CORNELL 
 
 UNIVERSITY 
 
 LIBRARY 
 
 GIFT OF 
 Sage ochool of Philosophy 
 

 V.I 
 
 260 
 
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/cletails/cu31924019527260 
 
THE CEITICAL PHILOSOPHY OP 
 IMMANUEL KANT 
 
THE 
 
 CRITICAL PHILOSOPHY OF 
 IMMANUEL KA:N^T 
 
 BY 
 
 EDWAKD CAIED, LL.D. 
 
 PROFEaSOE OF MORAL PHILOSOPHY IN THE tJNIVERSITY OF GLASGOW 
 LATE FELLOW AND TUTOR OF MERTON COLLEGE, OXFORD 
 
 IN TWO VOLUMES 
 VOL. I 
 
 NEW YORK 
 
 MACMILLAN AND CO. 
 
 1889 
 
Cornell University 
 
 The 
 
 f Sage School 
 
 of 
 
 Philosophy 
 
 Libraxy 
 
DEDICATED 
 
 TO THE MEMORY OF 
 
 GEOEGE EANKINE LUKE, 
 
 LATE SENIOR STUDENT AND TUTOR IN CHRIST CHUBOE, 
 OXFORD. 
 
lo veggio ben ohe giammai non si sazia 
 Nostro intelletto, se'l Ver non lo illustra, 
 Di fuor dal qual nessun vero si spazia. 
 
 Posasi in esso, come fiera in lustra, 
 
 Tosto che giunto I'ha ; e giugner puoUo ; 
 Se non, ciascun disio sarebbe/n/s^ra. 
 
 Nasoe per quello, a guisa di rampoUo, 
 Appi^ del vero il dubbio ; ed b natura 
 Ch'al sommo pinge noi di oollo in coUo. 
 
 Dante, Paradiso, iv. 123-32. 
 
PREFACE. 
 
 The object of this book is to give a connected view of the 
 Critical Philosophy, showing the relations of the three Critiques 
 to each other, and to the other works of Kant which may be 
 regarded as illustrations or developments of his main argument. 
 The first part, on the Critique of Pure Reason, deals with the 
 same subject as my former work, entitled The Philosophy of 
 Kant, but, except in a few passages, it is not a reproduction of 
 it. Since the date of its publication, many important contribu- 
 tions to the study of Kant have been made in Germany and other 
 countries. I wish especially to express my indebtedness to the 
 writings of Dr. Vaihinger, Benno Erdmann, Cohen, Paulsen, Ar- 
 noldt, Stadtler, Staudinger, and Eiehl ; also of the late Professor 
 Green and Professor Watson. In particular. Dr. Erdmann's publi- 
 cation of the Beflexion^n Kant's has thrown much new light on 
 the development of the Kantian philosophy, and Dr. Vaihinger's 
 acute statement of the main Kantian airoplai in his Com- 
 mentary is full of instruction and suggestion for every student. 
 I have sometimes given references to these and other writers ; 
 but often I have found it impossible to recall or to trace 
 what I owe to them, and must content myself in the main 
 with this general acknowledgment of my obligations. I also 
 
viii PREFACE. 
 
 owe much to the criticisms of my former book by various 
 writers, particularly by Dr. Hutchison Stirling, by Mr. Balfour, 
 and by Dr. Vaihinger ; and I have attempted to meet them by 
 giving a more carefuf and accurate statement of Kant's argu- 
 ment, and a fuller and more thorough estimate of its bearing 
 and value. In this way I have been led, among other things, 
 greatly to extend and modify my view of the Principles of 
 Pure Understanding. 
 
 The most important cause of the changes made in my former 
 representation of the Kantian doctrine has, however, been the 
 continued study of Kant himself, carried on mainly with the 
 view of tracing out the connexion of his different works. For 
 the very attempt to treat the Kantian philosophy as a whole, 
 and to show its unity and the method of its development, 
 made it necessary to restate the argument of the iirst Critique. 
 Thus, in the chapter on the Postulates of Empirical Thought, I 
 have entered at considerable length into the alterations made 
 in the second edition of the Critique, and I have endeavoured 
 to discover their causes. In doing so, I came to the conclusion 
 that these alterations are important, not, as has been main- 
 tained by Schopenhauer and others, because they show a 
 tendency in Kant to recoil to the point of view of the ordi- 
 nary common sense Eealism from an idealistic position similar 
 to that of Berkeley ; but because they indicate his progress 
 towards an Idealism in which the subjectivity of Berkeley's 
 theory is corrected. Here, in fact, we have one indication 
 among others that, as Kant advanced with his work, the ulti- 
 mate results of it came more clearly within his view, and even 
 had a certain reactive effect on his conception of the earlier 
 parts. Yet, on the other hand, we have to remember that the 
 
PREFACE. IX 
 
 work of Criticism was from the first conceived by Kant as a 
 whole, though in its execution it was divided into a number of 
 separate Critiques. The effect of this was apparently to give 
 undue prominence to special questions, and to hide for a time 
 their relation to other elements in the great problem which 
 Kant had set himself to solve. This concentration upon a 
 particular point has even at times affected Kant's own point of 
 view, producing a verbal contradiction between the statements 
 which he made at different stages of his work. But, in spite of 
 such inconsistencies, and of the reservations and cautions with 
 which he surrounds himself at every step in his progress, 
 I have attempted to show that there is an unbroken con- 
 tinuity in the movement of Kant's thought, and that the 
 lesson of his philosophy as a whole is definite and self- 
 consistent. 
 
 That lesson, however, he did not himself fully understand. 
 He suffered for his position as the discoverer of a new way of 
 dealing with the problems of philosophy. He had often to 
 invent his own language, or to use old terms in new senses ; 
 or, rather, we might say, he had to carry his readers through a 
 series of changes which collectively amounted to a complete 
 revolution of thought, and in doing so, he had again and again 
 to strain the language of a doctrine already received to inake it 
 express a new idea. Urged forward by a strong tendency to 
 the ideal, and by an original power of speculative insight which 
 was continually fertile of new views of truth, and held back by 
 a mind deeply imbued with the spirit of a sceptical age and 
 carefully trained in the school of Newtonian science, Kant had 
 a harder struggle with himself than he could possibly have had 
 with any critic or opponent of his philosophy. And, while he 
 
X PEEFACB. 
 
 expressed the result of his thought at each stage of its develop- 
 ment in the words that seemed most suitable, he seldom turned 
 back to compare them with what he had said before. It 
 might, without much risk of contradiction^ be contended that 
 he never repeated himself without introducing some new 
 modifying thought, which somewhat changed the aspect of his 
 previous statements. To understand him, therefore, is not 
 simply to combine different texts which exhibit the different 
 aspects of an unchanged thought. It is to detect a consistent 
 stream of tendency which, through all obstruction, is steadily 
 moving in one direction; to discern the unity of one mind 
 which, through all changes of form and expression, is growing 
 towards a more complete consciousness of itself. 
 
 In trying to discharge the task of a critic of Kant, it is 
 difficult not to seem " as if we were impertinently trying to ' pull 
 his work to pieces,' " (as Green remarked in relation to his own 
 criticism of Locke,) or to be ungratefully seeking an easy victory 
 over him from the vantage-ground of a later time — a vantage- 
 ground which he himself has helped to provide for us. But, as 
 Goethe has said, the main homage which a great man exacts 
 from those who follow him is the ever-renewed attempt to 
 understand him. And no one who recognises that progress in 
 speculative philosophy is a progress to self-consciousness, and 
 that such progress always involves a conflict between the con- 
 scious and the unconscious, even in the minds of those who are 
 its most prominent representatives, will fail to see that the 
 only valuable criticism is that which turns what is latent 
 in the thought of a great writer against what 'is explicit, and 
 thereby makes his works a stepping-stone to results which 
 he did not himself attain. It was those who stoned the 
 
PREPA.CE. XI 
 
 prophets that built their sepulchres. Those who really 
 reverenced them, showed it by following the spirit derived from 
 them to new issues. 
 
 One of the most difficult of the minor points which an 
 English commentator on Kant has to decide is the question of 
 the translation of technical terms. In most cases, owing to the 
 different genius of the language, only a compromise is possible. 
 This difficulty is found even in the most general word, used by 
 Kant for any mental modification whatever, the word Vor- 
 stellung. The term " idea " was employed in this sense by 
 Locke and most English writers on philosophy till a recent 
 date, and it seemed better to continue this usage than to adopt 
 any such formal term as ' representation ' or ' presentation.' 
 The main objection is that the term ' idea ' is employed by 
 Kant himself for another purpose, with distinct reference to its 
 original use by Plato. In this latter sense, wherever there was 
 any need of making the distinction, I have printed Idea with a 
 capital letter. Sometimes I have found it convenient to trans- 
 late Vorstellung by the word ' Consciousness.' Anschauung 
 1 have generally translated by ' Perception,' rarely by ' Intui- 
 tion,' as the term Intuition seems in English to carry with it 
 associations which are misleading. Sometimes I have used 
 ' Pure Perception,' where the context seemed to require it. 
 Wahrnehmung I have translated by ' Sense-perception,' where- 
 ever, as in the discussion of the Anticipations of Sense- 
 perception, it seemed necessary to call attention to its contrast 
 with Anschauung. For JBegriff I have invariably used ' Con- 
 ception.' 
 
 In quoting the Critique of Pure Season, I have given refer- 
 ences to the pages of Kant's first and second editions as A and 
 
xii PREFACE. 
 
 B. In most of the German editions and in Prof. Max Mtiller's 
 translation of the Critique, the paging of one or other of these 
 editions is given. In quoting from Kant's other works I have 
 referred to the editions both of Eosenkranz and Hartenstein as 
 R. and H. 
 
 The proofs of the whole of this book have been read by 
 Professor Jones, of the University College of !N'orth Wales, 
 and a considerable part of them by Mr. John S. Mackenzie, 
 of Trinity College, Cambridge. To both of these gentlemen I 
 have to express my obligations for numerous valuable sugges- 
 tions and criticisms. 
 
 Univkrsitt of Glasoow, 
 August, 1889. 
 
CONTENTS. 
 
 INTRODUCTION. 
 
 CHAPTER I. 
 
 THE IDEA OF CRITICISM. 
 
 ' The Age of Criticism — Different Uses of the term Criticism— Kant's Definition 
 of it in Distinction from Dogmatism and Scepticism — How Dogmatism 
 leads to Scepticism and Scepticism to Criticism — The Principle of Philo- 
 sophical Criticism — Locke's Idea of it — That it is not Psychological hut 
 Metaphysical — Its Possibility — Objection that we cannot go beyond our 
 own Ideas answered — The Criterion to be found in the Universal Presup- 
 positions of Knowledge — Our Unconsciousness of these Presuppositions 
 Explained — The a priori as the Common Ground of all Consciousness of 
 Objects^ — That the Critical Regress discloses this Ground — That all Science, 
 at its Constitution, presupposes such a Regress — Kant's View that at the 
 Present Day there is a Similar Necessity in the Case of Metaphysic — This 
 Problem opened up by the Modern Advance of Science and the Collision of 
 its Results with the Moral and Religious Consciousness — How this Collision 
 reflects back a Doubt even upon Science — That Science is Hjrpothetioal 
 because Abstract — That this is true even of the so-called Concrete Sciences 
 
 That Nature is not to be Explained apart from Spirit — That the Highest 
 
 must Explain the Lowest, and not vice versd — That, therefore, every 
 Advance of Science involves a new Critical Regress — Parallel between 
 Kant's Criticism and the earlier Regress made necessary by the Advance 
 from Mathematical to Physical Principles— Importance of the Modern 
 Problem, and Difficulty of the Antinomy which it Involves— Necessity of 
 * its Solution to avoid absolute Scepticism, . . .Pp. 1-44. 
 
xiv CONTENTS. 
 
 CHAPTER II. 
 
 KANT'S RELATION TO HIS TIME. HIS LIFE AND CHARACTER. 
 
 Place of Kant in the History of PhUosophy as the Author of a new Synthesis 
 which put an End to the Age of Enlightenment-General Characteristics of 
 the Enlightenment, its Bad and Good Sides-Its Necessity as a Stage in the 
 Development of Thought-The Protest of the Religious Consciousness in the 
 Age of Enlightenment— Kant's Relation to his Precursors-Defects of his 
 Method due to that Relation— Slowness of his Mental Development— His 
 Relation to P. A. Schultz— Influence of Pietism and of Classical Studies— 
 His Relation to Martin Knutzen— His Scientific Studies— His Life as a 
 private tutor, a Privat Docent, and a Professor— Political Interests— The 
 Seven Years' War, the American War, and the French Revolution— His 
 Philosophical Works— His Character— Heine's Account of him— His 
 Method in Teaching— Three Periods in his Mental Development, 
 
 Pp. 45-67. 
 
 CHAPTER III. 
 
 KANT'S PRECURSORS— DESCARTES — SPINOZA — LEIBNIZ AND 
 WOLFF— THE LATER WOLFFIAN SCHOOL AND MARTIN 
 KNUTZEN. 
 
 The Idea of Evolution as a Key to the Philosophies of the Past, and their Con- 
 nection with the Present — That it is not used by Kant, but helps us to 
 understand him — His Philosophy a Result of the Principle of the Refor- 
 mation, as it shows itself in Opposition to the Mediaeval Church and the 
 Scholastic Philosophy — Mode of reconciling Subjective Freedom and 
 Objective Truth adopted by Luther and Bacon — Subsequent Tendency to 
 dwell on the Subjective — Descartes' Cogito ergo mm — His Transition from 
 the Self to God by Removal of the Negative Element present in the 
 Finite — That his Principles lead logically to Spinozism — Spinoza's Hesita- 
 tion between the Negation of all Difference and its Reassertion in God — 
 Reassertion of the Finite in the Eighteenth Century — That its Subjec- 
 tive Individualism leads to Scepticism, as against which Kant reasserts 
 the Universal as presupposed in the Particular — That Leibniz prepares the 
 way by his Monads, which ideally include, really exclude, the World — ^The 
 Pre-Established Harmony — The Scale of Being — Man's Place in it — Rela- 
 tion of Sense and Thought — The Principle of Sufficient Reason — Incon- 
 sistencies of the Leibnizian Philosophy — Wolff's Attempt to Reduce it to 
 a consistent Individualism — His implicit Denial of the Constitutive Power 
 of Reason — The Pre-Established Harmony, reduced to a Harmony of Body 
 and Soul — Baumgarten's Modification of it — Knutzen's Attempt to unite 
 Leibniz and Newton, ..... Pp. 68-lO.S. 
 
CONTENTS. XV 
 
 CHAPTER IV. 
 THE PRB-CRITICAL PERIOD IN KANT'S MENTAL HISTORY. 
 
 Kant's Early Essays on Physics — His Attempts to Reoouo'ile Descartes 
 with Leibniz and Leibniz with Newton — That he starts, like Leibniz, 
 with the Individual Substance or Monad, but admits the Negative and 
 Positive Relativity of the Monads — His first Metaphysical Treatise, the 
 Dilucidatio Nova (1755) — The Ratio antecedenter determinans, and the 
 Ratio conseque.nte.r determinans — The Real and the Ideal Ground — Question 
 whether there is a Real Ground for Everything — That there can be no 
 such Ground for God — Kant's Modification of the Ontological Argument — 
 That there is a Real Ground for Changes in the Einite World — Principles 
 of Succession and Co-existence — Kant's Compromise between Individualism 
 and Pantheism — His Imperfect Development of the Idea of the Relativity 
 of Individual Substances — His Approximation to the Spinozistic Idea of 
 God — Import of the Change of Form in the Ontological Argument — Sum- 
 mary of the Dilucidatio Nova — Break with Wolffian Philosophy in the 
 Works published in 1762-3 — The False Suhtilty of the Syllogistic Figures — 
 That it shows the Movement of Thought to be purely Analytic, and 
 carries us back to many Indemonstrable Principles — The only Possible 
 Basis for a Proof of the Being of God — The Question whether Existence 
 can be included in the Conception of an Object — Difference of Absolute 
 and Relative Position, and of Reality and Possibility — That God is for 
 Kant the Unity of Reality and Possibility, Being and Thought, but not 
 the Unity of all Affirmative Predicates — Distinction of Logical and Real 
 Negation — Essay towards the Introduction of the Idea of Negative Quantity 
 into Philosophy — Real Position = Negation of an Opposite Position — The 
 World as a Unity of Opposites — The Analytic Movement of Thought con- 
 trasted with the Synthetic Movement of Knowledge — Question whether 
 Kant in this Essay follows Hume — That Kant does not yet deal with the 
 General Problem of Causality — Transition to the Problem, whether 
 the Synthesis, Impossible to Thought, is Possible to Sense— Treatise Ore 
 tlie Eoidence oftlie Principles of Natural Theology and Morals— BiSerence 
 between the Methods of Philosophy and Mathematics— Agreement of 
 Kant with Locke's Doctrine as to Mathematics — Kant's View as to the 
 Relation of Mechanism and Teleology — Summary of the Results reached 
 by Kant in 1764 — Prominence given to the Question of Method— Essay 
 Ore the Dreams of a Ghost-Seer, as illustrated hy the Dreams of Metaphysic 
 (1766) The Case for Spiritualism and for the Empirical Rejection of Spir- 
 itualism—Critical Solution of the Problem— Philosophy conceived as the 
 Theory of the Limits of Knowledge— Experience as the Source of Know- 
 ledge—Morality asserted to be independent of Metaphysics— Relation 
 of the Critical Empiricism of this Essay to the later Critical Philosophy- 
 Mendelssohn's Objections and Kant's Reply— End of the Pre-Oritical 
 Period about 1768 Pp- 104-160. 
 
xvi CONTENTS. 
 
 CHAPTER V. 
 
 THE DISSERTATION OF 1770, AND KANT'S STUDIES EOR THE 
 CRITIQUE OF PURE REASON IN THE YEARS 1770-81. 
 
 Kant's Reconciling Tendency and his Polemical Method — Kant's Letter to 
 Garve on his Philosophical Awakening by the Antinomies of Reason — Re- 
 lation of the Polemical Method to Criticism — Essay On the Rational Basis 
 for the Distinction of Regions in Space (1768) — Priority of Space Involves 
 its Apriority as a Condition of Perception — Dissertation on the Form and 
 Principles of the Sensible and Intelligible World (1770) — The World as Per- 
 ceived, and the World as Conceived — Opposition of Sense and Thought, 
 of the Phenomenal and Noumenal Worlds — Subjectivity of Sense and 
 Objectivity of Intelligence — Distinction of the Form and Matter of Sense 
 — Distinction of the Logical and Real Use of the Intelligence — General 
 Conception of the Absolute Reality as it Exists for the Intelligence — That 
 it cannot be Realised in Particular through Sense — That, nevertheless, 
 the Things of Sense also are Objects of Knowledge — Proof that Time and 
 Space are Forms of Sense — Nature of the Priority of Time and Space — 
 Their Continuity — Comparison of the Newtonian and Leibnizian Ideas of 
 Time and Space— Difference and Relation of Time and Space — That Pure 
 Conceptions are Necessary for the Determination of Objects in Relation to 
 each other as Parts of one World — Kant's View of the Results of the Dis- 
 sertation — Tliat he Explicitly Determines the Perceived World as Pheno- 
 menal from the Point of View of Intelligence, but also Implicitly 
 Condemns the Conceived World as Unreal from the Point of View of 
 Perception— Source of this Difficulty— Two Ways of Solving it by the 
 two Methods of Abstraction and Concretion — That Kant seems to Adopt 
 the Former, but that his Results often Imply the Latter— Criticism 
 Possible only on the Latter Method— How, from this Point of View, 
 the Relativity of the Object to the Subject must be understood— Differ- 
 ences between the Dissertation and the Critique of Pure Reasoji— 
 Kant's Lfetter to Herz (1772) — Question whether Pure Conceptions 
 can be Objectively Valid— Unsatisfactoriness of the Dogmatic Answers 
 —Answer of the Critique Anticipated— Kant's Position at this 
 Time — Question, when Hume broke Kant's Dogmatic Slumber — 
 Hmne's Discussion of Causality, and Kant's Understanding of it 
 —Kant's Search for a Systematic List of the Categories— That he 
 rejects Aristotle's List, and derives another List from the Logical 
 Analysis of Judgment— Substitution of the Self for God as the Principle of 
 Unity in Knowledge— Relation of the Dogmatic and Critical Explanation 
 of Knowledgei-Elemeut of Truth in the Dogmatic Explanation- Sense in 
 which Kant admits an Ultimstte Unity of Thought and Being— His use of 
 the Judgment and the Syllogism, as Analysed by Formal Logic, in the 
 Discovery of the Categories and the Ideas respectively— Relation of 
 Synthetic to Analytic Judgment, and of Synthetic to Analytic Syllogism 
 —Kant's View of the Imperfection of the Synthetic Judgment, and the 
 Failure of the Synthetic Syllogism— Source of the Kantian Opposition of 
 Thought to Knowledge— Summary Account of the Development of the 
 Critical Philosophy before the Critique of Pure Reason, Pp. 161-226. 
 
CONTENTS. xvii 
 
 BOOK I. 
 
 THE CRITIQUE OF PURE REASON. 
 
 CHAPTER I. 
 THE PEOBLEM OE THE CRITIQUE OE PURE REASON. 
 
 Different Views of the Problem— That the Critique of Pure Reason was origin- 
 • ally part of a Larger Scheme, which LQcluded the Suhjeots of the other 
 Oritiqms— Results of the Aesthetic, the Analytic, and the Dialectic — 
 Relation of the Critique of Pure Reason to the other Critiques — Eirst 
 Form of the Problem of the Critique of Pure i?easo»— That Doubt of the 
 a priori Principles of Knowledge arises through their being used beyond 
 the Limits of Experience — Necessity thence arising that the Conditions of 
 their Validity should be Discussed — Limitations of the Doubt which the 
 Critique seeks to.Remove — That Kant seeks to Limit a priori Knowledge 
 to Experience, in order that he may set free what is beyond Experience 
 from Empirical Conditions — Question whether there is any Change in Kant's 
 View of the Relation of Knowledge to Thought in his Later Works — 
 That the Problem of Kant is itself Dialectical^That in its Second Form 
 it is a Proof that there is a priori Knowledge, and that through it alone 
 a posteriori Knowledge is Possible — That Kant's View of Thought as Op- 
 posed to Knowledge Limits his Argument — The Distinction of Analytic 
 and Synthetic Judgments as Suggested in Different Ways by Locke and 
 by Wolff— ^Influence of Hume — That neither Thought nor Sense by itself 
 Explains S^thesis — The Difficulty of Mathematical Synthesis as Treated 
 by Locke and by Hume — Relations of Ideas and Matters of Fact — Kant's 
 Understanding of Hume — His View of Mathematical and of Dynamical 
 .Synthesis, their Agreement and Difference — A priori Synthesis as Treated 
 in the Aesthetic and the Analytic respectively — The Unity of the Self as 
 Manifesting Itself in the a priori Principles of Synthesis — That the 
 Analytic Unity of Thought cannot Supply such Principles, though it 
 Furnishes a Clue to them — Relations of the Synthesis of Understanding 
 to that of Reason — The Deduction of the Categories, and their Limitation 
 to Experience — Union of a priori Perceptions and Conceptions in Ex- 
 perience—Impossibility of a Pure Synthesis of Thought— That in 
 Synthetic Judgment the Mind goes beyond itself, and the ideas it has 
 already United with the Consciousness of Itself— The Possibility of a 
 Pure Analytic Judgment — Importance of the Distinction of Synthesis and 
 Analysis for Kant— Nature of the Ultimate Synthesis of Thought and 
 Perception— Difference of Mathematical and Dynamical Principles— Per- 
 ception and Conception as Elements, not Kinds of Knowledge— How it 
 was necessary for Kant to maintain both the Distinction and the Relation 
 of Perception and Conception — Difference in the Deductions of the Mathe- 
 matical and the Dynamical Principles- Summary of Kant's Views as to 
 the Problem of Criticism, Pp- 227-280 
 
XVlll CONTENTS. 
 
 CHAPTER II. 
 
 THE AESTHETIC. 
 
 Relation of the Aesthetic to the Dissertation— Timculty caused by Confusion/ C 
 of the Different Stages in Kant's Critical Regress— Relation of the AestJietu: 
 to the Analytic— Tha.t the Distinction of Sense and Thought in the Aesthetic 
 is the Presupposition of the Problem as to their delation in the Analytic— 
 That this Problem involves two Subordinate Problems which are dealt 
 with in the Analytic and Dialectic respectively- The Metaphysical and 
 Transcendental Exposition of Space and Time— Their A -priority proved by 
 their Priority— That this proof is directed against Newton, not against 
 Hume— Proof that Space and Time are Perceptions, not confused Concep- 
 tions, as Leibniz believed— In what sense Space and Time are spoken of by 
 Kant as Infinite Given frAo^cs- Correction of the inaccuracy of this State- 
 ment in the Dialectic— Tha.t the view of Space and Time as a priori Eorms 
 of Perception explains the possibility of Mathematics— How Mathematics 
 can anticipate Experience— How far the Aesthetic modifies the ordinary 
 ' Dogmatic View of Things— Empirical Reality and Transcendental Ideality 
 of Space and Time— Sense in which Kant maintains the Distinction of the 
 Primary and Secondary Qualities of Matter— Lambert's Objection to the 
 Ideality of Time— Kant's View of the Controversy between the New- 
 tonians and Leibniz — Trendenburg on Kant's Dilemma — Impossibility of 
 escaping it onKant's Presuppositions — Change of Kant's View of Perception 
 in the Analytic — The Question raised how Sensations become Perceptions 
 — That it is not merely by means of the Forms of Perception — That the 
 unconscious Synthesis of Imagination in Perception and the conscious 
 Synthesis of Understanding in Conception must be combined — How these 
 two Processes are related — Knowledge conceived in the Analytic and 
 Dialectic as a Process of Synthesis that cannot be Completed — New Con- 
 ception of the Relation of the Phenomenon to the Thing in itself thence 
 arising — Summary View of the Problem of the Analytic and the Dialectic. 
 
 Pp. 281-319. 
 
 CHAPTER III. 
 
 THE TWO LOGICS AND THE DISCOVERY OR METAPHYSICAL 
 DEDUCTION OF THE CATEGORIES. 
 
 That the Analytic and Dialectic involve a New Reflexion on the Presuppositions 
 of the Aesthetic — The Distinction of the Metaphysical and Transcendental 
 Deduction — Use of Formal Logic in the former — The Analytic Processes 
 of Judgment and Syllogism, and their Difference from the corresponding 
 Synthetic Processes, which are based on a priori Conceptions and Ideas 
 — Difference of Conceptions and Ideas — Both Imaginative and Conceptual 
 Synthesis necessary for Synthetic Judgment — Kant's First Statement of 
 the relation of Analytic and Synthetic Judgment — Difficulties in it — Kant's 
 Derivation of the Transcendental from the Logical System — How far it 
 corresponds to Kant's own Idea of a Metaphysical Deduction — That a purely 
 
CONTENTS. xix 
 
 Formal Judgment would disappear in the simple identity of Subject and 
 Predicate— That neither Judgment nor Syllogism is ever simply Analytic 
 —Consequences of Kant's View of Pure Thought as Analytic— That Kant 
 is obliged to Transform the Logical System in order to adapt it to his 
 Purpose— That the three Categories in each Class are not Exclusive Alter- 
 natives, and that Kant does not treat them as such— Distinction of the 
 Mathematical and Dynamical Categories — That there are really Three 
 Classes of Categories, because Three Stages of Reflexion— Kant's gradual 
 Advance towards an Organic System of Categories, . Pp. 320-346. 
 
 CHAPTER IV. 
 
 THE TRANSCENDENTAL DEDUCTION OP THE CATEGORIES. 
 
 Relation of the Metaphysical to the Transcendental Deduction- Change of the 
 latter in the Second Editions of the Critique — Necessity of the Categories for 
 a Consciousness of Objects as such — Self-Consciousness possible only through 
 aconneoted Consciousness of Objects — Pre-established Harmony of Percep- 
 tion with Conception — Kant's Illustrations of it — That all our Ideas must 
 be Consistent with the Possibility of Self- Consciousness — That this implies 
 a Conformity of all Perception to certain a priori Synthetic Principles — 
 Kant's Interpretation of the Pre-established Harmony of Leibniz — General 
 Purport of the Deduction in the First Edition— The Synthesis of Appreheu- 
 , sion in Perception, of Re-production in Imagination, and of Recognition in 
 , Conception — That the two Former are conformed.a priori to the Latter, as 
 necessary to the Possibility of Self-Consciousness — General Relation of the 
 Consciousness of Objects to Self- Consciousness — That reference to an Object 
 means Subsumption under a Rule, and that this Explains the Constraint 
 which the Object seems to put upon our Consciousness — That the Subjective 
 Unity is prior to the Consciousness of Objects, but the Consciousness of that 
 Unity is posterior to it — Difficulty as to the priority of the Synthesis of 
 the Imagination to the Synthesis of Recognition — That it is not Priority in 
 Time — Similar difficulty as to the Relation of Opinion to Science in Greek 
 Philosophy — Plato's Solution of it — That Knowledge of the Particular is 
 through the Universal does not imply that the Universal is present as such 
 to Consciousness — Kant's first Solution of the Difficulty as to the Relation 
 of Perception and Conception: that Conception is necessary to Generalise 
 Perception, and therefore, to Objectify it — His Second Step : that even a 
 Particular Judgment involves Conception, and therefore no Judgment of 
 mere Perception is possible — Question whether even the Consciousness of 
 Particular Images is possible without the Determination of these by Con- 
 ception — That even Consciousness of an Appearance, as such, involves a 
 Judgment — That Perception cannot be the Consciousness, or even the 
 Presence, of an Image, if isolated from Conception — ^That, on the other 
 hand, a Synthetic Principle cannot be derived from Pure Thought— That, if 
 Conception and Perception are thus absolutely Divided.Knowledge becomes 
 impossible— That, therefore, they can exist only as Factors in the Unity 
 of the Judgment — Imagination as the Mediator between them— Kant's 
 way of reaching Unity from a Pre-supposed Difference — The Pre-estab- 
 lished Harmony of Conception and Perception carried back to Unity — 
 
XX CONTENTS. 
 
 That Perception is not prior to Conception in the Sense it has as Subsumed 
 under Conception— Sense in which it is Prior to Conception— Kant's 
 Imperfect View of the Organic Unity of the Intelligence, as connected with 
 his Negative View of the Return of Consciousness upon itself— Question 
 whether his Ideal of Knowledge is an Abstract Identity which is neither 
 Conception nor Perception, or an Intuitive Understanding which is both— 
 That if the latter alternative be adopted, Pure Thought cannot be Analytic 
 —That Kant's Argument implies that it is Synthetic— How the recognition 
 of a Synthetic Movement in Self-Consciousness leads to a Transformation of 
 our View of Objects and of the Self— The Dualism of Descartes rejected 
 by Kant, though only in relation to Phenomena— That the Regress upon 
 the Unity of Self-Consciousness ought to be recognised by Kant as a 
 Progress — Consequences of so regarding it — How this View of it throws 
 Light upon the Deduction — Relation of Inner and Outer Sense, as exhibited 
 in the Second Part of the Deduction— The Determination of Inner Sense 
 by the Categories — That Kant treats Inner Experience as more abstract 
 than Outer, but enables us to see that the Opposite is the Truth — 
 Partial Survival in Kant of the Error of Berkeley, which he yet enables 
 us to Correct — The Development of the Consciousness of Objects not to be 
 separated from the development of Self- Consciousness, though logically 
 Self -Consciousness is Posterior — Summary of the Argument of the First 
 and Second Part of the Deduction and of the Results of the above 
 Criticism of them, Pp. 347-430. 
 
 CHAPTER V. 
 
 THE SCHEMATISM OF THE CATEGORIES. 
 
 The Categories as Principles for the Determination of aU Objects in Time — The 
 Judgment of Experience defined as the Subsumption of an Image of 
 Perception under a. Schematised Category — Necessity of the Schema — 
 The Distinction between a Schema and an Image — How the Schema 
 determines the Conception — Schemata of Pure Conception — The Schemata 
 as supplying what is wanting in Kant's View of the Relativity of Per- 
 ception and Conception — How Kant's Argument leads to an Alteration 
 of his Premises, and so to a truer View of the Relation, of Conception 
 V and Perception in Knowledge — Real Meaning of the Assertion that the 
 Categories are Species of Apperception — Relation of the Kantian and 
 Hegelian Logic — Steps in Kant's Transition from the Analytic to the 
 Synthetic Unity — How the Third Category in Quality and Quantity 
 comes into Existence in the Process of Schematism — Why Totality appears 
 as Degree, and Limitation as Number — Reason for the Imperfect Form 
 in which these Categories appear — Contradiction involved in this View 
 of the Synthesis represented by the Third Category — The Categories 
 and Schemata of Relation — That they carry us beyond the Analytic 
 View of Thought — The Peculiar Character of these Categories and 
 their Importance in the Development of Kant's Philosophy — Origin of 
 A the Necessity expressed in them — Peculiarities of the Categories of 
 Modality — Impossibility of Explaining them on Kant's Principles— That 
 Kant, by his View of the Schematism, imperfectly supplies a Defect in 
 his Theory of Knowledge— That Judgment involves Three Momenta, ex- 
 
CONTENTS. xxi 
 
 pressed in the Three Classes of Categories — That, therefore, it always 
 Involves all the Categories, though it depends on the Predicate which of 
 them is made Explicit — Judgments of Simple Position, of Relation or 
 Reflexion, and of Modality or Ideal Unity — When the Matter of Judg- 
 ment is adequate to its Form— How the Categories of Modality lose 
 Objective Validity for Kant — Reasons for regarding them as the only 
 Adequate Categories of Reality — Transition from the Schematism to the 
 Principles of the Pure Understanding, . . . Pp. 431-470. 
 
 CHAPTER YI. 
 
 THE SYSTEM OP PRINCIPLES OP THE PURE UNDERSTANDING. 
 
 Object of the Deduction of these Principles — Kant's Contrast of the Principles 
 of Thinking with the Principles of Knowing— Relation of the Schematism 
 to the Deduction of the Principles — Logical Character of this Deduction 
 — How Experience can supply a Basis for the Deduction of the Principles of 
 its own Possibility — Ambiguity in the Argument as stated by Kant — That 
 it is partly caused by the Ordinary Use of those Principles with the Con- 
 scious Use of them — That Science involves a Transcendental Regress, 
 which, however, is not carried so far back as Kant's Regress — Kant's Con- 
 fusion of the Transcendental Regress with a Psychological Account of the 
 Genesis of Experience out of independent Paotors-^That if we take away 
 the Principles we reduce Experience to a Chaos of Sensations, while, if 
 we Reflect on their Presence in Experience, we raise it into the Porm of 
 Science — The Mathematical and the Dynamical Principles — Necessity of 
 further Distinguishing the Postulates from the Analogies, Pp. 471-488. 
 
 CHAPTER VII. 
 THE MATHEMATICAL PRINCIPLES. 
 
 Nature of the Schemata of Number and Degree, as stated in Chap. V.— General 
 Purport of the Deduction of the Mathematical Principles — Deduction of 
 the Principle of the Axioms of Perception— That Objects are known as in 
 Space and Time only by a Synthesis under the Schema of Number, and 
 by the Consciousness of that Schema — Positive and Negative Aspect of 
 this Deduction— Imperfection of the Determination of Objects as Exten- 
 sive Quanta— Deduction of the Principle of the Anticipations of Percep- 
 tion—Three Points to be Distinguished in that Deduction- That Positive 
 and Negative are Viewed by Kant as reciprocally Exclusive in Pure 
 Thought, but not in Knowledge, and therefore as reciprocally Exclusive in 
 Things in Themselves, but not in Phenomena— Application of this Principle 
 to Matter in Space in the Metaphysical Rudiments of Physics— The Union 
 of Opposites as Mediated by Space and Time— Difficulty of Combining 
 Kant's different Statements as to Limitation and Degree— The True Rela- 
 tion of the Porms of Time and Space to the Categories of Totality and 
 Limitation— Criticism of Kant's View of the Relation of the Ideas of 
 Extensive and Intensive Quantity— That the latter is necessarily related 
 to the fbrmer- Kant's Criticism of the Explanation of Degree "by more or 
 
xxii COHTBNTS. 
 
 less of racmim— That the Principle of the Anticipations of Sense-Percep- 
 tion iustifies the Application of the Higher Mathematical Calculus to 
 Experience, Pp. 489-513. 
 
 CHAPTER VIII. 
 
 THE ANALOGIES OF EXPERIENCE. 
 
 Why Kant found it necessary to Deduce the Mathematical Principles before 
 dealing with the Deduction of Causality and the other Analogies— Why 
 he requires a Special Deduction of the Analogies— Explanation of the 
 Statement that the Analogies determine Objects as Existing— WhaX Kant 
 seeks to Prove— Necessity of the Analogies to Universalise the Determina- 
 tion of Phenomena in Time— Ambiguity in Kant's Statement of this 
 Necessity -Meaning of his Statement that the Principle of Causality can- 
 not be Analytically Derived from the idea of an Event— That it Implies 
 the Reduction of an Event to a Succession of Impressions— That of the 
 three Determinations of Time only Sequence can plausibly be Derived 
 from our Impressions— Hume's Treatment of the Analogies in the Treatise 
 on Human Nature—How, in the Inquiry, he makes the Problem easier 
 for himself by Confining it to Causality — That Kant, on the other hand, 
 Deals with the Three Analogies as essentially Connected with each other 
 — Characteristics of the Transcendental Deduction of the Analogies — The 
 Analogies as Conditions under which Objects as Existing in Time, and the 
 Sequence and Co-existence of , their States, can become known to us 
 — Ambiguity in this Deduction— How it Corrects our first Abstract View 
 of Objects — Particular Objects in Particular Time-relations known through 
 Universal Principles — How the Consciousness of these Principles adds to 
 our Previous Knowledge — Position and Negative Aspects of the Dedtiction 
 — Statements of Principle of Substance in the First and Second Editions 
 of the Critique — That the Statement of the Second Edition agrees more 
 exactly with the Form of Kant's Deduction — Reasons for that Form (1) in 
 Hume's Statement of the Problem of Causality ; (2) in the View of Time 
 V given in the ^es^/icfe'c — Meaning of the Term 'Analogy' — Defect of the 
 Deduction — How Kant seeks to Remove it in the Second Edition by 
 Introducing the Determination of Objects as in Space — That only the 
 Object of External Sense can be regarded as a permanent Substance — 
 Changes in the Deduction of the Principle of Substance which this would 
 Necessitate — That the Principle of Substance is Deduced merely as the 
 Pre-Supposition of the Principles of the other Analogies, because Substance 
 is Nothing apart from its Accidents — That this Limitation of the Principle 
 of Substance is Regarded as due, not to the Laws of Thought, bxit to the 
 Sensuous Conditions of Knowledge — Error in this View due to Kant's 
 Conception of Pure Thought as purely Analytic — That the Conception of 
 a Thing-in-Itself apart from Relations, is due to the Abstract Nature of 
 our first Consciousness of Objects ; and that, therefore, it should be Cor- 
 rected by going onfrom Substance to the other Categories of Relation — State- 
 ment of the Principle of Causality in the First and Second Editions of the 
 Critique — That the Statement of its Relation to the Principle of Substance 
 is already a Partial Answer to Hume — That the Succession of Perceptions 
 
CONTENTS. xxiii 
 
 cannot originate the Consciousness of Succession, any more than that of 
 Objective Permanence or Co-existence — That Objects are not Perceived 
 by us as Changing, except through a Synthesis of the Imagination and the 
 Recognition of the Principle of that Synthesis— That it is the Principle of 
 Causality by which the Imagination is Bound to a Definite Order of 
 Succession in the Perception of an Event as such — That the Principle of 
 Causality can be Derived from Phenomena only because the Understand- 
 ing has previously Determined them by it — Double Aspect of it : (1) that 
 Causality is not to be got by Analysis of the idea of an Event ; but yet 
 (2) that an Event cannot be Known by us as such, except through the 
 Principle of Causality — Explanation of the Apparent Contradiction in this 
 Deduction — Special Difficulty in the Deduction of Causality, due partly to 
 Kant's Absolute Distinction between Analytic and Synthetic Judgment, 
 and partly to the Separation of Causality from Substance — Cause and 
 Substance as Different Phases of One Conception — Schopenhauer's Objec- 
 tion to Kant's View of the Relation of Succession to Causality — Kantian 
 Answer — How Casvial Sequence is to be Explained on Kant's Principles — 
 That Causality cannot be taken as Directly Applying to Inner Experience 
 — That, in Relation to Outer Experience, the Relativity of Motion 
 implies that Causality must ultimately be regarded as Reciprocity — 
 Kant's Statement of the Principle of Meciprocity in the First and Second 
 Editions — The Commercium of Substances, as represented in the Disserta- 
 tion and the Critique — That Reciprocity is Implied in Co-existence as 
 Known — Positive and Negative Aspects of this Deduction — Reciprocity as 
 the Analogon of the Principle of the Disjunctive Judgment — That all the 
 Three Principles involve a given Difference, which has to be Reduced to 
 Unity ; that, therefore, they contain a Contradiction which can be Solved 
 only by a Higher Principle, .... Pp. 514-587. 
 
 CHAPTEE IX. 
 
 THE POSTULATES OP EMPIRICAL THOUGHT— KANT'S VIEW OF 
 
 IDEALISM, AND OF THE RELATION OF INNER AND 
 
 OUTER SENSE. 
 
 Import of the Categories of Modality — That in the Postulates of Empirical 
 Thought, the Results of the Analytic are Summed up — Difficulty as to the 
 Relation of Possibility and Actuality— Question whether it is Perception 
 that enables us to refer our Conceptions to Objects, or Conception that 
 enables us to refer our Perceptions to Objects — Why both these Modes of 
 Expression can be justified by Kant— Why Kant never completely Solves 
 the Contradiction between them— That Conception Pre-supposes and Tran- 
 scends Perception, as Self-Consciousness Pre-supposes and Transcends 
 the Consciousness of Objects- Reasons for the Imperfect Development 
 of this Idea in Kant— That the Principles of the Possibility of Experience 
 cannot themselves be regarded as merely Possible— That the Modal 
 Principles must be regarded as Expressing the Organic Unity of Objects 
 with each other, and with the Intelligence— Why Kant rejects this View- 
 General and particular Contingency of Objects in Relation to the 
 
 • Intelligence— That, the Idea of the General Contingency of Objects is 
 
xxiv CONTENTS. 
 
 modified in the Second Edition by a new Conception of the relation of Inaer 
 and Outer Experience-That Lambert's Difacultyas to Time, and ako the 
 Difficulty as to the Connection of Mind and Body, are Answered in the 
 First Edition by equalising Inner and Outer Experience-That Inner 
 Experience cannot be Co-ordinated with Outer Experience, as it is the 
 same Experience in a, further Keflexion-Kant's gradual Discovery of 
 this-His View as to the imperfectly Scientific Character of Psychology— 
 His Twofold Answer to those who treated him as a Supporter of the 
 Berkleian Idealism: (1) That Transcendental Idealism vindicates the 
 empirical Keality of the Objects known as Phenomena ; (2) That it does 
 not affect the Distinction between Phenomena and Things in Themselves 
 —Germs in the First Edition of the View that Outer Experience Pre- 
 supposes Inner Experience— Its Development in the Second Edition- 
 Kant's Final Analysis of Experience— How it is to be brought in Belation 
 to his Treatment of the Principles of Pure Understanding— His Admission 
 that the Deduction of these Principles requires them to be viewed 
 Primarily as Principles of the Possibility of External Experience— Tho 
 Refviation of Idealism— Ajnhigaity of the way in which it opposes ideas 
 to Objects- Logical Defects of the Eefutation—ln what sense it can be 
 said to treat Outward Objects as Things in Themselves— How Kant 
 
 * proceeds from the Determination of the External Object as an idea, to 
 the Determination of it as an Object in Distinction from our ideas— That 
 the Defect of the Argument is due to Kant's Confusion of the Eegressive 
 Method with the Method of Abstraction- That ideas cannot be treated 
 as States of the Thinking Subject, nor the Inner Life as a series of 
 such States— The Mistake of Berkeley, and Kant's Correction of it— Inner 
 
 K Experience — Outer Experience, considered in Eelation to Feeling and 
 Thought — Change of our View of the External World when we regard it as 
 the Environment of a Sensitive Organism — Further change when we view 
 it as a World of Objects existing for a Subject, and as the means whereby 
 its Self -Consciousness is Developed — Kant's View of Self-Consciousness, and 
 its Relation to the Idea of the Noumenon — Necessity of a Principle to 
 combine Analysis and Synthesis, Thought and Knowledge — That we cannot 
 stop with the Determination of the Real as Empirically Necessary, but must 
 go on to Eecognise ihe Self as the Principle of Unity implied in the Con- 
 sciousness of Objects, and realised in Self-Consciousness — The Correction 
 of our First Consciousness of Objects under the Mathematical Principles by 
 the Analogies ; and of the Consciousness of them under the Analogies by 
 the Principles of Modality or Ideal Unity. . . Pp. 588-654. 
 
THE CRITICAL PHILOSOPHY. 
 
 INTRODUCTION. 
 CHAPTEE I. 
 
 THE IDEA OF CRITICISM. 
 
 " rpUE present age may be characterised as the Age of Criti- Necessity of 
 
 I , , . Criticism. 
 
 cism, a criticism to which everything is , obliged to 
 submit. Eeligion, on the ground of its sacredness, and Law, on 
 the ground of its majesty, not uncommonly attempt to escape 
 this necessity. But by such efforts they inevitably awaken a 
 just suspicion of the soundness of their foundation, and they 
 lose all their claim to the unfeigned homage paid by reason to 
 that which has shown itself able to stand the test of free 
 inquiry." 
 
 What is meant bv Criticism in these words of Kant, and wtat is criti- 
 
 •^ cism ? 
 
 what reason had Kant for regarding Criticism as specially the 
 characteristic of his own age ? Not unfrequently the term 
 " Criticism " is applied to a process which has scarcely any 
 rule or principle, a process which consists simply in raising 
 manifold objections from any point of view that may suggest 
 itself to the theory or doctrine criticised. Such criticism may 
 have a certain relative value because it awakens the mind from 
 an attitude of passive reception and stimulates it to play 
 freely round the subject in hand. But it is not a scientific 
 » process. It has no definite standard of judgment which it 
 VOL. I. A 
 
2 INTRODUCTION. chap. 
 
 consistently applies. Hence its use is to prepare for a better 
 method ; and if it does not make way for such a method, it 
 soon passes into a sophistry which can prove or disprove any- 
 thing, just because it has no principle to which it steadfastly 
 adheres. It is criticism without a criterion, and in the end it 
 throws light upon nothing, except perhaps upon the mind of 
 the critic. Even upon this it throws light only in so far as it 
 enables us to penetrate to the unconscious presuppositions on 
 which his judgments are based. 
 Do^aS.*" ^o"^ ^^"^^ undoubtedly meant something quite different 
 from this when he called his own age the age of Criticism, and 
 when he spoke of his own Critique as carrying out the work 
 of the age to its legitimate result in the province of philosophy. 
 What he meant we may best understand if we consider how 
 he opposes Criticism to two other forms of philosophy. Dog- 
 matism and Scepticism. " Dogmatism,'' he declares, " is the 
 positive or dogmatic procedure of reason without previous 
 criticism of its own faculty;" i.e., it is a system which is pro- 
 duced in the direct effort to understand and interpret the world 
 — the effort of a mind which is as yet troubled by no scruples 
 as to its own competence, or as to the efficiency of the methods 
 and principles it uses. Such a mind, indeed, is generally un- 
 conscious of any method or principle whatever. It is too busy 
 with its object to attend to itself An early philosopher is 
 described by Aristotle as looking up at the expanse of heaven, 
 and declaring that "all is one." So by a direct effort of 
 intuitive thought, the mind which as yet is troubled with no 
 doubts as to the possibility of knowledge, seizes upon some 
 general principle that seems to be as wide as the universe 
 itself, and uses it to explain, or to explain away, all ap- 
 pearances. Such immediate, unhesitating action of the 
 intelligence does not of necessity fail of a good result. Nay, 
 it is to such action that man's first insight into the nature of 
 things is always due. But it invariably, in the first instance 
 at least, overshoots its mark. Lighting up one aspect of things 
 
!• THE IDEA OP CRITICISM. 3 
 
 with the vividness of intuitive presentment, it leaves the other 
 aspects in the shade. Grasping a principle of limited range, it 
 applies that principle fearlessly to objects which it cannot 
 explain, and which, therefore, it only serves to distort. Especi- 
 ally is this apt to be the case with minds of little originality 
 • which work by an impulse, and on the lines of a thought which 
 they have received from another. For men of creative and 
 original insight have generally a sense of the whole, a con- 
 sciousness of the unity of things, which restrains them from 
 unseasonable applications of a principle, though theoretically 
 they may have laid it down without qualification or limitation. 
 It is not from the Newtons or the Darwins that we get abso- 
 lute mechanical explanations of the universe, or scholastic' 
 schemata of the whole possibilities of animal existence. But 
 there are many men of a high, though not the highest, order of 
 intelligence who advance unhesitatingly on any intellectual 
 road on which they have once set out, and are able to render 
 great service to science so long as that road leads to anything, 
 i.e., so long as they apply their method only to objects to which 
 it is adequate. Yet it is often the fate of such men, by the 
 very fearlessness and good faith with which they apply their 
 key to everything, to awaken the first doubt whether it can 
 open every lock. Pressing on under what Goethe called the 
 "daemonic" influence of the idea that possesses them, they try 
 to bring every region of existence under its sway, until the 
 common intelligence refuses to follow them in their Procrustean 
 treatment of facts. Thus the consciousness that a principle is 
 not universal arises out of the very attempt to universalise it, 
 and it is well if the recoil of thought does not awake scepti- 
 cism as to its value and truth even within its own limited 
 
 sphere. 
 
