent is the substratum of our empirical re- 
 presentation of time itself, and expresses this time as the 
 constant correlatum of all existence and change in pheno- 
 mena. For alternation affects not time itself, but pheno- 
 mena in time. By comparison with the permanent, and 
 thus alone, does existence in various parts of the ever-fleet- 
 ing equable time-series become a quantum, which we call 
 duration. Consequently, as time itself cannot be appre- 
 hended, this permanent in phenomena is the substratum of 
 all time-determinations, hence of the synthesis of pheno- 
 mena, hence the condition of the possibility of experience 
 itself. In all phenomena the permanent is the object itself, 
 the phenomenal substance ; all that alters is but a deter- 
 mination of this substance. 
 
 Philosophers have at all times agreed with the vulgar in 
 assuming this distinction, but expressed themselves a little 
 
176 ASSUMPTION OF THE PRINCIPLE chap. 
 
 more precisely, and have said : in all changes the substance 
 remains, and the accidents only alter. But, says Kant, I 
 can find nowhere any trace of proof for this very synthetical 
 proposition ; it is also seldom placed, as it ought to be, at 
 the head of the pure a priori laws of nature. To say that 
 substance is permanent, is a mere tautology. For this per- 
 manence is the sole ground for our applying the Category 
 of substance to phenomena. It should rather have been 
 proved that there was something permanent in all pheno- 
 mena. As this synthetical a priori proposition could never 
 be demonstrated dogmatically, from mere concepts, and as 
 no one ever considered that such propositions are limited 
 to the sphere of possible experience, and so provable only 
 by its possibility being shown to depend on them — under 
 these circumstances it was of course assumed, as being a 
 necessary want, but never proved. 
 
 The philosopher who determined the weight of smoke 
 by substracting that of the ashes from that of the wood 
 burnt, postulated that even by fire substance or matter 
 cannot be destroyed, but only changed in its form. 
 Similarly the proposition ex nihilo nihil fit, is but a conse- 
 quence of the Principle of Permanence, or of the constant 
 presence of the subject proper in phenomena. Being the 
 substratum of all time-determination, it must have been so 
 in all past time, as well as now and henceforth. We can 
 therefore only call a phenomenon substance because we 
 postulate its existence in all time — a thing inadequately 
 expressed by permanence, except that future duration is 
 inseparably bound up with the necessity of permanence 
 a parte ante. Thus the ancients never separated these 
 statements : 
 
 ' Gigni 
 De nihilo nihil, in nihilum nil posse reverti,' 
 
KANT AND BERKELEY 177 
 
 which we do, because we falsely apply them to things per se, 
 and therefore suppose they might shake the dependence of 
 the world on an ultimate cause. But they do apply strictly 
 to phenomena, and to them alone, as the unity of these in 
 experience would be impossible if new substances originated. 
 We have no other way of representing the unity of time 
 than by the identity of the substratum, which unites with 
 its bond all alteration. This permanence is accordingly 
 nothing more than the way we have of representing to 
 ourselves the existence of things (in phenomenon). 
 
 If We call particular attention to this argument on the 
 nature of the substratum of phenomenal objects, as of the 
 last importance when we come to consider Kant's Refuta- 
 tion of Idealism. It contains the really fundamental con- 
 trast between his system and that of Berkeley ; nay, it 
 constitutes, perhaps, the most peculiar feature of the whole 
 Kritik. All other idealists, or preachers of the Relativity 
 of knowledge, had assumed that the substratum of pheno- 
 menal objects was identical with the things per se. Berke- 
 ley in particular denied any such substratum, on the grounds 
 that being heterogeneous from phenomena, it could never be 
 perceived by any human faculty. Now Kant used this 
 very argument against our knowledge of things per se, yet 
 here he supports and explains our notion of substance, as 
 phenomenal, though a substratum, and separate from each 
 of our represet^^Bis. It is the abolishing of this sub- 
 stratum, because Re confounded it with things per se, 
 which brings Berkeley under Kant's censure of making our 
 external experience mere illusion — a censure which has 
 often puzzled commentators, and made them charge Kant 
 with injustice to his predecessor. But we must carefully 
 separate in Kant's system three things — (1) representa- 
 tions ; (2) the substance or substratum in which they inhere 
 
I7§ THE SUBSTRATUM OF PHENOMENA chap. 
 
 as phenomena, which must be in space and time, and is 
 therefore itself phenomenal ; (3) a hidden thing per se, 
 which by its action, likewise occult, produces in us both 
 the changing representations and the notion of their sub- 
 stratum in space, but is totally unlike either, and may be 
 one or many, the same for internal and for external pheno- 
 mena, or diverse — this we can never tell. 1 It will at once 
 be seen that Kant's originality lies in the position given to 
 the second element. Its permanence need not be a per- 
 ception, but a representation that we always can have 
 perceptions, because we are always in time, and we 
 can only know this by having perception in time. It 
 is then a representation of the permanent, not a per- 
 manent representation, 2 which forms the phenomenal sub- 
 stratum of the qualities of bodies. In a former exposition 
 of this passage, appended to Kuno Fischer's Commentary, 
 this all-important doctrine was not made clear. We trust 
 there will no longer be any doubt or difficulty as to Kant's 
 meaning. 
 
 The particular ways in which a substance exists are 
 called its accidents, which are always real, as affecting the 
 existence of the substance ; negations are merely assertions of 
 their absence. If we choose to assert a separate existence 
 for any of these real accidents, as, for example, for motion, 
 
 1 In direct proof of this subdivision of substrata, cf. vol. ii. p. 223, 
 where, in discussing the relation of Phenomena and Noumena (First 
 Edition), he says : ' But as to the reason why we, not satisfied with the 
 substratum of sensibility, have added noumena to the phenomena.' etc. 
 Perhaps we should not have spoken of these latter (noumena) as sub- 
 stratum, an expression which Kant uses in the phenomenal sense, but 
 U will serve to put our meaning more clearly, with this remark as to 
 Kant's use here added, for the sake of accuracy. He shows in the 
 sequel (p. 225) how loosely we must understand the word noumenon, if 
 applied to things per se. 
 
 2 Cf. note to second Preface, Kritik, p. xli. sub fur. 
 
xii CONTINUITY OF CHANCE 179 
 
 as an accident of matter, it is called inherence, as opposed 
 to the subsistence of the substance. As many mistakes 
 may thus arise, it is better and more accurate to speak of 
 accidents as above, and call them the ways in which the 
 existence of substance is positively determined. But as it 
 is impossible to avoid separating them logically from sub- 
 stance, and contrasting them with it, the Category of Sub- 
 stance is placed under the head of Relation — rather as con- 
 taining the conditions of such, than as being itself a 
 relation. 
 
 From the same point of view we can now clear up the 
 notion of change, which is not the origination or extinction 
 of the object changed, but a mode of its existence, following 
 upon another. The object is constant, and its condition 
 only alters. This paradox therefore is true : that only the 
 permanent (or substance) is changed, the transient suffers 
 no change, but alternation, or substitution of one deter- 
 mination for another. It follows that changes only can be 
 perceived in substances, never absolute origination or ex- 
 tinction. For if we would pass from not-being to being, 
 there must be a time when the object which originates is 
 not yet. As empty time is no object of perception, we 
 could only perceive it by placing in it things which last up 
 to the moment when the new thing originates. If so, it 
 would only be a determination of these things regarded as 
 permanents. The same difficulties meet us if we try to 
 conceive extinction. There is only one time, in which all 
 different times are placed not as co-existent, but as success- 
 ive, and accordingly permanence is a necessary condition, 
 under which alone we can determine phenomena, as things 
 or objects, in possible experience. The empirical criterion 
 of this substantiality will be discussed in connection with the 
 next analogy. 
 
1S0 UNIFORMITY OF SUBSTANCE chap. 
 
 If Perhaps Kant should have here stated as a legitimate 
 inference of the Principle, that we regard substance in 
 general as one, and of the same kind. Being the substratum 
 which represents to us the single time in which all our ex- 
 perience takes place, and not being identical with its mani- 
 festations, which constantly vary, we regard this sum of real 
 substance in the world as one, identical and indestructible. 
 We shall see that as to the criterion of its presence he 
 agrees with Hamilton in making resistance and its modes 
 the proper proof of permanence. But we must first under- 
 stand clearly his notions as regards action, before this feature 
 in his system can be clearly explained. 
 
 B. Second Analogy. 
 
 Principle of succession in time according to the law oj 
 Causality. 
 
 Alt changes come under the law of Cause and Effect. 1 
 
 Proof. 
 
 All apprehension of multiplicity is with us successive, 
 but whether the parts are successive in the object, as well 
 as in our apprehension, is another question. We may 
 doubtless call all conscious representations objects, but 
 what this word means, apart from mere representation, re- 
 quires deeper investigation. As mere representations they 
 can in no wise be distinguished from our apprehension of 
 
 1 The original form was : Everything which happens (begins to be) 
 presupposes something, which it follows according to a rule. With his 
 usual love of repetition, Kant added in the Second Edition two paragraphs 
 entitled Proof, which merely repeat what follows after in the original 
 edition. We therefore omit commenting on them. The same remark 
 holds good ot almost all the paragraphs inserted in the Second Edition 
 immediately after the statements of the Principles. 
 
xii SUCCESSION IN OBJECTS 1S1 
 
 them, and are always successive. Were these objects things 
 per se, no one could possibly guess from their representa- 
 tions in us how matters stood in the objects themselves, for 
 we can only know them by our successive representations. 
 It remains for me to determine what combination in time 
 belongs to the manifold of intuition apart from our apprehen- 
 sion, which is always successive. For example, I apprehend 
 the parts of a house successively, and yet no one thinks that 
 these parts are themselves successive. Yet remember that this 
 house is no thing per se, but a phenomenon. What, then, is 
 understood by the question : How is multiplicity connected 
 in the phenomena, which are in themselves nothing ? For 
 here the sum of our representations is contrasted with our 
 separate representations, and we desire to investigate their 
 agreement. It is evident that on Kantian principles we are 
 concerned here with nothing but the formal conditions of 
 empirical truth, or of correspondence between cognition and 
 its object ; and phenomena can only be considered separate 
 from the acts of our apprehension, if they come under some 
 necessary rule (imposed, of course, by the understanding), 
 which distinguishes them from every other apprehension and 
 makes a specific sort of combination necessary. That in 
 the phenomenon which contains the condition of this 
 necessary rule of apprehension, is the object. 1 
 
 As we cannot empirically perceive creation or extinction, 
 any apprehension of an event is a perception following 
 another (positive) perception. But though all our perceptions 
 are successive, in some we perceive that the sequence cannot 
 
 1 This 'condition,' Mr. Stirling tells us, is the Category (p. 298, 
 loc, cit.) It is not so. There must be something in a Sense 
 Apprehension which determines us to apply one Category rather than 
 another to it ; and this ' cue ' is what Kant means by the ' condition 
 of the necessary rule of apprehension,' viz. that in the sensation units 
 which demands any specific Category. 
 
182 DEPENDS ON A RULE 
 
 be reversed. When, for example, I see a boat carried down a 
 river, I cannot reverse the order, and apprehend the boat at 
 first below, and then above in the parts of the river 
 observed. In apprehending a house, on the contrary, I can 
 begin with any part indifferently. Here there is no rale. 
 But in the other case, which is an event that happens, there 
 is necessary order, and therefore a rule. 1 
 
 In this case, then, the subjective sequence of apprehension 
 must be deduced from the objective sequence, which gives a 
 necessary rule to the arbitrary order of the former. Thus I 
 can distinguish phenomena from my apprehension of them, 
 and say that it is determined by them. The rule implies 
 that in the antecedents of every event are to be found the 
 conditions upon which the event necessarily follows ; but the 
 rule will not allow me to reverse the order, and determine 
 the antecedents from the event, for the order in time can- 
 not be reversed, and this is a necessary condition of the 
 rule. Were there no such antecedents, and no such rule, 
 then all the sequences of our perception would be merely 
 subjective, and we should possess a mere play of representa- 
 tions referring to no object, as the order might be reversed 
 
 1 Mr. Stirling's objection to Kant's proof comes to this. The 
 irreversibility of the sequence, and therefore the necessity, lies in the 
 facts themselves, and hence the Category is superfluous {contradictory, 
 as Von Kirchmann says). Now the cue, viz. that in the sense apprehen- 
 sion which calls for the Category, is in the facts themselves ; but the 
 necessity is not and cannot be. A child has the same sense experience 
 as an adult, but there is no notion of necessity for it, as it has not ap- 
 plied the Category. When irreversibility has been cognised, the Cate- 
 gory is already applied. 
 
 Schopenhauer misinterpreted the whole proof, for he thought that 
 Kant's position was that sequence implied causality. All that Kant 
 said was that those sequences are causal which are irreversible. Atten- 
 tion to this point will enable the student to answer Mr. Caird's criti- 
 cisms, which have been fully discussed by Prof. Watson, as well as by 
 Mr. Stirling. 
 
IS A NECESSARY LAW 183 
 
 at pleasure. We should then not have two states of the 
 object in succession, but one apprehension following another 
 subjectively, without determining any object for us. It is 
 only by making my subjective apprehension objective by 
 means of the rule that I can ever have experience of an event. 
 
 This view opposes the old empirical derivation of causes 
 from the observation of uniform consequences [in fact from 
 what J. S. Mill calls the Method of Agreement]. Were this 
 the real derivation of the notion, it would of course lose, 
 with its a priori character, its objective universality and 
 necessity. But the case is quite similar to that of space 
 and time, as regards which we only draw clear notions from 
 experience, because we have ourselves put them into ex- 
 perience, and even made experience possible by doing so. 
 We shall not deny that experience is of use in giving us 
 clear examples of the rule ; but the existence of the rule as 
 a condition of the synthetical unity of phenomena in time 
 is nevertheless the a priori basis of this very experience. 
 
 We are therefore called upon to show by an example 
 that even in experience we never attribute the sequence to 
 the object, nor separate it from our subjective apprehension, 
 except for a rule compelling us to adopt one peculiar 
 order of perceptions ; nay, more, that this compulsion is 
 what makes succession in the object possible. Seeing that 
 all our representations are equally internal modifications 
 given to us in the lapse of time, how is it that we add to 
 them an object, — in other words, that over and above their 
 subjective reality, as modifications (of mind), we give them 
 an objective reality ? It cannot be a mere reference of 
 some representations to some other, for then the same 
 question must still arise as regards that other. What, then, 
 is this new dignity, this reference to an object, which we give 
 our representation ? On investigation it will be found to 
 
GENESIS OF OBJECTS 
 
 affect representations in no other way than combining them 
 necessarily in a particular way, and so subjecting them to a 
 rule ; it is only by means of this rule that they on the other 
 hand obtain an objective signification. As all my apprehen- 
 sion is successive, it is only when I assume a necessary 
 reference to an antecedent state that I assert an event, or 
 that something has happened ; that is to say, I cognise an 
 object, which I put into a definite place in time, a place 
 fixed by the foregoing circumstances. To say, then, that 
 something happens, implies an antecedent, which defines 
 the time. But this time must be fixed by a rule, which 
 cannot be reversed, and moreover the antecedent must be 
 necessarily followed by the event in question. Hence 
 arises an order among our representations, in that the present 
 points to an antecedent state, which, though an undeter- 
 mined correlate, refers to its consequent as determining it, 
 and necessarily connected with it in the series of time. If 
 it be a formal condition of our perceptions that the prior 
 moment of time determines the posterior, it is of course a 
 necessary latv of the empirical representation of time that 
 phenomena, by which alone we can know empirically this 
 continuity in time, should be similarly determined. 
 
 No experience is possible without the understanding, 
 and its first duty is to render objects not distinct, as former 
 philosophers have said, but possible. This is done by trans- 
 ferring to phenomena and their existence the order of time, 
 and assigning to them a fixed place in it. This place is 
 fixed not by the relation of phenomena to absolute time, as 
 it cannot be apprehended per se, but to one another. An 
 event, then, is a perception belonging to possible experience, 
 which becomes actual by my regarding the phenomenon to 
 be determined in place as to time, and so by my regarding 
 it as an object, which I can always find by a rule in the 
 
xii CAUSALITY WITHOUT SUCCESSION 1S5 
 
 connection of my perceptions. The rule is, that there arc 
 antecedent conditions, upon which the event necessarily 
 follows. Consequently, says Kant, the Principle of Sufficient 
 Reason is [in this sense, viz.], as to succession in time, the 
 basis of possible experience. 
 
 After again recapitulating, for about the tenth time, the 
 momenta of his proof, Kant approaches a new difficulty. 
 We have hitherto confined the law of Causality to succes- 
 sions, whereas, in fact, it applies to simultaneous pheno- 
 mena. A room is warmed by a fire now present in it. 
 Here, then, is cause and effect, but no succession in time. 
 The greater part of the efficient causes in nature are of this 
 kind, and their effects are only consequences, because the 
 cause does not produce its whole effect in an instant. But 
 in the moment, when the effect first originates, it must 
 be [as it were, in contact, and] simultaneous with the 
 causality of its cause, for had this causality ceased for 
 an instant, the effect could not have arisen. It is not, 
 then, the lapse, but the order of time which is of import- 
 ance ; the interval may be evanescent, but still the time- 
 relation remains. If I place a bullet on a soft cushion, 
 it produces a cavity, which is simultaneous with the 
 bullet. But I nevertheless distinguish them by the 
 time-relation of their dynamical connection. The cavity 
 follows upon my laying the bullet down ; the bullet will not 
 follow upon the cavity, supposing it otherwise produced. 
 
 This Causality leads us to the notion of action, action to 
 that of force, and so we reach that of substance. Kant 
 declines to turn aside in his discussion of synthetical a priori 
 cognition for the sake of expounding mere analyses, but 
 here makes an exception as regards the empirical criterion of 
 substance, so far as it appears to manifest itself better and 
 more easily by action than by the permanence of the pheno- 
 
ACTION 
 
 menon. Where action is — that is, efficiency and force — there 
 must we seek for substance, the seat of a fruitful source of 
 phenomena. But if we proceed to explain what we mean 
 by substance, it is hard to avoid reasoning in a circle. 
 How can we directly infer from an action the permanence of 
 that which acts, and yet this is the essential and peculiar 
 attribute of phenomenal substance ? The problem cannot 
 be solved by the ordinary analytical method, but presents 
 to Kant no peculiar difficulty. Action implies the relation 
 of the subject of the Causality to its effect. 1 Now, as all 
 effects consist in events, which are changeable and success- 
 ive, the ultimate subject of the changeable is the permanent, 
 as being the substratum of all that changes, and this is sub- 
 stance. For according to the Principle of Causality actions 
 are always the first cause of all alterations of phenomena, 
 and cannot therefore be in a subject which itself alters, as 
 this alteration must then be determined by other actions, 
 and another subject. Action is accordingly a sufficient 
 criterion of substantiality, without my requiring to seek out 
 permanence by a comparison of perceptions, in which way 
 we could hardly obtain the strict universality required in 
 the notion of substance. That the first subject of the 
 causality of all origination and disappearance cannot itself 
 originate and disappear in the field of phenomena, is a cer- 
 tain inference, and coincides with the notion of substance 
 in phenomena. 2 
 
 1 This is Kant's consistent definition of action, which shows that he 
 distinguishes between the cause and its causality in a manner foreign to 
 Hamilton and Mansel. 
 
 2 The result of this argument is that the permanent is the simul- 
 taneous condition, not the cause, as Kuno Fischer seems to understand 
 it {Comm. p. 119), for Kant expressly states {Kritik, p. 259), that sub- 
 stance and accident do not stand in the relation of a regressive series. 
 But the question remains, What effects suggest to us action, as dis- 
 
HOW INFERRED 187 
 
 When something happens, the mere happening, apart 
 from the particular thing that happens, the passage from 
 the non-existence of a state to its existence, is worth in- 
 vestigating. Affecting not substance, but its states, it is 
 only a change, not an origination from nothing. If this 
 origination is regarded as proceeding from a foreign cause, 
 we call it creation, which cannot be admitted in phenomena, 
 as its very possibility would destroy the unity of experience. 
 What may happen if I regard things per se is not here 
 under discussion. But how a thing can change its state in 
 time, we have not the least notion a priori. We require 
 actual forces empirically given to us. But the form of 
 every change, as opposed to the matter, this can be deter- 
 mined by the law of Causality and the conditions of time 
 a priori. Observe, says Kant in a note, that I am not 
 speaking of the change of relations generally, as when a 
 body moves uniformly, but of a change of state, as when it 
 changes the quantity of its motion. 
 
 We now approach another inference. When a substance 
 changes its state, the new state b, even when only differing 
 in quantity from the first a, is related to it as to o or non- 
 
 tinguished from mere causality ? This question Kant has answered in 
 pp. 169, 379 of the Kritik, where he distinctly suggests impenetrability. 
 He is still more explicit in p. 193 (when discussing Leibnitz's system) : 
 ' Substance in space we only cognise through forces operative in it [the 
 term force is with him equivalent to action] either drawing others to- 
 wards itself (attraction) or preventing others from entering into itself 
 (repulsion and impenetrability) ; we know of no other properties that 
 make up the notion of substance phenomenal in space, and which we 
 term matter.' In other words, while we attribute other qualities of a 
 substance to causes acting upon it from without itself, we consider im- 
 penetrability (Locke's solidity) as the action of the substance itself upon 
 us. This is the plain, common sense belief. The reader will see that 
 Hamilton might have found both his division of the Qualities of bodies 
 into three classes, and the importance of Resistance and its modes, in 
 Kant's system. 
 
CONTINUITY PROVED 
 
 existence, for even in this case b - a, which did not before 
 exist, has come into being. How does a thing pass from 
 this a to b ? Between any two moments there is time, and 
 between any two states a distinction, which has a quantity 
 (by our Axiom of Intuition). Every change, then, takes 
 place in the time-interval between the moment which deter- 
 mines the former and the moment which determines the 
 latter condition. But every change has a cause showing 
 its causality through the whole time in which the effect is 
 being originated. It follows that all changes are produced 
 not suddenly, but gradually, and through lesser degrees 
 continuously up to the full result. The action of Causality, 
 so far as it is homogeneous, Kant calls a moment. The 
 change does not consist in these moments, but is gene- 
 rated by them as their effect. This is the law of Continuity 
 of all changes depending on the fact that neither time nor 
 anything appearing in time can be a minimum. No dis- 
 tinction of state can be the smallest possible, and so there 
 are infinite gradations from o to any given state a. 
 
 The uses of this Principle in the science of nature do 
 not concern us ; but how such a Principle, however right 
 and obvious, can be possible a priori, this is worthy of note, 
 as there are so many pretended extensions of our knowledge 
 by pure reason, without proper deduction, in our critical 
 sense. 
 
 It is simply this. Every increase in our empirical know- 
 ledge is nothing but an additional determination of our 
 internal sense, a progress in the determination of time, 
 whatever the objects may be. Accordingly, every transition 
 in perception to some succeeding perception is a gradual 
 generation of the latter in time, which is always a quantity, 
 through degrees, which are consequently never absolute 
 minima. Thus, then, we can assert a law a priori con- 
 
WHAT IS SIMULTANEITY ? 1S9 
 
 cerning changes as regards their form. The understanding, 
 by means of the unity of its apperception, enables us to 
 determine all the places in this time continuously, through 
 the series of causes and effects, which impose a necessary 
 order and sequence. 
 
 C. Third Analogy. 
 
 Principle of Co-existence, after the law of mutual actinti 
 in community. 
 
 All substances, so far as they can be perceived as being 
 together in space, are in systematic (durchgangiger) mutual 
 action, or community? 
 
 Proof. 
 
 Things are said to be simultaneous empirically when 
 their perceptions can follow mutually, as, for example, I may 
 perceive the earth and then the moon, or vice versa ; or 
 I can consider the various parts of a house in any order I 
 please. This means that they co-exist in the same time. 
 But I cannot perceive time itself, and take these perceptions 
 out of it, and all perception is successive ; so, then, without 
 some further clue, I should merely infer that the one is 
 there, when the other is not, and vice versa, but not that 
 the objects were simultaneous, that where one is the other 
 must be also, in order that the perceptions may succeed 
 each other mutually. We require, then, a concept of the 
 understanding to declare this simultaneity to be objective. 
 But the relation of substances, one of which contains deter- 
 minations based upon the other, is the relation of influence ; 
 
 1 The First Edition is not at all so cautiously worded: 'All substances, 
 so far as they are together, stand in systematic community (mutual 
 action).' 
 
190 COMMUNITY 
 
 and if this be reciprocal, it is the relation of Community, or 
 reciprocal action. The simultaneity of substances in ex- 
 perience cannot be cognised, except under this supposition, 
 which is accordingly a condition of the possibility of things 
 as objects of experience. 
 
 If It will be noted that Kant carefully avoids stating 
 this law as one of reciprocal causation — an error into which 
 Kuno Fischer and other commentators have fallen. The 
 statement of the Principle in the Second Edition confines 
 Community to relations of space, and we think rightly. 
 For there it is strictly true that if our experience is to have 
 any unity or connection, the places of all objects are mutually 
 determined by each other. I determine the place of a 
 table, a room, a house, a town, by its relation to other 
 simultaneous objects — nay more, the very earth itself by its 
 position in the Solar System. These objects do not stand 
 in the relation of reciprocal cause and effect, but in that of 
 mutual influence, or community, as Kant has above strictly 
 and accurately defined. The exposition of the First Edition, 
 which follows that inserted in the Second under the head of 
 proof, is not so accurate, though of greater length, and con- 
 tains a statement which has led to this misunderstanding. 
 ' That only can determine the position of anything in time 
 which is the cause of it, and [or ?] of its determinations. 
 Therefore must every substance (for it can only be a con- 
 sequence as regards its determinations) contain within itself 
 the causality of certain determinations in another, and at 
 the same time the effects of the causality of this other — in 
 other words, they must be in dynamical community, or 
 reciprocal action.' It is not correct to speak of the place 
 occupied in space by a substance as part of its causality, 
 and yet this is the principal ' ground of determination ' 
 which it possesses as regards other substances. We hold, 
 
COMMUNIO AND COMMERCIUM 191 
 
 then, that Kant, when inserting his new paragraph, should 
 have modified his old proof. But to resume. 
 
 The word community may mean either communio or 
 commercium. We use it in the latter sense, as that without 
 which the former could never have been ascertained. We 
 can see in experience that but for mutual relations, but for 
 a continuity of perceptions, influencing each other (at least 
 as regards place) in space, we could never proceed from 
 one perception to another. There may be such a thing as 
 empty space if you like, but its occurrence in the field of 
 phenomena would break up and destroy the unity and 
 system of our experience. This commercium, then, controls 
 our ever successive apprehensions, and compels us to place 
 those substances which are subject to it, not as successive, 
 but as simultaneous, and objectively so. Thus phenomena 
 are brought together into a composiium reale, or system of 
 simultaneously existing objects. 
 
 The three dynamical relations, which are the sources of 
 all others, are therefore Inherence, Consequence, and Com- 
 position. 
 
 By Nature, in the empirical sense, we understand the 
 connection of phenomena in their existence, according to 
 fixed rules. There are therefore certain laws or rules, 
 which are a priori conditions of nature ; on those the em- 
 pirical rules must depend. Our analogies express the unity 
 of nature in this connection of phenomena under certain 
 exponents, which express nothing but the relation of time 
 (as comprising all existence) to the unity of apperception. 
 They declare as their whole result : that all phenomena 
 must be comprised within a single Nature, since without 
 their a priori unity, no unity of experience or determination 
 of objects in it would be possible. 
 
 Kant's method of proof in establishing these transcend- 
 
192 CRITICAL METHOD OF PROOF CHAr. XII 
 
 ental laws of nature is declared by him a model according 
 to which all such demonstrations must proceed. He con- 
 trasts it with the dogmatic method, by mere analysis of con- 
 cepts. By this method such synthetical Principles could 
 never be attained, for where is the medium, the x (cf. above, 
 p. 36) in which we are to seek the evidence of such syn- 
 thesis ? It is in the possibility of experience, as a system 
 or complex of cognitions, or acts of knowing, under which 
 all the objects of cognition must come. Here we found 
 rules of synthetical unity a priori, and so anticipated nature. 
 It was for want of this clue that so many abortive attempts 
 have been made to prove the Principle of Sufficient Reason. 
 The other Analogies, though often assumed, were never 
 even stated, which could not have been the case had the 
 infallible clue of the Categories been known. Thus the unity 
 of the world, he adds in a note, is a mere inference from 
 our Third Analogy ; and were not the mutual influence of 
 objects a necessary condition of their very existence, the 
 unity of their connection could not be inferred from it. 
 
CHAPTER XIII 
 
 (4.) The Postulates of Empirical Thinking Gene- 
 rally. 
 
 1. What agrees with the formal conditions of experience 
 (intuition and concepts) is possible. 
 
 2. What is connected with the material conditions of 
 experience (sensation) is actual. 
 
 3. What has its connection with the actual, determined 
 according to universal conditions of experience, is (exists) 
 necessary. 
 
 § 1. Exposition. The Categories of Modality have this 
 peculiarity, that as predicates they do not increase the attri- 
 butes of the subject, but only alter its relation to our know- 
 ing faculty. Let the concept of a thing be ever so perfect, 
 I may still inquire whether it is possible or actual, or if so, 
 whether it be necessary ; and each of these mean, what re- 
 lation has it to our faculties of experience ? 1 These Prin- 
 ciples of Modality are, then, nothing but expositions of the 
 
 1 Hence it is that Kant afterwards calls propositions of existence 
 subjectively synthetical, as adding no objective predicate ; and hence it 
 is, probably, that older philosophers sought to discover existence analy- 
 tically in the notion of God. Locke, however, saw the distinction, un- 
 consciously perhaps, but nevertheless clearly, and accordingly separates 
 judgments of existence into a fourth class, distinct from both analytical 
 and objectively synthetical judgments. 
 I O 
 
194 POSSIBILITY 
 
 notions of possibility, actuality, and necessity in their em- 
 pirical use, and also restrictions of all Categories to mere 
 empirical use. For if these Categories are to be anything 
 more than mere analytical expressions of the form of thinking, 
 and are to apply to things, they must be applied to possible 
 experience, in which alone these things can be given us. 
 
 § 2. T/ie Postulate of Possibility. The objective form 
 of experience in general contains all the syntheses necessary 
 for cognising objects. A concept implying a synthesis is 
 void, and applies to no object, except the synthesis is 
 borrowed from experience (as in empirical concepts), or is an 
 a priori condition of experience (as in pure concepts), which 
 still belongs to experience, as its object must be there found. 
 It is, of course, a necessary logical condition that the con- 
 cept shall contain no contradiction, but this is far from 
 proving the objective reality of the object denoted by it. 
 For example, a bilinear figure is not self-contradictory ; its 
 impossibility depends on the conditions of space, in which 
 we must construct it ; these conditions have objective reality, 
 because they contain in them the general form of experience. 
 The Postulate is therefore of great importance. Supposing 
 I represent to myself the pure Categories of Substance, of 
 Causality, of Community, how can I infer from these arbitrary 
 syntheses themselves that they are to be found in the field 
 of experience ? I can only discover their objective reality 
 and transcendental truth by perceiving that they express a 
 priori the relations of all our perceptions, and they are thus 
 independent indeed of experience, but not of the form of 
 experience and its synthetical unity. 
 
 But if we endeavour to construct new notions of sub- 
 stances, forces, and community from the matter of our per- 
 ception, without finding in experience itself examples of 
 such connections (syntheses), we are adopting mere hallucina- 
 
ACTUALITY 195 
 
 tions, with no evidence of their possibility. Such imaginary 
 notions cannot be deduced from the a priori conditions of 
 experience, but must have their possibility shown a posteriori, 
 or not at all. A substance permanently present in space, 
 without filling it, like the tertium quid of philosophers, a 
 peculiar power of intuiting the future, without inferring it, 
 or a power of being ; in a community of thought with other 
 men, however distant from us — these are concepts whose 
 possibility has no basis, as not being founded on the known 
 laws of experience. They are not self-contradictory, but 
 have no claim to objective reality, or the possibility of such 
 an object as we here imagine. Here Kant is not concerned 
 with possibilities derived from experience, but with those 
 through a priori concepts, and such cannot be obtained 
 from mere concepts, but from the objective conditions of 
 experience generally. 
 
 It might be thought that the possibility of a triangle 
 could be inferred from our concept of it, which is quite 
 independent of experience, and to which we can give an 
 object by a pure a priori construction. But this is after all 
 only the form of an object, and would remain a creature of 
 the imagination, the possibility of which is doubtful, until 
 we see that space is the formal condition of all external 
 phenomena, and that the very same formative synthesis by 
 which we construct the triangle a priori is exercised in 
 apprehending a phenomenon and obtaining from it an 
 empirical concept ; then such an object becomes possible. 
 The same can be shown in any other case, such as that of 
 continuous quantity, etc. 
 
 § 3. The Postulate of Actuality requires perception or 
 conscious sensation, not indeed of the object itself, but of 
 its connection with some actual perception, according to the 
 analogies of experience. In the mere concept of a thing not 
 
196 NECESSITY 
 
 a single mark of its existence can ever be discovered. 
 However complete internally, all this has nothing to do 
 with the question, Is it given to us? — given in such 
 manner that the perception might come to us before the 
 concept. For perception is the matter of the concept, and 
 alone gives it actuality. Prior to the perception of the 
 thing, however, and therefore comparatively a priori, we can 
 assert it to be actual, when we infer it by the analogies 
 from other actual perceptions ; but these tell us that if we 
 pursue the clue given us, or that if our senses were more 
 acute, we should have actual perception of it. Such, for 
 example, is the case of magnetic particles, which our senses 
 are too coarse to perceive directly. But we must strictly 
 follow the laws of the empirical connection of phenomena. 
 
 If These rules for proving actual existence mediately 
 are directly opposed to idealism in the ordinary sense, which 
 asserts the actuality of mental phenomena, but will not con- 
 cede it to anything else. Kant accordingly turns aside here 
 to give (in the Second Edition) his celebrated and much 
 decried Refutation of Idealism. As the clear understanding 
 of this argument will require a comparison of other passages, 
 and also some polemical discussion, we shall postpone it to 
 a separate chapter, and shall here follow the remainder of 
 the exposition as regards the Postulates of Modality. 
 
 § 4. The Postulate of Necessity of course regards 
 material necessity in existence, not logical necessity in the 
 combination of concepts. But as actuality can be inferred 
 only comparatively a priori by the connection of perceptions, 
 so this necessity must be inferred, as necessarily implied by 
 it ; and as there is no existence necessarily given on con- 
 dition of others, except the effects of given causes, it is not 
 the existence of substances, but of their states, which we 
 can know to be necessary, and this in consequence of other 
 
XIII FOUR A PRIORI DYNAMICAL LAWS 197 
 
 states which imply them as their effects. Hence the 
 criterion of necessity is simply the law of possible experience, 
 that every event is determined a priori by its cause in 
 phenomena. So we cognise the necessity of effects in nature, 
 of which the causes are given us, and the law reaches no 
 further, not even to the existence of things as substances, 
 for we cannot perceive them as empirical effects. The law, 
 therefore, that every event is hypothetically necessary is a 
 rule of necessary existence, without which we can have no 
 nature. In other words, in mundo non datur casus is an a 
 priori law of nature, and this necessity being hypothetical 
 is rational. So is the other brocard, non datur fatum. 
 These laws bind the myriad variety of changes into one 
 nature, or synthetical unity imposed by the understanding. 
 The first is properly a consequence of the Principle of 
 Causality, the second of those of Modality. The Continuity 
 of time shown in the earlier Principles adds another, in 
 mundo non datur saltus ; that of space, a fourth, non datur 
 hiatus, as we saw that a vacuum was inadmissible. These 
 Principles being of transcendental origin can be ranged 
 according to the clue of the Categories, and Kant says any 
 .practised reader can do it for himself. He seems to regard 
 all these propositions as of the dynamical, not the mathe- 
 matical order. If this be so, saltus, casus, and hiatus are 
 respectively excluded by the Principles of Substance 
 (p. 179), Cause, and Community (p. 191); whilst the Prin- 
 ciples of Modality exclude fatum. 
 
 § 5. Whether the field of possibility be greater than that 
 of actuality, and this again than that of necessity — these 
 are interesting questions, and purely within the jurisdiction 
 of Reason. They mean to inquire, whether the whole 
 complex of phenomena belong to one experience, and to it 
 alone, or whether they may on the contrary belong to some 
 
198 THE POSSIBLE AND THE ACTUAL CHAi>. 
 
 other experience as well. Our Understanding lays down 
 its rules subject to the conditions of our sensibility. 
 Different forms of this sensibility, and different forms of 
 the understanding, even were they possible, we cannot 
 grasp or explain in any possible way ; but could we do so, 
 they would certainly not belong to that experience in which 
 objects are given to us. The Understanding, therefore, 
 has nothing to say to such questions. Apart from these 
 considerations the proof usually urged in favour of a very 
 extended field of possibility is palpably weak. ' All that is 
 actual,' say they, ' is possible, and then by conversion per 
 accidens, some possible is actual, which indicates that the 
 actual does not exhaust all the possible. It seems, then, 
 as if we added something to the possible to make it actual.' 
 Such a view Kant disputes. Whatever, he says, could be 
 added to the possible must be the impossible. What is 
 added is not in the possible, but in my understanding — that 
 is to say, a connection with some actual perception is added, 
 and this does not restrict the field of the possible, but brings 
 some part of it into a new relation towards me. All that 
 is possible, then, can become actual, and there is no sphere 
 of phenomena apart from our experience to be inferred from 
 this shallow argument. Let it be remembered that what 
 is possible only under conditions which are themselves 
 possible, is in no sense absolutely possible, or possible in 
 every respect {in aller Absicht). Yet this is the only sense 
 in which we can employ the term possibility, when inquiring 
 into the possibility of things beyond experience. These 
 questions, as transcending experience, belong properly to 
 the Reason, and will be considered more fully hereafter. 
 
 But in what sense are the Principles of Modality 
 Postulates ? Certainly not in the sense lately given it 
 by philosophers, and opposed to that of the mathematicians 
 
xni SUBJECTIVELY SYNTHETICAL PRINCIPLES 199 
 
 from whom they borrowed it — I mean in the sense of 
 assuming a proposition as immediately certain, without 
 proof or justification. Such a proceeding must ruin our 
 whole Kritik, in which for every synthetical judgment we 
 demand either a proof, or at least a deduction of its 
 claims. 
 
 But as was already observed, these Principles are synthet- 
 ical in a peculiar sense, not adding to the concepts of pheno- 
 mena any attribute, but merely adding a particular relation 
 to the mind, and hence subjectively synthetical. Thus a 
 thing, without in the least changing its attributes, is con- 
 sidered by us as possible, actual, or necessary, according as 
 it is connected with the formal conditions of experience, 
 with perception, or with the connection of perceptions by 
 concepts. These Principles, then, express the action of 
 the understanding, by which a certain concept is generated. 
 Now this is precisely the mathematical notion of a postulate. 
 We are told to describe a circle with a given line about a 
 given point, and this requires no proof, because the very act 
 postulated creates the notion of the figure which we want. 
 In like manner our present Postulates direct the mind how 
 to put together it concepts of things. 
 
 § 6. Kant adds in the Second Edition a General Remark 
 on the System of the Principles, which recapitulates the 
 general result of the discussion. In the first place, it was 
 shown that from the pure Categories we can never ascertain 
 the possibility of a thing, but we must always have recourse 
 to intuition, whenever we desire to prove objective reality. 
 Without doing so we have no cognitions, but merely forms 
 of thought. For the same reason no synthetical proposition 
 could ever be constructed, far less proved, from mere Cate- 
 gories. Take, as an example of the latter, the proposition : 
 Every contingent existence has a cause. We could never 
 
THE CONTINGENT 
 
 get beyond this point, that without such relation to a cause 
 we cannot comprehend the existence of the contingent. 
 But how does it follow that it is also a condition of things 
 as well as of our thoughts ? Our Principle of Causality 
 was proved for things of experience, and for them alone ; it 
 was valid of objects given in empirical intuition, and not 
 proved from mere concepts. It is indeed true that the 
 proposition, Everything contingent has a cause, is plain 
 enough from mere concepts, but only by so constructing 
 the concept of contingence as to contain not the Category 
 of Modality (something, of which we think the non-exist- 
 ence), but that of Relation (something, which can only 
 exist as the effect of something else). It is then an iden- 
 tical proposition to say, what can only exist as an effect 
 must have a cause. So it is that all examples of contingent 
 existence are taken from changes, and not from the possi- 
 bility of conceiving the reverse ; because a change depends 
 upon its cause as a condition, and this may not exist, hence 
 it is an identical proposition to say, that a contingent thing 
 must have a cause. On the other hand, the ancients, who 
 could very well conceive the non-existence of matter, did 
 not therefore make its existence contingent. Even the 
 successive being and not-being of any given state of a thing 
 does not prove the contingency of this state, except it be 
 shown that at the same moment in which it was in one 
 state it might have been in the opposite. Ex. gr., to show 
 the motion of a body to be contingent, we must prove that 
 it could at the same moment have been at rest. 
 
 It is still more remarkable, that in order to show the 
 possibility of things in accordance with the Categories, and 
 so show the objective reality of the latter, we require external 
 intuitions. To show something permanent in intuition 
 corresponding to the Category of Substance we require an 
 
APPLICATIONS 
 
 intuition of matter in space, as time is always fleeting. So 
 as to Causality we must take change in space or motion, to 
 explain to us the combination of contradictory states in the 
 existence of one and the same thing. For the same reason, 
 we can only represent time by a line, and internal succes- 
 sion by the drawing of a line, as the internal sense gives no 
 permanent intuition, whereby to perceive changes. So as 
 to Community : we cannot understand it or illustrate it 
 except by the relation of substances in space, as before ex- 
 plained. Thus Leibnitz, when he gave to substances, as 
 thought by the understanding, a community, was compelled 
 to bring in the Deity to produce it, as being in itself incon- 
 ceivable. 
 
 The importance of this remark will be felt in the Re- 
 futation of Idealism, which we shall now discuss, and it is 
 also of the last importance in refuting the false notions as 
 to our knowledge of self, which we shall meet in rational 
 psychology. 
 
If CHAPTER XIV 
 KANT'S attitude towards idealism 
 
 § i. There is no part of the Kantian philosophy clearer 
 or more precise than his careful distinction of his own posi- 
 tion as a mere critical idealist from that of Descartes the 
 sceptical, and Berkeley the dogmatical idealists. There is 
 no more masterly discussion in all the Kritik than his 
 refutation of the fourth paralogism of rational psychology, 
 as it stands in the First Edition, 1 and yet, owing to this 
 fuller statement being greatly curtailed, and also transposed, 
 in the later editions, its importance has been overlooked, 
 and its import strangely falsified. Even now, the German 
 Kantians are in the dark on the subject, and Kuno Fischer 
 in his latest work {Critique of Kant, 1883) repeats his for- 
 mer blunders, but brings forward no new arguments. 
 
 The passage which misled the Germans was the well- 
 known Refutation of (material or problematical) Idealism 
 {Kritik, p. 167), not introduced into this edition as a novelty, 
 but transferred to a new place from its old place among the 
 Paralogisms. Internal experience was preferred to external 
 by Descartes, and was called certain as opposed to the 
 uncertainty of the knowledge gained from external sense. 
 In the First Edition, Kant, refuting this (supposed) superior 
 
 1 Cf. vol. ii. pp. 242 sqq. 
 
chap, xiv THE TWO EDITIONS 203 
 
 certainty of internal experience, showed it to be merely 
 phenomenal, and therefore of the same kind as external 
 experience. This is the gist of his long discussion on the 
 fourth Paralogism of the ideality of external relations. In the 
 Second Edition, Kant, supporting the original dignity and 
 importance of external experience, showed it to be not only 
 equal to internal experience in certainty, but logically prior 
 to it, in that it is presupposed by internal experience. This 
 discussion naturally comes into the Analytic, being an im- 
 portant question in the Metaphysic of phenomena. Hence 
 it becomes necessary to transfer the argument from its 
 original place in the Dialectic to this earlier position. But 
 the identity of the two discussions is perfectly obvious to 
 anybody who will study them, though it has been ignored 
 by most commentators. This oversight was the main 
 cause of their subsequent errors. Regarding the new form 
 as a new argument, believing Schopenhauer's audacious 
 assertion about the retracted idealism of the First Edition, 
 they imagined that Kant had inserted this passage to prove 
 the existence of things per se in space, — an absurdity so 
 monstrous, in Kant's system, that it must indeed be, as they 
 said, the sign of a broken-down intellect. The absurdity, 
 however, is not in Kant, but in his critics. A short com- 
 parison of the two passages, and an exposition of their true 
 import, will show that they are consistent, logical, and 
 strictly necessary to Kant's system. 
 
 § 2. There are other passages, in which he officially 
 approaches the question, and in a polemical attitude. In 
 his Prolegomena (ii. pp. 56-62), as well as in an appendix 
 to his Prolegomena, where he replies to the strictures of 
 Garve in the Gottingen Gelehrte Anzeigen, he gives a short 
 and precise sketch of his attitude as regards Berkeley and 
 other idealists ; and we shall quote the relevant part of it 
 
2o 4 PARALLEL PASSAGES chap. 
 
 verbatim in the course of this exposition. The discussion 
 is so full and clear in its earliest form, that we need do 
 little more than point out the slight modifications in 
 form with the identity of matter, in the corresponding 
 places. 
 
 But there is one more passage, of capital importance, in 
 which Kant repeats his assertion of critical idealism. It is 
 the sixth section of the Antimony of the Reason in the 
 Kritik, and is entitled Key to the Solution of the Cosmological 
 Dialectic. In this discussion, which maintained its form 
 unaltered in the two editions, the reader will find a precise 
 and official re-assertion of the views contained in the fourth 
 Paralogism of the First Edition, and it was probably this 
 very recurrence which made him curtail that discussion. 
 There are no other passages of importance in Kant's works 
 which bear directly on the question. We have then these 
 six : (i) the discussion of the fourth Paralogism in the First 
 Edition; (2) the substance of this argument transferred 
 from the Dialectic to the Analytic, and entitled Refutation 
 of Idealism, in the Second Edition ; (3) a note appended to 
 the preface of the Second Edition, in which he expands and 
 explains this Refutation ; (4) the sixth section of the Anti- 
 nomy of picre Reason, in both editions; (5) his polemical 
 statement under the heading Remarks 2 and 3, at the 
 conclusion of the section, ' How is pure Mathematic pos- 
 sible,' in the Prolegomena (vol. ii. pp. 54-62); (6) Kant's 
 statement of his doctrine in reply to Garve, published at 
 the end of his Prolegomena, between the appearance of the 
 first and Second Editions of the Kritik. 
 
 ' Let us then see,' says Kant, in this last passage, ' what 
 sort of idealism it is which runs through my whole work, 
 though it be far from constituting the soul of the system. 
 The attitude of all genuine idealists, from the Eleatic school 
 
xiv KANT AND BERKELEY 205 
 
 to Bishop Berkeley, is contained in this formula : all cogni- 
 tion through sense and experience is nothing but mere illu- 
 sion, and only in the Ideas of the pure understanding and 
 Reason is there truth. The fundamental principle, on the 
 contrary, which thoroughly rules and determines my idealism, 
 in this : all cognition of things from mere pure under- 
 standing or pure Reason is nothing but illusion, and only 
 in experience is truth. As this is the exact opposite of that 
 proper idealism [which I have just described], how did I 
 come to use the term for an opposite purpose ? The solu- 
 tion can easily be found from the context of the book. 
 Space and time, with all they contain, are not the things, 
 or their properties per se, but merely belong to their pheno- 
 mena ; so far I am in agreement with these idealists. But 
 they, and among them especially Berkeley, considered space 
 a mere empirical representation, known to us, with all its 
 determinations, like the phenomena in it, only by means of 
 experience or perception. I show on the contrary that 
 space (and time, which Berkeley overlooked), with all their 
 determinations, are known a priori, as the pure form of our 
 sensibility. Hence it follows that as truth depends upon 
 universal and necessary laws as its criteria, experience can- 
 not have for Berkeley any criteria of truth, for he puts 
 nothing a priori at the basis of its phenomena. It follows 
 further that experience is nothing but illusion ; whereas in 
 my system space and time, combined with the Categories, 
 prescribe their laws a priori to all experience, and this 
 affords a sure criterion for distinguishing truth from illusion 
 in experience.' 
 
 1 Idealism proper,' he adds in a note, ' cannot but have a 
 visionary object, whereas mine is merely intended to explain 
 the possibility of our knowing objects of experience a priori, 
 — a problem never yet proposed, far less solved. By it we 
 
206 KANT'S IDEALISM 
 
 remove all this visionary idealism, which, as we can see 
 even from Plato, inferred from our a priori cognitions, even 
 of Geometry, some other intellectual intuition than that of 
 the senses, because it was never suspected that the senses 
 could intuite a priori. 
 
 ' My so-called (critical) idealism is accordingly quite 
 peculiar, both in upsetting the ordinary idealism, and by 
 giving to all a priori cognitions, even those of Geometry, 
 their objective reality, which, without my proving the ideality 
 of space and time, could not be asserted by the most 
 zealous realist. I wish I could change the title of my doc- 
 trine ; but as this seems hardly possible, I may be allowed 
 in future, in order to avoid misrepresentation, to call it 
 formal, or better, critical idealism, as distinguished from 
 the dogmatical of Berkeley and the sceptical of Des- 
 cartes.' x 
 
 § 3. We do not think any comment can make either 
 this statement or the other passage in the Prolegomena 
 clearer, nor is it needful to do much more than call the 
 reader's attention to the excellent resume in the Antinomy 
 of the Pure Reason {Kritik, pp. 307, sag.), where he repeats 
 that this idealism of his, which he calls indifferently formal, 
 critical, and transce?idental, is not only totally distinct from 
 material, problematical, and sceptical idealism, as he calls 
 that of Descartes, but even subversive of it. He shows 
 that the objects given in external intuition are indeed only 
 phenomena, but as such fully as real and certain as those of 
 internal intuition, and that these latter in no sense give us 
 anything more than phenomena in the sensibility of a self 
 unknown apart from them. Internal phenomena, because 
 in time, are as far from being things per se, as external 
 phenomena are, because they are in space. Actuality, apart 
 1 Cf. the analogous note in the Kritik, p. 307. 
 
THINGS PEP SE 207 
 
 from our present perceptions, means that in the progress of 
 experiences we should come upon the object, or it means 
 nothing at all. 1 
 
 Why, then, should Kant ever mention things per se ? — 
 why should we speak as if such things were in any sense 
 existent? Because, though to us such things per se, as 
 opposed to things of experience, must for ever remain com- 
 pletely unknown, there is yet a feature in our experience 
 which necessarily suggests them. Our sensibility is a re- 
 ceptivity, as opposed to ' the spontaneity of thinking. We 
 can only conceive this receptivity as being affected by some 
 foreign cause, and therefore 'we may call this intelligible 
 cause of phenomena generally the transcendental object (or 
 thing per se), in order to have something to correspond to 
 our receptivity.' 2 But, of course, this is no object in the 
 only positive sense in which we can use the term. It is 
 not in space, not in time, not under a category — not in fact 
 definitely conceivable. It may be the same for all pheno- 
 mena, or different ; it may be of the same kind with the 
 substratum of our thoughts, called the Ego, or it may not. 
 All this we can never know ; in fact, we are so completely 
 ignorant of its conditions that we cannot possibly deny its 
 existence, any more than affirm it. So, then, while as to 
 the possibility of our knowing things per se, Kant is a strong 
 and thorough (critical) idealist, as to the possibility of the 
 existence of things per se, he is a problematical Realist, 
 especially guarding himself against the folly of denying 
 
 1 Here is a strictly Berkeleian statement. 
 
 " ' Consequently, I grant by all means that there are bodies without 
 us, that is things, which, though quite unknown to us as to what they 
 are in themselves, we yet know by the representations which their in- 
 fluence on our sensibility procures us, and which we call bodies, a term 
 signifying the appearance of the thing which is unknown to us, but not 
 therefore less real.' — Proleg. (vol. ii. p. 54.) 
 
208 THE PARALOGISM OF IDEALITY f -»AP. 
 
 it — a thing, he says (Proleg. p. 61), 'which it never came 
 into my head even to doubt.' 
 
 This subject naturally came before him in the Paralo- 
 gisms of Rational Psychology, of which, perhaps, the most 
 salient, and even now widely extended, is that of the cer- 
 tainty of internal experience, and the doubtfulness of the 
 information given us by external sense (ideality of external 
 relations, ii. pp. 242 sqq.). ' Cogilo, ergo sum,' said Des- 
 cartes, ' is alone certain ; all else is doubtful, being a mere 
 inference from affections of self, of which the causes may be 
 various, and are therefore necessarily doubtful.' This is 
 problematical or sceptical idealism, which Kant calls em- 
 pirical idealism, because it calls in question the reality of 
 our experience, which he strongly asserts. These empirical 
 idealists, however, strangely enough, when they go beyond 
 experience, are found to be transcendental Realists, as they 
 hold that the external objects of our senses, though not 
 accurately known by us, are, nevertheless, really existent in 
 space and time, apart from our perceptions. Kant, on the 
 other hand, owing to his doctrine of the necessity, of the 
 subjectivity (and therefore of the objectivity), and of the 
 mutual independence, of space and time, is an empirical 
 Realist, asserting our experience to be as real as anything 
 can possibly be to us ; he is an Empirical Dualist, asserting 
 phenomena in space to be essentially different from those 
 in time, both immediately perceived, and both equally real ; 
 he is also a transcendental Idealist, asserting that (whatever 
 may be the case concerning unknown things per se) the 
 things of experience can have no existence beyond it, and 
 are in no sense transcendentally real. And yet it is in this 
 sense that ordinary idealists have been asserting the reality 
 of the things which we imperfectly know. He repeats 
 again and again, in this his First Edition, that he does not 
 
xiv SOURCE OF BERKELEY'S IDEALISM 209 
 
 deny an unknown cause witlwut us, in the transcendental 
 sense, as a thing per se, but he insists that this is not the 
 point at issue, which is rather the nature of things empiric- 
 ally without us, which he declares to be immediately per- 
 ceived and really known phenomena, while the sceptical 
 idealists think them mediately perceived and doubtfully known 
 things per se. 
 
 § 4. But hitherto we have confined ourselves to Kant's 
 criticism of Descartes : what does he say about Berkeley ? 
 Apart from the sceptical idealist, says he, there is the 
 dogmatist, who affirms the objects suggested by external 
 sense to be non-existent, and that the reality of our external 
 experience is an illusion. This theory proceeds, says Kant, 1 
 from finding in the notion of matter insoluble contradictions. 
 These again proceed from the false theory of regarding 
 space as an attribute of things, and not a form of sensibility. 
 This kind of idealism will be disposed of, says he in his 
 First Edition (ii. p. 251), by my discussion of the Antinomies 
 of the Reason by and by ; it has already been disposed of, 
 says he in his Second Edition (Kritik, p. 1 66), by my tran- 
 scendental Aesthetic. How are we to justify this view of 
 Berkeley's philosophy, and to explain this refutation of 
 Berkeley's doctrine ? 
 
 After considering with all possible care the works of 
 Berkeley, viewed from the aspect suggested by Kant's criti- 
 cism, we are of opinion that Kant has not clearly stated 
 the source of Berkeley's idealism, and also the exact differ- 
 ence between Berkeley's system and his own. Berkeley is 
 explicit in telling us that although his system does answer 
 many objections and difficulties raised by men of science, 
 by sceptics, and by atheists, nevertheless, his case rests 
 upon the a priori impossibility of conceiving matter or sub- 
 
 1 Vol. ii. p. 251 ; cf. ibid, p. 61, and Kritik, pp. 41 and 167. 
 
 I P 
 
THE CHARGE OF PREACHING ILLUSION chap. 
 
 stance apart from its attributes. It is in the imperceptible 
 substratum (which he identified with the thing per se), and 
 not in the attributes derived from space and time, that he 
 finds his difficulty. He expressly makes extension, in the 
 perception of space, subjective, and within the mind ; he 
 never once, so far as we can find, makes space and time 
 essential attributes of substance, while holding its other 
 qualities to be within us and subjective ; J and lastly, there 
 is no assertion which he has repudiated in all his philo- 
 sophical discussions with more warmth than the imputation 
 that his theory was one of illusion as to external bodies. 
 On this point he speaks in the very same tone in which 
 Kant did (above, page 64, and Proleg. pp. 56-57), in 
 answer to the very same charge. Perhaps the most ornate 
 passage in all his works 2 is devoted to showing what he 
 considers its absurdity. The very fact, then, of Kant 
 making this charge repeatedly against Berkeley, while he 
 was suffering under it himself, seems to prove that he had 
 done what is perpetually being done with philosophers, and 
 with none more than with Kant himself — he had taken some 
 second-hand account of Berkeleianism for Berkeley's own 
 views, and this he applies himself to refute. 3 It is just like the 
 
 1 Nay, his very argument, asserting for primary qualities that sub- 
 jectivity which had already been proved for secondary qualities, is an 
 almost verbal anticipation of one of Kant's arguments. Cf. Berkeley's 
 Principles, §§ 14-15, and Kant's Proleg. p. 55. 
 
 2 Second Dialogue between Hylas and Philonous, near its commence- 
 ment, vol. i. pp. 158-160 (Ed. 1820). 
 
 3 It was quite competent for Kant to argue that although Berkeley 
 repudiated the charge of holding our external perceptions to be illusions, 
 he was nevertheless justly liable to that charge. This is what many of 
 Kant's opponents have said as regards Kant's repudiation. There is no 
 doubt that according to Berkeley our internal intuitions justly and 
 necessarily suggest to us the presence of mental substance, whereas he 
 will not allow a similar inference in the case of external intuitions, which 
 
REAL DIFFERENCES 
 
 celebrated critique which Cousin wrote on Locke, in which 
 he refuted, not that philosopher, but his natural daughter, 
 the French sensual school. The vulgar Berkeleians certainly 
 did and do regard objects in space as an illusion, and when 
 they give reasons are apt to cite those very difficulties about 
 the infinity of finite things, and so forth, which Berkeley 
 has discussed towards the close of his Principles, and which 
 Kant explains in his Antinomies. But the doctrine of 
 Berkeley himself was far nearer to Kant's doctrine than 
 Kant imagined — in fact, it requires a careful weighing of 
 the two systems to discover the deeper differences. There 
 are, however, three capital points of contrast, which it may 
 be well to set forth here in a few words. 
 
 In the first place, Berkeley, while asserting clearly and 
 positively the subjective nature of space and time, while 
 asserting that our perceptions of extension, as well as of 
 succession, could be only in the mind, and nowhere else, 
 yet did not explain or appreciate the universality and 
 necessity of these two forms of our intuition, and hence 
 failed to assign them an origin in any wise different from 
 that of secondary or contingent qualities of bodies. Here 
 Kant's superiority is incontestable, as is fully explained by 
 him in the passage above cited from the Prolegomena. 
 Secondly, Berkeley, in his zealous polemic against the doc- 
 
 nevertheless suggest material substance with equal cogency. So far, 
 then, Berkeley, whether he will concede it or not, is guilty of preaching 
 illusion in two directions — first in overrating the evidence for an im- 
 material Ego given by internal phenomena ; secondly, in underrating 
 the evidence of external phenomena in comparison with it. Kant 
 showed clearly that both stand exactly upon the same footing. But in 
 charging Berkeley with preaching illusion, he was bound to notice 
 Berkeley's repudiation of it, and show that such repudiation, however 
 well meant, was not warranted. This omission, in such a man as 
 Kant, seems to prove his imperfect acquaintance with Berkeley's own 
 exposition. 
 
REAL DIFFERENCES 
 
 trine of an unknown and unperceived substratum, though 
 in much of his argument, and even in his very words, he 
 agrees literally with Kant's disproof of things per se in space 
 and time, yet falls into the error of denying that our sensa- 
 tions may be the effect of a heterogeneous thing per se as 
 their cause. He was probably bound by the old fallacy, 
 which made men believe the cause to be necessarily like its 
 effect, or homogeneous with it. Here, too, Kant is more 
 cautious and more philosophical. He maintains a critical, 
 not a dogmatical attitude. Thirdly, in disproving the sub- 
 stratum of qualities, Berkeley confused and identified the 
 thing per se with the substratum of phenomena, and con- 
 sequently failed to give a satisfactory account of Permanence, 
 as the criterion ordinarily applied to phenomenal substances. 
 He holds in fact that there is no such thing, except so far 
 as the ideas we have perceived continue to exist in other 
 minds. But these ideas, which render our perceptions per- 
 manent, by continuing them in other minds, must be either 
 numerically identical with those we have had or not so. If 
 not so, the permanence is only an illusion produced by the 
 occurrence of ideas exactly similar in other minds. If, on 
 the contrary, it be held that the ideas preserved in other 
 minds are numerically identical with ours, then their per- 
 manence is doubtless secured, but at the expense of their 
 spirituality, and in contradiction to Berkeley's own repeated 
 statement, that their esse is per dpi ; for that my bare per- 
 ception should become another man's bare perception, and 
 still maintain its numerical identity, is evidently inconceiv- 
 able. 
 
 Here the contrast of Kant's views is really striking, and 
 shows the immense advantages of a sounder theory on the 
 origin of space and time. Holding these latter not to be 
 empirically given, as Berkeley did, but necessarily to 
 
xtv PERMANENCE IN SPACE 213 
 
 accompany all our sensations, Kant first showed that we 
 can form no notion of Permanence except in space, thus 
 getting rid of all speculations as to the permanence of un- 
 known things per se. He next showed that Permanence 
 was no illusion, but the necessary condition of change, and 
 therefore an objective and necessary element of our ex- 
 perience. To what do we apply this notion of Permanence, 
 and whence do we obtain it ? From the fact that all our 
 experience is comprised in one time, which time cannot be 
 perceived in itself, but only when occupied by some per- 
 ception. Hence we infer the permanence of the matter of 
 experience, of phenomenal substance, the changing states of 
 which correspond to the various portions of changing time 
 comprised in the one great complex of time. Thus we 
 represent to ourselves the permanent, even though we have 
 no permanent representation ; and as an empirical criterion 
 of this permanence in time, we use impenetrability, or modes 
 of Resistance in space, for reasons already expounded 
 (above, p. 187, n), and thus this feature in our experience, 
 the belief in the permanence of phenomenal matter, and 
 even in the invariability of its quantity at all time, is vindi- 
 cated and explained. 
 
 This remarkable analysis was completely beyond the 
 range of Berkeley's mental vision. But still Kant's assertion 
 that he had overthrown Berkeley is only true as regards this 
 special point. Here Kant is right in saying that Berkeley, 
 considering things per se to be in space, denied them 
 altogether, because of the absurdities which resulted. But 
 he should have added that this remark applies only to 
 Berkeley's argument about the substratum of things in space, 
 and not to their qualities. It is by distinguishing two sub- 
 strata, one phenomenal, and the other a thing per se, that 
 Kant answers this difficulty, positively in his Aesthetic, 
 
2i 4 REFUTATION OF DESCARTES chap. 
 
 negatively in his Antinomies, where he solves the apparent 
 contradiction which results by means of his distinction. 
 He has therefore rectified and modified the idealism of 
 Berkeley, making it critical where it had been dogmatical, 
 but embodying all the truth and the soberness of that cele- 
 brated system. 
 
 § 5. We now return to the controversy between Kant 
 and Descartes. This sceptical idealism, says Kant, which 
 demands that our inferences from the data of sense shall be 
 proved, is a thoroughly philosophical proceeding, and of 
 great value in compelling men to sift the foundations even 
 of their most ordinary experience. But the transcendental 
 Aesthetic has already for us settled the whole question. 
 Both space and time are within us, but are representations 
 distinct in kind, and of equal authority. If internal pheno- 
 mena suggest a mental substance in which they inhere, it 
 must be a phenomenal substance, in time, and subject to 
 the conditions of all phenomena. If external phenomena 
 suggest a material substance in which they inhere, this 
 matter must be in space, and phenomenal substance, and 
 subject likewise to the universal conditions of phenomena. 
 In neither case do these phenomenal substances give us the 
 least clue to the nature of any thing per se, or supra-sensible 
 and hidden cause of phenomena, itself existing out of space 
 and time. Descartes, no doubt, thought that things per 
 se might exist in space ; he merely wanted proof that our 
 intuitions are trustworthy representations of this external 
 reality per se. Kant shows that this is a completely false 
 basis, that the only positive notion we have of reality is 
 phenomenal, and that if hidden things per se are indeed the 
 causes of phenomena, they are such absolutely heterogeneous 
 causes that phenomena can give us no clue whatsoever to 
 their nature. 
 
INTERNAL PRESUPPOSES EXTERNAL 215 
 
 These are the considerations adduced by Kant in his 
 First Edition against sceptical or problematical idealism. He 
 enters upon them when discussing the supposed superior cer- 
 tainty alleged by all rational psychologists (and Descartes 
 among them) in the case of internal experience. They all 
 consider mind as given immediately and certainly, whereas 
 matter is only given mediately, through what they call ideas, 
 and therefore with inferior certainty. Kant shows (if we 
 may be allowed to repeat this important point) that our ex- 
 ternal experience is as immediate, and therefore as certain, 
 as our internal experience. He also shows that neither can 
 give us any information concerning things per se. 
 
 But in his Second Edition he advances to a still bolder 
 position, as regards the dignity of external experience, and 
 asserts not only its equality with internal experience, but in 
 some respects its priority. Hence he transposes his refuta- 
 tion of idealism from the place where he denied the per- 
 manence and reality of self to be given, to the place where 
 he asserts the actuality and reality of our external experience. 
 It had been formerly an appendix to his refutation of the 
 superior certainty claimed for our knowledge of the pure Ego; 
 it now becomes an appendix to his demonstration of the 
 equal certainty of our knowledge of all empirical phenomena. 
 
 The argument can be thus briefly summarised. We can 
 prove against Descartes, that internal experience, which he 
 thought so certain, is only possible by presupposing ex- 
 ternal experience, which he thought uncertain. But this 
 proof will not be intelligible if we have not mastered Kant's 
 discussion on the Principle of Permanence. It was there 
 shown that we cannot possibly perceive change or indeed 
 any determination in time, without having something per- 
 manent wherewith to contrast the changes. Now when I 
 consider my own existence, I cannot but be conscious of it 
 
216 NECESSITY OF SOME PERMANENT 
 
 as determined in time. 1 If so, it must be determined by 
 some permanent, in regard of which alone, the changes of 
 any internal experience can be known as changes. ' This 
 permanent can be no intuition within me, for there are no 
 determinations of my existence within me but representations, 
 and these are the very phenomena which await their deter- 
 mination from some permanent distinct from them.' But 
 what permanent can there be distinct from our representa- 
 tions, and yet possible in intuition ? Of course, the old 
 theory suggested the permanent Ego, in contrast to its own 
 passing manifestations. If we had an intellectual intuition, 
 which could be applied to the pure Ego, it might doubtless 
 serve this important purpose. But as we shall show more 
 fully hereafter, this / am is no intuition, but a logical con- 
 
 1 Kant lays stress on the difference between a mere play of re- 
 presentations and an empirical experience which determines them in 
 time. ' If it be objected to my proof,' says he, in substance (note to 
 second preface, p. xl. ) 'that it is only the representation of external 
 things of which we are immediately conscious, and not [phenomenal] 
 external things themselves, I answer that the very representations which 
 you admit are determined in time, and therefore demand some permanent 
 as the necessary condition of their determination. It is therefore im- 
 possible to escape my conclusion that internal experience presupposes 
 external experience, whether you can understand the thing or not.' 
 To put the matter into a simpler form. Suppose you say to me that 
 after all I cannot have anything before my mind except intuitions of 
 external things, and these are, of course, within me, I answer, with 
 Kant, that it is not true that your external experience is a mere set of 
 mental modifications, for you have omitted this important additional 
 element — that they are fixed or at least determined in time. Your ex- 
 istence can be marked out in years and clays ; your external experience 
 does not come at random, but in fixed order, and in an order which you 
 can only very partially determine for yourself. In many cases these 
 your representations are forced upon you. What does all this mean ? 
 Simply that your fleeting existence is in relation to a permanent external 
 something, which determines it in time, and without which all your re- 
 presentations would form no legitimate experience. 
 
xiv WHAT THIS PERMANENT IS 217 
 
 dition accompanying every act of the understanding, and 
 only called permanent improperly, because it must ac- 
 company each of these acts. We must have the permanent 
 of which we are in search homogeneous with the changes 
 which it determines. If it be not in the same plane of 
 knowledge, so to speak, it is for our purpose useless. Then 
 again, a thing per se would be perfectly useless for the pur- 
 pose, as it could not possiby be represented in space or in 
 time, and therefore could not determine representations 
 taking place in space and time. 1 What, then, can this 
 permanent be ? It can be nothing but that phenomenal 
 substratum, or substance, expounded under the Principle of 
 Permanence, which represents to us the whole of time, and 
 so makes its determination possible. It is the way we have 
 of representing to ourselves the existence of phenomenal 
 things. It is, of course, no separate intuition, as there is 
 none such perpetually accompanying us ; but it is a possible 
 intuition, and being at the same time a necessary Principle 
 underlying experience, we use the empirical criterion of 
 solidity (or impenetrability) when we wish to find it in our 
 ordinary experience. This is what Kant means when he 
 says in his explanatory note that this representing to our- 
 selves something permanent in existence is not necessarily 
 having any permanent representation. These latter are 
 fleeting and various ; the former is an external thing, dis- 
 tinct fro?n all my representations — that is to say, the per- 
 
 1 Yet this is the absurdity with which Kant has been charged. In 
 fact his argument has been absolutely reversed by his critics. E.g. , when 
 Kant says : ' The perception of this permanent is only possible through 
 a thing external to me and not through the mere representation of a 
 thing,' Kuno Fischer {Critique, Eng. Trans, p. 98), tells us that this 
 Ding ausser mir is the thing in itself. When Kant speaks of external 
 objects in space, he does not necessarily speak of objects external to the 
 mind ; this his critics have failed to observe. 
 
2i8 APPEAL TO COMMON SENSE chap. 
 
 manent substratum of phenomena, which occupies the 
 whole of time, which never originates or decays, which can 
 never be increased or diminished in nature. 1 This is the 
 only suitable permanent, which can determine our internal 
 sense in time, and hence our internal experience is only 
 possible through external experience. 
 
 Before proceeding, let us turn all this abstruse argument 
 into the language of common life. How do we practically 
 determine the time and place of our own existence, the 
 duration of our thoughts, the reality of our waking thoughts, 
 as opposed to dreams and visions ? Is it not by showing 
 that they are simultaneous with external permanents, and 
 therefore determined by them ? Does the poet contrast the 
 change and current of his thoughts with his permanent self, 
 or with the permanence of external nature around him ? 
 Did not the most vulgar impersonation of common sense 
 reply to Berkeley, by appealing to the ' empirical criterion 
 of substance ' — to solidity, in proof that our internal ex- 
 perience was necessarily combined with our equally certain 
 external experiences ? Lastly, when a man closes his eyes, 
 and shuts out, as far as he can, external impressions, — when 
 he lets his fancy wander, or his thoughts apply themselves 
 to abstract subjects, — do we not then find that he rapidly 
 loses the sense of determination in time ; that he has no 
 measure within himself, no comparison with any fixed 
 
 1 Cf. R'ritik, p. xl. The reader will not have read this exposition 
 without being struck with the analogy to Mill's permanent possibilities 
 of sensation, which play so important a part in his latest philosophical 
 work. The true basis of this substratum of sensibility was shown long 
 ago in the works of Kant, to which Mill's argument bears in some 
 respects a strong verbal resemblance. See especially his account of our 
 passage from the permanent possibilities (or substratum of sensibility, as 
 Kant would say) to a belief in existence distinct from them (or to things 
 fey se, as Kant would say), Exam, of Hamilton, p. 230 (3d Edition). 
 
xiv EXPERIENCE AND FANCY 219 
 
 standard, no Atlas upon whose fixed support his thoughts 
 can perform their revolutions ? Is not this again an empir- 
 ical example of Kant's truth : that internal experience — that 
 is, the regulated, legitimate connection of internal phenomena 
 — is only possible by presupposing external experience ? 
 
 Of course, it is not asserted that all so-called external 
 experience has this dignity. There is such a thing as fancy, 
 and how its objects are to be distinguished from those of 
 the external sense, is to be determined by the rules which 
 distinguish experience generally (even internal experience) 
 from fancy. 1 But the very distinction would be impossible, 
 if there were not such a thing as real external experience, 
 nor could we ever have fancied the reality did it not exist. 
 Our fancies are imitations of the receptivity of sense by the 
 spontaneity of the imagination. 2 The position, therefore of 
 the idealists is completely overthrown. If either of the two 
 departments of our experience is more immediate or more 
 important than the other, it is not internal but external ex- 
 perience. They are, however, as we have amply seen, both 
 immediate, and both equally real and certain. There is 
 nothing either inconsistent, or illogical, or unreasonable in 
 this argument. It is the language of common sense trans- 
 lated into Kantian dress, and its basis in the understanding 
 explained. Neither can there be the smallest doubt that this 
 is the true meaning and purport of the passage in Kant's 
 system. 
 
 § 6. The reader who has followed the argument of this 
 chapter will now be in a position to estimate the comments 
 commonly made upon Kant's refulation of Idealism. It has 
 
 1 Cf. vol. ii. p. 250, for this argument in the First Edition. 
 
 2 The question how a thinking' subject can have external intuitions, 
 he lays aside in both editions as insoluble. Cf. Second Preface, p. xli., 
 note, sub An., and vol. ii. p. 263. 
 
KANT'S CRITICS 
 
 been persistently assumed, in spite of his explicit denial, 
 that it was directed against Berkeley. 1 It has been per- 
 sistently assumed, in spite of his explicit statement that it 
 affected external experience, that it was intended to prove 
 the existence of things per se in space. Even the careful 
 and laborious Ueberweg, who could not avoid seeing in 
 Kant's text that he was directing himself against Descartes, 
 invents a parallel reference to Berkeley, and labours to 
 show, what no one disputed, that Kant's refutation of prob- 
 lematical idealism does not refute dogmatical idealism. 2 
 It must be allowed that Kant's want of direct reference to 
 his first Analogy of Experience throughout the argument is 
 likely to mislead the unwary reader, and when the philo- 
 sophers came upon such expressions as, 'an external thing 
 apart from our representation,' and, ' a representation of the 
 permanent distinct from a permanent representation,' they, 
 so to speak, met an ugly fence in a rough hunting country. 
 Yet the discussion of the substratum in phenomena had 
 occurred but lately, its empirical criterion had been discussed, 
 and so they might fairly have been expected to recollect 
 what they had read only a few pages before. 
 
 But the substratum of phenomena had been tacitly 
 identified with the thing per se (which Kant had expelled 
 from experience), even though a permanence of substance 
 
 1 Fischer, Commentary, p. 132. 
 Cf. his Gnmdriss der Gesch. der Philosophic, iii. p. 195, note. 
 Ueberweg is far the most accurate and trustworthy of the commentators 
 we have met. If he had laid proper stress on Kant's explanation of the 
 Permanence of phenomenal substance, he might have seen the proper 
 bearing of the argument now under discussion. He has especially 
 vindicated Kant against the charge of inconsistency in his various 
 editions, but has, we think, only obscurely apprehended the whole of 
 his case. Mr. Matthew Arnold's estimate of the Germans is as true in 
 philosophy as it is in philology; they are rather erudite and pains- 
 taking than acute or clear in criticism. 
 
KANT'S USE OF THING 
 
 under changing attributes is a distinctly empirical notion, 1 
 and if not embodied in Kant's account of experience, would 
 leave a fatal and disgraceful gap. The critics were never 
 able to grasp, though it was written down plainly before 
 them, that Kant could hold the ' inner possibility of matter ' 
 (vol. ii. p. 244) to be nothing but phenomenon ; that he 
 considered permanence a necessary condition in experience, 
 though it be not itself given directly as an object of experi- 
 ence ; that he used substance, and substratum, in a pheno- 
 menal sense ; 2 lastly, that he properly used the word thing 
 for the permanent substance given in space, as well as for 
 the unknown transcendental cause of phenomena, always 
 appending 'per se ' to the latter. But so accustomed were 
 they to speaking of things per se, that they hastened to set 
 down the ' thing apart from our representations ' as a thing 
 of this peculiar kind. It will hardly be expected of us to 
 prove from the Kritik Kant's perpetual use of the term 
 things for objects of sense and experience. The term 
 occurs a hundred times. 3 Nor do we expect that any one 
 who lays stress on such a point will be likely to appreciate 
 the weightier arguments already adduced. It is worth 
 mentioning, lest it should be a stumbling-block to younger 
 students, and for that reason alone. 
 
 1 Cf. vol. ii. pp. 256, 257, a most important passage. 
 
 2 Viz. substantia phenomenon repeatedly through his work. 
 
 3 Cf. in the Kritik, §§ 22 and 23 of the Dedtution (p. 90), also the con- 
 clusion of the Postulates (p. 1 76), the Distinction of Phenomena and Nou- 
 mena, p. 182; the appendix on concepts of Reflection, p. 191 ; the 6th sect, 
 of the Antinomy, p. 308 ; the Prolegomena (vol. ii. ), pp. 54 sqq. , etc. etc. 
 
 We should perhaps not have concluded this chapter without noticing 
 that Kant himself, strongly and explicitly, in his second preface, asserts 
 the identity of the two editions in substance, and invites comparison of 
 them, as expanding different parts of the same arguments, without 
 differing either in propositions or in proofs (cf. Kritik, pp. xxxix. -xli. ), 
 save only in the method of proof there added to the former edition. 
 
CHAPTER XV 
 
 THIRD CHAPTER OF THE ANALYTIC OF PRINCIPLES. 
 
 § i. On the Distinction of all Objects generally into 
 Phenomena and Noumena. 
 
 ' We have not only now travelled through the land of 
 pure understanding, and carefully reviewed each part of it, 
 but even surveyed it, and assigned to each thing its place 
 in it. But the land is an island, and by nature itself shut 
 in with unchangeable bounds. It is the land of truth (a 
 charming name), surrounded by a wide and stormy ocean, 
 the proper home of illusion, where many a bank of fog, and 
 many a fast-melting iceberg lies to us of new lands, and by 
 ever deceiving with false hopes the exploring mariner, in- 
 volves him in adventures, which he can never either abandon 
 or bring to a successful close. Before risking ourselves 
 upon this ocean, in order to make certain whether it can 
 afford us any hopes, it may be desirable to cast one more 
 glance upon the chart of the country we are about to leave,' 
 and ask (a) whether we cannot rest content with this land 
 in default of other domains ; (b) on what title we hold this 
 country against all opposing claims ? A summary may be 
 of use, even though a full answer to these inquiries has 
 already been given in the course of the Analytic. 
 
 After so many repetitions, however, we shall spare the 
 
chap, xv LIMITS OF COMMON SENSE 223 
 
 reader the additional recapitulation, and proceed to the 
 passage in which he speaks of the Definitions of the Cate- 
 gories. The results indeed of the Analytic have hitherto 
 been so sober and shabby, in confining all our knowledge 
 to experience, that many might be disposed to question 
 the utility of so long and dry a discussion. It may be 
 urged in reply, says Kant, that no presumption is more 
 offensive than that of demanding beforehand the results of 
 an inquiry, when they would not even be intelligible, if at 
 once stated. 
 
 § 2. But yet there is an aspect which can be made 
 clear to the most dull and discontented learner. It is this, 
 that however well plain common sense may get through the 
 world, without analysing the sources of its knowledge, upon 
 one point it must be in the dark, and that is on the limits 
 of its use, and on the spheres which lie within and without 
 it respectively. This requires Metaphysic for its settlement, 
 and without it the plain man may be wandering into blunder 
 and illusion, which he cannot foresee or avoid. So then 
 our conclusion was really of the last importance, when we 
 showed that the Categories are only of empirical use. The 
 case was similar with the notions of Mathematic, and they 
 will serve us as a good example. These axioms about space 
 and about magnitude are generated a priori in the mind, 
 and yet require to be constructed in intuition, without which 
 they have no sense or meaning. We draw the straight line 
 between two visible points. We measure magnitude by 
 number, and this we learn through sight by our counting 
 boards, or strokes used as units. 
 
 That the Categories and Principles are similarly circum- 
 stanced appears moreover from this, that we cannot give a 
 real definition of any of them — that is to say, make the 
 possibility of its object intelligible, without descending 
 
224 CATEGORIES ONLY OF EMPIRICAL USE CHAP. 
 
 forthwith into the conditions of sensibility, where alone 
 they can have meaning. 1 No one can describe magnitude 
 except by saying that it is the determination of something 
 by thinking how often a unit is repeated in it. But this 
 how often is based upon succession, which implies time and 
 the synthesis of homogeneous units. So Reality, Permanence, 
 and Cause are mere idle words except we take in Time, as 
 has been amply shown in the discussion on the Analogies. 
 So the Contingent can only be grasped by seeing the exist- 
 ence of something follow upon its non-existence, or vice 
 versa, in time. Possibility, Actuality, Necessity, these too 
 can only be explained from the pure understanding by 
 manifest tautologies. The logical possibility of concepts 
 is very different from the transcendental possibility of things. 
 Our Principles are accordingly nothing but Principles of the 
 exposition of phenomena, and we must substitute for the 
 proud title of an Ontology, which asserts a synthetical a 
 priori knowledge of things generally, the more modest one 
 of an Analytic of the pure understanding. 
 
 We may express this truth by saying that the pure Cate- 
 gories, deprived of the formal conditions of sensibility, have 
 merely a transcendental meaning, but not a transcendental 
 use. They are nothing but pure forms, indicating how the 
 understanding must proceed in knowing any objects ; but 
 in no case can these pure forms of themselves give or sug- 
 gest to us any objects. 
 
 1 He develops the difficulty about defining the Categories in some 
 passages, which we have given from his First Edition, in vol. ii., 
 Appendix B, a and b. As an additional difficulty he tells us that the 
 Categories can only be defined by judgments, whereas all such judg- 
 ments imply and presuppose the Categories. It is the old difficulty of 
 endeavouring to explain a necessary element of our knowledge by means 
 of that knowledge which presupposes this element in fusion with other 
 elements. Cf. above p. 96, note. 
 
ORIGIN OF BELIEF IN KOUMENA 225 
 
 § 3. How is it, then, that the belief in nonmena has 
 arisen ? And yet this belief is an old and well-established 
 one in philosophy. We here prefer to follow the exposition 
 of the First Edition, which is translated in the second volume, 
 as it brings out more clearly the contrast between the pheno- 
 menal substratum and the thing per se, than his later ex- 
 position. 
 
 The ordinary objects of experience are called phenomena} 
 If we assume things which are merely objects of the under- 
 standing, and might be given to some other, say intellectual, 
 intuition, these intelligible (as opposed to sensible) things 
 should be called noiimena. At first sight it might be sup- 
 posed that the very limited notion of phenomena established 
 in the Aesthetic indicated plainly the objective reality of 
 noumena, as distinguished, not logically, by greater and less 
 clearness of knowledge, but generically. For if the senses 
 only represent things as they appear, some other sort of 
 knowledge should be at least possible, which should compre- 
 hend them as they are, apart from our sensibility. This 
 opens a new field, a world of thought apart from sense, nay, 
 perhaps, even intuited, as Schelling afterwards maintained, 
 by some intellectual intuition. There is indeed, as has 
 been shown in the Analogy of Permanence, something per- 
 manent conceived, as distinct from our representations, and 
 to which they all refer. This is, however, an x, or unknown 
 quantity, which brings our unity of apperception into 
 connection with the unity of diversity in sensuous intuition. 
 Accordingly, this transcendental object is inseparable from 
 
 1 Kant has hitherto used the word Erscheinungen, appearances, as 
 the German synonym. As there is no danger of error in using pheno- 
 mena in English, and as there is no other naturalised term which fully 
 corresponds to it, we have preferred using it throughout the former part 
 of the work. Appearance seems to imply illusion, which would be quite 
 false, according to Kant's notions. 
 I Q 
 
226 THEIR POSITIVE AND NEGATIVE SENSE chap. 
 
 sensuous data, without which it would never have come 
 into our minds. It is merely the representation of pheno- 
 mena under the notion of object generally — a notion deter- 
 mined, or rendered definite, only by the diversity of the 
 phenomena. 
 
 Why, then, have we not remained content with the sub- 
 stratum of sensibility, but added noumena to phenomena 1 
 Simply because no confusion was simpler, and because men 
 could not rest satisfied with so negative a notion. When 
 sensibility is removed, the understanding is still able to 
 think by means of its forms, and these are naturally identi- 
 fied with the thing that appears, as opposed to the appear- 
 ance. No doubt, the inference from phenomena to noumena 
 is so far valid that it is necessarily suggested by phenomena. 
 But this affords us a negative meaning for noumena — they 
 are not objects of sensuous intuition. If we proceed to 
 describe them positively as objects of non-sensuous intuition, 
 we dogmatise as to whether there be intuition distinct from 
 that with which we are endowed, and, moreover, as none of 
 our Categories apply to such an intuition, even were it not 
 imaginary, we cannot advance one step towards any know- 
 ledge of its objects. 
 
 Kant means by a problematical concept, ' one that con- 
 tains no contradiction ; that is connected with other cogni- 
 tions as a limitation {JBegrenzung) of given concepts, but of 
 which the objective reality cannot be in anywise known.' 
 Such is the notion of a noutnenon. It is in itself not con- 
 tradictory. It is necessary to show the limits of sensuous 
 intuition and its objective validity, and indicates that there 
 is a field to which that intuition does not extend But 
 yet its possibility cannot be made out, and though our 
 understanding extends problematically farther than pheno- 
 mena, there is no intuition or notion of an intuition hv 
 
XV MUNDUS SENSIBILIS AND INTELUG1BILIS 227 
 
 means of which our understanding could be used assertoric- 
 cilly beyond this field. The Noumenon is then a bounding 
 concept (Grenzbegriff), repressing the pretensions of sensi- 
 bility, not invented at random, but necessarily and un- 
 avoidably connected with the limitation of sensibility. 
 
 Concepts may be divided into intellectual and sensuous, 
 but we cannot admit a similar division of objects into pheno- 
 mena and noumena, in a positive sense , we cannot even call 
 the noumenon a peculiar intelligible object of our under- 
 standing ; we should rather say that the understanding to 
 which it belongs is itself a problem, of which we can never 
 conceive the possibility. 1 
 
 Kant objects to a recent use of (inundus) sensibilis and 
 intelligibilis, whereby the sum total of phenomena, if in- 
 tuited, is called the world of sense — if thought, as con- 
 nected by rational laws, the world of reason. 2 The mere 
 observation of the starry heaven would be the former, the 
 system of Copernicus or Newton the latter. This is mere 
 fencing with the real difficulty, which is to determine whether 
 understanding or reason have any uses when their objects 
 are no longer phenomena ; a question quite beyond any 
 astronomical theory, however scientific, and which we have 
 answered in the negative. Such a distinction between 
 things as they appear and as they are, is to be taken 
 merely in an empirical sense, or within experience, and 
 
 1 ' The notion of a noumenon is therefore no concept of an object, 
 but the problem unavoidably connected with the limitation of our sensi- 
 bility, viz. whether there may not exist objects quite independent of its 
 intuition — a question which can only be vaguely answered, by saying 
 that as our sensuous intuition does not apply to all things indiscrimi- 
 nately, there is room for more and for other objects, so that they cannot 
 be absolutely denied, but neither, in the absence of any definite notion, 
 can they be affirmed as objects for our understanding' (Kritik, p. 206). 
 
 2 Verstandeswelt, properly world of understanding ; but this expres- 
 sion is not English. 
 
228 TEST PROPOSED chap, xv 
 
 does not touch the question whether the pure understanding 
 can give us objects. We have shown, however, that sense 
 and understanding can only give us objects in combination, 
 and that when isolated, their representations are for that 
 purpose idle. 
 
 If any reader still feels any difficulty on the subject, 
 Kant proposes to him his usual test ; let him try to employ 
 the pure Categories by themselves in framing a synthetical 
 assertion, analytical propositions being of no avail for 
 increasing our knowledge. Take for example : all that is 
 here, exists as substance ; or : all that is contingent exists 
 as the effect of some other thing, which is its cause. 
 Whence can he get, or how can he use these notions, apart 
 from any reference to possible experience ? Where is the 
 medium (above, p. 36) in which they can be combined? 
 If he cannot answer this difficulty, he must confess that the 
 region of noumena is a mere vacuum, of use as a limit, but 
 totally devoid of positive objects, and foreign to human 
 cognition. 
 
CHAPTER XVI 
 
 Appendix on the Ambiguity of the Concepts of Reflection, 
 produced by a confusion of the empirical and transcendental 
 use of the understanding. 
 
 § i. Reflection is not directly concerned with ob- 
 taining notions of objects, but it is the mental attitude of 
 discovering the subjective conditions by which we can 
 attain notions. Of course, the first question which it raises 
 is this : to what faculty is the notion due ? Is it in sense, 
 or in understanding, that notions are to be compared ? 
 Thus many a judgment is due to. habit or inclination, and 
 is assumed, for want of reflection, to originate in the under- 
 standing. Judgments which are immediately certain, like 
 those of Mathematic, require no investigation of their truth ; 
 but even these, and a fortiori all others, require reflection, 
 or a consideration of the faculty from which they arise. 
 The act of making this comparison, and of distinguishing 
 whether they arise from sensibility or understanding, Kant 
 calls transcendental Reflection. The relations in which 
 notions stand to one another in our mental states are those 
 of identity and diversity, of agreement and opposition, of 
 internal and external, of determinable and determining 
 (matter and form). It makes a great difference in what 
 faculty these comparisons are carried out. 
 
2 3 o LOGICAL AND TRANSCENDENTAL REFLECTION CHAr. 
 
 In seeking objective judgments we compare concepts, 
 and thus we obtain universal judgments by means of their 
 identity, we obtain particular by their diversity, we obtain 
 affirmation by their agreement, etc. The concepts adduced 
 might accordingly be called concepts of comparison. But if 
 we look beyond mere logical form, to the content of the 
 concept, we find that their relation to one another is not to 
 be determined without regard to the faculty in which they 
 arise, and thus before we use the mere comparison (or 
 logical reflection) we must apply the transcendental reflection 
 just described, which is the real basis of any objective com- 
 parison of notions. This imperative duty we shall now 
 undertake, and apply it to the operations of the under- 
 standing. 
 