 The direct dogmatic or uncritical use of the understanding ^^^™ °*^ 
 is sure at some point to find itself checked and thwarted by 
 the nature of things. For the simple principles which first 
 •present themselves for the explanation of the worid are neces- 
 
4 INTRODUCTION. chap. 
 
 sarily imperfect and one-sided. If they explain phenomena, it 
 is only within a limited range, and when they are extended 
 beyond that range they come into contradiction with facts and 
 even with themselves. The category which forms a sufficient 
 guide so long as it is applied to the investigation of one 
 definite part of the world or one definite phase of reality, is 
 found inadequate when it is employed as a universal principle. 
 Hence, one^sidedness here calls forth an opposite one-sidedness 
 there, dogmatism is met by an opposite dogmatism, and in the 
 interminable controversy which arises between the champions 
 of apparently opposed but really complementary ideas, each 
 finds that the. sharp dialectic which he directs against his 
 opponent is retorted upon himself. Besides, even apart from 
 its being assailed in this way from without, a half-truth is its 
 own Nemesis. A one-sided dogmatism "has the opposite 
 dogmatism latent in itself. It needs only to be developed and 
 it destroys itself. A part setting itself up as a whole, an 
 abstraction claiming to be a complete reality, is in contra- 
 diction even with itself ; and this contradiction in the end must 
 be fatal to it. Thus the ancient dialectic gave the coup de 
 grdce to the theory that "all is one," (which was interpreted so 
 as to deny the reality of all difference,) when it showed that 
 absolute unity is no unity at all, or that unity means nothing 
 except in relation to difference.' 
 scf'tSa'm^'" ^^® ^^^*' sff^^*' °f *li^ failure of Dogmatism is naturally the 
 rise of Scepticism. The conflict of opposite dogmas produces a 
 sense of hopelessness, and even, it may be, a conviction that 
 " whatever can be asserted may with equal reason be denied." 
 Such scepticism may be of a deeper or of a shallower nature. 
 It may be only that superficial doubt which is the result of 
 observing many differences of opinion, and listening to much 
 argument on either side. It may be the sophistic consciousness 
 that a plausible case may be made out for anythincr or 
 against anything. Or finally it may be the deeper scepti- 
 cism of a reasoned despair of knowledge, arising out of the 
 
J- THE IDEA OF CRITICISM. 5 
 
 consciousness that every dogmatism has latent in it an opposite 
 dogmatism, and that the contradiction which it enqounters 
 from its opponent is only the recoil of its own logic upon 
 itself. It is especially this last kind of scepticism to which 
 Kant refers when he speaks of the sceptics as " those nomads 
 of the intellectual world who will not permit any steady culti- 
 , vation of the soil." Such scepticism, as Bacon said, is itself 
 a dogmatism ; " Ephedicus aKaTciXri-^lav dogmatizavit." The 
 sceptic seeks rest in negation, in the conviction that no truth 
 can be reached. He endeavours to make his mind content 
 itself with its subjective certitude of itself, and to repel as 
 slavery every objective belief. In this sense the ancient 
 Sceptics used the proverb that " he who shuns suretyship is 
 sure," interpreting it to mean that he who has committed him- 
 self to nothing, who rests his faith and trust on nothing either 
 intellectually or morally, but falls back upon the bare con- 
 sciousness of himself, is thereby raised above all disappoint- 
 ment or vexation. " Ich hale meine Sache auf Nichts gestellt " 
 is the wise man's motto. 
 
 Is such scepticism self-consistent, or is it vexed with a con- The sceptic a 
 
 Dogmatist. 
 
 tradiction like the dogmatism it opposes ? Kant maintains 
 that the latter is the truth, and that scepticism, like dogmatism, 
 carries in it the principle of its own refutation. It must do so, 
 because, as already said, it is only another kind of dogmatism. 
 When it stops short at a negative result and refuses to turn its 
 weapons against itself, it is guilty of the very inconsequence, 
 of which it accuses its enemy. It is really as a dogmatist 
 that the sceptic is strong against dogmatism: it is only as 
 asserting some principle which is common to the contending 
 parties, that he is able to show that they refute ^ach other. 
 For a purely negative position is an impossibility : even a 
 question involves an assumption ; even a doubt, still more a 
 negative conviction, must have a positive certainty behind it. 
 If I say that all I know is appearance and that I do not, 
 and cannot, know the reality which is beyond appearance, I 
 
6 INTRODUCTION. chap. 
 
 must have some positive reason for the distinction which I 
 make between appearance and reality; I must have some 
 criterion of the latter which enables me to deny its unity with 
 the former. If I say that I am conscious of myself and my ideas 
 but not of objects, I assume the reality or possibility of objects 
 independent of mythought, and alsothatlcan sever the conscious- 
 ness of myself from the consciousness of these objects. Absolute 
 scepticism thus destroys itself. The Sceptics themselves said 
 that it was a medicine which purged out itself as well as the 
 disease, but they did not recognise the force of their own 
 saying. For a scepticism that recognises its own inconsistency 
 has at the same time recognised that it is impossible to rest in 
 scepticism ; or in other words, that scepticism ends in disclosing 
 a fundamental belief in relation to which it is impossible 
 to be sceptical. If the first work of scepticism is to carry 
 us beyond opposite dogmatisms, the last work is to disclose 
 the basis of truth on which after all it, as well as they, 
 must rest. But when it takes this last work in hand, it 
 has ceased in the proper sense to be scepticism, and has 
 become Criticism. 
 so'ipticSmto T^is last statement may be illustrated by a remarkable 
 
 Criticism. . _ _y. . , „ . . ,. , 
 
 expression of Kant. " Scepticism," he says, " would have 
 been a useful regress, if it had gone back over the ground 
 traversed by the dogmatists to the point where their wanderings 
 began." ^ Criticism is a deeper kind of scepticism, which does 
 thus go back to the beginnings of our thought — or at least to 
 a point logically prior to that at which the opposite dogmatic 
 systems diverge from each other — and so gets into the straight 
 road again. In other words, its aim is to bring the controversy 
 to an end by detecting its sources and presuppositions. For 
 in every controversy there must be some ground common to ' 
 the controversialists, little as they may recognise it themselves. 
 If this were not so, assertion and denial, attack and defence, 
 would be equally unmeaning. And the value of scepticism is 
 1 E. i. 492 ; H. viii. 523. 
 
I- THE IDEA OF CRITICISM. 7 
 
 just this that, while using the arguments of each of the parties 
 to refute the other, it suggests that the question at issue has 
 certain presuppositions without the examination of which it 
 cannot be decided. To put the same thing from another point 
 of view, every great conflict of thought, such as that between 
 the Ionics and the Eleatics, or betw.een the Platonists and 
 Aristotelians, or between the Stoics and the Epicureans, or 
 again in modern times between the school of Locke and the 
 school of Leibniz, is really due to the fact that opposite but 
 complementary aspects or elements of a truth are taken for 
 absolutely contradictory views of that truth. Such contro- 
 versies arise out of the attempt to settle by a simple " yes " 
 or " no " questions which cannot be thus simply answered. 
 Hence each answer involves an absurdity and is open to an 
 irresistible attack from the other side, and that disputant will 
 be victorious who can secure the attack and force his opponents 
 to act on the defensive. Meanwhile the sceptic draws the con- 
 clusion that truth is unattainable, or, to put it in Kantian 
 language, that the question involves an insoluble " antinomy " 
 of reason. The , true interpretation of the facts is, however, 
 different. A dogmatism is an attempt to explain the whole 
 universe by a, principle which applies only to a fragment or 
 phase of it, while the opposite dogmatism denies that that 
 principle has any validity whatever, and puts an opposite 
 principle in its place. Finally the resulting scepticism is 
 simply the unlimited rejection of both the opposite dogmas. 
 But if this is so, it becomes obvious that all the combatants 
 are fighting within a closed arena where no conclusive victory 
 can be gained. The only way to put an end to the dispute is 
 to break through the narrow conditions under which it has 
 been carried on. And it is just this which Criticism seeks to 
 do. In other words, its aim is to penetrate to the principle 
 which underlies the controversy, to discover the more compre- 
 hensive conception which puts each of the opposing theories in 
 *its place as an element of the truth. This is a process which 
 
/. 
 
 Philosophical 
 Criticism. 
 
 8 INTKODUCTION. chap. 
 
 combines dogmatism and scepticism, yet is different from either. 
 It is dogmatic, in so far as it recognises the partial truth of 
 each of the dogmatic theories ; sceptical, in so far as it limits 
 each to a part ; and dogmatic again, in so far as it discovers 
 the unity which is manifested in their difference and relative 
 opposition. It limits the validity of secondary principles which 
 have been supposed to he primary and universal, by searching 
 out some prior truth which is the condition of their relative 
 validity. It is thus at once regressive and progressive, or 
 rather, it is regressive in order to be progressive. It goes back 
 to a principle of unity presupposed in the division of opposite 
 schools, in order to reach forward to a comprehensive idea in 
 which their difference is reconciled. 
 
 Criticism, (then, Vn the highest sense of the word, essentially 
 involves an effort to get beyond the sphere in which a contro- 
 versy is carried on, and to throw new light upon it from a 
 point of view which is above that of either of the disputants, 
 though it is also a point of view which both the disputants 
 tacitly acknowledge. That is a true criticism which lifts a 
 \ subject into the region of principle, and so frees it from the 
 I mere attack and rejoinder of ordinary controversy. A enticed 
 1 philosophy, in the sense of Kant, goes beyond this only in so 
 i far as it is an attempt to reach principles, which are prior not 
 \ only to a particular controversy, but to all controversy. As 
 j he describes it, it is a " criticism of the very faculty of know- 
 I ledge," the aim of which is to determine the most general con- 
 j ditions of the knowable. 
 Its Principle. It is obvious that of all that is knowable we must be able 
 to predicate whatever is involved in its being knowable, and 
 that such predicates will take precedence of all others, and will 
 determine or limit the sense in which they are to be understood. 
 From this point of view, therefore, there seem to be certain 
 assertions which we may make in regard to the world and to 
 
 every object in it, independently of its being actually known, 
 
 assertions which will not be altered or modified by any increase 
 
^- THE IDEA OP CRITICISM. 9 
 
 of our actual knowledge, or by any change of our view of those 
 particular objects which we already know. " The faculty of 
 knowledge " in this sense is the presupposition of anything 
 known ; and the criticism of that faculty, if successfully carried 
 out, must lead to the establishment of principles which are 
 universal, and which therefore can be used to determine the , 
 value and place in a scheme of knowledge of all secondary 
 principles. 
 
 The problem thus set before us by Kant seems at first sight Locke's, idea of 
 to be identical with that which Locke endeavoured to solve in 
 his Essay on the Human Understanding. On consideration, how- 
 ever, we find that there is an important difference, the explana- 
 tion of which may serve to throw some light both on the nature 
 of philosophical criticism and on the possibility of its object 
 being attained. Locke conceives the question as one of Psy- 
 chology. " I thought," he tells us, " that the first step towards 
 satisfying several inquiries the mind of man was very apt to 
 run into, was to take a view of our own understanding, examine 
 our own powers, and see to what things they were adapted. 
 Till that was done, I suspected we began at the wrong end, and 
 in vain sought for satisfaction in a quiet and sure possession of 
 truths that most concerned us, whilst we let loose our thoughts 
 ■ into the vast ocean of being, as if all that boundless extent were 
 the natural and undoubted possession of our understanding, 
 wherein nothing was exempt from its decisions, or escaped its 
 comprehension. Thus men extending their inquiries beyond 
 their capacities, and letting their thoughts wander into those 
 depths where they can find no sure footing, it is no wonder 
 that they raise questions and multiply disputes, which, never 
 coming to any clear resolution, are proper only to continue and 
 increase their doubts, and to confirm them at last in perfect 
 scepticism. Whereas, were the capacities of our understandings 
 well considered, the extent of our knowledge once discovered, 
 and the horizon found which sets the bounds between the en- 
 lightened and dark parts of things, between what is and what 
 
10 INTEODUCTION. ''HAP. 
 
 is not comprehensible by us, men would perhaps with less 
 scruple acquiesce in the avowed ignorance of the one, and em- 
 ploy their thoughts and discourse with more advantage and 
 satisfaction in the others." ^ Locke, in fact, bids us examine our 
 faculties in order that we may discover the, nature and limits 
 of our knowledge, very much as we might examine a telescope 
 in order to discover whether there was any flaw in its construc- 
 tion which might distort our vision of the objects seen through it, 
 or any limitation of its range, which, beyond a given distance, 
 might render it useless. Now, if this were the sense in which 
 Kant bade us criticise our faculty of knowledge, it would not be 
 unreasonable to object that we cannot examine the mind except 
 with the mind, and that any defect in the mind, which might 
 hinder us from knowing other objects or from knowing them 
 correctly, would equally hinder us from knowing the mind itself. 
 To this, from a purely psychological point of view, there seems 
 to be no answer.^ Mind, as one of the objects of knowledge, in 
 the sense in which we contrast it with matter, cannot be taken 
 as the object the knowledge of which is at the basis of all other 
 knowledge. On the contrary, to understand mind in that 
 sense implies that we already understand the material world. 
 For man is a being who doubly presupposes nature, as he is a 
 spirit which finds its organ in an animal body, and as it is in the 
 system of nature that he finds the presupposition and environment 
 of his life. To base knowledge on psychology would, therefore, be 
 to base it on what is its latest and most complex result. If we 
 knew mind in this sense, we should not need to look for the 
 principle on which knowledge is built, for we should already 
 have attained every end with a view to which such a principle 
 is sought after. If the copestone had been firmly and securely 
 
 * Essay, I. i. 7. 
 
 ' I.e., if Psychology deals with mind as one object among others. Locke 
 no doubt, like Berkeley, goes on the supposition that the mind knows itself 
 by direct introspection, and that the only difficulty is how it should look 
 beyond itself. This view, which to some extent influences even Kant, will be 
 fully discussed in the sequel. 
 
^- THE IDEA or CEITICISM. 11 
 
 put on the temple of knowledge, we should not be searching for 
 its foundations. 
 
 Now this difficulty arises from the twofold aspect in which '^ criticism 
 
 ^ based on 
 
 mind presents itself, as at once the beginning and the end of '^''^'''^^"syi 
 knowledge. The thinking being is not merely an object in the 
 known or knowable world, he is also a subject of knowledge, 
 and it is only for such a subject that an object or a world of 
 objects can exist. Hence we may speak of man's knowing 
 himself in two ways : of a knowledge of himseK in which he is 
 regarded simply as the self, the thinking subject which is im- 
 plied in all objects of knowledge ; and of a knowledge of him- 
 self as a human being, distinguished from other human beings 
 from the animals and from nature in general, and standing in 
 definite relations to each of them. With the labter kind of know- 
 ledge of himself, which is the subject matter of Psychology, 
 Criticism, in the primary aspect of it, has nothing to do ; for 
 this knowledge of mind, as has been already said, is not the 
 beginning, but rather the end, of science ; and it cannot be 
 used as a test or criterion • for that which is more simple than 
 itself. Criticism has to deal with the knowledge of mind only 
 in so far as mind is presupposed in everything known or know- 
 able ; or, in other words, in so far as the principles which are 
 involved in the relation of objects to a conscious self are 
 the latent presuppositions of all knowledge, the principles 
 through which everything else must be known, and by means 
 of which, therefore, every other kind of knowledge must be 
 tried. Psychology is only interested in such inquiries in 
 so far as the fact that a conscious being, as such, is a subject 
 of knowledge, must essentially modify our view of his relation 
 to all other objects in a world which cannpt logically be con- 
 sidered as existing apart from such a subject. 
 
 Now, leaving out of consideration for the present certain in- °hyi%aif '"*" 
 adequacies in Kant's statement, it is in the sense just explained' 
 that we must take his assertion that every philosophy is a dog- 
 matism which does not begin with a criticism of our faculty 
 
Defect of 
 
 Psychological 
 
 Criticism. 
 
 12 INTRODUCTION. chap. 
 
 of knowledge. For, as aforesaid, all knowable objects are in 
 necessary relation to a thinking subject ; they are essentially 
 objects-for-a-self : and this relativity makes it impossible to 
 treat them as external to the consciousness of such a subject, 
 or as things to which that consciousness is external, (as if 
 objects merely happened to be in one world with certain beings 
 capable of knowing them). From this point of view, mind is 
 not one thing and matter another : for if by mind we mean 
 the conscious subject, such a subject is the presupposition of the 
 material and spiritual worlds alike. Or, to put it otherwise, we 
 cannot speak of things as qualified in themselves apart from the 
 accident that makes them possible objects to a subject. They 
 have from the first involved in them the characteristics in virtue 
 of which I know them. They could never come into my con- 
 sciousness at all, unless they were from the first essentially 
 related to consciousness. Mind does not condition them 
 merely in the sense that thinking beings are part of the world 
 of nature, and in that world externally act on other beings and 
 things and are reacted on by them ; but in the sense that mind 
 is the condition of there being such a world of related objects 
 at all. For it is manifest that, if all existence has to be defined 
 as existence for a thinking self, in bringing to self-consciousness 
 what is involved in this relation to a self, we are not merely 
 explaining the relation of matter and mind as two separate 
 objects, but showing what are the preconditions of there being 
 any objects or knowledge of objects at all. "We are, in other 
 words, going back to the beginning of Imowledge, and at the 
 same time of the known or knowable world, when we thus ^o 
 back upon ourselves. 
 
 The defect of the Lockian or psychological theory of know- 
 ledge, then, may be thus stated. It treats the faculty of know- 
 ledge merely as an attribute of certain beings in the world, by 
 which they are characterised and distinguished from other beincfs 
 so that, e.g., as weight is the attribute of a stone, thought is the 
 attribute of man ; or, to use the famous saying of Cabanis, 
 
I- THE IDEA OF CRITICISM. 13 
 
 " the brain secretes thought just as the liver secretes bile." 
 But such a view of the faculty of knowledge is essentially 
 inconsistent with the very possibility of knowledge. It implies 
 that acts of the mind, by which .we perceive and think of 
 objects, are to be regarded as nothing more than states of the 
 individual conscioiisness as such. But if that were the true 
 account of them, the knowledge of objects through such states, 
 would be obviously impossible. For, how could the individual 
 in the mere states of his own being ever find a reason for say- 
 ing anything about things which ex hypothesi are not such 
 states ? If minds are just one sort of real things in the world, 
 and all the acts of knowledge are merely expressions of their 
 characteristic faculty, then they can know nothing but them- 
 selves. Knowledge, for them, must not only begin in psycho- 
 logy, but it must end there. To the being thus imprisoned in 
 himself, there would be no escape ; or if, by a miracle, — such 
 as is involved in the " occasional causes " of Descartes or 
 the pre-established harmony of Leibniz, — the ideas within 
 should correspond with the objects without him, he could 
 never know that they did so correspond. To take up the 
 position of Locke, i.e., to regard the mind as a separate object 
 and to treat its ideas and perceptions as mere subjective phe- 
 nomena, which may be discussed apart from all question as to 
 an objective world presented through them, is to assume the 
 impossibility of knowledge. For if the mind could know itself 
 and its operations apart from the knowledge of anything else, 
 it would ipso facto be shown to be a substance purely external 
 to that world and incapable of knowing it. In other words, it 
 would be shown to be a mind whose development to self-con- 
 sciousness did not imply, and which therefore could not by any 
 possibility attain, any consciousness of a world outside of it. It 
 would be imprisoned in itself, not in the sense of having an 
 imperfect or distorted vision of the wdrld through its own sub- 
 jective states, but in the sense that it would have no vision of 
 •the world at all. To make Psychology the ]prius of all other 
 
14 INTRODUCTION. •^^^''■ 
 
 sciences, is, in other words, to say that there shall be no science 
 but Psychology. The principle of the Lockian criticism is such 
 that it leaves us nothing to criticise. Its attempt to explain 
 knowledge must become inevitably, what it did become with 
 Hume, an attempt to explain how such a thing as knowledge 
 could be imagined to exist. It is a criticism which leads back 
 
 into scepticism. 
 Possibility of j5„(. if j+ ig imTjossible to take the mind as a separate existence 
 
 Metaphysical j-'mj, 4. j- 
 
 Criticism. Qu^jgi^e of the rest of the known world, and, by examination of 
 it, to secure a point of view from which we can look down upon 
 our consciousness of the world and criticise it : if we cannot, 
 apart from knowledge, examine our means of knowing, — how 
 then, it may be asked, can we criticise knowledge at all ? On 
 what shall we take our stand in examining the process whereby 
 we have acquired our experience and developed our present 
 beliefs ? Is it not absurd to speak of criticising our whole 
 view of life — all that we call knowledge — when we cannot 
 find any means of doing it except in that knowledge itself? 
 Archimedes could not move the world, because he could find 
 no support for his lever outside of the world itself. And it 
 seems at first as if the idea of a criticism of knowledge in 
 general contained a similar contradiction, the contradiction, 
 namely, of supposing that there is a point beyond consciousness, 
 which yet consciousness can reach, and from which it can judge 
 itself. 
 
 Limits of Now in a sense it is true — and it is of the highest import- 
 
 ance that we should recognise it — that we cannot get beyond 
 the cycle of " our own ideas." We can never know anything 
 except as it is related to the conscious self within us ; whatever 
 we deal with, we are still dealing with our own consciousness of 
 things. If anything is excluded by that, it is absolutely excluded. 
 Of things in themselves, altogether out of relation to conscious- 
 ness, if there are such things, we can know nothing.^ We 
 
 1 Whether Kant's things in themselves are thus unrelated to consciousness 
 we shall afterwards consider. 
 
 Knowledge. 
 
'■ THE IDKA OP CRITICISM. 15 
 
 could not criticise our consciousness of things except by a 
 second consciousness, and why should this second consciousness 
 have any more authority than that consciousness in which they 
 were first presented to us? The intelligible world in this sense 
 is a closed circle within which all things, or at least all things 
 that are objects for us, are included, and to explain anything 
 within this circle by reference to what is without it, is to use 
 words to which no meaning corresponds. The knowable universe 
 has no tortoise to rest on, no external handle by which it can be 
 grasped. If any one, therefore, should choose to direct a battery 
 of scepticism against the reality or possibility of Icnowledge, we 
 cannot refute him, except by showing that the battery itself is 
 planted within that very world of knowledge against which he 
 pretends to direct it. 
 
 And in this is already involved the solution of the whole The criterion 
 
 of Knowledge 
 
 difficulty. There is, indeed, no possibility of finding any -"^^^j^^^j^^^*! 
 criterion of knowledge outside of knowledge itself, no possi- 
 bihty of rising to another kind of consciousness which com- 
 mands or looks down upon our ordinary consciousness of the 
 world. If we raise the question of the criterion in the way in 
 which it was raised in the Stoic and Epicurean and also in some 
 modern schools of philosophy ; if, in other words, we suppose 
 consciousness as existing on the one side and the object on the 
 other, as independent things which can only externally act on 
 each other ; and if we ask how the mind, conscious primarily 
 of its own affections as such and of them only, is to get beyond 
 itself to apprehend the object, or where it is to find the criterion 
 by which it may test whether any, and which, of its subjective 
 ideas represent objective reality, the answer invariably must be 
 the sceptical one, that no such criterion can be found. We have, 
 in fact, made the problem insoluble by the very way in which 
 we have stated it. For we have been asking for a criterion 
 vdthin the mind of that of which the one assumed characteristic 
 is that it is without the mind, in such a sense that it cannot 
 cpme into any relation with mind at all. On such a question, 
 
16 INTRODUCTION. chap. 
 
 however, the true criticism is, not that the answer is unattain- 
 able, but that the question itself is meaningless. The objects 
 of which it speaks would cease to be what they are defined as 
 being, if they could be known. Even to speak of them is to 
 suppose a breach between intelligence and reality which cannot 
 be healed, but which, for the same reason, could not be known 
 to exist, could not even be spoken of, without an absurdity. 
 The question thus involves the same contradiction which has 
 just been pointed out in absolute scepticism, i.e., the assertion 
 of a point of view outside of the intelligible world from which 
 doubt may be directed against it. 
 
 The Particular But wliilc it is impossible to find a criterion within our con- 
 known 
 
 univSsai'"^ sciousuess by which we may test its correspondence with a 
 world, which is supposed to be outside of that consciousness ; 
 nay while the idea of such a criterion involves a contradiction, 
 this does not imply that it is impossible to find in our con- 
 sciousness a criterion of the validity of knowledge of those objects 
 which are present to that consciousness. Tor all our knowledge 
 of particular objects is based upon certain general principles, 
 principles which flow from the nature of consciousness itself 
 and its relation to objects in general. And if these principles 
 are once brought to light, they may be used to test and to 
 correct our special ideas and beliefs. Thus, e.g., in all conscious- 
 ness of the world, in the ordinary as well as in the scientific 
 consciousness, we find it represented as a unity, and even, with 
 more or less definiteness, as a systematic unity. As it is one self 
 
 to which all our consciousness is related, so it is as in one world 
 
 in one space and one time — that all objects of consciousness are 
 present to us. . And the things, beings, and events of that world 
 are, therefore, all conceived as standing in some kind of relation 
 to each other. This, at first, may not appear to be true of the 
 scientific consciousness, because science is continually discover- 
 ing new difficulties in the interpretation of facts, and these are 
 continually leading it to the adoption of new theories. And it 
 may not appear to be true of the ordinary consciousness, because 
 
I- THE IDEA OF CRITICISM. 17 
 
 it makes little or no effort to interpret facts on general prin- 
 ciples, but rather seems to take them as a confused mass of 
 particulars, associating them just as they present themselves in 
 space and time in individual experience. But a clearer view 
 makes us aware that the scientific consciousness is based upon 
 a belief in law and order which is never disturbed by the diffi- 
 culty of finding a definite place for particular plienomena, but 
 sees in such difficulty only an occasion for remoulding certain 
 of its subordinate views of nature in accordance with fixed 
 general principles. And as to the popular consciousness, it is 
 just the scientific consciousness in an inchoate state, working in 
 ignorance or at least without distinct knowledge of the prin- 
 ciples it uses, but none the less presupposing such principles in 
 all its rough and ready interpretations of particular facts. Thus 
 the categories of substance, of cause, and of reciprocal influence, 
 are imbedded in the very grammatical structure of language, 
 which is itself a result of the unconscious working of reason ; 
 and it is not difiicult to show that the " plain man " uses them 
 in every account of facts which he gives to himself or to others, 
 however little he may have reflected on them, and however in- 
 consistent or uncertain may be his application of them. They 
 are the framework, so to speak, upon which his view of the 
 world is laid down, the forms according to which his intelli- 
 gence acts in all its acquisition of knowledge. And the man 
 of science, with his canons of induction, by which he tries to 
 discover or verify the true causes and interdependence of phe- 
 nomena, does not essentially differ from the " plain man " in the 
 principles by which he guides himself, but merely in the cer- 
 tainty and clearness of consciousness with which he applies 
 them, deliberately employing them for a foreknown purpose, 
 instead of simply letting himself be led by them. J3ut both to 
 the scientific and to the ordinary consciousness, the world is 
 one in its manifoldness, permanent in its changes, inter-related 
 in its coexistences ; and to both, this general consciousness of 
 
 » unity takes effect in special attempts to connect and explain 
 VOL. I. B 
 
18 INTRODUCTION. chap. 
 
 particular things and events as causing or influencing each 
 other in definite ways. "When we realise what this means, we 
 see that it implies nothing less than this, — that all forms of 
 rational consciousness are built on one plan, according to some 
 ultimate principle of unity, which manifests itself in different 
 ways of connecting phenomena in space and time as coexistent 
 or successive, and which thus constitutes for each thinking being 
 a world of objects and events, standing to each other and to the 
 self that is conscious of them in definite and permanent relations. 
 bSd?n°° Now it is just here that Criticism seeks to find the standard 
 principles, by which to prove all things, the criterion by which to test all 
 ideas that present themselves as knowledge. If human experi- 
 ence is built on such general principles, criticism is possible ; if 
 it is not so built, then it is impossible. For in the latter case 
 no principle, in view of which we could criticise our conscious- 
 ness of the world and of ourselves, could have more than a 
 temporary and relative value. Criticism, in other words, is 
 based on the idea that below all special phases of knowledge, 
 there is a general form of knowledge, or a general " schema," — 
 to borrow an expression from Kant, — which we carry along with 
 us, and by means of which, ajl, even the least instructed of men, 
 impart a kind of unity to their experience. Every self-conscious 
 being has at least some rough tests of that which he recognises 
 as a fact ; and when a new phenomenon presents itself, every 
 such being is obliged in some way to find a place for it, to give 
 it a local habitation and a name, in relation to all other facts 
 in the one world of his consciousness. And this means that he 
 has within him the general plan for a self-consistent natural 
 system, and that he compels all things that claim to be real to 
 take up a definite position in it. It is true, as already shown, 
 that this plan may be, in a sense, unconscious ; ie:, it may never 
 be reflected on, or made an object of attention for itself, it may 
 reveal itself only in its effect, and not in the activity that pro- 
 duces it.i Few know that they have it in their minds at aU, and 
 
 1 A. 104. 
 
I- THE IDEA OP CRITICISM. 19 
 
 fewer still would be able to define or describe it. As in the 
 case of language the consciousness of relations shows itself in 
 grammatical structure long before these relations are named or 
 thought about for themselves, so it is with man's knowledge or 
 experience of the world, of which, indeed, language is but the 
 earliest expression. Human experience betrays its ideal character 
 in the way in which the unity of consciousness maintains itself 
 in and through the diversity of its contents and objects, long 
 before there is any belief or even thought of a reign of law. 
 Such a unity must in some way be present to the mind of man 
 if he is to have an intelligible experience at all ; or rather, we 
 might say, it is his mind ; as, on the other hand, the absence of 
 it, the incapacity to put particular facts in thei,r places in relation 
 to others in one consciousness, is exactly what we mean by 
 idiocy or madness. And this, indeed, is just the reason why we 
 do not at first recognise this unity, and why, even when our 
 thoughts are directed to it, it is so difficult to realise how much 
 is involved in it. The very fact that it is the ground upon 
 which all intelligible experience must proceed, hides it from our 
 view. It is already behind us, so to speak, when we begin to , 
 be conscious of objects as such, since it constitutes the very 
 faculty by which we know them. We look outwards he/ore we 
 look inwards, and we cannot look inwards till we look outwards 
 any more than we can be conscious of the faculty of sight with- 
 out first seeing something. And hence, although the faculty of 
 knowledge is in a sense prior to actual knowledge of objects, the 
 consciousness of that faculty, the consciousness of the self which 
 knows, is posterior to such knowledge and presupposes it. 
 Such consciousness is the product of a return of thought from 
 the knowledge of objects upon the unity implied in it, though 
 the self upon which we thus return is the presupposition of all 
 our objective consciousness and therefore of all objects what- 
 ever. 
 
 Now it is just this presupposed unity and the principles of apHoh and 
 knowledge to which it gives rise, which criticism must seek to 
 
20 INTKODUCTION. chap. 
 
 discover, in order that by them it may test the particular ele- 
 ments of the knowledge or supposed knowledge which we already 
 possess. This may be expressed by saying that Criticism has to 
 discover the a priori elements of knowledge. This expression, 
 it is true, is usually taken in another sense, as meaning that 
 which is attributable to the mind as opposed to what it gets 
 from experience. And when we come to treat of the special 
 form in which the problem of criticism presented itself at first 
 to Kant, we shall have to take account of this division of the 
 parts of knowledge, and to discuss its validity more fully. But 
 here, where we are dealing with the problem of Criticism in 
 general, without reference to the peculiarities of Kant's statement 
 of it, it may be sufficient to point out that the broad division of 
 a priori and a posteriori elements of knowledge, according to 
 which the former is derived from the mind itself and the latter 
 from without, must disappear, or at least must altogether change 
 its meaning, so soon as the problem of criticism is understood. 
 For the opposition of a priori to a posteriori, as different parts 
 of knowledge, really rests on a confusion of the distinction of 
 the subject and the object of knowledge with the distinction of 
 mind and matter as different objects of knowledge. But so 
 soon as it is recognised that no object of knowledge is given 
 apart from its relation to the subject, it becomes impossible to 
 say that any part of knowledge is purely a posteriori, in the 
 sense of being due to the object as apart from the consciousness 
 for which it is. And so soon as it is recognised that the subject 
 comes to the knowledge of its faculty, i.e., of itself, only in and 
 through the knowledge of objects, it becomes equally impossible 
 to say that any part of knowledge is purely a priori, in the sense 
 of being present to the mind apart from all consciousness of 
 objects. In the sense of criticism, therefore, the name a priori 
 can be applied only to those elements of truth which are pre- 
 supposed in all consciousness of objects — which are, so to speak, 
 the first stones in the foundation of the temple of knowledge 
 or, to take another metaphor, the seeds from which all know- 
 
I- THE IDEA OP CRITICISM. 21 
 
 ledge has to grow. They are, in other words, the principles 
 through which all other truth is and must be seen, and which, 
 therefore, are not capable of being treated like special facts or 
 laws which are to be put on equal terms with other facts or 
 laws. 
 
 This view of the a priori element of knowledge carries us Basis of unity 
 
 " in all con- 
 
 back to Kant's dictum about scepticism, which he declares to be 0^™.°^° °' 
 
 no true criticism, because in refuting dogmatism, it did not go 
 back to the point at which dogmatism diverged from the true 
 road of knowledge, i.e., it did not go back to the very idea of 
 truth on which all doctrines, which profess to be true, implicitly 
 claim to be founded. The idea before Kant's mind is that all 
 kinds of thinking consciousness, must proceed up to a certain , 
 point along a common path ; and that, therefore, there is a basis 
 of common understanding between all minds, whether they be 
 dogmatists or sceptics, and whatever they affirm or deny. This 
 might be otherwise expressed by saying that all assertion and 
 denial must take place within the limits of the intelligible. We 
 caii disprove a particular dogma, but in doing so our attitude 
 cannot be purely negative, any more than when we prove it. 
 For, in the former case as in the latter, we inevitably imply, 
 even if we do not express, some idea of truth by which we test, 
 or in relation to which we accept or reject it. The sceptic, like 
 all other rational beings, has his presuppositions, and he gets 
 his apparent advantage over his adversary only because he con- 
 ceals, or even perhaps is not aware of them. His advantage is 
 simply that he strikes from the dark at an enemy in the light. 
 Drag him from his covert, and you find that every weapon 
 which he uses can be retorted against himself. All rational attack 
 and defence must rest on, and appeal to, certain general prin- 
 ciples which make the assailant and the defender intelligible to 
 each other : and the sceptic, so soon as he begins to speak, 
 takes his stand along with his opponent upon the general basis 
 of intelligence. To attempt, as the Sceptic proposes to do, to deny 
 the very idea of knowledge, — ^hich alone makes his statement 
 
22 INTEODUCTION. chap. 
 
 intelligible to himself and to his opponent, and furnishes the 
 only common ground upon which they can meet, — is like 
 attempting to wrestle with an opponent while our feet are in 
 the air. The intelligence can no more hoist itself out of the 
 intelligible world by any process of argument, than the body 
 can lift itself out of the material world. On the contrary, as I 
 have already indicated, the very effort after absolute denial 
 which the sceptic makes, must tend to bring to light principles 
 which his scepticism does not and cannot assail, principles which 
 it seems able to assail onlj^ from a confusion of the universal 
 with the particular, of the idea of truth with a particular truth, 
 re^es^s""*^"*' The conscious need of a criticism, which shall disclose these 
 necessary. ultimate bases of truth and thus give definiteness to the idea 
 of knowledge, arises, as we have seen, out of the failure of the 
 first immediate constructive effort of thought, which Kant calls 
 dogmatism.! It is to the consciousness of such failure, indeed, 
 that we owe even the origin of science, which begins in doubt 
 or wonder,^a doubt or wonder which is produced by the 
 apparent inconsistency of phenomena with each other. For 
 what such doubt or wonder betrays is that the mind has been 
 proceeding on certain principles or presuppositions in the con- 
 struction of its experience into a whole, and that in doing so, 
 it has met with an obstacle, and has found itself unable to 
 combine the new experience with previous experience without 
 a change of these presuppositions. A phenomenon excites 
 wonder because it is not what we expected, because it will not 
 fit into a place in our general plan of things, but comes into 
 colHsion with other phenomena according to the view we have 
 been used to take of them. If we had no expectations, nothing 
 could surprise us. Our surprise means that there is a difficulty 
 in interpreting appearances according to the mental scheme or 
 plan of their connexion we have hitherto adopted. But it is 
 
 1 When we speak of ordinary opinion as ' dogmatic,' we are not using the 
 term with strict accuracy. A dogma, properly speaking, must be a general 
 principle or law. 
 
I- THE IDEA OF CRITICISM. 23 
 
 just such a difficulty which also for the first time directs 
 attention to the existence of that scheme or plan, or to some 
 element in it, which has hitherto been in operation without 
 being specially attended to. To' take an example which carries 
 us back to an early period in the history of science. The 
 determination of things as to their quantity must have begun 
 with assertions as to their relative size or weight, or the rela- 
 tive intensity of some of their qualities. But such assertions 
 were soon found to be wavering and uncertain, affected by 
 every change in the circumstances of the individual who made 
 them. The dogmatism of individual opinion, and the scepticism 
 as to the possibility of objective quantitative judgments, which 
 naturally arose from such collisions of opinion, ended in direct- 
 ing attention to the universal implied in them. A regress of 
 thought upon this universal was therefore made by some early 
 critical philosopher, who set himself to consider the nature of 
 quantity ia general and the principle of its determination. 
 Arithmetic and Geometry arose out of this effort of reflection.. 
 The relations of discreet and continuous quanta were dis- 
 entangled from the mass of concrete detail in which they had 
 been concealed, and the abstract development of these relations 
 furnished a basis on which the first accurate knowledge of 
 things, so far as they are quanta, could be founded. For so 
 soon as the general principles presupposed in all quantitative 
 judgments had been brought to light, a rule was found, accord- 
 ing to which observations could be connected together and new 
 results developed out of them. The a priori synthesis of the 
 mathematical sciences, to use an expression of Kant, furnished 
 a means for the anticipation of particular phenomena, the rela- 
 tions of which must be conformable to the principle through 
 which alone particular quantitative judgments could be made. 
 This critical regress upon the universal, therefore, at once put 
 an end both to dogmatism and to scepticism, and made mathe- 
 matical knowledge enter upon the " secure path of science." '• 
 1 Cf. the preface to the second edition of the Critique of Pure Reason, B. x. 
 
24 INTEODUCTION. ^^■*^- 
 
 sif^ofersu'ch ^0''^ similar remarks might be made in the case of all that 
 regress. ^^ ^^y^ scieiice : its first occasion always lies in some collision 
 of appearances with each other in our first synthesis of them. 
 And the only way of escape from the doubt and difficulty thus 
 produced has always been to discover the principle implied in 
 such synthesis, and to develop it into a system of relations, 
 which might serve as a guide in all particular judgments in 
 relation to the class of objects which fell under that principle. 
 Each science is thus the development of some general principle 
 which was already involved in our first unreflective judgments 
 about the subject matter of the science. As, however, in such 
 judgments the principle in question was not present to the 
 mind as an object, but only as an unregarded presupposition, 
 so its application was naturally capricious and uncertain, and 
 could not fail in the long run to give rise to difficulty and 
 even contradiction, and such difficulty and contradiction could 
 be removed only by bringing the presupposition to light and 
 discovering all that was involved in it. Now all the principles 
 of the sciences are ultimately particular developments of the one 
 general presupposition of all science, viz., that the world is an 
 intelligible whole. In other words, the latent assumption which 
 every intelligence carries with it, that all phenomena form parts 
 of one consistent system relative to thought, is the stimulus 
 which forces us to seek for some way of reconciling apparently 
 inconsistent facts : and it is our failure in our first attempts at 
 such reconciliation which makes us turn our attention to the 
 ■universal involved in these facts, as a principle by which the 
 relations of particulars may be determined. It is thus that 
 the critical regress becomes the means of a new progressive 
 movement of science and enables it to strike into a hitherto 
 untried path, along which it may proceed securely and rapidly 
 without being troubled with the misgivings of scepticism or 
 the conflict of dogmatisms. 
 ^new'cSa-'' Science, however, soon forgets the doubt in which it arose. 
 Once started on a definite line of inquiry, each science seems 
 
 cism. 
 
I- THE IDEA OP CRITICISM. 25 
 
 to go on accumulating truth upon truth without being obliged 
 to retract anything it has once ascertained. It is in fact only 
 when checked in its course that the mind turns back upon 
 itself, nor would it ever occur to it to criticise a principle and 
 a method which was every day being applied with success in 
 the extension of knowledge. The question whether knowledge 
 is possible, is precluded in such a case by the fact that it 
 exists, and all doubt disappears before the solvitur amlulando 
 of advancing science. In modern times this claim of success 
 has been, put forward with special emphasis in behalf of the 
 mathematical and physical sciences, which have long over- 
 passed the struggles of their youth. And a contrast is often 
 drawn between their steady advance and the perpetual alter- 
 nations of dogmatism and scepticism without any apparent 
 progress, which has been seen in the sphere of metaphysic. 
 How are we to account for this difference ? Are we to take it as 
 an indication that the objects of metaphysic are, essentially and 
 in the nature of things, beyond the possibility of knowledge ? 
 And if so, how are we to account for the inextinguishable life 
 which seems to belong to such investigations, which constantly 
 spring up again even in partibus.infidelium, — even in the works 
 of those who profess to have renounced metaphysic ? Or 
 shall we take it as an indication that in this region scepti- 
 cism has never gone deep enough to become criticism, and 
 to define that universal, the reality of which seems to be evi- 
 denced by our continual tendency to use it in particular judg- 
 ments, though its definition has never been ascertained in 
 such a way as to bring metaphysic into " the secure path of 
 science " ? 
 
 It is somewhat in this form that Kant presents the problem Contrast be- 
 
 ■^ tween tne 
 
 to us in the preface to the second edition of the Critique. The |[°l?°°g'y,^£* 
 sciences of Mathematics and Physics seem to be secure in 
 themselves, because they exist, and because they are continually 
 developing new results ; while Metaphysic evinces a tendency 
 to dogmatism, which is balanced by an equal tendency to an 
 
26 INTEODUCTIOlSr. chap. 
 
 opposite dogmatism, and is, therefore, continually brought back 
 to its beginning by scepticism. 
 Theprobiem jf y^Q igok a little closcr at this result, we find, as Kant 
 
 due to the ' ' 
 
 sote™e!/^\ intimates, one aspect of it which may at once lead us to a 
 ' clearer view of the difSculty, and put us in the way of solving 
 it. The metaphysical problem, which gives rise to the conflict 
 of opposite dogmatisms with each other and with scepticism, is 
 not entirely distinct from the problem of physical and mathe- 
 matical science. On the contrary, it is in the attempt to 
 universalise the principles of such science that the metaphysi- 
 cal diiSculty makes its appearance. These principles have a 
 sphere in which they are continually verifying themselves by 
 making new conquests, and therefore, the intelligence is only 
 following a natural impulse when it tries to use them as guides 
 in other spheres. So long as we are dealing with the material 
 world, we find no reason to doubt their applicability ; why 
 should we hesitate to apply tliem to the solution of questions 
 as to the spiritual world, — questions as to the nature of the 
 soul, the mode of its presence in the world, and the way in 
 which it acts and is acted upon by other beings and things ? 
 When, however, we attempt thus to extend them, when we try 
 to apply in this new region the principles which have led to 
 such successes in mathematics and- physics, we are at once 
 brought into collision with some of the most deeply rooted con- 
 victions of man in regard to the spiritual world ; we are 
 even led to doubt the very existence of such a world. Now, it 
 cannot be denied that, besides and beyond the objects of the 
 external or physical world, there are stHl many objects or sup- 
 posed objects round which the thoughts of man have turned in 
 aU ages. The higher interests of man, those interests that are 
 most distinctive of man, centre, according to Kant, in the three 
 great ideas of God, Freedom, and Immortality : or, to put it in 
 another point of view, in the ideas of the soul, as a spiritual 
 and self-determining subject ; of the world, as a system which 
 in all its variety and change yet maintains rational order and 
 
'• THE IDEA OF CRITICISM. 27 
 
 unity with itself ; and finally of a divine unity, which is at 
 once the source and the end of the intelligence and the intelli- 
 gible world. But any attempt to apply to the objects of these 
 ideas the principles which enable us to make the material 
 world intelligible, is equivalent to a denial of their existence. 
 To suppose the existence of such objects under conditions of 
 space and time, — and they must fall under such conditions if 
 Mathematics is to be applied to them, — involves a contradiction. 
 To treat them as objects externally determined by other objects, 
 according to the law of physical causation, — as would be 
 required if they were to be brought into the sphere of Physics, 
 — involves a still more obvious contradiction. If the soul be 
 taken as such an object, it must be conceived simply as an 
 attribute of the body or a series of phenomena occurring in it. 
 We must then treat it exactly like other objects in the external 
 world, and deny to it any independent, self-determining power. 
 And if on this view we cannot regard the soul as standing in 
 any exceptional position in the world, neither can we regard 
 the world as corresponding in any special way to it, or as 
 having in itself any unity other than that of an endless aggre- 
 gation of externally determined and externally determining 
 objects in space, which pass in consequence of this determina- 
 tion through definite changes in time. N'or lastly, can God be 
 anything but another name for this aggregate. If, on the other 
 hand, these ^conclusions are resisted ; and if, on the ground of 
 any spiritual functions of man's Kfe, on the ground of the facts 
 of knowledge or of moral action, it be maintained that the con- 
 scious subject is more than one of the facts or objects he knows, 
 more than one of the links in the chain of natural causation, 
 still more if it be maintained that that chain itself is but the 
 phenomenon of a higher reality ; then the principles of Mathe- 
 matics and Physics themselves seem to be brought in question; 
 and in particular the principle of physical casuality, which ^:^^i»«<«*i^ 
 underlies all the achievements of modern science, seems to be 
 deprived of its universal validity. And if it be admitted to be 
 
28 INTRODUCTION. ''^■^^• 
 
 not universally valid, a doubt is thereby cast on its validity 
 even in the sphere in which it is still allowed to prevail; at 
 least until some higher principle is found by which that sphere 
 may be limited and defined. In any case, the intelligence can- 
 not rest in the conception of a dual or divided world without 
 some principle to mediate between the different laws which are 
 supposed to rule it; some principle which may make the 
 dualism intelligible and at the same time carry us beyond it. 
 It is impossible that we can be content to put two distinct 
 principles and the spheres of their application side by side 
 without seeking for a deeper principle to which their difference 
 is subordinate. The same impulse, which in the first instance 
 forced men to ask for some explanation of the inconsistency of 
 appearances and. which thereby gave rise to science, must also 
 impel them to reconcile their conflicting consciousness of 
 natural and spiritual reality, either by a negation of the 
 difference between them (in other words, by the assertion that 
 " all is spiritual," or that " all is natural "), or by the discovery 
 of the nature of that unity, which is already in a certain sense 
 present to our minds in so far as we are conscious of each in 
 relation to the other. 
 Doubt A doubt of the principles of mathematical and physical 
 
 reflected back 
 
 upon Science, ggience Kant considered to he irrational, if not impossible, at 
 least so long as our minds remain within the peculiar sphere of 
 these sciences. At the same time he acknowledged that a reflex 
 doubt is cast both upon these principles and upon the science 
 that is built upon them, so soon as an attempt is made to carry 
 them beyond that sphere.^ For this attempt, as he argued, gives 
 rise to a conflict of dogmatisms which shows that the mind was 
 no longer moving in the "secure path of science." Such reflex 
 doubt, which properly affects only the universality of the 
 principles involved, cannot be set aside, until we find a deeper 
 principle, which shall at once explain the relative validity of 
 these subordinate principles in a particular sphere and their 
 
 lA. 87; B. 120. 
 
I- THE IDEA OF CRITICISM. 29 
 
 restriction to that sphere. For, so long as such a deeper prin- 
 ciple is not forthcoming, and so long as by means of it the line 
 is not definitely drawn, so as securely to determine the relative 
 limits of 'the subordinate principles, so long doubt must hover 
 on the borders of physical and mathematical science, and even 
 at times make incursions upon it. Thus a doubt which would 
 not affect such sciences taken in themselves, in so far as they 
 make up a coherent whole of knowledge completely rounded in 
 itself and determined by its own principles, is awakened against 
 them so soon as we are led to suspect that the world they 
 define is not a complete whole, but only a part which stands in 
 necessary relation to other parts that cannot be brought under 
 the same laws. For if this should prove to be the case, then 
 the truth which is found in these sciences must be regarded as 
 a truth of abstraction ; or in other words, as the necessary 
 logical result of an hypothesis, which is not absolutely or uni- 
 versally true. If the physical world is not really a separate") 
 world, not a complete whole limited and terminated by its own 
 principles, these principles cannot be regarded as true, but only 
 as the expression of a partially false, though it may be necessary, 
 hypothesis. 
 