 § 2. i. Identity and Diversity. If an object be 
 presented to us repeatedly [in space] with the same internal 
 features, such as quantity and quality, then as an object of 
 the pure understanding, it is always the same numerically 
 identical thing, and not many ; but as a phenomenon, the 
 accurate comparison of our several notions of it is of no 
 avail to prove its identity, for in spite of the most accurate 
 sameness of internal qualities, the difference of place at 
 the same time suffices to prove numerical diversity. So 
 from Leibnitz's point of view, who regarded phenomena 
 as things per se, but confused by our senses, his Law of 
 indiscernibility (principium identitatis indiscernibilinni) was 
 irrefragable. But we have shown that phenomena are 
 objects of sensibility, and that the understanding is, as 
 regards them, of empirical use only, and thus numerical 
 diversity is given by space, of which the parts, however 
 similar, can never be identical. Leibnitz, accordingly, 
 was in error for want of employing transcendental reflec- 
 tion, and his law is no real law of nature. It is merely 
 
REFUTATION OF LEIBNITZ 231 
 
 an analytical rule, when we compare things by mere 
 concepts. 1 
 
 2. Agreement and Opposition. If reality be considered 
 merely through the understanding (realitas noumenon), no 
 conflict of realities can be conceived, as, for example, that 
 in which they mutually cancel each other, as : 3 — 3 = 0. 
 Among phenomenal realities this is of constant occurrence, 
 such as opposing forces, or joys balancing sorrows. General 
 Mechanic even gives an a priori rule for the empirical con- 
 ditions of this opposition, depending on direction, which is 
 quite foreign to the transcendental notion of reality. 
 Though Leibnitz hardly insisted upon this denial of 
 opposition between substances explicitly as a new principle, 
 his followers expressly adopted it. They hold for example 
 that all evil is nothing but the consequence of the limita- 
 tions of creatures, negation being the only possible opposition 
 to reality. In the same way they were able to unite 
 all reality in one being, without fear of such reciprocal 
 limitations as must take place in phenomena. They 
 opposed to realities nothing but their logical negations. 
 
 3. The Internal and External. In an object of pure 
 understanding that alone is truly internal which has no 
 relation (as to existence) to anything different from it. The 
 inner determinations of a Phenomenal substance in space 
 are on the contrary nothing but relations, and it is itself 
 nothing but a mere complex of relations — such as attraction, 
 repulsion and impenetrability ; we know no other qualities 
 
 1 We have brought together under this and the following heads the 
 two repetitions of the argument in the Kritik, pp. 196 and 202, and 
 incorporated them with Kant's present statement, with the view (ever 
 present to us) of curtailing the prolixity of his exposition. The reader 
 will find that we have omitted no material point in any of these repeated 
 discussions. There is no part of the Kritik so prolix. Kant comes back 
 over and over again to his refutation of Leibnitz with great complacency. 
 
232 MONADOLOGY OF LEIBNITZ chap. 
 
 which make up substance, as we find it in space, and they 
 all relate directly to something apart from them. Such 
 qualities will not serve for a substance as an object of pure 
 understanding. The only internal accidents we can con- 
 ceive are those given us by internal sense — either thinking 
 or something analogous to it. Thus Leibnitz made of all 
 substances, which he regarded as noumena, simple subjects 
 with a faculty of representation, and called them monads. 1 
 Hence, too, he could not conceive the community of these 
 substances as anything but a pre-established harmony, and 
 rejected all physical influence. For as each substance is 
 purely internal, it cannot be related to any other, and there 
 must be some third thing causing them to correspond 
 mutually, not by continual and special influence (the system 
 of Divine assistance), but by their depending on an original 
 cause, from which they obtain their existence and per- 
 manence, and with it their mutual correspondences. 
 
 4. Matter and Form. These are two concepts which 
 are at the basis of all other reflection, so inseparably are 
 they bound up with every use of the understanding. The 
 former means the determinable, the latter its determination, 
 both in transcendental meaning, abstracting from all other 
 questions. The old Logicians called the universal matter, 
 but the specific difference the form. In every judgment 
 the given concepts are the matter, their relation by the 
 copula the form of the judgment, and thus it is all through 
 our thinking. Hence, according to the pure understanding, 
 matter precedes form, and accordingly Leibnitz assumed 
 first monads and an inner faculty of representation in them, 
 and then based upon this their external relations and 
 
 1 His argument was : Every real object has an inner nature, in- 
 dependent of and prior to its relation to other beings ; only mind has 
 such an inner nature, and so each monad has perception. 
 
LEIBNITZ S FALLACY 233 
 
 community. Space was only possible as the relation of 
 substances, and time as the connection of them as reasons 
 and consequences. This is correct enough if space and 
 time were attributes of things per se. But as sensuous 
 intuitions, which they are, the form of the intuition, or 
 subjective constitution of sensibility, must precede sensation, 
 or matter. And so this doctrine of space and time originated 
 in the same delusion of transcendental reflection. Leibnitz 
 then gave a false account of the peculiarity of these forms 
 when he attributed it to the confusion of our notions about 
 objects, so that we translated the mere dynamical relations 
 of things into independent intuitions. Nay, he even wished 
 to make his view, which could only suit noumena, valid for 
 phenomena, which were to him cognitions differing from 
 those of the understanding only bya lesser degree of clearness. 
 As Kant afterwards observes (p. 202), Leibnitz's whole 
 system is based upon a manifest logical fallacy. 1 For 
 though by the dictum de omni, etc., everything that belongs 
 to or contradicts any general concept, also belongs to or 
 contradicts all the particulars contained under the concept, 
 yet it is absurd to modify this principle so as to say that 
 whatever is not contained in the general concept is not con- 
 tained in the individuals or particulars under it, for in this 
 consists the difference of the particular and the universal, 
 that the former contains additional comprehension. Yet this 
 is the fallacy on which Leibnitz built his system. Because 
 two notions of things in general have no differences, therefore 
 the things themselves are really identical ; that is to say, the 
 absence of distinctions, in the absence of intuition, is supposed 
 to hold good when things come to be intuited. The same 
 criticism is applied to all the other heads by Kant. 
 
 1 I.e. inferring from the negation of the antecedent in hypothetical 
 reasoning the negation of the consequent. 
 
234 LEIBNITZ AND LOCKE chap. 
 
 § 3. Remarks on the Ambiguity of the Concepts of Re- 
 flection. When we have assigned to a notion its position in 
 the sensibility or pure understanding, this may be called its 
 transcendental place, and when this is generally done we have a 
 Transcendental Topic, like the Logical Topic of Aristotle, used 
 by teachers and orators, which indicated under certain heads 
 where they could find proper materials for the subject in 
 hand. 
 
 The transcende?ital topic, however, contains merely the 
 above-mentioned four heads of comparison and distinction. 
 They differ from the Categories, in that they do not expound 
 objects according to their several concepts, but rather 
 expound, in all its variety, the comparison of representations, 
 which precedes our notion of things. This comparison 
 requires the reflection above described, otherwise we fall, as 
 Leibnitz did, into a transcendental nn^MJinly of these 
 concepts. Our principle has enabled us to analyse his 
 famous intellectual system of the universe, and show that it 
 depends on a fundamental confusion. 'Thus he in- 
 telleclualised phenomena, as Locke sensualised the Cate- 
 gories according to the system of noogony, if I may so call 
 it, in deriving them all from reflection.' 1 Neither of these 
 great men saw that the two faculties were distinct in kind, 
 but only able to produce knowledge by conjoint action. 
 
 The practical result is as follows : Matter is the substantia 
 phenomenon. I seek its inner qualities in all parts of space, 
 and in all effects which can only be external intuitions. 
 We have accordingly no absolute, but only a relative, 
 internal, which consists of external relations. Any other 
 internal is a mere hallucination, and no object for our 
 understanding ; the transcendental object, which is at the 
 
 1 This is the old and once received view of Locke's philosophy, 
 which has been exploded, at least in our University, by the teaching of 
 Dr. Webb (fntellectualism of Locke, passim). 
 
RESULTS 235 
 
 basis of the phenomena, and which we call matter, being a 
 mere something, which we could not even comprehend, 
 were it described to us. In this sense, then, the complaint 
 that we cannot discover the internal constitution of things is 
 mere idle talk. Observation and analysis of phenomena 
 penetrate into the secrets of nature, and how far this may 
 be done in the lapse of time is hard to say. But (as Locke 
 said long ago) such investigations will not bring us a whit 
 nearer to the origin of our sensibility, its relation to things, 
 and the transcendental ground of this unity. For even 
 ourselves we know only through internal sense, and as 
 phenomena. Hence our Kritik of the process of reasoning 
 by mere reflection is of great use and importance. For when 
 we apply the notions gained by ordinary logical reflection 
 to objects generally, without determining their sources, there 
 forthwith appear limitations which distort the empirical use 
 of them, and show that the representations of things in 
 general are not merely insufficient, but even self-contradictory, 
 apart from sensuous conditions. 
 
 § 4. Before leaving the analytic, Kant adds a table 
 which he considers desirable for completeness's sake. In 
 the Prolegomena, he mentions (p. in) that he did it to 
 show how universally the Table of the Categories is applic- 
 able, seeing that the most abstract notions of something 
 and nothing range themselves under these heads. 
 
 The highest concept usually set down in transcendental 
 philosophy as a starting point is the division into the 
 Possible and the Impossible. But this, as every division 
 does, implies a concept to be divided, which is here the 
 concept of something in general, taken problematically. To 
 this corresponds the notion of nothing, of which the subdivi- 
 sions here follow. 
 
 (1.) To the concept of all, many, and one (in Quantity) 
 
236 ANALYSIS OF NOTHING CHAP, xvi 
 
 is opposed that of none, which removes them, and is a con- 
 cept without object, such as the noumena above spoken of, 
 which cannot be classed among possibilities, while on the 
 other hand they cannot be affirmed impossible. Such are 
 imaginary forces assumed in nature, not self-contradictory, 
 but not provable. They are to be called entia rationis. 
 
 (2.) Negation or nothing may be the denial of reality 
 (Quality) or of the absence of an object, such as cold, 
 which is the negation of heat. This is a nihil privativum. 
 
 (3.) We may have the mere form of intuition without 
 matter, such as pure space and time, which are mere formal 
 conditions (in Relation) of something. Such a nothing is 
 to be called ens i?naginarium. 
 
 (4.) The object of a self-contradictory concept is of course 
 nothifig, and is (as to Modality) impossible. This is a nihil 
 negativum. 
 
 Thus we have nothing subdivided into (a) empty con- 
 cepts without objects (ens rationis) ; (fi) Empty object of a 
 concept (nihil privativum) ; (y) Empty intuition without 
 object (ens imaginarium) ; (S) Empty object without any 
 concept (nihil negativum). (a) differs from (8) by being not 
 impossible. (f3) and (y) are on the contrary empty data for 
 concepts. Pure form is not an object any more than negation. 
 
 We have now completed the analysis of the positive side 
 of the great Kritik. The reader who has had the patience to 
 consider with care Kant's argument up to this point may 
 expect to find his difficulties almost at an end. The 
 Dialectic is so completely a consequence of the positive 
 principles established in the Analytic, that we shall content 
 ourselves with a much briefer exposition of it, delaying on 
 some details which are perplexing, but referring constantly 
 to the principles already laid down, which we presume to 
 be now familiar to the student. 
 
CHAPTER XVII 
 Transcendental Logic. — Part ii. The Dialectic 
 
 I. — Of Transcendental Illusion. 
 
 We called Dialectic in general a logic of illusion (Sc/iein) in the 
 Introduction to Transcendental Logic (p. 76, supra), and this 
 is the strict sense in which we shall use the word. We do 
 not mean by it a logic of probability 1 ( IVahrscheinlichkeit), for 
 probability is truth, though it is truth cognised on insufficient 
 grounds ; and though the information we gain by probable 
 reasoning is imperfect, it is not necessarily deceitful ; hence 
 then this branch of logic belongs not to Dialectic at all, 
 but to Analytic. Again, illusion is not the same thing as 
 phenomenon ; in order that we may understand this, let us 
 ask ourselves what is, in general, the nature of error? 2 
 The senses by themselves cannot err in a judgment, for 
 they do not judge at all ; likewise the understanding cannot 
 err without extraneous influence being brought to bear 
 on it, because the formal element of all truth consists 
 in accordance with its laws. In fact no power in nature 
 
 1 Aristotle uses the term in this sense. Cf. Topica, i. 1, 3, etc. 
 
 2 Cf. with what follows, Hamilton Lectures, iv. p. 135. 'Error 
 does not lie in the conditions of our higher faculties themselves, and 
 these faculties are not by their own laws determined to false judgments 
 or conclusions. If this were otherwise, knowledge would be impossible, 
 — the root of our nature would be a lie.' 
 
ILLUSION OF DIFFERENT KINDS 
 
 can of itself deviate from its own laws. Hence, since we 
 have no other source of cognition but these two, sensibility 
 and understanding, and since neither by itself can go 
 wrong, it follows that error is caused solely by the unob- 
 served influence of the sensibility on the understanding; so 
 it happens that the subjective T grounds of a judgment blend 
 with and are mistaken for the objective, and thus illusion 
 arises. We may illustrate this mathematically ; an erron- 
 eous judgment is like the resultant of two forces : in order to 
 expose its falsity, we must resolve it into its two components. 
 In pure a priori judgments, as we saw before, 2 this must be 
 done by means of transcendental reflection, by which the right 
 place is assigned to each representation, and the influence of 
 the sensibility upon such representation is made apparent. 
 
 But further, there are different kinds of illusion. There 
 may be empirical illusion, as often occurs in optical appear- 
 ances ; in this case the judgment is misled by the faculty 
 of imagination. Or there may be logical illusion, caused 
 by want of attention to logical rules ; this at once ceases 
 when the fallacy is exposed. But we are not concerned with 
 either of these; we consider here only transcendental illusion, 
 which, in spite of all warnings of criticism, tempts us beyond 
 the empirical use of the categories, and deludes us with the 
 dream of an extension of the sphere of pure understanding. 
 
 Here we must carefully distinguish the words tran- 
 scendent and transcendental.' 1 ' That which precedes experi- 
 ence as its necessary condition is transcendental , a principle 
 which, on the other hand, tends to transgress the bounds 
 of experience, is transcendent? as contrasted with one 
 
 1 Cf. vol. ii. p. n6. s Supra, p. 229. 3 Supra, p. 41. ■ 
 
 1 A transcendent object then is one absolutely beyond experience. 
 
 We shall see in the sequel that what is constitutively transcendent may 
 
 be regulative!? transcendental. 
 
KANTS CRITICS 239 
 
 whose application is entirely confined within the limits 
 of possible experience, which is called immanent. The 
 principles of the pure Understanding are immanent, of 
 empirical use only, as we saw in our Analytic , what we 
 now wish to do is to expose the illusion arising from the 
 transcendent principles of pure Reason, and to show the 
 grounds of this ever-recurring though futile tendency to 
 seek for supra-sensuous knowledge. And one remarkable 
 point about this transcendental illusion is that it does not, 
 like logical illusion, cease after it has been exposed ; l it has 
 its basis in the nature of Reason itself, and so is natural 
 and unavoidable. There exist in Reason certain subjective 
 rules and maxims of its exercise, which have all the appear- 
 ance of objective principles ; we cannot do away with this 
 misleading similarity : all we can do is to point out that it 
 is misleading and illusory. 
 
 11 This statement has been absurdly misinterpreted by 
 many of Kant's critics. They have spoken of Kant setting 
 up the reason against itself, as if he had made it the source 
 of errors which it could not unveil. Because Kant showed 
 that these delusions could be made to contradict each other 
 (and were therefore irreconcilable with reason) the critics 
 who read that they were unavoidable and natural, also 
 inferred that they were perpetual and not to be refuted. 
 We shall see in the sequel that Kant solves them all. But 
 just as in the case of certain optical delusions, when the 
 error, though exposed, still mocks us, so here the anti- 
 nomies or contradictory conclusions of the pure reason still 
 have their force, and are still apt to make us trip, even 
 after a sound criticism has thoroughly exposed them. It is 
 therefore desirable that the reader should hold fast this 
 illustration of Kant's, and not vilify reason because it some- 
 
 1 Cf. k'ritik, p. 264. 
 
240 REASON THE FACULTY OF PRINCIPLES chap. 
 
 times misleads him, any more than the senses, which also 
 mislead. For both afford him remedies for their own 
 imperfections. 
 
 II. — Of pure Reason as the seat of Transcendental Illusion. 
 
 A. Of Reason in general. — There is of Reason, as there 
 is of Understanding, a purely formal or logical use in which 
 no account is taken of the content of Knowledge; but there 
 is also, as we shall see, a real or transcendental use in that 
 it contains the origin of certain concepts and principles 
 {Grundsdtze), not borrowed from Sense or Understanding. 
 We may expect, as in the case of the Understanding, that 
 the logical concept will give us the key to the transcend- 
 ental. 
 
 We before defined the Understanding as the faculty of 
 Rules ; we now distinguish Reason by calling it the faculty 
 of Principles (Principien). But this requires explanation, for 
 the term principle is ambiguous. In logic the term is used 
 to denote any general proposition, inasmuch as it may be 
 made the major : premiss of a syllogism. And cognition 
 from principles may be described as knowing the particular 
 in the general by means of concepts? In this sense every 
 syllogism is a form of deducing some kind of cognition 
 from a principle. But when we look into the matter, we 
 see that if the major of our syllogism be a mere general 
 proposition of Understanding (whether it be a priori or 
 empirical), we cannot with propriety speak of the reasoning 
 as cognition from concepts, for, to establish the major, we 
 
 1 Cf. Bacon, De Aug. Sci. v. 2. , ' In syllogismo enim fit reductio 
 propositionum ad principia per propositiones medias,' etc. 
 
 2 Cf. Kritik, p. 435. 
 
LOGICAL USE OF REASON 241 
 
 must bring in experience or pure intuition, or the conditions 
 of a possible experience. So for definiteness we may speak 
 of such general propositions as comparative principles, prin- 
 ciples properly so called being synthetical cognitions from 
 concepts such as the Understanding cannot supply. We 
 shall thus reserve the name Principle for those propositions 
 which form the beginning of all knowledge, which depend 
 on no others, and on which all others depend. As the. 
 Understanding then is a faculty for producing unity among 
 phenomena according to rules, Reason is the faculty for 
 producing unity among the rules of the Understanding 
 according to Principles. 
 
 B. Of the logical use of Reason. — The distinction between 
 immediate and mediate inference is a familiar one ; it is 
 only with the latter, however, that Reason is, strictly speaking, 
 concerned, Understanding being competent to perform the 
 former ; i.e. the proper sphere of the Reason is syllogism, 
 and it acts in this way. In every syllogism there is a 
 rule or major supplied by the Understanding, then the 
 Judgment subsumes a particular case under the condition 1 
 of this rule, and finally the Reason draws the conclusion. 
 So then different kinds of rules will give rise to different 
 kinds of syllogisms, and hence syllogisms may be divided, 
 just like judgments, into categorical, hypothetical, and 
 disjunctive. The importance of this we shall see here- 
 after, but we make one remark now. It often happens 
 that a conclusion is put forward as a judgment to see 
 whether it does not follow from other judgments by which 
 a perfectly different object has been thought. If I suc- 
 ceed in showing that it does, then I have brought my con- 
 clusion under a rule which is valid for other objects of know- 
 
 1 The ' condition' of the rule is in fact the middle term of the 
 syllogism. 
 
 I R 
 
242 PURE USE OF REASON chap. 
 
 ledge also ; in other words, I have reduced my number 
 of principles; that is, Reason tries to reduce the great 
 variety of the knowledge of the Understanding to the 
 smallest number of principles, and thereby to produce in it 
 the greatest unity. 
 
 C. Of the picre use of .Reason. — We ask now : Is there a 
 pure use of Reason as well as a logical ? Is it an independ- 
 ent source of concepts and judgments ? Does it prescribe 
 a law to objects, or is it merely a subjective law of economy 
 applied to the stores of the Understanding. Does it con- 
 tain synthetical (not merely analytical) a priori principles 
 ( Grundsiitze) and rules ? If so, what are they ? We first 
 repeat what we said as to the relation between Understand- 
 ing and Reason in their logical use. Reason applies itself 
 to concepts and judgments, the Understanding applies to 
 intuitions ; the unity of the former is a collective unity, that 
 of the latter is distributive — the unity of a possible experience. 
 In short the Reason is related to the Understanding very 
 much as the latter is to Sense. 
 
 What does the ordinary logical procedure of Reason 
 imply ? In it Reason seeks to reduce any given judgment 
 to a general rule ; but this rule itself will again be subjected 
 to the same process, and its conditions sought by means of 
 a prosyllogism, so that we may in fact say that Reason seeks 
 to find for any conditioned cognition of the Understanding 
 the unconditioned, whereby complete unity of knowledge 
 may be attained. But a synthetical a priori principle lies 
 at the basis of this, viz. we are assuming that ' if the con- 
 ditioned be given, all its conditions, i.e. the Unconditioned, 
 are also given.' This principle J is a priori, for it is the 
 necessary condition of the ordinary logical process above 
 described; it is also synthetical, for it asserts the uncon- 
 
 1 Cf. p. 51, supra. 
 
xvii HIGHEST PRINCIPLE OF PURE REASON 243 
 
 ditioned to be given with the conditioned 1 — not merely 
 that a condition is given, which would be an analytical 
 proposition, if we merely judge it of the concept condi- 
 tioned, but would be the principle of causality if we judged 
 it of objects of experience. 
 
 Our problem in the Transcendental Dialectic is then to 
 determine whether this principle has objective validity, or 
 whether it is a merely logical precept which bids us ever 
 seek complete unity. AVe propose to show that it has not 
 objective validity — in other words, that there are no syn- 
 thetical a priori propositions of pure reason ; or, as we said 
 before (p. 36, note), that Metapbysic is impossible. The 
 truth is that Reason, as regards experience, can only be 
 used analytically (whereas the Understanding is used 
 synthetically). Any attempt to apply the Reason to objects 
 synthetically must result in illusion and error. The use of 
 the principles {Grundsdtze) of the understanding is immanent 
 only, the principles {Grundsdtze), on the contrary, resulting 
 from this highest principle (Princip) of pure Reason would 
 be transcendent ; but unfortunately the principle itself is not 
 objectively true — it is a mere demand. 2 
 
 1 This is the synthetical a prioi-i principle lying at the basis of the 
 Ideas of Reason, just as the synthetical unity of apperception lies at 
 the basis of the Categories (see p. 121). The Ideas are its various 
 phases. Many critics have confused it with the Category of causality, 
 which merely asserts that every phenomenal object has a condition and 
 is thereby perfectly satisfied without necessarily ascending any higher. 
 Now the special peculiarity of the Reason is, that, guided by this 
 synthetical proposition, it necessarily seeks for higher and higher con- 
 ditions ; so that, as Kant says (Kritik, p. 233), ' all pure concepts in 
 general are concerned with the synthetical unity of representations, but 
 concepts of pure Reason (Ideas) with the unconditioned synthetical unity 
 of all conditions in general. ' 
 
 2 Cf., however, carefully, Kritik, p. 399. 
 
CHAPTER XVIII 
 
 BOOK I 
 
 Of the Concepts of Pure Reason 
 
 The Concepts of Pure Reason are not obtained simply by 
 reflection but by inference, in this differing from Concepts 
 of the Understanding. The objective validity of the latter 
 depends upon this fact, viz. because they constitute the 
 intellectual form of all experience, it is necessary that their 
 application should always admit of being exhibited in ex- 
 perience. But concepts of Reason are not so limited ; their 
 very name seems to refer to a kind of cognition related 
 to empirical cognition as a whole to its part, its object 
 being the whole of possible experience to which no 
 actual experience can be adequate. Concepts of Reason, 
 then, serve for comprehending (pegreifen), 1 just as by means 
 of concepts of understanding we understand (versteken) 
 perceptions. If they contain the unconditioned they refer 
 to something to which all experience may belong, but 
 which can never itself be an object of experience. Now, 
 according as they do or do not possess objective validity 
 we shall call them conceptus ratiocinati (concepts legiti- 
 mately inferred) or conceptus ratlocinatites (sophistical 
 
 1 Cf. Kant's Introd. to Logic, p. 55, Abbott's Trans. 
 
chap, xvin THE PLATONIC IDEAS 245 
 
 concepts). However, to this we shall again return, merely 
 remarking now that as we called the concepts of the pure 
 Understanding, Categories , so we shall call the concepts of 
 pure Reason, Transcendental Ideas. 
 
 § 1. Of Ideas in Genera/. — In order that we may fully 
 appreciate the meaning of this term we must go back to Plato. 
 He meant by an Idea something which cannot be derived 
 from sensibility, and which, moreover, far transcends even 
 the concepts of understanding (with which Aristotle occupied 
 himself 1 ), inasmuch as in experience nothing corresponding 
 to it can be found. Ideas are, according to him, archetypes 
 of things themselves, and not merely keys to possible ex- 
 periences like the Categories. He saw that our faculty of 
 knowledge has a much higher vocation than merely to spell 
 out phenomena according to a synthetical unity, and thus to 
 read them as experience. He noticed that our Reason 
 naturally tries to reach cognitions to which no object of 
 experience can correspond, but which nevertheless have 
 reality, and are not mere cobwebs'of the brain. Such ideas 
 he found especially in the practical sphere, in Teleology as 
 well as in Ethics, and herein he was right. However, we do 
 not follow him in every particular, e.g. in his extension of 
 these Ideas to the speculative sphere, in his mystical deduc- 
 tion of them, and in the exaggerations which led him to 
 hypostatise them ; in all these points we regard him as mis- 
 leading. But as regards the principles of Ethics, of Legisla- 
 tion, and of Religion, his work is specially valuable, as it is only 
 by means of Ideas that these principles can be laid down : 
 to deduce them from experience is illusory. We cannot 
 now enter upon such inquiries; the point here to be remarked 
 is that we should, in the interests of philosophy, strive to pre- 
 serve the term Idea in its original signification. There is 
 1 In his treatise on the Categories. 
 
246 TABLE OF REPRESENTATIONS chap. 
 
 no lack of names adequate to express every kind of repre- 
 sentation without our having to encroach on terms which 
 are proper to others. The following is a graduated list of 
 them. The genus is Representation in general. Under it 
 stands Representation with consciousness ' (peireptw), the 
 forms of which may be thus tabulated. 
 
 Perception 
 
 1 
 
 
 1 . 
 sensation (subjective) 
 
 i 
 
 cognition (objective) 
 1 
 
 L 
 intuition 
 
 1 
 concept 
 
 1 
 
 i 
 
 empirical 
 
 1 
 pure 
 
 i 
 
 1 
 
 of Sensibility 
 
 (concept of a pure sensuous image) 
 
 i . i 
 
 of Understanding of Reason 
 
 (notion) (idea) 
 
 An idea or a concept of Reason is then a concept formed 
 from notions which transcends the possibility of experience ; 
 it is 'a necessary concept of Reason to which no corre- 
 sponding object can be given in the world of sense.' 
 
 § 2. Of Transcendental Ideas. — We saw in our Analytic 
 that the forms of judgments converted into a concept 
 of the synthesis of intuitions, gave us Categories ; now 
 we naturally inquire, Will not the forms of syllogisms 
 
 1 Apparently the other sub-class would be Representation without 
 consciousness, but Kant does not say so. He is always very careful not 
 to commit himself on the subject of the existence of latent mental 
 
 modifications. His definition of notion should be remarked ' a pure 
 
 concept, in so far as it has its origin in the Understanding, not in the 
 pure image of the Sensibility.' It is noteworthy that this is exactly the 
 sense in which our own Berkeley used the term. Cf. Fraser's Selections 
 from Berkeley, pp. 95 and 335. We may compare with the table gene- 
 rally the table of Degrees of Knowledge given, Introd. to Logic, p. 55. 
 
THE SYLLOGISTIC PROCESS 247 
 
 applied in like manner to the synthetical unity of intui- 
 tions give us the pure forms of Reason, i.e. Ideas ? To 
 unfold this, it is necessary to examine the form of a syl- 
 logism somewhat minutely. The function of Reason in 
 its logical use consists in the universality of the cognition 
 according to concepts, and the conclusion is a judgment 
 which is determined a priori in the whole extent of its con- 
 dition. The proposition ' Caius is mortal ' might be taken 
 from experience by means of the understanding only ; but 
 to reduce it to syllogism I have to find a concept (man) 
 which contains the more general condition under which the 
 predicate of this judgment is given ; then subsume the 
 minor, Caius is a man, under this condition taken in its 
 whole extent (all men are mortal), and so arrive at the know- 
 ledge of my object — Caius is mortal. ' Hence in the con- 
 clusion of a syllogism we restrict the predicate to a certain 
 object, after having thought it [the predicate] under a certain 
 condition in the major in its whole extent.' l To speak 
 logically, the major must be universal. Now to this univer- 
 sality (universalitas) corresponds totality (universitas) of 
 conditions 2 in the synthesis of intuitions. So the tran- 
 scendental concept of Reason is the concept of the totality 
 of the conditions of a given conditioned, i.e. the uncon- 
 ditioned. 
 
 And, as there are three Categories of relation, so we 
 
 1 This is a difficult sentence, because the doubt at once arises, ' How 
 can we say that the predicate in an affirmative syllogism of the first 
 figure is universal?' But the point is, it is not mortal that is taken 
 universally in the premiss, but tnortal men — the major term (mortal), 
 under a certain condition, viz. that it is to be regarded for the moment 
 as applicable only to human creatures. In no other way can we explain 
 Kant's language, which at best is misleading. Of course the predicate of 
 the major premiss, though not logically universal, is still used in a far 
 wider application than in the conclusion. - Cf. note, p. 241. 
 
ABSOLUTE POSSIBILITY 
 
 shall find three Ideas ; we must therefore seek for an un- 
 conditioned (i) of the categorical synthesis in a subject, (2) 
 of the hypothetical synthesis of the members of a series, 
 (3) of the disjunctive synthesis of the parts in a system. 
 There are exactly as many kinds of syllogisms, each of 
 which tries to advance through prosyllogisms to the un- 
 conditioned ; the first to the subject which itself is no 
 longer a predicate ; the second to the presupposition which 
 presupposes nothing else ; the third to that aggregate by 
 means of which complete division of any concept can be 
 determined. Hence the pure rational concepts of totality 
 in the synthesis of conditions are necessary; at least as 
 problems, in order to raise the unity of the Understanding 
 if possible to the Unconditioned, and they are founded 
 in the nature of human Reason even though they may 
 have no valid application in concreto. We must here 
 guard against an ambiguity which presents itself. The 
 phrase absolutely possible may signify one of two things — 
 either (1) that which is possible (interne) in itself (the least 
 that could be said of any object), or (2) that which is possible 
 in every relation, which is the most that could be said of an 
 object. Now the latter signification is the one in which we 
 are going to employ the term. Absolute necessity, in our 
 usage, is not synonymous with internal or intrinsic necessity. 
 Thus, though we can say that the opposite of that which is in- 
 trinsically impossible is absolutely necessary (for if a thing is 
 intrinsically impossible it will be impossible in every relation), 
 we cannot conversely say that the opposite of the absolutely 
 necessary is intrinsically impossible. 
 
 With this provision, then, we may say that the tran- 
 scendental concept of Reason has for its object nothing less 
 than absolute totality in the synthesis of conditions — that is 
 the Unconditioned. Reason has for its aim to collect into 
 
THE IDEAS 249 
 
 an absolute whole all acts of the Understanding, to give a 
 rational unity ( Vernunfteinluit) to the synthetical unity ( Vcr- 
 standeseinheii) which is thought in the Category. Hence 
 the objective employment of the Ideas is transcendent, that 
 of the Categories being immanent. We have established 
 now three points about the Ideas: l (1) they are concepts of 
 pure Reason regarding all empirical cognition as determined 
 by means of an absolute totality of conditions ; (2) they are 
 not arbitrarily assumed, but supplied to us by the very 
 nature of Reason, and have therefore a necessary relation 
 to the whole use of the Understanding; (3) they are transcend- 
 ent ; in this differing from the Categories that no 3 object can 
 be presented in experience which shall be adequate to them. 
 When we use the word Idea we say as regards the object 3 
 (as of an object of the pure Understanding) very much, but 
 as regards the subject (i.e. in respect of its actuality under 
 empirical conditions) very little, because an Idea as the 
 concept of a maximum can never be given adequately in 
 the concrete. Now because in the speculative employ- 
 ment of Reason the latter is the whole aim, and as in this 
 case merely to approximate to a concept which can never 
 be attained in practice is the same as if the concept were 
 non-existent, we are wont to say of a concept of this 
 kind (e.g. the absolute totality of phenomena), 'it is only 
 
 1 This may be regarded as a Subjective or Metaphysical Deduction of 
 the Ideas, corresponding to the Metaphysical Exposition of Space and 
 Time (supra, p. 49), and to the Metaphysical Deduction of the Categories 
 (p. 81), answering the question, ' If there be a priori forms of Reason, 
 what are their characteristics ? ' 
 
 2 And for this reason, as Kant remarks in § 3, no objective deduction 
 of the Ideas is possible. 
 
 3 Apparently this contrast between object and subject means nothing 
 more than that though an Idea may be defined and its nature explained, 
 its relation to experience cannot be determinately given inasmuch as it 
 is quite beyond the sphere of experience. 
 
250 THEIR VALUE- 
 
 an Idea,' in a disparaging kind of way. But on the other 
 hand, because in the practical employment of the Under- 
 standing we have only to do with action according to rules, 
 the Idea of practical Reason can always be given actually, 
 though only partially ; it is, in fact, the indispensable con- 
 dition of all practical employment of Reason. And so 
 the practical Idea is always fruitful, and we have no right 
 to say slightingly of Wisdom (e.g.) that 'it is only an Idea,' 
 for because it contains the Idea of the necessary unity of all 
 possible aims, it must determine all practice as an original, 
 and at least a limiting, condition. The Ideas then are 
 valuable in two respects — first, that though not themselves 
 determining any object, they guide the Understanding, and 
 supply it with a canon of procedure ; secondly, that they 
 furnish a connecting link, as it were, between Physics and 
 Ethics : the latter view, however, we here disregard. We 
 are only concerned with the logical form of all knowledge 
 of Reason, and are considering whether or not Reason is a 
 source of concepts enabling us to regard objects in them- 
 selves as determined synthetically a priori in relation to one 
 or other of its functions. We have before gone through the 
 details of the syllogistic process, and there is no need to do 
 so again, but we observe here that the series of syllogisms 
 (ratiocinatio polysyllogistica) to which Reason leads can be 
 continued either on the side of the conditions (per pro- 
 syllogismos) or of the conditioned (per episyllogismos) to an 
 indefinite extent. But the ascending and descending series 
 are very differently related to the faculty of Reason. For 
 if a conclusion is given (as it always is) only as conditioned 
 by its premises, Reason, which seeks complete knowledge, 
 will demand all its conditions — in other words, will require 
 the ascending series to be complete, given in its totality, 
 unconditionally true. But there can be no such necessity 
 
xvm PROGRESS AND REGRESS 251 
 
 for completing the descending series ; Reason is perfectly 
 indifferent as to how far // proceeds. The sole aim of 
 pure Reason is absolute totality of the synthesis on the side 
 of the conditions, not absolute completeness on the part of 
 the conditioned. If we once have a completely and uncon- 
 ditionally given condition, no concept of Reason is required 
 to complete the series, for the Understanding can take 
 by itself every downward step. Thus the transcendental 
 Ideas serve only for ascending in the series of conditions, till 
 they reach the unconditioned — that is, the Principles. With 
 regard to descending to the conditioned, there is no doubt 
 a widely extended logical use which Reason may make of 
 the laws of Understanding, but no transcendental one ; and 
 if we were to form an idea of the absolute totality of such a 
 synthesis (progressus) as, e.g. of the whole series of all future 
 changes in the world, such would be a mere ens rationis, 1 an 
 arbitrary fiction of thought, but not a necessary presupposi- 
 tion of Reason. 
 
 § 3. System of Transcendental Ideas. — We have seen 
 that there are three kinds of dialectical syllogisms, corre- 
 sponding to the three modes of conclusion by which Reason 
 may arrive at cognitions from principles, and that in all these 
 it is the object of Reason to ascend from the conditioned 
 synthesis to which the Understanding is always restricted, to 
 an unconditioned which the Understanding can never reach. 
 Now the most general relations which our representations 2 
 ( Vorstellungen) can have are (1) Relation to the Subject ; (2) 
 Relation to Objects, either as phenomena or as objects of 
 
 1 Cf. supra, p. 236. 
 
 2 Cf. with this the similar statement quoted in the App. to vol. ii. 
 p. 266, where he distinguishes these three syntheses as respectively 
 syntheses of the conditions — a, of a thought in general ; b, of empirical 
 thinking ; r, of pure thinking. 
 
252 THE THREE PROBLEMS CHAP. 
 
 thought in general. So that in fact the conceivable relations 
 of our representations are three in number. First, the Re- 
 lation to the subject ; Second, to the manifold of the object 
 as a phenomenon ; Third, to all things in general. Hence 
 all transcendental Ideas arrange themselves in three classes, 
 the first of which contains the absolute (unconditioned) 
 unity of the thinking subject ; the second the absolute unity of 
 the series of conditions of phenomena ; the third the absolute 
 unity of the conditions of alt objects of thought in general. 
 
 The thinking subject is the object-matter of Psychology, 
 the sum total of all phenomena (the world) the object matter 
 of Cosmology, and the thing which contains the highest 
 condition of the possibility of all which can be thought (the 
 Being of all Beings), Theology. So it is that pure Reason 
 presents us with a transcendental doctrine of the soul 
 (psychologia rationalis), a transcendental science of the 
 world (cosnwlogia rationalis), and a transcendental know- 
 ledge of God (theologia transcendentalis). These Ideas 
 follow the thread of the Categories, as we might expect, for 
 pure Reason never relates directly to objects, but to the 
 concepts of objects framed by the Understanding. 
 
 And it is noteworthy that there is 1 among them a 
 certain connection and unity, and that pure Reason, by 
 means of them, collects all its cognitions into one system. 
 From the knowledge of ourselves to a knowledge of the 
 World, and through it to a knowledge of God is a progress 2 
 
 1 Cf. Caird, Philosophy of Kant, pp. 533 sqq. 
 
 2 Kant added a note in the Second Edition, in which he says that 
 the above analytical order, though essential to be observed in a pre- 
 liminary inquiry, is not the order in which Metaphysic would treat of 
 the Ideas synthetically. The three great objects of Metaphysic are God, 
 Freedom, and Immortality, and it would aim at showing that the first 
 and second concepts, considered together, will inevitably lead to the 
 third as a necessary conclusion. 
 
THEIR CONNECTION 253 
 
 so natural that it looks like the logical march of reason 
 from premises to a conclusion. Whether, however, there 
 does exist here a relationship, such as does between the 
 logical and transcendental procedure of Reason, is a question 
 which can only be answered further on. It is sufficient 
 now for us to have shown the nature, origin, and number of 
 the transcendental concepts of Reason, as distinguished 
 from those which properly belong to the sphere of Under- 
 standing. 
 
CHAPTER XIX 
 
 BOOK II 
 
 Of the Dialectical Syllogisms of Pure Reason 
 
 The transcendental (subjective) reality of these pure con- 
 cepts of Reason rests on the fact that we are led to them 
 by a necessary syllogism. Of these dialectical syllogisms 
 there are three kinds (the categorical, the hypothetical, and 
 the disjunctive), 1 which we shall here distinguish briefly, 
 and afterwards develop in detail in the succeeding chapters. 
 
 Si. 'I conclude from the transcendental concept of 
 the Subject which contains no manifold, the absolute unity 
 of the subject itself, of which I have in this way no 
 concept,' — i.e. though of the subject, apart from its relation 
 to experience, I have no concept. This I call the Tran- 
 scendental Paralogism. 
 
 § 2. ' I conclude, from the fact that I have always a 
 self-contradictory concept of the unconditioned synthetical 
 unity of the series on one side, the correctness of the 
 opposite unity, of which I have nevertheless no concept 
 either,' — i.e. I do not perceive that the argument can be 
 retorted. This I call the Antinomy of pure reason. 
 
 § 3. 'I conclude from the totality of the conditions 
 1 Cf. p. 248. 
 
chap, xix THE ABSOLUTE SUBJECT 255 
 
 under which objects in general, so far as they can be given 
 to me, must be thought, the absolute synthetical unity of 
 all the conditions of the possibility of things in general,' — 
 i.e. whether they are given in sense or not. In fact ' I 
 conclude from things which I do not know according to 
 their mere transcendental concept, a Being of all beings, 
 which I know still less through a transcendent concept, and 
 of whose unconditioned necessity I can form no concept 
 whatever. This I call the Ideal of pure reason.' 
 
 IT It may be well, before beginning the chapter on the 
 Paralogisms, to give an explanation of Kant's statement 
 (p. 248, supra) that the Categorical Syllogism leads us 
 inevitably to the idea of the Absolute Subject. It seems 
 to be this. Reason always seeks completeness of knowledge, 
 and whereas Understanding would be satisfied with and 
 would rest in the mere judgment ' Caius is mortal,' Reason 
 asks, Why ? Primarily because Caius is a man and all men 
 are mortal. But why is he a man ? Because he is a Roman 
 citizen, and so on ad inf., through an indefinite series of pro- 
 syllogisms ; so that complete explanation is impossible, for the 
 furthest step would be ' because he is Caius,' which would 
 convey no information at all. That is, to quote Kant's own 
 language in the Prolegomena (vol. ii. p. 124), 'Pure Reason 
 requires us to seek for every predicate of a thing its proper 
 subject, and for this subject, which is itself necessarily nothing 
 but a predicate, its subject, and so on indefinitely (or as far 
 as we can reach). But hence it follows that we must not 
 hold anything at which we can arrive to be an ultimate sub- 
 ject ; and that substance itself never can be thought by the 
 Understanding, however deep we may penetrate, even if 
 all nature were unveiled to us. For the specific nature of our 
 Understanding consists in thinking everything discursively 
 — that is, representing it by concepts, and so by mere predi- 
 
256 BASIS OF RATIONAL PSYCHOLOGY chap. 
 
 cates, to which therefore the absolute subject must always 
 be wanting.' Nothing short of a knowledge of the absolute 
 subject could give us completeness ; this we cannot have, 
 though we are always striving to attain to it. A logical 
 paralogism consists in falsity of a syllogism in form irrespective 
 of its matter. A transcendental paralogism is one in which 
 we are led into this formal falsity through a transcendental 
 ground, and so it has its foundation in the very nature of 
 human reason, and is the parent of an unavoidable though 
 not insoluble illusion. 
 
 We have seen that the categorical syllogism leads us to 
 the Idea of the Absolute Subject ; that is, we have now 
 come to a concept of great importance, though a perfectly 
 content-less one, as we shall find hereafter — the concept of 
 the Ego. This I think must be regarded as a tran- 
 scendental concept of judgment, though it can have no place 
 in the list of categories, inasmuch as its only use is to 
 indicate that all thought belongs to consciousness. 1 And 
 it is upon this I think that the whole pretended science of 
 Rational Psychology is based ; from this starting point an 
 attempt is made to investigate the nature of the soul, 
 independent of all empirical aid. For if the smallest 
 empirical element of thought, or any particular perception 
 of my internal state, were mixed up among the grounds of 
 cognition of this science, it would no longer be a rational 
 but an empirical doctrine of the soul. It might be objected 
 at the outset that such a science is on the face of it im- 
 possible, for the very proposition from which it starts ' I 
 think' postulates an internal experience, and therefore 
 introduces an empirical element. But this objection is not 
 well founded, for it is not any particular act of experience 
 that is postulated by Rational Psychologists, but merely the 
 
 1 Cf. supra, p. 140. 
 
xix THE TWO EDITIONS 257 
 
 apperception ' I think,' without which all concepts, even 
 transcendental ones, would be impossible. Internal ex- 
 perience in general and its possibility must not be regarded 
 as empirical cognition, but as cognition of the empirical in 
 general ; l and so its discussion comes under the investiga- 
 tion of the possibility of all experience, which is certainly 
 transcendental. We proceed to show in other ways also that 
 the so-called science of Rational Psychology is quite illusory 
 and fallacious. 
 
 11 We have now two separate discussions of the Para 
 logisms before us, that of the First Edition (given very 
 briefly over again in the Second), the synthetical discussion 
 • — and that of the Second Edition, the analytical discussion. 
 It will be convenient to keep them distinct, for even 
 in the short synthetical discussion given in the Second 
 Edition there are some variations from the statements of 
 the First Edition, especially in connection with the fourth 
 Paralogism. 
 
 1 Cf. supra, p. 44. 
 
CHAPTER XX 
 
 SYNTHETICAL TREATMENT OF THE PARALOGISMS 
 
 Let us start, not with the judgment ' I think,' but with 
 the concept ' I, who am a thinking being,' and endeavour, 
 as Rational Psychologists, therefrom to develop the tran- 
 scendental doctrine of the soul. We shall, as always, follow 
 the guidance of the Categories, though we shall reverse 
 the usual order and begin with the Category of Substance 
 (by which a thing is represented in itself). The points 
 to be established are i 1 
 
 i. 
 The Soul is Subject. 
 
 2 - 3- 
 
 As regards its quality, it As regards the different 
 
 is Simple. times in which it exists, 
 
 it is numerically identi- 
 cal — that is, Unity not 
 Plurality. 
 
 4- 
 It is in relation to possible objects in Space. 2 
 
 1 Compare carefully his second and more systematic statement of 
 these propositions, App. to vol. ii. p. 271, which was omitted in the 
 Second Edition. 
 
 2 That is, as Mr. Caird points out, ' It is an actual object in relation 
 to possible objects in Space.' 
 
CHAP. XX THE STARTING POINT 259 
 
 If these propositions were established, all would follow 
 that we want. Props. I. II. III. would give us respectively 
 the Immateriality, the Incorruptibility, and the Personality 
 of the Soul, and all three together would give us Spirituality. 
 Prop. IV., giving us the soul in relation to objects in space, 
 gives us its connection (commeirium) with a body, i.e. 
 Animality ; and this last, limited by Spirituality, gives us 
 Immortality. Before we consider in detail the proofs 
 which have been offered for these propositions, let us 
 again . repeat that we have nothing to start from but this 
 perfectly contentless representation, ' I, who think,' which 
 cannot even properly be called a concept ; it is only a 
 consciousness which accompanies all concepts. By this I, 
 or He, or It which thinks, nothing more is represented than 
 a transcendental subject of thought = .r, which is known 
 only through the thoughts that are its predicates, and of 
 which, apart from them, we cannot form the least concept, 
 so that we are really turning round it in a perpetual circle, 
 having already to use its representation before we can 
 form any judgment about it. But next it might be asked, 
 Why is this common to all thinking beings because (suppose) 
 true of myself? Why must every thinking existence be 
 regarded as a self-conscious being ? The reason of it is that 
 we are constrained to attribute to things a priori all the 
 attributes which constitute the conditions under which 
 alone we can think them ; and the only way in which 
 I can get a representation of a thinking being is through 
 self- consciousness. That is, the argument is only hypo- 
 thetical ; ' If there be an /, who thinks; anywhere, his soul 
 will possess these properties. The / think is here taken in 
 a problematical sense, not in so far as it may contain the 
 perception of an existence, like the Cartesian Cogito ergo 
 sum, but in regard to its mere possibility. Lastly, this 
 
26o CONDITIONS OF KNOWLEDGE chap. 
 
 general remark gives us the key to the whole discussion : 
 cognition implies not only thought, but intuition. It is not 
 merely through my thinking that I cognise an object, but 
 only through my determining a given intuition in relation 
 to the unity of consciousness in which all thinking consists. 
 And so I cognise myself, not through my being conscious 
 of myself as thinking, but only when I am conscious of the 
 intuition of myself as determined in relation to the function 
 of thought. All the modi of self-consciousness in thought 
 are hence not concepts of objects (categories), but simply 
 logical functions which do not present to thought an object 
 to be cognised, and cannot therefore present my Self as an 
 object. Not the consciousness of the determining, but only 
 that of the determinable self, i.e. of my internal intuition, is 
 the object. 
 
 First Paralogism — of Substantiality. 
 
 Proof : 1 That which cannot be thought otherwise than 
 as subject does not exist otherwise than as subject, and is 
 therefore substance. 
 
 A thinking being, considered merely as such, cannot be 
 thought otherwise than as subject. 
 
 Therefore it exists also only as such, i.e. as substance. 
 
 1 We have taken, for this paralogism, the form of proof given 
 in the Second Edition. In the case of Paralogisms II. and III., the 
 proof is given more fully in First Edition. Paralogism IV. is stated 
 differently in the two editions. In the First Edition the proof of the 
 First Paralogism runs thus : That of which the representation is the 
 absolute subject of our judgments, and which consequently cannot be 
 used to determine anything else (as predicate), is substance. 
 
 I, as a thinking being, am the absolute subject of all my possible 
 judgments, and this representation of myself cannot be used as the 
 predicate of anything else. 
 
 Therefore I, as a thinking being (soul), am substance. 
 
AMBIGUITY OF SUBSTANCE 261 
 
 Critique of First Paralogism. 
 
 This argument is a sophisma figunc dictionis, for the 
 words substance and thought are taken in different senses 
 in it all through. 1. Substance in the major premiss 
 means logical subject — subject of a judgment ; but sub- 
 stance in the conclusion means metaphysical subject — 
 subject in a judgment, substratum, self-subsistent being. 
 If it be said that this is not meant, but that in the con- 
 clusion also, substance = logical subject, the whole argument 
 is useless, for we could never thence deduce the. permanence 
 of the soul, which is really the point at issue. In the 
 General Remark on the System of the Principles (above, p. 
 199), we saw that from mere thought without intuition we 
 cannot establish the objective reality of any concept, e.g. from 
 mere thought we cannot say that the concept of a thing 
 which can exist for itself only as subject, but not as pre- 
 dicate, possesses any objective reality ; in fact, we cannot 
 argue from the logical to the metaphysical sense of the 
 word substance. In order to do this we must have at the 
 foundation of our cognition a permanent intuition ; but in 
 internal intuition there is nothing permanent, for the Ego 
 is but the consciousness of my thought. We may indeed 
 perceive that this representation of the Ego is ever recurring 
 in every act of thought (above, p. 216), but not that it is 
 the fixed and permanent intuition in which thoughts (being 
 transient) alternate. 
 
 2. The word thought is also taken ambiguously. For 
 in order that the major may be true, it must be a general 
 term equivalent to ' thought or intuited ' ; but in the minor 
 the word must be used in its restricted sense as opposed to 
 ' intuited,' for by virtue of internal intuition the soul is 
 presented as an object to thought, not as a subject. Or 
 
262 PARALOGISM OF SIMPLICITY CHAr. 
 
 again, in the major the word is used as it applies to objects 
 generally, consequently to objects as given in intuition ; 
 but in the minor only as it relates to self consciousness. 
 ' Here no object is thought, but there is only represented 
 the relation to self, as subject (as the form of thought).' 
 
 So that we may say that the first syllogism of tran- 
 scendental psychology only palms off upon us a pretended 
 discovery, by setting up the continual logical subject of 
 thinking, as the cognition of the real subject of inherence. 
 That the Ego, the determining self, is the logical subject of 
 thought is an analytical proposition ; but to say that it is 
 substance, and therefore permanent, is a synthetical proposi- 
 tion, and one impossible to prove. ' The statement that 
 " the soul is a substance," then denotes a substance only in 
 Idea but not in Reality.' 
 
 Second Paralogism — of Simplicity. 
 
 Proof : A thing, of which the action cannot be regarded 
 as the concurrence of the action of several things, is simple. 
 Now the soul or thinking Ego is such a thing. 
 Therefore the soul is simple. 1 
 
 Critique of Second Paralogism. 
 
 This is the ' Achilles ' 2 of all the dialectical syllogisms 
 of pure psychology : its argument may be put into a popular 
 form as follows : 
 
 ' Every composite substance is an aggregate of many ; 
 and the action of any composite, or that which inheres in it 
 
 1 Cf. a similar argument in Butler's Analogy, pt. I. ch. i. Compare 
 also Mendelssohn's Phado, where this argument is to be found in some 
 detail. 
 
 - That is— the strong, though not invulnerable argument. 
 
ITS ARGUMENT 263 
 
 as such, is the aggregate of many actions or accidents, 
 divided among a number of substances. Now, an effect 
 which arises from the concurrence of several acting sub- 
 stances is possible when this effect is merely external (as, for 
 instance, the motion of a body is the joint motion of all its 
 parts). But the case is different with thoughts, which are 
 accidents belonging internally to a thinking being. For 
 supposing that this composite did think, each part of it 
 would contain part of the thought ; but all of them only 
 when combined, the whole thought. Now this is con- 
 tradictory. For since the representations which are con- 
 tained under the different parts (suppose the individual 
 words of a verse) are never [by themselves] a whole thought 
 (a verse), so thought cannot be inherent in a composite as 
 such. Thought, therefore, is only possible in a substance 
 which is not an aggregate of many substances, but absolutely 
 simple,' for different representations in different subjects 
 would as little produce a single thought as a number of 
 single words at random would a line of poetry. 
 
 The so-called nervus proband 7 of this argument lies 1 in 
 the proposition : that many representations must be con- 
 tained in the absolute unity of the thinking subject to make 
 up one thought. Now this is a synthetical a priori 
 proposition; how is it to be proved? As synthetical it 
 cannot be proved from mere concepts ; as a priori it cannot 
 be deduced from experience, for experience cannot give 
 necessity, and besides the concept of absolute unity is 
 beyond its sphere. The fact is that here, as in the previous 
 paralogism, the formal proposition of apperception, / think, 
 is the whole basis on which rational psychology ventures to 
 extend her cognitions ; and the mistake is the old one of 
 exalting a merely subjective condition into an objective. That 
 J I.e. in the minor premiss. 
 
264 LOGICAL AND REAL SIMPLICITY 
 
 is, it is true that knowledge is impossible unless I think 
 myself as simple ; but that does not prove that I am 
 absolutely simple ; logical simplicity does not prove real 
 simplicity. The fact of the simplicity of the representation 
 of a subject is not, for that reason, a cognition of the 
 simplicity of the subject itself. So I may be allowed to 
 say ' I am a simple substance ' only in this sense, that the 
 representation of my Ego never contains a synthesis of 
 multiplicity ; but this does not give me the least information 
 with regard to the Ego as an object of experience. 
 Just as the proposition ' the soul is substance,' in the only 
 sense in which it can be proved, is not of the slightest 
 concrete value, and does not enable me to prove the soul's 
 permanence; so the proposition 'the soul is simple,' when 
 understood properly, is of no value in proving that the soul 
 is not corporeal. 
 
 But further, if this cardinal proposition which we are 
 considering (that all which thinks is simple substance) were 
 conceded all objective reality, in the pure meaning of a 
 mere judgment of Reason, we shall now show that we could 
 not even then deduce any sound conclusion respecting the 
 difference between body and soul. 
 
 i. In one sense it is plainly true that the soul is not 
 corporeal, viz. : that as it is represented to us as an object 
 of the internal sense it cannot, so far as it thinks, be an 
 object of the external senses, or a phenomenon in space, i.e. 
 soul as a phenomenon is different from matter as pheno- 
 menon. 2. It is plainly improper to ask whether the 
 soul, as a thing per se, is or is not different from matter as 
 phenomenon ; for a thing per se must be of a different 
 nature from the determinations which merely constitute its 
 states. So, 3, the real comparison is between the soul as 
 a thing per se and the intelligible something at the basis of 
 
XX MATTER AND MIND 265 
 
 external phenomena. Now we know nothing whatever of 
 this latter ; phenomenal matter is composite, but for aught 
 we know this intelligible something may be simple. 1 And 
 therefore, even admitting the simplicity of its nature, the 
 human soul is not at all proved to differ internally from 
 matter, as regards their respective substrata. 
 
 So this concept of simplicity is useless in the only case 
 where it would be of service (i.e. to determine the peculiar 
 and distinguishing feature of our subject, when I compare 
 myself with the objects of external experience) ; it has no 
 application extending to real objects and cannot possibly, 
 therefore, enlarge our knowledge. 2 
 
 Third Paralogis7n — of Personality. 
 
 Proof : That which is conscious of its own numerical 
 identity at different times is, so far, a Person. 
 Now the soul has this consciousness. 
 Therefore, it is a Person. 
 
 1 Cf. his strong statement in the Second Edition, Kritik, p. 252. 
 ' Both kinds of objects differ from each other, not internally, but only 
 so far as the one appears (erscheinl) externally to the other ; possibly 
 what is at the bottom of phenomenal matter as a thing in itself, may not 
 be so heterogeneous after all as we imagine.' He here approaches 
 Monism very nearly. Cf. also in First Edition, vol. ii. p. 253. 'The 
 transcendental object which lies at the basis of internal intuition, as well 
 as of external phenomena, is neither matter nor a thinking being per se, 
 but a basis of phenomena unknown to us, and these give us the em- 
 pirical concept of the first as of the second.' 
 
 3 Kant remarks in the Second Edition, that it would be very curious 
 if the simplicity of substance which is so difficult to detect in the phe- 
 nomena of external intuition, should be presented immediately in the 
 poorest representation of all, i.e. in thought. The whole discussion 
 practically overthrows Mansel's theory of the Ego, which is very similar 
 to the one under consideration. 
 
266 TIME AND SELF 
 
 Critique of the Third Paralogism. 
 
 The point to be considered is, in what sense is the 
 minor premiss true ? This we proceed to discuss. When 
 I regard my own internal phenomena, I find them to be all 
 subject to the condition of time , but this time, again (and 
 the phenomena in it), I perceive always as in me, as a form 
 of my internal sensibility ; hence in [internal] apperception 
 self is the highest condition to which time is subject. For 
 this reason the identity of self has been regarded as the 
 necessary condition of my existence in time. This is true 
 subjectively, but not so objectively or absolutely ; for, 
 suppose another man perceives me, he perceives me 
 through his external sense, and I am [also] to him in time. 
 But, though he readily admits and believes in my conscious- 
 ness being accompanied with a full consciousness of identity, 
 this identity is not to him the condition of the time in which 
 he places me. He places me in time, instead of placing 
 time in me. And the feeling of identity which he allows in 
 me is to him no proof that my self is objectively permanent; 
 for it is not necessarily implied by the time in which he 
 places me. The identity, then, of the consciousness of 
 myself at different times is only a formal condition of my 
 thoughts and their connection, and does not demonstrate 
 the numerical identity of my subject, in which, notwith- 
 standing the logical identity of the Ego, such a change 
 might have taken place as to preclude its [numerical] 
 identity. We may illustrate this physically. A perfectly 
 elastic ball which strikes full upon a similar one imparts to 
 it all its motion (or all its state). Now conceive a whole 
 series of substances, each of which imparts its state and the 
 consciousness thereof to the next in order. In such a case 
 the last substance would be conscious of all the states of 
 
xx PARALOGISM OF IDEALITY 267 
 
 the previously changed substances as its own, since those 
 states were transferred to it along with the consciousness of 
 them ; nevertheless, it would not have been the very same 
 person in all these states. That ' all fleets ' then is a tenable 
 hypothesis ; at least there is nothing in the mere fact of 
 self consciousness which necessarily contradicts it. 1 To sum 
 up : the proposition of the [logical] identity of the subject 
 amidst all the manifold representations of which I am 
 conscious is an analytical one ; but this does not prove the 
 identity of the Person, ' by which is understood the conscious- 
 ness of the identity of its own substance as a thinking being 
 in all change and variation of circumstances.' To prove 
 this we should require not merely analysis, but synthetical 
 judgments based upon a given intuition. Nevertheless the 
 concept of personality (as well as that of substance and 
 simplicity) may remain, so far as it is transcendental, and 
 means a unity of the subject otherwise unknown to us, 
 but in whose states there is a thorough-going connection 
 through apperception. And so far this concept is both 
 necessary and sufficient for all practical uses : but we can 
 never depend upon it to extend our self cognition through 
 pure Reason. 
 
 Fourth Paralogism — of Ideality (of external relation)} 
 
 Proof : — Whatsoever can only be inferred to exist, as the 
 cause of given perceptions, has but a doubtful existence. 
 
 Now all external phenomena are of such a kind that 
 their present existence cannot be perceived immediately, 
 but we infer them to exist as the causes of given per- 
 ceptions. 
 
 Consequently the existence of all the objects of the 
 
 1 First Edition. 
 
268 THE CARTESIAN IDEALISM CHAP. 
 
 external senses is doubtful. This uncertainty I call the 
 ideality of external phenomena ; and the doctrine which 
 holds this ideality is Idealism, in contrast to which the 
 assertion of a possible certainty of objects of the external 
 senses is called Dualism} 
 
 Critique of the Fourth Paralogism. 
 
 U The reader who has worked through chapter xiv. will 
 have no difficulty whatever in understanding Kant's argu- 
 ment here. There are really only two points to be attended 
 to: (i) the difference between the Cartesian (or empirical) 
 and the Kantian (or transcendental) idealism, and (2) the 
 definition of actuality given in the second Postulate of em- 
 pirical thought. 
 
 In the first place it is quite natural that a Rational 
 Psychologist or Transcendental Realist should be forced into 
 this empirical idealism; for of course if the objects of our 
 senses are things per se existing quite independently of our 
 sensibility, there can be no guarantee for the actuality of the 
 representations which our senses give of those things. On 
 the other hand the Transcendental Idealist may quite 
 consistently call himself an Empirical Realist, for ' since he 
 considers matter, and even its internal possibility, 2 to be 
 
 1 The thesis to be proved is : ' The soul is in relation [an actual 
 object] to possible objects in space' i.e. it is not only substance, simple 
 and personal, but it independently exists. We have now got to the 
 Category of Modality, and Rational Psychology desires to show that the 
 existence of the soul is quite independent of the existence of external 
 things. The line of proof adopted here is the attempt to show that the 
 existence of external things is only doubtful. As Kant says in the 
 Second Edition (p. 247), problematical Idealism is perfectly unavoidable 
 in this rationalistic system, ' for if the existence of external things is not 
 requisite to the determination of the existence of a substance in time, 
 the existence of these external things is a gratuitous assumption.' 
 
 - This phrase possibly means the occult forces or elements, which 
 
xx TWO SENSES OF EXTERNALITY 269 
 
 nothing but phenomenon, which apart from our sensibility 
 is nothing at all, he only considers it as a kind of repre- 
 sentations (intuition), which are called external, not as if 
 they referred to objects external in themselves, but because 
 they refer perception to space, in which all things are 
 reciprocally external, while space itself is within us.' 
 
 The whole force of this Paralogism rests on the false 
 assumption that the objects of sensibility are things per se. 
 According to the true doctrine laid down throughout the 
 Kritik, external things exist just as much as I myself do ; 
 with regard to their actuality, I have just as little need of 
 inference as with regard to the actuality of my thoughts ; 
 for they are both nothing but representations, the immediate 
 perception of which is a sufficient proof of their actuality. 
 This is quite in accordance with our former definition of 
 actuality 1 given in the second Postulate of Empirical 
 Thought, 'that which is connected with a perception 
 according to empirical laws is actual' (above, p. 193). 
 
 So then internal experience is not a whit more immediate 
 and certain than external, and therefore we cannot base 
 on the supposed doubtfulness of external experience any 
 argument which would establish the independent existence 
 of the soul. 
 
 Discussion in Second Edition. 
 
 II So far the First Edition. The proof of the paralogism 
 hinted at in the Second Edition {Kritik, p. 242) is some- 
 may be discovered by experiment or observation. Cf. Kritik, pp. 384, 
 414, and 419 for similar phrases. The last of these passages, how- 
 ever, would point to a different explanation. 
 
 1 It may be here observed that the argument has been confused 
 throughout by translators rendering Kant's ' Wirlilichkeit ' by reality. 
 Reality comes under the second class of Categories, not the fourth ; 
 and we are here concerned with the actuality of the Categories of 
 Modality. Cf. p. 94, note, and p. 195, supra. 
 
270 INTERNAL PRESUPPOSES EXTERNAL chap, xx 
 
 what different. The paralogism, as given there, would 
 proceed thus : The soul independently exists, for we are 
 conscious of ourselves as distinct from external phenomena. 
 The critique on this is obvious : That my consciousness 
 of myself is distinct and different from my consciousness of 
 other things is an analytical proposition, but from such 
 mere analysis I cannot at all infer ' whether or not this 
 consciousness of myself is possible without external' things 
 through which representations would be given to me, and 
 whether or not, therefore, I can exist merely as a thinking 
 being (without being man).' And in fact, in the Refu- 
 tation of Idealism (cf. p. 215), we showed that ex- 
 ternal experience was not merely on a par with internal, 
 but that the latter required the former to guarantee its 
 actuality. 
 
 As Mr. Caird points out, Kant goes further here than he 
 did in the Refutation of Idealism, for he here questions 
 the possibility of even a pure consciousness of self, without 
 external perception {Phil, of Kant, p. 547); whereas there 
 he had only maintained that the knowledge of the pheno- 
 menal self through internal sense presupposes the know- 
 ledge of other objects through external sense. 
 
CHAPTER XXI 
 
 analytical treatment of the paralogisms (second 
 edition) 
 
 So far we have been developing the propositions of Rational 
 Psychology in synthetical order, applying successively the 
 categories of relation, quality, quantity, and modality to the 
 concept ' I, who am a thinking being,' and we have seen 
 that in all four cases the results arrived at are fallacious. 
 We proceed to show that the analytical method will lead to 
 equally useless results. 
 
 Let us now start with the. judgment 'I think,' in order 
 that we may discover, by means of it alone, how the Ego 
 determines its existence in time and space ; that is, we 
 lead off with an actuality (modality), and the properties of 
 a thinking being in general are then to be deduced from 
 the mode in which this actuality is thought, after everything 
 empirical has been abstracted. Rational Psychology says : 1 
 
 i. 
 
 / think 
 
 2- 3- 
 
 as subject as simple subject 
 
 4- 
 as identical subject 
 
 in every state of my thought. 
 
 1 It may be noticed that this analytical discussion is especially 
 directed against Descartes's Cogito ergo sum. 
 
272 ANALYTICAL METHOD USELESS chap. 
 
 Critique. 
 
 With reference to the statement 'I think, as subject' (2), 
 we have already shown that the concept of subject can here 
 be taken only in a logical and not in a transcendental 
 sense ; the fact that I must be the logical subject of my 
 thought does not prove that I must be a substance. And 
 as regards statements 3 and 4, how am I to establish them ? 
 How am I to go beyond the unity of apperception, indis- 
 pensable for the possibility of experience, and so pass the 
 bounds of experience ? If I am a materialist I cannot 
 establish (3) the simplicity of the subject, for in space 
 everything is composite. ' Apperception is something real, 
 and its simplicity lies in its very possibility. Now in space 
 there is nothing real which is simple ; for points, which are 
 the only simple things in space, are merely bounds, but not 
 constituent parts of space.' 1 Again the identity or perma- 
 nence of the Ego (4) during every act of thought can no 
 better be proved by analysis of this ' I think ' if I am a 
 spiritualist ; for there is nothing permanent given in internal 
 intuition, as we have before shown, p. 216 (cf. vol. ii. p. 252). 
 In fact we may say generally that the proposition ' I think ' 
 can afford no sure basis to Rational Psychology, and that 
 for two reasons. 
 
 1. The proposition is empirical and contains the 
 determinability of my existence merely in reference to my 
 representations in time. Of course it is not meant that the 
 Ego in the proposition is an empirical representation ; 2 (it 
 is purely intellectual) but merely this, that without some 
 empirical representation which presents to the mind material 
 for thought, the judgment ' I think ' could not be formed. 
 
 1 Cf. Kritik, p. 272, also Hamilton's Reid, pp. 922, 923. 
 2 Cf. Kritik, p. 253. 
 
THE I THINK 273 
 
 And so the proposition is simply equivalent to ' I exist 
 thinking.' Descartes was wrong in inferring the 'I exist' 
 from the ' I think,' for his major premiss must be ' Every 
 thinking being exists,' which would not be true, as it would 
 assert that the property of thought constitutes all beings 
 possessing it necessary beings. The one proposition is not to 
 be regarded as an inference from the other, just because, 
 when rightly understood (i.e. as limited by time), they are 
 identical. 
 
 2. The proposition is synthetical as involving existence. 
 However, the existence in the proposition is not a category, 
 inasmuch as the category never applies to an indeterminately 
 given object, but only to one of which we have a concept, 
 and of which we wish to know, apart from this concept, 
 whether it does or does not exist. But existence is here 
 given quite indeterminately, merely as something real in 
 sensation which awaits determination from the Categories, 
 and therefore I cannot say, ' I exist as substance, accident, 
 etc.,' which would be synthetical additions to my merely 
 existential judgment ; the mode of my existence cannot 
 then be determined by means of this simple self conscious- 
 ness. 
 
 So our conclusion must be that Rational Psychology 
 does not exist as a doctrine furnishing any addition to our 
 knowledge of ourselves ; it is nothing more than a discipline} 
 fixing unpassable bounds to speculative reason in this field 
 of thought, guarding against a soul-less materialism on the 
 one hand, against a vague spiritualism on the other. This 
 refusal of our reason to give satisfactory answers on certain 
 points should teach us to abandon fruitless speculation and 
 turn our self-knowledge to practical uses. 2 The whole 
 fallacious science has its origin in a mere misunderstanding. 
 1 Cf. Kritik, p. 432. a Cf. Locke, Essay, Introduction, § 6. 
 
 I T 
 
274 MENDELSSOHN'S ARGUMENT chap. 
 
 ' The unity of consciousness, which lies at the basis of the 
 categories, is taken for an intuition of the subject as an 
 object ; and the category of substance is applied to it' Or 
 to put the same thing in another way, 'The dialectical 
 illusion in Rational Psychology arises from our confounding 
 an Idea of Reason (that of a pure intelligence) with the 
 altogether undetermined concept of a thinking being in 
 general. I think myself in relation to a possible experience, 
 whilst I abstract from all actual experience ; and infer there- 
 from that I can be conscious of myself apart from experience 
 and its empirical conditions. I consequently confound the 
 possible abstraction from my own empirically determined 
 existence with the supposed consciousness of a possible 
 separate existence of my thinking self; and I believe that I 
 am cognising the substantial in myself as a transcendental 
 subject ; when I have nothing more in thought than the 
 unity of consciousness, which lies at the basis of all deter- 
 mination as the mere form of cognition.' 
 
 Refutation of Mendelssohn 1 s proof of the Permanence 
 of the Soul. 
 
 A common argument for the immortality of the soul 
 is this. The soul is simple, hence it possesses no plur- 
 ality, no parts, is indivisible and therefore cannot perish 
 by discerption. Mendelssohn saw that this argument is 
 incomplete, for it may be said that the soul does not perish 
 by discerption, but by vanishing; and so he offered in 
 his Phcedo the following proof of the impossibility of the 
 latter hypothesis. It is admitted that the soul cannot dis- 
 appear gradually by discerption ; now if it does so suddenly 
 — between the moments of its existence and non-existence 
 there is no time. But between any two moments there is 
 
xxi INTENSIVE AND EXTENSIVE 275 
 
 always time ; otherwise the law of continuity is violated. 
 Hence the soul can no more disappear suddenly than 
 gradually, and therefore is permanent. 
 
 The flaw in the argument consists in the assumption 
 that 'the soul has no parts.' Even supposing it proved 
 that the soul does not possess extensive quantity, it may yet 
 aAraitoi intensive quantity — that is, degrees of reality in regard 
 to all that constitutes its being. Consciousness 1 certainly 
 admits of degrees (above, p. 166), and of an infinite number 
 of degrees down to its total disappearance. Consequently 
 the faculty of being conscious admits of like diminution, 
 and may, if not by discerption, at least by gradual remission 
 of its powers, be finally reduced to nothing. This involves 
 no breach of the law of continuity. So the permanence of 
 the soul beyond life remains undemonstrated and in- 
 demonstrable. 
 
 A further remark may here be made. Some people 2 
 say that we must admit at least the possibility of thought, 
 after this life has ceased, though we have no knowledge of 
 it except when it is limited by its present empirical con- 
 ditions. It is very easy to answer these people by intro- 
 ducing other possible theories which are in no way bolder 
 than theirs. ' Such, for example, is the possibility of the 
 division of a simple substance into several, or conversely the 
 coalition of several into one simple substance. For although 
 divisibility presupposes a composite, it does not necessarily 
 require a composite of substances, but only of degrees (of 
 the manifold faculties) of one and the same substance. 
 
 1 This may be seen at once if we consider that we may have re- 
 presentations present to consciousness, which are not clear — that is, 
 which are not fully distinguished from other representations, and of this 
 obscurity there may be any number of degrees. Cf. Leibnitz on clear- 
 ness and distinctness of concepts. 
 
 2 E.g. Mendelssohn, in his Plnvdo. Discourse I. sub. fin. 
 
276 COUNTER HYPOTHESES chap. 
 
 Now we can think all the powers and faculties of the soul, 
 even that of consciousness, as diminished by one half, the 
 substance still remaining. So also we can represent to 
 ourselves without contradiction this extinguished half as 
 preserved, though not within the soul yet outside it ; and 
 we can hold that, as in this case, everything that is real in 
 the soul and has a degree — consequently its entire existence 
 — has been halved, a particular substance would arise outside 
 the soul' In the same manner several simple substances 
 might coalesce into one, nothing being lost thereby except 
 the plurality of subsistence, inasmuch as the one substance 
 would contain in itself the degree of reality of all the former 
 substances. When once we begin with such vague and 
 unprofitable speculation, there is no end to it. 
 
 Thus in this region of thought, criticism has shown that 
 it is impossible to make any dogmatic assertion concerning 
 an object of experience, beyond the bounds of experience. 
 But it is to be remarked that the necessity of admitting a 
 future life upon the principles of the speculative and practical 
 Reason combined has lost nothing by this criticism ; for 
 the mere speculative proofs which we have been discussing 
 never had any influence on the common reason of men. 
 The really weighty, popular, and practical proof remains 
 untouched, viz. the proof arising out of the consideration 
 that man's faculties, desires, and natural gifts reach far 
 beyond earthly use — they are out of all proportion to our 
 terrestrial needs ; and as we find from the analogy of nature 
 that means and ends are always proportioned, we must 
 only conclude that hereafter there will be a sphere in which 
 these faculties and desires will have full scope. 
 
XXI GENERAL REMARK 277 
 
 General Remark on the Transition from Rational 
 Psychology to Cosmology. 
 
 We stated (p. 272) that 'I think' is an empirical pro- 
 position. Now such a proposition is based on an empirical 
 intuition, and its object is phenomenal ; so that it might 
 seem as if, according to our theory, the soul was changed 
 altogether even in thought into phenomenon, and so our 
 consciousness (being in fact mere illusion) would rest upon 
 nothing. Here then we must make a distinction. Thought, 
 taken by itself, is merely the logical function which binds 
 together the manifold of a possible intuition, and so is 
 pure spontaneity ; it does not represent the subject of con- 
 sciousness as phenomenon, because it does not consider 
 whether the mode of intuiting it be sensible or intellectual. 
 When I represent myself as the subject of thought or as the 
 ground of thought, these modes of representation are not 
 the categories of Substance and Cause. If I wished to 
 know myself, then these categories would have to be applied 
 as usual to my sensuous intuition ; but here I only wish to 
 be conscious of myself as thinking ; in this consciousness I 
 am Being itself ( Wesen selbst), and of this Being nothing 
 further is given in thought. 
 
 But secondly, if the proposition ' I think ' is to be 
 equivalent to ' I exist thinking,' it is no longer the mere 
 representation of a logical function ; it determines the 
 subject in relation to existence, and so the aid of internal 
 sense is required. We have now not only the spontaneity 
 of thought, but we have also the receptivity of intuition. 
 It is plain then that any cognition of myself of this kind 
 must be phenomenal, for the internal empirical intuition 
 is sensuous and presents us only with phenomenal data, 
 which do not furnish anything to the object of the pure 
 
278 PSYCHOLOGY AND PHYSIOLOGY chai\ 
 
 Consciousness for the knowledge of its own separate 
 existence, but can serve the purpose of experience 
 only. 1 
 
 Consideration on the whole of pure Psychology as an 
 appendix to the Paralogisms! 1 
 
 'If we contrast the doctrine of the soul [psychology] as 
 the physiology of the internal sense with the doctrine of 
 bodies, — as a physiology of the objects of the external sense, 
 — we shall find (in addition to the fact that in both we 
 know a great deal empirically) this remarkable difference, 
 that in the latter much can be cognised a priori from 
 the mere concept of an extended incompressible being ; 
 whereas in the former, from the concept of a thinking being, 
 nothing can be cognised synthetically a priori. The cause 
 is this. Although both are phenomena, yet the pheno- 
 menon presented to the external sense has something per- 
 manent or fixed, which gives a substratum lying at the basis 
 of changeable determinations, and so gives us a synthetical 
 concept, namely, that of space and a phenomenon in it. 
 (above, p. 217). Time, on the contrary, which is the only 
 form of our internal intuition, has nothing permanent in it ; 
 so that it only lets us know the change of determinations, not 
 the determinable object.' And so, as we have seen, Rational 
 Psychology gives us no synthetical propositions about the 
 soul. But still, though it does not add to our knowledge 
 of the properties of the soul, it serves this valuable purpose 
 
 1 Briefly : That I think, I am assured of by pure consciousness ; but 
 litno I thin]; or hmo I exist I can only tell by calling in the aid of 
 internal' sense, and so bringing clown the question to the region of 
 phenomena. 
 
 2 We insert this here ; it was left out in the Second Edition of the 
 Kritik. 
 
xxi THREE PROBLEMS 279 
 
 that it secures us from the danger of Materialism. ' This is 
 done by the rational concept of our thinking self, which we 
 have set forth ; for instead of there being any danger that if 
 matter were taken away, all thought — and even the existence 
 of thinking beings — would consequently vanish, it is rather 
 clearly shown that, if I take away the thinking subject, the 
 whole world of matter must vanish, being only what appears 
 in the sensibility of our subject as a species of its re- 
 presentations.' 1 
 
 We are now in a position to consider three additional 
 dialectical questions (cf. p. 259) which form the proper 
 object of Rational Psychology, viz. (1) the possibility of the 
 community of the soul and an organic body, i.e. the 
 animality and condition of the soul in this life ; (2) the 
 commencement of this community, i.e. the state of the soul 
 at and before birth [the question of pre-existence] ; (3) the 
 end of this community, i.e. the state of the soul at and 
 after death (the question of immortality). 
 
 1. As question (1) is commonly understood it comes to 
 this : How can there be a community between two hetero- 
 geneous substances, mind and matter ? And three different 
 hypotheses have been made in order to answer this, viz. 
 the doctrines (a) of Physical Influence, (ji) of Pre-established 
 Harmony, (y) of Supernatural Assistance.' 2 Now if we con- 
 sider the problem critically we shall see that there is a 
 fallacy in its very statement, and hence in all the solutions. 
 For the all-important distinction between phenomena and 
 noumena is overlooked. All three theories make the 
 fundamental presupposition that soul and body are distinct 
 per se ; and, as theories, are only possible under this 
 
 1 As Mr. Caird points out {Philosophy of Kant, p. 549), this sentence 
 contains the real answer to all theories of Materialism. 
 
 2 Respectively due to Descartes, Leibnitz, and Malebranche. 
 
2So SOUL AND BODY CHAP, xxi 
 
 supposition. But how could we know this to be true? 
 We have seen before (p. 265) that at least such an as- 
 sumption is perfectly gratuitous. And so our Kritik leads 
 us to the conclusion that such a statement of the problem 
 involves gross dualism. The real problem is this : ' How 
 can external intuition (viz. that of space) be at all possible 
 in a thinking subject ? ' And this problem is insoluble ; 
 we can never supply this gap in our knowledge — we can 
 only indicate it by ascribing external phenomena to a tran- 
 scendental object as their cause, of which however we have 
 not and cannot have any notion. 
 
 2 and 3. So also is it with the problems of pre- 
 existence and immortality. Stated in their proper form they 
 will be as follows : — Before the commencement of that kind 
 of sensibility by which something appears to us in space, 
 how were those transcendental objects intuited, which in 
 our present condition appear as bodies ? And when that 
 kind of sensibility ceases by which transcendental — and 
 now wholly unknown — objects appear to us, will all intuition 
 of them consequently cease ? We cannot answer such 
 questions ; we cannot reasonably pretend to know on what 
 the actuality of external phenomena depends, even in the 
 present state, and therefore we cannot assert positively 
 whether it will or will not continue after death. 1 So that 
 the whole dispute about the nature of the soul and its 
 connection with the external world merely arises from our 
 supplying the gaps in our knowledge by paralogisms of the 
 Reason, in that we make our thoughts to be things and 
 hypostatise them, whence arises an imaginary science, both 
 as regards its affirmations and its negations. 
 
 1 Cf. the strikingly similar language of Butler, Anal. I. cli. i. 
 
CHAPTER XXII 
 
 THE ANTINOMY OF PURE REASON 
 
 We now proceed to consider the dialectical procedure of 
 pure Reason in reference to hypothetical syllogisms. Here 
 our aim is to reach the unconditioned unity of the objective 
 conditions in the phenomenon ; as in the preceding dis- 
 cussion our argument was concerned (following the analogy 
 of the categorical syllogism) with the unconditioned unity 
 of the subjective conditions of representations generally. 
 
 We shall term all transcendental ideas, in so far as they 
 relate to the synthesis of phenomena, cosmical concepts l 
 ( Weitbegriffe), partly on account of this unconditioned 
 totality, on which the concept of the world-whole rests — 
 a concept which is itself only an Idea — partly because they 
 relate solely to the synthesis of phenomena, which is 
 empirical ; 2 while, on the other hand, the absolute totality 
 in the synthesis of the conditions of all possible things gives 
 rise to an Ideal of pure reason, which is quite different from 
 the cosmical concept, although it stands in relation with it. 8 
 
 1 Cf. Kritik, pp. 263, 508. 
 
 - I.e. the word Welt on the one hand applies to phenomena, and on 
 the other, in the transcendental sense, signifies the absolute totality of 
 the content of existing things (ko<j>«5s). 
 
 3 It is essential to bear this distinction in mind all through. Cf. 
 Kritik, pp. 232-236, etc. 
 
282 COSMOLOGICAL IDEAS 
 
 And just as the paralogisms of pure reason formed the 
 foundation of a dialectical psychology, so it will be seen 
 that the antinomies, which here result, place before us the 
 transcendental principles of a pretended pure Rational 
 Cosmology. 
 
 § i. System of Cosmological Ideas. — In order to 
 enumerate with systematic precision these Ideas according 
 to a principle, we must remark two things. — 
 
 ist, That pure and transcendental concepts arise from 
 the Understanding alone ; the Reason does not give birth 
 to any new concept, but only frees the concept of the 
 Understanding from the unavoidable limitation of a possible 
 experience (above, p. 26), and thus tries to raise it beyond 
 the bounds of the empirical, though still in connection with 
 it, as was fully explained above (p. 242). Hence then we 
 may say that the transcendental ideas are properly nothing 
 but categories raised to the unconditioned, and they may 
 be arranged in a table corresponding to that of the cate- 
 gories. 
 
 2nd, All the categories do not serve for this, 1 but only 
 those in which the synthesis constitutes a series, and a series 
 of subordinated (not of co-ordinated) conditions. And, as 
 we saw before (p. 250), absolute totality is required of 
 Reason only in so far as concerns the ascending series of 
 the conditions of a conditioned, and not when we are 
 descending to the consequences. The descending or pro- 
 
 1 It must be remembered that Kant is here only speaking of Cos- 
 mological Ideas ; the categories in which the synthesis does not furnish 
 us with a series may of course be raised to the Unconditioned, but 
 when they are, they do not give the Cosmological Ideas, but the 
 Pyschological or Theological, leading not to the absolute unity of a 
 series of conditions of phenomena, but to the absolute unity of the 
 thinking subject, or of the conditions of all objects of thought in general. 
 Cf. Kritik, p. 234. 
 
QUANTITY 2S3 
 
 gressive series, proceeding in consequentia, cannot be regarded 
 as given, but only as capable of being given (dabilis). The 
 Cosmological Ideas on the other hand are occupied with 
 the totality of the regressive synthesis, and proceed in ante- 
 cedentia. 
 
 Let us now try to construct our table of cosmological 
 ideas in accordance with that of the Categories ; in other 
 words, let us seek the complete series of the conditions of 
 a given phenomenon. 
 
 1. As to Quantity, the two original quanta of all our 
 intuition are Time and Space. Every object of intuition is 
 composite, i.e. extended in time and space. 1 As regards 
 time, it is in itself a series, and so in this respect the 
 transcendental idea of the (regressive) series of the con- 
 ditions of a given conditioned is merely the whole of 
 past time. But it may be said a similar remark cannot 
 be made of space, for how can the extension of space 
 be regarded as a regressive series of conditions ? Are not 
 the parts of space simultaneous ? To this we reply : 
 Though all the parts of space are co-ordinated, yet they 
 must be apprehended' 2 successively, and are hence (a) 
 a series. As in the measurement of space we take units 
 successively, and as these units are evidently limited and 
 conditioned by the succeeding units, we have (/?) a series of 
 conditions and conditioned. And as no space can be given 
 or produced by another as a result, but is only limited by it 
 as a condition, this limitation is (7) rather a regressive than 
 a progressive series. Hence then I have a right to demand 
 the absolute totality of phenomena in space, just as in time. 
 Whether my demand can be satisfied is a question to be 
 answered in the sequel. So then my first Cosmological 
 
 1 Cf. p. 163, supra. 
 - Cf. Kritik der Urtheilskraft , § 27. 
 
QUALITY AND RELATION 
 
 Idea is the absolute Completeness of the Composition of the 
 given totality of all phenomena. 
 
 2. As to Quality. All matter, as being in space, is 
 divisible or consists of parts. Its reality depends on the 
 reality of each of its parts ; they are its conditions : the parts 
 of parts are its remote conditions. So here is our regressive 
 synthesis, the absolute totality of which is required by 
 Reason. This can only be obtained by a completed divi- 
 sion, and so our Idea is the absolute completeness of the 
 Division of a given totality in a phenomenon. 
 
 3. As to Relation, (a) The category of Substance will 
 not furnish us with a cosmological idea, as we can get no 
 regressive series. For accidents (in so far as they inhere in 
 a substance) are co-ordinated with each other, and do 
 not constitute a series : and in relation to substance they 
 are not properly subordinated to it, but are the modes of 
 existence of the substance itself, (b) The same holds good 
 of substances in community, which are mere aggregates and 
 have no exponent of a series. For they are not sub- 
 ordinated to each other as conditions of its possibility, 
 which may be said of spaces, as above explained. (c) 
 Hence, therefore, it is only in the category of Causality 
 that we can find such a series as we want, a series of causes 
 to a given effect, and our third Cosmological Idea is that 
 of the absolute completeness of the origination of a pheno- 
 menon. 
 
 4. As to Modality, the concepts of the possible, the actual, 
 and the necessary conduct to no series, except so far as the 
 contingent in existence must always be regarded as condi- 
 tioned, and as pointing according to a rule of the Under- 
 standing to a condition under which it is necessary to rise 
 to a higher condition until, in the totality of the series, 
 Reason finds unconditioned necessity. So our fourth Cosmo- 
 
ABSOLUTE AND INFINITE 285 
 
 logical Idea is that of the absolute completeness of the De- 
 pendence of the Existence of the changeable in a phenomenon. 
 
 This Unconditioned may be thought either (a) as exist- 
 ing merely in the entire series, in which all the members 
 are conditioned and only the totality 1 (Das Game) of 
 them quite unconditioned — and in this case the regress 
 is called infinite ; or (b) the absolutely unconditioned is 
 only a part of the series to which the other members arc 
 subordinated, but which itself stands under no other con- 
 dition. 2 In the former case the series is a parte priori 
 without bounds, without beginning, i.e. infinite, and al- 
 though entirely given, yet the regress in it can never be 
 completed, and can only be called potentially infinite. In 
 the latter case there exists a First in the series, which is 
 called, with reference to past time, the beginning of the 
 world; with reference to space, the boundary of the world , 
 with reference to the parts of a given limited whole, the 
 simple ; with reference to causes, absolute spontaneity 
 (freedom) ; and with reference to the existence of change- 
 able things, absolute necessity of nature. 
 
 Now the word nature has been used in two senses, 
 which we shall distinguish. (1) Nature, understood adjective 
 (formaliter), signifies the complex of the determinations of a 
 thing, according to an internal principle of causality, i.e. 
 stands for a Mathematical whole. 3 On (2) the other hand 
 we understand by nature substantive (materialiter), the sum 
 total (Inbegriff) of phenomena in so far as they by virtue of 
 an internal principle of causality are connected with each 
 
 1 Cf. p. 284. 
 
 2 In the one case, in the language of Hamilton, we may say that the 
 Unconditioned is infinite, negating extrinsic dependence ; in the other 
 case it is absohde, positing intrinsic independence. 
 
 3 Cf. above, p. 135, and Prohg. p. 99. 
 
286 WORLD AND NATURE chap. 
 
 other throughout. This is a dynamical whole. If we wish 
 to use language exactly we will use IVorld ( Welt) to signify 
 the mathematical whole of phenomena and the totality of 
 their synthesis, whether by composition or division ; and 
 we will term the world Nature when it is regarded as a 
 dynamical whole. If then we adopt this terminology, we 
 shall restrict the phrase x cosmical concepts ( Weltbegriffe) to 
 the first two Ideas of Rational Cosmology, calling the latter 
 two transcendent concepts of nature (transcendente Natur- 
 begriffe). 
 
 § 2. Antithetic of Pure Reason. — Transcendental Anti- 
 thetic is an investigation into the antinomy of pure Reason, 
 its causes and its result. If we apply our Reason not only 
 to objects of experience according to the laws of under- 
 standing, but venture to extend it beyond these bounds, 
 there arise sophistical (iierniinftelnde) propositions which 
 can neither hope for confirmation nor fear refutation in 
 experience. Each of them is not only in itself self-consistent, 
 but can point to conditions of its necessity in the very 
 nature of Reason, only that unfortunately its opposite can 
 produce equally valid and necessary grounds for its support. 
 The questions which naturally arise in such a Dialectic 
 of pure Reason are therefore — (i) In what propositions is 
 pure Reason unavoidably subject to antinomy? (2) On 
 what causes does this antinomy depend ? (3) Whether 
 and in what way can Reason in spite of this contradiction 
 find a way to certainty ? 
 
 A dialectical theorem (Lehrsatz) is different from a 
 sophistical, firstly, in that it is not an answer to an arbitrary 
 question, but to one naturally and necessarily forced upon us ; 
 and, secondly, in that it carries with it an illusion which can- 
 not be dispelled, 2 even when we are no longer deceived by it. 
 1 See p. 281. 2 Cf p _ 23 g 5 sufra _ 
 
xxii SCEPTICAL METHOD 2S7 
 
 Now when we meet with dialectical propositions of this 
 kind, which seem to contradict each other, and yet which 
 seem to be equally susceptible of rational proof, the method 
 to be adopted for finding the truth is what we call the 
 sceptical method. That is, we set ourselves to discover 
 whether the object of the struggle is not a mere illusion, 
 which each side strives to reach, but which would be of no 
 gain even if it were reached. 1 This is quite a different 
 thing from scepticism, or that artifical and scientific 
 agnosticism which undermines the foundations of all 
 knowledge in order to destroy our belief and confidence in 
 it. For the sceptical method aims at certainty by endeavour- 
 ing to discover in a conflict of this kind, conducted honestly 
 and intelligently on both sides, the point of misunderstand- 
 ing. Such a method may be dispensed with in mathematics 
 or in physics, where we can appeal either to pure intuition 
 or to experiment ; but in transcendental philosophy, where we 
 have no such sure criteria for testing the falsity of an 
 assertion, its use is indispensable. 
 