 The meaning of this statement will be better understood, if f°^^°l^ 
 we consider that, even within the limits of the sciences in *^° '^^■'"'o"- 
 question, we find it necessary to work upon abstract hypotheses, 
 which we have to correct as we advance by the re-introduction 
 of qualifying truths which were for the moment neglected. In 
 other words, we have, for the purposes of a particular science, 
 to state as absolute or universal truth what we are afterwards 
 obliged to acknowledge to be true only within conditions which 
 are never exactly fulfilled. Or we are obhged to treat as whole 
 realities, as, to use Spinoza's language, res completae, existences 
 which we are afterwards obliged to recognise as incapable of 
 being severed from their connexion with other elements in a 
 wider whole. Thus Arithmetic, e.g., is an abstract, and so far a 
 > hypothetical, science — a science based upon a hypothesis which 
 
30 INTKODTJCTION. °^-*-^- 
 
 is not absolutely true. The computations of Arithmetic would 
 be exactly true, only if the beings and things to which they have 
 to be applied, the beings and things of the natural and spiritual 
 world, were mere units, having no relations except that they 
 are capable of being externally added to one another ; and if, 
 when thus added, they produced no further effect upon each 
 other. Now this is not strictly true of anything in the world, 
 not true even of two pieces of what we call dead matter, which 
 are what they are just because they attract and repel each other 
 chemically or mechanically, and which, when combined, are 
 never merely the sum of their parts. It is still less true of , 
 organic beings either in their relation to each other or in the 
 relations of their different parts. Por in the case of such 
 beuigs, there is a sense in which, in spite of Arithmetic, the 
 whole is in every part. The value of the category of number 
 descends as we ascend in the scale of existence, and, though it 
 never ceases to have some place in our knowledge even of the 
 highest spiritual relations, yet in regard to them it merely 
 enables us to give a first superficial characterisation, through 
 which we pass rapidly to more adequate views. Often, in such 
 cases, the idea of number appears to be introduced rather to 
 make its inadequacy an epigrammatic starting-point for the 
 apprehension of that which is beyond number, as when it is 
 said that " one with God is a majority '' ; or as in the dictum 
 of Novalis, " it is certain that my conviction gains infinitely the 
 moment one other soul believes with me." In all these cases, 
 of course, the arithmetical relation is present, .but in no case is 
 it present alone, as in the numbers of Arithnaetic. Hence the 
 value of Arithmetic, in helping us to explain any phenomenon, 
 is in inverse ratio to the complexity and comprehensiveness of 
 the phenomenon itself In the same way, the physical and 
 chemical aspects of matter, as in their increasing complexity 
 they presuppose and transcend each other, force,&pon us a con- 
 tinual correction of the abstraction which was necessary in order 
 CO the first development of special sciences. And, even to ex- 
 
I- THE IDEA OP CRITICISM. 31 
 
 plain the simplest substance of the inorganic world, we need to 
 reinstate the unity which we have broken up for the purposes 
 of investigation. 
 
 It may, however, be argued that this abstraction, and there- Are the so- 
 
 . p . p ■ T called concrete 
 
 fore imperfection of science, disappears, when we reach the sciences an 
 
 ^ ' J. J. J exception? 
 
 more complex sciences which deal with actual objects in all 
 their physical qualities. Thus it might appear that if pure 
 Mathematics is abstract, its abstraction is corrected by Physics, 
 and the correction is completed by Chemistry; for in the 
 range of these sciences we deal with every quality which 
 belongs to what we call dead matter, and though these quali- 
 ties may be said to form only one part of what is to be 
 explained in the nature of living beings, yet they form the 
 whole nature of things that are inorganic. And the same 
 remark might be applied to all the successive complexities 
 introduced in the life of plants, of animals, and of men. In 
 each stage it might be said that the scientific principles, which 
 are abstract when regarded as explaining only part of the 
 nature of the more complex being, and which, therefore, in 
 that reference may be treated as hypothetical, are concrete and 
 capable of expressing the whole truth that can be ascertained 
 in reference to the being which is less complex. Thus the 
 thing or being, which has no higher quality than that which it 
 is the object of a particular science to investigate, may be fully 
 defined without trespassing beyond the limits of that science, 
 or introducing any categories or principles not employed 
 in it. 
 
 To this, however, it may be answered that, even if we Even they are 
 
 ' " abstract. 
 
 confine ourselves to the natural world, it is impossible to 
 conceive the inorganic without reference to the organic which 
 determines it, and by which it is determined. It is true that 
 there are objects which exhibit in themselves only what we 
 caU physical properties, but it is not true that the physical or 
 merely inorganic can be completely and adequately compre- 
 Jhended without reference to the organic, with which it is 
 
32 INTEODUCTION. chap. 
 
 united in one world. To say that the inorganic world might 
 exist without the organic, is to turn a convenient abstraction 
 into a res completa. The distinction of secondary from 
 I primary qualities, the former of which are supposed to be 
 relative to a sensitive subject and the latter to belong to things 
 in themselves, and the use made of this distinction in Physics, 
 already shows us that the physical world, regarded apart from 
 its relations to organisms, is no longer the physical world 
 of our experience, but a world of pure mechanical attractions 
 and repulsions, of motions that never reveal themselves as 
 colours or sounds, as pressure or heat, as taste or smell. And 
 if we are still able to separate the relations of sensible objects in 
 space, as moving and acting and reacting on each other, from 
 their relation to the organs of sense, it is not because we can 
 treat matter in its primary qualities as a thing in itself, 
 independent of relation to anything not inorganic, but because 
 these qualities more obviously involve a relation to a subject 
 which is not merely sensitive but conscious.^ It follows 
 therefore, that the truths of Physics are hypothetical, not 
 merely as regards the living organism, which is only partially 
 explicable through them, but also as regards the inorganic 
 world itself We are taking for absolute truth a false abstrac- 
 tion — though an abstraction which is serviceable for the 
 purposes of a particular science — when we regard the inorganic 
 
 ^ I cannot expect that what is said above -will he of itself clear or convincing. 
 To say that the inorganic is essentially relative to the organic, and that the 
 neglect of that relation, however useful and even necessary for the purposes of 
 physical science, leads to an abstract and incomplete view even of matter, is a 
 statement which involves such an inversion of ordinary modes of thought and 
 ordinary methods of explanation that it cannot be expected to pass without 
 ^ challenge. And to say further that both inorganic and organic alike are 
 relative to the unity of consciousness, in such a sense that no ultimate 
 explanation of either can be given apart from this relation, involves not only 
 the whole argument of Kant, but carries it to a result which even Kant 
 never completely accepted. The main objections to such a theory and 
 the main reasons for it will be discussed in the sequel. Here it is stated 
 merely with a view to meet an objection and to complete the preliminary 
 expression of that point of view which, as I think, a consistent critical philo- 
 sophy must take up. 
 
J- THE IDEA OP CRITICISM. 33 
 
 as necessary for the organic, but deny that the organic is 
 equally necessary for the inorganic. If, therefore, there be 
 any new principles necessary to explain the organic world, 
 which go beyond the principles of pure Physics and Mathe- 
 matics, it is a mere illusion to say that we can completely 
 explain the inorganic world without these principles. On the 
 contrary, all the parts of the one world, which includes both 
 the organic and the inorganic, ultim.ately need for their 
 explanation the highest category which is necessary to explain 
 any one of them. 
 
 The same idea may be still more obviously seen to hold dan nature be 
 
 '' explained 
 
 good in relation to all attempts to explain the natural world 3^^*^°°' 
 apart from its relation to the principle manifested in the 
 life of self-conscious or spiritual beings. There is no possi- 
 bility of explaining nature apart from spirit if spirit is more 
 than merely a part of nature on a level with the other parts, 
 or if there is anything in it that goes beyond the limits of 
 what is in them. We cannot explain the latter without the 
 former, unless we can explain the former by the latter. If 
 man is not merely the child of nature, capable of complete 
 explanation by its physical and vital agencies, then nature 
 cannot be taken as a system which is complete in itself apart 
 from man, or in which the presence of man is but an accident. 
 The strange conclusion of those Physicists who, finding them- 
 selves unable to explain consciousness as one of the physical 
 forces, were driven by the necessity of their logic to the 
 hypothesis that consciousness produces no result at all in the 
 world which it contemplates, illustrates this difficulty. Science 
 must inevitably treat the spiritual either as natural or as non- 
 existent, if it is not prepared to admit the imperfect or merely 
 abstract truth of its own principles. There are no alternatives 
 but either to press the physical explanations j to their , last 
 result and so to reduce the spiritual world to the natural ; or, 
 on the other hand, to admit that there is, properly speaking, 
 no such thing as a merely natural world, though in the > 
 VOL. I. c 
 
34 INTRODtrCTION. CS&.P. 
 
 necessary abstraction of science, we must speak as if there were 
 one. If, therefore, it be once admitted, that there is a limit to 
 the validity of physical explanations, and that therefore they 
 are in a sense hypothetical, we are forced by the very interest 
 out of which science arises, to seek for a deeper principle, in 
 which that limit finds its explanation. 
 Keed to 'level xhc idea lust suffs'ested is one which we cannot stop to 
 
 up m Science. '^ ~^ 
 
 consider fully, still less to justify ; but it may be useful to 
 give it clearness by comparing it with the idea which is most 
 frequently set up in opposition to it. The prevailing method 
 of explaining the world may be described as an attempt to 
 "level downwards," i.e., to take the lowest forms of existence 
 as the explanation of those that stand higher in the scale, and 
 to universalise the principles which are implied in the scientific 
 determination of these forms. In this way it is sought to 
 carry back Psychology to Biology, Biology to Chemistry, and 
 Chemistry to Physics. The doctrine of development, inter- 
 preted as that idea usually is interpreted, supports this view, 
 as making it necessary to trace back, higher and more complex 
 to lower or simpler forms of being ; for the most obvious way 
 of accomplishing this task is to show analytically that there is 
 really nothing more in the former than in the latter. And 
 this again seems to find a kind of empirical support in the 
 geological or astronomical regress, which carries us back to a 
 time when the earth was yet incapable of supporting even the 
 simplest forms of life and when, therefore, the inorganic 
 existed without the organic. This view of things, however 
 plausible it may be, cannot, for reasons already partially 
 stated, be allowed to be more than plausible. For though, by 
 an act of abstraction, it may be possible to treat the inorganic 
 world as if it were essentially unrelated to the organic, and 
 though we may picture to ourselves a time when the world 
 ^ was stin uninhabited even by the simplest forms of life, the 
 world which we thus represent is a world of which we can 
 say nothing, except in terms that necessarily imply a relation 
 
I- THE IDEA OF CRITICISM. 35 
 
 to that sensitive and conscious being which we pretend to 
 remove from it. If we could carry out our abstraction to its 
 legitimate result, we should ijpso facto annihilate the very- 
 thought of an inorganic world, or of any world or object at all. 
 We must, therefore, recognise that the inorganic world is only 
 by abstraction separated from the organic, and the unconscious ^ 
 object from the consciousness for which it is object. Not 
 only, therefore, must we deny that the explanation, which 
 seems to be sufficient for matter, is sufficient for life and mind, 
 but, since matter is necessarily related to mind, we must deny 
 that the explanation in question is sufficient even for matter. 
 We must, therefore, invert the method of explanation which 
 has just been referred to, and we must say that the ultimate 
 interpretation even of the lowest existence in the world cannot 
 be given except on principles which are adequate to explain the 
 highest. We must " level up " and not " level down " : we 
 must not only deny that matter can explain spirit, but we 
 must say that even matter itself cannot be fully understood 
 except as an element in a spiritual world. 
 
 In the above statement I have anticipated the result of an f-^^^^''^^^^. 
 inquiry which has yet to be made. For the present I shall Svanoe!™"'^ 
 use it only to illustrate the way in which the progressive 
 movement of science, as it advances from the lower to the 
 higher forms of existence, ever gives new occasion for a critical 
 regress, in search of categories and principles higher than those 
 which have been already used. Such a regress, as we have 
 seen, was involved even in the first movement of thought which 
 originated mathematical science. It was repeated in the 
 determination of the conceptions that furnish the basis and 
 rule for the investigations of physical science. And it has 
 become necessary to repeat it again at the present time, 
 because the great progress of physical science has led to an 
 attempt to extend its principles not only to life, but even to 
 mind. The first regress brought order into the mathematical 
 Sciences by directing attention to the pure idea of space and 
 
36 INTRODirCTION. chap. 
 
 to the principles of its determination, as well as to all the 
 other laws that determine the relations of things as mere quanta. 
 The Pythagorean doctrine, that reality consists in numbers, 
 expressed the consciousness of a time when this first step in 
 science was achieved, and when it was not yet perceived that 
 it is an explanation, not of the world as a whole, but only of 
 one aspect of it abstracted from all the others. The category 
 of quantity alone seemed to be needed to make all nature 
 intelligible. It might be said with no great unfairness that 
 the science of the Ancients never got beyond this stage, 
 and even that the mistake was a fortunate one for the 
 progress of that mathematical science, which was required as 
 the first condition of all other science. But physical science 
 could not be prosecuted with any success until the abstractness 
 of this method of explaining things was recognised, until it 
 was made manifest that its solutions did not apply to the 
 material world as we know it, but only to an 'ideal world of 
 pure quantity, pure space, and pure time. "When this was 
 seen, the transition from mathematical to physical conceptions 
 became a necessity; in other words, a principle had to be 
 sought which should at once limit the mathematical conception, 
 of matter and bring that conception into subordination to a 
 more complex view of it, as not merely extended but resistent, 
 not merely passively capable of being divided and re-united, 
 but possessed of active energy to repel and to attract. The 
 modern philosophers, Descartes and Spinoza, were still to some 
 extent under the sway of the mathematical synthesis of the 
 Ancients, as is shown by their identification of matter with 
 extension, and by their treatment of it as essentially inert.' 
 But the abstract discussions in which they were engaged, and 
 especially the controversy which arose between the followers 
 of Descartes and those of Leibniz as to the idea of force, eon- 
 
 ^ Spinoza still speaks of res exteima, though in one of his letters (70) he. 
 objects to the Cartesian view of matter as inert. To this he was naturally led 
 by the parallelism of' extension and thought. 
 
I- THE IDEA OP CRITICISM. 37 
 
 tain many traces of the critical process by which a new 
 conception of the material world established itself. Such 
 discussions were partly occasioned by the advance of physical 
 science, but partly also they assisted that advance by making 
 scientific men conscious of the new categories and methods 
 necessary for the new region of investigation into which they 
 were entering. It is only a narrow view of results without 
 reference to the categories or principles that make them 
 possible, that could overlook the great though indirect effect of 
 these abstract discussions in giving distinctness of aim to 
 science. ' For nature only answers the questions that are asked 
 of her, and the reason why discoveries are made at a particular 
 time lies, not merely in the increased knowledge of facts, — -which 
 in themselves have no meaning unless they are collected with 
 a view to some particular problem to be solved by them,— but 
 in that ripening of the intelligence to self-consciousness which 
 causes certain questions to be asked, a ripening which ex- 
 presses itself above all in the progress of philosophy.^ It is 
 perhaps a significant fact, that it was on the controversy 
 between the Cartesians and the Leibnizians as to the nature of 
 ^, force that Kant wrote his first treatise. He who in his youth 
 discerned the importance of the transition from mathematical 
 to physical conceptions, and the necessity of subordinating 
 the former to the latter, was destined in his age to point the 
 way to a more difficult transition, the transition from the 
 conceptions which are sufficient to explain the material world, 
 taken as an independent thing in itself, to that idea of it 
 which is necessary when it is discovered that that world has 
 essential relations to a conscious subject. 
 
 Now there is a certain parallelism between the two transi- J^^l'];^;^,^ 
 tions iust spoken of The consciousness of the necessity of "^''S"™'' 
 
 . . _ . regress. 
 
 advancing from mathematical to physical prmciples arose m 
 
 I Another instance of this may be foiind in the development of the idea of 
 Evolution in the writings of Kant and his Idealistic successors prior to the 
 • great scientific movement which that idea has inspired. 
 
38 INTRODUCTION. CH'^^- 
 
 connexion with an attempt to extend the former beyond their 
 proper sphere, and to treat them as adequate for the complete 
 explanation of the material world. For this led immediately 
 to a perception of the inadequacy of these principles, or in 
 other words, of the way in which facts must be distorted in 
 order to make them explicable by such principles. The 
 extravagance of the Cartesian hypothesis itself awakened a 
 protest and a controversy, which was not settled until the 
 limits of a merely mathematical explanation of physical pheno- 
 mena were established, and the idea of quantity was 
 subordinated to the idea of force or physical causality. In 
 like manner, the necessity for an advance beyond physical 
 principles was first felt, when the attempt was made to extend 
 them to a sphere in which they ceased to be adequate. Physical 
 science had gained its triumphs mainly under the guidance of 
 the principles of the permanence of substance, of the necessary 
 connection of the successive states of substances under the law 
 of cause and effect, and of the reciprocal determination of the 
 states of coexistent substances by each other. It was by 
 following out the chains of reasoning suggested by these 
 principles, that all the great discoveries of the nature of the 
 material world had been made. But the impulse of success 
 naturally carried scientific men beyond the inorganic, and even 
 the material, world, and made them reluctant to admit that 
 there was any subject to which they were inapplicable or 
 inadequate. Hence it became necessary to inquire whether 
 these principles were really universal, or whether they had 
 only seemed to be so, because they were hitherto applied to 
 the material world alone, and even to it only from an abstract 
 point of view from which its whole nature could not be under- 
 stood. For the extension of physical principles into the region 
 beyond Physics, was not unresisted; but on the contrary, it 
 was met by an opposite dogmatism, already in possession of 
 the field and supported at once by morality and religion. 
 Further it was found that in this region the principles so 
 
I- THE IDEA OE CRITICISM. 39 
 
 fruitful in Physics could no longer be used with the same 
 effect, but that they seemed themselves to create a dogmatic 
 system of opinions, quite as much open to attack as that to 
 which it was opposed. The intensity of the conflict which 
 thus arises finds its measure only in the importance of 
 the interests arranged on either side. On the one side, the 
 intolerance, with which the scientific man regards any refusal 
 to admit the universality of his method, really springs from 
 one of the deepest intellectual instincts of man, which will not 
 let him treat anything as truth that is not universally valid ; 
 and with this instinct is combined that confidence in his own 
 principles which a long experience of their power to unlock the 
 secrets of nature has naturally produced. On the other hand, 
 those who resist the extension of these principles to the new 
 region, though often unable to oppose to them any distinct 
 principle of their own, are supported by a consciousness of the 
 facts of their own spiritual life, and by a perception that these 
 facts, which to them are the most certain of all, must be 
 treated as illusory if the claims of science be admitted. Yet 
 they cannot but see that a merely defensive attitude is weak 
 and intellectually untenable, and that it is impossible con- 
 clusively to repel the scientific explanation of mind, except by 
 showing that it is not a sufficient explanation even of matter. 
 The world is one, and admits of only one ultimate principle of 
 explanation, and if the claims of science are to be repelled, 
 they must not only be resisted in one sphere but refuted in 
 all. Each mode of thought is thus driven by an inner neces- 
 sity, not only to maintain its own ground but to assail its 
 rival, and out of the conflict arises that " Antithetic of 
 Eeason," which has been the deepest source of modern scepti- 
 cism. Nor can such scepticism be overcome except by a 
 critical regress whiclr shall discover the basis and the limits of 
 the physica,l conceptions of nature, and so bring nature itself 
 into its true relation to spirit.^ 
 
 ^ Of. the chapter of the Gritique upon " The interests of reason in this its 
 antinomy." A. 462 ; B. 490. 
 
40 INTRODUCTION. chap. 
 
 Greater im- The necessitv of Criticism is, then, in a sense, peculiar to 
 
 portanceofthe 
 
 modern criti- ^jje modcrn time. It is true, indeed, that all the great advances 
 
 cal regress. " 
 
 of knowledge have been made on the principle of " reculer pour 
 viieiox sauter." They have been regresses made necessary by 
 the fact that the principles previously treated as universal were 
 beginning to be found ineffectual, because they were used 
 beyond the limits within which they are valid. But the 
 necessity for such a regress has never before been felt so 
 strongly, because it was never before the case that the principles 
 claiming to be universal could appeal to such a long record of 
 successes in the explanation of facts, and because, on the other 
 hand, these claims never before came into collision with in- 
 terests and beliefs which had so deep a hold upon the human 
 spirit. The very development of our knowledge of the material 
 world has forced us to ask with a new meaning whether the 
 world is merely material, or at least, whether it can be com- 
 pletely explained on the principles which have been found 
 adequate to the explanation of the material world as such. 
 In earlier times, the consciousness of the natural and of the 
 .spiritual were, so to speak, fairly balanced against each other, or, 
 if there was a preponderance, it was rather on the side of the 
 spiritual. Their own consciousness occupied men so fully that 
 nature seemed to be a mere attendant of their lives, without 
 any independent being or power. A facile anthropomorphism 
 either clothed natural phenomena with a vesture of humanity, 
 or reduced them into secondary instruments of spiritual powers 
 similar to those that ruled the life of man. Now, however, in 
 consequence of the necessary order in which the sciences are 
 developed, the tables have been turned. The natural sciences, just 
 because of the greater simplicity of the principles on which they 
 are founded, have been earlier in striking into the sichere Gang 
 cler Wissenschaft. And their steady advance has redeemed so 
 many and so great departments of study from the anarchy of 
 the intdlectus sibi permissus, and turned them into, secure and 
 intelligible possessions of human thought, that it seems all but 
 
THE IDEA OP CRITICISM. 
 
 41 
 
 impossible to resist their claims to pass the boundary which 
 has hitherto been maintained against them — maintained, too, 
 as it has mainly been, by the undrilled battalions of unscien- 
 tific opinion, rather than by any rival army of science. Yet 
 the passing of this boundary means nothing less than that 
 spirit shall be included in nature, and that the methods and 
 principles, which have been found sufficient to explain the 
 latter, shall be treated as universal, and used also to explain 
 the former. Even the maintenance of the status quo, still more 
 the restoration and permanent confirmation of the old supre- 
 macy of spirit over nature, is impossible, except by a regress of 
 thought, which shall discover at the very basis of the concep- 
 tion of nature a still higher principle of interpretation. If it 
 is true that science has raised an abstraction, which includes 
 only part of the elements of experience, into a principle for its 
 universal explanation, this can only be shown by a deeper 
 examination of those general principles, which are involved in 
 the nature of the intelligence and of the intelligible world, on 
 which science and ordinary experience alike are built, and by 
 which 'all their special truths must be ultimately interpreted. 
 
 Criticism, in this sense, has its source ultimately in the anti- Antinomy™ 
 nomy between the principles of physical science and that un- 
 scientific consciousness of spiritual reality which is expressed 
 in rehgion and morality ; and immediately in the scepticism 
 which is due to this antinomy, and which, rejecting both the 
 unscientific and the scientific view (because of their opposition 
 to each other), takes refuge in Agnosticism. It is an attempt 
 to solve this antinomy by seeking out the sources of it or the 
 unity that transcends it. It therefore seeks, in a sense, to get 
 beneath both the ordinary and the scientific consciousness, but 
 yet it cannot, as we have seen, propose to itself to go outside 
 of them. It cannot find any point of view outside of experi- 
 ence from which to criticise it. It can only go back, in 
 Kantian language, upon the " conditions of possible experience 
 • or knowledge," upon the principles that are involved in every 
 
 wN- 
 
42 INTRODUCTION. *^S-*-^- 
 
 intelligible consciousness of things. It can only retrace the 
 road of thought to the point from which the divergence of 
 opposite dogmatisms begins, and so endeavour to find, in the 
 general presuppositions under which we know both nature and 
 spirit, in their opposition and in their unity with each other, 
 the key to the inconsistency of the views which are presented 
 i to us from two opposite quarters. But, this being presupposed, 
 we see at once the twofold bearing of the proposed inquiry, as 
 an attempt to find a way heyond natural science to some con- 
 clusion or conviction, whether scientific or not, as to the things 
 of the spirit, by going hack upon the preconditions of natural 
 science itself. It is an attempt to find a key to the difference 
 and opposition of the two divergent forms in which knowledge 
 or belief presents itself, by asking for a definition of the genus 
 within which both species must fall. Por it is obvious that 
 if there is any reason for the assertion that the principles of 
 physical science are not capable of being applied, e.g., to man's 
 moral and religious experience, this reason must be sought in 
 the unity which embraces all forms of experience ; and, if con- 
 versely there is no room left in the very idea of experience 
 for such an opposition of its forms, the opposition must itself 
 be pronounced to be an illusion. The freedom of spirit can 
 vindicate itself against the necessity of nature, only if it is 
 possible to lift the controversy into a region in which those 
 two are no longer left dogmatically opposed to each other, but 
 placed in due relation through the one principle which explains, 
 the possibility of each kind of experience, or even, if it be so, 
 , of seeming experience. 
 Regreaa ^^ The rcsult of all that has been said is thus to show the two- 
 
 necessary to 
 
 progress. f^j^ naturc of the work of criticism. Criticism is always the 
 result of the fact that the intelligence has found its way blocked 
 by some difficulty, which has awakened a suspicion against 
 the universal applicability of the categories or methods which 
 it had been using. In this sense criticism was at the very 
 birth of science, and it has mediated every transition to a new 
 
I- THE IDEA OF CRITICISM. 43 
 
 point of view, by which science has widened the scope of its 
 investigations and brought the concrete fact of the world in its 
 diversity and unity more definitely within the reach of the in- 
 telligence. But in a peculiar sense it may be said to be the 
 special intellectual task of the present age, just because the 
 special obstacle to science with which we have now to deal, 
 lies in the opposition, and — for the ordinary consciousness at 
 least — the fixed opposition, between the material and the 
 spiritual world. This last and greatest division of thought 
 against itself cannot present itself to us without awakening a 
 perception of the greatness of the interests, which are here 
 apparently set against each other. But as there can be no 
 absolute oppositions within the intelligible world, i.e., no oppo- 
 sitions which have not a principle of unity beneath their dif- 
 ferences, — and as, indeed, such a unity is implied in the state- 
 ment of any difference as an intelligible difference, — so we are 
 obliged to think that the key to the problem, the means of 
 reconciling the opposition or removing it, will be found, if we 
 can clearly determine what that principle of unity is. For 
 such a principle must enable us, and it alone can enable us, to 
 define the opposing elements in their relation to each other, or 
 to determine the limits of their respective validity. The critical 
 regress thus raises a new question for philosophy^ or it raises 
 the old question in a way in which it was put by no previous 
 philosophy. It is a regress upon the beginnings of the know- 
 ledge which we have, with a view to a kind of knowledge 
 which we have not, at least in the form in which we desire it. 
 For the ultimate aim of the metaphysical or ontological inves- 
 tigations which immediately seek only for the basis of Physics 
 and Mathematics, in other words, of the science of nature, is to 
 prepare the way for a new Psychology, Cosmology, and Theo- 
 logy — or at least, for some new determination of the natural 
 and the spiritual worlds in their relations to each other, which 
 shall take the place of the so-called sciences that have 
 hitherto borne these lofty names. In short, the ultimate aim of 
 
44 INTRODUCTION. chap. i. 
 
 criticism is to settle the possibility of an idealistic interpreta- 
 tion of the universe, and, if it is possible, to determine the 
 form which such an interpretation must take. Immediately, it 
 has to do with what have been called " the first things," but 
 its ultimate aim is " the last things " of the intelligible world. 
 It is a new Logic which, preparing the way for a new view of 
 man and God, casts a new light also upon nature ; for, as we 
 have seen, nature must take a new aspect if it be conceived as 
 standing in a necessary relation to spirit and not as including 
 it. A nature so related can be no closed system of purely 
 physical relations ; it must be conceived as part of a greater 
 whole, and it may even be the case that, in the ultimate 
 account of it, we may have to regard it as the necessary mani- 
 festation of spirit. To say this, however, is to anticipate the 
 conclusion of the inquiry which we are now beginning. 
 
45 
 
 CHAPTEK II. 
 
 kant's relation to his time, his life and chaeactee. 
 
 HEG-EL speaks of certain great writers who are like knots in Kace o£ Kant 
 ■^ ° in the History 
 
 the tree of human development, at once points of con- "' Philosophy, 
 centration for the various elements in the culture of the past 
 and starting points from which the various tendencies of the 
 new time begin to diverge. In the history of thought there is 
 no one to whom this saying can be applied with more 'confi- 
 dence than to Kant. In the German phrase, he " makes an 
 epoch " — the end of one mode of thought and the beginning 
 of another. His works form a kind of bridge by which 
 we pass from the ruling conceptions of the eighteenth to 
 those of the nineteenth century. And the reason is, that 
 he brings together all the elements of the thought of the 
 eighteenth century in such a way that a new and higher 
 thought springs from their union. To use the words of 
 Green, he "read Hume with the eyes of Leibniz and Leib- 
 niz with the eyes of Hume,'' and therefore, " he was able to rid 
 himself of the presuppositions of both, and to start a new method 
 of philosophy." i In other words, he effected such a synthesis 
 of the different tendencies of his time as carried him beyond 
 their one-sidedness, and thereby lifted philosophical discussion 
 to a new level. There is even some excuse for a German 
 writer who refuses to take account of any philosophical thinker 
 1 Green's Works, I. 3. 
 
46 INTKODUCTION. CHAP. 
 
 after Kant, unless he can be shown to have listened to Kant's 
 lesson. A modern philosophy may not be Kantian, but it 
 must have gone through the fire of Kantian criticism, or it will 
 almost necessarily be something of an anachronism and an 
 ignoratio elenchi. 
 TheAgeofEn- How are we to describe the great change which came over 
 
 lightenment. ... 
 
 human thought towards the end of the last, and the begmnmg 
 of this century ? In general terms we may say that it was 
 a change from division to reconciliation, from Individualism and 
 Atomism to a renewed perception that the whole is prior to the 
 parts, and that individual independence must rest on social unity. 
 Or, to put all in a word, it was the substitution of the idea of 
 organic unity and development for the idea of the mechanical 
 combination of reciprocally external elements. The eighteenth 
 century called itself the " Age of Enlightenment," nor can there 
 be any doubt that it loved light and hated darkness. But this 
 love and this hate often misled it into denying the existence of 
 anything which it could not see clearly. It insisted on abolish- 
 ing mystery, and it regarded as mystery everything which was 
 not finite, everything which could not be set by itself and 
 clearly pictured by the sensuous imagination or defined by the 
 logical understanding. It favoured a way of thinking which 
 was clear and definite, but at the same time deficient in depth 
 and suggestiveness. Art-critics speak of pictures as "wanting 
 atmosphere," when the figures and forms in them stand out in 
 hard outlines as if broken off from each other, instead of show- 
 ing the graded transition and continuity of nature. A similar 
 charge may be brought against many of the most powerful 
 writers of the eighteenth century, who carry the desire for 
 clearness to such a point, that they seem to make it a rule of 
 composition that every sentence should be completely intelligible 
 by itself, even if isolated from its context. And the same 
 tendency limits the harmony and dries up the inspiration of 
 poetry, producing a kind of verse, of which Pope's is the most 
 consummate type, in which rhetorical point and emphasis takes 
 
II- LIFE AND WORK OP KANT. 47 
 
 the place of imaginative insight, and every thought and 
 image, and almost every couplet stands stiffly by itself, so that 
 the whole can scarcely be said to be more than the sum of 
 the parts. The great words of the Shakesperian age, winged 
 with music and laden with meaning, are silenced, and in place 
 of them we have a definite lesson, enforced in lucid and pointed 
 language, which can be fully understood at the first hearing. 
 
 It would be absurd at this date to repeat the invectives of Good and bvu 
 
 of the En- 
 
 Carlyle against the godless age, in which religion and poetry lightenment. 
 were extinguished and mechanism took the place of life. It is 
 almost a law of human development that men should be 
 unjust to their immediate predecessors, from whose yoke they 
 have had to emancipate themselves ; but it is irrational to con- 
 tinue the injustice after spiritual independence has been secured. 
 We can now afford to recognise the great step in the liberation of 
 humanity which was taken in the eighteenth century ; and we 
 can do so without denying the truth of much that has been 
 said on the opposite side. It was the age of Individualism, of 
 Secularism, even, in a sense, of Atheism. It sought truth by 
 dividing and isolating its parts, with the result often that the 
 spiritual unity of truth disappeared in the process. This 
 method was fatal to its insight into the higher life of men, and 
 it was not without an unfavourable influence on its view of 
 nature. For though it is that which is organic, that which has 
 the unity of hfe, and above all, of conscious life, which suffers 
 most injustice when it is treated as a mere sum or col- 
 lection of parts externally related to each other, it is not 
 possible that even the inorganic should be completely explained 
 apart from any reference to its connexion with life and mind. 
 For the world is one world, and it is impossible to reach the 
 ultimate interpretation of any part or aspect of it, if we neglect 
 its unity. Or, to put the same thought from another side, every 
 object or system of objects must, in the last resort, be regarded 
 as a microcosm, which is not only connected with the whole, 
 •■ but in some sense reflects the whole ; and if we insist on treat- 
 
48 INTRODUCTION. chap. 
 
 ing it as a " thing in itself," as a thing which is completely 
 determined apart from all such connexion or reflexion, we 
 empty it of its highest meaning. The last word cannot be said 
 of anything except in the light of the relation of all things to 
 each other and to the mind that knows them, and the thought 
 that neglects this ultimate relativity must in the long run 
 narrow and externalise our view of everything. For, while 
 there is nothing superficially so distinct and intelligible as the 
 finite separated from the infinite, there is nothing, when we 
 come to think it out, so obscure and unintelligible. In truth, 
 the clearness of isolation is really due to the ignoring of the 
 difficulty. We can escape the endless relativity which is hid 
 beneath the separate existence of things, only by resolutely 
 keeping to the surface. In this sense Kant justly says that 
 " many a book would have been more intelligible, if its author 
 had not laid himself out to be intelligible at all costs." A 
 writer of true insight will not be vague and indefinite; he will 
 set one object distinctly before us, but he will not tear it from 
 its place in the whole; he will make us feel that it reaches 
 beyond itself and takes hold of other things. Like the healthy 
 eye, his thought will embrace much more than that which is the 
 immediate object in the focus of clearest vision. The world 
 will be for him a continuum, and not a mere collection of in- 
 dependent and externally related objects. 
 "ev?lopme°nt'" ^ct wHle all this is true, while the method of division and 
 of Thought. jgQi^(;jQjj necessarily tends to superficiality, and while it is im- 
 possible ultimately to separate finite things from each other and 
 from the unity of the whole without at least a temporary loss 
 to spiritual life, there is undoubtedly a sense in which such 
 isolating thought is a necessary stage in the development of the 
 human mind. The immediate, intuitive solutions of the problem 
 of existence, which are the results of the first movement of con- 
 structive imagination and religious feeling, have this peculiarity, 
 that they deal with the whole problem at once. They orasp 
 the unity of the world with itself and with the inteUio'ence 
 
"• LIFE AND WORK OP KANT. 49 
 
 without any regard to what Bacon calls ■ axiomata media.' They 
 are everything at once or nothing. They meet every question 
 as to the finite by relating it to the infinite. In this lies their 
 strength and their weakness ; their strength, in so far as they 
 go at once to the centre and deal directly with the ultimate 
 truth of things ; and their weakness, in so far as it is 
 impossible' for the ultimate truth of things to be ade- 
 quately apprehended, when it is thus taken at a bound, 
 or when it does not come as the last interpretation of all 
 other trutl^ The highest unity can be reached only 
 through a fully articulated difference, and the method of 
 division and simplification is a necessary step in the progress of 
 knowledge. Divide et impera is the motto of science. It is by 
 abstraction and isolation of parts and aspects of the manifold 
 world, that it becomes possible to deal successfully with the 
 various problems it presents to us. And to separate the finite 
 from the finite is necessarily at the same time to sep&,rate the 
 finite from the infinite. Hence the individualism, the secular- 
 ism, the dividing and dissecting methods of the eighteenth cen- 
 tury, if in one point of view they narrowed and externalised the 
 thoughts of men, were not without compensating advantages. If 
 they were chargeable with an apparent, and in some respects a 
 real, superficiality of view in relation to the deepest problems of 
 man's existence, they quickened his apprehension of finite in- 
 terests and objects. If they made his consciousness of himself 
 • and of the world unpoetic and irreligious, at least they put to 
 flight the spectres by which his higher life had so long been 
 haunted. If they tended to substitute the mechanical for the 
 vital and the spiritual, yet in doing so they opened up the way 
 for the great achievements of physical science and the industrial 
 arts ; and we have to remember that though life and spirit are 
 more than mechanism, they are not without mechanism. To 
 make man. a free possessor of the finite world, to enable him to 
 understand and appreciate what is present to him in immediate 
 ♦experience, to banish the supernatural from the natural world' 
 VOL. I. i> 
 
50 INTRODUCTION. *^^^^- 
 
 and bring about a secure conviction of the reign of law — these 
 were no small gains, even if they were purchased by a temporary 
 weakening of the consciousness of the ideal meaning of life. And 
 even that consciousness could only gain in the long run by a 
 process that freed man's higher beliefs from the fatal alliance of 
 a fanaticism which confused the spiritual with the supernatural. 
 In separating the finite from the infinite, "Enlightenment" 
 prepared the way for a consciousness of their relations which 
 was at once purer and more rational, less dependent upon 
 illusions of imagination and less tainted with the base alloy 
 of " other-worldliness." 
 Protest of The eighteenth century was primarily the age of Enlighten - 
 
 u'ht'enmeli't mcnt. It is, howevcr, necessary to remember that, though this 
 was its prevailing character, the opposite aspect of the truth was 
 never left without a witness. On the contrary, throughout the 
 whole period we fifed a chain of writers and preachers, who kept 
 up a continuous protest against the spirit of the time. The one- 
 sidedness of those who held that "the proper study of mankind 
 was man " and his world — and not God, and that the highest 
 morality of man was to " cultivate his garden," was balanced by 
 a mysticism which regarded the things of this world as nothing. 
 Methodism in England, like Pietism in Germany, expressed that 
 revulsion which the Enlightenment produced in all deeply 
 religious minds. This movement, indeed, prevailed mainly 
 among the less educated classes, and it had no great literary 
 representatives. All the prominent literary men of the time 
 were enlisted in the service of the reigning ideas. But it is 
 the province of impulse, with its implicit reason, to counterbalance 
 the defects of conscious reason. A religious doctrine which 
 made man the passive vessel of divine grace, was the natural 
 counterpart and complement of the isolating individualism which 
 recognised no vital or organic relations between man and man 
 or between man and God. This undercurrent, which ran in an 
 opposite direction to the main stream of tendency during the 
 whole century, cannot be left out of view in any account of its 
 
II- LIFE AND WORK OP KANT. 51 
 
 main characteristics. For it was by means of the collision and 
 conflict of the two opposite influences, that the more comprehen- 
 sive ideas of the following era were developed. 
 
 Towards the end of the century we find many indications that f^^f^ relation 
 this new era was at hand. In France the Eevolution, which 'P''^'"^'^°^^- 
 put an end to so many political compromises, was heralded by a 
 series of writers- who broke through the theoretical compromises 
 of English Individualism. In Diderot's unlimited protest 
 against law and authority, and his unhesitating proclamation of 
 the doctrine that impulse is its own justification, we see the 
 individualistic tendency carried to an extreme in which it refutes 
 itself ; whUe Eousseau's somewhat hesitating enunciation of the 
 ideas of a raison commune and a volontd gdndraU, manifesting 
 themselves in and through the varying opinions and wills of 
 individuals, contains the germs of a new philosophy. In 
 Germany the hard dogmatism of Wolff, who had popularised, 
 and we might say vulgarised, the Monadism of Leibniz by leaving 
 out all its deepest speculative ideas, gradually gave place to 
 a vague and varying Eclecticism, which sought to combine the 
 elements of many philosophies without any definite principle of 
 reconciliation. Kant alone had the constructive power, the 
 speculative insight, the patience, to form a true estimate of these 
 various ideas and systems, and by combining the different tend- 
 encies of the eighteenth century, to initiate a new philosophical 
 movement. He alone saw how it was possible to unite the 
 characteristic ideas of the Enlightenment, and especially it? 
 rational conceptions of the order and connexion of finite experi- 
 ence, with a new vindication of those higher beliefs which the 
 Enlightenment had rejected. In this sense, we may fairly say 
 that, with the single exception of .Goethe who worked in another 
 field, Kant was the most potent of all the agents in the transition 
 from the ideas of the eighteenth to those of the nineteenth 
 century. 
 
 It must be added, however, that in some points Kant suffers DefootsofMs 
 for his position. That his writings, as above said, form a bridge 
 
52 INTRODUCTION. chap. 
 
 between the old and the new era, implies that he cannot be re- 
 garded as strictly belonging to either. He was all his lifetime 
 struggling to divest himself of the conceptions with which he 
 started, and to develop another order of ideas to which, how- 
 ever, he was never able to give free and unambiguous utterance. 
 It would not be altogether unfair to say that he was constantly 
 trying to pour new wine into old bottles. He uses the dead 
 scholastic phraseology of the "Wolfifian philosophy to express the 
 living thought of the new era, and there is, therefore, a fre- 
 quent struggle between his ideas and his expression of them. 
 Hence the many controversies about his meaning, some of 
 which it is almost impossible to settle, because that meaning is 
 continually in process of changing and deepening, so that the 
 letter halts behind the spirit. Sometimes his words would 
 induce us to give him credit for all that subsequent writers 
 have found in them, or developed out of them; sometimes they 
 tempt us to reduce his thought almost to the level of the dog- 
 matism he was combatting. Such ambiguities, however, are no 
 great hindrance to the student, if only he remembers their 
 cause ; for the important thing is not so much to find a precise 
 interpretation for every passage, important as that may be, as 
 to observe the direction and the manner in which his thought 
 is developing. Often the variation of his expression at different 
 stages is full of instruction, as it forces us to retrace the path 
 by which he advanced and to realise the difficulties involved in 
 the subsequent transition. Mr. Sidgwick, speaking of Political 
 Economy, says that in that science students are apt to overrate 
 the importance of finding correct definitions and to underrate 
 the importance of seeking them. With still greater truth it 
 may be said that in Philosophy results can mean little or 
 nothing to those who have not understood the process by which 
 they were reached. And in the in,terpretation of Kant, we are 
 not merely discovering interesting facts about the individual 
 development of a great philosopher, not merely following the 
 steps of one of the most patient and methodical of thinkers, 
 
"• LIFE AND WORK OF KANT. 53 
 
 who never took a step in advance until he was forced onward 
 by the growing thought within him; we are, so to spealc, 
 watching the process by which the new roads of modern thought 
 were made. We are thus shaken out of the tendency to take 
 for granted the ideas which are kindred with the spirit of our 
 time instead of reproducing them for ourselves, and we are en- 
 abled to appreciate the real value of these ideas and their rela- 
 tion to the past. The very imperfections of Kant's statement 
 are thus full of lessons, which we might not so readily gather 
 from the more consistent language of those who came after 
 him. 
 
 A short sketch of Kant's life may be useful as a preparation slowness of 
 
 Kant's mental 
 
 for a more definite discussion of the steps by which he found defeiopment. 
 his way to the idea of Criticism. He was born on the 22 nd 
 April, 1724, in the city of Konigsberg in the province of East 
 Prussia ; and during his long life of nearly eighty years he 
 never once crossed the borders of that province. He lived , a 
 quiet, studious life of teaching and writing, never taking part in 
 any of the great events of his time or coming into contact, 
 except officially, with any of the important actors in the 
 political world. He was slow in his mental, and specially in 
 his philosophical, development ; for, though in his early man- 
 hood he published several short treatises of some importance, 
 especially one in 1755 on the General Natural History avd 
 Theory of the Heavens, in which he anticipated the ideas of La- 
 place as to the formation of the solar and sidereal systems, his 
 serious philosophical writing did not begin till he was nearly forty 
 years of age, and it was not till his fifty-seventh year that he 
 published his first great work, the Critique of Pure Reason ; 
 and it was during the following twenty years that he displayed 
 his greatest literary activity in applying the principles of Criti- 
 cism to Morals, to Physics, to Aesthetics and to Theology. 
 Kant was no brilliant intellectual adventurer like Berkeley, who 
 had already conceived and uttered almost all his distinctive 
 ideas before he left the University. He was a slow, ■ deep- 
 
54 INTEODXTCTION. chap. 
 
 mining thinker who had not only one great thought, but a 
 whole philosophy almost full grown in his brain, before he 
 began to set any part of it before the public. 
 Influence of Kaut had a long and hard struggle with circumstances before 
 
 Schultz. ^ 
 
 he gained a position of independence. The son of a poor strap- 
 maker, who had a large and delicate family, he seems to have 
 owed his early education to the notice which he attracted from 
 Franz Albrecht Schultz, an eminent pastor and professor who 
 was also the head of the most important school of Konigsberg, 
 the Collegium Fredericianum. By his piety, intelligence and 
 practical energy, Schultz had gained the thorough confidence of 
 that strange conscientious despot, Frederic William, the father 
 of Frederic the Great ; and for about ten years he was main- 
 tained by that king as a kind of dictator of Konigsberg and 
 the province of East Prussia in all matters educational and 
 theological. A man like Schultz, of such original force and in 
 such a position, could not but impress the stamp of his character 
 on the intellectual life of a city like Konigsberg, cut off as it was 
 by its remote situation from extraneous iniluences. And when 
 we remember the mediating and reconciling work which Kant 
 was to undertake, we cannot but attach some importance to the 
 fact that his early training was directed by a teacher who was 
 the first, or one of the first, to conceive the possibility of com- 
 bining the two great spiritual forces which were then striving 
 for the victory in the German Universities, Pietism and the 
 Wolffian Philosophy. Throughout Germany these two forces 
 were opposed, and even bitterly opposed, to each other. For, 
 as has been already indicated, the Wolffian Philosophy was a 
 dogmatic Individualism in which the speculative elements of 
 the Leibnizian Philosophy were discarded, and all the truths of 
 the reason were brought to the bar of the understanding. On 
 the other hand, the Pietists, in recoil from the rationalistic 
 tendencies of the time, threw themselves into a religion of 
 feeling, and denounced the natural understanding as incapable 
 of measuring divine truth. Schultz, however, with something 
 
"• LIFE AND WORK OF KANT. 55 
 
 of a practical man's indifference to logic, endeavoured to com- 
 bine the good elements in each of these opposites. He was 
 a favourite pupil of Wolff, who is said to have singled him out 
 as the pupil who best understood his principles, and he was at 
 the same time deeply imbued with the spirit of Pietism. From 
 the former he derived enlightened views of education, and a 
 desire for a philosophical explanation or proof of his ideas ; 
 from the latter a fervid religious spirit and a belief in the effi- 
 cacy of a strict and somewhat ascetic, moral discipline. His 
 pastoral office brought him into contact with the pious strap- 
 maker's family, and led him to interest himself in the promising 
 boy, whom he found there and whom he admitted to his school. 
 It is probable, though we have no evidence of the fact, that he 
 watched over Kant's training, and did something to help him 
 in his hard struggle with circumstances. Kant at any rate had 
 so strong a feeling of obligation towards his former teacher, 
 that it was one of his latest projects, a project unfortunately 
 unrealised owing to the increasing weakness of age, to write 
 some account of Schultz's life and services. 
 
 In speaking of his own boyhood, Kant was wont in later ^^^^°°? "' 
 years to lay stress mainly upon two important influences : on the 
 religious spirit of his parents, and on the good classical training 
 which he received in the Collegium, Frederieianum. Of the 
 former, especially of the first simple lessons in religion which 
 he received from his mother, he could never speak without 
 emotion, "The religious ideas of those times," he said to a 
 friend on one occasion, " and the prevalent notions of virtue , 
 
 and piety could hardly be said to be either clear or satisfactory, 
 but the root of the matter was in them. Say what you will of 
 Pietism, no one can deny the sterling worth of the characters 
 which it formed. It gave to them the highest thing that man 
 can possess — that peace, that cheerful spirit, that inner harmony 
 with self which can be disturbed by no passion. No pressure 
 of circumstance or persecution of man could make them discon- 
 ^ tented, no rivalry could provoke them to anger and bitterness. 
 
56 INTRODUCTION. chap. 
 
 Even the casual observer was touched with an involuntary- 
 feeling of respect before such men. I yet remember what 
 happened on one occasion when difficulties arose between the 
 strapmakers and the saddlers in regard to their respective 
 rights. My father's interests were seriously affected : yet even 
 in conversation the difference was discussed by my parents with 
 such tolerance and indulgence towards the opposite party, and 
 with such a fixed trust in Providence, that, . boy as I then was, 
 the memory of it will never leave me." ^ 
 cai studies.^^^' The narrowncss of the strict discipline and pietistic teaching 
 of Kant's home, which was continued in the school, found a 
 corrective in the efficient classical training which he received 
 from one of the schoolmasters called Heydenreich. Kant and 
 two friends, David Euhnken and Martin Cunde, were inspired 
 with such enthusiasm that in the later part of their school years 
 they used to meet several times a week for the reading of 
 Latin authors who did not form part of the regular course ; 
 and the three boys in common resolved to devote themselves 
 to a career of classical study. Euhnken realised this youthful 
 dream and became in later years a distinguished scholar ; and 
 Kant was so far influenced by it that, when he went to the 
 University, he did not inscribe himself as a member of the theo- 
 logical or any other special faculty. And from these early 
 studies he derived a good knowledge of Latin literature and a 
 power of ready quotation from the Latin poets, which he retained 
 to his latest years. 
 
 K^ton. ^^^^^ ^^^ sf^^ool years were over, at the age of eighteen, 
 
 Kant's Kant began to attend the University, supporting himself mainly 
 
 scientific i, j. i • -.i ,. , , 
 
 studies. by teachmg, with some little help from an uncle. Of his col- 
 lege life we know very little, except that he gained the friend- 
 ship of the ablest of the Konigsberg professors, Martin Knutzen, 
 a man of no little importance in the history of the development 
 of the Wolffian school. Knutzen followed his teacher Schultz 
 in the effort to unite the formal method of the Wolffian philo- 
 ^ Rink's Ansichten, p. 13. 
 
"• LIFE AND WORK OF KANT. 57 
 
 sophy with the spirit of Pietism ; but while Schultz was 
 satisfied with an external combination, Knutzen, a man of 
 much greater speculative ability, made commendable advances 
 towards a real philosophical synthesis of the two elements. On 
 Kaint his influence feeems to have been considerable, though 
 Kant's thought went through so great a development in subse- 
 quent years that we can scarcely find any definite traces of the 
 ideas of Knutzen, beyond the general mediating tendency which 
 characterised them both. What is recorded as to their relations 
 is, that Knutzen urged Kant to the study of Newton, and gave 
 him the use of his own large library. The advice seems to 
 indicate that Kant's abilities showed themselves in the first 
 instance in the direction of mathematical and physical science, 
 rather than of philosophy. And this may also^ be inferred from 
 the fact that his first work was an essay, published in 1749, 
 three years after he had finished his University curriculum, in 
 which he discussed the opposing theories of the Cartesian and 
 the Leibnizian schools as to the way in which the force of 
 moving bodies is to be estimated, and tried to find a method of 
 reconciling them. Some six years after this youthful produc- 
 tion, his thoughts still ran mainly in the same direction. In the 
 General Natural History and Theory of the Heavens, published 
 by Kant in his thirty-first year, he extends the theory of Newton 
 as to the present movements of the solar and sidereal system to 
 the original formation of those systems. At the same time, he 
 attempts to show that this extension of the mechanical explana- 
 tion of the Universe does not affect the argument from design 
 or involve the entire rejection of final causes. 
 