 *[ We now proceed to develop the conflicts of the 
 transcendental ideas in order ; but before doing so we state 
 Kant's position. You may affirm or deny the complete- 
 ness of composition, division, origination, or independence. 
 Here are ' eight propositions which contradict one another 
 in pairs, but each of which necessarily belongs to Meta- 
 physic, which must either accept it or refute it {although 
 there is not one that has not in its time been held by some 
 philosopher).^ Now you may choose any of the eight you 
 please, and Kant undertakes to prove its opposite. ' That 
 is to say, we cannot have, with regard to the pheno- 
 menal series, the completeness of antecedence which 
 
 1 Cf. Kritik, p. 450, and also cf. supra, p. 18, on the distinction be- 
 tween the dogmatical method and dogmatism. 
 
288 KANT'S POSITION chap, xxii 
 
 reason demands. In brief a cosmical metaphysic is im- 
 possible, and this is the indirect proof of Transcendental 
 Ideality.' (Prof. Maguire's Lectures on Philosophy, p. 160.) 
 The proofs of these different propositions now follow : their 
 critical solution comes afterwards. 
 
CHAPTER XXIII 
 
 THE ANTINOMIES 
 
 First Antinomy (of Quantity) 
 
 Thesis. — The world has (a) a beginning in time, and is (/3) 
 also limited in regard to space. 
 
 Proof. — For (a) if it had no beginning in time, an 
 eternity must have already elapsed. But the infinity of a 
 series consists in the very fact that it can never be com- 
 pleted by means of a successive synthesis ; therefore an 
 infinite series already elapsed is impossible, and the world 
 has a beginning in time. 1 (/3) If it were unlimited in 
 regard to space, then it must be an infinite given total 
 of co-existent things. But an infinite can never be given, 
 for qua infinite it can neither be given in a single intuition, 
 nor can it be given in a series of successive intuitions of 
 
 1 It has been objected by Dr. Erhardt in his Kritik der Kantischen 
 AiUittomienlehre, that this proof is fallacious, for infinite is used in two 
 senses — (a) as that which has no beginning ; (i) as that which can never 
 be completed through a successive synthesis. But it is not easy to see 
 how the synthesis of a series which has no beginning could ever be com- 
 pleted, and probably Kant would say that the transition from meaning 
 (a) to meaning (b) is therefore perfectly legitimate. Somewhat similar 
 objections were raised to the proofs of thesis and antithesis in all the 
 antinomies by Prof. Maguire as far back as 1874. 
 I U 
 
2go NOTION OF INFINITY CHAP. 
 
 its parts, for a complete synthesis of this kind would 
 require an infinite time, already shown to be impossible. 
 
 Kant adds that he might have proved the thesis also 
 by starting with a false notion of the infinity of a given 
 quantity — the mathematical conception — that it is a 
 quantity greater than any other ; but of whatever number 
 of parts we take a quantity to be, we can add to this 
 number. Therefore, etc. But this notion of an infinite 
 quantity assumes a given unit, and the argument based 
 on it would equally well disprove the possibility of a finite 
 body having an infinite number of component parts. The 
 proper transcendental notion of infinity is that adopted in 
 the text ; it does not answer the question, How large ? but 
 the question, How often must I repeat a unit of any she 
 T choose to assume, to obtain it ? Hence in proportion to 
 the unit assumed, infinites may vary greatly in quantity. 
 
 Antithesis. — The world has no beginning and no 
 boundaries in space, but is infinite in respect both (a) to 
 time and (/3) space. 1 
 
 Proof. — (a) If it be not infinite with respect to time, it 
 must have had a beginning ; therefore there must have been 
 a time when the world did not exist, i.e. a void time. But in 
 a void time nothing can originate, for there is no reason why it 
 should originate at one moment rather than at another ; and 
 therefore the world could never have come into existence at 
 all. Hence the world cannot have had a beginning. (/3) If 
 it be bounded in space, it must exist in a void space (which 
 is not bounded). We then have not only a relation of 
 
 1 It may be observed that all the proofs (cf. the subsequent remarks 
 on the fourth antinomy) of thesis and antithesis are indirect. And it 
 must be constantly borne in mind that, on the assumption that pheno- 
 mena are things in themselves, all the proofs are regarded as perfectly 
 valid and conclusive by Kant. The whole force of his subsequent 
 reasoning depends on this. 
 
xxiii MUNDUS PHENOMENON 291 
 
 things in space, but a relation of things to space. But as 
 the world is an absolute whole, out of and beyond which no 
 object of intuition can be discovered, this relation to void- 
 space is virtually relation to no object, is nothing. Hence 
 then the world cannot be regarded as bounded in space. 1 
 
 In fact, if we assume the existence of a bound to 
 the world either in space or in time, we are forced to 
 assume two nonentities {Undinge), void space and void 
 time. On this argument Kant makes the following remark. 
 It may be said by a philosopher of the school of Leibnitz, 
 this assumption is not necessary ; 2 space and time are only 
 the relations of things and not supposable beyond them. 
 Instead of a ' first beginning,' all that you need assume is 
 an existence which postulates no other condition in the world ; 
 instead of ' bounds of extension,' all that you require is ' limits 
 of the world-whole ;' 3 you get rid of time and space in this 
 way. In reply we need only say we are not talking of. a 
 mundus intelligibilis ; we are talking of a mundus phenomenon 
 and its quantity, and we cannot ignore the conditions of 
 sensibility without destroying its very being. If we ignore 
 space in general as an a priori condition of the possibility of 
 phenomena, the whole world of sense vanishes, which alone 
 forms the object of our inquiry. 
 
 Second Antinomy {of Quality) 
 
 Thesis. — Every composite substance in the world con- 
 sists of simple parts, and there exists nothing but the simple 
 or what is composed of it. 
 
 p R00F . — if composite substances do not consist of 
 simple parts, then if all composition were annihilated in 
 
 1 Contrast carefully with this the critical solution, p. 318, infra, which 
 is quite different. 
 
 2 Cf. p. 49, supra. 3 Cf. Proleg. p. 1 54. 
 
292 COMPOSITUM AND TOTUM 
 
 thought nothing would remain. But with substances com- 
 position is only a contingent relation, an external state, apart 
 from which they must still subsist as permanent beings. 
 Hence our hypothesis is false, and therefore there must 
 be simple parts as the first elements of all composition. 
 
 It must, however, says Kant, be observed that the above 
 inference of the simple from the composite is only true of 
 self-subsistent beings. That is, ' if I speak of a whole which 
 consists necessarily of simple parts, I understand thereby 
 only a substantial whole, as the true compositum, i.e. the con- 
 tingent unity of the manifold, which, given as separate (at 
 least in thought), is brought into a mutual connection, and 
 thus constitutes one thing.' For example, the above argu- 
 ment does not apply to (a) space or to time. Space is 
 not a compositum in the above sense, it is rather a totum , 
 for its parts are possible only in the whole, and not the 
 whole by means of the parts. It might perhaps be called 1 
 a compositum ideale, but not a compositum reale. And so 
 also with time. Again the argument (b) will not apply 
 to that which only belongs to the state of a substance, even 
 though this state possesses quantity, i.e. a certain degree 
 of change does not arise from the accumulation of many 
 simple changes. 
 
 So we must not suppose that the inference of the simple 
 from the composite is valid of everything composite. As a 
 matter of terminology, Leibnitz's word monad should be 
 reserved for 'the simple,' given immediately as simple sub- 
 stance (e.g. in self- consciousness) ; while the word atom 
 should signify merely the simple, as an element of the 
 composite. So then this thesis might well be called 
 Transcendental Atomistic ; but as the word has been used 
 to designate a particular theory of corporeal phenomena, 
 
 1 Cf. p. 50, supra. 
 
xxiii INFINITE DIVISIBILITY 293 
 
 it is better to call it the dialectical principle {Grundsatz) of 
 Monadology. 
 
 Antithesis. — (a) No composite thing in the world 
 consists of simple parts, and (f3) there exists nowhere in the 
 world anything simple. 
 
 Proof. — (a) For suppose that a composite does consist 
 of simple parts. Composition is only possible in space, 
 therefore each simple part must occupy space. 1 But every- 
 thing which occupies space contains parts external to each 
 other, and is therefore composite. Hence it follows that 
 the supposed simple element must be a substantial com- 
 posite, which is self- contradictory. (/3) Grant that an 
 absolutely simple object were found in experience. The 
 empirical intuition of such an object must contain no 
 manifold whatever. But of this we could never be sure ; 
 for though we were not conscious of any manifold in an 
 object, we could not therefore assert that no manifold was 
 there. So then absolute simplicity cannot be inferred from 
 any perception whatever ; as, therefore, an absolutely simple 
 object cannot be given in any experience, and as the world 
 of sense must be regarded as the sum total {Inbegriff) of 
 all possible experience, nothing simple exists in the world. 
 
 This second proposition (as banishing the simple from 
 nature) goes further than the first, which merely banishes 
 it from the intuition of the composite. Hence we were 
 unable to prove it from the concept of a given object of 
 external intuition (of the composite), but were obliged to 
 prove it from the relation of such given object to a possible 
 experience in general. 
 
 Remarks. — The Monadists do not see that it is one 
 thing to say that the concept of the composite requires 
 the concept of the simple ; quite another thing to dis- 
 
 1 Vide p. 163, supra. 
 
294 PECULIARITY OF ANTITHESIS chap. 
 
 cover for the i?itaiiion of the composite, the intuition of 
 the simple. It is this latter that is quite impossible. Of 
 course, as in Antinomy I., the Leibnitzians may try to evade 
 the argument by denying that space is the condition of the 
 possibility of objects of external intuition ; but let us again 
 remind them that we have only to do with phenomena ; 
 their position would be tenable enough if bodies were 
 things in themselves. 
 
 ' The second dialectical assertion (in the antithesis) 
 possesses the peculiarity of having opposed to it a dogmat- 
 ical proposition, which, alone among all sophistical asser- 
 tions, undertakes to prove evidently, in the case of an 
 object of experience, the actuality of that which we 
 counted before as merely belonging to transcendental ideas, 
 viz. the absolute simplicity of substance.' The proposition 
 is that the Ego is a simple substance. We need not now 
 repeat what we said before on this subject (cf. pp. 262-5) ; 
 we merely remark that even if the Ego in relation to itself 
 were absolutely one, still, when this subject is regarded 
 externally as an object of intuition it would, as pheno- 
 menon, be composite. And it must always be regarded 
 in this way, if we wish to know whether there is or is not 
 contained in it a manifold whose parts are external to each 
 other. 
 
 Third Antimony {of Relation) 
 
 Thesis. — Causality, according to the laws of nature, is not 
 the only causality from which the phenomena of the world 
 can be deduced. To explain them, we must necessarily 
 assume, in addition, a causality through freedom. 
 
 Proof. — Suppose that there is no other kind of causality 
 than that according to laws of nature. Then everything 
 that happens presupposes an antecedent which it follows 
 
xxin TRANSCENDENTAL FREEDOM 29s 
 
 inevitably according to a rule. But this antecedent itself 
 was also something that once happened, and therefore pre- 
 supposes another antecedent. Hence, then, if everything 
 happens solely in accordance with the laws of nature, there 
 can never be any first beginning ; i.e. there is never a 
 sufficient cause for anything, which contradicts the law of 
 causality itself. So that to suppose that " all is nature " is 
 self- contradictory, and therefore we must admit another 
 kind of causality, through which something happens without 
 its cause being determined according to necessary laws by 
 a preceding cause. That is, we are forced to admit an 
 absolute spontaneity of causes, which originates of itself a 
 series of phenomena proceeding according to laws of nature, 
 i.e. we must have Transcendental Freedom. 
 
 Remarks. — It is important to bear in mind that this 
 transcendental Idea of Freedom is something quite different 
 from the psychological concept of the same. 1 Freedom, in the 
 transcendental sense, is merely the faculty of beginning 
 an event spontaneously ; it is the stone of stumbling in 
 philosophy, because people are not careful to preserve this 
 strict sense of the word. Now what we have proved so 
 far is this — that a free cause is absolutely required to 
 account for the origin of the world. Hence [as freedom 
 is a vera causa] we feel ourselves authorised to admit a 
 faculty originating a series in time, not only at the begin- 
 ning of things, but in the ordinary course of nature : i.e. 
 we attribute free action to substances. We must not think 
 the fact that everything has a beginning in time any 
 difficulty ; for an origin as to causality is a different thing. 
 Such an event as the act of a free agent must of course 
 succeed, but does not proceed from, its temporal antecedents. 
 
 1 Cf. Proleg., p. 141 n., and cf. for a full account of the difference 
 Theory of Ethics, p. 190. 
 
TRANSCENDENTAL PHYSIOCRACV 
 
 It may be observed that this need of a free first cause to 
 account for the origin of natural phenomena was felt by all 
 philosophers of antiquity (the Epicureans excepted), and led 
 them to the doctrine of a primum mobile. 
 
 Antithesis. — There is no freedom, but everything in 
 the world takes place solely according to laws of nature. 
 
 Proof. — If there be such a thing as a faculty originating 
 a state of itself, we are led to strange consequences. For 
 every beginning of an act presupposes a state in which the 
 cause is not yet active, and a dynamically primary beginning 
 of an act presupposes a state which has no causal con- 
 nection with the preceding state of that cause — that is, in no 
 wise follows from it. Hence we have a post hoc, in no way 
 determined by a propter hoc, by which the natural law of 
 causality is totally destroyed. If then we admit tran- 
 scendental freedom we are admitting something which can 
 be met with in no experience, and which actually destroys 
 the unity of experience ; it is consequently a mere fiction, 
 and we have nothing but nature to fall back upon. Freedom 
 (independence) from the laws of nature is doubtless a 
 deliverance from restraint, but also from the guidance of all 
 rules. 1 Nature and transcendental freedom differ from 
 each other as conformity to law and lawlessness ; and if we 
 assume the latter we are at once giving up the possibility of 
 any system at all. 
 
 Remarks. — The argument of the advocate of this 
 Transcendental Physiocracy may be put thus. If you do 
 not admit a mathematical first in relation to time, you Jiave 
 710 need to seek a dynamical first in regard to causality. 
 
 1 Kant points out very clearly the inaccuracy of such a statement, 
 Theory of Ethics, p. 65 — "Although freedom is not a property of the 
 will depending on physical laws, yet it is not for that reason lawless," 
 etc. 
 
xxin A NECESSARY BEING 297 
 
 There is no evidence that there is any such dynamical first. 
 The unity of experience requires us to suppose that sub- 
 stances have always existed in the world, and so there is no 
 difficulty in believing that the change of their states, i.e. a 
 series of changes, has always existed too. This infinite 
 series is certainly incomprehensible, but so are all ultimate 
 facts, e.g. the fact of change. And, further, even if we do 
 admit the existence of a transcendental faculty of freedom, 
 it can only exist out of and beyond the world. To attribute 
 such a faculty to substances in the world is to substitute 
 chaos for cosmos, to contradict the law of causality, and so 
 to take away our sure test for distinguishing actual experi- 
 ence from visionary dreams. 
 
 -Fourth Antinomy (of Modality) 
 
 Thesis. — There exists an absolutely necessary being 
 belonging to the world, either as a part, or as the cause of it. 
 
 Proof. — The world of sense, as the sum total (das 
 Game) of all phenomena, contains a series of changes. 
 (For without such a series the representation of the series of 
 time itself, as a condition of the possibility of the world of 
 sense, could not be given to us.) 1 Now every change 
 stands under its condition, which precedes it in time and 
 renders it necessary. But [see p. 242] the existence of a 
 given conditioned presupposes a complete series of condi- 
 tions up to the absolutely unconditioned, which alone is 
 absolutely necessary. Hence something absolutely necessary 
 must exist if change is to exist as its consequence. 
 
 1 Objectively [logically], time, as the formal condition of the possibility 
 of change, precedes all change ; but subjectively [chronologically], and in 
 the actuality of consciousness, the representation of time, like every 
 other, is given only on occasion of perception. 
 
298 EMPIRICAL CONTINGENCY chap. 
 
 Further, this necessary thing must belong to the world 
 of sense. For suppose it did not ; then, as the beginning 
 of a series in time can be determined only by that which 
 precedes it in time, a necessary being outside the world 
 could never originate a series of changes in the world. Its 
 causality must belong to time, and therefore it must itself 
 be in the world, and so our supposition that our necessary 
 being exists outside the world leads to absurdity. Hence, 
 then, we have proved that ' there is contained in the world 
 [of sense] something that is absolutely necessary (whether 
 it be the whole world-series itself, or only a part thereof).' 
 
 Remarks. — The above argument, which ascends from 
 the conditioned in phenomena to the unconditioned in 
 concepts, is commonly called the cosmological proof of the 
 existence of a necessary being, and is the only one which 
 can be here employed. Taken by itself it does not show 
 whether this necessary being belongs to the world or not ; 
 but still we have seen that if we take as the foundation of 
 it the fact of change in the world, we shall have to conclude 
 that the necessary being must also be in the world, other- 
 wise we shall be guilty of making an unwarranted saltus 
 (/x£Ta/3a<ris eU aXXo yeros). And the truth of this may be 
 shown in another way, as follows. 
 
 As we saw before (page 200), the word contingent is 
 ambiguous. The empirically contingent is that which can 
 only exist as the consequence of something else. The 
 intelligibly contingent (and this is the proper meaning of the 
 word) is that, the contradictory of which is possible — that 
 of which the non-existence can be conceived. Now change 
 cannot be proved contingent in the latter sense ; it is only 
 empirically contingent, and therefore can only lead to a 
 necessary being in the time-series, and hence the Uncondi- 
 tioned at which we arrive must be regarded as in the world. 
 
REMARKABLE CONTRAST 299 
 
 Antithesis. — There does not at all exist any absolutely 
 necessary being, either (a) in the world, or (b) out of it, as 
 its cause. 
 
 Proof. — For if (a) be true there are only two possible 
 cases. Either the absolutely necessary being must be a 
 First Cause, unconditionally necessary, and therefore un- 
 caused, which would be at variance with the dynamical law 
 of the determination of all phenomena in time [as was 
 proved before in the antithesis of Antinomy III.] ; or else 
 it must be identified with the infinite series of world causes, 
 in which case we shall have a series contingent and con- 
 ditioned in all its parts, and yet as a whole absolutely 
 necessary. But this is a self-contradictory supposition, for 
 the existence of an aggregate cannot be necessary if no 
 single part of it possesses necessary existence. 1 If (b) be 
 true we are led into the absurdity of holding that the 
 causality of a cause, which itself is out of time, can begin 
 in time, as we showed more fully under the thesis. So 
 the conclusion of the antithesis is, that a necessary being 
 exists nowhere. 
 
 Remarks. — In this Antinomy there is a very remarkable 
 contrast. The very same grounds of proof which established 
 in the thesis the existence of an original being, proved in 
 the antithesis with equal strictness its non-existence. We 
 found first that a necessary being exists because the whole 
 time past contains the series of all conditions, and with it, 
 therefore, the unconditioned ; secondly, that there does not 
 exist any necessary being for the same reason, viz. that the 
 whole time past contains the series of all conditions — 
 
 1 It might be objected that this is not obvious, and further, that it is 
 contradictory to Kant's own statement (p. 285) that ' the unconditioned 
 may be thought as existing merely in the entire series, in which all the 
 members without exception are conditioned, and only the totality of 
 them quite unconditioned, in which case the regress is called infinite.' 
 
300 FOURTH ANTINOMY chap. 
 
 which are themselves, therefore, in the aggregate, con- 
 ditioned. The explanation is this. We attend in the 
 thesis solely to the absolute totality of the series of con- 
 ditions, the one of which determines the other in time, and 
 thus arrive at a necessary unconditioned. In the anti- 
 thesis we consider on the contrary the contingency of every- 
 thing that is determined in the time-series. 
 
 11 A good deal of misapprehension as to this fourth An- 
 tinomy has arisen through not attending to Kant's language, 
 which is, as usual, very precise. In the first place we 
 must establish the differences between Antinomies III. and 
 IV. We believe that the main point of importance is to 
 be found in the statement (Kritik, p. 345) that here the 
 series is not a series of intuitions, as in the other Anti- 
 nomies, but of concepts. This marks out the fourth 
 Antinomy as quite different from the others. Moreover, 
 in the Antinomy of Relation we are discussing the nature 
 of the causality of substances : Is this causality ever free, 
 or is it always in accordance with the law of nature ? 
 In the Antinomy of Modality we ask : Is there such a 
 thing as a substance independently existing, a causa suil 
 Without considering now the way in which substances act, 
 we inquire, Is there any substance anywhere possessed 
 of independent existence ? The thesis asserts that such 
 a necessary being (for necessary it must be if uncon- 
 ditioned) must exist in the world, while the antithesis 
 denies that it exists anywhere. And l there are three 
 possible ways in which the thesis may be true — this neces- 
 sary being may be (a) a part, or (b) the cause of the world, 
 or (c) may be identified with the whole cosmical series. 
 
 1 For a confirmation of this, vide p. 309, infra. ' If we assume the 
 existence of an absolutely necessary being, whether it be the world 
 itself, or something in the world, or a cause of the world,' etc. 
 
xxni POINT OF THE THESIS 301 
 
 The antithesis denies the possibility of any of these supposi- 
 tions, and shows in detail that neither (b) nor (c) can be 
 true. Case (a) hardly requires separate refutation, for any 
 part which is unconditionally necessary will be a causa sui, 
 and hence uncaused ; and so this opinion is practically open 
 to the same difficulties which beset case (b), in which really it 
 must merge. So far all is clear, but several objections have 
 been raised, e.g. (1) How, it is asked, are we to explain 
 the fact that both thesis and antithesis refute the notion of 
 the necessary being existing outside the world as its cause ? 
 As they agree here perfectly, might not this argument be 
 eliminated from both sides? (2) How does it happen 
 that the proof of the thesis in this Antinomy is direct, and 
 so different from the seven other proofs, which are all ad 
 absurdum ? 1 (3) It seems no proper part of the thesis to 
 urge that the whole cosmical series is necessary, for in the 
 other antinomies the thesis always asserts the unconditioned 
 to be absolute, while the antithesis says that it is infinite 
 (cf. supra, p. 285). 
 
 To answer these — (1) Let us remember exactly what we 
 want to prove in each case. In the thesis we desire to 
 show not only that a necessary being exists somewhere, 
 but that it exists in the world. That this latter point is 
 essential may be seen from the statement of the thesis 
 given in the Proleg., p. 133. 'In the series of the world- 
 causes there is some Necessary Being.' And in the anti- 
 thesis we are striving to show not only that a necessary- 
 being does not exist in the world, but that it does not exist 
 at all. Hence, then, it is requisite in both cases to disprove 
 the notion that a necessary being may exist outside the 
 world. The above statement of the argument will enable 
 the reader to judge of the correctness of Professor E. Caird's 
 1 Cf. Fischer's Commentary, p. 219. 
 
302 MISTAKES OF CRITICS chap, xxiii 
 
 statement {Phil, of Kant, p. 574). 'The parallelism 
 between thesis and antithesis would have been more 
 complete, if Kant had not introduced under the former 
 the proof that the necessary being must be in the world,' 
 etc. This misses the real point of the argument. (2) 
 The proof of the thesis is not direct. We begin by stating 
 the general principle (p. 242), underlying all the arguments 
 of the Dialectic, and then we show indirectly that this 
 Unconditioned to which the given conditioned leads must 
 be in the world, for if it were out of the world its causality 
 could not be in time. (3) In the thesis we say : There 
 is an unconditioned causa sui in the world ; in the anti- 
 thesis we say : All is contingent. This is exactly parallel 
 to Antinomy III., where we say in the thesis, there is 
 such a thing as free causation ; and in the antithesis, all 
 causation is in accordance with the law of nature. Critics 
 have not been careful enough in distinguishing Antinomies 
 III. and IV.; hence these mistakes, e.g. Professor Wallace 
 tells us (Kant, p. 186) that the fourth Antinomy 'is only a 
 slightly altered form of the third.' 
 
 Professor Maguire has pointed out (Lectures, p. 163) 
 that opinions (a) and (6) were held respectively by Des- 
 cartes and Leibnitz, and that the complicated form of the 
 arguments arises from the fact that Kant is stating opinions 
 that have actually been held. 1 
 
 1 Cf. infra, p. 309, note. 
 
CHAPTER XXIV 
 
 THE INTEREST OF REASON IN THESE CONFLICTS 
 
 § 3. The dialectical assertions which we have been con- 
 sidering are so many attempts to solve four natural and in- 
 evitable problems of reason. There can be neither more nor 
 less than four, because there are no other series of synthetical 
 hypotheses which limit Reason in endeavouring to grasp 
 the Unconditioned. Let us now briefly consider what side 
 in the controversy w T e should most willingly take if we were 
 guided only by considerations of interest, leaving altogether 
 out of account any logical test of truth. 
 
 We may remark in the assertions of the antithesis a 
 complete uniformity in the mode of thought and a perfect 
 unity of maxims, viz. we here adopt the principle of pure 
 Empiricism not only in the explanation of the phenomena 
 in the world, but also in the solution of the transcendental 
 ideas of the universe itself. The assertions of the thesis, 
 on the contrary (apart from the empirical mode of explana- 
 tion employed within the phenomenal field), are grounded 
 on intellectual beginnings, and its maxims are so far not 
 simple. 
 
 On the side of the thesis, or of dogmatism, we find — 1. 
 A practical interest, inasmuch as the different theses are so 
 many foundation stones on which morality and religion are 
 
304 INTEREST OF EMPIRICISM chap. 
 
 built up, supports of which the counter-assertions would 
 deprive us. — 2. A speculative interest, inasmuch as the dog- 
 matist at least has a solution to offer, while the holder 
 of the opposite opinion cannot do much more than say that 
 the problem at issue is insoluble, as each question ever 
 points to another question, and so on ad infinitum. 
 
 But on the other hand empiricism holds out to the 
 speculative interests of reason certain attractions and 
 advantanges far exceeding . those that the dogmatist can 
 offer. 1 For with the empiricist the understanding is always 
 on its own proper ground, namely, the field of possible ex- 
 perience, the laws of which it can investigate, and by means 
 of which it can attain certain and clear knowledge without 
 end. While it stays on this ground it is safe from error ; 
 but if, as often happens, empiricism, in relation to Ideas, 
 becomes dogmatic in its turn and boldly denies that which 
 is above the sphere of its intuitive knowledge, it becomes 
 guilty of a want of modesty all the more blamable because 
 an irreparable inquiry is thereby inflicted on the practical 
 interests of Reason. Here, then, is the difference between 
 Epicureanism 2 and Platonism ; the former encouraging 
 
 1 Kirchmann remarks that the side of the antithesis would have much 
 greater speculative interest now since Darwin's labours, which have 
 enormously increased both the number of antecedents and the length of 
 time involved in the series, than when Kant wrote. 
 
 - It is doubtful, says Kant, whether Epicurus ever put forward these 
 propositions as objective assertions. If he meant them to be no more 
 than logical maxims he would have shown thereby a more genuine 
 philosophic spirit than any of the philosophers of antiquity. That, in 
 the explanation of phenomena, we must proceed as if the field of inquiry 
 were enclosed by no Boundary or Beginning of the world ; that we 
 should accept the material of the world as it must be, if we want to learn 
 anything about it from experience ; that we must look for no other 
 origination of events than that brought about by the law of nature ; and 
 that no cause distinct from the world must be appealed to — are all 
 correct and indispensable scientific maxims, but we must take care not 
 
xxiv ARCHITECTONIC INTEREST 305 
 
 Science at the expense of the Practical ; the latter, though 
 supplying excellent practical principles, neglecting physical 
 investigation and substituting therefor idealistic explanations 
 of natural phenomena. 
 
 3. The side of dogmatism has also the advantage of 
 popularity ; for the common understanding x likes to have 
 some firm and definite basis on which its theories may be 
 founded, and does not find the least difficulty in the idea 
 of the unconditioned beginning of all synthesis. And, 
 further, it thus finds itself in a situation where not even 
 the most learned can have the advantage of it ; for one 
 peculiarity of the concepts which it assumes is that no one 
 knows anything about them. So indolence and vanity are 
 strongly in favour of dogmatic principles. Again, human 
 Reason is by nature architectonic, i.e. it regards all cogni- 
 tions as belonging to a possible system, and therefore 
 accepts only such principles as are capable of being placed 
 along with others in a general system. But the propositions 
 of the antithesis are of such a character that they render 
 the completion of an edifice of cognition impossible. Hence, 
 then, the architectonic interest of Reason (which demands 
 a priori not empirical but pure rational unity) serves as a 
 natural recommendation for the assertions of the thesis. 2 
 
 to turn them into objective laws without very good reason. [However, 
 as Fischer points out (p. 224), Epicurus was an atomist, and atomism 
 is in every case nearer to asserting simple substances than to denying 
 them.] 
 
 1 Cf. Bacon, De Aug. Set'., v. 4. ' Magno studio appetunt homines 
 aliquem habere intra se cogitationum Atlantem, aut polos, qui intellectus 
 fluctuationes et vertigines aliquando regent.' 
 
 2 Fischer paraphrases well : ' Our reason is interested in knowing 
 the unconditioned object or absolute unity of things which is the uni. 
 verse ; it is also interested in systematically combining its knowledge 
 into a whole of science. The former may be called the speculative, the 
 latter the architectonic interest.' Commentary, p. 222. 
 
 I X 
 
3 o6 A SOLUTION POSSIBLE CHAP - 
 
 § 4. The transcendental problems of pure Reason, in so far 
 as they must absolutely admit of a possible solution. 
 
 To attempt to solve all problems and answer all questions 
 would be ridiculous, but nevertheless there are certain 
 sciences of such a nature that every question arising within 
 their sphere (ouastiones domeslicce) must be capable of an 
 answer from the same sources as those from which the 
 question springs. Transcendental Philosophy is such a 
 science ; all questions raised by the pure Reason must be 
 answerable by the pure Reason, and we cannot here excuse 
 ourselves by the plea of unavoidable ignorance — a solution 
 is absolutely requisite. 1 In particular cosmological ques- 
 tions can even be answered as regards the constitution of 
 their object, for it must be given empirically, and leaves 
 only the question of adequateness to its Idea. This is 
 a peculiarity of the cosmological Ideas, for they alone 
 can postulate their object and the empirical synthesis 
 requisite for its concept as given ; and the question which 
 springs from them relates only to the progress of this 
 synthesis, so far as it is to contain absolute totality — which 
 is no longer empirical. The general question, ' What is the 
 constitution of a transcendental object,' would be unanswer- 
 able ; in fact the question itself is nothing, for it is concerned 
 with the constitution of a something which can be thought 
 through no determinate predicate — being quite beyond the 
 sphere of objects which can be given to us. So to such 
 questions, e.g. as those which would arise in Transcendental 
 Psychology, no answer is the proper answer. But in our 
 present inquiry we cannot by complaints of the narrow 
 limits of that Reason escape the obligation to give at least 
 a critical solution ; for all these questions relate to an 
 
 1 Cf. Preface, p. 4, supra. 
 
REASON NOT DELUSIVE 307 ' 
 
 object which can be given nowhere else than in our 
 thoughts, viz. the absolutely unconditioned totality of the 
 synthesis of phenomena. If we are not able to say any- 
 thing certain about this from our own concepts, we must 
 not throw the blame on the thing as obscure (for no such 
 thing can be given in experience), but on our Idea, cor- 
 responding to which we obstinately assume an actual 
 object. In all possible perceptions we are always sub- 
 ject to conditions ; we never come to the Unconditioned so 
 as to be able to decide whether it is to be placed in an 
 absolute beginning of the synthesis or in an absolute totality 
 of the series without beginning. 1 Phenomena can only be 
 explained so far as the conditions of that explanation are 
 given in perception ; but the sum of that which is given, 
 considered as an absolute whole, is itself 2 no perception. 
 Yet it is this very whole the explanation of which is required 
 in the transcendental problems of Reason. As, therefore, 
 the solution of these problems can never occur in experience, 
 a dogmatical solution is impossible. The critical solution, 
 at which we shall soon arrive, does not consider the ques- 
 tions objectively, but only with reference to the basis of 
 the cognition on which they rest. 3 
 
 1 Cf. Hamilton's Law of the Conditioned. 2 Read keine for tine. 
 
 3 The heading of the above section sufficiently refutes the old and 
 now exploded charge against Kant, that he made the Reason a source 
 of unavoidable and insoluble delusion. Sir Wm. Hamilton says, 
 e.g. (Lectures, ii. p. 543) ' He makes the Reason a complexus of 
 antilogies,' etc. Kant held what none of his critics would have denied, 
 ' that there are some principles of our nature perpetually leading us 
 astray' (Mansel. Proleg. Log. p. 153), but he also held that when 
 Reason leads us into error, it must also be able to lead us out again. 
 And, in fact, in the discussion which succeeds, he gives a most elaborate 
 critical solution of these antinomies into which Reason inevitably falls. 
 Cf. Kritik, pp. 394, 410. 
 
•308 SCEPTICAL ACCOUNT 
 
 § 5. Sceptical Account of the Cosmological questions 
 in all four Transcendental Ideas 
 
 We should no doubt cease to ask for a dogmatical answer 
 to our questions if we understood beforehand that, be the 
 answer what it may, it would only serve to increase our 
 ignorance, and to drive us from one incomprehensibility 
 into another. Let us then apply the sceptical method which 
 was described above (p. 287). If we can show that the 
 cosmological Idea, whatever direction the regressive syn- 
 thesis of phenomena towards the unconditioned may take, 
 will always be either too large or too small for any concept 
 of the understanding, then the Idea, referring as it does 
 to an object of experience which must correspond to a 
 possible concept of the understanding, is void and without 
 significance, for want of an adequate object. 
 
 I. Suppose that the world has no beginning; this is 
 too large for our concept, which, consisting in a successive 
 regress, can never reach the whole eternity that has elapsed. 
 Suppose that it has a beginning ; this is too small, for as a 
 beginning presupposes a preceding time, it is not uncon- 
 ditioned, and the law of the empirical use of the under- 
 standing obliges us to ask for a higher time-condition ; so 
 that the world is evidently too small for this law. In 
 the same way to say that the world is infinite and un- 
 bounded is too large for any possible empirical concept. 
 If it is finite and bounded, what determines these bounds? 
 Void space is not a self-subsistent correlate of things and 
 cannot be a final condition — still less an empirical condition, 
 forming a part of a possible experience ; for how can there 
 be any experience of the absolutely void ? But the absolute 
 totality of the empirical synthesis requires that the uncondi- 
 
DIFFICULTY 509 
 
 tioned be an empirical concept. Hence a bounded world is 
 too small for our concept. 
 
 II. If every phenomenon in space (Matter) consist of 
 an infinite number of parts, the regress of the division is 
 always too large for our concept. If the division of space 
 stop at any member (the simple) it is too small, for that 
 member admits of a regress to more parts contained in it. 
 
 III. If all events happen in accordance with the law of 
 nature, the series of conditions a parte priori must be 
 prolonged for ever, and this is too large. If we admit 
 origination from freedom, we are compelled to ask the why 
 of the first cause ; and so to go beyond it, so that this 
 supposition is too small. 
 
 IV. If we admit an absolutely necessary being (whether it 
 be the world itself, or something in the world, or the cause 
 of the world), we must place it in a time infinitely distant 
 from any given moment ; for otherwise it must be dependent 
 on some other and older existence. , But such an existence 
 is unapproachable by our empirical concept, and too large 
 to be reached by a continued regress. If, on the other 
 hand, we believe that everything belonging to the world — 
 whether as condition or conditioned — is contingent, every 
 given existence is too small for our concept, for it compels 
 us to seek for some other existence on which it depends. 1 
 
 So in all cases the cosmical Idea is 2 either too large or 
 
 1 There is a difficulty here. In the first three antinomies, the 
 object of the antithesis in each case is too large, and that of the thesis 
 too small, for the concept of the understanding ; but when we come to 
 the fourth antinomy we find the very reverse. This is very curious in 
 a book so systematic as the Kritik, nor is it easy to see any way in 
 which the discrepancy can be explained. Cf., however, p. 300 for a 
 possible explanation. 
 
 2 We express ourselves thus (says Kant), and do not say conversely 
 that the empirical concept is too large or too small for the cosmological 
 Idea, because a possible empirical concept is the only standard by which 
 
3 io TRANSCENDENTAL IDEALISM 
 
 too small for the empirical regress, and consequently for 
 every concept of the understanding, and hence we are led 
 to the well-founded suspicion that the errors we have been 
 considering arise from subjective causes. 
 
 § 6. Transcendental Idealism as the key to the solution 
 of pure Cosmological Dialectic 
 
 The doctrine of Transcendental Idealism proved in our 
 /Esthetic asserts that all things intuited in space and time, 
 i.e. all objects of an experience possible to us, are nothing 
 but phenomena or mere representations ; while the tran- 
 scendental realist regards these representations as things in 
 themselves. This is not to be confounded with empirical 
 or material idealism, which, while admitting the actuality of 
 space, doubts the existence of objects extended in it, and so 
 leaves us without a sufficient criterion to distinguish between 
 dreaming and truth. Transcendental or formal Idealism, 
 on the contrary, allows that the objects of external intuition 
 as intuited in space, and all changes in time as presented 
 to the internal sense, are actual. But, nevertheless, time 
 and space, with all the phenomena therein, are not in 
 themselves things , they are nothing but representations, 
 and cannot exist apart from our minds. The objects of 
 experience, then, are never given in themselves, but only in 
 experience, and have no existence apart from it. That 
 there may be inhabitants in the moon, though no one has 
 ever perceived them, must certainly be admitted ; but this 
 assertion means only that we may meet with them in the 
 possible progress of experience. For [according to the 
 Second Postulate, p. 193] everything which stands in connec- 
 tion with a perception according to the laws of the progress 
 
 we can judge an idea. Thus we do not say the man is too large for his 
 coat, but, the coat is too small for the man. 
 
xxiv CRITERION OF ACTUALITY 311 
 
 of experience is actual. In fact, to call a phenomenon an 
 actual thing, prior to perception, can only mean that we 
 must meet with such a perception in this empirical 
 progress. 
 
 The faculty of sensuous intuition is properly a re- 
 ceptivity — a capacity of being affected in a certain manner 
 with representations, the relation of which to each other is 
 a pure intuition of time and space. These representations, 
 in so far as they are connected and determinable in this 
 relation according to the laws of the unity of experience, 
 are called objects. We may, if we like, term the merely 
 intelligible cause of phenomena in general, the transcend- 
 ental object — but merely in order to have something to 
 correspond to sensibility ; this non-sensuous cause is com- 
 pletely unknown to us, and cannot be intuited as an object. 
 It is essential to insist that the only criterion of actuality 
 we have is that contained in the Second Postulate. Things 
 in themselves, without any relation to possible experience, 
 are for me non-existent, consequently are not objects, for 
 they are not contained in the series of the empirical regress. 
 
CHAPTER XXV 
 
 THE CRITICAL SOLUTION 
 
 § 7. Critical decision of the Cosmological Conflict of the 
 Reason with itself 
 
 The whole antinomy of pure Reason rests on this 
 dialectical argument : If the conditioned is given, the whole 
 series of its conditions is also given ; but objects of sense 
 are given to us as conditioned, therefore, etc. (see p. 242, 
 supra). On this reasoning we make one or two remarks. 
 
 Firstly, So much is obviously true : If the conditioned 
 is given, a regress in the series of all its conditions is 
 required. For this is but a logical postulate of Reason — to 
 pursue, through the Understanding, as far as possible, that 
 connection of a concept with its conditions which is 
 inherent in the concept itself. 
 
 Secondly, If both conditioned and condition are things 
 in themselves, and if the former be given, not only is the 
 regress to the latter required (aufgegeben), but the latter is 
 actually given (gegelmi) with the former ; and so in this 
 case the complete series of conditions (and consequently 
 the unconditioned) is given, the conditioned being only 
 possible through that series. 
 
 But, on the other hand, if I am dealing with phenomena 
 
CHAr. xxv FALLACY 313 
 
 which, as mere representations, are not given at all, if I do 
 not attain to a knowledge of them, I cannot draw any such 
 inference. If the conditioned is given, I can only say that 
 a regress to the conditions, i.e. a continuous empirical syn- 
 thesis in that direction, is required, and that conditions 
 must exist which are given through that regress. 
 
 So. we may now see that there is in the above dialectical 
 syllogism a sophismafigura dictionis. The major premise uses 
 ' conditioned ' in the transcendental signification of the pure 
 category, but the minor in the empirical signification of a 
 concept of the understanding applied to mere phenomena. 1 
 The fallacy is natural, but still a fallacy. The synthesis of 
 the conditioned with its condition, and the whole series of 
 conditions in the major, are in no way limited by time, and 
 are free from any concept of succession. The empirical 
 synthesis, on the contrary, and the series of conditions in 
 phenomena, subsumed in the minor, are necessarily suc- 
 cessive and given as such in time only. There is then this 
 fatal mistake in the argument adopted by both parties as 
 the foundation of cosmological assertions. However, we 
 may go still further and show not only that their premises 
 are unsound, but that they are disputing about nothing, 
 and that a certain transcendental illusion has been mocking 
 them with visions of actuality where there is none. 
 
 Of two contradictory propositions one must be true 
 and the other false. This we may call analytical opposi- 
 tion. But if two opposite judgments pre-suppose an 
 inadmissible condition, both will fall away, because the 
 condition has itself disappeared, and we cannot in such 
 case infer the truth of one from the falsity of the other. 2 
 
 1 Briefly, in the major we speak of the conditioned as cogitable, in 
 the minor as intuitible. 
 
 - Cf. Proleg. p. 136, for a very clear statement of this. 
 
3H DIALECTICAL OPPOSITION 
 
 This we may. call dialectical opposition. For example, 
 take the two statements. The world is finite ; the world 
 is infinite. These may both be false. For what right 
 have we to assume, as we do when we make either of 
 these statements, that the world may in itself be determined 
 in regard to quantity ? If we regard them as analytically 
 opposed, we are assuming that the world — the whole series 
 of phenomena — is a thing in itself. This we have no right 
 to do. We know it only in the empirical regress of the 
 series of phenomena, and not in itself. This series is 
 always conditioned, and so the world is never completely 
 given : it is not an unconditioned whole, and so does not 
 exist as such either with an infinite or finite quantity. 1 
 And the same may be shown of the other antinomies. 
 The number of parts in a given phenomenon is in itself 
 neither finite nor infinite ; because phenomenon is nothing 
 existing in itself; and its parts are given only in the regress 
 of the decomposing synthesis, which is never given as 
 absolutely complete. The same is true of the series of 
 subordinated causes, or of the conditioned up to uncon- 
 ditioned necessary existence.; because as a series of sub- 
 ordinated representations, it consists only in the dynamical 
 regress, prior to which it cannot exist as a self-subsistent 
 series of things in themselves. 
 
 Thus the antinomy of pure Reason is merely dialectical, 
 and is the conflict of an illusion produced by applying the 
 idea of absolute totality — admissible only as a condition of 
 things in themselves — to phenomena which exist only in 
 our representations and, when constituting a series, in a 
 
 1 Kant compares Zeno's paradoxes to the antinomies, and seems to 
 think that when Zeno denied that God was either at rest or in motion, 
 what he had in view was to show that properly neither motion nor 
 rest could be predicated of God as an absolute being. The two state- 
 ments were diakclically not analytically opposed. 
 
INDIRECT PROOF 315 
 
 successive regress. And this furnishes us with an indirect 
 proof of the Transcendental Ideality of phenomena, if we 
 are not satisfied with the direct proof set forth in the 
 Esthetic. E.g. If the world is a whole existing in itself, 
 it must be either finite or infinite. But the antithesis and 
 the thesis respectively show that these assertions are false. 
 Therefore the world, the sum -total ( Inbegriff) of all 
 phenomena, is not a whole existing in itself : phenomena 
 are nothing but mere representations, and this is what we 
 mean by Transcendental Ideality, as expounded in § 6. 
 
 It is important to observe, once more, that the proofs of 
 the four antinomies are not mere sophistry ; they are 
 perfectly valid on the supposition that phenomena (or a world 
 of sense which includes them all) are things in themselves. 
 So, though transcendental dialectic does not favour scepticism, 
 we see how useful is the sceptical method. 
 
 § 8. Regulative Principle l of Pure Reason in ?-elation to 
 the Cosmological Ideas 
 
 As through the cosmological principle of totality no 
 maximum of the series of conditions in the world of sense 
 considered as a thing in itself is given, but is only required 
 in the regress, so that principle cannot be regarded as an 
 axiom, 2 but merely as a problem. That is, it is not a 
 principle of the possibility of experience and of the 
 empirical cognition of objects of sense, and so not a 
 principle of the Understanding; for every experience is 
 confined within its own bounds (according to a given 
 intuition). Neither is it a constitutive principle of Reason 
 
 1 To distinguish Grundsatz from Princip, for which we have only 
 one word in EngHsh, we have generally italicised principle when it 
 stands for Princip. 
 
 - See p. 368, infra. 
 
316 PROGRESS AND REGRESS 
 
 authorising us to extend our concept of the world of sense 
 beyond all possible experience, and asserting the absolute 
 totality of the series of conditions to be given in the object. 
 It is merely a principle for the extension of experience 
 as far as possible, forbidding us to regard any empirical 
 bounds as absolute ; as a rule, telling us how we ought to 
 proceed in our empirical regress, but not anticipating, prior to 
 that regress, what is given in the object. Let us call it, there- 
 fore, a regulative principle. Now, if we were dealing with a 
 progress (from condition to conditioned), it would make no 
 matter whether we spoke of a progress in infinitum or in 
 indefinitum, for such a possible progress in the phenomenal 
 series always proceeds to infinity ; but in the case of a 
 regress we must make a distinction. 
 
 (i.) If a whole is given in empirical intuition, the regress 
 in the series of its internal conditions proceeds in infinitum, 
 e.g. the division of any given portion of matter. For that 
 matter is complete, and so, with all its possible parts, is 
 given ; the more remote members of the division are given 
 prior to the division. That is, the division proceeds to 
 infinity. 
 