 During the nine years before the publication of this treatise^ fri"*t-Docent 
 Kant resided as a private tutor in various families in the district 
 near Konigsberg. Of his relation to his employers or pupils we 
 know little or nothing. He himself declared afterwards that he 
 was one of the worst of tutors; for, though he had a clear theory 
 as to the method of teaching, he was unable to acquire the art of 
 faking himself intelligible to children. In 1755, he returned 
 
58 
 
 mTRODUCTIOlSr. 
 
 CHAP. 
 
 His political 
 interests. 
 
 Frederic the 
 Great. 
 
 to the University, took the degree of Doctor of Philosophy, and 
 produced a dissertation on metaphysics as the necessary quali- 
 fication for teaching. He had, however, even after this, to pass 
 through eleven years of poverty and hardship ere he received 
 his first small University appointment as a sub-librarian; and it 
 was fifteen years before the struggle was ended by his appoint- 
 ment as a Professor of Logic and Metaphysics. During this 
 period his name was gradually rising. At first he lectured on 
 the mathematical and physical sciences as well as on philosophy, 
 and his earlier publications related mainly to the former. But 
 gradually he confined his teaching to philosophy, to which he 
 subsequently added courses of popular lectures on physical 
 geography and anthropology. The former of these subjects 
 seems to have been introduced for the first time into the studies 
 of the University by Kant himself, who, though he never left 
 Konigsberg, was an eager reader of voyages and travels. His 
 vivid realisation of the geographical features and physical 
 characteristics of each country, as well as of the ways of life of 
 the people who inhabited it, made his lectures attractive to 
 many beyond the regular students of the University, among 
 whom may be mentioned specially the officers of the Eussian 
 garrison which held Konigsberg during the third Silesian war 
 (1756-63). 
 
 A word or two may be here introduced as to Kant's relation 
 to the political life of his time. Kant was no politician in a 
 practical sense : so far as we know he never had to perform 
 one directly political act. But he was a diligent student of 
 contemporary politics, and he watched with keen interest the 
 various phases of the political movement of the eighteenth 
 century. He was a witness of three of the acts of what we 
 may call the great modern political drama : he almost saw 
 with his own eyes the Seven Years' war, in which the great 
 Frederic stood at bay against the combined forces of the 
 Austrians, the French, and the Russians : he followed with the 
 warmest sympathy the phases of the distant conflict which 
 
"• LIFE AND WORK OP KANT. 59 
 
 ended in the assertion of the independence of the United States 
 of America : and, he lived long enough to see the French 
 Eevolution and the beginning of the career of Napoleon. The 
 first of those great events was brought home to him, as already 
 indicated, by the Kussian occupation of Konigsberg ; and we 
 cannot suppose that it had less influence upon him than it had 
 upon others of his nation. It was, indeed, the heroic struggle 
 of Frederic which first awakened modern Germany to a con- 
 sciousness of its powers. It made Germans begin to think of 
 themselves as a distinct people and to take pride in their 
 language. It roused the national genius from the long sleep 
 which had ■ held it inactive ever since the disaster of the 
 Thirty Years' war, and stirred it to that vigorous fermentation 
 out t)f which came a new national literature. Frederic himself, 
 indeed, was deeply imbued with French culture, and he thought 
 of the German language as a semi-barbarous dialect, which 
 would not be expected to produce literary fruit of any ex- 
 cellence. But even this, as Goethe tells us, was an additional 
 stimulus to German authors to prove their worth to him, whom 
 all regarded with admiration as the national hero. At any rate, 
 it is certain that after Frederic's victory a new spirit seemed 
 to pass into Germany, and that, for the first time since the 
 close of the Thirty Years' war, works of permanent literary 
 value began to be produced. First came Lessing, the scholar 
 and dramatist, the master of literary criticism, and the beginner 
 of philosophic theology: then Elopstock, the "German Milton," 
 author of an imposing, though somewhat pompous and rhetori- 
 cal epic and of many vigorous lyric poems ; then Winckel- 
 mann, the modern Greek, whose marvellous intuition of beauty 
 first revealed to the modern world the full meaning of Hellenic 
 art ; and Wieland, the German Frenchman, whose half romantic, 
 half classical novels have now almost lost their interest, but 
 who in his own time did no little to awaken the literary 
 faculty of his countrymen. And this first generation of authors 
 ^ was followed by a second of still greater power, among whom 
 
60 INTRODUCTION. chap. 
 
 we need only mention the names of Herder and Jacobi, of 
 Goethe and Schiller.^ 
 v^rmd^the™ The American war, which arose out of the attempt made in 
 volution. 1765 to enforce taxation in the American colonies, was an 
 event in which we know from many indications that Kant 
 took a lively interest. It is recorded that one day when in a 
 public garden he was maintaining earnestly the rightfulness of 
 the resistance of the colonies, an Englishman sprang up and 
 challenged Kant for insulting his country. Kant answered 
 firmly but with such calmness and persuasive force, that his 
 assailant speedily became paciiied and ended by shaking hands 
 with him. This Englishman was a merchant, called Green, 
 who settled in Konigsberg, and became one of Kant's closest 
 friends. Until Green's death Kant was wont to spend part of 
 every Saturday with him, and he is even said to have submitted 
 some of his works to Green's criticism. The French Eevolu- 
 tion, the last great political movement which took place in 
 Kant's time, came too late in his life to affect him greatly; but 
 he was deeply influenced by the works of Eousseau, in which 
 that Eevolution may be said to have been anticipated. An 
 era of political change is favourable to philosophy. It liberates 
 the mind from the yoke of custom, and encourages it to specu- 
 late freely upon the moral principles which underlie the social 
 order. And it was of no little importance for Kant's works 
 that he witnessed so much of the emancipating struggle of the 
 eighteenth century. For the Critical Philosophy is not a 
 product of the mere study of books : it is the work of one who 
 was alive to the spirit of the time, and who reproduced in his 
 thought the great movement for the liberation of humanity 
 which he saw going on without him. We may even say, 
 without much exaggeration, that in Kant's philosophy the 
 
 1 The following dates may be useful to mark the position of Kant with 
 reference to contemporary literature :— Kant, 1724-1804; Lessing 1729-81- 
 Winokelmann, 1717-68; Klopstock, 1724-1803; Wieland 1733-1815- 
 Herder, 1749-1803; Jacobi, 1743-1819; Goethe, 1749-1832- Schiller' 
 1759-1805. ' ' 
 
"• LIFE AND WOEK OF KANT. 61 
 
 reason or principle of that movement was first brought to 
 light. 
 
 Kant was, > as I have said, slow in finding his work. For, Kant's profes- 
 
 sorship and 
 
 though as early as 1763 he published several important essays, ^^)^f 
 — which indicate that he had already broken away from the 
 Wolffian dogmatism, and that he was seeking for light in 
 various directions, and especially from the English philosophy 
 of Locke and his followers, — it was not till the year 1 7 7 , not till 
 the forty-seventh year of his life, that we can trace in his works 
 the beginnings of the Critical Philosophy. In that year, how- 
 ever, in entering upon his duties as a Professor, Kant produced as 
 his inaugural discourse an essay on the Form and Principles of 
 the Sensible a-Ad the Intelligiile World, which contains (almosy all 
 the thoughts afterwards embodied in the Aesthetic, the first 
 part of the Critique of Pure Reason, and also some hints of the 
 ideas expressed in the other sections of that work. But the 
 Critique itself did not appear till eleven years afterwards. 
 After 1781, and for the next ten years, in a continuous flow 
 of literary production, Kant poured forth the treasures of 
 thought which he had so patiently accumulated, and worked 
 out the application of his principles to every department of 
 philosophy. In 1783 came his Prolegomena to Every Future 
 Metaphysic, a new exposition of the ideas of the Critique from 
 a somewhat different • point of view: in 1785 the Foundation 
 of the Metaphysic of Ethics, in which the moral aspects of his 
 philosophy were first exhibited: in 1786 the Metaphysical 
 Budiments of Physics: in 1788 the Critique of Practical Peason: 
 and in 1790 the Critique of Judgment. This last work, which 
 treats of the grounds of our aesthetic judgments, and also of 
 the application of the idea of final cause to nature and especially 
 to the organic world, completed the series of Critiques. During 
 the next ten years his only works of great importance were the 
 treatise on Beligion within the hounds of Pure Reason (1793), 
 and the works on the Doctrine of Right and the Doctrine of 
 J^irtue (1797), in which Kant's moral principles are worked 
 
62 INTRODtrCTIOK. chap 
 
 out in their applications to law, politics, and ethics. Besides 
 these larger works he published a number of essays on special 
 subjects in almost every branch of philosophy. The last books 
 published by himself were his essay on the Conflict of the 
 Famlties and his lectures on Anthropology (1798), though even 
 after that date his lectures on Logw, on Physical Geography, and 
 on Paedagogio were edited under his authority by certain of 
 his pupils. By this time Kant had become a permanent 
 invalid. In 1797 he ceased to lecture and his memory began 
 to fail him, and two years later he became so weak that he 
 expressed a longing for death. A young man called "Wasianski, 
 who resided with him and took charge of his affairs, has 
 recorded with almost too much detail his gradual decay in 
 mind and body. In these last days Kant still tried to work 
 at a book on the Metaphysic of Nature, but he could not 
 distinctly express his thought and fell into continual repeti- 
 tions and confusions. The slow process of death ended on the 
 12th of February, 1804, when he was close upon his eightieth 
 year. 
 His character. Kant was a man of feeble physique, hollow-chested and of 
 small stature ; but by adherence to strict hygienic rules which 
 he had laid down for himself, as well as by the utmost order 
 and regularity of life, he long preserved unbroken health and 
 was able to get through an almost incredible amount of work. 
 He was of a cheerful, unassuming disposition and extremely 
 modest in his personal claims. "While he deeply sympathised 
 in the aspirations of his time after greater social and political 
 freedom, his temper inclined him to avoid any thin o- like 
 rebellion against constituted authority : and the most dubious 
 act of which he was guilty was that, in submission to an 
 order from the government of the pietistic successor of the 
 great Prederic, he promised during that king's reign to be 
 absolutely silent on theological subjects. Kant met with no 
 great outward success, and indeed almost no practical 
 recognition of his great powers, till he had reached a com- 
 
II- LIFE AND WORK OF KANT. 63 
 
 paratively advanced age : and when in later life attempts were 
 made to draw him away from his native Konigsberg by the 
 offer of more lucrative employment elsewhere, he refused 
 to avail himself of them. He had arranged and regulated his 
 life, had formed his habits of living and working, and he would 
 not let himself be disturbed by change. His long struggle with 
 fortune taught him to be severely economical, but he never 
 showed a trace of avarice ; on the contrary he was markedly 
 generous to those who had claims upon his help. In his 
 earlier life he was fond of company, and especially of the 
 company of refined and educated women. At a later period 
 he ceased to go into society, but he almost invariably enter- 
 tained one or two friends at his table. Otherwise he sought 
 no relaxation from the never-hasting, never-resting work, 
 which he pursued day after day and year after year, without 
 turning aside from the fixed rules of life which he had made 
 for himself. "With no lack of kindliness or social interest, he 
 was one whose life always reminds us of the " categorical 
 imperative of duty," which was for him the kernel of morals. 
 Of no one can it be said more truly, that he purchased inner 
 freedom by strict obedience to law and even to every limit of 
 convention or authority, which he could recognise as reasonable. 
 Heine draws a laughable contrast, in which there is some germ 
 of truth, between the quiet and sober tenor of his life as 
 an exact University official and a respectable citizen, and the 
 world-moving power of his writings. 
 
 " The life of Immanuel Kant is hard to describe : he had geine on 
 
 Kant. 
 
 indeed neither life nor history in the proper sense of the words. 
 He lived an abstract, mechanical, old-bachelor existence in a quiet, 
 remote street of Konigsberg, an old city at the north-eastern 
 boundary of Germany. I do not believe that the great cathedral 
 clock of that city accomplished its day's work in a less passionate 
 and more regular way than its countryman, Immanuel Kant. 
 Eising from bed, coffee-drinking, writing, lecturing, eating, 
 i ♦ walking, everything had its fixed time ; and the neighbours 
 
 u 
 
64 INTBODTJCTION". CHAP 
 
 knew that it must be exactly half -past four when they saw 
 Professor Kant in his grey coat with his cane in his hand step 
 out of his house door, and move towards the little lime tree 
 avenue, which is called after him the Philosopher's Walk. 
 Eight times he walked up and down that walk at every season 
 of the year, and when the weather was bad or the grey clouds 
 threatened rain, his servant, old Lampe, was seen anxiously 
 following him with a large umbrella under his arm, like an 
 image of Providence." 
 
 " Strange contrast between the outer life of the man and his 
 world destroying thought. Of a truth, if the citizens of 
 Konigsberg had had any inkling of the meaning of that 
 thought, they would have shuddered before him, as before an 
 executioner. But the good people saw nothing in him but a 
 professor of philosophy, and when he passed at the appointed 
 hour, they gave him friendly greetings and set their watches." 
 His method in In his earlier days he seems to have been a very effective 
 
 teacliing. ^ 
 
 lecturer, as is evidenced by the enthusiastic words of Herder : 
 but as the life of thought absorbed him more and more, the 
 interest of personal address became less keen, and his prelec- 
 tions lost something of their living interest. As to his 
 methods of teaching, Jachmann gives us a hint which is not 
 without value in relation to the interpretation of his writings. 
 In lecturing, Kant, he says, was wont " as it were to conduct 
 an intellectual experiment before his audience, as if he were 
 himself beginning to meditate on the subject. First, he set up 
 a rough definition of the subject to be discussed, then by 
 degrees he introduced new conceptions to modify it : step by 
 step the explanations which had been tentatively presented, 
 were corrected, until at last the finishing touch was given 
 to the definition, which had been elucidated in every point of 
 view. In this way an attentive listener not only was made 
 acquainted with the subject, but also received a lesson in 
 systematic thinking. But a hearer who was ignorant of the 
 method of the teacher, and who took his first explanation 
 
"■ LIFE AND WORK OF KANT. 65 
 
 of the subject for a final and exhaustive statement, to the 
 neglect of the subsequent steps, was likely to carry away only 
 half truths." This method was used by Kant not only in his 
 lectures but also to some extent in his books, and it has given 
 rise to the same misunderstanding in the latter as in the 
 former case. It was, indeed, almost a necessity that one who 
 had to open up a region of thought so new and unfamiliar as 
 that of the Critical Philosophy, should begin by using the 
 ordinary conceptions of his time, and should gradually trans- 
 form them by explanation upon explanation till he lifted his 
 readers to his own point of view. The method is one which 
 makes Kant's works very instructive to any one who will 
 patiently follow him in every new step, but almost incompre- 
 hensible to those who expect their author to do their thinking 
 for them, and to present them always with cut and dry results. 
 Like Socrates, Kant forces his pupils to co-operate with him, 
 and the conclusions to which he brings them are of almost no 
 value apart from the process. Nay they may become even 
 misleading ; for Kant, as the discoverer of a new principle and 
 a new method of speculation, was not, and could not be 
 expected to be, aware of the full power and bearing of his own 
 thought. The scaffolding obscures the edifice even from its 
 builder, and of Kant even more truly than of any other 
 philosopher it may be said that to understand him is to go ^ 
 beyond him. On the other hand there is at least this amount 
 of reason in the call to " return to Kant," that there is nothing 
 which does so much to enable us to understand any principle 
 or way of thinking as the consideration of the conditions of its 
 first expression. 
 
 In the beginning of this chapter attention has been drawn Three periods in 
 
 ° ° ^ Kant's Mental 
 
 to the general characteristics of the age of Enlightenment in development. 
 which Kant's powers ripened to maturity ; it only now remains 
 to bring what is there said into more definite relation with 
 Kant's own speculative development up to the period when the 
 Critical idea began to take form in his mind. In that develop- 
 
 VOL. I. E 
 
66 INTRODUCTION. °^^^- 
 
 ment we may distinguish three periods. In the first of these 
 Kant is still in the main an adherent, though a somewhat rest- 
 less and dissatisfied adherent, of the Wolffian philosophy, or 
 rather of a modification of that philosophy which had resulted 
 from the speculative movement within and without the school 
 of Wolff. Of this stage of his thought the main record is 
 found in the Mw Exposition of the First Principles of Meta- 
 physic with which in 1755 he "habilitated," or qualified as a 
 Frivat-Docent, in the University of Konigsberg. In 1763 this 
 stage ends with the publication of three essays, which may be 
 regarded as Kant's declaration of independence. In these 
 essays he breaks with the Wolffian philosophy, and shows a 
 tendency, which is even more strongly marked three years later 
 in the treatise on the Breams of a Ghost-Seer illustrated hy the 
 Ih-eains of Metaphysic, to adopt the principles of the Empiricism 
 of Locke. Finally, about the year 1768-9, there is evidence 
 of a second recoil from Empiricism towards Eationalism, and 
 the commencement of an effort to reach a higher point of view 
 from which the opposition between it and Empiricism may be 
 reconciled, an effort the first fruit of which was the Disserta- 
 tion on the Form and Principles of the Sensible and the Intelligible 
 World. That treatise, as above stated, was published in 1770, 
 but it took eleven years of what we may call intellectual fer- 
 mentation before the process of synthesis thus initiated was 
 brought to a definite result in the first edition of the Critique 
 of Pure Reason. And even the Critique of Pure Reason must 
 itself be regarded as presenting to us only one side or aspect 
 of a comprehensive thought which was subsequently worked 
 out in the other Critiques, and that not without reactive influ- 
 ence on the first Critique.'^ 
 
 ' In relation to the first two stages our main authority must be the published 
 works of Kant, though some additional illustration is supplied by the Reflex- 
 ionen Kant's, published by Dr. Benno Erdmann. For the third stage, the 
 main authorities are the Dissertation and Kant's letters to Lambert and Herz. 
 But valuable sidelights are supplied by the Reflexionen (with Dr. Erdmann's 
 introductions) and by Politz's editions of Kant's lectures on Metaphysic and 
 
"• LIFE AND WORK OF KANT. 67 
 
 In the three following chapters, I propose lirst to give an t^^foUow'iM 
 account of Kant's precursors, or in other words, to describe '''''^'''™" 
 the earlier modern philosophical movement as far as is neces- 
 sary to exhibit Kant's relation to it, and especially to show at 
 what point he took up the philosophical problem ; in the next 
 place, to follow Kant through the first two stages of his de- 
 velopment, or, in other words, through the speculations of the 
 precritical period ; finally, in the last chapter of this Intro- 
 duction, to trace the steps in the development of the idea, of 
 criticism from its first dawn to the publication of the Critique 
 of Pure Reason. 
 
 Natural Theology and Dr. Erdmann's articles in the Philosophische Monatshefte 
 (vols. 19 and 20). Cf. also Cohen, Die systematische Begriffe in Kant's Vor- 
 kritischen Schri/ten, and Paulsen, Die EntwicJcelungsgescMchte der KantiacMu 
 Erkenntnisslehre. 
 
68 
 
 CHAPTEK III. 
 
 Kent's peecuesors, descaetes, spinoza, leibniz, and wolff — 
 
 THE LATEE wolffian SCHOOL AND MARTIN KNUTZEN. 
 
 EvoSn'as a T^HE idea of evolution is now so familiar, and it has lent 
 h£tory*o/ such a living interest to the history of the past, that it 
 
 is not easy to realise the point of view of those who were 
 without that idea. Especially is this the case in relation to 
 philosophy. To us the history of philosophy has become a 
 part of philosophy itself, because we have learned to look on 
 the speculations of earlier times, not as dogmatic systems to be 
 accepted or rejected, but rather as the first stages in the pro- 
 gressive evolution of a thought of which, in a further stage, we 
 ourselves are the organs and interpreters. Hence follow two 
 important consequences. On the one hand, we are freed to 
 some extent from historical partisanship, since we do not 
 expect to find direct support for our own ideas in any past 
 system : yet on the other hand, we are enabled to feel a living 
 interest in all such systems, as containing aspects or elements 
 of the truth which we seek to discover. We are pledged to 
 show that the system which we regard as true, is the result of 
 a synthesis in which those aspects or elements are combined. 
 But to those who looked upon the history of philosophical 
 opinion without the guiding light of the idea of evolution, that 
 history necessarily took the aspect of a series of unrelated 
 facts the knowledge of which could have no scientific value. 
 
CHAP. III. KANT's precursors from DESCARTES TO WOLFF. 6 9 
 
 They could feel a living interest only in the opinions and 
 reasonings of those writers who were near in time to them- 
 selves, and who, therefore, put the problem of philosophy in the 
 same way. What lay further off they were obliged to distort 
 into an artificial resemblance to their own ideas ; otherwise it 
 had no philosophical meaning for them. Hence, with few ex- 
 ceptions, philosophic writers were wont to pay little heed to 
 any speculations save those of their immediate predecessors, 
 and the doctrines of earlier writers were remembered, if at all, 
 only as a dead tradition, a cipher to which the key was lost. 
 
 This indifference to the past was specially characteristic of Hii^t^of"""^ 
 the eighteenth century, partly because of its violent recoil from 1^"°*=*™°- 
 the ideas of the previous period, partly because of the character 
 of the ideas in which that recoil landed it. The individualistic 
 tendencies of the age of Enlightenment, which separated each 
 man from the unity of the social organism to which he belonged, 
 separated him also from the past out of which his intellectual 
 life had grown. Hence to the writers of that time the inde- 
 pendence of philosophical thought seemed to involve that each 
 thinker must begin the work of speculation de now : and to 
 admit the possibility or necessity of a mediation of truth to 
 the individual by the communis sensus of humanity was in 
 their eyes the same thing as to accept the dictation of an 
 external authority. In this respect Kant shared in the indi- 
 vidualistic and unhistorical modes of thought characteristic of 
 his time, though it may be also said that it was a result of his 
 work to change them, or at least to prepare the way for a 
 deeper conception of the relation of the individual to humanity. 
 No more than his contemporaries had he a vital hold of the 
 remoter past. Even of Leibniz, the great thinker to whom he 
 stood nearest, he can scarcely be said to have had any direct 
 vision : he sees him almost entirely through the spectacles of 
 Wolff: and if he ever gains a deeper apprehension, it is only as 
 the advance of his own thought gTadually leads to a revival of 
 ^ome of the elements of the Leibnizian doctrine. Spinoza's 
 
70 ■ INTRODTJCTION. CHAP. 
 
 thought, as he confessed, remained an enigma to him, even 
 after Jacobi had paved the way to a better knowledge of it. 
 And his references to Plato and Aristotle seldom go much 
 beyond the limits of the ordinary inaccurate tradition, which 
 makes them respectively types of the a priori, and the em- 
 pirical, modes of thought : though he is conscious of a certain 
 affinity to Plato, especially in relation to his moral ideas. The 
 historical relations of Kant's philosophy, as described by him- 
 self, are thus limited to the Wolffio-Leibnizian system on the 
 one side, and the philosophy of Locke, Newton, and Hume on 
 the other. The only other great writer to whom he confesses 
 an obligation is Eousseau, in whose ideas of a " common reason " 
 and a "general will," we may discern some anticipation of the 
 intellectual revolution which Kant was to carry out. 
 Historical Yet it would be a mistake entirely to limit our view of the 
 
 relations or *' 
 
 ^ptyf ™'°" historical setting of Kant's philosophy to the relations of which 
 Kant himself was conscious. Kant's Critique was the beginning 
 of an intellectual revolution by which a synthesis was made, 
 not only between the different forms of Individualism which 
 arose in Germany and England, but also between the Individual- 
 ism of the eighteenth century and what may be called the 
 abstract Universalism of an earlier period. It is no mere 
 accident, but an indication of the same spirit of the time which 
 was manifested in the Kantian philosophy itself, that an inter- 
 est in Spinoza revived shortly after the publication of the 
 Critique of Pure Beason. The way in which Spinoza's Ethics 
 took hold of Goethe, who was above all the representative of 
 the modern spirit, and of its characteristic ideas of organic unity 
 and evolution, is sufficient to show this. And the fact that 
 Kant did not recognise any such affinity, is explained partly by 
 the manner of his own mental development, which made him 
 regard Spinoza merely as a representative of the Dogmatism 
 against which he had to contend. Yet it is impossible to con- 
 sider that development carefully without seeing that Kant is 
 deeply influenced by the Spinozistic idea of a universal unity. 
 
HI- KANt's precursors from DESCARTES TO WOLFF. 71 
 
 a unity of knowing and being, which through Leibniz passed 
 as a tradition into the Wolffian philosophy, and which saved 
 that philosophy from the Individualism, or even Atomism, to 
 which it tended. A few words on this subject may enable us 
 better to appreciate the modified Wolffianism of Kant's first 
 period. 
 
 The history of modern thought begins with the declaration ^^^"Sorml. 
 of the spiritual independence of the individual, and the rejection ''™" 
 of the principle of authority. The Eeformation was in principle 
 the negation of the claim of any doctrine to be accepted by the 
 individual, which could not find its evidence in the movement 
 of his own reason ; of any law to be obeyed by him, which could 
 , not be shown to spring from his own will. It was a return of 
 man's spirit to itself and a rejection of all that is merely exter- 
 nal and foreign. A truth which cannot become a conviction, 
 one with my very consciousness of self, is no truth for me. A 
 power, which directs my activity to an end which I cannot 
 regard as my own good, is an external tyranny, to which I may 
 be forced to submit but which I cannot regard as a legitimate 
 authority. Before this assertion of the principle of freedom, of 
 self-determining reason and self-legislative wiU, the dogmatic 
 system of thought and the external system of government, which 
 were characteristic of the middle ages, crumbled into dust. The 
 dualistic ideas of a revelation coming entirely from without, of 
 a transformation of man's life and nature by a foreign power, of 
 an expulsion of nature that spirit might reign in its place, were 
 once for all exploded. Old things had passed away : the years 
 of man's nonage were completed, and he was no longer under 
 tutors and governors, but master, of himself and of his fate. 
 
 No one who looks beneath the surface of modern history can its power over 
 
 thought and' 
 
 doubt the power of this idea. We all, as Hegel once said, '"^■ 
 fight " under the banner of the free spirit." Again and again, in 
 religion, in science, in politics and in social life, it has stirred 
 the modern world to convulsive and even fanatical efforts to rid 
 « itself of the weight of the past that was oppressing it. Before 
 
72 INTRODUCTION. 
 
 CHAP. 
 
 this principle the strongest systems of political and ecclesiastical 
 order have gone down in utter ruin at the hour when they 
 seemed most impregnable. It has been the secret of that dis- 
 ease of subjectivity which has infected the modern world, and 
 has filled literature with the voice of its discontent. And on 
 the other hand, it has been the essence of that " freshness of 
 the earlier world " which has purified and transformed modem 
 poetry and art and renewed the exhausted springs of emotion. 
 Above all, it has given rise to that effort to simplify life and 
 thought, to get rid of the meaningless and the unreal, to come 
 face to face with nature and man, and to banish as much 
 as possible the merely conventional and unintelligible, — in 
 short, to extirpate everything incapable of being brought into 
 living relation with self-consciousness, — which is the distinguish- 
 ing characteristic of the modern spirit. The whole history of 
 modern philosophy may be regarded as an endeavour to work 
 out the results of the same principles. Kant's protest against 
 the admission of the " Transcendent," of that which is incapable 
 of being brought in relation to human experience, is but a further 
 step in the same process which was initiated by Descartes, 
 when he sought to base all truth upon the Cogito ergo sum of 
 self-consciousness. 
 Meaning of From this point of view we may recognise a bond of kinship 
 
 the revolt 
 
 against bctwecn the Eeformers and those who originated the modem 
 
 Scholastic *= 
 
 Philosophy, scientific movement. The explanation of what is within the 
 range of the human consciousness by what is supposed to be 
 utterly beyond it, is the one thing which they are agreed in 
 rejecting. Luther condemns an authoritative church as stand- 
 ing between the soul and God in the same spirit in which 
 Bacon afterwards condemned the abstractions of Scholastic 
 philosophy as standing between the human mind and nature. 
 Both demand the closest contact of subject and object, and both 
 are equally convinced that by means of this contact, subject and 
 object may be brought into unity without any need for a 
 mediator. In both cases, indeed, a process is held to be 
 
in.. KANT S PRECURSORS FROM DESCARTES TO WOLFF. 73 
 
 necessary ; a- process in the one case, whereby the natural man, 
 to whom the life of Christ is an external fact is to be converted 
 into the spiritual man to whom the belief in Christ is one with 
 the consciousness of himself; and a process in the other case 
 whereby the student of nature is to free his mind from all pre- 
 suppositions and prejudices, and turn it, in Bacon's language, 
 into a pure mirror of the world. But, in the one case as in the 
 other, the process is one which is to be carried on from beginning 
 to end in and through the consciousness of the individual him- 
 self, without any dependence on an external authority. The 
 circle is to be completed within the range of human perception 
 and human thought, and to rest upon nothing beyond it. The 
 living experience is to be its own sufficient evidence. 
 
 In this way the independence of the individual mind was subjective 
 
 ^ J freedom and 
 
 asserted without shutting it out from objective truth. It was ?ruth'"^ 
 admitted that knowledge is possible only through the surrender 
 of the intelligence of the individual to a truth which in the 
 first instance seems to be quite independent of it. " Into the 
 kingdom of man which is based on science, as into the king- 
 dom of Heaven, we can enter only sm& persona infantis." The 
 first step towards the knowledge of the world was to give up 
 all preconceived opinions and subjective idols which stand be- 
 tween us and the object. The *first step towards the knowledge 
 of God was to empty the mind of self. But, on the other 
 hand, this simplicity of reception was not in either case supposed 
 to imply a mere passivity of spirit. On the contrary, our first 
 impressions of things were viewed as furthest from the truth, 
 and an acquiescence in immediate appearance was denounced, 
 both by Luther and Bacon, as the great hindrance to true 
 religion and true science. In both cases an active transforming 
 process was required, through which the mind must pass ere it 
 could reach the truth and identify itself therewith, so as to find 
 the evidence for that truth in itself. By this process of in- 
 tellectual or spiritual activity the object ceased to be, — what in 
 its first apprehension it necessarily was, — a merely external 
 
Empiricism. 
 
 74 INTRODUCTION. chap. 
 
 object, a mere fact unrelated to the intelligence : it was seen in 
 its principle or law, and so became one with the mind which 
 apprehended it. Hence what in one point of view was a change 
 of the object was, in another point of view, a change of the 
 subject. To sacrifice the mere subjective presuppositions or 
 idols which stand between the intelligence and nature, was at 
 the same time to break through external appearances and to 
 discover the essential relation of nature to the intelligence. 
 The surrender of reason to the teaching of Christ was in 
 another point of view the discovery of reason in Christianity. 
 o^Miion'* This conception of the unity of thought with its object — ^the 
 Bacon^s'''"^ solc conception which makes it possible to reconcile the possi- 
 bility of objective knowledge of God or of the world with the 
 subjective principle of freedom — was involved in Luther's idea of 
 Eeligion, and it was involved also in Bacon's Empiricism, but it 
 was not distinctly recognised by either of them. Bacon, while 
 he describes the method of physical inquiry and lays down the 
 principles by which it is to be guided, has no idea of connecting 
 that method or those principles with the nature of self-con- 
 sciousness. He generally speaks as if facts were given through 
 sense without any aid from reason, and as if any contribution 
 of reason to the data of sense must necessarily involve a dis- 
 tortion of those data. If he admits the necessity of the activity 
 of the intelligence in the apprehension of truth, it is in his 
 conception a merely formal activity. The image of the mind as 
 a mirror of the world seemed to him a sufficient account of its 
 relation to its object : nor did it ever occur to him to attribute 
 any constitutive power to thought. And Luther, while he 
 allowed that the truth must be spiritually discerned, thinks of 
 such discernment as the result of a divine influence, under 
 which the soul is purely passive. It is, therefore, fair to say of 
 both that, while they virtually asserted the unity of the mind 
 with the object which it apprehends, they were content to feel 
 that unity, without attempting to understand it. The implicit 
 reason of the mind's assent to truth was not by them made ex- 
 
III. KANT's PRECURSORS TROM DESCARTES TO WOLFF. 75 
 
 plicit. How the individual, as such, can transcend his individu- 
 ality ;- how his assurance of that which is not himself can be 
 vindicated ; how it is possible that a finite being, himself a part 
 of the partial world, can in thought overpass the limits of his own 
 subjective consciousness and become a " spectator of all time 
 and existence," conscious of the world and of God ; these were 
 questions which they did not ask. Ultimately both Bacon and 
 Luther rested on an immediate certitude of feeling, or faith. As 
 the one felt he was dealing with reality when he was dealing 
 with the external world, so the other felt he was dealing with 
 reality when he was dealing with God: and the processes which 
 were necessary to develop this knowledge, or to bring the object, 
 which at first appears as external, into unity with the subject, 
 were justified merely by the result. As_a-jnatter of fact, nature 
 ceasestobea merely external existence for us when we dis- 
 c over its law s. As a matter of fact, God ceaseTto be a mere 
 name for the absolute Being, when we receive into our minds 
 the Christian idea of Him. But still, even after this process, 
 self, the world, and God have the aspect of three elements, which 
 we find together in our minds, but which are connected by no 
 necessary relations, or at least by relations which are felt only 
 and not comprehended. Hence the immediate and unreflecting 
 consciousness in all its forms is exposed to the assaults of doubt, 
 — a doubt which may assail even the very existence of its ob- 
 jects. The consciousness of self may still be turned against the 
 consciousness of an external world, or the consciousness of an 
 external world against the consciousness of self, so long as they 
 are not seen to be necessary to each other. Or again the con- 
 sciousness of the finite may be turned against the consciousness 
 of the infinite, and either may be used to suppress the other, so 
 long as they are not seen to be at once distinct and necessarily 
 connected. 
 
 So soon as such doubts arise, the immediate consciousness of Dangers of 
 
 reflexion and 
 
 reality ceases to be sufiicient for itself, and philosophy becomes analysis. 
 a necessity. For that consciousness necessarily takes its objects 
 
76 INTRODUCTION. °^^^- 
 
 for granted. To the religious man God, as to the scientific man 
 the external world, are realities immediately bound up with his 
 consciousness of self, and he has no need to seek for a bridge 
 from the one to the other. But reflexion breaks up this im- 
 mediate unity, and forces philosophy to undertake the task of 
 showing that the different elements of consciousness are con- 
 nected in one system of belief, and that it is impossible to 
 admit one of these without being driven to admit them all. 
 The business of philosophy is thus to cut away the ground from 
 scepticism by exhibiting the reciprocal implication of all the 
 principles on which the world, as an intelligible world, must 
 rest. For scepticism, as was suggested in the first chapter of 
 this Introduction, is never complete. Its strength always lies 
 in turning some element of truth, the certitude of which is ■ 
 assumed, against tha other elements. And the answer to it can 
 lie only in showing the necessary relation of the element of 
 truth which is thus assumed to the other elements of the system 
 of knowledge. Thus philosophy is a kind of reasoning in a circle, 
 but that is no argument against it ; for it is a circle beyond 
 which nothing lies, and in particular it is a circle which 
 includes the position of the sceptic himself. 
 Theaubjective xhe form wMch sccpticism has generally taken in modern 
 
 principle of 1.0./ 
 
 Sd'D^art™ timcs is determined by the subjective principle of Protestantism. 
 The consciousness of self has been turned against the con- 
 sciousness of the external world, or against the consciousness of 
 God, or even against both at once. The question, therefore, 
 has been how we are to know anything but ourselves and our 
 own ideas ; how our consciousness is to go out to other finite 
 objects or to rise to the infinite. This point of view is clearly 
 indicated by Descartes when he finds in the Cogito ergo sum the 
 primal and invincible certitude, to which doubt can reduce us, 
 but of which it can never deprive us. The external world is 
 to Descartes a world not only extended and external to itself 
 but also external to our consciousness. On it therefore, con- 
 sciousness, which has for its immediate object only its own 
 
Ill- KANT S PRECTJRSORS FROM DESCARTES TO WOLFF. 77 
 
 ideas, has no direct hold. Hence in spite of the clearness and 
 distinctness of our ideas as to the relations of the extended,— a 
 clearness and distinctness which gives rise to the certitude of 
 mathematical science, — ^our belief in the existence of the 
 extended is unable to resist the shock of doubt. As Descartes 
 naively expresses it, we can imagine that some untruthful 
 spirit is playing upon our minds and awaking in them the 
 ideas of things that have no reality. Thus we are driven back 
 upon the ideas which are the immediate objects of our con- 
 sciousness, and upon the conscious self for which they are objects. 
 In relation to these, it is supposed, no lying spirit can deceive 
 us. The conscious self asserts its existence even in the very 
 act by which it abstracts from everything else. Its self- 
 affirmation is involved in the very possibility of doubt, and 
 cannot itself be an object of doubt. What we can doubt is 
 only what we can separate from the self-affirming ego, which is 
 as it asserts itself and asserts itself as it is. But the end of 
 doubt is the beginning of knowledge, and in the Oogito ergo sum 
 we have a first unity of thought and being, out of which a 
 complete reconciliation of them may spring. 
 
 Self-consciousness, then, is the starting point or principium Descartes' 
 cognosceTidi, from which we must start and on which we must infi™i?e"'*° *" 
 base all other forms of consciousness. As, however, Descartes 
 begins by separating the consciousness of self from the im- 
 mediate consciousness of the external world, it is obvious that 
 he cannot bring them together again except through some 
 tertium quid. And of this tertiu^n quid, which is to form the 
 connecting link between the consciousness of self and the con- 
 sciousness of the world, there must be a certitude as immediate 
 as that of the consciousness of self. Descartes, therefore, is 
 necessarily driven to the assertion that the consciousness of 
 God is directly involved in the consciousness of self and indeed 
 is prior to it. "Let us not imagine,'' he declares in one 
 remarkable passage, " that the conception of the infinite is got 
 , merely by negation of the finite, just as we conceive rest to be 
 
78 INTRODUCTION. chap. 
 
 the negation of movement and darkness to be the negation of 
 light. On the contrary we obviously think of the infinite 
 substance as having more reality in it than the finite substance: 
 nay, it may even be said that our consciousness of the infinite 
 is in some sense prior to our consciousness of the finite or, in 
 other words, that our consciousness of God is prior to our con- 
 sciousness of self. For, how could we doubt or desire, how 
 could we be conscious that anything is wanting to us, and that 
 we are not altogether perfect, if we had not in ourselves the 
 idea of a perfect being in comparison with whom we recognise 
 the defects of our own nature ? " In this passage the method 
 of abstraction which Descartes followed, when he separated the 
 consciousness of self from the consciousness of the object, is 
 carried a step further. He now, in effect, bids us abstract from 
 the distinction of the ego from the non-ego, on the ground that 
 that distinction is a mere limit or negation ; and he maintains 
 that the consciousness of the absolute unity which we thus 
 reach is prior to, and presupposed in, the consciousness of any 
 finite object, even the self. In apprehending this absolute 
 unity, therefore, we are lifted beyond the distinction of self and 
 not-self, of the thinking and the extended substance, and 
 through it we are enabled to connect the latter with the former. 
 Thus the difference, which was absolute from the point of view 
 of the finite, disappears altogether from the point of view of the 
 infinite. The effect of this method of reasoning was, it must 
 be acknowledged, very imperfectly seen by Descartes himself, 
 who saves himself from self-contradiction by many loose and 
 popular ways of expression. Thus failing to recognise that, 
 according to his own statement, the opposition of thought and 
 being disappears in God, who as an object of consciousness is 
 prior even to the self, he proceeds to speak of the thought of 
 God as only one of the ideas of the self, which requires an 
 adequate reality as its cause. And instead of saying that in 
 the consciousness of God, who is the absolute unity, the division 
 between the self and the not-self disappears, he tells us that 
 
III. KANT'S PRECURSORS FROM DESCARTES TO WOLFF. 79 
 
 the truthfulness of God is our pledge for the objective reality 
 of our " clear and distinct " ideas of that extended substance 
 which is the object of our perceptions. But here as often, the 
 disciples betray the secret of the master. When Malebranche 
 spoke of " seeing all things in God," he was only giving a more 
 vivid expression to the idea that the consciousness of God is 
 the connecting link between the consciousness of self and the 
 consciousness of the world. And Spinoza was only following 
 out the same thought to its necessary result, when he put the 
 idea of God at the head of his system as at once the principium 
 essendi and the privcipium cognoscendi of all that is and is 
 known. (^^^ 
 
 The philosophies of Descartes and Spinoza, in one point of Relation ot 
 
 '- ^ i. ' i Spmozato 
 
 view, are polar opposites, but they are only the opposite poles of Descartes, 
 the same thought. They exhibit, in the language of pure 
 thought, the same dialectical movement by which Luther, the 
 apostle of the rights of the individual conscience, became also 
 the asserter of the absolute passivity of man in relation to 
 God. The necessity of resisting all merely external authority 
 was forced on the individual by the consciousness of a divine 
 voice speaking within him, which he could not disobey : he could 
 only gain sufficient power to resist the world by regarding 
 himself as the mere organ of the divine will. His strength 
 against men was the counterpart of his absolute surrender of 
 himself to God. In philosophy this connexion of ideas 
 reappears in a slightly different form. The individual has 
 become conscious of himself as an independent thinking being, 
 and refuses to accept anything as true for him which is not 
 mediated' by his own thought : but this very return upon him- 
 self — this refusal to admit any belief that comes to him merely 
 from without — is possible for him only because he is not a 
 merely individual being who stands in an external relation to 
 other individuals, but one who has the universal principle of 
 knowledge bound up with his consciousness of himself. There 
 t would be no possible escape from the limitations of self- 
 
80 INTRODUCTION. ''^^^■ 
 
 consciousness, if it were not in some sense true, as Descartes 
 puts it, that the consciousness of God is prior to the conscious- 
 ness of self, i.e., that the unity, which all consciousness pre- 
 supposes, is a principle of unity for all things and not for the 
 inner life of the individual alone. It is in virtue of this that 
 the individual is able to abstract not only from external objects 
 but from himself also as a mere individual subject : as, on the 
 other hand, it is in virtue of this that he is able to know these 
 external objects in relation to himself and himself in relation 
 to them. Hence, we cannot know ourselves apart from, or 
 prior to, other things : nor is it possible that we should be con- 
 scious of our own ideas as our own ideas prior to, and indepen- 
 dent of, other objects. So far Spinoza was right' in holding 
 that the true starting point of thought is not in the consciousness 
 of self as a priiicipium cognoscendi separate from God who is the 
 principmm esscndi. It is only as the " spectators of all time 
 and existence " that we can know anything even of our own 
 existence : and the beginning of knowledge cannot be the 
 separate consciousness of our own inner being, which has no 
 meaning except in relation to an outer being from which we 
 distinguish it. We do not know ourselves first and the world 
 through ourselves ; but we know ourselves only in relation to, 
 and in distinction from, the world : nnd we know both through 
 their relation to the one principle of unity which underlies all 
 knowledge, 
 negation of Spiuoza, therefore, is not without justification in his substi- 
 the Finite. ^.^j-Jq^ gf ^ metaphysical for a psychological principle as the 
 basis of philosophy. At the same time it has to be acknow- 
 ledged that the process by which this substitution is made, is 
 defective in its logic, and that its result is, therefore, to sup- 
 press, rather than to solve, the problem. Por, as will be re- 
 membered, the problem was to find a synthetic principle by 
 aid of which the consciousness of self should be connected with 
 the consciousness of objects which were supposed to be external 
 to the self. And the solution was found in a unity which was 
 
in. KANT S PRECURSORS FROM DESCARTES TO WOLFF. 81 
 
 reached by abstraction from all differences, even the difference 
 of self and not-self. The principle- already laid down by Des- 
 cartes, but seen in its full bearing only by Spinoza, that omnis 
 determinatio est negatio, i.e., that all definition and determination 
 is the limitation of a presupposed positive being by a negation 
 or an unreality, led directly to the conclusion that the only 
 thing real in the proper sense of the word is that being which 
 is absolutely indeterminate, without distinction or limit ; and 
 that the only truth which is unmixed with illusion is the 
 thought which apprehends such being. Abstraction then 
 becomes the sole method of knowledge, and all distinctions, 
 including the distinction between self and not- self, between 
 thought and extension, are transcended only because they are 
 all set aside. Spinoza struggles against this result, and, indeed, 
 we may say that he has an intuitive perception of the way to 
 escape it, when he converts, the idea of a substance or purely 
 indeterminate being into the idea of a causa sui or self-deter- 
 mined subject " with an infinite number of attributes." But 
 he never saw the distinction between these two conceptions of 
 the ultimate unity, which in the subsequent development of 
 his system are treated as interchangeable. Thus thought and 
 extension appear as absolutely unrelated expressions of the one 
 substance, which is simply identical in both its expressions 
 (though at the same time thought is conceived as a conscious- 
 ness of extension as well as of itself). The result is that the 
 principle of unity, placed at the beginning of the Ethics, does 
 nothing to explain the difference subsequently introduced. 
 Being and knowing, or extension and thought, are set side by 
 side, and the reality of their difference is denied, but nothing 
 is done to mediate between them. Even in the definition of 
 substance as id quod in se est et per se concipitur there is no 
 connexion between the two members. The abstract assertion 
 of the unity of all things in God does nothing to break down 
 the absolute dualism of the world. To see things sub specie 
 q^ternitatis is simply to forget a difference which is found to 
 
 VOL. I. r 
 
82 INTRODUCTION. chap. 
 
 be as hard and insoluble as ever, when we return to it again. 
 It is like a Sunday confession that the things of this world are 
 naught, while we treat them as absolute realities all the other 
 days of the week. 
 theTnitein the '^^^ truth of the philosophy of Descartes and Spinoza lay in 
 oentur™' their assertion that the unity of the consciousness of self and 
 of the world must be mediated by the consciousness of God, or 
 in other words, that it is only in so far as the unity of the 
 consciousness of self is a principle which is presupposed in 
 " all objects of all thought," that knowledge is possible. The 
 error of that philosophy was, that this principle was reached 
 by abstraction and conceived in an abstract way " as substance 
 not as subject," as a mere common element or logical genus, 
 and not as a principle of unity in difference. But such a 
 direct effort to suppress difference only makes the difference 
 develop into contradiction. For a common element is no 
 bond of union, and the abstraction that separates it from the 
 elements that are not common only causes the latter to fall 
 asunder in hopeless opposition. Hence the result of this first 
 effort at an immediate reconciliation between the object and 
 the subject, through consciousness of the unity which is above 
 their difference, was only to discredit that consciousness. The 
 idea of God or the infinite, which in Spinoza had all but sup- 
 pressed the consciousness of the finite world and of the finite 
 self, begins with those who follow him to empty itself of its 
 contents ; and what is lost to the infinite is gained by the 
 finite. Nominalism takes the place of Eealism ; the individual 
 becomes everything, the unity in which all individuals are held 
 as parts of one whole becomes nothing. The universe is repre- 
 sented as a collection of isolated beings or things with no vital 
 connexion ; for the God who holds them together is conceived 
 as a mere external creator and governor, and rapidly sinks 
 into an unknown and unknowable Mre Supreme, which it 
 matters little to assert or to deny. Hence also the answer to 
 the problem of knowledge is sought in a different way. The 
 
III. rant's precursors from DESCARTES TO WOLFF. 83 
 
 psychological point of view gains at the expense of the meta- 
 physical, and instead of " seeing all things in God," it now 
 becomes received almost as an axiom that we must see all 
 things, — and God also, if we can see Him at all, — ^in ourselves. 
 Whether it is through the sensations which outward objects 
 have produced in our minds, or through the ideas which spring 
 directly out of our own consciousness, that we come to a know- 
 ledge of other things, is still a much debated question : but no 
 question is raised as to the truth that the individual has 
 immediately to do only with that which belongs to his own 
 individual self-consciousness, and that the world of finite things 
 which is close to him, as well as the existence of God which is 
 more remote, can be reached merely by inference. 
 
 Such an individualistic way of thinking was predestined to individualism 
 
 ** o i necessarily 
 
 end in Scepticism, whether it took a sensationalistic or an g\fgp°t*°j3„_ 
 idealistic form. If it took the former, it must end in resolving 
 our consciousness of the world as given into a mere flux of sen- 
 sations without connexion or relation ; if it took the latter 
 form, ib must end in the admission that the conscious self by 
 the mere repetition of its identity with itself can never manu- 
 facture any knowledge of objects. In either case the excliision 
 of any consciousness of the universal, as a principle which 
 manifests itself in the difference of the subjective and the 
 objective consciousness and at the same time binds them to- 
 gether, is fatal to the possibility of knowledge. Without such 
 a principle " things cannot migrate into our consciousness," nor 
 can our consciousness go out of itself to enter into them. 
 Neither the passivity nor the activity of the mind can enable 
 it to escape from the prison of individuality to which it has 
 condemned itself. It cannot by thinking " add a cubit to its 
 stature," nor can any power in the universe draw it beyond the 
 limits of its own finite existence. Unless the consciousness of 
 the not-self be from the first bound up with the consciousness 
 of self, it is hopeless to attempt to join them. The primary 
 ^ulf between the self and the not-self ca,nnot be bridged by 
 
84 INTRODUCTION. CHAP. 
 
 one who remains ex hypothesi fixed at one side of it. On the 
 other hand, if this hypothesis is not tenable, and if the con- 
 sciousness of self has the consciousness of the not-self bound up 
 with it, it is obvious that the subjective principle must, when 
 fully worked out, come into contradiction with itself. Even the 
 consciousness of self must become illusory to one who divorces 
 it from its necessary counterpart. 
 Assertion of ^hc wholc history of both the great individualistic schools 
 aslmpUedln of the eighteenth century is an illustration of the logic by 
 
 the particular. . , ^ 
 
 which the individual, conceived as immediately conscious of 
 himself and of himself alone, is gradually driven to surrender 
 all hold upon objective reality. And the critical reaction of 
 Kant may be described in general terms as a reassertion, though 
 it may be a still imperfect reassertion, of the truth that there 
 is a universal principle underlying all consciousness, in virtue 
 of which it can transcend itself and apprehend objective reality. 
 This reassertion was, however, not a simple revival of the point 
 of view of earlier philosophy. For, in the first place, the indi- 
 vidualistic protest against a merely abstract and formal philo- 
 sophy had not been wasted. It was no longer possible to be 
 content with a speculation which dissolved the finite self and 
 all that is finite in the infinite. By the advance of physical 
 science and of the industrial arts, by the development of a free 
 social and political life, the consciousness of the worth and 
 dignity of the individual and of all the immediate interests of 
 his existence had been so fully developed, that it had now be- 
 come impossible to return to mediaeval Eealism, even by the 
 circuitous path of Descartes and Spinoza. The consciousness 
 of the universal might be re-established as the basis of the con- 
 sciousness of the particular, but the former could no longer be 
 set in opposition to the latter, or used to suppress it. The only 
 universal which could now be accepted was one which should 
 furnish a principle of synthesis among the particulars, a prin- 
 ciple of unity in difference. If any idea corresponding to the 
 consciousness of God in the Cartesian and Spinozistic system 
 
HI- KANt's precursors prom BESCARTES to WOLFF. 85 
 
 was again to be placed at the head of the system of philosophy, 
 it must not be the idea of an ens realissimum, a substance 
 without limit or determination, but the idea of a self-determin- 
 ing subject. In the second place, as the system of Kant grew 
 up in close relation to the subjective individualism of the 
 previous period, even to the last it retained traces of its origin. 
 Its modes of expression, its whole armoury of metaphorical 
 formulas, all its external framework or scaffolding were bor- 
 rowed from the philosophies in room of which it came. 
 Furthermore, it was the work of one who fully appreciated 
 the valuable results of the "Enlightenment" both on the 
 scientific and the political side, and who had not the slightest 
 wish to undermine it, but rather sought to make manifest the 
 deeper foundation upon which it rested. Kant was called by 
 Hamann the Alles-zermalmender, "the man of crushing dialectic," 
 and Heine, in a passage already quoted, speaks as if this were 
 his main characteristic. But his crushing dialectic was used 
 only against the hollow logic of Wolffian dogmatism, the re- 
 moval of which was necessary to make room for a great con- 
 structive effort of speculation ; and in regard to the genuine 
 scientific work of the eighteenth century, he was only anxious 
 to show that it rested upon principles, which carry us beyond 
 its own limitations. Kanb was no revolutionist who rode to 
 success on' the top of a wave of reaction by which one half 
 truth was set up in the place of its opposite. He was, so to 
 speak, an organic reformer, whose aim was to remould the ac- 
 cepted doctrine by bringing into prominence the neglected truth 
 which was its necessary complement/ The main accu s ation 
 we can make against him is, not that he did not sufficiently 
 respect the souriar"eIenients in the system of thought which he 
 inherited, but rather that he did not see the full extent to 
 which they must be modified by the new elements he intro- 
 duced/ Thus he stops short of the necessary end of the process 
 of transformation which he initiates. But this, which is the 
 , defect of his philosophy as an ultimate result, is also that which 
 
86 INTRODTJCTION. "HAP. 
 
 gives it most value as bringing before us the whole process of 
 the great transition, of that synthetic movement by which philo- 
 sophy at the end of the eighteenth century was raised to a new 
 point of view. 
 The Leibniz- Kant appears in the first period of his intellectual develop- 
 
 ian Individ- ^-^ r icfi 
 
 uaiism. jjjg^^. j^g ^j^g supporter of a peculiar modification of the Wolffian 
 philosophy, a modification produced by the effort to bring its 
 principles into harmony with scientific ideas derived from 
 Newton. Although, therefore, we shall have in the sequel to 
 deal with the special points of contact between the critical and 
 other philosophies, it is necessary to say a few words here to 
 show the exact point at which Kant took up the speculative 
 problem. The history of the development of philosophy from 
 Leibniz to Wolff, like the history of its development from Locke 
 to Hume, is a history of the progress of Individualism to its 
 necessary consummation in Scepticism. This negative movement 
 of thought, indeed, is not so obvious and manifest in the former 
 case as in the latter. Still, in spite of the formal completeness 
 and comprehensiveness which is maintained to the last in the 
 system of Wolff, we have not much difficulty in unveiling the 
 character of the process by which the individual was gradually 
 separated from the universal, till he ceased to have any content 
 or meaning in himself. Leibniz developed his individualism 
 under the immediate influence of Spinoza. Hence he saw the 
 necessity of providing for the unity of the individual with the 
 whole of which he is a part, and especially for the relation of the 
 individual to the world which he knows. His individual sub- 
 stances or monads are, therefore, conceived as in some sort 
 universal. Each monad ideally includes, while it really excludes 
 the whole universe : if it is independent of the world, it is be- 
 cause in a manner it contains the world in itself. Leibniz lays 
 great emphasis on both of these two points of view. On the 
 one hand, the universe is merely a collective unity, and in all 
 its apparent complexity and continuity there is no reality except 
 that which is found in the isolated individuals of which it is 
 
Ill- KANT S PRECURSORS FROM DESCARTES TO WOLFF. 87 
 
 made up. " There can be nothing real or substantial in the 
 collection, unless the units be substantial." ^ And these units 
 are absolutely impenetrable to all influences from without. 
 " They have no windows through which anything might go into 
 them, or go out of them." ^ Each is a little world developing 
 under its own laws " as if there were nothing in existence but 
 itself and God." ^ And the exception made in favour of God 
 is, as we shall see, an unexplained mercy. On the other hand, 
 each monad in itself " represents," or " expresses," the whole 
 universe. It is a " living mirror, gifted with an internal 
 activity, whereby it represents the whole universe according to 
 its particular point of view, and in such a way that its ideal 
 universe has all the regularity of the real one." * It is " like 
 a separate world, sufficient for itself, independent of every 
 other creature, enveloping the infinite, expressing the Universe; 
 and it is as durable, self-subsistent and absolute as the universe 
 itself" ^ Finally the connecting link between these two 
 opposite aspects of the world is the " pre-established harmony," 
 or, in other words, it is God, the universal principle of unity, 
 who is Himself the first of monads, and who so constitutes the 
 other monads that their inner lives shall move in unison, and 
 that the ideal picture of the world which is present in each 
 shall correspond to the reality. 
 
 These are the three main aspects of Monadism, and it will Son and^Ideai 
 enable us to understand the weakness as well as the strength monads. 
 of the system, if we devote a few words to each of them. If 
 we follow out the idea of the real separateness of the monads 
 from each other we are obliged to conceive the world as 'a col- 
 lection of absolutely unrelated units. Eeal continuity, i.e., a 
 complexity which is not resolvable into indivisible units, is on 
 this principle impossible. Hence Leibniz denies the reality of 
 space and extended matter. Por the extended as such is con- 
 tinuous and infinitely divisible. The Atomists, attempting to 
 
 1 Erdmann's Leibniz, 714 ^ Id. 705. ^ 7^;. 127. 
 
 ^ Id. 714. ^ Id. 128. 
 
88 INTRODUCTION". chap. 
 
 reach individual substance, fell into the error of supposing the 
 existence of something in space, which yet was without parts. 
 But their failure only shows the impossibility of .finding a real 
 unit in space. The real units or monads, therefore, must be 
 unextended ; and the conceptions of space and extended sub- 
 stance are " confused ideas " : and they are connected with 
 other confused ideas, viz., the ideas of time, of motion, as well 
 as of the external determination of one substance by another. 
 In reality monads have only internal determinations, or in 
 other words, " perceptions," and their relations are limited to 
 correspondence in these perceptions. The world in time and 
 space is merely a phenomenal world, i.e., a world apprehended 
 through the confused perceptions of sense, and not through the 
 clear and distinct conceptions of the understanding. The 
 external world is, indeed, no mere illusion : it is a phenomenon 
 bene fundatum, a coherent and connected appearance, subject 
 to definite laws which are exhibited in substantial and physical 
 science. Nay, the connexion of phenomena in time and space 
 is in perfect correspondence with the connexion or harmony of 
 real things, though the former is determined in accordance 
 with the law of efficient, and the latter with the law of final, 
 causation. In other words, all monads are determined from 
 within, but their self-determination corresponds exactly with an 
 apparent determination from without, which takes place in their 
 phenomena as perceived by us in space and time. " Zes dines 
 agissent selon les loix des causes finales par appititions, fins et 
 moyens. Les corps agissent selon les loix des causes efiicientes, ou 
 des mouvements ; et les deux rtgnes, celui des causes efficientes, et 
 celui des causes. finales, sont harmoniques entre eux."^ They are 
 harmonious, because the latter is the reality of which the former 
 is the phenomenon, though Leibniz often speaks as if they were 
 two separate kingdoms of reality. The phenomenal world is 
 only the real world confused, or perhaps we ought rather to say, 
 refracted, in passing through the medium of sense ; but the 
 ' Monadologie, 79. 
 
Ill- KANT S PRECURSORS FROM DESCARTES TO WOLFF. 89 
 
 refraction follows a regular law of change, transmuting the 
 harmonious development of the inner life of the different self- 
 evolving monads into the external connexion of iDhenomena 
 which act and react on each other in space and time. 
 
 Whence, then, comes that '' confusion '" ia the inner life of "^^ oontinu. 
 
 ' ous scale of 
 
 the monad, by reason of which its representation of the ^^'°*' 
 universe falls short of reality and sinks to the phenomenal ? 
 In answer to this question Leibniz is obliged greatly to qualify 
 his doctrine as to the self-determination of the monads.. God 
 alone, he declares, is actus purus, but into the original constitu- 
 tion of every other monad he has introduced a passive element, 
 a. prima materia, a. limit: and its absolute spontaneity^ therefore, 
 does not mean that it is unlimited, but that it is limited only 
 by the negative element of its own nature. !Now as the activity 
 of the monad is representative or perceptive, this limit takes 
 the form of a certain confusion in its perceptions. The whole 
 world is present to each monad, but present always with more 
 or less confusion or complexity, owing to the passivity or nega- 
 tion which belongs to them as finite. With this is connected 
 the Leibnizian view of the universe as a graduated scale of 
 being. All monads are perceptive beings, distinguished from 
 each other by the degree of clearness of their perceptions ; and 
 the chain of life extends downwards without a break from God, 
 who apprehends all things in the clearness of pure thought 
 through finite spirits like men, who apprehend the world partly 
 in the light of thought and partly in the confusion of sense, to 
 animals which have only sense perception, and from these 
 again down to monads whose'perceptions are too confused even 
 for sense. But perceptions there must stUl be wherever there 
 is existence : for otherwise there would be no internal deter- 
 mination of the monad, no ideal centre or self to which its life 
 could be referred. The world is thus through and through 
 organic : and, as we have already seen, inorganic matter exists 
 only in the confused ideas of sense. g^^^ ^^^ 
 
 . Man, then, stands midway in the scale of being, having not ^SJ.^''*'" 
 
90 INTRODUCTION. chap. 
 
 only perception with its " confused " apprehension of things in 
 their phenomenal appearance in time and space, but thought or 
 apperception, which enables him to grasp their real nature as 
 they are in themselves. What, then, is the relation between 
 those two ways of knowing ? The answer which seems most 
 naturally to follow from the principles of Leibniz is that, as 
 each monad represents the universe, a clear consciousness of 
 itself would enable any monad to see all things in their true 
 nature. Hence, a perfect intelligence would be ruled in all its 
 consciousness by the law of identity and all things would be 
 revealed to it through pure apperception, i.e., through the 
 analytic consciousness of itself. That it is not so with us but 
 that we are conscious of objects as external to us and to each 
 other in space, and that we are obliged to argue from the one 
 to the other according to the law of sufficient reason (or 
 causality) is due to the infinite complexity and confusion in 
 which they are given to us as phenomena through sense. Still 
 the difference of sensation and thought is merely a difference of 
 degree and not of kind. Sensation is but confused thought, and 
 thought but distinct sensation. And there would seem to be 
 no reason why our knowledge should not be raised from con- 
 fusion to distinctness alid so from the phenomenal to the real 
 by the continued progress of analysis, or why at least we should 
 not be continually approximating to the ideal of a determina- 
 tion of things in pure apperception according to the law of 
 identity and without any aid from the principle of sufficient 
 reason. Yet we find that, wherever Leibniz passes to a more 
 exact consideration of the nature of the two ways of knowing, 
 he treats the distinction between perception and apperception 
 as one of kind and not of degree. The opposition between 
 contingent and necessary truth, between truth of fact and truth 
 of reason, is for him not fluctuating, but fixed and unchange- 
 able.' The latter can, he thinks, be carried back by analysis to 
 
 ^This view is developed by Leibniz, especially in the Nouveaux Bssais, 
 where, in answer to Locke, he draws a wide distinction between the iutelli- 
 
HI. KANT's PRECURSOES prom DESCARTES TO WOLEF. 91 
 
 self-evident principles and indeed to identical propositions, 
 while the former is entirely incapable of being so analysed'. 
 " I use," he declares, " two principles in demonstration : one of 
 them is the principle that whatever contains a contradiction is 
 false ; the other is the principle that for every truth which is 
 not an identical proposition, a reason can be given. In other 
 words, the notion of the predicate is always explicitly or 
 implicitly contained in the notion of the subject, and this is 
 the case not less in contingent than in necessary truth. The 
 distinction between contingent and necessary truth very closely 
 resembles the distinction between commensurable and incom- 
 mensurable numbers. Just as we can always find a common 
 measure for commensurable numbers, so we can always demon- 
 strate necessary truth, i.e., we can always carry them back to 
 identical propositions. On the other hand, just as the analysis 
 of a ratio of incommensurables produces an infinite series, so 
 contingent truths require an infinite analysis which God only 
 can complete, "Wherefore it is by him alone that they are 
 known certainly and a priori : for though a reason can always 
 be found for the state that succeeds in the state immediately 
 before it, yet this reason requires another reason, and so on 
 ad infinitum. And this processus ad infinitum takes in our 
 knowledge the place of a sufficient reason, which can only be 
 found outside the series in God, on whom all its parts, prior 
 and posterior, depend, far more than they depend on each other. 
 For, when a truth is incapable of final analysis, and cannot be 
 demonstrated from its own reasons, but derives its final reason 
 and certitude from the divine mind alone, it is not necessary. 
 
 gence of man. and that of the animals. Apperception, it is there asserted, 
 involves that apprehension of universal and necessary truths which gives rise 
 to science, while the animals, being confined to mere perception, are necessarily 
 " pure empirics." It is only another way of stating the same idea when it is 
 said that the other monads represent or express rather the world than God, but 
 that spirit-monads represent or express rather God than the world. Leibniz, 
 however, goes even beyond this and practically adopts the view of Malebranohe 
 when he says that the thinking monad has one immediate external object, viz., 
 ♦God. (Erdmann's Leibniz, 222. ) 
 
92 INTRODUCTION. <^^-"'- 
 
 Such are all those I call truths of fact, and this is the root of 
 their contingency which I doubt whether any one hitherto has 
 explained." ^ 
 Is sufficient Qn this view it is obvious that the principle of sufficient 
 
 reason iden- ^ ■*- 
 
 oiuse°?*°'^ reason would merge in the higher principle of identity, if only 
 the reason were really sufficient : but the limitation of human 
 intelligence is just that such a perfect analysis is impossible to 
 it. Thus, though the distinction is really one of degree, it 
 remains one of kind for us : and this is why we need two 
 separate principles of knowledge. Leibniz, however, gives 
 another view of the principle of sufficient reason which iden- 
 tifies it with the law not of efficient, but of final, causation 
 and makes it the higher principle to which the principle of 
 identity is subordinated. According to this view, a distinction 
 must be drawn between metaphysical and moral necessity, i.e., 
 between the necessity of that which we cannot think otherwise 
 and the necessity of that which we can think otherwise but 
 which must be as it is, because it is for the best, — ^must be 
 because it ought to be. The former necessity, which depends 
 on the principle of identity, fixes the bounds of possibility : 
 even the power of God cannot give existence to that which is 
 self-contradictory. The latter necessity, which depends upon 
 the law of sufficient reason, determines the content of reality : 
 it is the necessity by which God's goodness makes him create 
 " the best of all possible worlds," or, as it is more definitely 
 expressed, the world in which there' is the highest sum of 
 ' compossible ' existence, i.e., of realities that can exist together 
 in one world. JBut'what is the highest sum of all compossible 
 existence ? To Descartes and Spinoza this question could 
 present no difficulty ; for they both accepted the principle that 
 the absolute reality was the unity of all affirmations to the 
 exclusion of all negations. And on that principle there could 
 be no possible discord between positive realities, and all 
 things that were possible would be ' compossible.' Thus no 
 
 ^Erdmann's Leibniz, 83. 
 
III. KANT S PRECURSORS FROM BESOARTES TO WOLFF. 93 
 
 room -would be left for the distinction between metaphysical 
 nnd moral necessity, between that which is because it is, and 
 that which is because it is for the best. But as we saw, the 
 necessary result of this way of thinking was a Pantheism, for 
 which all distinction and limitation, all finite being, disappeared 
 in the one absolute substance,' — a system which could not but 
 be rejected by Leibniz, the keynote of whose speculation is 
 individuality and self-determination. Hence the idea to which 
 Leibniz holds, though by no means in a clear and consequent 
 way, is that the absolute Being is not a mere ens realissimum 
 or unity of affirmative predicates, but a self-determining monad, 
 who reveals himself in a world of monads, each of which has 
 its own self-determined individuality, distinct and independent 
 of that of all the others, while yet all are held together in the 
 pre-established harmony of one world. In this world there 
 are no " indiscernibles," no monads absolutely like each other : 
 yet each is a microcosm in which the whole is represented. 
 But the p^'incipium melioris or principle of sufficient reason, on 
 this view of it, is not merely different from the principle of 
 identity but directly opposed to it. For while the latter is an 
 analytic principle which, if it were taken as absolute, would 
 force us to seek for identity under every appearance of differ- 
 ence, the former is a principle of synthesis which involves that 
 there is no identity which does not manifest itself in real 
 differences.' Leibniz, however, as he escaped the opposition 
 between true indi^^iduality and the universal relativity of the 
 monad by making the one real and the other ideal, so he 
 evades the opposition between the principium melioris and the 
 ■principle of identity by making the latter a principle of possi- 
 bility, and the former a principle of reality. And he hides 
 from himself the inconsistency of having two first principles 
 by the conception of a God who selects among possibilities 
 those which are capable- of combination,, and which, when 
 combined, will produce the greatest sum of reality. " It is yet 
 unknown to me," he declares, " what is the reason of the 
 
94 INTRODUCTION. <^H-*-P- 
 
 incompossibility of different things or how it is that the 
 natures of different things can be opposed to each other, seeing 
 that all purely positive terms seem to be compatible." This 
 assertion shows how uncertain is Leibniz's hold of the prin- 
 ciples by which he is guided. For the last clause can only 
 mean that he has no intelligible reason to give for departing 
 from the principles of Spinoza, though he is aware that they 
 are fatal to his characteristic doctrine of the independent 
 reality of individual substances. With similar inconsistency 
 he accepts the definition of God as the union of all affirmative 
 predicates, not seeing apparently any more than Spinoza that 
 such a definition is inconsistent with the idea of God as a 
 causa sui or self-determining monad, and still more inconsistent 
 with the idea that he is a self-revealing subject or good being, 
 who, because he is good, must create other beings like himself. 
 Again, while he admits the necessity of reciprocal negation, 
 limitation and conflict in the world of finite monads, he yet 
 does not seem to regard this merely as an accidental evil, 
 bound up with the attainment of the highest good. Hence 
 the highest good is for him not a positive good which is 
 realised in and through negation and evil, but an abstract sum 
 of reality which can be attained in spite of such negation. 
 Finally, as might be expected, in his conception of the relation 
 of God as the highest monad to the other monads, he is poised 
 between two inconsistent alternatives : for whenever he is in 
 earnest with the reality of the individual substances, God is 
 reduced to another word for the harmony that prevails between 
 them ; and whenever he is in earnest with the reality of God, 
 all the other monads shrink into momentary expressions or 
 " modes" of his unity. If " God alone is the primitive unity 
 or simple originative substance, of which all created or deriva- 
 tive monads are the productions, born, as it were of the 
 continual fulgurations of divinity from moment to moment,"^ 
 the independence of the monads shrinks within very narrow 
 ^ Erdmann's Leibniz, 708. 
 
III. KANT S PRECURSORS TROM DESCARTES TO WOLFF. 95 
 
 limits. If, on the other hand, " every spirit " (and therefore in 
 a sense every monad) " is like a separate world, sufficient for 
 itself, independent of any other creature, enveloping the in- 
 finite, expressing the universe, and as durable, as self-sufficient 
 and as absolute as the universe itself," ^ where is there room 
 for God ? 
 
 It may now appear what is the weakness and what is the summary 
 
 view of the 
 
 strength of the philosophy of Leibniz. Its strength lies in this £ei^°n?f''y "* 
 that, while he introduced into philosophy the principle of 
 individuality, which was characteristic of the following century, 
 he yet keeps hold of the principle of universality which was 
 characteristic of the previous age. His weakness lies in this, 
 that he achieved no real synthesis between the two elements 
 which he thus brought together. He puts side by side the 
 real individuality of the monad and its ideal relativity to the 
 universe ; the absolute independence of each substance and the 
 immediate relation of all substances to God; the analytic prin- 
 ciple of identity and the synthetic principle of sufficient reason ; 
 the idea of God as the ens realissimum, who absorbs all positive 
 existence into himself and the idea of Him as the self-revealing 
 spirit, whose nature it is to create other monads different from 
 himself and from each other and through their difference to 
 realise the highest unity. Nor does he ever attain anything 
 more than an external " harmony " between these different 
 sides of his philosophy. 
 
 A system whose parts were joined together with such un- 
 tempered mortar could not but yield to the dissolving force of 
 time, and it was natural that, in the first instance at least, the 
 inconsistency should be remedied not by seeking for a deeper 
 principle of reconciliation between the different elements, but 
 by dropping those of them which were least in harmony with 
 the spirit of the age. Wolff "swept and garnished" the 
 Leibnizian philosophy. Without changing the outer frame- 
 work of the system, he sought to make it self-consistent, and in 
 • ' Erdmann's Leibniz, 12S. 
 
96 INTRODUCTION. chap. 
 
 doing so he gradually eliminated all those speculative elements 
 which saved it from the emptiness of a formal Individualism. 
 At the end of the process it was found that the kernel had 
 been removed and only the husk left for show, and thus 
 the hour of triumph of the Wolffian philosophy was the hour 
 of its fall. 
 Wolffs con- Wolffs first step was to discard the synthetic principles 
 
 sietent Indi- ^ 
 
 diSyrthe which Leibniz had introduced, by reducing the law of sufQcient 
 phuSS'. reason to the law of identity ; and the way in which he reached 
 this result shows that he took that law in the strict Spinozistic 
 sense. He begins with two definitions : Nihilum, est cui nulla, 
 respondet notio and aliquid est cui notio aliqua respondet. Thus 
 " something " and " nothing " are contradictories between which 
 there is no middle term. But no repetition of nothings can 
 make them equivalent to something, or, in other words, ex 
 nihilo nihil fit : which is simply the converse of the proposition 
 that there is a positive reason for everything. It is evident that 
 this argument rests on that absolute separation of affirmation and 
 negation of which the necessary result was seen in Spinozism. 
 For, strictly speaking, it involves not only that nothing can 
 come of nothing, but that there is no " coming " or development 
 at all, either in thought or reality. Hence there is no meaning 
 left for the distinction of reason and consequent, which Wolff 
 suddenly produces upon us as by a stroke of logical sleight of 
 hand. Wolffs definition of " nothing," id cui nulla respondet 
 notio, would be sufficient to refute him: for it is a definition 
 of the indefinable : and as the definition of " something " has 
 significance only by contrast therewith, it too becomes unmeaning. 
 Thus Wolffs argument implicitly involves the principle that 
 " all determination is negation," the immediate corollary of 
 which is that positive reality can only be found in an absolutely 
 undetermined being or substance. But while Wolff thus goes 
 back to the formal principle of Spinoza, he is as far as possible 
 from coinciding with Spinoza's tendency to dissolve all finite 
 being in the universality of substance. On the contrary the 
 
Ill- KANt's precursors from DESCARTES TO WOLFF. 97 
 
 presupposition, upon which he always rests as an absolute basis 
 of certitude, is the pure individuality of all real being ; and he 
 rejects altogether the qualifying idea by which Leibiliz saved 
 himself from Atomism, that each monad is representative of the 
 universe. The only remnant of this which Wolff retains is the 
 idea that the soul is in pre-established harmony with the body, 
 so as to apprehend its affections and the world through them, 
 and it is only in this sense that he can still call the soul a vis 
 repraesentativa universi. Even this part of the Leibnizian theory 
 he accepts with some hesitation, as being the most probable 
 Recount of the relation of two substances so heterogeneous as 
 soul and body, between which it seems impossible to conceive 
 of any direct relation. In like manner he retains the Leibnizian 
 idea of God, as an external power who combines into one 
 universe the individual substances which in themselves have 
 no necessary or essential relation to each other. In truth, 
 without these two assumptions the unity of things with each 
 other and with the mind that knows them must have altogether 
 disappeared in the Wolffian philosophy. For in it the harmony 
 of soul and body is the one link between consciousness and the 
 world — the one principle which saves the individual from being 
 imprisoned in his own self-consciousness. And the idea of a 
 Grod, who keeps together the disjecta membra of the world and 
 fits them to each other by an external process of arrangement, 
 is a necessary expedient to correct the isolation to which things 
 were reduced by the individualistic principle. For, if the 
 world is merely a collection of independent things, it is obvious 
 that its parts must be held together by a foreign hand, and if 
 the mind and its object are essentially disparate, their relation 
 can only be an externally produced harmony. With Leibniz, 
 indeed, these consequences of individualism were partly con- 
 cealed. For, with him the ideal harmony of things was always / 
 on the point of passing into their real unity as organic elements 
 of one world, and the principle of identity, by which each 
 thing was referred to itself, had beside it the principle 
 VOL. I. G 
 
98 INTRODUCTION. CHAP. 
 
 of sufficient reason, by which it was related to all other 
 things. But, as with Wolff the whole process of reason 
 reduces itself to analysis, there is for him no possibility 
 of bringing into essential relations any elements of thought 
 or reality which have once been separated or even dis- 
 tinguished. It is only in so far as he derives from experience, 
 or from the common consciousness, the idea of a connexion 
 of things with each other or with thought that he has anything 
 left to analyse ; and if his analysis does not end in absolutely 
 destroying any such connexion, it is an unexplained mercy, 
 or a mercy which can be explained only by the imperfect way 
 in which Wolff realised what he was doing. As it is, all con- 
 ' nexion or unity is reduced by him to an external composition. 
 All, therefore, that was necessary, in order to bring down the 
 whole edifice of the Wolffian system like a house of cards, was 
 that scepticism should be directed against the principle of con- 
 nexion which it still retained ; or in other words, that it should 
 be pointed out that on Wolff's principles there was no rational 
 ground for any belief in the unity of objects with each other 
 or with the knowing subject. 
 priM^pStho What has just been said may be illustrated from another 
 power of '^° point of vicw by a short regress. So soon as the principle of 
 
 reason must 
 
 be denied. identity ccascs to be taken in the Spinozistic sense as the 
 negation of all real differences ; so soon, in short, as it ceases to 
 be understood as a principle of analysis, which can only be 
 used to exhibit more definitely the different elements contained 
 in any given subject of predication, it becomes obvious that no 
 process of thought founded upon it can enable us to go beyond 
 the data with which we start. Either, . therefore, we must 
 suppose that there is some other principle of thought by which 
 its defects are supplied, some principle of synthesis, like 
 Leibniz's principium melioris, which may be brought to the aid 
 of the principle of analysis ; or we must suppose that' all 
 the connexions of things are given in experience, leaving 
 to thought no task except to analyse them : or, lastly, we must 
 
Ill- KANT S PRECURSORS FROM DESCARTES TO WOLET. 99 
 
 suppose that the synthetic unity of knowledge arises in some 
 way out of the combination of the manifold data of sense 
 passively received with an activity of thought. Now the third 
 of these solutions of the difficulty we may leave out of account, 
 as it was not conceived, or at least clearly worked out, by any 
 one before Kant. The first solution was formally rejected by 
 Wolff, when he reduced the principle of sufficient reason to the 
 principle of identity. Hence it would seem inevitable that he 
 should have accepted the second alternative. And this he 
 practically did, though with no distinct consciousness of what 
 he was doing. It is true that he draws a wide distinction 
 between the a priori and the a posteriori, the pure and the 
 empirical, parts of each department of philosophy, between 
 rational and empirical Psychology, rational arid empirical 
 Cosmology, etc. But this distinction has no meaning in a 
 system in which reason is reduced to the function of analysis, 1 
 and is supposed to be unable to establish any relation between 
 ideas except that they contain, or do not contain, each other. 
 In fact, Wolff is obliged frequently to appeal to experience, not 
 only for all his data, but also and above all for the connective 
 principles by which these- data are related to each other : and 
 even where the appeal is not directly made, we can easily see 
 that such principles are tacitly borrowed from the ordinary 
 consciousness of his time. Now, so soon as the critics of the 
 Wolffian system began to detect this secret, the effect was to 
 discredit reason as a source of knowledge and to reduce it to a ' 
 merely formal activity. And as reason was discredited, 
 its place could be taken only by experience. The result 
 of the Wolffian method was, therefore, the very opposite of that 
 which its author proposed. He had been possessed with a 
 noble confidence in the power of intelligence to clear up all 
 darkness and to penetrate all the mysteries of things ; but in 
 the end he seemed only to prove the barrenness of thought in 
 itself,, and the necessity that everything should be given to it 
 • from without. He sought to show that the world was an 
 
100 INTRODUCTION. "HAP. 
 
 intelligible system, in which the mind could find the counter- 
 part of its own transparent unity ; but, if he proved anything, 
 it was that the whole nature of the world must be taken as an 
 unintelligible fact, and that the bond of union between its 
 elements is something upon which thought can throw no light, 
 something which must be accepted by a faith which is either 
 above, or below, reason. Wolff thus played into the hands of 
 those to whom he was most opposed, the Empiricists and 
 the Pietists. For each of these proposed to supply the 
 evidence of immediate experience for certain of the prin- 
 ciples which Wolff was obliged to assume, the former for 
 the principles supplied in his physical, and the latter for the 
 principles implied in his metaphysical, philosophy. On the 
 other hand, if reason were to reject such assistance, if it were 
 to refuse to accept the external aid of faith and experience, 
 what must be the result ? The analytic philosophy which 
 separated object from object, and subject from object, must end 
 in depriving all these elements of their meaning. For an 
 (individual thing, separated from all its relations, becomes 
 I a " thing in itself," of which nothing can be said ; and a thinking 
 subject which has no consciousness of any object but itself, is 
 necessarily limited to what Kant calls the " analytic unity of 
 apperception," i.e., the mere tautological " I am I " of self- 
 consciousness. In short, the abstract individual, separated from 
 all other individuals, is as empty and indeterminate as the 
 abstract universal, and for the same reason. For a philosophy 
 which excludes all synthetic principles of 'reason is forced in 
 the end either to accept the truth as an unintelligible datum 
 from without, or to reduce it to a blank form without any 
 content or matter. 
 P,l''?L°,*™ °* It cannot be said that these inferences were distinctly drawn 
 
 Wolff s pre- .' 
 
 harmonytf ^^^her by those who supported, or by those who opposed, the 
 
 body and soul. -Yyoifgan philosophy. But the logical effect of a system is 
 
 often felt or anticipated long before it is explicitly stated. In 
 
 this case it showed itself in the growing influence of its rival. 
 
Ill- KANT S PREOTJRSQRS FROM DESCARTES TO WOLFF. 101 
 
 the empirical philosophy, in the development of several forms 
 of Eclecticism, and finally in the gradual transformation of the 
 Wolffian philosophy itself at the hands of its adherents. This 
 transformation began, as might be expected, by a criticism of 
 the principles of connexion between individual substances, which 
 •were still retained from Leibniz in the Wolffian philosophy. 
 For Wolff, insisting upon the conception of the monads as 
 isolated individuals and as, therefore, in all the changes of their 
 states independent of any action from without, had rejected 
 Leibniz's qualifying idea that these monads perceive or repre- 
 sent the world. Perception and representation were for him 
 the attributes of the spirit-monad only, and even it was 
 regarded as, in the first instance, representative only of the 
 body attached to it. But with this change of the Leibnizian 
 doctrine, the inner life of those monads which were not souls 
 became a blank, and the spirit-monad was so widely separated 
 from all the others that its perception of them became some- 
 thing anomalous and almost miraculous. It was this that 
 caused Wolff to be so hesitating in his assertion of the pre- 
 established harmony between body and soul. His hesitation, 
 however, only betrayed the weak place in his system, upon 
 which all its opponents at once directed their attack. The 
 crude form of the theory of an injluxus physicus, i.e., of a mere 
 determination of the passive soul from without as by material 
 impact, could easily be refuted by those who had been taught 
 by Leibniz that there is no such passivity even in matter, but 
 that every material element as such is a centre of. force. But, 
 on the other hand, the science of mechanics seemed to show 
 that there is no force or determination which is purely internal 
 and is not correlative with a determination from without ; and 
 the same principle seemed applicable mutatis mutandis to the 
 relation of soul and body, in the sense that there is no 
 consciousness of self which is not a reaction upon a stimulus 
 from without. Baumgarten, who finally summed up the result 
 • of the controversy, preserved the Wolffian doctrine in name 
 
102 INTRODUCTION. ™'*^^- 
 
 while sacrificing its substance, when he denied real, and 
 admitted only ideal, influence of monads upon each other; 
 defining a real influence, however, as one to which the object 
 acted on contributes nothing, while an ideal influence is one in 
 which the effect produced in one substance by the action of 
 another involves an activity in the substance affected.' Accord- 
 ing to this view it was still possible to assert that there are 
 none but ideal relations between substances, in the sense that 
 one substance cannot act upon another substance except to 
 stimulate it to the activity that is characteristic of it ; and that 
 in particular a material substance can be stimulated only 
 to motion and a thinking substance to thought. It is, how- 
 ever, obvious that by this interpretation the idea of the pure 
 self-determination of substances, which Wolff had inherited 
 from Leibniz and had even exaggerated, was altogether 
 surrendered.^ 
 Knutzen's It was Knutzcu, the teacher of Kant, who took the foremost 
 
 synthesis of 
 
 Ne-Jrton'"* part in the controversy which led to this result — a result 
 which really meant the surrender of the central principle of the 
 Wolf&o-Leibnizian philosophy, i.e., the pure inward self-deter- 
 mination of the individual substance. Knutzen, however, does 
 not seem to have reached his conclusions by simple reflexion 
 upon that philosophy. He did not begin by recognising that a 
 purely inward self-determination is empty or self-contradictory, 
 and then proceed to argue that an individual substance cannot 
 determine itself except in relation to another substance which 
 acts upon it. Eather, his view was suggested by the study of 
 Newton's Princi^ia, and it was developed in the effort to com- 
 bine Newton's ideas with the principles he had previously learnt 
 from Leibniz and Wolff. At the same time, it was the inner 
 necessity of this new idea, — the necessity for it arising from the 
 development of the Wolffian philosophy itself, — which made 
 the acceptance of it so rapid and irrevocable. With this syn- 
 
 1 Benno Erdmann [Martin Knutzen und Seine Zeit, oh. 4) has given the first 
 clear account of this controversy and its result. 
 
in. KANT's precursors FROM DESCARTES TO WOLFF. 103 
 
 thesis of Knutzen began the process of organic combination 
 between the hitherto separated and opposed streams of English 
 and German thought, between the Empiricism of Locke and the 
 Eationahsm of Leibniz: a process which could not end till it 
 had transformed the problem of philosophy by raising it to a 
 point of view above that of either of these one-sided schools. A 
 short sketch of the earlier period of the philosophical develop- 
 ment of Kant will show us how he attained to this new point 
 of view. An account of the later history of it will show how 
 from it he gradually accomplished the consequent transforma- 
 tion o,f philosophy. 
 
104 
 
 CHAPTEE IV. 
 
 THE PRE-CEITIOAL PERIOD OF KANT'S MENTAL HISTORY. 
 
 Kanfs earliest TZ'ANT, as we have seen, was educated during the period in 
 
 writiogs. rV ' 
 
 «eoMU? *° which the Wolffian philosophy was in process of disintegra- 
 
 Leibniz. ''" tion even at the hands of its own disciples, and he was the pupil of 
 one who did much to hasten that process. Knutzen sought to 
 correct Wolff by the aid of Newton ; and in doing so he was led 
 to give the coup de grdce to the theory of pre-established harmony, 
 and to revive in a better form the theory of an inflivxus physicus. 
 In the first period of Kant's career his thoughts ran in the same 
 groove with those of his teacher; i.e., he was still under the general 
 influence of Wolffian Eationalism, attempting only to modify it 
 in such a way as to make room for the mechanical conception 
 of nature initiated by Newton. Thus in his youthful essay on 
 vis -viva, he not only rejects the Cartesian idea of matter as a 
 passive substance which is determined purely from without, in 
 favour of the Leibnizian idea that it is made up of elements 
 which are active centres of force, but he goes on to turn this 
 conception against Leibniz himself, who had supposed that sxich 
 force belongs only to moving bodies. He contends, on the con- 
 trary, that it must be conceived as itself a source of movement. 
 Connected with this is the idea that the individual substances 
 are possessed of attractive, and not only, as Leibniz had main- 
 tained, of repulsive, force in relation to each other. This idea 
 Kant, of course, received from Newton, of whom he was an 
 
CHAP. IV. PEE-CRITICAL WRITINGS OF KANT. 105 
 
 enthusiastic student. Nor did he content himself with repeat- 
 ing Newton's ideas, he also tried to develop them, and that in 
 two directions. On the one hand, in the Natural History and 
 Theory of the Heavens, he tried to show that the mechanical 
 principles which Newton had used to explain the solar system 
 would account also for its genesis ; i.e., they explain how matter 
 diffused through space would, by the reciprocal attraction and 
 repulsion of its particles, be turned into a system of planetary 
 bodies revolving round a central orb in the way exempli- 
 fied in the solar system. On the other hand, in his Monado- 
 logia Physica, he attempted by a regress from the Newtonian 
 theory to determine the ultimate constitution of matter. The 
 idea of actio in distans, which was rejected by Leibniz and 
 which even Newton accepted only as a provisional hypothesis, 
 was asserted by Kant as a necessary part of the dynamic idea 
 of matter. Kant argued that if, according to the Leibnizian 
 view, the primary substances are points of force, they must be 
 simple unextended substances which, taken individually, cannot 
 be conceived as occupying space at all. But they may occupy 
 it in relation to each other, in so far as each monad excludes 
 or repels the others (with a force which as it acts in every 
 direction, must be supposed to vary inversely as the cube of the 
 distance between them). On the other hand, if this were the 
 only force which the monads possessed, they could not consti- 
 tute matter as we know it ; for under its action they would 
 isolate themselves from each other and dissolve all continuity : 
 they could never retain any definite volume or mass. "We 
 must, therefore, suppose that the monads are held together by 
 an opposite force, which, as it depends upon the spherical 
 superficies over which the action is extended, varies inversely 
 as the square of the distance. 
 ■ In this theory it is to be observed that Kant, like Leibniz, Kant begins 
 
 •' ' ' with the 
 
 begins with the individuality of things as separate substances S,t'proceed8 
 or monads existing in simple self- identity ; he then goes on to andpMitTve, 
 conceive them as having repulsive force, i.e., as negatively 
 
106 INTRODUCTION. chap, 
 
 related to each other, each, so to speak, asserting itself and 
 negating the others : and finally he shows that this negative 
 relation implies a positive relation, an attractive force which 
 they exercise upon each other ; for otherwise they could not be 
 kept in relation so as even to repel each other. Both forces 
 are thus necessary to the occupation of space. The external 
 way in which these different elements are added to each other 
 exhibits clearly the process of self-correction by which Kant 
 reaches his final conception. And, as we shall have many 
 occasions of observing, it is characteristic of Kant that, while 
 he qualifies the individual self-identity of the monad, first by 
 the conception of its negative, and then of its positive rela- 
 tivity, the result he thus reaches does not lead him to go back 
 upon his starting point, or to set aside the absolute individuality 
 of the monad with which he began. 
 His first Kant's Dilucidatio Nova, published in 1755, in the same 
 
 Metaphysical 
 
 Beassertion of 7^^^ '^^^^^^ ^^^ Monodologici PJhysica, enables us to see more 
 of suEeS'^ exactly the degree to which he still retained his hold upon the 
 general principles of the system of Wolff. In that treatise he 
 starts, like Wolff, with the principles of identity and contradic- 
 tion as the highest principles of truth. But he immediately 
 goes on to treat as a principle at least partially independent 
 the law of sufficient, or rStner, as he prefers to call it, of deter- 
 minant reason. Determinant reason, he adds, appears in two 
 different forms : as ratio antecedenter determinans, or as ratio 
 conseguenter determinans. The latter is a mere ratio cognoscendi, 
 which enables us to connect a given predicate with the subject, 
 but does not tell us why it is so connected in reality. The 
 former is a ratio essendi ml fiendi, which not only explains the 
 connexion but is the cause of its existence (non, solum explanat, 
 sed ejicit). Thus, to take Kant's example, the eclipses of Jupiter's 
 satellites may be the means of our discovering the rate of the pro- 
 pagation of light, but no one would say that these eclipses' are 
 the real reason of the rate being such as it is. On the contrary, 
 the phenomena of these eclipses could be what they arfe only 
 
 reason. 
 
l^- PEE-CRITICAL WRITINGS OF KANT. ,107 
 
 on' the presupposition of the actual constitution of light. But 
 if we were to adopt the Cartesian theory, that the reason of the 
 propagation of light lies in the definite elasticity of the globes 
 of ether, and if we were able on this basis to calculate the rate 
 of its propagation, we should be tracing it to its real reason or 
 ratio antecedenter deter minans} 
 
 From this it would appear that the ratio consequenter deter- '^^^^ ^"* 
 ininans is the fact or datum by analysing which we discover a s™""'^- 
 principle ; whereas the ratio antecedenter determinans is the 
 principle in which we find the reason why the datum or fact is 
 what it is. Under the head of ratio antecedenter determinans, 
 however, Kant brings both the ratio veritatis and ratio existen- 
 tiae, the ideal and the real ground, pointing out that in the case 
 of the former the ground may be either that the predicate is 
 involved directly in the nature of the subject, or that it is con- 
 nected with it by some tertium quid. We may, therefore, say 
 that no true proposition lacks a ratio antecedenter determinans. 
 It is somewhat different in the case of the ratio existentiae, for 
 we cannot say that everything has a real reason or cause. We 
 can only say that all contingent existence has a real reason, but 
 that ultimately the reason for all such existence must be found 
 in that necessary being which has no ratio antecedenter deter- 
 minans. This distinction of the real and ideal reason, which > 
 Kant adopts from Crusius, leads him decidedly to reject the 
 idea that existence can ever be included as an element in the 
 conception of an object. If we speak, in the language of 
 Wolff, of the essences of things as absolutely necessary, all 
 that we can mean is not that the things in question necessarily 
 exist, but that, given their existence, they necessarily have 
 certain predicates. " The essence of a triangle, which consists 
 in the construction of three straight lines in a certain way, is 
 not per se necessary. For what sane person would argue that 
 the three sides must be always conceived as meeting in that 
 way ? All that is necessary by the principle of identity is that, 
 iR. I, 11; H. I. 373. 
 
108 INTRODUCTION. chap. 
 
 if you think of a triangle, you are bound to think of the sides 
 as so united."^ This consideration is fatal, as Crusius had 
 shown, to the ontological argument for the being of God in its 
 usual form. For all it proves is that, if we include existence 
 in our idea of God, we are bound to determine him as existing. 
 In order to make it a valid argument, we are obliged, therefore, 
 to give it a new form. We cannot prove God's existence from 
 the essence or conception of Him, as from a ratio antecedenter 
 determinans, but we may show that His existence is presupposed 
 in the conception of Himself, and indeed in the conception of 
 anything. Datur ens cujus existentia praevertit ipsam et ipsius et 
 omnium rerum possibilitatem. Vocatur Deus? The proof of 
 this is that the question of possibility can only arise when we 
 have certain determinate data which we can compare, and as to 
 which we can ask whether it is, or is not, possible to combine 
 them without contradiction. But this means that we take the 
 data as real in considering the possibility of their combination. 
 Hence, ia considering the question of possibility, we are always 
 dealing with the modified or determinate forms of a presupposed 
 primary reality, to which we are necessarily carried back as the 
 " material of all possibility," the underlying unity of which all 
 conceivable existence is some special limitation. This, there- 
 fore, is the ratio antecedenter determinans of all existence, for 
 which it is absurd to seek a reason. On the other hand, 
 everything but this primary reality is contingent and deter- 
 mined to exist by a reason other than its own nature. Hence 
 Kant distinguishes the ratio veritatis and the ratio adualitatis 
 s. exisfentiae in the following way. " In the former the ques- 
 tion is only as to the assertion of a predicate, which we are 
 enabled to assert by reason of its identity with the subject 
 regarded either in itself or in its connexion with other thiags." 
 In this case, therefore, the ratio antecedenter determinans 
 reduces itself to the principle of identity, and " all we 
 do is to throw light on a ' connexion of subject and 
 ^ R. I. 15 ; H. 1. 376. ^^ R. I. 14 ; H. I. 376. 
 
IV. PBE-CRITICAL WRITINGS OF KANT. 109 
 
 predicate which is. already given. But in the latter 
 case we are examining not merely whether the predicate 
 belongs to the subject or not, but why it does so. If the 
 opposite predicate be excluded by the absolute position of the 
 subject, we have va. the subject a being whose existence is 
 necessary absolutely and per se. But if not, then there must be 
 something else which, by positively determining the subject so 
 and not otherwise, excludes the existence of the opposite." ^ A 
 ratio antecedenter determinans or real ground is, therefore, required 
 only in the case of the existence of finite substances and their 
 states. Yet of the general reason for the existence of finite 
 substances Kant gives no account and could give none ; for the Kant does not 
 
 give a real 
 
 God he has proved is merely the indeterminate substance of 1^°™?'"*''^ 
 
 ^ >/ finite m 
 
 Spinoza, which contains in itself no reason for the existence of on?y for''"' 
 anything but itself. The self-limitation by which God is sup- phenomena. 
 posed to create finite things is, therefore, a conception for 
 which Kant provides no justification (any more than Spinoza 
 could explain the existence of attributes and modes).^ In place, 
 therefore, of the inference which Kant actually draws, viz., that 
 the " quantity of reality in the finite world cannot be naturally 
 increased or diminished," it would have been more logical to 
 infer that the absolute substance is without change or limit. 
 When, however, the existence of finite substances has once been 
 assumed, Kant is able to give a better account of the reasons or 
 causes of their changing states. For, from the general principle 
 of determinant reason, he derives two subordinate principles, 
 which he calls the principle of succession and the principle of 
 coexistence. The principle of succession is that " no change principle of 
 can occur in substances unless they are connected with others, ' 
 in such a way that their reciprocal dependence determines a 
 mutual change of state." This is proved by the consideration 
 that a substance existing by itself can only have internal states 
 which are determined for it by reasons that flow from its own 
 nature. But this nature is, ex hypothesi, simple and incapable 
 1 R. I. 19 ; H. I. 380. ^ R. I. 31 ; H. I. 389. 
 
 succession. 
 
110 INTRODUCTION. chap. 
 
 of being changed by'itself . It follows that all reasons or causes 
 of change must come from without, from other substances with 
 which it is brought into connexion. If therefore a commercium 
 of substances is established, it cannot be accounted for by their 
 bare existence as individual substances. On the contrary such 
 individuality would rather imply that each of them should have 
 an existence which is completely intelligible apart from the 
 coexStence. othcrs. Hcncc wc have the second ' principle of coexistence,' 
 that "finite substances by their mere coexistence are not de- 
 termined as having any relation to each other, and are held 
 together in no mutual commercium except so far as they are 
 maintained in it by a common principle, i.e., by the divine 
 intelligence." " The schema of the divine intelligence in which 
 all creatures take their origin is a continuous act commonly 
 called preservation : and if in that act there were any substances 
 which were conceived by God as solitary and without any 
 relation between their determinations, no connexion or mutual 
 reciprocity could ever arise between such substances ; if, 
 however, substances be conceived relatively in the divine 
 intelligence, then, conformably to this idea, their determination 
 will always thereafter have a mutual reference during all the 
 continuance of their existence, i.e., they will act and react, and 
 each of them will have an external state, which could not be 
 accounted for by their existence alone apart from the uniting 
 principle."^ Kant goes on to say that space and position are 
 names for relations of substances which have mutual com- 
 mercium through their common dependence on God, relations 
 which would otherwise have no existence. Hence we may 
 gather that the Newtonian attraction, or universal gravitation 
 of matter, is " probably the effect of the same nexus of sub- 
 stances whereby they determine space." In any case this view 
 enables us to mediate between the theories of pre-established 
 harmony and occasional causes on one side, and the crude idea 
 of an influxus physicus on the other ; for while, according to 
 '■ E. I. 41 ; H. I. 397. 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 1 1 1 
 
 the view above stated, we maintain that there is a real action 
 and reaction of substances, we yet hold that that action and 
 reaction cannot be explained by the nature of individual 
 substances as such, apart from. their dependence upon the one 
 absolute principle, which at once gives them existence and binds 
 them together in one world. 
 