 (2.) On the other hand, if only one member of the 
 series is given, from which the regress is to proceed to 
 absolute totality, the regress is only in indefitiitum. For 
 example, if we want to construct a genealogical pedigree, 
 the regress proceeds from any one member to one still 
 higher, and so on, and nowhere meets with any empirical 
 boundary ; but the series of ancestors is not given in its 
 absolute totality in any experience prior to the regress, 
 which then only proceeds in indefinitum. Of course, in 
 neither case is the series of conditions to be regarded as 
 given as infinite in the object : they are not things in 
 themselves, but only phenomena, which, as conditions of 
 
xxv CRITICAL SOLUTION 317 
 
 each other, are only given in the regress itself. The 
 distinction may be variously stated by saying that in case 
 (1) it impossible to recede to infinity : as more members are 
 empirically given than I can arrive at in the regress, it is 
 ever necessary to find the other members of the series ; in 
 case (2) it impossible to infinity to recede; as no member is 
 empirically given as absolutely conditioned, and so a higher 
 member is always possible, it is necessary to inquire for 
 other members. 
 
 § 9. Of the empirical use of the Regulative Principle of 
 Reason with regard to all Cosmological Ideas. 
 
 We have now established the validity of the principle of 
 Reason, as a rule for the extension of experience, not as an 
 axiom , as regulative, not constitutive. In fact we have 
 changed what was a dialectical principle into a doctrinal one, 
 just as useful in science as if it were an axiom for the a 
 priori determination of objects. We proceed in detail to the 
 critical solution, by means of it, of the several antinomies. 
 
 I. The ground of the regulative principle of Reason here, 
 as in the other cases, is the proposition that in our em- 
 pirical regress no experience of an absolute boundary is 
 discoverable. So it merely remains to ask whether this 
 never bounded regress is to be called indefinite or infinite. 
 And it is plainly only indefinite, for the reason explained in 
 the last section, viz. the world is not given in its totality in 
 any intuition. So that to the question concerning the world- 
 quantity, the first and negative answer is : The world has 
 no beginning in time, and no absolute boundary in space. 
 For if it had, it would be bounded by void time or void 
 space; now, being phenomenal, its boundaries must be 
 phenomenal, and therefore objects of perception. But we 
 
3 ,8 THE WORLD-QUANTITY 
 
 can have no perception of anything that has no content. 
 Therefore an absolute boundary of the world is empirically, 
 and so absolutely, impossible. 1 From this follows the affirm- 
 ative answer : The regress in the series of phenomena, as 
 determining the world -quantity, proceeds in indefinitum. 
 But it does not give us any determined concept of it; 
 still less a concept of an infinite quantity. [Briefly, Is the 
 world finite ? No. Is it infinite ? No, for it has no absolute 
 quantity. We can only say that the empirical regress must 
 go on ad indefinitum.\ 
 
 II. Here we are dealing with the totality of the division 
 of a whole given in intuition, and therefore, as we saw (p. 
 317), the regress will proceed in infinitum. But still we 
 are not entitled to say of a whole of this kind, that it 
 consists of an infinite number of parts , for although all the 
 parts are contained in the division of the whole, the whole 
 division is not contained therein. The division is contained 
 only in the regress, which first makes the series of parts 
 actual. So we can only say in this case, the whole is 
 infinitely divisible. We may fairly argue in the case of space 
 as a quantum continuum, that division can never surpass 
 the limits of composition, for otherwise space, which is 
 not self-subsistent, would cease to exist, and this is im- 
 possible. But, it may be objected, this argument would not 
 apply to a body in space, for it may be said that substance, 
 the subject of the composition, must remain in its elements, 
 even after their conjunction in space — which constitutes a 
 body — is annihilated. The answer to this is obvious. 
 Phenomenal substance (and this is all we can talk about) is 
 
 1 Kant calls attention to the fact that this proof is entirely different 
 from that given in the antithesis of Antinomy I. ; for there it was taken 
 for granted that the world is a thing in itself. And so too, the 
 conclusion drawn was different, for there the actual infinity of the world 
 was inferred. 
 
xxv MATHEMATICAL AND DYNAMICAL 319 
 
 not an absolute subject, and so the argument will not hold. 
 Further, this infinite divisibility can only be predicted of 
 a quantum continuum , it depends altogether on the nature 
 of space. We cannot assert of a quantum discretum, such 
 as an organised body, that every organic part is itself 
 organised ad infinitum. For in such a case the number of 
 organic parts is always definite and can only be determined 
 from experience. 
 
 With this caution, then, we may sum up briefly thus : 
 In a whole given in intuition is there any simple part ? 
 No. Is everything composite, is there an infinite number 
 of parts ? We do not know, though we must believe the 
 whole to be infinitely divisible. 
 
 Transition from the Mathematical to the Dynamical 
 Antinomies. 
 
 Before going further we must notice a remarkable 
 difference between the Cosmological Ideas, the first 
 two indicating a mathematical, the last two a dynamical 
 synthesis of phenomena. 1 Now, in a mathematical series 
 of phenomena, no other than a sensuous condition is 
 admissible ; heterogeneous members cannot be united 
 mathematically, and it is for this reason, as we have seen, 
 that both thesis and antithesis in the first two Antinomies 
 must be regarded as false 2 [for they attribute phenomenal 
 attributes to a thing in itself]. But it is not equally true 
 to assert that what can be dynamically united must be 
 homogeneous ; a dynamical series of sensuous conditions 
 may admit also a heterogeneous condition, which is not a 
 member of the series, but as intelligible lies outside it. 
 
 1 Cf. p. 162, supra. 
 - A quantitative unconditioned is an absolute contradiction. 
 
320 TRANSCENDENTAL FREEDOM 
 
 And from this fact we shall find a result to follow which we 
 should not have expected from an Antinomy, viz. that in 
 the Dynamical Antinomies both the thesis and antithesis 
 may be regarded as true from different points of view, and 
 thus satisfaction is done to the Understanding on the one 
 hand and to the Reason on the other. 
 
 [In short, in no case are the thesis and antithesis logical 
 contradictories ; in the Mathematical Antinomies they are 
 more like contrary, in the Dynamical like sub-co?ilrary 
 propositions.] 
 
 III. We can conceive only two kinds of causality — 
 through Nature and through Freedom ; and the tran- 
 scendental (not merely physiological) problem before us, 
 is to determine whether the latter is necessarily inconsistent 
 with the former. We 1 already have defined freedom in 
 the pure transcendental sense as the faculty of beginning 
 an event spontaneously ; and it is now to be observed that 
 this is postulated by the common practical concept of 
 freedom, viz. the independence of the will 2 from coercion 
 through sensuous impulses. Without transcendental we 
 cannot have practical freedom. Now, on the hypothesis of 
 the absolute reality of phenomena — if phenomena are things 
 in themselves — nature is a sufficient cause for every event, 
 and freedom is impossible. But if, on the contrary, 
 phenomena be mere representations connected according 
 to empirical laws, they must have a ground which is not 
 phenomenal. From this consideration we may be able to 
 arrive at a solution of the Antinomy, to which we now 
 address ourselves. 3 
 
 1 P. 295, supra. "- Cf. Theory of Ethics, p. 188. 
 
 3 It is to be remembered that the question before us about Freedom 
 is not psychological but cosmological. It is not sufficient to oppose the 
 indirect testimony of consciousness by itself to a principle like that of 
 
CAUSE AND CAUSALITY 
 
 Possibility of Causality through Freedom in harmony with 
 the universal law of Natural Necessity 
 
 All phenomena are representations, and as such suggest 
 a transcendental object or noumenon. Now, if a subject 
 possesses in itself a power which is not an object of 
 sensuous intuition, but by means of which it can be the 
 cause of phenomena, the causality x of this being may be 
 regarded from two points of view, (a) It is intelligible 2 
 as regards its action — the action of a thing .itself ; (b) 
 It is sensible as regards its effects as a phenomenon. And 
 since every cause must have a law of its causality, ac- 
 cording to which the effects follow, which we call its 
 character, we may express this by saying that the causality 
 of every phenomenal subject has both an intelligible and 
 empirical character. The acting subject in its intelligible 
 character is subject to no conditions of time — in it there 
 is no before or after — and so the intelligible causality 
 does not form part of the series of empirical conditions 
 which necessitate the event in the sensuous world, and 
 therefore can never be immediately cognised. In its 
 empirical character the subject will obey the law of nature 
 — all its actions will be explicable according to natural 
 laws ; while in its intelligible character, on the other hand, 
 it is free and independent of natural necessity, which exists 
 
 causality. What we want to show is that a cosmological principle 
 corresponding to this testimony of consciousness is not necessarily 
 contradictory to causality. And except freedom be overthrown in this 
 way and declared a priori impossible, our internal consciousness of 
 ought and the moral law must be allowed to have weight. 
 
 1 The causality is the action of the cause. The student should be 
 careful not to confuse the two. 
 
 2 " That in an object of the senses which is not itself phenomenon, 
 I term intelligible" says Kant at the beginning of this section. 
 
 I Y 
 
322 INTELLIGIBLE CHARACTER 
 
 only in the field of sensibility. Now all this is mere 
 general theory without proof, but at all events we may say 
 this. There is nothing impossible in the supposition that 
 there are some x natural causes which have a power that is 
 only intelligible, inasmuch as it is not determined to action 
 by empirical conditions, but solely upon grounds of the 
 understanding; so, however, that the action in the phe- 
 nomenon of this cause must be in accordance with all the 
 laws of empirical causality. 
 
 If we apply this to actual experience, what do we find ? 
 Man, to whom the rest of nature reveals itself only through 
 sense, cognises himself, not only by his senses, but also by 
 mere apperception ; and this in actions and internal 
 determinations which he cannot regard as sensuous im- 
 pressions. 2 He thus is to himself on the one hand a 
 phenomenon ; but on the other, in respect of certain 
 faculties, viz. Understanding and Reason, an intelligible 
 object. Now that this Reason of his does possess the 
 faculty of causality, or that at least we are forced so to 
 think, appears -from the imperatives which present them- 
 selves in the moral sphere. The word ought indicates a 
 
 1 Kant does not say that all natural causes have an intelligible 
 character. 
 
 2 Hence, what might be a mere invention in the case of the nou- 
 menon, supposed to be the basis of external phenomena, is raised to a 
 more conceivable position in the case of internal phenomena ; for here 
 we are convinced that there is a noumenon acting through intelligible 
 faculties, being conscious of these faculties, and also not being conscious 
 that they are determined from without. And the moral imperatives 
 show that we ascribe causality to this noumenon — not only the power 
 of becoming a phenomenon, but also the further causality of deter- 
 mining its empirical character. But it is to be observed that there 
 seems no reason to think, because free action is conceived by us only in 
 the Reason, that therefore Reason must be the only free cause or 
 intelligible character in nature. [This is the inference of Fischer and 
 Schopenhauer.] 
 
ELECTIVE WILL 323 
 
 possible action, the ground of which is nothing but a mere 
 notion. Whatever number of motives nature may present 
 to my wish, the moral ought is beyond its power to 
 produce. So that it would seem quite possible (from these 
 considerations as to the nature of the moral imperative) 
 that Reason should stand in an actually causal relation to 
 phenomena. If it does, it must of course exhibit an 
 empirical character. For every cause supposes a rule, 
 according to which certain phenomena follow as effects, 
 and 1 every rule requires uniformity in these effects — (this 
 is the very ground of the concept of cause as a power). 
 Hence then the elective will 2 of every man has an empirical 
 character; i.e. a certain causality of his reason manifests in 
 its phenomenal effects a rule, 3 according to which his 
 actions and their motives may be explained. Viewed on 
 this phenomenal side, then, no human action is free ; so if 
 we only consider the question physiologically, by simple 
 observation, we shall be led to regard the law of nature as 
 the only law of the human will. But if we consider a 
 human action, from the other point of view — in relation to 
 Reason as its producing cause — it would appear as if we 
 had come upon a rule and an order very different from 
 that of nature. For Reason often tells us that what has 
 happened ought not to have happened; and indeed we 
 sometimes fancy that we have met with a case in which an 
 action was done on moral, i.e. rational grounds. Of this, 
 however, we can never be quite certain, as our estimate 
 of an action can only relate to its empirical character; 
 
 1 Is this the Principle of the Uniformity of Nature, that the same 
 cause will always produce the same effect ? 
 
 3 Following Mr. Abbott {Theory of Ethics, p. 268, note), we 
 translate VVillkiihr (arbitrium) elective will, as opposed to Wille (ra- 
 tional will). 
 
 3 This shows why he chose the word character. 
 
324 REASON IS CAUSAL chap. 
 
 that it is truly moral could never be proved from observa- 
 tion. 
 
 It may, however, be said . Suppose that the possibility 
 of reason standing in a causal relation to phenomena be 
 admitted, how does that save freedom, if the action in its 
 empirical character is to be regarded as necessary ? The 
 answer is : because the empirical character is itself determi?ied 
 by the intelligible character, of which it is but the sensuous 
 schema. An action, so far as it is to be ascribed to 
 Reason as its cause, does not follow from it according to 
 empirical laws. That is, not the conditions of pure Reason, 
 but only their effects in the phenomenon of the internal 
 sense, precede the act. We cannot say that the causality 
 of Reason in its intelligible character begins to be, as it is 
 not subject to time-conditions at all ; but we may say this. 
 If Reason is causal, it is a faculty through which the 
 sensuous condition of an empirical series of effects first 
 begins ; the rational condition, being non-sensuous, cannot 
 be said to have any beginning. 1 
 
 1 Kant anticipates {Kritik, 254) an objection which might be based 
 upon his theory of the intelligible and empirical character. He holds 
 that the Idea of duty, which can never be found in experience, furnishes 
 us with the a priori law according to which we legislate for our actions, 
 and by which we determine their merit or demerit. Hence every action 
 may be regarded as not only resulting from what preceded it in time, but 
 also from the intelligible character of the man, which itself indeed can 
 never be cognised, but the effects of which are cognised in the empirical 
 character. Supposing, then, that the objector says to Kant : You say 
 that I cannot know the Ego to exist except I make it an object ; and to 
 do this I must have it determined in some way (more than the mere 
 vague subject of the Cogito) ; and you add that all such determination 
 must be sensuous, and so phenomenal. On your own principles you 
 have supplied me with an a priori determination of self through the 
 intelligible character, and so a priori and intellectual ; on this, then I 
 may base a rational psychology. To this Kant answers : It is true 
 that I allow an a priori determination of self through the moral law, 
 
xxv AN INSOLUBLE PROBLEM 325 
 
 And although we can never pro-ve from experience that 
 Reason is causal (for no empirical proof is sufficient for a 
 transcendental proposition), yet our ordinary judgments on 
 the morality of human actions presuppose that it is so. 
 For example, suppose a man tells a lie. We may proceed 
 to examine the empirical character of the offence by tracing 
 it to its sources, e.g. a defective education, bad company, 
 want of reflection, strong temptations, etc. Now, although 
 we believe the action to have been completely determined 
 by all these circumstances, yet we blame the offender all 
 the same. Why ? Because we consider that his Reason 
 is a cause which could have and ought to have made him 
 speak the truth, independently of all empirical conditions : 
 it is completely free, and so, no matter what the tempta- 
 tion, if the action be estimated according to its intelligible 
 character, it must be regarded as blameworthy. That is, 
 we regard the causality of Reason not merely as a co- 
 operating agency, but as complete in itself. It is the 
 permanent condition of all actions, for it does not itself 
 exist in time ; it is, relatively to new states, determining but 
 not determinable. Hence we cannot ask : Why did not 
 Reason determine itself differently ? The question ought 
 to be thus stated : Why has not Reason, through its causal- 
 ity, determined certain phenomena differently ? But this is 
 a question which admits of no answer ; for a different intel- 
 ligible character would have generated a different empirical 
 character. As to why the intelligible character generates 
 such and such phenomena, and exhibits such and such an 
 
 and that the intelligible character does determine us ; but how ? only 
 by producing the empirical character. Though the determining comes 
 from within, is intellectual and a priori, the determination is wholly — 
 so far as we can know it — phenomenal, and subject to the laws and 
 restrictions of phenomena. 
 
326 SCHOPENHAUER CHAP - 
 
 empirical character under certain circumstances, 1 it is be- 
 yond our power to say ; nor is it the question at issue here. 
 
 To sum up : it must be noted with care in the first 
 place that we have not attempted to prove the actuality of 
 Freedom. For such an argument would not be tran- 
 scendental — it would not be limited to the discussion of 
 mere concepts ; and further, it is obvious that we can 
 never infer from experience what cannot be thought in 
 accordance with its laws. Further, we have not even 
 tried to prove the possibility of Freedom ; for we can in 
 general cognise the possibility of no real ground, and of no 
 causality from mere concepts a priori. All that we have 
 attempted is this — to show that nature and freedom are not 
 contradictory, and this we have certainly done. 
 
 IT Fischer thus explains Schopenhauer's position ( Comm. 
 p. 247). 'If the intelligible cause be nothing but a neces- 
 sary will, it must be the Will which must be at the basis of 
 all phenomena — of all representations. And this is the 
 point of the Kantian philosophy from which Schopenhauer 
 deduces his own. The real solution of the cosmological 
 problem, which Kant declares to be insoluble, and therefore 
 avoids, is, according to Schofenhauer, " the world as ivill,"' 
 This is a false inference from Kant, and part of that perverse 
 system of interpretation which would force absolute idealism 
 upon him. He uses the will as illustration of what may 
 not impossibly be the case with all phenomena, but does 
 not say that it is the only intelligible cause. He is always 
 most careful not to assert dogmatically that the noumenon 
 at the basis of external phenomena is identical with that at 
 the basis of internal. That the logical result of Kant's own 
 principles is absolute idealism may be true enough ; but 
 
 1 Cf. for similar questions which cannot be answered, as dealing 
 with ultimate facts, Kritik, pp. 89 and 377. 
 
FOURTH ANTINOMY 327 
 
 most certainly Kant himself believed the reverse. The diffi- 
 culty of the whole discussion arises from the fact that he is 
 so careful not to dogmatise on this point, and that in con- 
 sequence he treats the subject with great generality as a 
 cosmological problem. 
 
 IV. In the case of the fourth Antinomy we have to do 
 not with the unconditioned causality, but with the uncon- 
 ditioned existence of substance (as was before pointed out) ; 
 and consequently the series before us here is not one of 
 intuitions, but of concepts. As in the third Antimony, if 
 phenomena were things in themselves the antithesis would 
 have to be adopted, and a necessary being as condition of 
 the existence of sensuous phenomena could not be arrived 
 at. But, on making this important distinction between 
 phenomena and things in themselves, we may see, as in the 
 last section, that both thesis and antithesis may be regarded 
 as true from different points of view. For the regulative 
 principle of Reason guiding us is, that everything in the 
 world of sense must possess an empirically conditioned 
 existence. But this does not in the least exclude the 
 existence of a non-empirical condition of the whole series, 
 i.e. an unconditioned necessary being, which, as uncon- 
 ditioned, would not itself be a member of the series. And 
 the difference between this solution and that given for 
 Antinomy III. may be stated thus. There the thing itself 1 
 as cause (substantia phenomenon) belonged to the series of 
 conditions, and only its causality was regarded as intelligible ; 
 whereas here the necessary Being must be regarded as quite 
 outside the sensuous series (ens extramundanum) and merely 
 intelligible, for otherwise it would be subject to the 
 phenomenal laws of contingency and dependence. 
 
 As before, we do not attempt to prove the existence or 
 1 Not, of course, the thing in itself. 
 
3 2S INTELLIGIBLE OBJECTS chap. 
 
 even the possibility of such a being ; we only desire to show 
 that the contingency of natural phenomena is quite con- 
 sistent with the assumption of such an intelligible condition 
 of the whole empirical series. Of course, too, the empirical 
 employment of Reason will not be affected by such an 
 assumption ; it will still continue its operations in the field 
 of nature on the hypothesis that everything is contingent. 1 
 
 Concluding Remarks. 
 
 So long as the object of our rational concepts is the 
 totality of conditions in the world of sense, our ideas are 
 transcendental, but yet cosmological. If, however, we place 
 the Unconditioned — with which we have especially to do — 
 outside that world, our ideas become transcendent : for they 
 serve not only for the completion of the empirical use of 
 Reason, but they separate themselves entirely from it, and 
 make for themselves objects, the material of which is not 
 taken from experience, and the objective reality of which is 
 not based on the completion of the empirical series, but on 
 pure concepts a priori. The cosmological Idea correspond- 
 ing to the fourth Antimony is the only one which urges us 
 to take this step ; it requires us to look for an object 
 different from phenomena — an intelligible object — with 
 which all contingency must cease. Now once having 
 admitted, outside experience, a self-subsistent actuality, 
 phenomena can only be regarded as contingent modes of 
 representing intelligible objects employed by beings which 
 are themselves intelligences. But of intelligible objects 
 
 1 The solution then briefly is — Phenomenally, all is contingent ; but 
 still there may belong to the world of sense, as its noumenal ground, an 
 absolutely necessary being. Both thesis and antithesis may thus be 
 considered true. 
 
A NECESSARY BEING 329 
 
 experience can tell us nothing, and so the very first step we 
 have taken beyond the world of sense obliges us to begin 
 with the investigation of the absolutely necessary Being, 
 and to derive from our notions of it the notions of all 
 things in so far as they are intelligible. This we shall 
 attempt in the next chapter. 
 
CHAPTER XXVI 
 
 THE IDEAL OF PURE REASON 
 
 § i. Of the Ideal in general. — By the Ideal I understand, 
 says Kant, the Idea not merely in concreto but in individuo, 
 i.e. an individual thing determinable or even determined by 
 the Idea alone. We have already seen that no Idea can be 
 phenomenally represented in concreto, and it would appear 
 that the Ideal is still farther removed from objective reality. 
 It has a prominent place in Plato's philosophy as an Idea of 
 the divine mind — an individual object present to the pure 
 intuition of the Deity, the most perfect of every kind of 
 possible beings, and the primary ground of all phenomenal 
 copies. But without entering into this, it is plain that there 
 are Ideals in the moral sphere which serve as archetypes, 
 just as there are ideas which afford rules for conduct ; the 
 wise man of the Stoics will occur to every one. And although 
 we cannot ascribe objective reality to these Ideals, yet are 
 they indispensable as standards by which to judge of any 
 action, because Reason requires the concept of that which 
 is perfect in its kind in order to estimate the defects of that 
 which is imperfect. Here we must not confuse Ideals of 
 the Reason with Ideals of Sensibility, or creatures of the 
 imagination : the former are always based upon determinate 
 concepts and serve as models for imitation ; of the latter no 
 
chap, xxvi COMPLETE DETERMINATION 331 
 
 concept can be had, and they are based upon no rule — they 
 can never afford a standard for judgment. In fact, Reason 
 aims at complete determination according to rules a priori , 
 and hence it presupposes an object which must be com- 
 pletely determinable according to Principles, though without 
 the sufficient conditions in experience ; the concept of this 
 object — the Ideal — is therefore transcendent. This we 
 shall now develop in detail. 
 
 § 2. Of the transcendental Ideal {Prototypon transcendent- 
 ale). — Every concept, in relation to that which is not con- 
 tained in it, must be subject to the principle of delermin- 
 ability [Law of Excluded Middle], viz. to the logical 
 Principle that of any two contradictory predicates only one 
 can belong to a concept. But further, every thing, as 
 regards possibility, is subject to the principle of complete 
 determination} viz. that of every possible pair of contradictory 
 predicates one must belong to it. This is no mere analy- 
 tical law, but is based on a transcendental presupposition, 
 viz. that of the totality of all possible predicates — the material 
 for all possibility which must contain a priori the data for 
 this or that particular possibility. We may put the case in 
 this way. Reason always seeks complete knowledge ; to 
 obtain complete knowledge of any particular thing we must 
 know which of every possible contradictory pair of predicates 
 belongs to it ; and so the Idea which guides our procedure 
 must be the Idea of the sum-total of all possibility. Or 
 again, in order to attain to complete knowledge of any 
 particular thing x, we must proceed by disjunctive syllogisms 
 thus, x, is either A or not A ; it is either B or not B ; it is 
 either C or not C ; and so on, until we have exhausted the 
 field of possibility, gone through every possible pair of con- 
 
 1 This, accordingly, is the transcendental parallel to the Law of 
 Excluded Middle (cf. pp. 128 and 159, supra). 
 
332 OMNITUDO REALITAT1S chap. 
 
 tradictory predicates, and decided which of each pair belongs 
 to the object in question. Our guiding idea is plainly 
 A + B + C + . . . the sum-total of all possible positive pre- 
 dicates — i.e. the transcendental Ideal, for it represents an 
 individual object completely determined by and through 
 the mere idea. As all its attributes are positive, it must be 
 one, for it will be thus determined in every respect. Again, 
 as negatives but indicate the absence of reality, this primi- 
 tive concept ( Urbegriff), containing no derived or negative 
 concepts, may be described as the idea of the sum-total of 
 reality (pmnitudo realitatis). All true negatives are then 
 nothing but limitations, a term which could not properly 
 be applied to them did not the unlimited (the all) form 
 the basis of one primitive concept. By this complete pos- 
 session of reality is the concept of a thing in itself repre- 
 sented as completely determined ; and the concept of an 
 ens realissimum is the concept of an individual being, 
 inasmuch as of all possible contradictory predicates, that 
 one which absolutely belongs to being is formed in its 
 determination. 
 
 It is obvious that in this procedure Reason does not 
 presuppose the existence of such a being, but only the Idea 
 of it — for the purpose of deducing from an unconditioned 
 totality of complete determination the conditioned, i.e. the 
 totality of the limited. The Ideal is therefore the prototype 
 of all things, which as imperfect copies or ectypes derive from 
 it the material of their possibility, approaching it more or 
 less nearly, but never reaching it. All the manifoldness of 
 things only comes from the various ways of limiting the 
 concept of the highest reality which forms their common 
 substratum ; just as all figures are only various modes of 
 limiting infinite space. Hence this omnitudo realitatis is 
 is also called the Primal Being (ens originariwii) ; as having 
 
NOT AN AGGREGATE 
 
 no existence superior to it, the Supreme Being (ens suminum) ; 
 and, as everything conditioned stands under it, also the 
 Being of all beings {ens entiuni). 
 
 It should not be described as an aggregate of many other 
 derived beings, for these latter presuppose it, and therefore 
 cannot be said to make it up ; and so it follows that the 
 Ideal must be thought as simple. It is, then, in this view, 
 regarded rather as the ground than as the sum-total of things ; 
 and their manifoldness depends not on any limitation of the 
 primal Being itself, but upon that of the complete series of 
 effects which flow from it. To this latter belongs our whole 
 sensibility, together with all reality in phenomenon, which 
 certainly cannot belong to the Idea of the supreme Being 
 as a constituent element. 
 
 All this, however, does not indicate the objective relation 
 of an actual object to other things, but only of an Idea to 
 concepts, and so leaves us in perfect ignorance of the 
 existence of a Being of such surpassing excellence. If we 
 proceed to hypostatise this Idea, as we inevitably do, by 
 introducing existence into it, we arrive at the concept of 
 God in its transcendental sense, and so the Ideal of pure 
 Reason becomes the object-matter of a transcendental ] 
 Theology. But by such an employment of the Idea we 
 are overstepping the boundaries of its determination and 
 admissibility. Reason merely placed it as the concept of 
 all reality, at the basis of the complete determination of 
 things in general, without requiring that this concept be 
 given objectively. It is necessary to point out the source 
 of dialectical illusion which Reason thus falls into when it 
 not only regards the possibility of all things as deduced 
 
 1 We should naturally expect transcendent here ; but Kant is not very 
 strict in his use of this word in this part of the Krilik. Cf. p. 238, 
 supra. 
 
334 PROGRESS OF ERROR chap. 
 
 from a single possibility (viz. of the highest reality), but 
 also presupposes this as existing in a particular Primal 
 Being. The illusion arises thus. The possibility of objects 
 of sense (as we saw in the Analytic) is a relation of them to 
 our thought, by which something (the empirical form) may 
 be thought a priori ; while that which constitutes the 
 matter (the reality in the phenomenon, corresponding to 
 sensation) must be given from without. 
 
 Such an object is completely determined only when it 
 has been compared with all phenomenal predicates, and 
 represented by means of these either positively or negatively. 
 So then the material of the possibility of all objects of 
 sense must be presupposed as given in one whole, and only 
 on the limitation of this whole can the possibility of em- 
 pirical objects, their difference from each other, and their 
 complete determination, be founded. Hence it follows that 
 nothing can be an object to us unless it presupposes the 
 sum-total of all empirical reality as the condition of its 
 possibility. Now, a natural illusion leads us to consider 
 this principle, which is valid only for objects of sense, as 
 valid with respect to all thi?>gs iti gene?-al ; thus we are led 
 to regard a merely empirical principle of our concepts of 
 the possibility of things as phenomena, as a transcendental 
 principle of the possibility of things in general. And 
 then we proceed afterwards to hypostatise this Idea by 
 changing dialectically the distributive unity of the empirical 
 exercise of the understanding into the collective unity of 
 an empirical whole, and by regarding this whole of phe- 
 nomena as an individual thing containing in itself all 
 empirical reality. 
 
 If Kant says in a note that a further mistake is made in 
 personifying this ideal. ' For the regulative unity of ex- 
 perience is not based upon phenomena themselves (Sensi- 
 
xxvi KANT'S STARTING-rOINT 335 
 
 bility alone), but upon the connection of the manifold by 
 the Understanding (in an apperception), and thus the unity 
 of the supreme reality and the complete determinability 
 (possibility) of all things seem to reside in a Supreme 
 Understanding, and consequently in an Intelligence.' 
 
CHAPTER XXVII 
 
 PROOFS OF THE EXISTENCE OF GOD 
 
 § 3. Reason soon becomes aware of the illusion which it 
 falls into when it regards a mere creation of thought as an 
 actual being, and introduces existence' into the concept 
 of the Ideal ; but nevertheless it is compelled to seek some 
 resting-place in its regress from the given conditioned to 
 the unconditioned. And the following seems to be the 
 natural course of its procedure. If we admit the existence 
 of any one thing, we must also admit that something exists 
 by necessity, for the contingent always presupposes the 
 necessary : in this way we persuade ourselves that some 
 necessary being must exist. We next ask ourselves, ' What 
 is this necessary being?' Which of all our concepts of 
 possible things is most agreeable to the concept of a being 
 which exists unconditionally ? It is the concept of the ens 
 realissimum, the sum-total of all reality, the primal basis of 
 things ; and consequently Reason, in its ordinary course, 
 regards this ens realissimum as the absolutely necessary 
 being which it seeks. 
 
 There is a certain amount of truth in this conclusion, 
 but it is not exact. If we were compelled to make up our 
 mind one way or the other as to the nature of necessary 
 being, we should inevitably decide that it is the ens realissi- 
 
chap, xxvn A PRACTICAL POSTULATE 337 
 
 mum . but we are not so obliged, in a speculative point of 
 view. All we can say is that if a being is the sum-total of 
 all reality it must be necessary, but we have no right to 
 infer that because it is necessary therefore it is identical with 
 the sum of all reality. Even allowing the inference from 
 contingent to necessary existence, it still remains to be 
 proved that the ens summum is the only necessary exist- 
 ence (cf. p. 343, infra). 
 
 In a practical regard, however, we must make up our 
 minds : certain moral obligations lie upon us, which would 
 be obligations without motives except upon the supposition 
 of a Supreme Being to give effect to practical laws ; and 
 that being so, it is natural that we should place the ultimate 
 causality where we place the supreme causality, viz. in that 
 Being which contains originally in itself the sufficient ground 
 of all possible effects. In this way, though in the specula- 
 lative sphere Reason be incompetent to prove the Being of 
 God, yet is this a necessary postulate of Reason in the 
 practical sphere. In speculative proofs 1 we must either try 
 to demonstrate of the most real being that it necessarily 
 exists, or of necessary existence that it constitutes the 
 most real being ; and according as the proof takes its 
 departure from the rational concept of the ens summum 
 or from the empirical concept of conditioned existence, 
 it will be a priori and transcendental or a posteriori 
 and empirical. The empirical proof again may start from 
 two different points : either it begins from the existence 
 we can experience, quite independent of its form and 
 order, or it begins from the consideration of the order 
 of natural existence. The first starting point is the exist- 
 
 1 As Hegel puts it : ' Either we may begin with Being and proceed 
 to the abstraction called Thought ; or the movement may begin with the 
 abstraction and end in Being' (Wallace's Logic of Hegel, p. 86). 
 I Z 
 
338 THREE POSSIBLE PROOFS 
 
 ence of the world (indeterminate experience) ; the second, 
 the existence of the order of the world (determinate ex- 
 perience). 
 
 Rational theology then may put forward three proofs 
 of the Being of God : the transcendental {ontohgical), the 
 cosmological, and physico-theological ; but these three con- 
 stitute all possible proofs. 
 
 § 4. The Ontohgical Proof. — We have seen that the con- 
 cept of an absolutely necessary being is a mere Idea, which 
 serves but to indicate a certain unattainable perfection and 
 rather to bound the Understanding than to extend it to new 
 objects. People have been foolish enough to attempt to 
 prove the existence of this absolutely necessary being, 
 without first considering whether it is even cogitable. It is 
 easy to define it as that of which the non-being is impossible ; 
 but this does not help us much, for what tests of possibility 
 or impossibility can we apply to an abstraction like the 
 Unconditioned ? Some philosophers again have thought to 
 explain this concept of the absohctely ?iecessary by an appeal 
 to examples ; such a proposition, they said, as a triangle 
 has tlu-ee angles sufficiently illustrates its character. But all 
 these pretended examples are taken from judgments, not 
 from things ; and the unconditioned necessity of judgments 
 is not at all identical with the absolute necessity of things. 
 Thus, if in an identical judgment I reject the predicate and 
 retain the subject, there arises a contradiction ; and hence 
 I say, the former belongs necessarily to the latter. But if I 
 reject the subject as well as the predicate no contradiction 
 arises, because there is nothing left which could be con- 
 tradicted. Take the proposition, ' the ens realissimum is 
 absolutely necessary ' ; both subject and predicate may be 
 simultaneously rejected in thought without any contradiction 
 
ACTUAL AND POSSIBLE 339 
 
 However, it has been said l [by Wolff and his school], 
 there is one concept in which the non-being or the removal 
 of the object mould be self-contradictory, and this is the 
 concept of an ens realissimum. For the ens realissimum as 
 containing only positive predicates cannot be self-contra- 
 dictory, and therefore must be possible. It possesses all 
 reality (the argument proceeds), and therefore must include 
 in its concept the notion of being ; hence it exists. In this 
 argument there are two fallacies. (1) The absence of 
 contradiction is a sufficient criterion of the logical possibility 
 of a concept , it is not a sufficient guarantee of the real 
 possibility of a thing. Kant himself has given quite a 
 different test (p. 193, above). (2) All existential judgments 
 are synthetical ; you cannot infer the being of a thing from 
 the mere concept of it. Being is not a real predicate — 
 that is, a predicate which can be added to the concept of 
 a thing ; it merely brings the thing into relation with my 
 concept of it. The synthetical element in an existential 
 judgment is not the enriching the subject with a new 
 predicate, but the bringing the subject into a new relation 
 to my experience. There is no more in a hundred actual 
 dollars than in a hundred possible dollars, as far as 
 the content of the concept goes, though there is a very 
 real practical difference. As far as the pure category is 
 concerned, we have no test by which to distinguish possi- 
 bility from actuality ; and consequently it is idle to attempt 
 to infer the latter from the former. If we were dealing with 
 an object of sense, this difficulty could not arise, nor could 
 such a mistake be made as to confound the existence of a 
 
 1 Cf. for a similar answer Descartes — ' L'existence du triangle ne 
 doit pas etre comparee avec l'existence de Dieu ; parcequ'elle a mani- 
 festement en Dieu une autre relation a l'essence qu'elle n'a pas dans le 
 triangle' (Cousin's Edition, 1824, vol. ii. p. 291). 
 
34Q HEGEL'S CRITICISM 
 
 thing with the concept of it ; for by the concept the object 
 is thought only as in harmony with the general conditions 
 of a possible empirical knowledge, while by its existence it 
 is thought as contained in the context of collective experience. 
 In truth, all our knowledge of existence relates but to the 
 field of experience, and any existence outside that field, 
 though it cannot be declared absolutely impossible, is a 
 hypothesis which we have no means of justifying. The 
 celebrated ontological or Cartesian argument for the existence 
 of God is therefore insufficient ; and we might as well hope 
 to increase our knowledge by the aid of mere ideas, as a 
 merchant to increase his capital by adding a few noughts to 
 his cash account. 
 
 H The popularity of Kant's criticism of this argument, 
 Hegel has remarked, results probably from his homely 
 illustration of the ioo dollars. Every one can see that in 
 the case of the dollars you cannot deduce the being from 
 the mere notion ; but it is important to remember that the 
 illustration is not quite apt. ' The very nature of a finite 
 object is expressed by saying that its Being in time and 
 space is discrepant from its notion. God, on the contrary, 
 ought to be what can only be "thought as existing"; His 
 Notion involves Being. It is this unity of the Notion and 
 Being that constitutes the notion of God." 1 What Kant 
 has shown is that on the supposition that Sensibility is 
 different in source from Understanding, you cannot infer 
 existence in space and time from a mere concept. But 
 Hegel saw that this supposed difference in source was a 
 
 1 Wallace's Logic of Hegel, p. 92. Kant points out very clearly 
 {Kritik der Urtlteilskraft, pp. 414, 415, Hartenstein's edition, 1867) the 
 connection between his doctrine of Possibility and Actuality, and his 
 theory as to the difference in source of Sensibility and Understanding. 
 The whole passage deserves careful study. 
 
KANT'S DEFECT 341 
 
 fiction ; Sensibility as well as Understanding is but a phase 
 of Thought, and so Kant's laborious argumentation here is 
 not worth much. Kant's great merit was to point out the 
 necessary and a priori element in the Sensibility ; his great 
 defect was to isolate it from and oppose it to the Under- 
 standing. Had Sensibility and Understanding a common 
 root (as he hints in some passages), the distinction between 
 analytical and synthetical propositions would be only one 
 of degree, and his criticisms would have no force. We have 
 not had special occasion to remark this before, but unless 
 this be seen, the whole drift of the Dialectic is lost ; and, 
 we will add, the position which Kant occupies in the his- 
 tory of Philosophy cannot be appreciated. 
 
 § 5. The Cosmological Proof proceeds in a much more 
 natural way than the ontological, the subtlety of which is 
 due to the scholastic philosophy [Anselm] ; it infers from 
 the unconditioned necessity of some being its unbounded 
 reality. It was called by Leibnitz the argumentinn a con- 
 tingentia mundi, and may be stated as follows. If anything 
 exists, an absolutely necessary being must exist. But I 
 myself exist ; therefore there exists an absolutely necessary 
 being. And this necessary being must be the ens realis- 
 simum, for such a being alone rests on itself and has all the 
 conditions of its existence in itself; the ens realissimum 
 alone can be completely determined a priori through its 
 concept. 
 
 This proof (called cosmological, because the object of all 
 possible experience is the world) begins in the minor 
 premise with experience, unlike the ontological proof; and 
 inasmuch as it takes no account of any peculiar property of 
 objects of sense by which this world may be distinguished 
 from other possible worlds, it differs also from the physico- 
 theological proof. We may readily see that we have an old 
 
342 FALSE ASSUMPTIONS 
 
 sophism here in a new dress, for the only use the cosmo- 
 logical proof makes of experience is in the first step, to tell 
 us that a necessary being exists ; but experience cannot tell 
 us what are the properties of this necessary being, nor can 
 it identify it with the ens realissimum. Indeed, when we 
 go into details we find in this argument a perfect nest of 
 dialectical assumptions, as follows : — 
 
 (i) The transcendental principle — everything that is 
 contingent must have a cause, which is the basis of the 
 major premise in the syllogism at the beginning of the 
 proof — has no significance except in the world of sense. We 
 have before pointed out (p. 298) that the intellectual concept 
 of contingency (which is all that we can here apply) is simply 
 that of which the non-existence is possible ; but the con- 
 tingent, in this sense, we have no right to refer necessarily 
 to a cause. 
 
 (2) There is involved in the proof the inference to a 
 first cause from the impossibility of an infinite series of 
 causes being presented one after another in the world of 
 sense. This is an inference which Reason does not permit 
 us to employ as a principle even within experience, still 
 less to extend it beyond experience where there is no chain 
 of causes at all. [In fact the error committed is the 
 same as that which is apparent in the thesis of the 4th 
 Antinomy.] 
 
 (3) Supposing this series of conditions to be completed, 
 this could never be done by a being which itself lies totally 
 beyond and apart from the senses ; the ens realissimum does 
 not complete the series, being separated therefrom by an 
 impassable gulf. 
 
 (4) The whole procedure is an ignoratio elenchi, because 
 it brings us back to, and assumes the validity of, the onto- 
 logical argument, which it professes to avoid. For, passing 
 
xxvn COSMOLOGICAL ARGUMENT 343 
 
 to the second stage of the proof, what right have you to say 
 that the absolutely necessary being arrived at must be the 
 ens realissimum ? To put it strictly, you have assumed that 
 all absolutely necessary beings are entia rcalissima ; but no 
 ens realissimum is different from any other — what is valid of 
 some is valid of all. Hence the assumption really is, in 
 a converted form, that all entia realissima are necessary 
 beings. But this was the very crux of the ontological 
 argument, and you are here again guilty of the old blunder 
 exposed in § 4. 
 
 ' The logical possibility of a concept of all united reality 
 (the test of which possibility is the absence of contradiction) 
 is confounded with the transcendental, which requires a 
 principle for the practicability of such a synthesis ; but such 
 a principle is applicable to the field of possible experiences 
 only.' The cosmological argument thus only tries to evade 
 by a mere trick the proof of the existence of a necessary 
 Being a priori by mere concepts. 
 
 The truth is that though it is quite allowable to admit 
 the existence of an all-sufficient Being, a cause of all pos- 
 sible effects, in order to assist Reason in the quest after 
 unity, yet we cannot assert that such a Being exists neces- 
 sarily. This unconditioned necessity which we require 
 as the ultimate support of all things, is the true abyss 
 of human Reason. Eternity itself, terrible and sublime 
 as it is, does not produce the same giddy impression, for 
 it only measures the duration of things, but does not sup- 
 port them. 
 
 Many powers of nature, says Kant in conclusion, which 
 manifest their existence by certain effects, remain for us in- 
 scrutable ; they evade our observation. The transcendental 
 object which forms the basis of phenomena, and, in connec- 
 tion with it, the reason why our Sensibility possesses these 
 
344 TWO REGULATIVE PRINCIPLES CHAP. 
 
 rather than any other ultimate conditions, remain for us 
 inscrutable. But an Ideal of pure Reason cannot be termed 
 inscrutable ; because it has to adduce no credential of its 
 reality further than the need felt by Reason to perfect all 
 synthetical unity by means of it. It is not given as a 
 thinkable object ; being a mere Idea it must find in the 
 nature of Reason its place and its solution ; for the very 
 essence of Reason consists in its ability to give an account 
 of all our concepts, opinions, and assertions, 1 — either 
 as objective, or, if they are illusory, on subjective 
 grounds. 
 
 Explanation of the dialectical illusion in these tran- 
 scendental proofs. We naturally inquire next, What is the 
 cause of this dialectical illusion by which we connect the 
 concepts of necessity and supreme reality, and hypostatise 
 that which can only be an Idea ? First of all, it is a per- 
 fectly natural (though as we have seen not a reliable) 
 procedure to infer that if anything exists, something must 
 exist necessarily; and it is on this inference that the 
 cosmological argument rests. But again I find that I can- 
 not think the existence of any individual thing as necessary; 
 nothing prevents me from thinking its non-existence. In 
 other words, I cannot complete the regress to the conditions 
 of existence, without admitting a necessary Being, but I can 
 never begin from such a Being. Now if (i) I must think 
 something necessary as the basis of all existing things, and 
 yet (2) am not justified in thinking of anything as in itself 
 necessary, the inevitable conclusion is that necessity and 
 contingency do not concern things themselves; otherwise 
 these principles would contradict one another. It is plain 
 indeed that they cannot be regarded as objective; for if all 
 that we perceive in things must be regarded as conditionally 
 1 Cf. p. 306, sii/ra. 
 
MISTAKES 345 
 
 necessary, it is impossible that anything which is empirically 
 given should be absolutely necessary. 
 
 It follows from this that the absolutely necessary Being 
 (which we require by our first rule) must be regarded as outside 
 the world, inasmuch as it is useful merely as a Principle of 
 the highest possible unity of phenomena, and we cannot 
 meet with any such necessary existence in the world (by our 
 second rule). As an example of the mistakes that have 
 been made by not attending to both these rules, it may be 
 noted that the philosophers of antiquity regarded all form 
 in nature as contingent, but matter as original and necessary. 
 Now they only attended to the first regulative principle, — 
 they overlooked the second ; if they had regarded matter, 
 not relatively as the basis of phenomena, but according to 
 its being — in itself — they would have seen at once that it 
 was not absolutely necessary ; it can be annihilated in 
 thought without any contradiction — every one of the real 
 properties of matter, being derived, are only conditionally 
 necessary. Matter (and generally all that belongs to the 
 world) is not congruous with the Idea of a necessary primal 
 Being, even as a Principle of the greatest possible empirical 
 unity ; this Being, as we have shown by our second regulative 
 principle, must have its place assigned outside the world. 
 