 In the treatise iust analysed we see the compromise at which compromise 
 
 ^ 'J r between Iiidi- 
 
 Kant had arrived at the beginning of his philosophical career. Thelstlcunity* 
 It is not accurate to say that . in it Kant shows himself a 
 thoroughgoing supporter of the dogmatic Eationalism of Wolff, 
 any more than to say that he has distinctly broken away from 
 it. There is as yet no trace of acquaintance with those ideas 
 of Locke and Hume, which were afterwards to infliaence him so 
 powerfully. But the study of Newton has introduced a new 
 principle which is already transforming his philosophy from 
 within. Thus Kant assumes as an axiomatic principle that the 
 world is a collection of simple individual substances with a 
 nature of their own which is determined purely from within ; 
 but he immediately modifies his view by the conception thab 
 that which manifests itself in the phenomenal relations of the 
 substances is not their inner nature but only their changing 
 states which are determined by the action and reaction going 
 on between them. Yet this external influence of the substances 
 upon each other is asserted to be not merely accidental but due 
 to the same creative act which gave them existence. Thus, in 
 conformity with Leibnizian principles, there is a certain ideal 
 priority given to the individuality of the substances as com- 
 pared with their relativity ; and even when that relativity is 
 brought in, as in the Monadologia Physica, the repulsive force 
 is treated as prior to the attractive, i.e., the negative relation of 
 substances to the positive.^ But as in that treatise it is argued 
 that both forces are necessary to the constitution of matter, so 
 here the unity of the monads as elements in the same world is 
 
 ^It is: noticeable that this priority of the repulsive force maintains itself in 
 *Kant's fjhought to the last, i.e., in the Metaphysical Rudiments of Physics. 
 
112 INTRODUCTION. chap. 
 
 supposed to be communicated to them along with their exist- 
 ence as individuals.^ Indeed, although Kant speaks of an inner 
 state of the monads determined for them apart from their 
 relations, he neither here nor in the Monadologia Physica makes 
 any attempt to define it. At a later period, in the Critique of 
 Pure Reason, we find him saying that, if monads have an inner 
 state of being, it must be that which Leibniz attributed to 
 them, a status representativus : for thought or consciousness is 
 the only conceivable inner state which a substance can have.^ 
 Here, however, Kant says nothing of the Leibnizian theory, and 
 the inner state or independent being of the monads apart from 
 all external relations remains an empty abstraction, which merely 
 indicates Kant's individualistic starting point. But this ab- 
 straction leads to a further result. Just because Kant con- 
 ceives the individuality or independence of the substances and 
 their relativity or dependence upon each other as separate 
 moments, which have no necessary relation to each other, he 
 needs a tertium quid to mediate between them. The dev^ ex 
 machina has to be introduced, since otherwise it would be im- 
 possible to conceive of the individual things or monads as, 
 notwithstanding their individuality, acting upon each other, 
 or even as contained in one space : for as contained in space 
 there already is a relation between them which is not implied 
 in their mere existence. God is needed to account for the 
 unity of the world. " Since therefore," Kant argues, " the 
 mere existence of substances is . plainly insufficient to account 
 for their mutual connexion and reciprocal influence, and an 
 external nexus of things independent implies a common prin- 
 ciple through which their existence is determined with relation 
 to each other : and since, without such a common principle, no 
 general bond of union can be conceived, we have here a most 
 
 ^ E. I. 44 ; H. I. 399. Est realis mbstantiarum in se invicem facta actio, sen 
 commercium per causas vcro efficientes, cpwniam idem quod existentiam reram 
 stabilit principium, ipsas liuic legi alligaias exhibet. 
 
 2 A. 274; B. 330. 
 
IV. PRE-CEITICAL WRITINGS OP KANT. 113 
 
 evident proof of the existence of God, and, indeed, of the ex- 
 istence of one God — a proof which is far more convincing 
 than the common argument ex contingentia mundi."^ 
 
 In this theory there is a curious inversion of the point of impf'ect 
 
 •^ ^ developmeut 
 
 view maintained by Leibniz and Wolff, though the lingering ?he reiatMty 
 influence of their doctrines is still traceable. Leibniz had argued ° 
 that, because substances are individual there can be no real 
 relations between them. Kant argues that, unless the sub- 
 stances be in real relation to each other, this simple individuality 
 would make their changing states incomprehensible. At the 
 same time, their relations are explained not as necessary pro- 
 perties but as separable accidents of them, and God has to take 
 the place of a synthetic principle to bind together substances 
 which are not necessarily connected. Since, however, the rela- 
 tions of the substances are represented by Kant as real and 
 not merely ideal, and since the substances can manifest their 
 nature only in those relations, the opposition of their individu- 
 ality to their relativity is on the point of disappearing, and 
 with it of course must disappear the externality of the principle 
 that unites them. For, if the difference of the substances be 
 merely a relative difference, i.e., a difference of elements which 
 are n\Dthing apart from their relations to each other, the binding 
 principle cannot be regarded as an external link of connexion, 
 but must be taken simply as the unity which underlies the 
 differences of the substances, and which manifests itself in their 
 action and reaction upon each other. 
 
 Such a conception is not of course furnished to us by Kant, Kant'srecuv- 
 
 ■*- .J ' rence to the 
 
 but he makes a step towards it when he treats God as not fS ofood. 
 
 merely an external power who arranges the monads and binds 
 
 them to each other, but as the real being who is presupposed in 
 
 all possibility, i.e., the positive or affirmative unity which is the 
 
 basis or presupposition of all determination and distinction. 
 
 For it is to be noticed that, while Kant rejects the ontological 
 
 argument in the form in which it was stated by Wolff, 
 
 » iR. I. 42; H. I. 398. 
 
 VOL. I. H 
 
114 
 
 INTRODUCTION. 
 
 CHAP. 
 
 Importance of 
 this change. 
 
 he restores it in another form, i.e., in the form which was 
 characteristic of Descartes and Spinoza. "We may not, he 
 thinks, argue that the being whom we conceive as the omnitudo 
 realitatis necessarily exists, on the ground that if he did not 
 exist, existence, which is a species of reality, would be lacking 
 to the idea of him ; for in such an inference we make a leap 
 from the ideal to the real. To such logic, it is sufficient to 
 answer that " if all realities without exception are united in 
 any being, that being must exist : but if they are only con- 
 ceived as united, its existence will only be an existence for 
 thought." 1 Thus it appears that that which exists for thought 
 is only a possibility, and something more is required for 
 actuality. But Kant immediately turns round and asserts that 
 there is one actuality which is not something added to possi- 
 bility or thought, but the presupposition of it. This is " the 
 material of possibility," the positive being which we assume 
 as the basis for further determination, and which we actually 
 determine whenever we assert any one thing to the exclusion of 
 another. 
 
 Is there any value in this alteration of the form of the ontolo- 
 gical argument, or has Kant merely changed the foot upon 
 which his dogmatism is standing ? The answer is that it is 
 one thing to argue from the finite determined as such to the 
 infinite, and another thing to say that the finite cannot be con- 
 ceived except through, and in relation to, the infinite. Or, to 
 put it in another way, it is one thing to say that from my 
 thought I can argue to a reality which is supposed to be ex- 
 ternal to my thought and independent of it, and another thing 
 to say that the distinction between my thought and reality 
 cannot be made except by a consciousness which in a sense 
 embraces both. Hence it was an obvious case of reasoning in 
 a circle, when Wolff, who held that the inner life of the 
 " thinking substance " was simply a consciousness of his own 
 ideas, and that these ideas corresponded to objects only by the 
 1 R. I. 14 ; H. I. 375. 
 
IV. PRE-CKITICAL WRITINGS OF KANT. 115 
 
 external arrangement of a pre-established harmony, attempted 
 to prove from his own ideas the existence of the Being who 
 made this arrangement. For how could the monad, which ex 
 hy^othesi is confined to itself, prove' from its own ideas the 
 existence of a Being who connects these ideas with external 
 realities ? If, however, the "consciousness of our own ideas as (, 
 such.is possible only through the consciousness of objects which ) 
 are not our ideas, then we shall have a right to say that the 
 consciousness of self in distinction from the not-self presupposes 
 a unity beyond that distinction. And this is the real meaning 
 of Kant's assertion that the existence of God is presupposed in 
 consciousness of self. i << 
 
 So far, then, we may recognise that Kant's change of the ^°^°i "^^1 
 form of the ontological argument has a real and important spmomsm. 
 meaning. Unfortunately, however, Kant's way of working out 
 the true idea of the priority of the unity of being and thought, 
 of objeqt and subject, to their difference is simply the old way 
 of Descartes and Spinoza, which involves the negation of that 
 difference and the dissolving of all finite being in the infinite. 
 This is implied by Kant in the treatise we are examining, when 
 he says that all limited existences are ipso facto contingent, 
 and it is more directly expressed in his Essay upon Optimism, 
 published four years afterwards. In that essay he adopts the 
 argument of the Leibnizian Theodicy, that all positive reality is 
 combined in God, and that a world of finite beings, if it was to 
 exist at all, must have more or less negation mingled in each of 
 its parts, and, therefore, must be vexed with conflict and evil. 
 On this view the finite does, indeed, imply the infinite, but it has 
 no ground for its existence in the infinite : in other words, the 
 affirmative is presupposed hj the negative, but the negative 
 does not in any sense flow from the affirmative. But, then, 
 the principle " determination is negation " must be taken in 
 the Spinozistic sense that the only real is the indeterminate: 
 and the finite must be regarded as an illusory semblance. 
 
 ° "' Summary o£ 
 
 ft The result, then, of Kant's first treatise on metaphysic is to ^^„^^f """*'"'' 
 
116 INTRODUCTION. CHAP. 
 
 transform the "Wolffian philosophy by the assertion of real 
 relations between substances, which are therefore regarded 
 not as purely self-determined monads but as individual 
 objects in one world, the changing phenomena of which are 
 determined by their reciprocal action and reaction upon each 
 other. Further, this action and reaction is carried back to 
 certain repulsive and attractive forces, which belong to the 
 substances of which the world is composed and determine 
 their relations to each other. Finally, each of these substances 
 is taken as in itself individual and exclusive, but at the same 
 time as presupposing an absolute or universal substance in 
 relation to which their exclusive individuality disappears. 
 Thus Kant revives the old dilemma of Leibniz, that it is im- 
 possible to be in earnest with the reality of the individual 
 substances without denying their unity and the reality of the 
 universal substance, or to be in earnest with the reality of the 
 universal substance without reducing the individual to some- 
 thing illusory and unreal. 
 rejeotion"onhe The Special problem which naturally became prominent in 
 philosophy. Kant's thought as he reflected upon this result was the problem 
 of the connexion of thought and reality; and this problem 
 itself opened up into two subordinate questions — the question 
 whether, and how far, thought can go beyond itself to assert the 
 reality of its object, and the question whether, and how far, the 
 connexion of thought corresponds to, or differs from, the con- 
 nexion of objects. Accordingly, it is with these questions that 
 we find Kant engaged in a series of treatises published during the 
 years 1762-3.1 These are : — a short paper on The False Sultilty 
 of the four Syllogistic Figures; a longer treatise on The Sole 
 Ground for a Demonstration of the Being of God; an essay, 
 written for a prize offered by the Berlin Academy, On the Evi- 
 dence of the Principles of Natural Theology and Morals ; and 
 
 ' It has been proved by Dr. Benuo Erdmann that the four treatises were 
 published in the order stated in the text. (Reflexionen Kant's, Introduotion 
 to vol. II. p. 17 seq.) 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 117 
 
 finally, an Attempt to Introduce the Conception of Negative 
 Quantity into Philosophy. This series of writings, to which 
 may be added the Breams of a Ghost-Seer illustrated hy the 
 Breams of Metaphysic, published three years afterwards, naark a 
 decided advance upon those we have already examined. They 
 show that Kant had almost entirely shaken off the yoke of 
 "Wolffian philosophy, and the last mentioned of them might be 
 regarded even as a kind of declaration of war against it. They 
 prove also that he was gradually approximating to the point of 
 view of English Empiricism as represented by Locke, if he did 
 not fully accept it. Hence it has been proposed to call this 
 period of Kant's development the epoch of Critical Empiri- 
 cism, in distinction from the earlier period of Eationalistic 
 Dogmatism. If, however, we have given a true account of the 
 results which Kant had reached during that first period, no 
 such marked line can be drawn between them. The despairing 
 renunciation of Eationalism, which shows itself in the 
 " Dreams," is only the final result of a course of investigation 
 which is already begun in the Bilucidatio Nova ; and the inter- 
 vening treatises enable us to connect the latter with the former 
 almost without a break. 
 
 The essay on The False Subtilty of the Syllogistic Figures need S a"SgM ™* 
 not detain us long. It takes the first step towards a discussion analytic, 
 of the opposition between thought and reality by pointing out 
 the limits of the movement of thought, according to the idea of 
 that movement which was accepted by the Wolffian school, — 
 an idea, we may add, which was never questioned by Kant 
 himself, at least as regards the movement of pure thought. All 
 the syllogistic processes are, he argues, reducible to analysis, 
 though this is somewhat concealed by the artificial complexity 
 of the theory of the logical figures. The first figure is the only 
 one which fully expresses all that it implies, the only one, 
 therefore, which exhibits the process of thought in its simpli- 
 city; for the evidence of the other figures rests on the possi- 
 Ijjlity of reducing them to the first by the conversion of one or 
 
118 INTRODUCTION. chap. 
 
 both of their premises. But this conversion should be regarded 
 as a separate inference and not introduced to complicate the 
 syllogistic forms. The sole principles of syllogism are these 
 two : — Nota notae est nota rei ipsius, and Repugnans notae 
 repugnat rei ipsi : and these are themselves only corollaries 
 of the laws of identity and contradiction, which are the prin- 
 ciples of affirmative and negative judgment respectively. Hence 
 there is no ground for saying that reason, i.e., the faculty of 
 reasoning, is different from understanding, the faculty of judging. 
 Syllogism is just the activity of thought whereby a judgment is 
 made complete, as judgment is the activity of thought whereby 
 a conception is made distinct. The higher faculty of knowledge, 
 which is characteristic of man, may, therefore, be described 
 comprehensively as a faculty of judging or making his ideas 
 distinct. The animals have no distinct conceptions or, in other 
 words, make no judgments, and in this lies their inferiority. It 
 has, indeed, been contended that the ox has a clear idea of the 
 door of its stall, and can, therefore, use that idea as a means 
 of giving distinctness to its conception of the stall : just as a 
 man selects a particular attribute of an object, and makes it 
 the predicate of a judgment concerning that object. But this 
 is not the case. For it is one thing to have before the mind a 
 clear image of that which is the attribute or mark of an object, 
 and it is quite another thing to recognise it as the mark of that 
 object, i.e., to separate the two ideas and refer them to each 
 other, as in the judgment, "this door belongs to this stall." 
 And the same may be said of distinction which has just 
 been said of relation. It is one thing to distinguish objects, 
 and it is quite another thing to know the distinction of objects. 
 The latter is logical distinction and means the recognition that 
 A is not B; i.e., it implies a negative judgment. The former is 
 merely physical distinction, and it implies only that the animal 
 that makes it is moved to different acts by different ideas. 
 " We shall, therefore, be able to determine what constitutes the 
 distinction between rational and irrational beings, if we are 
 
IV. PEB-CRITICAL WRITINGS OF KANT. 119 
 
 able to determine what the secret power is which makes judg- 
 ment possible to us. My present opinion is that this power or 
 capacity is nothing but the faculty of inner sense, i.e., the 
 faculty which enables ns to make our own ideas the object of 
 our thoughts." ^ 
 
 It would seem as if Kant had already touched upon one of ne^o"ssity1f 
 the main ideas used in his Transcendental Deduction of the Sonstrabit 
 
 . - principles. 
 
 Categories, when he recognised that the faculty of judgment is 
 at the same time the faculty which enables us " to make our 
 own ideas the objects of our thoughts," or to refer our thoughts 
 to objects. But a passing suggestion is not a discovery : and 
 though Kant speaks of the activity of judgment as that which 
 turns a series of like, or unlike, presentations of sense into a 
 knowledge of objects, it does not occur to him to suspect that 
 there is anything more involved in that activity than mere 
 analysis. On the other hand, he sees clearly that, if the 
 operation of thought is merely analytic, it is absurd to suppose 
 that by means of it we can develop all truth out of one funda- 
 mental principle. " Those philosophers, who proceed as if there 
 were no fundamental truths which are beyond demonstration 
 except one, are as far from the truth as those who are ready to 
 assume indemonstrable principles without sufficient warrant. 
 Human knowledge is full of indemonstrable judgments." But 
 if this be the case, then the complex ideas expressed in such 
 judgments must, it would seem, be given independently of the 
 activity of thought. 
 
 Having thus reduced the process of thought to analysis, betocTud'Idin 
 Kant goes on to enforce the lesson that it is impossible by that of aroi^e^tT" 
 process to bridge over the gulf between thought and reality. 
 This is the leading idea of his next essay on the Onli/ Possible 
 Basis for, a Proof of the Being of God, in which, however, 
 he does not, except in one point afterwards to be mentioned, 
 get much beyond the ideas of the Bilucidatio Nova. As 
 in the latter treatise, he urges that the analysis of our ideas 
 ' K. I. 73 ; H. II. 68. 
 
120 INTRODUCTION. CHAP. 
 
 can never assure us of the reality of their objects. Being, or 
 Existence, is no part of any conception, whicli may be ex- 
 Lracted from it by analysis and used as its predicate. " Take 
 any subject you please, say Julius Caesar. Gather out of the 
 conception of him all his thinkable predicates, those of time 
 and place not excepted, and you will soon understand that, 
 though you have given him all these attributes, you have still 
 left it undetermined, whether he exists or not." "In the com- 
 plete determination of the possibility of a thing, no predicate 
 can be left out which it would have, if it existed." " Where in 
 our ordinary manner of speech, existence appears as a predicate, 
 we must understand it not as a predicate of the thing itself, but 
 of the thought of the thing. When, e.g., we say that the sea- 
 unicorn (or narwhal) exists, but not the land-unicorn, this 
 means only that the former is an empirical conception, a con- 
 I ception of a thing that actually exists. In order, therefore, to 
 
 show the truth of such a proposition, we have not to examine 
 what is contained in the conception of the subject (in which we 
 could find only predicates of its possibility), but we have to 
 inquire into the origin of the conception. The question, in 
 short, is whetB^we have seen the narwhal, or heard about it 
 from those who have seen it. If, then, we were studying per- 
 fect accuracy of language, it would be better for us to say — 
 not, ' The sea-unicorn is an existent animal,' but, ' An existent 
 marine animal has all the predicates which I unite in the con- 
 ception of a unicorn.' " i Hence also, the Cartesian proof of the 
 Being of God, which is based on the inclusion of being in the 
 conception of God, must be rejected. 
 fentfoaS ^^i^S ^^ ^ ^^pl^ idea, which we cannot explain by dissecting 
 poiuwn. j|. .j^^^ g^jj^ simpler ideas ; but we may help towards a clearer 
 understanding of it by means of the closely related idea of 
 position. When used as the copula, the verb of being implies 
 the relative position or assertion of something, i.e., the position 
 or assertion of a predicate in relation to a subject. Nothin" is 
 iR. I. 171-2; H. 11. 115-117. 
 
IV. PRE-CEITICAL WRITINGS OF KANT. 121 
 
 herein asserted as to the reality of the subject itself, which may 
 
 be entirely a creation of imagination. But when the same 
 
 verb is used as a predicate, it means the absolute position of the 
 
 thing, i.e., its assertion simply in relation to itself. Or, to put 
 
 it otherwise. Being does not supply a predicate for any subject, 
 
 but rather a subject to which predicates may be attached. 
 
 When, therefore, we ask for a proof of the being of God, we are 
 
 not asking for a middle term by means of which we may show 
 
 that the predicate of being is attached to God as a subject, but 
 
 we are asking for a proof that something existent has all the 
 
 predicates which are gathered into the idea of God. 
 
 From these considerations it follows that all the usual Kant's ex- 
 planation of 
 
 explanations of the idea of being or existence must be rejected „{ re^f^^a^ " 
 as inadequate, or even positively misleading. The definition po^^'^'^'^y- 
 given by "Wolff, that it is the " complement of possibility," is 
 inadequate ; for it does not tell us what is to be added 
 to possibility in order to constitute reality. Baumgarten, one 
 of Wolffs most eminent followers, explained that the possibility 
 of a thing was its logical essence or definition, with all the 
 properties deducible therefrom : while the " complement " 
 necessary for reality consisted of all the accidental predicates 
 which come to be added to it through its connexion with other 
 things in the world. But this is erroneous, for we may deter- 
 mine all the predicates of an object, whether necessary or 
 accidental, without passing beyond its possibility. Nay, we 
 may even say that, in order to be possible, an object must be 
 completely and individually determined; e.g., every possible " 
 man must be of a certain age, stature, figure, etc., and a man 
 in whom any of these points are undetermined is impossible. 
 Equally erroneous is the view of Crusius, that the determina- 
 tions of time and place make the difference between possibility 
 and reality. For, not to mention that this would involve the 
 questionable principle that all that exists is limited by 
 conditions of time and space, it is obvious that we can think of 
 possible, as well as of actual, time and place. The true 
 
122 INTRODUCTION. chap. 
 
 distinction of the real from the possible is found, not in the 
 matter or content of our conceptions, but in the way in which 
 ithey have been formed. In the case of all things but God, 
 (Kant in this essay maintains that the real is that which is 
 given in experience and the possible is that which is not 
 so given but merely thought. In the case of God, on the 
 other hand, he repeats and illustrates at great length the same 
 proof which we have already found in the Bilucidatio Nova, 
 that God is the ultimate reality implied in all possibility. 
 The only new point is the contrast which he draws between 
 the real ground of all possibility, and the formal or logical 
 ground of it which is found in the principles of identity and 
 contradiction. 
 Criticism The assertion that reality or existence is not a true 
 
 of that 
 
 explanation, predicate, and that a possible object may have all the 
 attributes of a real one without being real, is one that will 
 meet us again in the Critique of Pure Beason. It raises the 
 whole question of the relation of thought and reality, and 
 of the netture of the opposition between them. Here it is only 
 necessary to point out that when Kant interprets existence by 
 "position," he admits, though without seeing the effect of 
 his admission, that existence must mean existence for a self, 
 or as posited by a self. For if this be admitted, the possible 
 and the actual cannot be absolutely opposed, as that which 
 is posited by thought to that which is given independently of 
 its activity. And as the position of an object for a self can 
 only mean the determination of it in relation to the other objects, 
 which constitute the one objective world that exists for that 
 self, so absolute position cannot be regarded as essentially 
 different from relative position. It is the lingering associa- 
 tions of the individualistic point of view with which he 
 started which makes Kant separate so absolutely between 
 what the thing is and what it is for us, and again between 
 what it is in itself and what it is in relation to other things. 
 And the same almost ineradicable influence of his original 
 
IV. PRE-CBITIOAL WRITINGS OF KANT. 123 
 
 point of view also leads him to suppose the possibility of a 
 complete, analytic determination of the conception of a thing, 
 which yet leaves it open to us to consider whether or not it 
 exists ; and to forget that every step in such determination 
 is the assertion of a connexion of the thing with other things 
 in the one intelligible world. The complete explanation of the 
 possibility of a thing is, therefore, the same thing with the 
 proof of its reality. The possible which can be opposed to the 
 real is simply that which is determined for us only by a few 
 relations, and which for that reason remains hypothetical, 
 or merely possible. Hence, if our whole present consciousness 
 of the world may be regarded as in a sense ideal or merely 
 possible, it is only in so far as we recognise that that con- 
 sciousness is in process of development and not because, simply 
 as ideal, it is not real. So much it is necessary to say at 
 present, though the full discussion of the subject must be 
 reserved till we come to deal with the Critique of Pure Reason, 
 in which the same opposition appears in a modified form. 
 
 It has already been pointed out in relation to the Bilucidatio Kantthmksof 
 
 God as the 
 
 Nova that Kant himself was not altogether blind to the truth unityofreaiity 
 
 " and possi- 
 
 which is the necessary complement of his dualistic view of the and Though 
 reilation of thought to being. Such a complementary idea 
 appears, both there and again in this treatise, in Kant's version 
 of the argument for the Being of God. For when Kant argues 
 that there is an actuality implied in all possibility, a con- 
 sciousness of a primary reality implied in the conception 
 ' of anything as possible, he is putting a true thought into a 
 somewhat obscure and inaccurate form. We cannot oppose the 
 ideal to the real, that which is for us to that which is in itself, 
 unless we assume that that which is in itself is also for us. We 
 cannot be conscious of our ideas as our own in opposition to 
 objects, except on the presupposition of an idea which is also 
 objective. All such distinction supposes a relation of the 
 elements opposed and therefore a unity beyond the distinction. 
 In this sense, therefore, it is no mere " dogmatic slumber " that 
 
124 INTRODUCTION. C'i^- 
 
 makes Kant revive the outological argument. Unfortunately, 
 however, in reviving it, he does not yet improve upon that 
 form which it had with Descartes and Spinoza. Like them, he 
 reaches the unity beyond difference merely by abstraction from 
 the differences, and therefore he conceives- that unity merely 
 as a " material of possibility," i.e., an affirmative reality which 
 we limit or determine by negation, when we think of any 
 particular object as possible. Now, as we have seen, the 
 legitimate result of such a way of thinking is to merge all 
 finite reality in the absolute substance. In other words, 
 Kant has as yet found no escape from the dualism that abso- 
 lutely separates thought and its object, except in a unity 
 in which all distinction is entirely lost. At the same time he 
 is so far from seeing that this is the necessary result of his 
 logic, that he goes on to argue that the absolute Being must be 
 conceived as a spirit endowed with absolute power, absolute 
 wisdom, and absolute goodness. In fact, having once got his 
 absolute reality, he drapes it in all the '' old clothes " of 
 orthodoxy, without any consciousness of the opposition between 
 the Spinozistic and the Christian idea of God. 
 theide/tiat'* While, howcver, this is true, it is not the whole truth. 
 
 God unites in . ^ ., . . . . , ^^ 
 
 Himself au Already, m the treatise we are considering, Kant takes one 
 
 affirmative 
 
 predicates, important step towards the correction of the Spinozistic idea 
 of the ultimate unity of Being and Knowing, when he 
 lays down the principle, that though God is to be conceived 
 as the ens realissimum, who is at once the highest reality 
 and the ground of all other reality, he is not to be con- 
 ceived as embracing all reality in himself. For there are, 
 Kant argues, realities which it is impossible to combine in 
 one subject, e.g., extension and thought. " It is vain to try 
 to evade this conclusion by saying that the qualities excluded 
 are not true realities. The impulse of a moving body and 
 the force which it holds together are undoubtedly positive: 
 nor can the feeling of pain be regarded as a mere privation. 
 The attempt to explain away the opposition of positive 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 125 
 
 realities is due to a confusion. It is contended that two 
 realities cannot be opposed to each other because, as they are 
 both true affirmatives, they cannot be inconsistent as predi- 
 cates of one subject. But though I admit that there is no 
 logical contradiction between two positives, this does not 
 exclude a real repugnance or opposition between them. Such 
 repugnance on the contrary is actually found wherever one 
 thing, regarded as a cause, annihilates the effect of something 
 else. Thus the moving force of a body in one direction and an 
 equal tendency to move in the other direction are not in con- 
 tradiction : nay, they may actually exist in the same body at 
 the same time. But the one annihilates the real effect of the 
 other ; and whereas each severally would have produced an 
 actual motion, the effect of both together is zero : in other 
 words, the consequence of two opposite moving forces is rest. 
 But this shows that real opposition is different from logical 
 contradiction ; for nothing is possible which implies a contra- 
 diction. Now, in the most real of beings there can be no such 
 real repugnance or opposition of attributes, the consequence of 
 which would be privation or want ; and as there would be such 
 an opposition in Him if all realities were brought together in 
 Him as attributes, we are forced to conclude that, though all 
 realities must be referred to Him as their cause, some of them 
 must be regarded only as effects of his action and not as 
 attributes of his being." ^ 
 
 What is here suggested, it will be observed, is not a concep- Distinction of 
 
 logical and 
 
 tion of the ' polarity of opposites, or of the necessary relation of real negation, 
 affirmation and negation. All that Kant says is that positives 
 may be so opposed as to produce a negative result. In other 
 words, he does not deny the possibility of a negation, which is 
 nothing but negation, but he maintains that that is not the 
 only form of negation. But to one with Kant's generalising 
 tendency, the assertion that there is a negation which implies 
 position or affirmation could not be long separated from the 
 • 1 E. I. 189 ; H. II. 129. 
 
126 INTRODUCTION. chap. 
 
 assertion that there is a position or affirmation which implies 
 negation. By the side of logical affirmation and negation, 
 which are supposed absolutely to exclude each other, Kant 
 is, therefore, led to erect two new species, the position which 
 implies negation, and the negation which implies position. Or 
 to put it in a point of view which shows more clearly the 
 effect of this great advance of thought, alongside of the analytic 
 process of thought which moves by identity and contradiction, 
 and which, therefore, cannot by any possibility pass beyond the 
 limits of the conception with which it starts, Kant now places 
 a synthetic movement which proceeds from the assertion of one 
 thing to the negation of another, and through that to the 
 assertion of a third. In tliis way Kant evolves the great dis- 
 tinction of analytic and synthetic judgment, which has such 
 J important effects in his subsequent philosophy. 
 
 What is veal I* is in the Essay towards the Introduction of the Idea of 
 nega ion . jsfQg/^n^g Quantity into Philosophy that Kant for the first time 
 definitely expresses this important distinction — a conception 
 which probably, as Paulsen suggests, arose in his mind in 
 connexion with his difficulty about the idea of God, but when 
 once suggested, it found ready support and illustration in that 
 view of matter as the unity of opposite forces to which he had 
 already been led by Kewton. Kant begins by pointing out that, 
 according to the law of logical opposition, the opposite of any 
 predicate is its mere negative, but that in the mathematical 
 determination of quantity, the opposite of + A is —A, and pure 
 negation is only reached through their union, so that + A - A = 0. 
 In the former sense, a union of opposites. is an impossibility {nihil 
 negativum), in the latter it is a simple privation or zero {nihil 
 privativum). We should, therefore, be ■ careful to distinguish 
 between a negative quantity and the negation of quantity : for 
 while the latter removes quantity and puts nothing in its 
 place, the former is simply the opposite of the quantity that is 
 taken as positive. Hence it is often indifferent which of two 
 quantities we call negative and which we call positive : or 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 127 
 
 rather, we should say that both taken abstractly are positive 
 but that, when put together, they constitute a real opposition 
 in which either member may be taken as positive and the 
 other as negative. Thus impenetrability (= repulsion) may 
 be described as negative attraction, or attraction as negative 
 impenetrability, the one force being as positive as the other. 
 Or to take an example from mind, attention may be repre- 
 sented as negative abstraction or abstraction as negative 
 attention, since it requires the same positive effort to exclude 
 from the mind everything but one object and to concentrate it. 
 upon that object : indeed the one is but the necessary correla- 
 tive of the other. 
 
 These considerations open up a new vein of reflexion, for Eeai position 
 
 = negation of 
 
 they suggest that the world may he regarded as the theatre of -'n opposite 
 a conflict of opposite forces, in which the absence of a mani- 
 festation of activity in any one direction is the indication, not 
 of the absence of any tendency to act, but of the equipoise of 
 opposite tendencies. Thus every piece of matter, the elements 
 of which are at rest as regards each other, is a coherent unity 
 only as the result of the equilibrium of the forces by which its 
 elements repel and attract each other ; and in our moral 
 experience inactivity is, or may be, the result of a tension 
 between the consciousness of duty and the force of inclination. 
 Conversely, when activity in any one direction begins, we are 
 often obliged to recognise, not simply that one movement comes 
 in the place of another that has ceased, but that the force 
 which was previously keeping it back has been neutralised by 
 an opposite force. Thus when a resting body begins to move, ■ 
 
 it is because the equilibrium of forces acting upon it is dis- 
 turbed, and its so-called inertia is overcome by a greater force. 
 And, in like manner, when I begin to think of anything, it is 
 not merely that other objects cease to occupy me, but that they 
 are driven out by the greater power of the new object over the 
 mind. So one desire does not simply come into the- mind in 
 ;^oom of another, but the former yields because it is driven 
 
128 INTRODUCTION. chap. 
 
 out by what has been called " the expulsive power of a new 
 affection." 
 The world as These thoughts Kant applies not onlv to the different forces 
 
 a unity of o ^ ^ *' 
 
 opposites. acting upon an individual object, but to all forces actual and 
 potential in the world as a whole. For as every manifestation 
 of force may be regarded as the locking up of another equal and 
 opposite force (so that to make one force actual is to make an 
 equal and opposite force potential), it follows that, as Leibniz has 
 maintained, the sum of forces — meaning by that the sum of 
 actual and potential forces — is constant. And indeed that 
 sum, according to the same principles, is always zero, if 
 opposite forces be taken .from each other. Hence also, if we 
 conceive the Supreme Being as a Being in whom there is 
 no real opposition of elements, and therefore no such law 
 of compensatory reaction, we must suppose the nature of his 
 activity to be altogether different from any activity to be seen 
 in the natural world either outward or inward. 
 
 movlmentof From this vicw of the real opposition Kant finally derives 
 
 Thought and . j , • ■ n t • t , 
 
 synthetic an important prmciple, which he extends not only to the real 
 
 movement of ** 
 
 Knowledge, negation which is the result of an equilibrium of opposites, but 
 also to real position which is the result of the disturbing 
 of such equilibrium. This principle is that such negation and 
 such position are altogether different from all merely logical 
 negations or positions, which are simply the analytic develop- 
 ment of given premises. In the latter we cannot move from 
 positive to negative or from negative to positive; nor again 
 can we move from one position to another which is different 
 from it. In the former on the contrary we can and must 
 make such a transition. There is, therefore, a marked contrast 
 between the logical relations of ideas and the real relations of 
 things, and we must go beyond the former in order to under- 
 stand the latter. " I understand, e.g., how, when I assert that 
 God is infinite, I am forced to deny that he is mortal : for his 
 mortality would contradict his infinity. But how it is that by 
 the motion of one body the motion of another is stopped. 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 129 
 
 is quite another question. For we cannot say that the motion 
 of one body is the contradictory or logical negation of the 
 motion of another." And there is a parallel difficulty in the 
 case of logical and real position. If I look at things in their 
 logical aspect, I see that by the mere analysis of conceptions, 
 "I can find, e.g., in. composition a ground for the assertion of 
 divisibility, in necessity a ground for the assertion of un- 
 changeableness, in infinity a ground for the assertion of omni- 
 science, etc. In all these cases I clearly see the connexion of 
 reason and consequent, for the consequent is identical with a 
 part of the conception of the reason. But how one thing can 
 follow from another when it is not connected with it according 
 to the rule of identity, that is the point which I could very much 
 wish some one to make intelligible to me." The real problem 
 of knowledge thus escapes from the domain of Logic ; for what 
 is wanted for knowledge is not to explain how a conception i 
 remains identical with itself and repels its negative, but how, I 
 one thing being posited, the position or negation of something 
 else is the consequence. Kant ends with these words, " I have 
 carefully considered the nature of our knowledge as it is 
 expressed in judgments in relation to reasons and consequents, 
 and I shall shortly take an opportunity to communicate in 
 detail the result of my inquiries. But the sum and substance 
 of what I have to say is, that the relation of a real reason to its 
 positive or negative consequent cannot le expressed iy a judgment 
 hut only iy a conception. We may, no doubt, sometimes 
 reduce such a conception by analysis to simpler conceptions of 
 real reasons, but in the end all our knowledge of this relation 
 must terminate in simple and irreducible conceptions of real 
 reasons or causesjthe relation of which to their consequents cannot 
 be further explained."^ The meaning of this obviously is that 
 reason and consequent are given in connexion with each other, 
 and we must take them as they are given. The formal laws 
 of thought, as they show themselves in the act of judgment by 
 
 ♦., iR. I. 158-160; H. II. 103-6. 
 
 'vol. I. I 
 
130 INTRODUCTION. chap. 
 
 which we analyse the content of our conceptions, do not enable 
 us to explain any real or causal connexion of things, whether 
 negative or positive. We are obliged, therefore, to fall back 
 on an original unity of the conceptions themselves, i.e., on a 
 nexus in the matter of thought as it is given in experience, 
 which it is impossible either to expla,in or to explain 
 away, 
 following™^ III the concluding words of this essay Kant seems to 
 Reasons for a approximate so closely to the ideas and even to the language 
 
 negative 
 
 answer. of Humc, that it is difficult to regard the coincidence as merely 
 accidental. Kant seems to be insisting on the very point on 
 which Hume dwelt with such emphasis, viz., that there is no 
 link of necessary relation between the phenomena which we 
 regard as cause and effect. On consideration, however, we see 
 that, though there is a verbal parallelism, the premises from 
 which the two writers start and the conclusions to which they 
 are pointing are quite different. Hume was the interpreter of 
 a philosophy the first principle of which was that all that 
 is true in our ideas must be traced back to that which is 
 given to the passive mind, and that all merely subjective 
 additions to the facts presented must be fictitious and illusory. 
 Hence, when he had shown that in the impressions or immediate 
 experiences of the outer and the inner life there is no trace of 
 that necessary connexion of antecedent and consequent, which 
 is supposed to be involved in the idea of causality, he conceived 
 himself at once entitled to treat such necessity as an iLle<nti- 
 mate product of custom, a confusion of subjective association 
 with objective reality. The conception of a synthetic power of 
 thought was excluded ex 7iypot7iesi, and Hume, therefore, could 
 not in his Essays have any intention of refuting it. Kant, on 
 the other hand, inherited the tradition of a philosophy which 
 sought to demonstrate everything by a purely logical process, 
 ■ and so to raise all_sciences to the level of mathematics ; and 
 not only in this essay, but also in those earlier writings of 
 which we have already sppken, we have found him asking 
 
I^- ■ PRE-CRITICAL WRITINGS OF KANT. 131 
 
 himself the question how much the purely logical movement of 
 thought can do. In answer to this question he pointed out, in 
 the essay oh The False SwUilty of the Four Syllogistic Figures, 
 that demonstration guided by the laws of identity and ' contra- 
 diction can only analyse what is given, and must, therefore', 
 start with inany indemonstrable principles.. On the same 
 grounds he proved, in the essay On the only Possible Basis for 
 a Proof of the Existence of God, that pure thought cannot bridge 
 the gulf between itself, and objective reality. Finally, in the 
 essay which we are discussing, he shows that logical and real 
 opposition are essentially different, and that there is nO 
 parallelism between the movement of pure thought, in which 
 we proceed according to the law of contradiction from the 
 assertion of a predicate to the exclusion of its negative, and 
 the objective process of things in which one positive reality, 
 e.g., one force, neutralises another. It is, therefore, only 
 another step in the same direction, i.e., another step in the 
 proof that pure thought cannot by its analytic movement add 
 anything to its content, when at the end of the essay he goes 
 on to show that the causal relation, — involving as it does a 
 synthetic movement from the position of one thing to the 
 position of, a different thing, — cannot be explained by the 
 logical connexion of reason and consequent according to the 
 law of identity. The result to which Kant points in this 
 argument is just that which Hume is led by his argument to 
 reject, viz., that the causal connexion must be given in experi- 
 ence. For whereas Hume sought to prove that the causal 
 relation, because of its necessity, cannot be given in experience, 
 Kant seeks._.to_- show .that the same_ relation, because of its 
 syntiietic_oiL_.ampliative character, cannot be derived from pure 
 thought. When we take all this into consideration^ the 
 coincidence of expression becomes a matter of little moment : 
 for, even if we could suppose that Hume's words were present 
 to Kant when he wrote, — and the manner in which Kant 
 speaks seems to preclude such an idea, — we could explain 
 
answer. 
 
 132 IKTEODUCTION. CHAP. 
 
 Kant's use of them only by supposing that he did not under- 
 stand them, but read into them his own meaning. 
 positiTC ""^ " There is, however, one point that should be mentioned on the 
 other side, and that is that Kant himself did not always, even 
 in his Critical period, clearly distinguish between the two sides 
 of his argument, as against the German dogmatism and the 
 English empiricism respectively. Frequently he seems to mis- 
 understand Hume's reasoning and to suppose Hume to be 
 refuting the doctrine that reason has ampliative power, when 
 he was merely showing that no necessity of connexion is given 
 in sense. Thus, to take one passage for many, we find Kant 
 saying that Hume held " the increase of conceptions out of 
 themselves, and what may be called the spontaneous generation 
 of the understanding without being fertilised by experience, to 
 be impossible. He regarded, all supposed a 'priori principles 
 as imaginary, i.e., he supposed them to be nothing but habits 
 springing from experience and its laws, and he therefore argued 
 that they are merely contingent rules to which we falsely 
 attribute universality and necessity." ^ Here, as in many 
 other places, Kant puts together two aspects of synthesis, as 
 ampliative of a given unity of conception, and as connective of 
 given differences of perception, and argues that an a priori 
 synthesis in the former of these senses is necessary to explain 
 a necessary synthesis in the latter sense. And he attributes to 
 Hume a consciousness of the whole problem which he himself 
 sought to solve. Now, it is quite true that Hume dwelt upon 
 the fact that, in the idea of the object we regard as cause, there 
 is nothing to suggest that it has any connexion with its effect, 
 except the sequence of the latter upon the former : but as 
 I have already said, he points this out, not in order to disprove 
 the power of reason to amplify its conceptions, (which he 
 assumed to begin with,) but in order to show that there is no 
 impression from which the idea of the relation, as a necessary 
 relation, of cause to effect, can be copied. So far, therefore 
 1 A. 765 ; B. 793. ProUgomena, § 4. 
 
IV- PRE-CRITICAL WRITINGS OF KANT. 133 
 
 Kant seems to ha^■e always understood Hume in his own 
 sense ; a fact which is rather in favour of the views of those 
 who contend that Hume's influence is already traceable in the 
 Ussay on Negative Quantity. But, on the other hand, in the 
 critical period, the ampliative character of the judgment of 
 causality is always, as in the passage just quoted, immediately 
 connected with its necessary character as an a priori judgment, 
 a point to which no reference is made in the essay. 
 
 Connected with this is the fact that Kant as yet has no Kant does not 
 
 yet deal ■with 
 
 difficulty as to the general principle of causality, but only as to ^^^^il^^^ 
 the application of it in particular cases. " I am not going to °*"^*"*y- 
 let myself be put off with the words cause and effect, or with 
 the words force and action. Por if I regard a thing as a cause 
 or attribute force to it, I have already thought of it as standing 
 to its consequent in the relation of a real reason or antecedent; 
 and -then it is easy to explain my assertion of the existence of 
 the consequent according to the law of identity." Kant here 
 obviously omits to ask himself the question how a mind ruled 
 by the law of identity, as he interprets it, should ever conceive 
 the idea of such a relation as that of causality, or should 
 attribute to it objective reality. The question as to the univer- 
 sal principles presupposed in all experience had not yet been 
 suggested to him, and, therefore, he could not see the force of 
 Hume's attack upon the idea of necessary connexion, as he 
 afterwards saw it.-' 
 
 1 It will be necessary to return to this subject in the next chapter and also in 
 commenting upon the Critique. Here, however, it is necessary to refer to one 
 point on which there is some misunderstanding. It has been maintained that 
 Hume in his Essays deals merely with the relation of particular causes to their 
 effects and not with the general principle of causality, which he discusses in 
 the Treatise, — a point which is important because Kant seems to ha ve known 
 only the Essays. Now it is true that, in the Essays, Hume dwells upon the 
 fact that we do not perceive any necessary connexion between events either in 
 outer or in inner experience. "We are never able," he declares, "in any 
 single instance to discover any power or necessary connexion." But he 
 immediately infers from this that the general idea of necessary connexion is 
 objectively invalid, " seeing that all our ideas are nothing but copies of our 
 impressions." We need, therefore, to explain the idea of causality as a copy 
 
possible to 
 sense " 
 
 134 INTRODUCTIOHr. chap. 
 
 ^^^gSyUf^'^f^'^ In regard to the general problem of this essay, it may he 
 iastructive to refer by anticipation to the way in which Kant, 
 at a later period, dealt with the contrast between logical and 
 real opposition. For then he maintained that, while the con- 
 ception of real opposition cannot be derived from pure thought, 
 it can be made intelligible by means of the application of pure 
 thought to the a "priori forms of perception. But this way out 
 of his difficulty had not been suggested to him at the time of 
 the Essay on Negative Quantity. All that Kant had as yet 
 reached was the conviction that synthesis, negative or positive, 
 is impossible to pure thought. Before the ideas of the 
 Critiqice could be developed, it was necessary that Kant should 
 learn the lesson of the empirical philosophy, and especially the 
 lesson finally drawn from it by Hume, viz.. that all synthesis, 
 as a movement to integrate or connect, or at least necessarily to 
 connect, elements given as different, is impossible to isense. If 
 pure thought be confined to a movement of analysis, a move- 
 ment by mere identity, pure sense, as Kant was to learn, is 
 confined to a mere succession of isolated presentations without 
 any unity. If the former supplied no principle of . differentia- 
 tion, if it could not go beyond .the conceptions with which it 
 started so as to add to them any new elements, the latter on 
 the other hand supplies no principle of integration by which the 
 " manifold " of sense, its endless difference and variation, could 
 be brought back to unity. But both differentiation and inte- 
 gration are necessary for synthesis. The problem of Kant, as 
 it was subsequently stated, was to explain how the analytic 
 unity of thought could develop a synthetic power in relation to 
 
 of a "new sentiment or impression, to wit, the customary transition of 
 thought or imagination between one object and its usual attendant : and this 
 sentiment is the original of that idea." In other words, the idea of necessary 
 connexion is simply the generalised copy of this sentiment. Kant, on the other 
 hand, is here dealing not with the impossibility of finding the idea- of neces- 
 sary connexion in experience, but solely with the impossibility of thought 
 adding to its own content except by the aid of experience, — an impossibility 
 which he always maintained, though he found a way out of the difficulty by 
 the aid of his doctrine of the a priori forms of sense. 
 
IV. PRE-CRITICAL WRITINGS OP KANT. 135 
 
 the difference of sense, and how the differences of sense could 
 be correlated under the unity of thought. , In the essay which 
 we are considering, however, Kant has only grasped one 
 element in the problem. He sees that thought in itself is 
 condemned to a purely analytic movement, and he casts the 
 burden of supplying not only the manifold data but their con- 
 nexion upon experience, i.e., upon perception. At the same 
 time the clear distinction between the formal principles of 
 knowledge, derived from pure thought and material principles 
 derived from experience, which he has already attained, marks 
 a very important stage in his progress towards the critical 
 point of view. He has " seen Leibniz with the eyes of Hume," 
 or at least of Locke.; it remains only that he should " see Hume 
 with the eyes of Leibniz." 
 
 In Kant's prize essay On the Evidence of the Principles 0/1'™^'^™'=^ 
 Natural Theology and Morals, which was probably finished phuosophy 
 before the essay we have just been considering, the distinction matics. 
 between material and formal principles is more fully illustrated. 
 Following Crusius, and developing the thought already ex- 
 pressed in the essay On the False Subtilty of the Syllogistic 
 Figures, Kant points out that there is only one, or, if we 
 distinguish the principles of identity and, contradiction, two 
 formal principles, but that there are many material principles 
 of knowledge, i.e., many propositions which are incapable of 
 being proved. " Crusius is, therefore, right when he blames 
 other schools of philosophy for overlooking these material prin- 
 ciples and directing all their attention to the formal principles 
 of knowledge. By the latter alone nothing can be proved, for 
 such proof would require premises containing the middle terms 
 whereby the logical relation of subject and predicate in the 
 conclusion is mediated, and in looking for such middle terms 
 we must finally come to first premises, beyond which we cannot 
 go." ^ What, however, gives special point to Kant's essay is, 
 that in it hfe for the first time deals with that distinction be- 
 • 1,11.1. 103; H. II., 303. 
 
136 INTRODUCTION. °^^- 
 
 tween the method of philosophy and the method of mathema- 
 tics, a distinction on which he afterwards had so much to say. 
 The example of mathematics has, he thinks, been a great snare 
 to previous philosophy. Mathematical science was supposed to 
 be the type of demonstrative science, and its great success Of 
 extending human knowledge had naturally led Descartes and^^^ 
 Spinoza, Leibniz and Wolff to expect that the same method 
 might be employed with equally great results in another 
 department. But as Bishop Warburton had remarked, " nothing 
 could be more fatal to philosophy than the imitation of the 
 method of mathematics." i For the business of mathematics is 
 not to analyse given conceptions, but by arbitrary synthesis to 
 produce new conceptions that are not given. The few concep- 
 tions which it presupposes, the conceptions of quantity and 
 space, it does not need to analyse. It simply takes them as 
 they are given in the common consciousness. On the other 
 hand, its own special objects are produced by the very act of 
 mind that defines them : i.e., we give rise to these objects by 
 putting together given conceptions in an arbitrarily determined 
 way. Hence we can be quite sure that there is nothing in 
 these constructions of our thought except what we have put 
 into their definition. Thus the definition of a circle is just a 
 prescription of the process whereby the figure, to which we 
 attach the name of a circle, is to be drawn : the circle is not 
 given as an object iefoj'e the definition, but comes into existence 
 in virtue of it. Further, when we observe the method of 
 mathematics, we find that, though the truths it establishes are 
 universal, they are always demonstrated in an individual con- 
 crete instance, which is brought before the eyes or at least 
 before the imagination. Thus to prove that space is infinitely 
 divisible, the mathematician draws two parallel lines and one 
 cutting them at right angles ; then, from a point in one of the 
 parallels, he draws lines passing through the perpendicular to 
 various points in the other parallel ; and from the infinite 
 1 R. I. 88 : H. II. 291. 
 
IV. PRE-CRITICAL WRITINaS OP KANT. 137 
 
 number of possible sections of the perpendicular in this 
 one instance, he concludes to the infinite divisibility of space. 
 But in philosophy we have to proceed in quite a different way. 
 There we have not to produce an object by arbitrary synthesis, 
 but to discover what is the nature of the object as it is given; 
 not to combine elements apprehended separately, but to 
 distinguish elements which are already united. Definition, 
 therefore, is rather the ultimate end at which the philosopher 
 has to aim, than the first point from which he starts. And it 
 is far from easy for him to see things exactly as they are 
 given, or to be sure that, in defining them, he has omitted 
 nothing. Hence, most philosophers have sinned both by arbi- 
 trary combination of things which are given separately, and by 
 arbitrary separation of things which are united and perhaps 
 necessarily united. The fanciful synthesis which produced the 
 Leibnizian conception of the " slumbering monad " is not more 
 irrational than the dogmatism of other writers about the 
 possibility of a soul without a body. Again, the sign which 
 the philosopher must use to express any given conception is 
 not an individual instance in which the nature of its object is 
 fully represented, but a word, an arbitrary mark, which in 
 itself exhibits none of the characteristics of the object which 
 it designates. He is forced to deal with his conceptions 
 in abstracto, without being able, like the mathematician, to 
 " envisage " them in coTwreto, and , so to verify or correct them 
 by the manifest congruity or incongruity of the elements united 
 in them. From this it follows that philosophy cannot imitate 
 the method of mathematics. It must proceed by analysis, not 
 by synthesis ; and, indeed, it must often be content for the 
 time with -an analysis which is far from complete. What it 
 has to copy is not the method of mathematics, but the method 
 which Newton introduced with so great profit into physical 
 science. " According to his teaching it is the business of the 
 physicist to determine by sure experiences (aided, of course, 
 by geometry) the rules according to which certain natural 
 
138 INTRODUCTION. (JHAP. 
 
 phenomena take place. For, even though he may not thus be 
 able to penetrate to the primary cause of these phenomena in 
 the nature of material bodies, he can be certain that they act 
 according to the laws discovered : and it is already an explana- 
 tion of the complexity of natural events, vrhen it is clearly 
 shown that they fall under well attested rules. In like manner 
 in Metaphysic, we ought to seek by sure inner experience, i.e.) 
 by the immediate evidence of consciousness, to determine those 
 elements which are manifestly contained in the conception of the 
 properties of the things which we are examining ; and though this 
 may not reveal to us at once all that belongs to the definition 
 of such properties, yet we may safely avail ourselves of our 
 knowledge so far as it goes, as a sufficient basis for many in- 
 ferences." So far, then, is it from being the case that we can 
 proceed from a few simple notions to demonstrate everything, 
 that we must rather allow that philosophy has many starting 
 points : i.e., in other words, that there are an' indefinite number 
 of material truths which we must take just as they are given. 
 And if there is room to hope that some day we may be able to 
 reverse our steps and to proceed as in mathematics from the 
 simple to the complex, yet we must allow that that day is a 
 long way off. " We are still very far from the time when it 
 will be possible to proceed synthetically in metaphysic : only 
 when analysis has helped us to perfectly distinct and definite 
 conceptions, will it be possible by synthesis, as in mathematics, 
 to subsume all complex conceptions under the most simple 
 principles." ^ 
 Kantt'^TSw"* IVom the last instance we see that Kant has not yet adopted 
 Locke.''^' °* *lie idea that the synthetic method of mathematics is impossible 
 for philosophy. He sees, indeed, that mathematics has the 
 advantage of a sensuous presentation of the objects in concreto, 
 but he does not yet clearly oppose perception to thought as 
 subjected to altogether different conditions. On the contrary, 
 he thinks that the impossibility of applying the mathematical 
 1 R. I. 97 ; H. II. 298. 
 
•'^- PEE-CRITICAL WHITINGS OP KANT. 139 
 
 method is due to the fact that the analysis of conceptions has 
 riot yet been carried down to the simplest elements. As to the 
 nature of the synthesis by which reason has advanced in 
 mathematics, Kant tells us that it is an arbitrary synthesis ; but 
 it does not yet occur to him to ask how an arbitrary synthesis 
 can be friiitful 'ini the knowledge of objects. Kant says nothing 
 to indicate that ^ Locke had any special influence upon his 
 mental development at this time. But we cannot fail to be 
 struck with the coincidence between Locke's ideas and those of 
 Kant, both in regard to the method of mathematics and to the 
 possibility of its future extension to other subjects. Por Locke, 
 too, held that mathematical truths had " barely an ideal exist- 
 ence," its constructions not being - " ectypes " of anything pre- 
 sented in' sense, but " archetypes " constructed by "the mind 
 itself. But just for that reason they would seem to . have no 
 basis in, or relation to, experience, which is always particular 
 and never universal : as Locke himself doubted the possibility 
 of " general propositions on matters of fact." On the other hand, 
 in spite of their ideality, the principles of mathematics were not 
 to Locke mere definitions of nominal essences ; they were not 
 " trifling " or analytic propositions, but propositions which gave 
 "instructive real knowledge '' ;in Kant's language, they were 
 synthetic judgments. How such merely subjective synthesis 
 should yet give real knowledge of objects, Locke does not inform 
 us, any more than Kant tells us how the " arbitrary combina- 
 tion" of simple conceptions should yet be a scientific process, — 
 a process of the discovery of truth and not of the invention of 
 fictions. What is still more remarkable is that both writers 
 seem to throw out somewhat vague hopes of extending the 
 method of mathematics to other subjects. For while Kant 
 speaks, in words already quoted, of the possibility of applying 
 the synthetic method to metaphysics at some future time, when 
 the work of analysis has been completed, Locke hopes for 
 an extension of knowledge by " singling out and laying in 
 order intermediate ideas " in the same way that mathematics 
 
140 INTRODUCTION. <^H*P- 
 
 has singled out and laid in order the " intermediate ideas that 
 
 demonstratively show the equality or inequality of unappli- 
 
 cable quantities." i 
 
 m^^inicai One Other point is necessary to complete' the account of 
 
 logical views Kaut's speculatiou in this important year, 1763. I have 
 
 of the world. 
 
 already referred to the modification of the Ontological argu- 
 ment which he still accepted, and I have attempted to show 
 that in it we may find an imperfect expression of the idea that 
 knowledge presupposes an ultimate unity which is prior to all 
 difference, even the difference of the intelligence and the 
 intelligible world. This is an idea which Kant never com- 
 pletely lost sight of, though in the Critique of Pure Re,ason, it 
 essentially changed its form and reappeared as the transcen- 
 dental unity of apperception, which is presupposed in all ex- 
 perience. In the period we are considering Kant still regards 
 God as the centre to which all must be referred ; but it is God 
 conceived simply as Being in general, i.e., as a Spinozistic unity 
 in which all difference is lost. Hence, as one of Kant's earliest 
 commentators declared, it would be truer to say that Kant 
 proved Being in the abstract to be God, than that he proved 
 God to exist.^ From the idea of this abstract unity Kant 
 immediately advances to the idea of a spiritual God, on the 
 
 ' Locke's Essay, Book IV. 12, 7. It is necessary here, as Dr. Vaihinger 
 points out, to guard against the confusion of the synthetic and the analytic 
 methods in question, with the synthetic and analytic judgments which are the 
 starting points of either. At this time (as we see from the Eefiexionen Kant's, 
 (II. 292, saq.) Kant held that "all rational judgments are analytic, and all 
 empirical judgments are synthetic" ; or, in other words, that Philosophy starts 
 from complex ideas or principles, I.e., from ideas which contain distinguishable 
 elements and which therefore, are expressed in synthetic judgments, though in 
 another way they may he "simple" (Pi. I. 160; H. II. 106), i.e., in the sense that 
 their elements cannot be separated from each other. Both these, however, Kant 
 derives from experience, nor does he yet ask how it is possible that complex 
 ideas, the elements of which are inseparable, can be so derived. On the other 
 hand, it is his view that mathematics arbitrarily puts together elements which 
 are given in their simplicity, and therefore that it may be said to start from 
 analytic judgments. 
 
 ^ Tieftrunk, quoted by Cohen : Die Systematische Begriffe in Kant's Vor- 
 kritischen Schrifien, p. 33. 
 
i^- PEE-ORITICAL WKITINGS OF KANT. 141 
 
 vague ground that there is " most reality " in such a being. It 
 is scarcely necessary to say anything in criticism of this sudden 
 leap from the most simple and abstract to the most complex 
 and concrete of conceptions. What, however, gives interest to 
 the transition is the way in which Kant connects it with the 
 opposition between efficient and final causes. After reaching 
 the idea of God by the " high priori road," Kant proceeds to 
 supplement it by means of an argument from experience. In 
 doing so he first dwells, — in language which he repeated with 
 little change many years afterwards in the Critique of Judg- 
 ment, — on the way in which each geometrical figure, when its 
 properties are evolved by mathematics, becomes a key to mani- 
 fold problems in relation to the world of experience ; while, on 
 the other hand, it rests upon the idea of space as a unity in 
 which all the peculiarities of it and all kinds of figures are 
 predetermined. The principle of unity thus revealed in all the 
 forms of the manifold world does not involve design, for it is a 
 necessary consequence of the nature of space : yet it conforms 
 well to the idea that there is a purposive unity in the original 
 principle from which the world in space is derived. A similar 
 thought is suggested to us by the way in which the necessary 
 causal relations of things give rise to a certain harmony and 
 order among them. This Kant illustrates in various ways, and 
 particularly by his own discovery that the same mechanical 
 laws, by which Newton had explained the present movements 
 of the solar system, are also adequate to account for the forma- 
 tion of that system. Every such step towards the extension of 
 the mechanical explanation of things by the laws of their 
 action and reaction upon each other narrows the limits of the 
 accidental, and points back to a unity as the source from which all 
 things come. On the other hand, it also limits the range within 
 which the idea of final causes has to be applied. For a unity 
 of design is necessarily an accidental unity, i.e., a unity not 
 given in the separate natures of the elements, which are com- 
 bined in order to realise it. Such a unity we find in the 
 
142 
 
 INTRODUCTION. 
 
 CHAP. 
 
 Germs of 
 Kant's later 
 views. 
 
 Summary of 
 the results 
 reached, by 
 Kant in 1764. 
 
 organic world : for, except by reference to an ideal cause, 
 which does not lie in the separate elements of the body, we 
 cannot explain how their action and reaction should be sub- 
 ordinated to the life of the individual and the species. At the 
 same time, there is much even in the organic world which we 
 can account for by physical laws apart from special design, and 
 still more, which we may be able to account for in the same 
 way in future. And it is the business of science to extend such 
 explanation as far as possible. It is not piety but sloth that 
 would bid us rest in the ignava ratio of final causes. Those 
 who do so, forget that when we show that a purposive con- 
 nexion of things flows from their mechanical relations, we are 
 also doing much to prove that that mechanism itself is the 
 manifestation of a rational principle. On the other hand, 
 however far we may go in this direction, we can see that there 
 are, and will always be, limits beyond which the mechanical 
 explanation of life and mind cannot be carried. 
 
 This attempt to " redd the marches " between mechanism and 
 teleology need not detain us, as we shall find Kant afterwards 
 reproducing, the main ideas of it in a better form. It is, how- 
 ever, needful to notice it here, as it shows how Kant's mind 
 was already working on lines parallel to those drawn in his 
 Critiques. Especially it lets us see the germiiial form of the 
 conception that the unity of intelligence is implied in nature, 
 even regarded as a system of necessity, — a thought which is, 
 we might even say, the fundamental conception of the Critical 
 philosophy. On the other hand, we may observe that Kant 
 does not yet characterise the teleological principle. as an "idea 
 of reason," although he already insists upon those cautions 
 against the resort to final causes, which were afterwards brouo-ht 
 into connexion with the distinction of ideas of reason from 
 principles of understanding. 
 
 The position, then, which Kant had attained at the besinnin" 
 
 o o 
 
 of 1764, may be thus summarised. He had broken with the 
 Wolffian philosophy and especially with its method. He had 
 
I^- PEE-CEITIOAL WRITINGS OF KANT. 143 
 
 shown that mathematical demonstration is something quite 
 different from the kind of logical deduction which Wolff had 
 exemplified, and that the latter is merely a method of analysis. 
 For the pseudo-synthesis of the Wolffians, who pretended to 
 deduce all truth from general principles, he proposed to substi- 
 tute a method of analysis which should begin by accepting the 
 matter to be analysed from experience. The formal principles 
 of thought would thus find their complement in a large number 
 of " irreducible conceptions," and would necessarily give rise to as 
 many propositions or judgments incapable of being proved. On 
 the other hand, Kant did not renounce that idea of a unity of 
 knowledge and of the knowable world, which was expressed in 
 a theological form by earlier philosophers. All he did was to 
 change the Wolffian back into the Spinozistic form of the 
 ontological argument, or rather into a form which, even more 
 distinctly than the Spinozistic, brings out the idea that that 
 unity is the presupposition of all knowledge. Further, he 
 attempted, in a method not altogether unlike that of Leibniz, 
 to show that that unity expresses itself in our knowledge, at 
 once as a principle of unity underlying the mechanical necessity 
 of nature, and as a teleological principle which subjects nature 
 to ends not included in itself Finally, in all these investiga- 
 tions he was led to direct great attention to the question of the 
 method of thought, and there is good evidence that he regarded 
 all the special results yet attained by him as merely provisional. 
 He had, so to speak, drifted from his moorings, and taken up a 
 tentative and critical attitude, looking about for light in every 
 direction and testing and comparing the different methods by 
 which truth was to be attained. In the programme of his 
 lectures in the Winter- Semester of 1765-6 Kant writes: — 
 " The youth, just liberated from the discipline of the school, 
 comes to the University accustomed to be taught, and thinks he 
 will now be taught philosophy. VBut what he has really to be 
 taught is to philosophise, j Before any one can be taught philo- 
 sophy, there must be a philosophy to teach : but where is it ? Who I 
 
144 INTRODUCTION. CHAP. 
 
 can show me a book and say : — ' Here is wisdom and insight 
 that can be depended on : learn to understand and to compre- 
 hend what is in this book, and build new conclusions upon it, 
 and you will be a philosopher.' Till you can show me such a 
 book of universal wisdom which I can appeal to, as to Polybius 
 for history, and to Euclid for geometrj^ you must allow me to 
 say, that you are abusing the confidence of the public in pre- 
 , tending to offer such ready-made wisdom :| for all you really 
 can do is to exercise the faculties of the youth committed to 
 your care, with a view to a knowledge which they will have in 
 the future to acquire for themselves. The true method of the 
 instruction of youth is zetetic, as the ancients called it : and it 
 is only when reason has been thoroughly exercised that it can 
 in some cases become dogmatic." The tone of these sentences 
 might leave it uncertain whether Kant is thinking more of the 
 defects of the actual systems of philosophy or of the exigencies 
 of the young, who need to be exercised to think rather than to 
 be taught definite conclusions. But it was his conviction at 
 this time that in Philosophy young and old are equally confined 
 to the zctctic method. " Metaphysics," he had said not long 
 before, "is undoubtedly the most difficult of the sciences, but 
 it is a science which has not yet come into existence." ^ And 
 in writing to Lambert in 1965 he tells him that the main con- 
 clusions which he had as yet reached were only in relation to 
 method, and that even in that point of view his studies were 
 not completed. " I have for several years been turning my 
 philosophical reflexions in every possible direction, and after 
 much tacking and turning, in which I was always seeking for 
 the sources of truth and error in the manner of procedure, I 
 have at length reached an assurance as to the method which 
 it is necessary to follow, in order to get beyond that illusory 
 appearance of knowledge which makes us think at one moment 
 that we have reached a decisive conclusion and the next 
 moment forces us to go back to the beginning again, and which 
 ' R. I. 88 ; H. II. 291. 
 
IV- PRE-CRITICAL WRITINGS OF KANT. 145 
 
 is the source of all the hopeless discord in the opinions of 
 philosophers. My habit now is always, in relation to every in- 
 vestigation I undertake, to examine first of all what I must know 
 in orde-r to be able to solve the special problem before me, and 
 what degree of knowledge is made possible by the data that are 
 given. In this way my judgments are often more limited in 
 scope, but they are likely to be more definite and secure than 
 is common. All my endeavours, therefore, point, as the main 
 interest, to the determination of the proper method of meta- 
 physic and through it of philosophy in general." ^ 
 
 The question of method and of the theory of knowledge was ?f°™^t™5re 
 now Kant's main interest and in view of it he held himself meUioa? ° 
 ready to sacrifice his most cherished opinion and traditional 
 habits of mind. " If a man is in real earnest about the truth, 
 he will at last be brought to the point at which he ceases 
 to spare his own productions, even those which might seem 
 likely to bring him some scientific credit. Everything he has 
 learnt or taught he brings under an unsparing criticism."^ 
 Such are the words in which Kant describes his attitude of 
 thought at this time, and he adds that this method "soon 
 enabled him to see that the whole dogmatic theory was 
 dialectical." He sought, however, to reach " something certain, 
 if .not in view of the object, at least in view of the nature and 
 limits of philosophy." He had brought himself into an attitude 
 of critical detachment from all speculative conclusions, and was 
 ready to weigh the reasons for and against any of them with abso- 
 lute impartiality, nay, even to lend all the powers of his reason to 
 argue against the side towards which he was conscious of any 
 bias.* In short he practised that " sceptical method," which he 
 afterwards declared to be the true preparation for criticism. 
 This did not involve any real indifference towards the settle- 
 
 1 R. I. 303 ; H. VIII. 655. 
 ^ Beflexionen Kant's, II. 4. 
 
 'Cf. B. Erdmann's introduction to the second volume of the Reflexionen 
 Kant's, p. 41, where this point is illustrated in a striking way. 
 VOL. I. K 
 
146 INTRODUCTION. chap. 
 
 ment of those questions in relation to which, (as he once said) 
 " it is impossible for any rational being to feel indifferent." 
 On the contrary, there are many indications to show that the 
 ideas of " God, freedom and immortality," under which Kant 
 brings all the problems of philosophy, were then and always 
 the main ultimate interest of his speculation. But he had 
 learned more and more to distrust the philosophy of the 
 schools and all its dogmatic ways of solving these problems 
 without any preliminary inquiry as to method. He had 
 observed how easy it is to maintain either side in such a 
 controversy, or rather how easy it is to attack either; for 
 in such disputes, as he said subsequently, the victory is always 
 with the attacking party. And he had begun to suspect that 
 the proof of either the affirmative or the negative in many 
 cases lies beyond our reach. Nay, he had begun to ask 
 further, how far it is necessary for our higher life that we 
 should dogmatically solve such questions, and whether there is 
 not a practical faith which can serve the purposes of human 
 life better than the demonstrations of an intelligence which is 
 unaware of its own limitations. 
 Essay On (Ae It is this temper of mind which expresses itself in that 
 
 Dreams of ' ^ 
 
 Ghost-Seer. strange essay On the Breams of a Ghost-Seer as illustrated 
 hy the Dreams of Metaphysic, published in the year 1766, in 
 which he, in the first place, sets before us the conception of a 
 world of spirits alongside of the world of experience, then tears 
 that conception to pieces in the spirit of a narrow Empiricism, 
 and finally withdraws from the question as one the means for 
 dealing with which are beyond our reach. In examining the 
 Arcana Caelestia of Swedenborg, Kant had been greatly struck 
 with the symmetry and self-consistency of the system of the 
 spiritual world as there pictured, and with its similarity to the 
 universe of self-determined yet harmoniously acting monads, 
 which had been conceived by Leibniz. He had also detected 
 a curious coincidence between the view of Swedenborg as to 
 the connexion of the spiritual and the natural worlds, and the 
 
'^- PRE-CRITICAL WRITINGS OF KANT. 147 
 
 view of Leibniz as to tKe relation of the world of phenomena to 
 the real world of monadic substances, And this harmony he 
 sets himself in a half-serious, half-humorous way to trace. If 
 a spirit or a ghost be defined, according to the common notions 
 ' on the subject, as a being that can be present in a portion of 
 space without manifesting that property of resistance or solidity 
 which distinguishes material objects, no one can say that such 
 a being is impossible. To prove that such a being actually 
 exists is another thing : for though good arguments have been 
 drawn from the nature of thought for the simplicity and indi- 
 visibility of the ego, this does not necessarily involve that the 
 soul is quite different in character from the elements of 
 matter. On the dynamic theory, matter itself is made up of 
 points of force, which, taken in themselves, have no extension 
 and which occupy space only as they repel each other, On 
 the other hand, since we know that the elements of matter 
 have impenetrability or repulsive force merely as a fact of 
 experience, without being able to comprehend it or see its 
 necessity, we cannot deny that substances without such a force 
 are possible. But if we suppose the spirit or soul to be such 
 a substance, we are landed in many difficulties as to the 
 manner of its connexion with the body ; e.g., whether it acts 
 upon the organism at one point or is present equally in every 
 part of it. In the latter case the mode of its presence is incom- 
 prehensible, in the former case it is difficult to conceive it as 
 acting on material substances without being itself one of them. 
 Are we, then, to suppose that every element of matter has 
 itself an inner principle of activity which determines its outer 
 relations, and that the immaterial substance works upon that 
 principle ? There is something plausible in the idea of 
 Leibniz that all monads are " representative " ; for it is 
 difficult to conceive how a simple part of matter should be 
 possible without any inner state, and it is impossible to 
 think of any inner state which is not analogous to thought. 
 5ut, though we admit this supposition, it does not help us 
 
148 INTRODUCTION. chap. 
 
 to see how a spirit and a body can form a unity, or how, 
 if they form a unity, they can ever be separated. Still, 
 having entered upon this perilous path of speculation, we have 
 only to follow it out and we shall find ourselves rapidly 
 constructing a system of the spiritual and natural worlds, 
 which may take its place beside that of Leibniz or Swedenborg. 
 " It is then as good as demonstrated, or it could easily be demon- 
 strated, if the process were not too tedious ; or, still better, it will 
 be demonstrated at some future time, I know not when or where, 
 that the human soul even in this life is bound up in indissoluble 
 community with all immaterial natures belonging to the world 
 of spirits, and that it acts upon them and is acted upon by 
 them, though, as men, we are not conscious of such influences 
 so long as it goes well with us. And it is also probable, 
 on the other side, that purely spiritual natures have no 
 immediate sensitive consciousness of the material world: for 
 they are not bound up in one personality with any part of 
 matter, in such a way as to become conscious of occupying a 
 definite place in the natural world and to apprehend through 
 special organs of sense their relation to other extended, 
 existences, although they may be able to produce influences 
 upon, and receive influences from, the souls of men, as 
 beings of kindred nature with themselves. Hence, not- 
 withstanding the communion between "the two orders of 
 spirits, it is impossible that those ideas which are received 
 by the embodied spirit, as a being dependent upon the 
 material world, should pass without change into the minds 
 of purely immaterial beings, or that the thoughts of im- 
 material beings should without losing their peculiar charac- 
 ter, pass over into the consciousness of men ; for the contents 
 of these different kinds of consciousness are specifically 
 distinct."! 
 fopStnt * ^ow, this idea seems to shed no little light upon our con- 
 sciousness of ourselves as moral and spiritual beings. For that 
 'R. VII. 53; H. II. 341. 
 
 ' Spiritualism. 
 
IV- PRE-CRITIOAL WRITINGS OP KANT. 149 
 
 consciousness appears always to point to a unity of all spirits 
 with each other, which limits their individual lives, just as the 
 law of universal attraction limits and determines the movements 
 of the different material bodies. " Is it not then possible to 
 represent the phenomena ' of the moral impulses, which deter- 
 mine the reciprocal'relations of conscious beings, as the effects 
 of an active force which they exercise upon each other ? 
 May we not regard moral feeling as just the felt dependence of 
 the particular upon the universal will, and as a consequence of 
 the natural and universal interaction whereby the spiritual ■ 
 world attains its moral unity, as it forms itself according to the 
 laws of its own connexion into a system of spiritual perfection ? 
 If we allow this thought to have enough verisimilitude to make 
 it worth while to follow it out to its consequences, we shall find 
 that they are such as to produce a certain prejudice in its 
 favour. For, if it were a true thought, most of those difficulties 
 would vanish, which arise out of the contradiction of the moral 
 and physical relations of men upon earth. Thus the moral 
 character of men can never have its complete effect in their 
 outward life according to the order of nature, but it may realise 
 itself fully in the world of spirits according to spiritual laws. 
 The unrevealed goodness of many an effort which is thwarted 
 in its result by defective power, or again the malice hidden 
 beneath many an apparently good act, are for the most part 
 lost, if judged by their physical effects in the material world : 
 but, according to the idea suggested, they would have to be 
 regarded as fruitful of consequence in the world of spirits ; for 
 in a world under spiritual laws, the connexion of the private 
 will with the universal wiU, — that will which constitutes the 
 world of spirits into a united whole, — would be such that each 
 free act would produce an effect and call forth a reaction in exact 
 accordance with its moral character." ^ Supposing this view to 
 be true, it would follow that, even now in the present world, 
 the spiritual subject must take the place among the spiritual 
 iR. VII. 56; H. 11.343. 
 
150 INTEODTJCTION. chap. 
 
 substances of the universe which is appropriate for it according 
 to moral laws ; and it must take that place with the same neces- 
 sity with which material bodies determine their respective places 
 according to the laws of motion. And if in a future state the com- 
 munity between the soul and the material world should be broken 
 off, the moral laws that already determine 'its relations in this 
 world would continue to operate without a break. The only 
 difficulty that remains unexplained is, how we are to reconcile the 
 existence of such a spiritual community with the fact that we are 
 so seldom conscious of it. For the spiritual world is present to 
 man, if at all, only in occasional glimpses, which, besides, have 
 often a somewhat uncertain and even irrational character. This, 
 however, is already explained by what has been said of the 
 nature of the consciousness of man as contrasted with that of 
 purely spiritual beings. For what we experience as Spirits wiU 
 not naturally enter into that consciousness which we have of 
 ourselves as men ; or if it does so enter at all, it wiU only be 
 under abnormal conditions, and even then the intimations from 
 the spirit-world will necessarily take the form of the conscious- 
 ness into which they intrude. Spiritual realities will be 
 pictured as objects and events in the natural world, and all the 
 imperfections of the medium will affect the vision. For men 
 in general such perceptions will have something of the character 
 of disease : and if there are a few exceptional individuals who 
 are so constituted as to be continuously conscious of spiritual 
 influences, their minds will be so much drawn out of proper 
 balance as to the things of this world by the confusing presence 
 of another, that they will often be regarded by other men as 
 insane. In this way it only needs a little ingenuity to explain 
 all the facts of ghost-seeing in accordance with our primary 
 assumptions as to the relations of the two worlds. " For meta- 
 physical hypotheses have wonderful pliancy ; and it would show 
 a great want of ingenuity not to be able to adapt this hypothesis 
 to every story of supernatural visitations, and that without 
 taking any trouble to investigate its truth, which in many cases 
 
 J 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 151 
 
 it would be impossible, and in yet more would be discourteous, 
 to attempt." ^ 
 
 After thus playing the spiritualist, Kant in the next chapter The Empirical 
 
 anBwer to 
 
 turns the tables, and writes what he calls an Anticabala, in spiritualism. 
 which he endeavours from the point of view of ordinary com- 
 mon sense to show the hoUowness of all such theories and 
 hypotheses. " When we wake," it has been said, " we live in 
 a world that is common to all : when we dream, we live each 
 in a world of his own." This world of our own may be con- 
 structed by the intellect or the senses ; by the " dreamers of 
 reason " or by the kindred " dreamers of sense "*: in both cases 
 it is removed from the ordinary tests of experience : it is not 
 part of that order and connexion of things which is present to 
 our waking consciousness. Its only criterion is self-consistency, 
 and this criterion may be equally satisfied by many different 
 systems of speculation and superstition, by a world of Sweden- 
 borgian ghosts alike with a world of Leibnizian monads. When, 
 however, we ask what experiences can be appealed to in support 
 of either, we find nothing except some sensible, and generally 
 visual, appearances, which are presented in certain states of 
 body to many men, and which to certain peculiarly constituted 
 men are frequent or even constant. But the rational explana- 
 tion of such facts is not supernatural powers but disease. Thus, 
 a defect in the organ of vision may easily explain how the 
 'pictures of imagination, which we usually represent as merely 
 mental phenomena, should appear as objects in the outward 
 world. To suppose an influence of spiritual beings without us, 
 in order to account for appearances which can thus be readily 
 explained by an unhealthy state of the body (a state manifestly 
 existing in many cases of supposed supernatural vision) is to 
 neglect all the rules of scientific method. The ghost-seer ^, 
 should rationally be regarded, "not as a half-citizen of the 
 other world, but as a candidate for the hospital." 
 
 Between these opposite views Kant does not decide. In foiuuon'jfth'e 
 
 problem. 
 » ' R. VII. 63 ; H. II. 349. 
 
152 INTRODUCTION. chap. 
 
 fact, he seems to say that he regards the controversy mainly as 
 an illustration of the " sceptical method " of balancing opposite 
 theories against each other and so freeing ourselves from a 
 false dogmatic attitude in regard to questions which we cannot 
 settle. Enough for us is gained when we can set a purely 
 physical theory against a spiritual or spiritualistic one, and so 
 free ourselves from the strong bias of our own hopes and fears. 
 In this way, Kant declares, " I have freed my mind from pre- 
 judice, and extinguished the blind credulity which opened the 
 way to many a fictitious assumption of knowledge. 
 Formerly, I lo(5ked at the general intelligence of man from the 
 point of view of my own intelligence ; now I put myself in the 
 place of an alien and external reason and observe my own 
 judgments, with all their secret occasions, from the point of 
 view of others. The comparison of these different views shows, 
 indeed, many marked instances of what may be called mental 
 parallax, but it is the sole means of preventing illusion and 
 reducing our conceptions to the true place which they occupy in 
 relation to our faculties of knowledge." ^ It is our hopes and our 
 wishes, which throw themselves into one of the scales and make 
 us lean to the belief in a community of spirits : otherwise, our 
 theories would not be so exclusively taken up with the question 
 how the spirit of man goes out of the world, to the neglect of 
 the not less difficult questions how it comes into the world, and 
 how it is present in it. But the " polemical discipline of the 
 understanding " prepares us to admit that the whole subject 
 belongs to the region of opinion and not of knowledge, in which 
 dogmatic denial and dogmatic assertion are equally irrational. " I 
 venture to assert that if the reader makes a proper use of what 
 has been said, he will find in it the last word which philosophy 
 has to utter about such spirits, and that whatever variety of 
 opinions may appear in the future, no one will ever be able to 
 know any more. This may seem a bold statement to make, 
 when we find it impossible either by observation or by reason- 
 1 li. VII. 74 ; H. II. 356. 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 153 
 
 ing to exhaust the nature of any one object of sense, be it even 
 a grain of sand or a drop of water: so immeasurable to a 
 limited intelligence like ours is the complexity which nature 
 presents even in its smallest parts. But it is quite different 
 with the doctrine of philosophy in regard to spiritual beings. 
 That doctrine can be completed, though only in a negative way. 
 In other words, we can here ascertain with certainty the limit 
 of our knowledge, and convince ourselves that, while we are 
 able to observe the various phenomena of life and to discover 
 their laws, we are totally unable to reach any positive deter- 
 mination as to the principle of that life. We may, indeed, 
 suspect that there is a spiritual principle beneath the natural, 
 but we cannot know it ; for we find no data in our sensuous 
 perceptions on which such knowledge could be based. We are 
 driven, therefore, to resort to negatives, in order to determine 
 that which is absolutely separated from all that is given in 
 sense : though the possibility of even such negative determina- 
 tion rests neither on experience nor on reasoning, but solely on 
 a fiction to which reason is obliged to have recourse in default 
 of any other means of dealing with its object. On this footing, 
 philosophy can be nothing but a doctrine of our necessary 
 ignorance of a certain problematical class of beings ; and, if it 
 confine itself to this, there is nothing to hinder it from com- 
 pletely solving the problem set before it." ^ 
 
 In what remains of the essay Kant gives an account of the PMiosophy 
 
 *' ° conceived as 
 
 stories told of Swedenborg's prophetic powers, and also a sketch |^^ umitFrf 
 of the system of the spiritual world exhibited in the Arcana """^ ^^' 
 Caelestia, with the view of showing its general accordance with 
 the spiritualistic theory previously stated : and in the end he 
 returns to the lesson that metaphysic (" of which," he declares, 
 " I have always been a lover though I can boast of few of her 
 favours,") is unable to solve the problems which reason suggests 
 as to the hidden nature of things, but that it can do much to 
 teach us the limits of possible knowledge. The great import- 
 , ' K. VII. 77 ; H. II. 359. 
 
154 INTRODtrCTION. CHAP. 
 
 ance of this negative use of metaphysic lies in the fact that it 
 is only the known impossibility of gaining anything by attempt- 
 ing to go beyond these limits, that will induce men to confine 
 themselves within them. " So long as it is thought possible to 
 reach that distant goal of knowledge, so long it is vain for a 
 wise simplicity to protest that we can do very well without it. 
 The pleasure of advancing knowledge makes it readily take the 
 appearance of a duty ; and a deliberate self-restraint of reason 
 seems to show, not the simplicity of wisdom, but a stupidity 
 which hinders the elevation of our nature. For questions as to 
 the nature of spirit, as to freedom and predestination, as to the 
 future state, etc., at once set in motion all the powers of the 
 intelligence, and draw men by their importance into a fever of 
 speculation which subtilises and decides, dogmatises and con- 
 troverts, with every new semblance of insight. It is only when 
 such discussions give place to a philosophy which tests its own 
 procedure, and takes account, not only of objects, but also of 
 their relation to the mind of man, that the limits can be drawn 
 closer and the boundary stones laid, which wiR henceforth 
 prevent speculation from passing beyond its proper sphere. It 
 needs some philosophy to discover the difficulties that surround 
 many conceptions, which are treated by the ordinary conscious- 
 ness as easy and simple. A little more philosophy drives away 
 the illusion of knowledge which still remains, and persuades us 
 that such objects lie entirely beyond the. horizon of man's 
 intelligence."^ 
 Experience Kant then procccds to lay down the limits of knowledge in 
 
 the source of " 
 
 aUKnowiedge. ^j^g game Way as in his previous essays. We can analyse 
 phenomena, he thinks, according to the principles of identity 
 and contradiction, but ultimately such analysis always brings 
 us to certain primary connexions of phenomena which we cannot 
 get beyond. " The fundamental conceptions of things as causes 
 and of their forces and actions are quite arbitrary when not 
 taken from experience, and apart from experience we can neither 
 iR. VII. 101 ; H. II. .377. 
 
IV. PRE-ORITICAL WRITINGS OF KANT. 155 
 
 prove nor "disprove them. I know well that thought and will 
 move my body, but I can never by analysis reduce this pheno- 
 menon, as a simple experience, to anything more simple than 
 itself : I know that it is so, but I cannot tell why it is so. 
 Thatjny will should move my arm is not to me more intel- 
 ligible than that it should hold back the moon in its course : 
 the difference is only that I experience the former, but that 
 the latter has never come under the observation of my 
 senses." Hence, we can say nothing at all about the nature 
 or existence of spirits without bodies ; and of the actual 
 commercium of soul and body we can only discover that it 
 exists and that it has certain laws ; but we are totally 
 unable to explain the why or the how of this action and 
 reaction. 
 
 When we have realised the impossibility of answering such ^e°ndent if*^" 
 questions, we are prepared to find that it is unnecessary for us to ™° *^ ^^°^' 
 answer them. For, in the first place, the obligations of morality 
 are not in any way dependent on the belief in another world. 
 " Surely, the heart of man contains in it immediate moral pre- 
 scriptions ; nor does he require, in order to move himself to act 
 according to his vocation in this world, to fix his lever upon 
 another." The hope of immortality will not turn a knave into 
 a righteous man, though it is a hope which springs up in aU 
 good men, as a natural accompaniment of their elevation of 
 feeling. " Hence it appears more accordant with human nature 
 and the purity of morals to ground the expectation of a future 
 world on the feelings of a well disposed mind, than to attempt 
 to base good conduct on the hopes of another world." A 
 moral faith, i.e., a faith that springs out of morality instead 
 of pretending to prepare the way for it, is the only faith 
 useful to man. And in this sense we may adopt the saying 
 of Candide as true philosophy. Instead of seeking to know 
 what is too high for us, our business is to "cultivate our 
 garden." ^ 
 
 , iR. VII. 106; H. 11. 381. 
 
156 INTRODUCTION. ''°'^^- 
 
 criHoSEmt''* I have given a somewhat full account of this essay, because 
 ^J^aXtlf"^ it contains a curious anticipation of the critical philosophy on a 
 Jph^ ' °" lower level of speculation. The results are in many respects 
 analogous, still more the general tendency. Here, as in Kant s 
 later writings, we have the critical withdrawal from a problem 
 which gives rise to an antinomy of reason, because it contams 
 difficulties which are beyond the reach of the human under- 
 standing ; and here, too, the thesis and the antithesis of the 
 antinomy are derived from the dogmatic assertions of idealists 
 or spiritualists and the dogmatic denials of empiricists. And 
 to complete the likeness, we have here also a final recurrence 
 to the moral consciousness as that out of which alone any 
 rational faith can be drawn. On the other hand, while we 
 find Criticism in this essay, it is still merely an Empirical 
 Criticism. For in the first place, the antinomical results of the 
 application of the intelligence to certain subjects is taken 
 merely as a fact, and not explained. But, as Kant afterwards 
 pointed out, the mere fact that all previous speculation has 
 failed to deal successfully with certain problems does not of 
 itself give us reason to assert that they are beyond the limits 
 of our understanding : otherwise progress would be impossible. 
 Ultimately, we can be finally prevented from discussing such 
 subjects, only if it is shown from the very nature of our faculty 
 of knowledge that it is beyond our power to deal with them. 
 Now, it is true that in the present essay Kant supplies a kind 
 of theory of knowledge and its limits, when he says that we are 
 confined to the analysis of data which are given us in experience, 
 and that the primary relations of things as causing changes in 
 each other must be simply taken as facts. But this only shows 
 that Kant has not yet asked himself the question : What is 
 experience ? In particular, he does not yet see that, if experi- 
 ence is the knowledge of a connexion of effects and causes, it 
 involves something which is not given in sense. In other 
 words, he is not yet aware that there is a synthesis involved in 
 the assertion of the very principle of causality, as truly as in 
 
IV. PRE-CRITICAL WRITINGS OF KANT. 157 
 
 the act of judgment by which special phenomena are brought 
 together under that principle. And it is only another over- 
 sight of the same kind that, while he shows that antinomies 
 arise in the attempt to answer certain questions, it does not yet 
 occur to him to inquire how we come to ask such questions at 
 all. For an object must be in some sense within the reach of 
 the intelligence, when it is possible to ask a question about it ; 
 and if so, it is hard to see how we should be able to know that 
 the question is absolutely unanswerable. In all these respects 
 Kant as yet falls short of an adequate consciousness of the critical 
 problem, as he came afterwards to understand it. Such a con- 
 sciousness could only be attained by a deeper investigation of 
 the conditions which render knowledge possible within, and im- 
 possible beyond, certain limits, which yet are not absolute 
 boundaries to all human thought. 
 
 Meantime, the unsparing rigour with which Kant applied Mendelssohn's 
 
 J^ o o rr objections and 
 
 his elenchus and the occasional bitterness of his jests at the ^™*'^ ™P^y- 
 expense of the dogmatic philosophy gave some not unnatural 
 offence to Mendelssohn, who, though not strictly a representa- 
 tive of the dogmatism of the previous time, yet upheld many 
 of its traditions in a somewhat softened form. Mendelssohn 
 was the most eminent of the popularising Eclectics, who tried 
 to provide a philosophy adapted to a time of transition by a 
 somewhat inconsistent mixture of different elements, held 
 together only by a common-sense dislike of pushing anything 
 to an extreme. To such a man the ruthless and fearless dia- 
 lectic of Kant could not but be startling and annoying. In 
 answer to a letter expressing these feelings Kant admits that 
 his essay bore traces of the conflict of feeling under which it was 
 written. ' " It was difficult for me," he says, " to find a way of ex- 
 pressing my thoughts without exposing myself to ridicule. It 
 seemed to me, therefore, best to anticipate others by ridiculing 
 myself. Yet in this I have proceeded quite honestly, since the 
 state of my mind was in truth somewhat absurd. For I cannot 
 ri^ myself of a slight inclination to attach credit to such stories. 
 
158 INTRODUCTION. chap. 
 
 nor can I altogether regret the grounds of reason which support 
 their possibility ; 'and that in spite of the extravagances which 
 are connected with the former, and the hair-splitting subtilties 
 and irrationalities which deprive the latter of their value." As 
 regards metaphysic, Kant goes on to declare that he is so far 
 from intending to cast any scorn upon it that he is convinced 
 that " the true and lasting well-being of humanity rests upon 
 it." But just in order to make room for a true metaphysic, he 
 thinks that the time has fully come to drag off the dogmatic 
 garb with which a false and pretentious philosophy has clothed 
 itself. Of such a Katliartikon for perverted intelligence, it was 
 the aim of his essay to supply a specimen. The special point 
 to which it directed discussion might, perhaps, have been more 
 clearly indicated : it was the question of the nature of the rela- 
 tion between the soul and the world, i.e., both the natural and 
 the spiritual world in which it has a place. " In my opinion, 
 the first point necessary is to examine what data we have for 
 the solution of the problem, how the soul is present in the 
 world and how it comes into relation both with material 
 things and with other beings of its own kind. In other words, 
 it is necessary to discover what is that power of acting on 
 things without and of receiving impressions from them, which 
 is exemplified in the union of the human soul with a body. 
 Now such knowledge must be derived either from experience 
 or from reason. But there is no experience whereby we can 
 come to know such a subject in 'the various relations which 
 alone would reveal its external force and capacity and its har- 
 mony with the body ; no experience which can throw light on 
 the connexion of the inner state of the soul with the outer state 
 of the matter of gur body, and so on the connexion of an inner 
 activity with an outer activity. Is it then possible to deter- 
 mine these forces of spiritual substances by a priori principles 
 of reason ? This question reduces itself to another more 
 general question, viz., whether it is possible by reason to dis- 
 cover a primitive force, i.e., the first fundamental relation of a 
 
IV, PRE-CRITICAL WRITINGS OP KANT. 159 
 
 cause and an effect, I answer with certainty that it is impos- 
 sible. Hence I am reduced to the conclusion that, except in 
 so far as such forces are given in experience, they are only 
 fictions of imagination. Such fictions can never be proved to 
 be even possible, and to say that they are thinkable, — which 
 appears to be the case only because we cannot show that they 
 are impossible, — is purely illusory. For I have shown in my 
 essay that, on the same method, I can defend the dreams of 
 Swedenborg himself against any attack upon their possibility. 
 The analogy, which 1 have there drawn, between the moral 
 influence of spiritual beings upon each other and the universal 
 gravitation of matter, is not to be taken as an earnest opinion 
 of mine. It was intended only to show the ease with which 
 we can advance in philosophical fictions, when there are no 
 data to guide or check us ; and at the same time to make mani- 
 fest the necessity of considering what data are needed for the 
 solution of the problem and whether these data are forthcoming. 
 For, if once we were to set aside all proofs from the fitness 
 of things and the divine designs, and to ask simply whether 
 it is possible to derive from experience a knowledge of the 
 soul, sufficient to reveal to us the manner of its presence in 
 the universe both in relation to material things and to other 
 spiritual beings, it would become easy for us to determine 
 whether birth, life and death (in a metaphysical sense) are 
 things which are within the reach of our intellectual insight. 
 In short, the great question is whether at this point we do not 
 discover limits fixed for us once for all by the finitude of our 
 intelligence, as well as by the nature of the empirical data with 
 which that intelligence has to deal." -^ 
 
 We here see Kant at his nearest point of approximation to End of the 
 
 ■^ pre-critical 
 
 Locke. But, as he reached that point of view by a dis- p^"""*- 
 covery of the defects of the opposite school of thought, he was 
 the more ready to detect the incompleteness of the Lockian 
 criticism, and, indeed, the inherent contradiction of an empiri- 
 
 , iR. XI. 10; H. Vm. 674. 
 
160 INTRODUCTION. chap. iv. 
 
 cal criticism of experience. It was this discovery which was 
 soon to lead to. a partial recoil towards the Wolffian Eationalism 
 and to give rise to that first imperfect synthesis between it and 
 Empiricism, which is contained in the Dissertation of 1770 
 on the Form and Principles of the Sensible and Intelligible 
 World. 
 
CHAPTEE V. 
 
 THE DISSERTATION OF 1770, AND KANT'S STUDIES FOR THE 
 CRITIQUE OF PURE REASON IN THE YEAES 1770-1781. 
 
 IT has often been noticed that Kant, from the beginnins tolKant-s 
 ° ^ Ireconciling 
 
 the end of his career, shows a tendency to _gpek for some/^^i°p™g^i™i'^ 
 middle term or higher reconciling principle between opposite ™* 
 schools of. thought. " We are in a way defending the honoui 
 of human reason, when we reconcile it with itself in the persons 
 of different writers of high intelligence, and discover the truth, 
 which by such men is never entirely missed, even in their con- 
 tradictory utterances."^ This sentence quoted from the first 
 published essay of Kant, in which he endeavours to find a way 
 of combining the different views of Descartes and Leibniz upon 
 vis viva, gives the keynote of most of his subsequent writings, 
 To mediate between Leibniz and Ifewton was the aim of his 
 first philosophical essays ; to mediate between the English 
 Empiricism and the German Eationalism may be said to be one 
 of the main objects of the Critical Philosophy. The idea of 
 criticism itself, as was shown in the first chapter of this work, 
 springs out of the opposition of different dogmatisms and of 
 dogmatism to, scepticism, and it is essentially an attempt to 
 reconcile them.^ But this reconciling tendency involves a 
 
 1 R. V. 189 ; H. I. 144. 
 
 ^ Note, however, that Kant always demands a real mediation of the opposite 
 dogiftatiams by " going back to the point from which their divergence began," 
 VOL. I. L 
 
162 INTRODUCTION. °HAP, 
 
 willingness to do full justice to each side in the debate and to 
 follow out the reasonings of each to their utmost consequences. 
 Hence we do riot wonder that Kant, like Plato, was deeply 
 impressed with the advantages of what in the Critique he calls 
 " the discipline of pure reason in its polemical use." As early 
 as 1758 we find him announcing to the students who proposed 
 to attend his lectures on metaphysic, that on Wednesdays and 
 Saturdays he intended to treat polemically the doctrines ex- 
 pounded in the previous days ; because in his opinion this was 
 " one of the most excellent means to attain to profound views of 
 any question." And at the end of the last chapter, we have seen 
 that his essay on the Breams of a Ghost-Seer as illustrated hy the 
 Dreams of Meta]physic is just an example of this method as it was 
 practised by him in the period of what has been called his 
 Critical Empiricism. The often-quoted sentence in the Prolego- 
 mena to the effect that it was Hume who " first awakened him 
 from his dogmatic slumber " has probably a narrower reference ; 
 though Hume is .sometimes taken by Kant as the general 
 representative of the Scepticism that flows from the collision 
 of opposite dogmatisms. But we can now place beside it an 
 expression in a letter of Kant to Garve (dated September 21st, 
 1798), in which he gives a somewhat different account of this 
 " awakening." " The point," he says there, " from which I 
 started was not the discussion about the being of God or about 
 immortality : it was the antinomies of pure reason in general, 
 — from the first antinomy (' The world has a beginning : it has 
 not a beginning ') to the last ('There is freedom in man : there 
 
 i.e., by rising to a point of view which is above the opposition and from which 
 it may be explained ; and that he had no toleration for a mere " splitting the 
 difference," such as was common enough in the Eclectic systems of his time. 
 "Moderation," he declares, " which tries to hit the mean between extremes 
 {welcher auf die Halbschied geht) and thinks it can find the philosopher's stone 
 in subjective probability, and which by heaping upon each other many isolated 
 reasons, none of which is by itself convincing, supposes it can supply the place 
 of the sufficient reason, is no philosophy at all. " (R. I. 652 ; H. VI. 492. ) Dr. 
 B. Erdmann (see especially his Introduction to the second volume of the Reflex- 
 ionen Kant's) first showed the full importance of the " polemical discipline " of 
 the understanding in Kant's pre-critical development. 
 
V. DEVELOPMBISTT OP THE CBITICAL PHILOSOPHY. 163 
 
 is nothing in man but necessity') — that first awakened me from 
 my dogmatic slumber, and drove me to the criticism of pure 
 reason, in order to remove the scandal of the apparent discord 
 of reason with itself" And in one of the Eeflexions which Dr. 
 Erdmann has published, Kant tells us that it was in the 
 practice of this polemical method that the light of the critical 
 doctrine first dawned upon him. " I saw this doctrine at first 
 only in an obscure way. I endeavoured earnestly first to 
 prove a proposition and then to prove its opposite ; not in order 
 to set up a sceptical, doctrine, but because I suspected an 
 illusion of the understanding and sought to discover wherein it 
 lay. The year 1769 brought me great light." ^ 
 
 The exact effect of his use of the polemical method is in- Relation of 
 
 the polemical 
 
 dicated in another of these Eeflexions, which we have already 5'^'?'°'* *° 
 quoted. •" I discovered that the whole dogmatic theory is 
 dialectical ;" *.«i, according to Kant's use of the term " dia- 
 lectical," he discovered that the antinomies of reason are 
 not a mere accidental equipoise of reasons in particular cases, 
 but that in all metaphysical questions there is something in 
 the very nature of human reason which brings it into collision 
 with itself. His next step, therefore, was to go back upon the 
 subjective conditions of knowledge. " I sought to discover 
 something certain, if not in view of the object, yet in view of 
 the nature and limits of our- way of knowing > it." In other 
 words, his failure to determine the objects turned his atten-^ 
 tion to the method of knowledge. His subsequent progress is 
 thus described : " I gradually came to see that many of the 
 propositions which were generally regarded as objective, are 
 really subjective, i.e., they are expressive of the conditions 
 under which alone we can apprehend or compfehend the 
 object." ^ " Before the Disputation " {i.e., the Dissertation) " I ' 
 
 had already got an idea of the influence of the subjective con- d 
 
 ^ Sejkxionen Kant's, II. p. 4, § 4. In the introduotion to these Reflexions 
 Dr. Erdmann collects all the passages quoted in the last paragraph. 
 " M. § 3. 
 