 These remarks will make it plain that the Ideal of a 
 Supreme Being is but a regulative Principle of Reason, requir- 
 ing us to regard all means in the world as if it had its origin 
 from an all-sufficient necessary cause, in order to found upon 
 this the rule of a systematic and necessary unity, according 
 to general laws in the explanation of phenomena. By a 
 transcendental subreptio we regard this formal Principle as 
 constitutive, and hypostatise this unity. We make a similar 
 mistake as to space. Space originally makes possible all 
 shapes or figures, which are properly but different limita- 
 
346 THE DESIGN ARGUMENT 
 
 tions of it ; and thus, although it is only a Principle of 
 Sensibility, we take it for an absolutely necessary, self- 
 subsisting thing, for an object given a priori in itself. It 
 is in this way that Ideas are regarded as actual objects ; and 
 principles, only regulative, are regarded as if they were 
 constitutive. 
 
 § 6. The physico-theological proof. — Since then neither 
 the concept of things in general, nor the experience of any 
 one being in general, can give us what we require, it remains 
 to try whether a determinate experience, viz. the things of the 
 present world, their constitution and order, may not enable 
 us to establish the existence of a Supreme Being. We can 
 hardly expect, indeed, that any satisfactory proof will be 
 thus afforded, as we have seen so clearly in the preceding 
 sections that the peculiarity of an Idea consists in the fact 
 that no experience can be adequate to it. 1 If the Supreme 
 Being lies within the empirical series, it must itself be con- 
 ditioned ; if it be supposed outside the series, how are we 
 going to bridge over the chasm between it and experience? 
 But, inasmuch as everywhere around us we see a chain of 
 causes and effects, of means and ends, we are inevitably 
 led up to a Supreme Being, which, as the cause of the 
 origin of the world, may also secure its permanence. To 
 this, then, we attribute a degree of perfection beyond every- 
 thing else that is possible ; which we can easily do, though 
 only in the faint outline of an abstract concept, if we repre- 
 sent to ourselves all possible perfections united in it as in 
 one substance. Such a concept agrees with the Law of 
 Parcimony, is not self- contradictory, helps to extend the 
 empirical use of Reason, and does not conflict with any 
 experience we have. Now this proof will always deserve 
 to be treated with respect ; it is the oldest, the clearest, 
 1 CI Proleg. p. 1 1 6. 
 
xxvii MOMENTA OF PROOF 347 
 
 and most in accordance with human Reason ; but never- 
 theless it is not demonstrative, as we shall see if we detail 
 the chief points in it. 
 
 It may be divided into four steps, (i) There are in 
 the world clear signs of an arrangement full of definite 
 purpose, carried out with great wisdom, existing in a whole 
 indescribably complex in content and unbounded in extent. 
 (2) This end-adaptation (Zweckmassigkeii) is quite extrane- 
 ous to the things of the world, and belongs to them only 
 contingently ; that is, the nature of different things could 
 not through such a variety of means conduce to the same 
 ends, unless these means had been specially selected and 
 adapted thereto by a rational disposing Principle, in accord- 
 ance with fundamental Ideas. 1 (3) There exists therefore a 
 sublime and wise cause (or more than one), which must be 
 the cause of the world, not merely as a blind and all- 
 powerful Nature through unconscious fecundity, but as an 
 intelligence by freedom. (4) The unity of this cause may 
 be inferred from the unity of the mutual relation of the 
 parts of the world as portions of an artistic edifice — with 
 certainty, as far as our observation reaches, with probability 
 even beyond that range by all laws of analogy. 
 
 In this argument we assume plainly an analogy between 
 certain products of nature and those of human art; but 
 since these latter are the only ones of which we know com- 
 pletely both the cause and the way in which it produces 
 its effect, no other course is open to us. Reason would 
 have no right to pass over a causality which it knows, for 
 obscure and undemonstrable principles of explanation, 
 which it does not know. But there are two vital flaws in 
 the proof. In the first place, the end-adaptation and 
 harmony of so many works of nature proves contingency 
 1 Cf. all through Kritik der Urtheilskraft, pp. 379. 4 18. 
 
34§ CRITICISMS 
 
 merely in the form, not in the matter of the world. And 
 so the proof at most can give us a world-architect, who is 
 very much limited by the capabilities of his material ; it 
 does not give us a world-creator, to whose Idea everything 
 is subject. And so it fails to establish what it aims at — an 
 all-sufficient Original Being; to demonstrate the contingency 
 of matter, we must have recourse to a transcendental 
 argument, which is the very thing we have been trying to 
 avoid. So that, when analysed, the physico- theological 
 proof depends on the cosmological, just as we saw that this 
 last depended on the ontological. And, in the second 
 place, we infer from the order and end-adaptation visible in 
 the world, the existence of a cause proportionate thereto ; 
 but how can we vindicate an argument from finite order 
 and good, to infi?iite wisdom and goodness ? No one will 
 be bold enough to say that he has such a perfect insight 
 into the relation between the magnitude, order, and unity 
 of the world of observation and an Omnipotent, all-wise, 
 Individual Being. 1 Physico-theology can therefore give no 
 definite concept of the supreme world -cause, and so is in- 
 sufficient to serve as the basis of religion. The attainment 
 of absolute totality is thus quite impossible on the empirical 
 path : the ontological proof, through pure concepts of 
 Reason, is the only possible one, if any proof of a proposi- 
 tion so far transcending the empirical use of the under- 
 standing be possible at all. 2 
 
 § 7. Kritik of all Speculative Theology. — The different 
 
 1 Cf. Theory of Ethics, p. 236. 
 
 2 These criticisms on the cosmological and physico-theological proofs 
 both rest, it must be observed, on the supposed insufficiency of the 
 ontological argument. If that be admitted (and Hegel has shown that 
 Kant's objection to it is not sound), the other two proofs appear in 
 quite a different aspect. Cf. Wallace's Logic of Hegel, pp. 87-88, for 
 some instructive remarks on this subject. 
 
DEISM AND THEISM 349 
 
 methods of procedure we have been considering may be 
 classified in the following way. Leaving out of account 
 that theology which is based upon revelation, Rational 
 Theology may cogitate its object either by pure Reason 
 and through transcendental concepts only, as the e/is 
 originarium, ens realissimtim, ens entium ; or by a concept 
 borrowed from our own mind, as the Supreme Intelligence. 
 A man who adopts the former method (Transcendental 
 Theology) is called a Deist ; he holds that we may know 
 the existence of an Original Being by mere Reason, but 
 that our concept of it is transcendental only, as of a Being 
 which possesses all reality, but a reality which cannot be 
 further determined. He who adopts the latter method 
 [Natural Theology), on the other hand, is a Theist ; such an 
 one maintains that Reason is capable of determining its 
 object more closely in analogy with nature, namely as a 
 Being which through understanding and freedom contains 
 in itself the original ground of all other things. The Deist 
 admits a cause of the world only (whether through the 
 necessity of its nature or though freedom remains un- 
 decided) ; the Theist admits an author of the world. 
 Briefly, the Deist believes in a God, the Theist in a living 
 God. According as the transcendental theologian adopts 
 the Cosmological or Ontological line of argument, his science 
 will be (a) Cosmotheology or (b) Ontotheology. A twofold 
 course is also open to the natural theologian, according as 
 he looks at the Supreme Intelligence as the principle either 
 of physical or moral order and perfection ; in the one case 
 his science is (c) Physico-Theology, in the other (d) Moral 
 Theology. 1 Now Kant claims to have shown the impossi- 
 
 1 Not theological ethics; for in that science moral \a.v/s presuppose 
 the existence of a Supreme Governor of the World, while Moral 
 Theology on the other hand grounds this upon moral laws. 
 
350 VALUK OK TI11CKK 1'KOOKS c.-iiap. xxvii 
 
 bilily of (a), (I/), and (c) in the preceding discussion, and so 
 his conclusion is that Rational Theology can have no 
 existence unless it be founded upon Moral Laws (d). lie 
 does not develop this line of argument here, but refers to 
 his Kritik of the Practical Reason for its elucidation. The 
 spec ulative proofs are, however, not without their use. The 
 Design argument, for example, though insufficient of itself, 
 yet pre] tares the Understanding for theological knowledge 
 and gives to it the right direction. 
 
 The transcendental arguments likewise are useful in a 
 negative aspect ; they correct our concept of a Supreme 
 Being ; they show that all phenomenal elements must be 
 eliminated from it, and they put an end to all opposing 
 assertions, be they atheistic, deistic, or anthropomorphic. It 
 is plain that the same arguments which show the inability 
 of Reason to demonstrate the existence of a Supreme Being, 
 must be alike sufficient to prove the invalidity of its denial. 
 
 This Supreme Being is then for purely speculative 
 Reason a mere Ideal, but still a perfectly faultless one, which 
 completes and crowns the whole of human knowledge. And 
 if moral theology is able to supply what is deficient in specu- 
 lation, then will transcendental theology be no longer merely 
 problematic, but indispensable in the determination of the 
 concept with which we are concerned. Necessity, infinity, 
 unity, existence apart from the world (not as a world-soul), 
 eternity as free from time conditions, omnipresence as un- 
 affected by space conditions, etc., are transcendental predi- 
 cates ; and their purified concepts, needful for every system 
 of theology, can be derived from transcendental theology 
 alone. 
 
CHAPTER XXV111 
 
 THK REl'.l'l A 11VE VSK OF THE IDEAS 
 
 The Dialectic has now confirmed the truth of the negative 
 concluson of the Analytic, viz. that all inferences which 
 would lead us beyond the field of experience are fallacious 
 and groundless ; but it teaches us also that Reason has a 
 natural inclination to overstep these bounds, and that Ideas 
 are as natural to it as categories to the Understanding. 
 There is this difference, however, that the latter lead us to 
 truth, or the agreement of our concepts with their objects, 
 while the former produce merely an irresistible illusion, 
 which only the severest Kritik can remove. Still the Ideas 
 must have their own proper use, if we could only discover 
 what it is, for everything grounded in the nature of our 
 faculties must have some purpose. 1 When we mistake them 
 for concepts of actual objects, they become transcendent in 
 their application, and deceive ; but they probably have, too, 
 an immanent use, when they will be directed to the employ- 
 ment of the Understanding in the sphere of experience. 
 Thus all errors of siibreftio, of misapplication, are to be 
 attributed to defects of Judgment and not to Understanding 
 or Reason. 
 
 It was before pointed out (page 242, su/ru) that the 
 1 Of. ]■>. j„;S. «<//.;. 
 
352 PROBLEMATICAL METHOD chap. 
 
 Reason is related to the Understanding very much as the 
 latter is to sensibility. As the Understanding brings unity 
 into the manifold of objects by means of concepts, so the 
 Reason brings unity into the manifold of concepts by means 
 of Ideas, making a certain collective unity the aim of the 
 operations of the Understanding, which otherwise is occupied 
 with a distributive unity alone. Accordingly transcendental 
 Ideas should not be employed as constitutive ; they are not 
 concepts of actual objects ; when thus regarded they become 
 dialectical. But they have an admirable and indispensably 
 necessary regulative use in directing the Understanding to 
 a certain aim towards which all its rules converge. 
 An idea is like the imaginary focus of a pencil of rays, 
 which, though not a point from which they actually pro- 
 ceed, is regarded as such by a natural illusion ; it helps 
 us, however, to systematise our knowledge, to explain it 
 according to one principle, which always is the proper 
 function of Reason. It will make this matter clearer to 
 point out that there are two methods by which Reason may 
 proceed, the demonstrative and the hypothetical. In the 
 former case our universal is certain and we deduce by 
 syllogism our particular from it ; in the latter case the 
 universal with which we start is problematical only, a mere 
 Idea, while the particular is certain. The problem then is 
 to establish the universality of our problematical rule by 
 an examination of particular cases ; and if all particular 
 cases which can be produced follow from the rule, we may 
 conclude it to be universal. 1 But it is plain that we 
 
 1 This principle Kant lays down in his Introd. to Logic (Abbott's 
 Translation), p. 43 : ' If all the consequences of a cognition are true, 
 then the cognition itself is true. For if there were any falsity in the 
 cognition, some false consequence must needs exist.' Cf. also p. 75. It 
 is involved in his transcendental proof {oi. p. 37S, infra). 
 
chai\ xxvin LAW OF PARCIMONY 353 
 
 can never know all possible cases that may arise, and so 
 this hypothetical method of procedure does not lead to the 
 absolute universality of the rule ; this use of Reason is 
 merely regulative, intended to bring particular cognitions 
 into unity as far as possible, and such unity is the criterion 
 of the truth of the rule. From the above considerations 
 we can see that we can only assert this systematic unity of 
 cognitions to be a logical principle aiding the Under- 
 standing ; it can by no means be regarded as a transcend- 
 ental principle of Reason. But again, a logical law of 
 this sort would be devoid of meaning or application without 
 a transcendental principle at its basis ; otherwise Reason 
 would be proposing as its aim an Idea contrary to the con- 
 stitution of nature, viz. the Idea of the unity of the funda- 
 mental powers of nature. This can be best explained by 
 going into some detail. 
 
 1. The Principle of Homogeneity. — In order to reach 
 unity science must constantly unify its concepts — must 
 seek what is homogeneous in them and class it as their 
 higher genus. Hence we have as the logical law of genera, 
 Entia non sunt multiplicanda prceter necessitatem — the 
 well-known Law of Parcimony. The scholastic philo- 
 sophers erred in setting this up as a law of nature, of 
 things. The cause of the mistake was this, that the law is 
 not merely a subjective, economical plan of Reason adopted 
 to save trouble ; for Reason does not merely suggest the 
 rule, but commands the Understanding to follow it, although 
 we are quite unable to determine the proper bounds of the 
 unity sought. And further, if there existed among pheno- 
 mena so great a diversity of content that the Understanding 
 could not discover the slightest similarity between them, the 
 law would be inapplicable though commanded by Reason. 
 Hence, then, it is more than a logical law ; it presupposes a 
 
 1 2 A 
 
354 SPECIFICATION CHAP. 
 
 transcendental principle, not constitutive but regulative, ac- 
 cording to which homogeneity is necessarily presupposed 
 because without it no empirical concepts, and consequently 
 no experience, would be possible. 
 
 2. The Principle of Specification. — To the logical prin- 
 ciple of genera, which postulates identity, corresponds 
 another — that of species, which requires manifoldness and 
 diversity in things, notwithstanding their accordance in 
 the same genus. Reason here forbids the Understanding 
 to consider any species or sub-species as the lowest possible, 
 Entium varietates non temere sunt minicendoz. And this 
 logical law, like the last, must be based upon a transcend- 
 ental law of specification ; it cannot be deduced from 
 experience, for empirical specification very soon stops in 
 the distinction of diversities. But, as before, this transcend- 
 ental principle is not constitutive ; it does not require an 
 actual infinity of variety in the things that are to become 
 objects ; it is only regulative, prescribing to the Under- 
 standing x the duty of looking for sub-species under every 
 species, and minor varieties under every variety. 
 
 This Principle bids us notice differences, as that of 
 homogeneity bade us notice agreements ; and Kant notices 
 that, according as a man is an empiricist given to analysis, 
 or a speculator given to synthesis, he will be inclined to 
 lay stress on one or other of them. The one brings 
 out the faculty which corresponds to wit ; the other that 
 which we call judgment? 
 
 3. The Principle of Continuity of Forms. — This law of 
 
 1 Mr. Caird puts this well. ' Conception can never be adequate to 
 perception, although it must continually strive to make itself so. We 
 can never define the individual, yet the individual is the end which in 
 all definition we strive to reach.' — Phil, of Kant, p. 656. 
 
 2 Cf. Locke, Essay, ii. xi. 2. 
 
CONTINUITY 355 
 
 the affinity of all concepts requires a continuous transi- 
 tion from every species to every other by a gradual in- 
 crease of difference ; intermediate species are always to be 
 regarded as possible. Datur continuum formarum. This 
 unites the former principles by prescribing homogeneity in 
 the greatest variety, through the gradual transition from 
 one species to another : thus indicating a kind of relation- 
 ship of the different branches, as having all sprung from 
 the same stem. And as in this way there arises no void 
 in the whole extent of possible concepts, we may say : Non 
 datur vacuum formarum. As in the other cases, this logical 
 law of continuity presupposes a transcendental principle at 
 its basis (lex continui in naturd) ; and it is equally plain 
 that such continuity of forms is a mere Idea. For (i) the 
 species in nature are actually divided and form quanta 
 discreta, and (2) we cannot make any definite empirical 
 use of the law, as not the smallest criterion of affinity is 
 indicated by it to tell us how far we ought to pursue the 
 gradation of differences. 
 
 If we were to arrange these principles of systematic unity 
 in the order required for their empirical use, they would 
 stand thus : Variety, Affinity, and Unity. The remarkable 
 feature about them all is that they seem to be transcend- 
 ental ; and although they contain for the guidance of the 
 empirical use of Reason mere Ideas which Reason can only 
 approach as it were asymptotically, 1 yet they possess as syn- 
 thetical a priori propositions a more than subjective though 
 undefined validity, serving as rules for possible experience. 
 And also in the elaboration of experience they are of service 
 as heuristic' 1 ' principles; but with all this a transcendental 
 
 1 I.e. it never arrives at them, though always approaching them, as 
 a curve does its asymptote. 
 
 2 I.e. they enable us to ask a question, but do not supply the answer. 
 
356 ANA LOG ON TO THE SCHEMA chap. 
 
 deduction x of them in the sense already known to the 
 reader is quite impossible. 
 
 A difficulty occurs here. In our Analytic we distin- 
 guished (p. 173) the dynamical Principles of the Under- 
 standing, which are regulative principles of intuition, from 
 the mathematical, which are constitutive of intuition. These 
 laws are, however, all constitutive in relation to experience, 
 because they make the concepts without which experience 
 could not exist a priori possible. But our present prin- 
 ciples of pure Reason cannot be constitutive even in 
 regard to empirical concepts, because no sensuous schema 
 corresponding to them can be given, and therefore they 
 can have no object in concreto. How then can they be 
 regarded as having objective validity at all ? The answer 
 is to be found in the fact that the Idea serves as an 
 analogon to the sensuous schema, the difference being 
 that the application of the categories to the schema of 
 Reason is not a knowledge of the object itself (as in the 
 application of the categories to sensuous schemata), but 
 only a rule or principle for the systematic unity of all 
 employment of the Understanding. 2 We shall, in short, 
 
 1 But cf. next page. 
 
 J Cf. Theory of Ethics, p. 160. ' The physical law, being a law to 
 which the objects of sensible intuition, as such, are subject, must have a 
 schema corresponding to it — that is, a general procedure of the imagina- 
 tion (by which it exhibits a prion to the senses the pure concept of 
 the Understanding which the law determines). But the law of freedom 
 (that is, of a causality not subject to sensible conditions), and conse- 
 quently the concept of the unconditionally-good, cannot have any intui- 
 tion, nor consequently any schema supplied to it for the purpose of its 
 application in concreto. Consequently the moral law has no faculty but 
 the Understanding to aid its application to physical objects (not the 
 imagination); and the Understanding, for the purposes of the Judgment, 
 can provide for an Idea of the Reason, not a schema of the Sensibility, 
 but a law, though only as to its form as law ; such a law, however, as can 
 be exhibited in concreto in objects of the senses, and therefore a law of 
 
xxvm DEDUCTION OF IDEAS 357 
 
 best describe these principles by calling them maxims? i.e. 
 they are ' subjective principles, not derived from the con- 
 stitution of the object, but from the interest which Reason 
 takes in a certain possible completeness of our knowledge 
 of the object;' and we shall thus avoid the contradictions 
 which must arise if we regard them as objectively true in 
 any other sense. 
 
 Of the ultimate aim of the natural dialectic of human 
 reason. — The Ideas cannot, as we have seen, admit of the 
 same kind of deduction as the categories ; but if they are 
 to have objective validity and be anything more than 
 empty creations of thought, some kind of justification of 
 them must be possible. And this, their transcendental deduc- 
 tion [in a new and limited sense], may be said to consist in 
 showing that they produce systematic unity in the laws of 
 the empirical use of Reason. Except in the case of the Cos- 
 mological Ideas (which are antinomial), there is, of course, 
 nothing to prevent us from admitting the possibility of 
 their being objective ; but nevertheless all that we can 
 assert for them is, that they are schemata of the regulative 
 principle of the systematic unity of all cognition. Thus 
 the transcendental and only determinate concept of God 
 presented to us by speculative reason is in the narrowest 
 sense deistic , even if we admit a Divine Being we can 
 have no concept of the internal possibility of His supreme 
 perfection or of the necessity of His existence. But it is to 
 be observed that I may admit a thing relatively without 
 being obliged to do so absolutely ; for example, I look upon 
 all connections in the world as if they were ordered by a 
 
 nature. We can therefore call this law the Type of the moral law.' 
 Mr. Caird (Philosophy of Kant, p. 659) gives another illustration of this 
 schematism of the Reason — from the idea of a practical maximum. 
 1 Cf. Theory of Ethics, p. 105. 
 
358 IGNAVA RATIO 
 
 supreme reason, and I then may represent to myself that 
 Supreme Being through concepts, which, properly speaking, 
 are only applicable to the world of sense. For inasmuch 
 as I make only a relative use of my transcendental hypo- 
 thesis, as the substratum of the greatest possible unity of 
 experience, I am quite justified in representing it so ; 
 though absolutely, as existing by itself, I can apply no 
 sensuous predicates to it at all. 1 
 
 The highest formal unity, which is based upon natural 
 concepts alone, is the unity of purpose; and the speculative 
 interest of Reason forces us to regard all order in the world 
 as if it originated from the design of a Supreme Intelligence. 
 This principle opens out to Reason in the field of experience 
 new prospects, and invites it to unite the things of the world 
 according to teleological laws. If we keep to this as a 
 regulative hypothesis, we cannot go very far wrong ; error 
 can have no worse result than that, where we expected to 
 discover a teleological connection, only a mechanical or 
 physical connection appears. But if we employ the Idea of 
 a Supreme Being as a constitutive principle, instead of 
 restricting it to regulative influence, we shall be liable to 
 make serious mistakes. One error which will naturally 
 arise is that of ignava ratio ; reason will indolently refer 
 physical phenomena immediately to the Will of God with- 
 out taking the trouble of investigating their physical causes. 
 Another error into which we are likely to fall in such case 
 is that of perversa ratio ; we shall be prone to begin at the 
 wrong end, to give an anthropomorphic determination to 
 the concept of a Supreme Intelligence and then proceed 
 to impose aims upon nature. That is, by a sort of a 
 va-repov irporepov, we shall very probably be tempted to 
 distort empirical facts until they accord with the results of 
 1 Cf. Proleg. p. 162. 
 
xxvin CRITICAL ANSWERS 359 
 
 our a priori reasonings. Against such mistakes we shall be 
 guarded by this Kritik of Theology. 
 
 In conclusion, as we have frequently said that it is 
 always possible to answer the questions which pure Reason 
 raises, it may be well to point out the critical answers to a 
 few questions which may naturally occur. Suppose it be 
 asked (1), Whether there is something different from the 
 world, which contains the ground of cosmical order and 
 connection according to general laws ? our answer is, 
 Certainly. For the world is a sum of phenomena ; there 
 must therefore be some transcendental basis of these 
 phenomena — that is, a basis thinkable by the pure Under- 
 standing alone. If (2) we are asked whether that Being is 
 a substance, of the greatest reality, necessary, etc., our 
 answer is, This question is without meaning, for I cannot 
 apply the categories beyond the world of sense. If (3) 
 the question be whether we may not think this Being 
 according to the analogy of the objects of experience, the 
 answer is, Undoubtedly, but only as an Idea and not in 
 reality. Again it will be asked (4), Can we then admit a 
 wise and omnipotent Author of the world ? Without doubt, 
 and not only we can but we must. But (5) do we not 
 thus extend our knowledge beyond the field of possible ex- 
 perience? By no means, for we have only presupposed a 
 something of which we have no conception whatever as to 
 what is by itself; the idea is valid only relatively to the use 
 of our Reason in the world. However (6), Can I make any 
 use of this concept in the rational investigation of nature ? 
 Yes; that is its very purpose ; but you must be careful, 
 when regarding physical processes as the result of design on 
 the part of the Supreme Being, to remember that you have 
 no right to regard this last idea as anything more than 
 regulative. When using these analogies you must be always 
 
360 VALUE OF KANT chap, xxviii 
 
 aware that they are for you nothing more than analogies. 
 A certain dim consciousness of this true use of the idea, 
 says Kant, seems to have dictated to the philosophers of 
 all times the moderate lauguage used by them concerning 
 the cause of the world; we find them employing the 
 expressions divine wisdom and wisdom of nature indif- 
 ferently — nay, rather preferring to use the latter. 
 
 IT So then the conclusion of the whole matter for Kant is 
 that Reason contains only regulative principles ; illusion 
 arises through regarding them as constitutive. It is not 
 hard to see that this negative result is the only possible one 
 for him ; his rrpdrov rpevSos in the complete isolation of the 
 Sensibility and Understanding prevented him from rising to 
 the conception of an intuitive understanding on the one 
 hand, and on the other from reckoning seriously with the 
 capacity of pure thought to supply us with knowledge 
 which must be as true as anything we can attain. But the 
 object of this commentary being to expound, not to criticise, 
 we leave the Dialectic without further remark. The value 
 of the study of Kant as a propaedeutic, and this is all he 
 claims for himself, will remain even though we do not go 
 with him in all his conclusions. The service he did in the 
 past by clearing away the cobwebs of the Wolffian Metaphysic 
 can hardly be appreciated by us who, in this century, have 
 entered into his labours, and with difficulty understand what 
 attraction that Metaphysic had for the men of his time. 
 
CHAPTER XXIX 
 
 TRANSCENDENTAL DOCTRINE OF METHOD 
 
 If we regard the whole cognition of pure speculative reason 
 as an edifice, of which we have at least a general notion, it 
 may be said that in our Elementology we have examined 
 the materials and determined to what edifice they belong, 
 and what its height and stability. Our present task in 
 the Methodology relates not to the materials, but to the 
 plan of the building. So then we understand by ' Tran- 
 scendental Doctrine of Method ' the determination of the 
 formal conditions of a complete system of pure reason, 
 and we shall treat successively of its Discipline, its Canon, 
 its Architectonic, and its History, and this from the tran- 
 scendental point of view. Such a task has before been 
 attempted under the head of practical logic in vain, for 
 general logic, 1 not being limited either to any particular 
 kind of cognition or to any particular objects, can give 
 merely the heads of possible methods for the several sciences, 
 and employs a technical terminology in each case, the 
 meaning of which the pupil can only learn subsequently. 
 
 The Discipline of Pure Reason 
 We have seen before that the peculiar province of nega- 
 1 Cf. p. 72, supra. 
 
362 DISCIPLINE NECESSARY chap. 
 
 tive judgments is to prevent error, and therefore they are of 
 especial use in cases where the tendency to form judgments 
 transcending possible experience is great, and the error 
 resulting from such illusion dangerous. The restraint which 
 is employed to repress, and finally to extirpate, the constant 
 inclination to depart from certain rules is called discipline : 
 it is distinguished both from culture and histruction, in that 
 it is of a purely negative character. 
 
 That such talents as imagination and wit, which ask a 
 free and unlimited development, require in many respects 
 the corrective influence of Discipline, every one will admit : 
 it may seem strange, however, that reason, whose proper 
 duty is to prescribe rules of discipline to all the other powers 
 of the mind, should itself require this corrective. It does 
 not indeed require it, either in the field of possible experience, 
 where its principles (Grundsatze) are subject to the test of 
 observation, or in the sphere of Mathematic, where its con- 
 cepts must always be presented in pure intuition ; but, 
 when we can obtain neither of these tests, as in its tran- 
 cendental use, when it proceeds according to pure concepts, 
 here discipline is very necessary to keep it within the limits 
 of possible experience. 
 
 We must notice that in this, the second part of the 
 Kritik, discipline is directed not to the content, but simply 
 to the method of the cognition of pure reason ; the former 
 task has been completed in our Elementology. The latter 
 is also necessary ; for the employment of reason in the 
 transcendental sphere being quite unique, without such a 
 discipline errors would unavoidably arise from the follow- 
 ing of methods which, however suitable to reason in other 
 inquiries, would be out of place here. 
 
 § i. Discipline in the sphere of Dogmatism. Mathematic 
 is a brilliant example of the extension of pure reason 
 
xxrx CONSTRUCTION OF CONCEPTS 363 
 
 without the aid of experience. We ask, then, What is the 
 difference between the mathematical method and that 
 which in philosophy we call dogmatical 1 The answer is : J 
 Philosophical cognition proceeds from concepts, while 
 mathematical cognition can only proceed by means of the 
 construction of concepts. By the construction of a concept 
 we mean the a priori presentation of the corresponding 
 intuition ; we construct, e.g. a triangle, either by mere 
 imagination in pure intuition, or upon paper in empirical 
 intuition ; but in both cases quite a priori. The individual 
 figure drawn upon paper, though empirical, 2 serves to 
 indicate the concept in its universality, because we attend 
 to the act, not to the result of the construction, which is 
 performed under certain universal conditions. Even an 
 empirical figure on paper suggests the general procedure of 
 the understanding, and so forms the basis of apodeictic 
 proof. Philosophical cognition accordingly regards the 
 particular only in the general ; mathematical, the general in 
 the particular, or rather in the individual. 
 
 Some have said that the feature which distinguishes 
 Philosophy from Mathematic is, that the former has to do 
 with quantity, z the latter with quality. We are now in a posi- 
 tion to see their error ; they have mistaken the effect for the 
 cause. Only concepts of quantity can be constructed, and so 
 the cognition of qualities through reason must be discursive, 
 through concepts. The difference between the two methods 
 of cognition is formal ; it does not result from a difference 
 
 1 Cf. Prolegomena, p. 27. 
 
 2 We have here Kant's answer to the question, How can an in- 
 dividual figure and an individual demonstration be valid universally ? 
 Cf. for another answer, Locke, Essay iv. i. 9. We may also remark that 
 this passage illustrates the schematism, and enables us to see what 
 Kant's views were as to the meaning of general terms. Cf. p. 147, supra. 
 
 3 Cf. Kant's Introduction to Logic, p. 13. 
 
364 GEOMETRY AND ALGEBRA CHAP. 
 
 of object matter. Philosophy, as well as Mathematic, treats 
 of quantities, e.g. of infinity, but in a different way. 
 
 It may also be observed that in Mathematic we construct 
 not only quantities {quanta), as in geometry, but also mere 
 quantity (quantitas), as in algebra, where we abstract 
 altogether from the qualities of the object. In this science 
 general signs are adopted to denote symbolically the con- 
 struction of concepts (viz. such signs as + , - , «/), and 
 also arbitrary symbols (letters such as x, y, z) are substituted 
 for the concepts of pure quantity according to their various 
 relations. Constructions are performed in intuition with 
 this apparatus, just as valid as, and more general than, 
 geometrical conclusions. 1 
 
 If an a priori concept contains a pure intuition, it can 
 be constructed, and we can proceed intuitively : if it con- 
 tains but a synthesis of possible intuitions, it cannot be 
 constructed, and we must proceed discursively. Such 
 synthetical propositions with respect to things in general, 
 the intuition of which cannot be given a priori, are called 
 transcendental ; they only contain the rule according to 
 which we must look empirically for a certain synthetical 
 unity of what cannot be represented in a priori intuition. 
 The transcendental concept of substance [the indefinable 
 subject of qualities], for example, does not denote an in- 
 tuition, but merely the synthesis of empirical intuition 
 which cannot be given a priori, and so no determining 
 synthetical proposition can spring from it ; we can only 
 get a principle of the synthesis of possible empirical 
 intuitions. The same is true of cause ; in this case I go 
 beyond the empirical concept of an event, not however 
 
 1 This is Kant's full explanation of the way in which algebraic 
 demonstrations are to be brought under intuitions in space and time. 
 Cf. Mill, Logic, vol. i. p. 287. 
 
xxix EXACTNESS OF MATHEMATIC 365 
 
 to any intuition giving the concept of cause in concrete, 
 but only to the conditions of time, which may in experience 
 be found in accordance with that concept. 
 
 These transcendental efforts to get beyond the limits of 
 experience are so alluring that it may be well to show more 
 fully the difference between the mathematical and the 
 philosophical method, and so to prove that the employment 
 of the method of Mathematic cannot be of the slightest ser- 
 vice in the field of philosophy. The exactness of Mathe- 
 matic depends on definitions, axioms, demonstrations. We 
 proceed to show that none of these, in the strict mathe- 
 matical sense, can be obtained in philosophy. 
 
 1. Definitions. A definition is the representation, upon 
 primary grounds, of the complete concept 1 of a thing within 
 its bounds. Hence we cannot expect to obtain a definition 
 of (i.) an empirical concept, e.g. of gold, for the progress of 
 observation is constantly adding new predicates or removing 
 old ones, so that such a concept never remains within fixed 
 bounds ; in logical language, it is not precise. All we can 
 hope for here is what we may call a 2 description or explication. 
 
 ii. Nor can an a priori concept, e.g. substance, be defined, 
 for we can never be sure that our analysis is complete ; 
 here we may say that we cannot get beyond an exposition. 
 
 iii. As regards arbitrary concepts, 3 these are certainly 
 in one sense susceptible of definition, for they are made 
 by myself; but I can never say, when they depend on 
 empirical conditions, that I have thus arrived at a definition 
 of a real object, for the existence of a concept does not 
 
 1 Compare the laws of logical definition and cf. Introd. to Logic, pp. 
 
 5'-53- 
 
 2 In fact, in such cases, the denotation rather than the connotation 
 
 of the concept is the important point. 
 
 3 Cf. Locke on notions and complex ideas generally. Essay ii. 22. 2. 
 
366 DEFINITIONS ' " Ar - 
 
 guarantee the existence or even the possibility of anything 
 corresponding. Perhaps I ought here to describe my ex- 
 planation isa 1 declaration (of my project). 
 
 iv. There remains then, as alone susceptible of definition 
 properly so called, concepts containing an arbitrary synthesis, 
 which can be constructed a priori. Such are mathematical 
 concepts, and this because in Mathematic, and in it alone, 
 is such a construction possible. We see then that philoso- 
 phical definition (so called) is only exposition (analytical), 
 but mathematical definition is by construction of concepts 
 (synthetical). 
 
 Hence we observe that though we may — nay, must — begin 
 with definitions in Mathematic,- we end with them in Philo- 
 sophy. We must of course assume provisional and imperfect 
 definitions at the commencement of our inquiry even in this 
 case, but a complete analysis of them is the last step. And 
 further, mathematical definitions cannot be erroneous as 
 regards content, though mistakes may occur now and then 
 as to form. Analytical definitions, on the other hand, arc 
 very liable to error either of excess or defect. 
 
 2. Axioms. These, so far as they are immediately cer- 
 tain, are synthetical principles a priori. Now, philosophical 
 cognition is always discursive, by means of concepts, and 
 therefore can never be axiomatic ; for a synthetical connec- 
 tion between two concepts is only immediately apparent 
 through intuition. Mathematic, on the contrary, may 
 possess axioms, because in it, by means of the construction 
 of concepts, two predicates may be connected a priori, and 
 
 1 Apparently in this case, though Kant docs not say so, the third 
 condition of definition is not fulfilled — that is, the precise number of 
 predicates of an arbitrary concept cannot be determined on primary 
 grounds [urspriinglich). 
 
 2 In Mathematic definitions belong ad esse, in Philosophy ad melius 
 esse. 
 
xxix AXIOMS 367 
 
 united. Philosophical principles, then, are discursive and 
 deductive ; axioms are self-evident. It is true that in the 
 Analytic certain ' axioms of intuition ' were spoken of, but 
 the principle 1 there instanced, far from being an axiom, 
 served only to indicate the principle (Principiuni) of the 
 possibility of axioms in general, being itself only a principle 
 (Grundsatz) based on concepts. It was introduced in that 
 place, because it is necessary for transcendental philosophy 
 to establish even the possibility of mathematics. 
 
 3. Demonstrations. Apodeictic propositions may be 
 divided into two classes, possessing intuitive and discursive 
 certainty respectively, to which the names mathemata and 
 dogmata may be given. It is only to a proposition of the 
 former class that the term demonstration can be applied, 
 and hence it is plain, as before, that philosophical proofs, 
 which must be discursive and not intuitive, cannot be 
 demonstrative ; we may call them acroamatic, as being 
 taught by words only. 
 
 The method of philosophy cannot then be dogmatical 
 (though it may be systematical), for pure speculative reason 
 does not possess a single direct synthetical proposition from 
 concepts, i.e. does not possess a single 1 dogma to start 
 from. Even such a proposition as ' Every effect has a 
 cause ' is not a dogma, for it is not based merely on con- 
 cepts ; but, in order to establish it, we require to bring in 
 the possibility of experience. The proper term for such a 
 proposition is Principle (Grundsatz), not dogma or precept 
 (Lehrsatz), because, although it does not require to be 
 proved, it possesses the remarkable peculiarity of rendering 
 its own proof, experience, possible. 
 
 § 2. Discipline of Pure Reason in Polemics. — We saw 
 
 1 AU intuitions are extensive quantities. Cf. p. 163, supra. 
 - Cf. p. 36, supra. 
 
368 DISCIPLINE IN POLEMICS chap. 
 
 that the apparent contradictions into which reason fell in 
 the Antinomies could be reconciled with each other by 
 adopting the distinction between phenomena and noumena. 
 In certain other cases, however, reason falls into an anti- 
 thetic, which cannot be so easily accounted for. For 
 example, take the counter assertions : There is a God — 
 there is no God : the soul is mortal — the soul is immortal. 
 Here we have two pairs of contradictory propositions, of 
 which, in each case, one must be true and the other false ; 
 but we cannot tell which, for we have no sure basis for 
 an argument on either side. Hence, though we cannot 
 demonstratively prove the existence of God or the immor- 
 tality of the human soul, we have the consolation that no 
 one can prove the opposite. Now what we mean by the 
 Polemic of Pure Reason is this defence of a proposition 
 made by reason in opposition to the dogmatical counter- 
 assertions of other people. By such a Polemic, however, 
 we can only oppose and disarm our opponents — we cannot 
 attempt to defend our own side ; to do the latter would at 
 once be dogmatical. 
 
 But we need not always have recourse to scholastic 
 arguments. We may safely accept those propositions which 
 agree with the speculative interests of reason in its empirical 
 use, and which are the only means of reconciling them with 
 our practical interests. It is quite allowable to employ, in 
 the presence of reason, the language of a 'arm faith, though 
 we may have to give up all pretensions to knowledge. 
 
 Properly speaking, then, there is no insoluble antithetic 
 of pure reason, nor should there be any such thing as a 
 Polemic ; two people ought not to dispute about things the 
 reality of which neither can present in actual or even in 
 possible experience. Both parties in such case are fighting 
 with their own shadows : their combat is as unending as 
 
xxix LIMITS AND BOUNDS 369 
 
 that of the heroes in Walhalla. Nothing will put an end 
 to these foolish disputes but criticism, which will show that 
 knowledge is unattainable on such matters, and that we 
 should be content with a rational faith , we have then an 
 unassailable position. However, it must be observed that 
 it is impossible to satisfy Pure Reason in conflict with itself, 
 by a mere sceptical solution : scepticism cannot be a per- 
 manent condition of the human mind. Consciousness of 
 ignorance ought to be the strongest impulse to investigation. 
 Here we come upon an important distinction j ignorance 
 may be_ (a) of facts or (b) of the bounds of cognition. If 
 ignorance is accidental it should incite us (a) to investigate 
 facts dogmatically, (l>) to investigate the bounds of cognition 
 critically. With regard to the manner in which these 
 inquiries are to be pursued we may see a posteriori (a) that 
 reason, as a matter of fact, has * limits, but (b) the deter- 
 mination of its bounds can only be accomplished a priori, 
 and is a matter of science, not of mere perception. For 
 instance, every one knows that he can only see a certain 
 portion of the earth's surface at any given time, and so his 
 perception is limited ; but unless he knows that the earth is 
 a sphere, he cannot tell what are the bounds of possible 
 perception ; this can only be found out a priori. Reason 
 is not to be considered as an indefinitely extended plane, 
 the limits of which are only known in a general way, but 
 ought to be compared to a sphere, the radius of which can 
 be determined from the curvature of a portion of its surface 
 (i.e. from the nature of its synthetical a priori propositions). 
 The procedure of reason may be described as at first 
 dogmatic ; then it subjects its facts to censorship — it sees 
 that they may be doubtful beyond certain limits — it declares 
 itself to be limited: this is the sceptical stage. But lastly, 
 1 a.Preleg, p. 1 54. 
 I 2 B 
 
37° HUME'S POSITION 
 
 not merely the facts of reason, but reason itself in its whole 
 power and in regard to its fitness for a priori knowledge, is 
 placed under examination : this is the final and critical 
 stage, in which the actual bounds of reason are determined 
 a. priori. Now we are actually in possession of a priori 
 synthetical cognitions [as we proved in the Analytic], but 
 till we can point out the limits and bounds of their use, we 
 are dogmatists. And this must be possible to determine ; 
 reason originates the concepts in question, and therefore 
 will be able to solve all problems immediately concerning 
 them and which inevitably arise (cf. p. 307, supra). 
 
 We can now see Hume's position in philosophy : he 
 was a sceptic (perhaps the ablest of all sceptics) — he did 
 good service in showing the insufficiency of dogmatism, 
 and in pointing out that knowledge was limited, but he 
 never advanced to the critical standpoint — he never pointed 
 out the bounds of knowledge. And we may say that his 
 fundamental mistake arose as follows : — 
 
 A priori synthesis of the Understanding derives all its 
 validity from the fact that the judgments we form relate to 
 objects of possible experience; and though we cannot 
 proceed immediately beyond the contents of a given con- 
 cept, we can a priori discern the law of its connection with 
 other things, by reference to the possibility of experience 
 itself. 1 A priori synthesis of the Reason, on the other 
 hand, is impossible, because we have here no third term with 
 which we may a priori discern the connection of our 
 concepts, and so determine their connection with each 
 other. Now Hume did not distinguish between the valid 
 procedure of Understanding and the dialectical attempts of 
 Reason ; he regarded the spontaneous generation of Under- 
 standing and Reason, independently of the impregnation of 
 
 1 Cf. p. 159, supra. 
 
xxix DISCIPLINE IN HYPOTHESIS 371 
 
 experience, as alike impossible. His error lay, not in 
 denying the possibility of Metaphysic, but in refusing to 
 recognise the just claims of Physic. 
 
 For example, he thought that the Law of Causality 
 possessed only a subjective necessity — was only contingent. 
 Here he mistook the contingency of our determination 
 according to the law, for the contingency of the law itself. 
 It may be only contingent that an effect A has been pro- 
 duced by a particular cause B, but it is objectively neces- 
 sary that A should have had some cause. As was pointed 
 out before (p. 40), he did not take a large enough view ; 
 he viewed Causality as an isolated and unique principle, 
 whereas we have given in our Analytic a complete table 
 of similar principles. 
 
 § 3. The discipline of Pure Reason in Hypothesis. — 
 There are two conditions which must be fulfilled that an 
 hypothesis may be regarded as legitimate. i. Its object 
 must be possible : reason must not attempt to form con- 
 cepts of objects which do not harmonise with the conditions 
 of the possibility of experience, Space, Time, and the 
 Categories. ii. It must be adequate to determine a 
 priori all the consequences that require explanation ; it 
 must be independent of subsidiary hypotheses. Hence 
 then a valid hypothesis must be not only possible but 
 useful. 
 
 For example, to suppose that the soul is a simple 
 substance would be to violate the first condition ; though, 
 on the other hand, it is quite admissible to think it simple, 
 in order to employ, in our inquiries into mental pheno- 
 mena, the Idea of the complete and necessary unity of 
 the mental faculties. Such an hypothesis we call tran- 
 scendental. We may next notice that an hypothesis of this 
 nature is quite inadmissible in the sphere of phenomena , it 
 
372 
 
 MATERIALISM 
 
 might help to satisfy the Reason, but it would not help the 
 Understanding in dealing with objects. Natural events 
 must be explained by natural laws ; we must not introduce 
 the dcus ex machina into physics. 1 In fact, by hyperphysical 
 explanations of physical facts, we not only do not advance 
 Reason, but we deprive it of the fruits which spring from 
 the cultivation of its proper soil, Experience. 
 
 The only legitimate sphere for Transcendental Hypo- 
 theses is that of Polemic, not Dogmatic , and the reason of 
 this is, as we saw before, that all synthetic propositions of 
 pure reason possess the peculiarity that, though not suscept- 
 ible of proof, they are equally incapable of disproof. So, 
 plainly, if there is a practical necessity of reason to believe 
 in a certain proposition, which cannot be proved false, we 
 are perfectly justified in maintaining its truth, even though 
 we cannot on the other hand prove it to be true. Melior 
 est conditio possidentis. These hypotheses of pure reason, 
 though they are but leaden weapons, not having been 
 steeled by any law of experience, are quite as useful as any 
 that can be employed by opponents. To take another 
 example : Suppose a materialist to say that mind is a 
 function of the organism, and to appeal in support of his 
 doctrine to the admitted fact that mental powers seem to 
 be largely dependent on conditions of the body. We at 
 once weaken the force of his argument by the assumption 
 that the body is only the fundamental phenomenon, which 
 is the condition in our present state of sensibility and 
 therefore of thought. The body, far from being the cause 
 of thought, is only its restrictive condition. 
 
 Such an assumption cannot be disproved, and so we at 
 once show our opponent that the laws of experience and 
 nature do not exhaust the field of possibility, and that, 
 T Cf. p. 358, supra. 
 
xxix DISCIPLINE IN PROOFS 373 
 
 therefore, whatever he may think of our position, he has 
 no right to regard his as established, there being another 
 alternative which he cannot show to be impossible. It 
 must never be forgotten, however, that such hypotheses, in 
 the speculative sphere, are not valid as independent pro- 
 positions, but only in relation to opposite transcendent 
 assumptions. 
 
 § 4. The Discipline of Pure Reason as to Proofs. — 
 The proofs of transcendental synthetical propositions are 
 distinguished from the proofs of all other synthetical a 
 priori propositions 1 (e.g. in Mathematic) by this, that in 
 them Reason does not apply itself directly to an object 
 through its concepts, but has first to prove the objective 
 validity of those concepts and the possibility of their a 
 priori synthesis. The neglect of this preliminary con- 
 sideration has been the source of much error : attempts 
 have been made in vain to prove such propositions by 
 the principle of inseparable association. For example, all 
 previous efforts to prove the principle of Sufficient Reason 
 have been fruitless ; here philosophers seem to have 
 abandoned proof, and appeal boldly to the common sense 
 of mankind. 
 
 And if in synthetical a priori propositions in the sphere 
 of Understanding, a necessary preliminary to proof is the 
 deduction of the possibility of the synthesis, such a deduc- 
 tion must be still more indispensable in the case of an 
 assertion of Pure Reason, where I aim at going beyond 
 my empirical concepts by the aid of mere Ideas. For, 
 take the supposed proof of the simplicity of the soul 
 derived from the unity of apperception. It looks plausible ; 
 why does it fail ? It is because the notion of absolute sim- 
 plicity is not a concept which can be referred directly to a 
 1 Cf. p. 364, supra. 
 
374 TRANSCENDENTAL PROOFS 
 
 perception ; as an Idea, at best, it can only be inferred. 
 In fact, to argue from the simplicity of a thought to the sim- 
 plicity of the subject of thought is a step which requires 
 justification. We should first of all have established the 
 validity of this step — have established the possibility of con- 
 necting our concept of the soul a priori with the Idea of a 
 simple being. Now, how can our concept of simplicity be 
 thus expanded ? In the language of Spinoza, how can we 
 elevate an abstraction into a res completa ? The possibility 
 of this a priori synthesis cannot be established, and there- 
 fore the supposed proof is only a paralogism. 
 
 i. The first rule that we arrive at is, then, not to attempt 
 a Transcendental Proof before we have first considered 
 from what source we are to derive the principles on which 
 it is to be based. If they are principles of Understanding 
 they are only applicable to objects of experience, and there- 
 fore we cannot arrive at Ideas by their means. If they 
 are principles (Grundsatze) of Pure Reason they are only 
 regulative of experience. All this we have shown before. 
 
 2. A second peculiarity of transcendental proof is this : — 
 For every transcendental proposition only one proof can be 
 found. If I am drawing conclusions not from concepts, 
 but from intuitions, as in Mathematical or Experimental 
 Science, then I may arrive at the result by several different 
 methods. But, on the other hand, every transcendental 
 proposition sets out from one concept, and posits the 
 synthetical condition of the possibility of the object according 
 to that concept. There can, therefore, be but one ground 
 of proof, because, besides this concept, there is nothing else 
 by which the object could be determined. The proof can 
 contain nothing more than the determination of an object 
 in general according to that unique concept. For example, 
 in our Transcendental Analytic, we inferred that ' Every 
 
xxix THEIR PECULIARITIES 375 
 
 effect has a cause,' from the single condition of the objective 
 possibility of the concept of an event ; that is, we showed 
 that an event could not be determined in time unless it 
 were subject to such a dynamical rule. In the same way, 
 if we want to prove that ' Every thinking being is simple,' 
 we must pay no attention to what is manifold in thought, 
 but keep the concept of the Ego alone in view. So 
 also with the transcendental proof of the existence of 
 God, which rests entirely on the reciprocability of the two 
 concepts of a most real and of a necessary Being, and cannot 
 be sought elsewhere. 1 If more reasons than one are pro- 
 duced by the dogmatist for any transcendental proposition, 
 we may be sure that he has not got the one on which the 
 whole force of the proof must rest, that his reasons are false 
 and sophistical — mere special pleading. 
 
 3. It will hence follow, in the third place, that transcend- 
 ental proofs must be always direct, not indirect. In one 
 respect indeed, apagogic or indirect proofs have the advan- 
 tage of being more easily understood, but they can, at 
 best, only show that a proposition is true, not why it is 
 so : an osiensive or direct proof, on the contrary, not only 
 establishes the proposition, but gives it grounds. The 
 reason why apagogic proofs are so much used in the different 
 sciences appears to be as follows. In hypothetical reason- 
 ing we may legitimately argue from the negation of the 
 consequent to the negation of the antecedent ; but the only 
 case in which an argument from the affirmation of the 
 consequent to the affirmation of the antecedent is valid 
 is this — if we have enumerated all possible consequents. 2 
 If every possible consequence of an hypothesis be true, the 
 
 1 In fact a transcendental proof is an argument from an effect to its 
 only possible cause. Cf. p. 45, supra. 
 
 2 Cf. p. 352, note. 
 
376 APAGOGIC METHOD chap, xxix 
 
 hypothesis itself must be true ; but it is needless to point 
 out that it is not possible in practice to discover all possible 
 consequences of a given proposition. Hence the modus 
 Miens or indirect method is often substituted for the 
 direct modus ponens, being easier and just as valid. How- 
 ever, this apagogic method is only admissible in sciences 
 where it is impossible to mistake a merely subjective repre- 
 sentation for objective truth. Now in Mathematic such 
 subreption is impossible, and so in that science the apagogic 
 method is valid and useful. In Physic we may guard 
 against mistakes in its employment by multiplying our 
 observations; however, the method here is not rigid, and is of 
 little use. But when we come to Metaphysic, the sphere of 
 pure Reason, the proper region of dialectical illusion, the 
 method is quite useless. Here, in order to prove any 
 synthetical statement, it is not sufficient to establish the 
 inconceivability of the opposite, for error may arise in two 
 ways, (a) This supposed refutation may merely show the 
 inconsistency of the opposite opinion with the subjective 
 conditions of reason, e.g. in the Dynamical Antinomies ; it 
 does not follow that, because the thesis and antithesis are 
 subjectively opposed, they are not both true from different 
 points of view, (b) On the other hand, as in the Mathe- 
 matical Antinomies, both thesis and antithesis may be false 
 and dialectical, being based on an impossible concept (an 
 absolutely existing world of sense). In this latter case the 
 rule applies : Non entis nulla sunt prozdicata. 
 
CHAPTER XXX 
 
 CANON, ARCHITECTONIC, AND HISTORY 
 
 The Canon of Pure Reason 
 
 We have seen that the proper use of a philosophy of pure 
 speculative reason is negative only ; it does not supply us 
 with an Organon, only with a Discipline ; it rather guards 
 us from error than furnishes us with new truth. At the 
 same time, there must be some source of positive Knowledge 
 in the sphere of Pure Reason ; how else can we account for 
 the irrepressible desire of the human mind to get beyond 
 mere experience, to arrive at the Unconditioned ? May we 
 not hope that in the practical sphere, which we have not 
 considered as yet, Reason may meet with better success 
 than it has had in the speculative ? 
 
 Now, if there be any legitimate exercise of Reason at all, 
 there must be a Canon for its procedure. By a Canon is 
 meant the sum-total (Inbegriff) of the a priori principles of 
 the right use of the cognitive faculties. General Logic 
 supplies us with a merely formal Canon of both Under- 
 standing and Reason. 1 Our transcendental Analytic is the 
 Canon of the pure use of the Understanding. What we 
 now proceed to investigate is a Canon of Pure Reason, and 
 1 Cf. Introd. to Logic, p. 3. 
 
378 THE TWO QUESTIONS chap. 
 
 we shall confine ourselves to the practical sphere, for we 
 have seen that its speculative exercise is dialectical. 
 
 § i. Of tlie Ultimate End of the pure use of our Reason. — 
 The speculative interest which Reason has in its three Ideas 
 is very small, for they are transcendent, not immanent, and 
 so are of no use in explaining empirical facts. Hence, then, 
 since they are of no service as regards knowledge, and yet 
 are so strongly urged upon us by Reason, it is plain that 
 their real value and importance relate to our practical 
 interest. Now the moral laws alone, not being empirically 
 conditioned, but being absolutely imperative, are the pro- 
 ducts of pure practical reason, and hence they alone admit 
 of a canon such as we seek ; that is, there is a Canon of 
 practical Reason, if and if only there be moral laws. In 
 fact, Reason, in its pure use, is wholly directed to the 
 three problems of the Freedom of the Will, the Existence of 
 God, and the Immortality of the Soul. These themselves, 
 however, have a still further object, namely, to know what 
 ought to be done, if the will be free, if there be a God, and if 
 there be a future World. As this concerns our conduct, 
 with reference to the highest end of life, we see that the 
 last intention of nature was really, in the constitution of our 
 Reason, directed to the Moral alone. 
 
 We may further notice that the question of tran- 
 scendental freedom is entirely speculative, and that the 
 existence of practical freedom 1 can be proved from ex- 
 perience alone : hence in our Canon we have only to do 
 with two questions, viz. Is there a God? Is there a 
 future life ? 
 
 1 ' A will which is independent of sensuous impulses and can be 
 determined therefore by purely rational motives, is a Free Wilt: This 
 is the meaning of practical freedom ; for Kant's definition of tran- 
 scendental freedom, cf. p. 295, and Proleg. p. 141. 
 
xxx A MORAL WORLD 379 
 
 § 2. Of the Ideal of the Supreme Good, as determining 
 ground of the ultimate end of Pure Reason. — The whole 
 interest of my Reason is centred in the three following 
 questions: 1 What can I know? What ought I to do? 
 What may I hope ? 
 
 The first question is purely speculative, and Kant considers 
 that he has sufficiently answered it in the Kritik. The 
 second question is purely practical : the answer is, ' Do 
 that which will render thee worthy of happiness.' The 
 question, ' If I act so as not to be unworthy of happiness, 
 may I hope on that account to obtain it,' is both practical 
 and theoretical, and is really that one with which we are 
 here concerned. Now, before answering it, we assume that 
 there are Moral Laws which command absolutely (and 
 therefore can be obeyed), i.e. that actions done from purely 
 rational motives are possible. Assuming this is to say that 
 Pure Reason contains in its practical use principles (Prin- 
 ciple^ of the possibility of experience, namely, of such 
 actions as, in accordance with moral precepts, might be 
 met with in the history of man. That is, we establish the 
 systematic unity of nature and the objective reality of these 
 principles in the practical sphere, though we could not do 
 so speculatively. We call the world, thus regarded, a moral 
 world ; it is only an Idea, but it is an Idea which may have 
 (and ought to have) an influence on the world of sense. In 
 this intelligible moral world happiness would be necessarily 
 proportioned to morality ; freedom, as impelled or restrained 
 by the moral law, being itself the cause of general happiness. 
 But, in postulating this moral world, we must postulate the 
 condition under which alone it is possible, viz. the idea 
 of an intelligence in which the most perfect moral will 
 united with the highest blessedness is the cause of all 
 1 Cf. Introd. to Logic, p. 15. 
 
380 THE SUMMUM BONUM chap. 
 
 happiness in the world so far as it stands in strict relation 
 to morality, i.e. the Ideal of the Supreme Good. In plain 
 language, we are necessitated by Reason to regard ourselves 
 as belonging to a world in which our happiness is to be 
 exactly proportioned to our virtue. This is not the case in 
 the present world of sense, nor could it ever be the case 
 except under a perfect moral governor. Hence God and a 
 future life are two hypotheses which are necessary conse- 
 quences of the moral law ; if they were not true, the moral 
 law could be only an idle dream. Without them the ideas 
 of morality may be objects of praise and admiration, but 
 cannot be springs of purpose and action, because they 
 would fail to fulfil all the aims which are natural and 
 necessary for a rational being. In short, the Supreme 
 Good consists of two distinct elements, Virtue and Happi- 
 ness. 
 
 This moral theology has the peculiar advantage over 
 speculative, that it leads inevitably to the concept of a single, 
 most perfect, and rational First Cause, whereof speculative 
 theology does not give us even an indication on objective 
 grounds, far less a conviction. The only way in which we 
 can establish the existence of a single Being at the head of 
 all natural causes is to take our stand on moral unity as a 
 necessary law of the universe, and thence infer that there 
 must be one supreme will which comprehends all these laws 
 within itself. This systematic unity of ends in the world of 
 intelligences then leads to the teleological unity of all things 
 which make up this great whole according to universal 
 natural laws, just as the former unity is according to univer- 
 sal and necessary moral laws. In this way the practical 
 and speculative reason become united. 1 The world must 
 be represented as having originated from an Idea if it is to 
 
 1 This is the point of departure of Kant's idealistic successors. 
 
xxx APODEICTIC CERTAINTY 381 
 
 accord with that use of Reason, without which we should 
 be unworthy of the possession of Reason — the moral use. 
 Hence the investigation of Nature becomes teleology, and 
 further, this notion of design in Nature gives rise to a 
 transcendental theology which takes the ideal of the highest 
 ontological perfection as a principle of systematic unity. 
 
 § 3. Of Opinion, Knowledge, and Belief. When Reason 
 in its canon asserts apodeictically upon the basis of its moral 
 laws, the faculty of freedom, the existence of God, and the 
 immortality of the soul, it assumes these three propositions 
 with a certainty excluding all doubt, and yet, strange to say, 
 they cannot be demonstrated scientifically. The nature of 
 our assent to them, then, is a point that requires further 
 clearing up. 
 
 If a judgment is valid for every rational being, then its 
 ground is objectively sufficient, and we call our assent to it 
 conviction; but if it has its ground only in the peculiar char- 
 acter of the subject, the name persuasioti is given. The former 
 can always be communicated to others ; the latter cannot. 
 
 Or we may express ourselves thus : There are three 
 degrees of holding for true, three modi of assent, Opinion, 
 Belief, and Knowledge. Opinion is subjectively as well as 
 objectively inadequate, is a mere problematical judging. 
 Belief is objectively inadequate, but is subjectively adequate, 
 an assertorial judging. And Knowledge is both sub- 
 jectively and objectively adequate ; what I know I regard 
 as apodeictically certain. 1 I must never venture to be of 
 opinion without knowing at least something by which my 
 judgment, problematical in itself, is connected with truth, 
 which connection, although not perfect, is still something 
 more than an arbitrary fiction. Moreover, the law of such 
 
 1 It is worth noticing how clearly Kant distinguishes objective from 
 subjective necessity in this passage. 
 
382 PRAGMATICAL BELIEF chap. 
 
 connection must be certain. For if, with respect to this 
 law, I have nothing but opinion, it is all a mere play of the 
 imagination bearing no relation whatever to truth. 
 
 In the judgments of Pure Reason opinion has no place ; 
 for in them the principle of connection must be universal 
 and necessary. Hence the sphere of opinion can neither 
 be Mathematic nor Ethic ; here there must be knowledge. 
 We must not hazard an action on the mere opinion that is 
 allowed ; we must know it to be so. In the transcendental 
 use of Reason, on the contrary, mere opinion is insufficient, 
 but knowledge is too strong a term. Speculatively here we 
 cannot form a judgment at all, because the subjective 
 grounds on which we hold a thing to be true cannot be 
 admitted in speculative inquiries. Nor can the theoretically 
 insufficient acceptance of a truth be called belief except from 
 the practical point of view. And this practical reference is 
 either to skill 1 or to morality, the former being concerned 
 with any arbitrarily proposed ends, the latter with absolutely 
 necessary ends only. 
 
 I. As an example of the former case, a physician 
 believes, say, that his diagnosis is correct in a case of 
 disease, and acts accordingly ; such a belief, contingent 
 indeed, but nevertheless giving the ground of the actual 
 use of means in certain actions, is called pragmatical. In 
 proportion as a man holds this kind of belief strongly, he 
 is ready to bet on it, and this is a fair test of its strength. 
 But it must be observed, further, that in many cases where 
 our judgment is purely theoretical, in which we cannot 
 undertake any correspondent actions, we can still represent 
 to ourselves in thought the possibility of such actions, for 
 which we suppose ourselves to have sufficient grounds, if 
 
 1 Cf. the distinction between hypothetical and categorical imperatives. 
 Theory of Ethics, p. 32. 
 
MORAL BELIEF 383 
 
 any means existed of ascertaining the truth. And so there 
 is in purely theoretical judgments an analogon to this 
 pragmatical belief in practical matters, which we will call 
 doctrinal belief; my belief of the existence of God, which 
 is produced in my mind by the pursuit of physical science, 
 is of this character. As a condition of a contingent, but 
 not unimportant end, namely, that I may have guidance in 
 the investigation of nature, it is necessary to admit a wise 
 author of the world. 1 Mere doctrinal belief, however, is 
 rather unstable ; we often lose it on account of difficulties 
 in speculation (though in the end we always return to it). 
 
 II. We now turn to Moral Belief. Here action is absolutely 
 necessary, and I must obey the moral law in all points. 
 The end in this case is incontestably established, and, as far 
 as we know, 2 only one condition is possible under which 
 that end could agree with all other ends and thus have 
 practical validity, viz. the existence of God and of a future 
 world. Hence I inevitably believe this, and this belief is 
 moral. My conviction may not be logically certain, but it 
 is morally certain ; my belief in God and in another world 
 is inextricably interwoven with my moral nature. 
 
 To sum up, then, our Canon consists only of these two 
 articles of belief; in respect to the essential ends of human 
 nature, the highest philosophy cannot effect anything more 
 than that guidance which nature has vouchsafed to the 
 meanest understanding. 
 
 1 Kant means by belief, 'assent which is sufficient to influence 
 action' (Introd. to Logic, p. 58, note). Our conviction of the exist- 
 ence of God derived from our investigation of nature is such that 
 we act on it when occasion arises ; it is stronger than mere opinion, 
 and so we call it belief, but doctrinal belief, to distinguish it from moral 
 belief derived from the evidence of the moral law. Compare with this, 
 Butler's use of the phrase ' practical proof in the Introduction to the 
 Analogy. * ViJe P- 3 8 °> »'&*• 
 
384 ARCHITECTONICAL UNITY 
 
 The Architectonic of Pure Reason 
 
 Architectonic is the art of system, and reason always re- 
 quires that knowledge should be reduced to system — that is, 
 that our different cognitions should be ranged under one 
 idea. This leading idea is the rational concept of the form 
 of a whole in which both the extent of its manifold contents 
 and the plan belonging to each part are determined a priori. 
 When we view a science from such a standpoint, it is like 
 an organism (articulatio), not a mere aggregate (coacervatio). 
 It grows internally, but not externally — like an animal 
 body, the growth of which adds no new member, 
 but without changing the proportion of the parts renders 
 each member stronger and more efficient for its own 
 purpose. 
 
 This Idea requires for its realisation a Schema, i.e. an 
 essential variety and order of parts determined a priori from 
 the principle of the end in view. A Schema which is not 
 designed according to an Idea, but empirically in accordance 
 with contingent ends, can only give us technical unity in our 
 science. In order to get architectonical unity, the schema 
 must give a priori the outline (inonogramma) of the science 
 and the division of it into parts according to the Idea ; the 
 parts must be shown to have an essential affinity, in that 
 they all depend on one supreme and internal end, which 
 renders the whole possible. A science must have made 
 considerable progress before such a conception of it can 
 be grasped ; its originator seldom sees it in its full extent 
 and true character. Scientific systems seem to grow like 
 worms by a kind of generatio tzquivoca, by the mere con- 
 fluence of concepts ; though really they all have a schema, 
 like an original germ, in accordance with which they are 
 developing. And just because this germ lies in reason in 
 
HISTORICAL KNOWLEDGE 385 
 
 all cases, it is possible to systematise human knowledge, to 
 construct an Architectonic of it. 
 
 We may first notice that knowledge may be either 
 historical (cognitio ex datis) or rational {cognitio ex prin- 
 cipiis). With the former we have nothing to do : we are 
 only concerned to sketch the plan of the Architectonic of 
 cognition from pure reason. As to rational cognition, we 
 before divided it (p. 363) into Mathematical, which proceeds 
 according to the construction of concepts, and Philosophical, 
 proceeding from concepts themselves. Mathematical know- 
 ledge cannot be merely historical , A a man cannot really 
 accept a mathematical truth without appreciating the steps 
 by which it is arrived at, and so it may be said that mathe- 
 matics may be learned ; philosophy, on the other hand, 
 cannot be learned except in the historical sense. A 
 person who has made himself acquainted by reading with 
 the tenets of a philosophical system, but who has not 
 thought them out for himself, cannot be called a philo- 
 sopher ; ' he is only the plaster-cast of a living man.' 
 
 So far we have been regarding Philosophy as equivalent 
 to rational knowledge from concepts. Now this is only the 
 scholastic concept of the science; from such a concept 
 we could never attain more than mere logical perfec- 
 tion. In this regard the end of a science is skill ; it is 
 directed towards certain arbitrary objects, and its purpose 
 is fulfilled if these are attained. And so, in the scholastic 
 sense, we are philosophers if we can philosophise well, if we 
 are masters of the art of reason ( Vernunftkunstler). But 
 there is also the cosmical concept of philosophy, which 
 has much more dignity. According to it philosophy is the 
 science of the relation of all knowledge to the essential 
 
 1 Historical knowledge is the mere knowledge that a thing is, as 
 distinguished from knowledge why it is. 
 ! 2 C 
 
386 COSMICAL CONCEPT chap. 
 
 ends of human reason (teleologia rationis humance) ; the 
 philosopher is not a mere master of the art of reason, but 
 a lawgiver. The mathematician, the physicist, and the 
 logician are only masters of the art of reason ; but besides 
 these there is the philosopher, the ideal teacher who uses 
 them all as instruments for the advancement of the essen- 
 tial aims of human reason. We do not pretend that he 
 anywhere exists, but nevertheless the Idea of his legislation 
 exists everywhere in the reason of every rational being; and 
 so, according to this Idea, we may determine what kind of 
 systematic unity philosophy in its cosmical concept 1 re- 
 quires, and also accordingly we may make out its sub- 
 divisions. 
 
 Now, plainly, if the perfect systematical perfection of 
 reason has been reached, there can be but one highest end 
 — the destination of man, and so philosophy, in the highest 
 sense, is really Moral Philosophy, as in fact it was regarded 
 by the ancients. Indeed, even at present we popularly 
 speak of a man who has his passions under proper control 
 as a philosopher. 
 
 We now proceed to classification. Knowledge being 
 either empirical or a priori, we shall subdivide philosophy 
 into (a) pure, (b) empirical. To the latter we shall return 
 presently. As to the former it may be either Propadeutic 
 — that is, preparatory and critical, or a systematical body of 
 doctrine, usually called Metaphysic. It is however not un- 
 common to include both criticism and doctrine under this 
 head, but we adopt the above distinction for the sake of 
 clearness. Next, Metaphysic may be either of Nature or 
 
 1 'By a cosmical concept,' says Kant, 'I mean one in which 
 every man necessarily is interested.' Compare with this the use of the 
 same phrase on pp. 281, 286. With the whole passage the student 
 should compare Intiod. to Logic, pp. 14-15. 
 
CLASSIFICATION 387 
 
 of Freedom, may be speculative or p7-actical ; when practical 
 it is called Ethic. The extreme importance of having our 
 classification of the sciences thus constructed on definite 
 principles is plain, for Metaphysic has never before been 
 clearly explained. When it was spoken of, e.g. as the 
 science of first principles of cognition, this definition gave 
 no clue to its real nature, for there was no criterion given 
 which might distinguish first principles from those which 
 were acquired later. The differences in kind between a 
 priori and a posteriori knowledge, and between philo- 
 sophical and mathematical cognition, lie at the root of the 
 whole matter, and Kant proudly claims for himself that he 
 was the first to observe them. The Metaphysic of Nature 
 consists of two parts, Transcendental Philosophy (Ontology) 
 and Rational Physiology. The former treats only of Under- 
 standing and Reason themselves in a system of concepts 
 and principles which relate to objects in general ; the latter 
 treats of nature — that is, the complex of given objects. 
 This latter again may be either immanent or transcendent, 
 the first if it takes account of the conditions under which 
 alone objects can be given us, the second if it regards a 
 connection of the objects of experience which transcends 
 all experience. Transcendent Physiology treats then of the 
 external connection of the whole of Nature with a Being 
 above Nature (Theology), or postulates an internal tran- 
 scendental connection of its parts (Cosmology); and as the 
 objects of the senses are twofold, either of corporeal or 
 sentient nature, we may have two branches of Immanent 
 Physiology also, rational Physics l and rational Psychology. 
 Now such a division is prescribed by the fundamental 
 
 1 Kant points out that he does not mean by this what is commonly 
 called fhysica generatis, which is rather Mathematic than a philosophy 
 of nature. 
 
PSYCHOLOGY AND PHYSIOLOGY 
 
 idea of a philosophy of pure reason, and so it is architectonic, 
 adequate to its essential aims, and not merely technical and 
 accidental. 
 
 In conclusion, two questions may naturally be asked. 
 i. How can I expect knowledge a priori, i.e. Metaphysic, 
 of objects so far as they are given to our senses a posteriori! 
 how can I expect to have a rational physiology ? The 
 answer is : We take from experience nothing beyond what 
 is necessary to give us an object of external or internal 
 sense ; in the one case by the mere concept of matter 
 (impenetrable and lifeless extension), in the other by the 
 concept of a thinking being (in the internal empirical 
 representation / think), ii. How shall we classify empirical 
 psychology! We reply : It comes under applied philosophy ; 
 it has, properly speaking, nothing to say to Metaphysic. 
 It may, however, retain its present place under Metaphysic 
 for convenience sake, but as a mere guest until it is provided 
 with a proper home in a systematic Anthropology, the 
 proper pendant of Empirical Physics. 1 
 
 Such is the general idea of Metaphysic. It must be 
 observed that we have shown that though it does not supply 
 the foundation of religion, it must always remain its bulwark. 
 This is true philosophy ; it leads to wisdom, and the way to 
 it is the way of science. 
 
 The History of Pure Reason 
 
 We can only, of course, under this head, give the merest 
 sketch ; it is striking that Theology has always, in the past, 
 formed the beginning of speculative inquiry rather than 
 the end. 
 
 We may classify philosophers in three ways : 
 
 1 Cf. vol. ii. p. 253. 
 
HISTORY 389 
 
 i. With regard to the object of rational cognition, they 
 must either be Sensualists with Epicurus or Intellectualists 
 with Plato. 
 
 ii. In relation to the origin of pure rational cognition, 
 they are either Empiricists or Noologists ; they either follow 
 Aristotle and Locke, or Plato and Leibnitz. Locke is 
 peculiarly inconsequent in his empiricism, for having 
 derived all concepts and principles from experience, he 
 yet endeavours to apply them to prove the existence of 
 God and the immortality of the soul (though both lie 
 entirely outside the bounds of possible experience). 
 
 iii. In relation to method. If we pass over the natural- 
 ist of pure reason who relies altogether on what he calls 
 common sense, 1 the only methods in use in the past have 
 been the dogmatic {e.g. Wolff) and the sceptical {e.g. Hume). 
 The one scientific path which has not yet been tried is the 
 critical, and this Kant commends to his reader as a path of 
 great promise in the future. 
 
 1 Cf. Proles;, p. 182. 
 
 END OF VOL. I. 
 
 Printed by R. & R. Clark, Edinburgh. 
 
WORKS by the Rev. J. P. MAHAFFY, M.A., D.D., 
 
 FELLOW OF TRINITY COLLEGE, DUBLIN. 
 
 Social Life in Greece, from Homer to Menander. 
 
 Sixth Edition, enlarged. Crown 8vo. 9s. 
 
 The Athenanim says : — " Mr. Mahaffy has, in this little volume, given 
 us the results of considerable reading, and has given them in a form which 
 is both pleasant and interesting. . . . No omission greatly detracts from 
 the merits of a book so fresh in its thought and so independent in its 
 criticism. . . . One feels that the author is no mere compiler, but an 
 original thinker ; and whether we agree with his conclusions or not, we 
 at least respect the boldness and straightforwardness with which he holds 
 his own." 
 
 Greek Life and Thought, from the Age of Alexander 
 to the Roman Conquest. Crown 8vo. 12s. 6d. 
 
 The Academy says : — " These studies of Greek life and thought may be 
 regarded as complementary to more than one of Professor Mahaffy's other 
 volumes upon Greek affairs. They carry further the account of Social Life 
 in Greece ; they take note of authors later than the point at which the 
 History of Greek Literature ends ; and they supply the domestic and 
 literary features necessary to body-out the brilliant sketch of Alexander's 
 Empire. Mr. Mahaffy vindicates, in an eloquent introduction, the im- 
 portance of his subject. . . The subject is a grand one, and the author's 
 opportunity is the better because the subject has been comparatively 
 neglected. The political and social experiments of the time ; the spread 
 of Greek culture necessary for the future of humanity ; the actual achieve- 
 ments in architecture and sculpture, in poetry and in painting — these things 
 arouse instantly the interest of the historian and the artist, while the 
 philosopher will be curious to see how Mr. Mahaffy justifies his favourable 
 opinion of the morality and the daily life. This splendid subject has now 
 found an English historian competent to do it justice." 
 
 The Guardian says : — " . . At his own immediate subject Mr. 
 Mahaffy has worked with much skill and care. It is pleasant to visit 
 Alexandria, Antioch, and Pergamon in his company, and to make acquaint- 
 ance with their kings, their poets, and their philosophers. We see what 
 kind of literature, what kind of philosophy, grew up in a very artificial 
 and conscious age, an age which, as Mr. Mahaffy does not fail often to 
 remind us, has so very much in common with modern Europe." 
 
 Rambles and Studies in Greece. With Illustrations. 
 
 Third Edition, revised and enlarged. Crown 8vo. 10s. 6d. 
 
 The St. James's Gazette says :— " It is unnecessary to praise this book, 
 
 which the public has already decided to be one of the best of its kind. 
 
 The present edition is enriched with a good deal of additional information." 
 
 1] MACMILLAN AND CO., LONDON. 
 
WORKS by the Rev. J. P. MAHAFFY, M.A, D.D, 
 
 FELLOW OF TRINITY COLLEGE, DUBLIN. 
 
 The Art of Conversation, The Principles of. Second 
 Edition, revised. Crown 8vo. 4s. 6d. 
 
 The Guardian says : — "The result is a book which has abundance of 
 shrewd remarks." 
 
 The Saturday Review says : — " Full of shrewd observation. . . . Most 
 of his rules and recommendations are aimed at promoting conversation, 
 and directing it into the right channel." 
 
 The Nation (New York) says: — "Independently of these general 
 questions, which must be touched upon of necessity in a volume of first 
 principles, Mr. Mahaffy's dissertation is a very practical and useful one. 
 His treatment is light, but thorough, and he pursues his subject with un- 
 flagging spirit through all its compass. Hosts and guests may equally 
 learn from him." 
 
 Euripides. l8mo. IS. 6d. {Classical Writers. 
 
 The Academy says : — "No better book on the subject has previously 
 been written in English. Mr. Mahaffy is scholarly and not pedantic, 
 appreciative and yet just." 
 
 The Cambridge Review says : — " We can recommend this work to all 
 our readers ; it is very interesting, and, what is more important, very 
 suggestive." 
 
 Greek Antiquities. Illustrated. i8mo. is. 
 
 [Literature Primers. 
 
 The Saturday Review says: — "We can conceive no handbook of a 
 hundred pages which furnishes readers with so much light upon the civil 
 and domestic life of the Greeks." 
 
 The Decay of Modern Preaching. An Essay. Crown 
 8vo. 3s. 6d. 
 
 The Saturday Review says: — "The work is clever and sensible in 
 most of its criticisms and suggestions." 
 
 The Church of England Pulpit says : — " It is an excellent little book." 
 The Scotsman says : — " It is thoroughly well worth reading." 
 
 Euripides. Hippolytus. Edited, with Introduction and 
 Notes, by J. P. Mahaffy and J. B. Bury. Fcap. 8vo. 
 3S. 6d. [Classical Series. 
 
 MACMILLAN AND CO., LONDON. [2 
 
MESSRS. MACMI LLAN & CO.'S PUBLICATIONS. 
 
 With upwards of Eighty Illustrations. 
 Sketches from a Tour through Holland and Germany. 
 
 By J. P. Mahaffy and J. E. Rogers. With Illustrations by J. E. 
 
 Rogers. Extra Crown 8vo. ios. 6d. 
 This is an account of a tour through some of the less known towns in 
 Holland, Central Germany, and along the Baltic Coast. The titles of 
 chapters are as follows : Preamble ; Down the Thames ; Dordrecht ; 
 Leiden and Haarlem ; About the Zuider Zee — Enkhuizen ; Miscellaneous 
 Gleanings ; Central Germany, Brunswick — Romanesque Churches ; Helm- 
 stedt — Hildesheim; Marburg — Fulda ; In the Country; Some Baltic 
 Towns — Wismar ; Liibeck — Liineberg — Hamburg ; On the Way Home. 
 
 The Saturday Review says : "If this tour through Holland and part of 
 Germany does not largely increase our knowledge of those countries, it is, at 
 any rate, a very pleasant and amusing book." 
 
 The Scotsman says : "A cheery and delightful book of travel. . . . All 
 the book is bright and animated. The sketches join with the text in giving 
 just notions of bits of mediaeval architecture. The text itself is richer. Art, 
 history, economics, social habits, surface-politics, fashions in dress, philology 
 and the inaptitude of the Dutch vulgar to understand their own language when 
 spoken with a possible brogue — these and a thousand others are the topics of 
 the text. They are treated most lightly and pleasantly. The book will be 
 heartily enjoyed by all its readers." 
 
 The Glasgow Herald says : " This entertaining and most readable volume 
 is in no sense a guide book. It does not attempt to catalogue and label every 
 sight and every curiosity for the benefit of intending travellers. It contains 
 the jottings of two intelligent, cultured, and observant tourists who, in addition 
 to all that Baedeker and Murray point out, have seen much that is interesting, 
 but too often overlooked, and is, indeed, just such a book as will delight those 
 who already possess some acquaintance with the places described in it." 
 
 The Manchester Guardian says : ' ' Any one who is doubting where he had 
 better go for his next continental holiday will find this book pregnant of sug- 
 gestion. . . . On the whole, the book is slight — a record of casual impressions, 
 though the impressions of exceptionally intelligent and open-eyed observers. 
 It is, however, distinctly readable, and the architectural sketches of interesting 
 bits in old towns like Alkmar, or Hildesheim, or Liibeck, or Rostock, greatly 
 add to its attractiveness." 
 
 The Aberdeen Journal says : " The book contains a large fund of informa- 
 tion both valuable and interesting, especially to archaeologists devoted to the 
 study of Gothic and Renaissance architecture. . Mr. Mahaffy and Mr. 
 
 Rogers have indeed given their readers a treat." 
 
 Truth says : " It is a brightly written and beautifully illustrated book." 
 
 The Field says: "This book contains an interesting account of many 
 places in Holland and Central Germany, outside the track usually followed by 
 the English or American tourist ... the illustrations are remarkably good 
 ... the account given by the authors of the people and the things in general 
 that came under their notice are all pleasantly, and in many instances amus- 
 ingly, written ; and their remarks on the political state of Germany at the 
 present time are well worthy of notice." 
 
 The Dublin Evening Mail says : " Of all the books this season has pro- 
 duced within measurable distance of modest pockets, this sketch-book deserves 
 the palm." 
 
 3] MACMILLAN AND CO., LONDON. 
 
MESSRS. MACMILLAN & CO.'S PUBLICATIONS. 
 
 MENTAL AND MORAL PHILOSOPHY. 
 
 The Mathematical Analysis of Logic. Being an Essay towards a 
 Calculus of Deductive Reasoning. By George Boole. 8vo. Sewed. 5s. 
 
 Handbook of Moral Philosophy. By the Rev. Henry Calderwood, 
 LL.D-, Professor of Moral Philosophy, University of Edinburgh. Fourteenth 
 Edition, largely rewritten. Crown 8vo. 6s. 
 
 Seeing- and Thinking*. By the late Professor W. K. Clifford, F.R.S. 
 
 With Diagrams. Crown 8vo. 3s. 6d. {Nature Series.) 
 
 The Elements of the Psychology of Cognition. By the Rev. Robert 
 Jardine, B.D., D.Sc. (Edin.), Ex-Principal of the General Assembly's College, 
 Calcutta. Third Edition, revised and improved. Crown 8vo. 6s. 6d. 
 
 By the Late W. STANLEY JEVONS, LL.D., M.A., F.R.S. 
 Primer of Logic. New Edition. iSmo. is. {Science Primers,) 
 Elementary Lessons in Logic ; Deductive and Inductive ; with copious 
 
 Questions and Examples, and a Vocabulary of Logical Terms. New Edition. Fcap. 
 
 8vo. 3s. 6d. 
 The Principles of Science. A Treatise on Logic and Scientific Method. 
 
 New and Revised Edition. Crown 8vo. 12s. 6d. 
 Studies in Deductive Logic. Second Edition. Crown Svo. 6s. 
 
 Formal Logic, Studies and Exercises in. Including a Generalisation of 
 Logical Processes in their application to Complex Inferences. By John Neville 
 Keynes, M.A., late Fellow of Pembroke College, Cambridge. Second Edition, 
 Revised and Enlarged. Crown Svo. 10s. 6d. 
 
 Psychology. By James McCosh, D.D., LL.D., Litt.D. President of 
 Princeton College, Author of " Intuitions of the Mind," " Laws of Discursive Thought," 
 etc. Crown 8vo. 
 
 I. The Cognitive Powers. 6s. 6d. 
 
 II. The Motive Powers. Crown 8vo. 6s. 6d. 
 
 First and Fundamental Truths : being a Treatise on Metaphysics. 
 
 A Text-Book of Deductive Logic for the use of Students. By P. K. 
 Ray, D.Sc. (Lon. and Edin.), Professor of Logic and Philosophy, Presidency College, 
 Calcutta. Fourth Edition. Globe 8vo. 4s. rjd. 
 
 Works by HENRY SIDGWICK, M.A., LL.D., 
 
 Knightbridge Professor of Moral Philosophy in the University of Cambridge. 
 The Methods of Ethics. Third Edition. 8vo. 14s. A Supplement to 
 
 the Second Edition, containing all the important Additions and Alterations in the 
 
 Third Edition. Demy Svo. 6s. 
 Outlines of the History of Ethics, for English Readers. Second Edition, 
 
 revised. Crown 8vo. 3s. 6d. 
 
 By JOHN VENN, M.A., 
 
 Fellow and Lecturer in the Moral Sciences in Gonville and Caius College, Cambridge, 
 Examiner in Moral Philosophy in the University of London. 
 
 The Logic of Chance. An Essay on the Foundation and Province of the 
 Theory of Probability, with special Reference to its Logical Bearings and its Applica- 
 tion to Moral and Social Science. Third Edition, rewritten and greatly enlarged. 
 Crown 8vo. 10s. 6d. 
 
 Symbolic Logic. By the same Author. Crown 8vo. 10s. 6d. 
 
 Principles of Empirical or Inductive Logic. 8vo. 18s. 
 
 MACMILLAN AND CO., LONDON. [4 
 
MESSRS. MACMILLAN & CO.'S PUBLICATIONS. 
 
 Aristotle on Fallacies ; or, The Sophistici Elenchi. With a Transla- 
 tion and Notes by Edward Poste, M.A. 8vo. 8s. 6d 
 
 The First Book of the Metaphysics of Aristotle. Translated into 
 English Prose, with marginal _ Analysis and Summary of each Chapter. By a 
 Cambridge Graduate. 8vo. 5s" 
 
 The Politics. Translated, with an Analysis and Critical Notes. By J. 
 E. C. Welldon, M.A,, Fellow of King's College, Cambridge, and Head Master of 
 Harrow. Crown 8vo. Second Edition. 10s. 6d. 
 
 The Rhetoric. By the same Translator. Crown 8vo. 7s. 6d. 
 
 By THOMAS RAWSON BIRKS, M.A. 
 Late Professor of Moral Philosophy in the University of Cambridge. 
 
 First Principles of Moral Science ; or, First Course of Lectures 
 delivered in the University of Cambridge. Crown Svo. 8s. 6d. 
 
 Modern Utilitarianism ; or, the Systems of Paley, Bentham, and Mill 
 Examined and Compared. Crown 8vo. 6s. 6d. 
 
 Modern Physical Fatalism, and the Doctrine of Evolution. In- 
 cluding an Examination of Mr. Herbert Spencer's " First Principles." New Edition, 
 with a Preface, in reply to Mr. Herbert Spencer's Strictures, by Professor Pritcharu. 
 Crown 8vo. 6s. 
 
 By HENRY CALDERWOOD, LL.D., 
 Professor of Moral Philosophy, University of Edinburgh. 
 
 The Relations of Mind and Brain. Second Edition. Revised and 
 enlarged, with additional chapter on Animal Intelligence. 8vo. 12s. 
 
 The Relations of Science and Religion. The Morse Lecture, 18S0. 
 Connected with the Union Theological Seminary, New York. Crown 8vo. 5s. 
 
 The Metaphysics of the School. By Thomas Harper (S.J.) (In 5 
 
 vols.) Vols. I. and II. 8vo. 18s. each. Vol. III., Part I. 12s. 
 The Realistic Assumptions of Modern Science Examined. By 
 Thomas Martin Herbert, M.A., late Professor of Philosophy and Church History 
 in the Lancashire Independent College, Manchester. Second Edition. Crown 8vo. 
 8s. 6d. 
 
 By JAMES McCOSH, D.D., LL.D., Litt.D., U.S. 
 The Method of the Divine Government, Physical and Moral. 
 
 Fourteenth Edition. 8vo. 10s. 6d. 
 
 The Supernatural in Relation to the Natural. Crown 8vo. 7s. 6d. 
 
 The Intuitions of the Mind. A New Edition. 8vr.. 10s. 6d. 
 
 An Examination of Mr. J. S. Mill's Philosophy. Being a Defence of 
 Fundamental Truth. Second Edition. 8vo. 10s. 6d. 
 
 The Laws of Discursive Thought. Being a Text -book of Formal 
 Logic. Crown 8vo. 5s. 
 
 Christianity and Positivism. Lectures on Natural Theology and 
 Apologetics. Crown 8vo. 7s. 6d. 
 
 The Scottish Philosophy, from Hutcheson to Hamilton, Bio- 
 graphical, Expository, Critical. Royal 8vo. 16s. 
 
 The Emotions. 8vo. 9s. 
 
 Realistic Philosophy Defended in a Philosophic Series. 2 Vols. 
 Vol. I. Expository. II. Historical and Critical. Crown 8vo. 14s. 
 
 5] 
 
 MACMILLAN AND CO., LONDON. 
 
MESSRS. MACMILLAN & GO.'S PUBLICATIONS. 
 
 Critique of Pure Reason by Immanuel Kant. In commemoration of 
 the Centenary of its first Publication. Translated into English by F. Max Muller. 
 With an Historical Introduction by Ludwig Noire. 2 vols. Demy 8vo. 16s. (sold 
 separately) each. Vol. I. Historical Introduction, by Ludwig Noire, etc., etc. 
 Vol. II. Critique of Pure Reason. Translated by F. Max Muller. 
 
 Of Professor Max Muller's translation of the "Critique of Pure Reason" the Times 
 says : " Through this translation Kant's work has for the first time become international 
 —the common property of the whole world." 
 
 The Academy says: "A worthy piece of work. Its language is generally vigorous 
 and direct, and it preserves a considerable fidelity to the turns and structure of the original. 
 ... It may convey to English readers, as few translations from the German do, a real 
 idea of the original work. It will bring them nearer Kant's own thought." 
 
 The British Quarterly Review says: " Professor Max Muller has here performed a 
 service for his great countryman which, as years roll by, will_ be more and more recog- 
 nised by our own countrymen as having supplied a new starting-point for their thought- 
 lives, and given them the means of accounting to themselves for experience." 
 
 Nature says : "Professor Max Muller has given to students of philosophy what they 
 did not possess before, and that is far superior to the ordinary work of this sort. His 
 offering is indeed what it intended to be, a fitting commemoration of the centenary of the 
 date in which was published the treatise which was destined to revolutionise philosophy. 
 A faithful and literal translation of that treatise is a book for which he will not find the 
 public ungrateful to him." 
 
 The Nation (New'iYork) says: "It generally gives Kant's meaning with superior 
 precision. It is no inconsiderable advantage that the language from which, and that 
 into which, Professor Max Muller translates, are both alike his own. To English readers 
 his work is sure to bring Kant home better than anything we had before ; and even those 
 who read German readily will frequently gain something by looking up Max Muller's 
 understanding of an obscure passage. His translation ' serves in some passages even as 
 a commentary on the original.' For the first time, too, we have the ' Critique ' as it was 
 originally written." 
 
 The Athetupum says: "Professor Max Muller's translation is the best English 
 version of the whole ' Kritik,' . . . and it is likely to hold that position for some time." 
 
 Outlines of Cosmic Philosophy, Based on the Doctrine of Evolution, 
 with Criticisms on the Positive Philosophy. By John Fiske, M.A., LL.B., for- 
 merly Lecturer on Philosophy at Harvard University. 2 vols. 8vo. 25s. 
 
 Recent British Philosophy. A Review, with Criticisms, including some 
 Comments on Mr. Mill's Answer to Sir William Hamilton. By David Masson, 
 M.A., Professor of Rhetoric and English Literature in the University of Edinburgh. 
 Third Edition, revised and enlarged. Crown 8vo. 6s. 
 
 Moral and Metaphysical Philosophy. By Frederick Denison 
 
 Maurice. Vol. I.— Ancient Philosophy and the First to the Thirteenth Centuries. 
 Vol. II.— Fourteenth Century and the French Revolution, with a Glimpse into the 
 Nineteenth Century. Fourth Edition. 2 vols. 8vo. 16s. 
 
 Old -Fashioned Ethics and Common -Sense Metaphysics. By W. 
 T. Thornton, C.B., late Secretary for Public Works in the India Office. 8vo. 
 10s. 6d. 
 
 On Truth and Error. Thoughts on the Principles of Truth, and the 
 Causes and Effect of Error. By John Hamilton. Crown 8 vo. 5s. 
 
 MACMILLAN AND CO., LONDON. [6