164 INTRODUCTION. *'^*''- 
 
 ditions of knowledge upon our knowledge of objects, which was 
 soon followed by the discovery of the distinction of the sens- 
 ible from the intellectual conditions. As yet, however, this 
 .distiaction was viewed by me raerely on its negative side." ^ 
 foundtX"^ This last remark requires a little explanation. Kant's 
 dewrmination polemical method led him to consider the subjective conditions 
 objects as of knowledge. So far he had already advanced at least as early 
 as 1766, when he wrote the Breams, of a Ghost-Seer and laid 
 down the doctrine that Metaphysic must henceforth become a 
 doctrine of the limits of human reason. The next step, how- 
 ever, which brought him to the ideas of the Dissertation, was 
 the distinction of the sensitive and the intellectual conditions 
 of Tjnowledge. Now that distinction was manifestly suggested 
 by the results of an essay On the Rational Basis for the Bis- 
 tinction of Regions in Space, which was published in 1768. 
 For in that essay Kant finally emancipated himself from the 
 Leibnizian view of space and adopted a view kindred with that 
 of Newton. Leibniz had conceived space as a confused idea of 
 the relations of coexistent monads ; and Kant himself, though 
 he refused in some respects to adopt the Leibnizian theory, 
 had maintained in the Monadologia Physica that it is. a relation 
 of points of force, which, taken separately, are not in space. 
 Even in that treatise, however, it gave Kant some trouble to 
 reconcile the discoveries of Newton, — which as stated by 
 Newton himself were connected with a view of space as pre* 
 supposed in, and prior to, all the substances that exist 
 and move in it, — with his own modified Leibnizian theory.^ 
 
 1 Eeflexionen Kant's, II. § 6. 
 
 " See the fourth chapter of this Introduction, p. 105. In Kant's earliest essay- 
 he maintained that space is a relation of substances which taken individually 
 are not in space, and he even sought to deduce the characteristic property of. 
 our space, that it has three and only three dimensions, from the Newtonian 
 law that bodies attract each other with a force that varies inversely as the 
 square of the distance. Hence he thinks that we can quite reasonably suppose 
 that there are other worlds in which bodies act on each other- according to a 
 different law, and in which, therefore, space has a different number of dimen- 
 sions. "A science of all these possible kinds of space would certainly be the 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 165 
 
 And in 1768, he became convinced that such a reconciliation 
 is altogether impossible. In a treatise published in 1748, the 
 celebrated mathematician Euler had called attention to the 
 difficulty of giving any definite meaning to the highest laws of 
 motion, and especially to the law of inertia, if the idea of 
 space be taken as derived by abstraction from the relations of 
 things ; seeing that that law asserts the persistence of a material 
 object in its state of motion or rest, so long as it is not acted 
 upon by some other outward object. But this obviously 
 implies that the body in question stands in a certain relation 
 to space itself and therefore can be regarded as moving or 
 resting apart from, or independent of, its relation to other 
 material bodies. Starting from this suggestion, Kant endeavours 
 to show that a similar relation to absolute space is involved in 
 all determination of direction in space, and so in all the 
 " perceptive judgments in relation to extension." For in all 
 such judgments, it is presupposed that " absolute space has a 
 reality of its own, independent of the existence of matter, and 
 that it is itself the. first ground of the possibility of matter as 
 a complex or divisible substance."^ In other words all deter- 
 mination of things in space presupposes, in addition to the 
 determination of the relations of the different parts or objects 
 to each other, a determination of the relation of the whole 
 system of such parts or objects to an all-embracing space. 
 
 The argument is as follows. Space has three dimensions. Priority ot 
 
 Space. 
 
 and in it, therefore, we can draw three planes, cutting each 
 
 highest Geometry which a, finite understanding could attempt." (R. V. 27; 
 H. I. 22. ) But with the new view of space as a universal form of external 
 perception, Kant gives up all such speculations as to the possibility of a space 
 of more than three dimensions, though he does not exclude the idea that there 
 may be other forms of sense for other finite beings than men. Those who have 
 maintained the possibility of space of more than these dimensions should con- 
 sider what definition they could give of space that would leave open such a 
 possibility. If it is defined, as Locke and others define it, as that which has 
 partes extra, partes (in German, as the Aussercindnder), it seems impossible to 
 carry out the conception of a space with more than three dimensions without 
 contradicting the definition. But the subject is too wide to be discussed here. 
 »E. V. 294;H. IL 386. 
 
166 INTRODUCTIOlir. chap. 
 
 Other at right angles, " and as through our senses we know 
 what is without us only so far as it stands in immediate con- 
 nexion with ourselves, so it is in the relation of three such 
 planes to our body, that we naturally iind the first ground for 
 the definition of the different regions in space." The plane at 
 right angles to our body we call " horizontal," and by relation 
 to it we distinguish the regions we call " above " and " below." 
 The other planes enable us to distinguish the regions " before " 
 and " behind," " right " and " left." In these distinctions we 
 find a means of expressing differences which we can indicate in 
 no other way. Thus a screw that turns to the right and one that 
 turns to the left, may absolutely correspond in all their parts, 
 but the one cannot be substituted for the other, or placed so as 
 to coincide with it. The right and left hand, or the face and its 
 image in the mirror, are other instances of objects, which, 
 though perfectly corresponding in all their parts and in 
 all the relations of their parts to each other, yet cannot 
 occupy the same space ; such objects are what Kant calls 
 " incongruent counterparts " of each other. Now, it is to be 
 observed that in the distinction of these objects there is always 
 implied a relation of their parts not to each other but to space 
 beyond them. The order of the parts in any complex material 
 body may be inverted without making the least change in their 
 relations to each other, or to the body as a whole. The right 
 is not distinguished from the left hand, except in relation to 
 the space it occupies. And in like manner, the order of the 
 heavenly bodies might be inverted without any change so far 
 as their relations to each other are concerned. The difference 
 would lie only in their relation to absolute space. Thus the 
 complete ground for the determination of any material form 
 does not lie in the position of its parts relatively to each other, 
 but also in its relation to universal or absolute space : not, 
 indeed, that this relation can be immediately perceived, but 
 that distinctions of bodies can be perceived which rest only on 
 this presupposition. Hence, " absolute space, though it is no 
 
It ia, therefore, 
 an a priori 
 condition of 
 perception. 
 
 V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 167 
 
 object of external sense, is a fundamental conception which (jo^ 
 
 makes all such sense perception possible." ^ 
 
 The result thus arrived at seemed at iirst to be a complete 
 confirmation of the Newtonian view of space, as opposed to the 
 Leibnizian. But the words just quoted suggest a thought 
 which was soon to give a new turn to Kant's speculations. 
 The one universal space, which is presupposed in all determina- 
 tion of external objects, is not itself an object of external 
 sense, but a " fundamental conception that makes external 
 perception possible," It is, in other words, a presupposition of 
 external perception which the mind brings with it, and without 
 which it would be impossible to perceive any object as external. 
 It is, in Kant's subsequently adopted language, a form of such 
 perception, which is not derived from sense, but rather con- 
 stitutes it as a " capacity of being so affected by objects as to 
 perceive them." It is an " a priori " through which alone 
 there can exist for us any " a posteriori." Kant adds that this 
 must not lead us to " take it for a mere ens rationis (GedanJcen- 
 ding) " ; though, as he immediately goes on to say, there are 
 many difficulties which arise out of this conception when we 
 attempt to make its reality intelligible to ourselves through 
 ideas of reason, " evident as that reality may be to the inner 
 .sense." In other words, Kant thinks that the argument of his 
 essay is a complete refutation of the Leibnizian view, that space 
 is a confused idea of the relations of monads wljich are not 
 spatially determined at all. But at the same time, he sees that 
 by his acceptance of the Newtonian conception of it as a 
 reality presupposed in all external existence, he is landed in 
 many difficulties from which the Leibnizian view had saved 
 him. For space is continuous and divisible ad^ infinitum ; it 
 is, therefore, impossible to represent a world, of which space is 
 the primary condition, as a completed whole made up of real 
 or substantial parts. To Leibniz the world was a systematic 
 whole, definitely determined as the greatest " sum of com- 
 • 1 R. V. 301 ; H. II. 391. 
 
The world as 
 conceived. 
 
 168 INTRODUCTION. chap. 
 
 possible reality," and its constituent parts were monads, i.e., 
 individual and indivisible substances. But a world in space 
 cannot be conceived as a self-bounded universe beyond which 
 nothing can exist, or as a whole made up of individuals which 
 cannot be further divided. Though it presents itself to us as a 
 complex whole or unity of differences, the world in space is 
 incapable of being made intelligible to our understandings 
 either in its unity or in its differences : for we cannot carry it 
 back by analysis into its elements, neither can we mark out by 
 any limits the sphere of its totality. Nay, we can see that 
 from the nature of the case it is impossible to do either. On 
 both sides the world escapes into infinity, and refuses to be 
 known as we cannot but seek to know it. And the same 
 difficulties reappear, when we try to conceive the development 
 of the world as a series of states in time. 
 
 ITow the object of' Kant's Dissertation on the Form and 
 PriTiciples of the Sensible, and Intelligible World, with which in 
 1770 he entered upon his duties as a professor, is to follow out 
 to its consequences the antinomy just suggested, and to offer an 
 ' explanation of it. This explanation is derived from the distinc- 
 tion between the a priori conditions of perception and those 
 of thought, or in other words, between sense and understand- 
 ing, and it was suggested by the results of his essay upon The 
 Rational Basis for the Distinction of Regions in Space. Let us 
 first bring before us the precise nature of the difficulty or 
 antinomy as it is stated by Kant. What are the elements 
 involved in our conception of the intelligible world as we must 
 think it ? In dealing with this question Kant returns to the 
 old answer which he had given in the Dilucidatio Nova, his 
 first strictly philosophical work. For the first effect of the 
 distinction which Kant now drew between sense and thought 
 was to free his original Eationalism from the weights that had 
 been depressing it, and to restore it for a time to its former place 
 in his philosophy. He maintains, therefore, that when we think 
 of the world abstractly, we necessarily regard it as a unity, made 
 
y. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 169 
 
 up of a number of different substantial parts, which maintains its 
 identity through all the changes of these parts in relation to each 
 other. But this implies that we conceive it as having a certain 
 matter of which it is made up, and a certain form according to 
 which this matter is bound together so as to make up a world. 
 The matter lies in the elementary substances which are its 
 parts : the form lies in the way in which these parts are co- 
 ordinated. Now, it is to the form that we must look to explain 
 all possible influences of one substance upon another, and at 
 the same time, the limit of such influence. For in order that 
 the world may remain the same world, not only the identity 
 of the parts, or substances out of which it is made up, must be 
 maintained, but the changes through which they pass must be 
 consistent with, and indeed be determined by, the characteristic . 
 mode of combination by reason of which they make one whole. 
 Hence, finally, we must conceive the world as an absolute 
 whole, which is definitely determined in itself, both in the 
 multitude of its parts, and in the states through which they 
 pass. 
 
 Now, when we regard the world as it presents itself to us in The world a 
 time and space, we find that it in every way refuses to be 
 brought under such a , conception as we have just given. For 
 in space there can be no smallest parts, no parts which we can 
 regard as substances, and no whole which is limited and 
 bounded in itself. And in time, in like manner, there can be 
 no indivisible moments and no absolute beginning or end. Nor, 
 in the idea of the unity of space or time can we find any principle 
 to determine the connexion between the objects that exist and 
 the events that take place in them. For time and space are 
 forms in accordance with which phenomena may be placed in 
 relation to each other ; but they do not contain any principle in 
 virtue of which such relations can be actually established. 
 Nothing in the nature of space and time fixes any substanqe or 
 any world of substances to a particular place, or any event or 
 series of events to a particular date. For, wherever or when- 
 
170 INTRODTJCTION. chap. 
 
 ever we place or date these objects and events, there is always 
 left an infinity beyond, in relation to which the whole complex 
 of substances and events is brought ; while yet there is no 
 ground in this infinity for determining the relation of anything 
 to it. 
 s?Meand°* Now, whcn we look upon this picture and upon that, we 
 Thought. begin to see that there is a collision between the conception of 
 reality which we necessarily form to ourselves in pure thought, 
 and the nature of that reality which we actually apprehend by 
 sense as existing in time and space. How are we to explaiu 
 this antinomy between the aspects of the world, both of which are 
 necessarily presented to us ? We cannot but seek to reduce aU 
 complexity to its simple elements, yet it is demonstrable that we 
 never can divide any spatial or extended object into simple 
 parts. We cannot but regard all changes as completely deter- 
 mined through the relations of the different parts to each other 
 ' in the unchangeable whole ; yet it may easily be shown that 
 
 in time we can have presented to us only an endless series of 
 changes, which determine each other successively and in which 
 no absolute beginning or end is capable of being reached. We 
 cannot but regard the whole of things as limited and determined 
 by itself; but as in time and space, the world is either 
 absolutely limitless, or it hangs in a vacuum which yet con- 
 tains no condition to determine it to be when and where it is. 
 These, and similar difficulties meet us whenever we begin to 
 compare our ideas with each other ; and they force us to ask 
 the meaning of this great dualism which runs through 
 them, and which makes us take one view of the world when 
 we contemplate it as it is perceived under the forms of time 
 and space, and quite another view when in thought we regard 
 it sub specie aeternitatis. 
 ^rniSmenli ^ow, Kant's answcr in the Dissertation, — and we shall see 
 n^woridsr^' that, though with great modifications, it was his answer 
 always, — is to revive the old Platonic opposition of the 
 phenomenal and the noumenal worlds, of the w^orld as 
 
V- DEVELOPMENT OP THE CRITICAL PHILOSOPHY. 171 
 
 it is to sense and the world as it is to intelligence. The 
 antinomy arises, in other words, from the fact that we 
 have two separate sources of knowledge, which are not com- 
 pletely in harmony with each other, sense and thought. Our | 
 general idea of the world, that which directs and underlies all 
 the questions of the intelligence, is due to the activity of the 
 intelligence itself. But the particular data, with which we 
 strive to answer these questions, are due to sense and received 
 under its conditions, though these conditions are such that the 
 answer can never be adequate. " Quas mens ah intelledu 
 acceptas ferf ideas dbstraotas, illas in concreto exsequi et in 
 intuihis commutare saepenumero non potest."^ There is a 
 " subjective hindrance " in the nature of sense which makes it 
 impossible to realise in particular what we conceive in general ; 
 and the consequence is that instead of one world of objects, we 
 seem to have two ; and the intelligence appears to be at vari- 
 ance with itself as to the nature of reality. This appearance of 
 contradiction, however, merely arises from a confusion of that 
 which is incapable of being perceived or imagined with that 
 which is incapable of being thought. Por while that which 
 contradicts the laws of reason, that which is unthinkable, is 
 impossible, that which cannot be perceived or imagined may 
 quite well be possible. The general idea of the world, which 
 the intelligence presupposes and on the basis of which it asks 
 its questions, is not to be regarded as invalid, because, owing 
 to the manner of our perception, we can never verify it in 
 particular. On the other hand, if we suppose the mind to be 
 thus hindered from working out its pure conceptions of reality 
 by the defect of sense, it becomes possible to lay down at least 
 a negative principle, by which metaphysic may be guided in 
 its determination of the noumenal world as the absolute 
 reality of things. It must abstract from all the special 
 conditions under which sense and imagination are found to 
 represent their objects, and must coniine itself to those de- 
 • 1 R. I. .S05 ; H. II. 397. 
 
172 
 
 INTRODUCTION. 
 
 CHAP. 
 
 Necessity of 
 removing the 
 couditions of 
 sense from 
 things in 
 themselves. 
 
 terminations of reality which the intellect derives from itself 
 alone. 
 
 In this way then Kant reaches what as yet he considers an 
 adequate principle of criticism. Detecting that the determina- 
 tion of things as in time and space is not adequate to the 
 demands of thought, he infers that we must regard such deter- 
 mination as applicable only to phenomena ; and that, on the 
 other hand, in order to determine noumena, we must abstract 
 from time and space and so purify our metaphysic from those 
 general principles of sense which have hitherto worked con- 
 fusion in it. " There seems to be a need for a quite peculiar 
 though merely negative science, (Fhaenomenologia generalise 
 which should precede metaphysic, and which should determine 
 the validity and limits of the principles of sensibility, so that 
 they may not confuse our judgments in relation to the objects 
 of pure reason. For time and space, and the axioms that 
 determine our perception of things under the relations of time 
 and space, are, in reference to our empirical knowledge and 
 all the objects of it, very real, since they express the conditions 
 under which all phenomena and all empirical judgments are 
 possible. But when we are dealing with any object, not as an 
 object of sense but through a universal and pure conception of 
 reason, e.g., when we are regarding it as a thing or a substance 
 in general, we are led into many misconceptions, if at the same 
 time we bring it under the fundamental principles of sense." ^ 
 It is the sensitivae cognitionis cum intellectuali contagium that 
 has introduced the greatest confusion into metaphysic, bring- 
 ing with it insoluble dilemmas as to the soul, the world, and 
 God, and forcing us either to deny the reality of substances, 
 especially of spiritual substances, or else to subject them all 
 to the conditions of space and time. If we can free ourselves 
 from this fatal source of error, we shall at once get rid of all 
 those obsciire, and at present unanswerable, questions as to the 
 seat of the soul, as to the eternity of matter, as to the omni- 
 1 B. I. 360 ; H. VIII. ,663. 
 
objectivity of 
 intelligence. 
 
 V. DEVELOPMENT OP THE CRITICAL PHILOSOPHY. 173 
 
 presence of God, etc., which have perplexed philosophy. For 
 we shall be able to see that they all arise from our treating a 
 subjective condition of our apprehension of things, as if it were 
 an objective predicate of things in theinselves. And it may 
 well be that when we have cleared away such unanswerable, 
 because irrational, questions as to the nature of things, the true 
 questions, i,6., the questions which can be answered, will make 
 their appearance, and metaphysic will be established on a 
 sound basis. ^ 
 
 The main point then is to show that time and space are not of"sinse mS 
 conditions of real being, but only of our modes of perceiving it, 
 And this Kant attempts to do in a method almost identical 
 with that which he afterwards followed in the Aesthetic. It 
 may, however, be of advantage to follow his discussion of the 
 subject in this earlier form, as some of the difficulties of the 
 Critique arise from his never having substantially altered that 
 form, even when the new elements of thought subsequently 
 introduced rendered such alteration desirable. Kant, so to 
 speak, built up the Analytic of his Critique of Pure Reason 
 upon a repetition in the Aesthetic of the ideas of his earlier 
 Dissertation, instead of reducing these ideas to an element in a 
 new construction. Now the view which Kant presents to 
 us in the Dissertation starts from a fundamental opposition of 
 the phenomenal and the noumenal, or, stating it subjectively, 
 of sense and thought. Sense is passive and receptive, and what 
 is presented to it depends on the manner in which what he 
 calls the status representativus of the subject is affected by the 
 presence of the object. Intelligence is active, and its pure 
 activity is shown in its forming for itself conceptions of things 
 which are independent of all affections of sense by these things, 
 but which for that very reason are objective, i.e., refer to the 
 things as they are in themselves. In other words, these con- 
 ceptions have a real and not merely a phenomenal value, 
 because they do not depend on the special character of the 
 subject, which is shown in the modifications its sensibility can 
 
174 INTRODUCTION'. chap. 
 
 suffer, but on the universal activity of reason which is the 
 same in all rational beings. Hence it is clear that " sensitive 
 cogitata esse rerum repraesentationes uti apparent, intellectualia 
 autem sicuti sunt." 
 Diatinotion of This, however, as Kant 2oes on to point out, does not 
 
 form and ' ^ ^ 
 
 ^nst™"' exhaust the theory of the relations of sense and understanding. 
 
 To do that we have to distinguish between the form and the 
 
 matter of sense, and between the logical and the real use of the 
 
 understanding. As regards the former, the matter of sense 
 
 consists in the affections to which it is subjected, while the 
 
 form is the mode in which these affections are co-ordinated by 
 
 the mind in being apprehended. The former, therefore, may 
 
 be characterised as a posteriori and the latter as a priori. 
 
 Neither matter nor form of sense, however, can be taken as 
 
 expressing the real nature of things. The matter cannot be so 
 
 taken : for " though sensation witnesses to the presence of an 
 
 object which produces it, the quality of the sensation depends 
 
 not on the object in itself, but on the nature of the subject as 
 
 capable of receiving a certain modification from the object." 
 
 And the form, though it gives rise to a certain general relation 
 
 of the elements of the sensible, which is not contained in the 
 
 sensible affections themselves, is yet not independent of these 
 
 affections, but is merely " the innate law of the mind according 
 
 to which it co-ordinates the affections it receives from the 
 
 present object " : ^ in other words, it is the result of an 
 
 " internal principle of mind, according to which the various 
 
 objects that affect sense take on a certain aspect as in being 
 
 Distinction of presented to us they coalesce into one whole." Again as 
 
 the logical and ° 
 
 toteihgence. ^^gards the activity of intelligence, we have to distinguish 
 between its logical and its real use, i.e., between the formal 
 process whereby it brings one idea under another more general 
 idea, guided only by the logical laws of identity and contradic- 
 tion, and the real process, in which it determines objects 
 according to conceptions which are not derived from sense but 
 1 Dissertation, § 4. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 175 
 
 produced by the pure activity of the inteUigence itself. By 
 the former, the logical use of the intelligence, we can rise to 
 general or abstract ideas which express the common elements 
 of sensible appearances ; but this process, which is commonly 
 called experience, does not take us beyond phenomena and the 
 empirical laws of their connexion. On the other hand, the 
 intelligence in its real use is the source of what we may also 
 call abstract ideas ; not, however, in the sense that they have been 
 abstracted from the objects of sense, but in the sense that they 
 are produced by the mind in complete independence of them. 
 These conceptions are grounded in the very nature of the 
 intelligence, not indeed as " innate ideas," or ready-made con- 
 ceptions which we find in our minds prior to all experience, 
 but as laws of the activity of the intelligence, which it reveals 
 when it is brought into operation in experience. Such are 
 the conceptions of possibility, of existence, of necessity, of 
 substance, of cause, etc., with their opposites and correlates.^ 
 
 ■^ Here we may observe is the first distinct appearance of the a pHori in 
 Kant's later sense, i.e., the thing, though not yet the name. The genesis 
 of this conception may probably have been helped, as Vaihinger suggests, by 
 the Nouveaux Essais of Leibniz, which were first published in 1765 ; though we ' 
 cannot say that there is conclusive evidence of this in the language of the 
 Dissertation, and we must remember that Kant was always familiar with the 
 Leibnizian conception of the development of consciousness by inner self- 
 determination. In 1763 he wrote (K. I. 155 ; H, II. 101) " There is something 
 great, and, as I think, very true in the thought of Leibniz, that the soul 
 embraces the whole universe in its consciousness, though it is only an infin- 
 itely small part of that consciousness which is clear. In fact all kinds of 
 conceptions must rest solely upon the inner activity of our spirit as their 
 ground. External things may, indeed, contain the condition under which 
 they appear, but not the force that produces them." The idea of distin- 
 guishing a universal u, priori element from an element which is particular 
 and dependent upon an outward stimulus did not, however, as yet present 
 itself to Kant, who at that time referred all synthetic principles to experi- 
 ence. But the discovery of the priority of space, in 1768, brought him 
 into many difficulties which he could solve only by the idea of its a-priority ; 
 and the priority of the idea of God and of the general conceptions of 
 causality, reciprocity, etc., could not but be interpreted in the same way. 
 Thus Kant was forced to make a distinction, which even in 1766 he had over- 
 looked, between the relations of particular causes and eflfects, which are 
 supposed to be given a, posteriori, and the general principle of causality which 
 is derived from the mind itself. The element still wanting to the conception 
 
176 INTRODUCTION. <^i^^''- 
 
 Knowledge Now, it is of the utmost importance to observe that the 
 
 through sense, 
 
 kindtom distinction between the sensible (including both the form and 
 thr^gh^^ the matter of sense) and the intelligible, is one not of degree 
 intelligence. ^^^ ^^ ^.^^ ^^ ^^^^ ^.^^^ Leibniz that the sensible is that 
 
 which is confusedly known, and the intelligible that which 
 is distinctly known, is to treat a real distinction of the data 
 of thought as if it were merely a logical distinction. What 
 is known through sense may be distinct, and what is known 
 through intelligence may be obscure ; but the kinds of distinct- 
 ness and obscurity are different. The difference between 
 perception and conception is qualitative, as is that of the 
 objects with which they have respectively to do. Phenomena 
 and noumena are things essentially different. 
 Definition of Let US then scc what are the characteristics, the merits and 
 
 the absolute t» *- i . 
 
 reauty as it defects of cach of thcse kinds of science. Metaphysic exhibits 
 
 exists for the '^ •' 
 
 intelligence, •'.^^q general principles of pure intelligence," and these 
 culminate in a " certain archetype or perfcctio Twumenmi" 
 which we presuppose as the ultimate reality of things, and 
 which at the same time we set before us as the ultimate goal 
 of all our efforts, speculative and practical. On its speculative 
 side we call it God, on its practical side, moral perfection. It 
 may be described quantitatively as a maximum, which, how- 
 ever, we must remember, is not reached by adding, since rather 
 all that falls short of it is defined by limiting it. Thus it is 
 the common measure and principle of all knowledge, the same 
 that Plato meant by his ideas, which are at once the ideals we 
 aim at and the real being of things as distinguished from their 
 appearance. In this sense God is the principle of knowledge, 
 
 of a priori as it appears in the Critique, is the consciousness that a priori 
 synthesis is not possible to perception alone or to conception alone, but only- 
 through the determination of the form of sense in conformity with the cate- 
 gories of the understanding. Already in the Dissertation (§ 8) he speaks of the 
 pure conceptions as abstracted from the innate laws of the mind (by attending 
 to its action on the occasion of experience), and therefore acquired and not 
 innate ideas." See B. Erdmann, in Philosophische Monatsheft, XX. 67 where 
 he traces the change by which Kant substitutes his own idea of the a priori for 
 that of the Wolffian school : cf. also Vaihinger, I. 190, seq. 
 
 M 
 
V- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 177 
 
 the ideal which guides it, and which it seeks to realise, because 
 He is the' absolute reality, the principium essendi ei fiendi of all 
 things.^ 
 
 Of this ideal, however, which is also the absolute reality, we "^^ general 
 
 ' ' •/ ' definition can- 
 
 have only symbolic knowledge, for it is never presented to us l'n*pa?tio?Qarf 
 in perception. "We know it not intuitively but discursively ; in ™"^ '™°°' 
 the abstract by general conceptions, not in the concrete in its 
 individual determination. Hence our pure intellectual concep- 
 tion of it has, as we have already seen, to be realised or 
 specified by means of sense perceptions, which do not corre- 
 spond with it. For all our perception or sensuous intuition is 
 bound down to a certain formal principle, under which alone 
 we can apprehend the individual as such; and the matter, 
 which is apprehended under this form, is not the objective or 
 noumenal reality, but only the sensible affection which it pro- 
 duces in us. We must, indeed, think of the divine intelligence 
 as creating its objects, and so intuiting or perceiving them in 
 their individuality by the same activity whereby it thinks of 
 them in general. But our intelligence is ectypal not archetypal ; 
 it has to work with materials which it passively receives in 
 the various forms of sense-perception, and therefore it cannot 
 find itself again in the objects it apprehends, or, in other 
 words, cannot realise and find in them the idea of truth which 
 it brings with it.^ 
 
 On the other hand, though, for the reason just given, ^f*en°*^'°^' 
 phenomena are not things in themselves or Ideas in the knowledge. 
 Platonic sense, yet this does not prevent them from being 
 objects of scientific knowledge. For, in the first place, our 
 sensitive apprehensions by their passivity necessarily bear wit- 
 ness to the presence of real objects which cause them, and this 
 
 ' In saying that the principles of pure intelligence " culminate in this arche- 
 type," Kant seems to anticipate what he afterwards expresses in the 4th 
 section of the Dissertation, that the idea of God is the presupposition of the 
 unity of the world as a world of individual substances reciprocally acting and 
 reacting on each other. See below, p. 183. 
 
 i Dissertation, § 10. 
 
 VOL. I. M 
 
 / 
 / 
 
178 INTRODCJCTION. <^^-^P- 
 
 excludes any Idealism that would deny the existence of things 
 in themselves. And in the second place, though these things 
 are not present to us as they really are in themselves, we must 
 remember that true knowledge consists in the agreement of 
 predicates with the subject of which they are predicated ; and 
 that, when we speak of the things of sense, our objects are 
 phenomenal as well as the qualities by which we determine 
 them, being both equally conditioned by the forms of sense. 
 Hence the two great empirical sciences of Physics and Psycho- 
 logy are not illusory, though they deal only with phenomena : 
 and the same is true of the a priori sciences of Geometry and 
 Mechanics as well as of Arithmetic, which are based upon the 
 forms of sense, and which form the very types of scientific 
 exactness. So much may be said to show in what sense it is 
 true and in what sense it is false that there is no possible 
 science of the things of sense.^ 
 
 In order to develop more clearly the contrast we have been 
 exhibiting between the intelligible world and the world of 
 sense, the noumenal and the phenomenal, it is necessary for us 
 to show more definitely the nature of time and space which we 
 have seen to be the forms of sense and so of the phenomenal 
 world ; and on the other hand, in opposition to these, to give 
 more precision to the idea of the intelligible or real world and 
 the forms by which we determine it. 
 ttSe'and^space Now what wc havc to show is that time and space must be 
 sense and 80 of conceived as forms of sense which do not belong to things in 
 
 the phenome- , , , i.ii !_■ 
 
 nai world themsclves, but condition the perceptions which these things 
 produce in us, i.e., determine how the data of sense must be 
 combined together in our perception of sensible objects. In 
 order to prove this, we have to distinguish time and space at 
 once from the matter of sense to which they give a character- 
 istic form and from the conceptions of the understanding, the 
 universal forms of thought, by which, — apart from all sensible 
 affections and, therefore, from all special characteristics of our 
 1 Dissertation, §§ 11, 12. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 179 
 
 individual being, — we determine objects as they are in them- 
 selves. In the former point of view, it has been already 
 pointed out that the mere affections of sense cannot give rise 
 to a principle co-ordinating those affections ; nor can we, as 
 Leibniz supposed, rise from the special relations of objects in 
 time and space to the ideas of time and space in general ; for, 
 on the contrary, these special relations presuppose the ideas of 
 time and space and are particular determinations of them. 
 Thus the ideas of simultaneity and succession cannot be made 
 the basis of the idea of time, nor can we define time, as is 
 sometimes done, as " the series of actualities which exist after 
 each other " ; for this " after " means " in different times,", just 
 as the idea of simultaneous actualities is the idea of those 
 which exist at the same time. And it is to be observed further 
 that " though time is of one dimension, yet the ubiquity of 
 ' time (to use Newton's expression), according to which all 
 sensible phenomena are somewhere, adds another dimension to 
 the quantity of reality," and that the definition of time just 
 given would leave no room for this second dimension. In like 
 manner it is impossible for us to derive the idea of space from 
 objects given as external ; for in order to be given as external 
 to each other they must be represented as in space and as 
 occupying different positions in it. In other words, in deter- 
 mining objects as in space or events as in time, we are always 
 determining a space or time which is presupposed in, and not 
 itself produced by, the perception of the objects or events in 
 question. Space and time are thus not the objects of percep' 
 tion, but the conditions which make it possible,^ 
 
 On the other hand, the priority in question is not the Nature of the 
 
 ''-•''■ priority of 
 
 priority of a genus to its species, of a conception to the par- sp^^e anytime, 
 ticulars that fall under it. It is the priority of an individual 
 whole to the parts into which we divide it, and which cannot 
 be represented except as in it. All spaces are in space, 
 all times are in time: and if we speak of several times 
 • 1 Dissertation, §§ 13-15. 
 
180 INTRODUCTION. chap. 
 
 and spaces, they are not separate wholes which have a com- 
 mon element, or which become one merely by aggregation, 
 but special limitations of the one space and the one time. 
 Hence time and space are objects of perception, and of pure 
 perception ; of perception because individual, of jMre perception 
 because not derived from sensation but prior to, and pre- 
 supposed in, objects of sense, as supplying the basis for certain 
 relations which necessarily are established between all such 
 objects. For time is a necessary condition of all phenomena, 
 and space is a necessary condition of all external phenomena. 
 Only this view of time and space can make intelligible the way 
 in which mathematical science determines a priori the relations 
 of objects, relations which are not analytically developed from 
 the general conceptions of time and space, but intuitively 
 discerned when we make certain constructions in time and 
 space. Thus it is that we perceive that space has three dimen- 
 sions and time only one, that parallel lines can never meet, etc. 
 Hence Geometry does not demonstrate its universal proposi- 
 tions by thinking its objects through general conceptions, as is 
 the way with the objects of intelligence, but by bringing them 
 before the eyes or the imagination in individual perceptions, 
 i as is the way with the things of sense. And, as a con- 
 firmation of this view, the essay On the Rational Basis of our 
 Distinction of Different Regions in Space has shown also that the 
 difference between " incongruent counterparts," a difference 
 necessarily implied in all determination of objects in space, 
 cannot be defined except by bringing before our mind's eye the 
 space of which they are the limitations. 
 Continuity of The nature of time and space will be still further illustrated, 
 
 apaceandtime. 
 
 if we refer to their continuity, i.6., that characteristic which 
 makes it impossible to resolve them into simple parts, or take 
 away their complexity. The reason of this is, that through 
 time and space we can think nothing but relations, and in a 
 relation there must be a difference of terms. Hence to deprive 
 time and space of all complexity were to annihilate them. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 181 
 
 Moments in time, points and lines in space, are not parts of 
 time and space, but only limits which we cannot conceive 
 except in relation to that which is limited. 
 
 From all this we gather that time and space are not objective T°eTof the' 
 and real, not substances, nor accidents or relations of sub- andLeitnizian 
 
 ideas of space 
 
 stances ; but subjective or ideal conditions which the mind '^^^ '™«- 
 brings to its. perceptions, and by which it co-ordinates them. 
 For, e.g., space is not something in which things can be contained, 
 and by reference to which they can be placed. Things are in 
 space only as they or their parts stand in a certain relation to 
 each other. Yet we cannot identify space with this relation, 
 •else we should lose the distinction of regions in space. Space 
 thus seems to hover between an adjective and a substantive, 
 being the presupposition of certain relations without being any- 
 thing to which they can be attached, or which itself determines 
 them. And the same is true of time. Hence the Newtonian 
 view of space and time has an element of truth in it, in so far 
 as it leaves room for the priority of space and time, as the pre- 
 supposition of certain relations of things, and thus is consistent 
 with the possibility of Geometry as an a priori science, which 
 yet is objectively valid in relation to all external phenomena. 
 But while it thus secures a basis for mathematics, the New- 
 tonian view involves the absurd idea of " the real existence 
 of infinite relations apart from any things related,"^ and by 
 making space and time the conditions of all existence it lands 
 us in a materialistic theory of the universe. The Leibnizian 
 view, again, which regards space and time as abstracted from 
 the relations of things as given in sense, avoids these difficulties ; 
 but it fails to explain the priority of the ideas of space and 
 time and the a-priority of mathematical science. On the other 
 
 1 Cf. Kant's letter to Herz (R. XL 29 ; H. VIIT. 692) where it is said that 
 ^' space cannot be supposed to be objective, or (therefore) intellectual (i.e., a 
 conception of the pure intelligence) because when we analyse it, we find 
 neither the idea of objects (which can exist only in space) nor a real connecting 
 link between them (which at any rate could not be asserted apart from any 
 object connected), i.e., no causal relations in which one thing could be regarded 
 at the ground of determinations in another. '' 
 
182 INTRODUCTION. chap. 
 
 hand, the Kantian view explains how it is that time and space 
 should be prior to all objects of perception, as being the sub- 
 jective presuppositions under which alone objects can be 
 perceived, presuppositions which the mind brings with it to per- 
 ception, not indeed as innate ideas but as internal laws of the 
 mind that determine how it must co-ordinate the matter of its 
 affections so soon as its sensibility is stimulated. 
 Difference and Kant coucludes his theorv of the a priori forms of sense by 
 
 relation of J Jr j 
 
 M'foms 0?*°^ noticing that the two forms are not altogether independent of 
 ^™°°' each other. For if we compare them, we see that space is a 
 
 determination of objects as such, while time has to do with 
 states, and principally with the states of the subject as repre- 
 senting or perceiving objects. Hence time stands between the 
 conceptions of the understanding and the perceptions of sense. 
 Viewed in the latter relation, time depends upon space for the 
 completion of its intuitive determination as an object ; for we 
 have to represent time as a line, and its limits, the moments of 
 time, by points. On the other hand, " time approaches more 
 nearly than space to a conception of the understanding, as it has 
 greater generality and embraces all things in its relations, both 
 space itself and also properties which are not comprehended in 
 relations of space, to wit, the thoughts of the mind." Nay 
 more, time, though it does not dictate laws to the understanding, 
 yet conditions the action of the understanding in comparing its 
 conceptions. For though it cannot be asserted that there is no 
 impossibility except that a thing should be ..and not be at the 
 same time, it may safely be asserted that we cannot judge 
 that anything is impossible except as it involves this.^ Again, 
 though it cannot be asserted that there is no dependence of 
 effect upon cause except where the former is posterior in time 
 to the latter, it may safely be asserted that, apart from their 
 position in time as prior and posterior, it is impossible to dis- 
 cover their relation as cause and effect. Lastly, even the 
 quantity of space itself cannot be determined by us, unless we 
 ' A view which ia rejected by Kant in the Critique. A. 153 ; B. 192. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY, 183 
 
 express it by number in relation to a certain unit, and number 
 is a multitude which is distinctly* known by successively 
 adding one to one in a given time. 
 
 The general result of Kant's argument as to space and time P"re conoep- 
 
 *-" ox tions neces- 
 
 is to show that they are forms which objects must take on, as aetermination 
 objects of perception to us, i.e„ as co-ordinated in relation to delation to " 
 
 . • (. T 11 each other. 
 
 each other in our sense perception of them, but that they are 
 not principles of the co-ordination or relation of objects as 
 things in themselves. They do not explain how individual 
 things can form part of one world and act upon each other. 
 We cannot, consistently with the view we have taken of space, 
 say that things act on each other because they are in one 
 space : rather we must say that it is because they act on each 
 other, that they appear to us as in one space ; and we must, 
 therefore, look for some other principle of relation than space, 
 which may make the fact of their reciprocal action and reac- 
 tion intelligible to us. Now no such principle of connexion 
 between individual things is involved in their existence ; for 
 the conception of individual substances does not imply any 
 tf^elativity, or, at least any relativity except, if they are 
 contingent substances, to their cause. Thus if objects existed 
 necessarily, they could have no reciprocal relations whatsoever, 
 because no dependence on anything but themselves. Their 
 existence in mutuo commercio is, therefore, possible, only if 
 they are dependent substances and if they all depend upon 
 one cause. Unitas in conjunctione substantiarum Universi 
 est consectariv/m dependeniiae omnium ah Vno ; and the same 
 Being who creates the matter of the Universe, i.e., the 
 substances of which the world is made up, must also give 
 it that form by which all these substances exist in mutuo 
 commercio. Further, we must apply this principle also to 
 the relation of knowledge, the relation in which the sub- 
 stances we call minds are so connected with other substances 
 as to know them. We must explain such knowledge, in 
 other words, by the dependence of the knowing substance as 
 
184 INTRODUCTION. ' chap. 
 
 well as of those that are known upon the one Being who gives 
 them existence and binds them together as one world. Hence 
 the unity of space and the unity of time, which are the pre- 
 suppositions of our perceptions of the world, are themselves to 
 be explained as the forms in which this higher unity presents 
 itself to us. Space, which is the necessary and universal con- 
 dition of the coexistence of all things together with each other, 
 may be called Omnvpraesentia Phaenomenon. And time, which 
 is the universal and necessary condition under which all the 
 changing states of things are represented by us as states of one 
 permanent world, may be called Aeternitas PhaeTiomenon. Such 
 thoughts, however, seem to Kant to carry us beyond the limits 
 of our finite intelligence, and to be too closely akin to the 
 mystic " seeing all things in God " of which Malebranche had 
 spoken.i 
 ttiTrcBuiteof' ^^ *'^^ concluding remark to which we have just referred, 
 ^ii^^""""' Kant shows himself dissatisfied with the results of his work on 
 the positive side, i.e., in so far as it leads to the development 
 of a new metaphysic ; and in a letter to Lambert he declares 
 that the whole of those results must be regarded as provisional,^ 
 and that the real value of the Dissertation lies in its negative 
 side, in those sections that show that time and space should 
 be taken as determinations of phenomena and not of things 
 in themselves. At the same time, he declares that the positive 
 side is that for which ultimately the negative exists : in other 
 words, it is as a step towards a new metaphysic or science of 
 things in themselves, that Kant seeks to show that the de- 
 
 1 Dissertation, 22. Cf. Kant's lectures on Metaphysic published by Politz 
 (pp. 113, 338) to which Dr. B. Erdmann first directed attention in the Phil. 
 Monatsheft (XIX. 133). Also Rejhxionen Kant's (11. 106, 123, etc.). These 
 passages showthat the view above expressed, accordingto which God is conceived 
 as mediating between the mind and its object, was no mere passing thought, 
 but that for some time Kant rested in it, as supplying a means of connecting 
 his new idea of space and time with his earlier theory that the existence of 
 
 I God is the presupposition of all thought and being. How this conception 
 ! makes way for the idea that the unity of the conscious self is presupposed in 
 all experience, we shall soon see. 
 
 2 R. I. .360 ; H. VIII. 663. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 185 
 
 terminations of time and space are to be taken merely as 
 determinations of phenomena. This remark will be found 
 afterwards to be important in considering Kant's object not 
 only in the Dissertation but also in the Critique, which was 
 the ultimate product of the course of speculation upon which 
 he had now entered. 
 
 In order to appreciate the merits and defects of the Dis- ^J^l^^^^^ 
 sertation, we have to keep before us the fundamental conception worS°'™ 
 
 phenomenal 
 
 upon which Criticism rests. Self-criticism means self-tran- from the point 
 
 •*■ of view of 
 
 scendence. It implies that the critic has reached a point of ™*«^*senoe. 
 view from which he can look down upon the previous work of 
 his intelligence, and see what is its limit or defect. The 
 application of this to the present case is obvious. To talk of 
 phenomena except in relation to noumena, or of a consciousness 
 of the phenomenal as such, except in relation to a conscious- 
 ness which in some way reaches to the noumenal, would be 
 absurd ; for it is only through its opposition to the noumenal 
 that the phenomenal has any meaning. Now Kant's method 
 of thought in the Dissertation is in conformity with this 
 principle ; for he maintains that man has not merely a re- 
 ceptivity for impressions, but a universal faculty of intelligence 
 or understanding, which by its pure activity can determine 
 things as they really are. It is true that this universal faculty 
 is brought into play only in response to the stimulus of sense ; 
 but the conceptions, which are its products, are not the results 
 of that stimulus, nor can they be explained by it either in 
 whole or in part. Hence, when we direct our attention to these 
 conceptions and work out, their full meaning, they, in Kant's 
 language, " issue in a certain exemplar or archetype, which we 
 can perceive only by the pure intelligence, and which is the 
 common measure of all other things as realities;"^ i.e., they give 
 us a conception of what reality or the object of knowledge mws^ 
 be in order to satisfy our idea of knowledge. This conception 
 of that which reality must be, is what Kant tries to exhibit in 
 ♦ ' Dissertation, § 9. 
 
186 INTRODUCTION. chap. 
 
 outline, in the section of the Dissertation on the form of the 
 intelligible world, and by it he judges the ideas of the world 
 which are derived from sense. Thus the knowing subject, 
 in so far as it is passive and dependent upon sense, is regarded 
 by Kant as an object like other objects ; and its perceptions are 
 conceived to be merely states of its own particular being, 
 dependent on the way in which it is affected by other beings 
 and things. But in so far as it is an active intelligence, 
 it apprehends itself and all other objects in their true nature, 
 and can, therefore, criticise itself, and recognise the subjective 
 or particular character of its own perceptions. In this way, it 
 can see its own limits without being able to get rid of them ; 
 for, as active intelligence, it necessarily takes its stand upon 
 certain general ideas and conceptions which it cannot but 
 regard as true, while yet it finds itself altogether unable to 
 realise the idea of truth involved in these conceptions by means 
 of the only perceptions given to it. It is obliged to make 
 demands upon itself which it cannot fulfil, to conceive the 
 object of knowledge as that which yet it can nexer perceive or 
 inmgine. And if it insist on bringing together the two disparate 
 elements of knowledge and working out its conceptions in 
 relation to perception, it is necessarily involved in an antinomy 
 which is absolutely insoluble. The conclusion, therefore, which 
 Kant draws from his analysis of knowledge is that a true 
 metaphysic must be reached by abstraction, not only from the 
 particular data of sense, but also from the forms of time and 
 space, which have application exclusively to these data. 
 ft^p;jJtt%, he I* ^S' however, to be observed, that this conclusion involves 
 inteiiSwo ^ really a twofold criticism : a criticism of our knowledge of 
 
 world as un- 
 real from the thmgs through the conceptions of the intelligence, as well as of 
 
 point of view o o r o > 
 
 of perception, q^jj. knowledge of phenomena through sense-perception, though 
 it is only to the latter criticism that Kant here specially directs 
 our attention. For while he shows that our knowledge through 
 perception is imperfect, inasmuch as it does not correspond to 
 the demands of intelligence, at the same time he shows that our 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 187 
 
 knowledge through conception is equally imperfect, since it can 
 never be realised in concreto, can never become a knowledge of 
 things in their individuality. Indeed, he expressly asserts 
 that " a thing which cannot be known in any intuition or per- 
 ception, is unthinkable, and therefore, impossible " ; ^ and he 
 declares that the only organ capable of grasping the absolute 
 reality would be " a pure intellectual intuition " which appre- 
 hended things not only generally through conceptions but in 
 their individuality, which, therefore, was not only active but 
 creative. But even to speak of such an " intellectual intuition," 
 is to go beyond a criticism of the perceptions of sense in view 
 of the conceptions of the intelligence ; it is to criticise both 
 these elements of our knowledge in view of a still higher ideal ; 
 it is to compare our dualistic mind, with its two disparate 
 faculties of intelligence and sense, with the idea of a mind in 
 which this difference does not exist or is overcome. Yet Kant 
 says that an intellectual intuition, which is not limited by the 
 forms of time and space, is a thing which " nuUo mentis conatu 
 nefingendo quidem assequi possumus," which we cannot set before 
 us m thought even as a possibility. It would seem, therefore, 
 that Kant here takes his stand at a point of view which he at 
 the same time declares to be impossible. Or he sets before us 
 as the idea of truth or of absolute knowledge a conception, 
 which he declares in the same breath to be made up of elements 
 that are for us irreconcileable. 
 
 The key to this inconsistency, an inconsistency which affects The source of 
 
 this incon- 
 
 the whole argument of the Critique, lies iu this, that Kant's ^t^^y- 
 method, or rather his professed method, — for in reality, as we 
 shall find, he is guided by a higher principle than that which he 
 explicitly" admits, — is one of abstraction. The conception of 
 existence in time and space is not one upon which we can rest 
 as final; or, to put it otherwise, such existence, when taken as a 
 res completa, falls into contradiction with itself. Thus space can- 
 not be conceived as limited and bounded by itself, but only as 
 • ^ Dissertation, § 25. 
 
188 INTRODUCTION. chap. 
 
 that to which limits or bounds are set by another. It is the 
 infinitely extensible, the infinitely divisible, not the infinitely 
 extended or the infinitely divided. It cannot, therefore, be con- 
 ceived as determining the world which is in it; for, as Kant points 
 out, there is nothing in space which affords a condition for deter- 
 mining the world to exist at any particular place, and there- 
 fore a determination of the world in reference to pure space is 
 utterly impossible. Nor again is there anything in time which 
 can supply a rational ground for the beginning or ending of 
 that which is represented as in time. Hence, when we take 
 the world as simply a world in time and space, we fall into an 
 antinomy, because we take as self-bounded and self-determined, 
 as a complete reality, that which yet is only an abstraction. 
 The world in time and space inust be taken as limited, and 
 there must be an end and a beginning to it, because we have 
 set it up as an independent whole, a complete reality in and 
 for itself And it cannot be conceived as so limited, because 
 of the very nature of time and space ; for they, ex hypothesi, 
 are the only things that limit it, and they can supply limits to 
 nothing, but rather require to be limited by something else 
 than themselves. 
 
 orSng°th'«'^ What inference, then, are we to draw from this antinomy, 
 difficulty. Qj. T^Q^ g^j.g ^^ ^Q ggj^yg -J. 2 ,j,^Q possible ways out of the 
 
 difficulty suggest themselves. We may suppose that the key 
 to the difficulty is to be found in the fact that time and space 
 are in themselves forms of sense, in which reality cannot be 
 adequately given, and that therefore, we must abstract from 
 both in order to reach the conception of things as they are. 
 Or we may say that in the conception of the world in space 
 and time, as a thing in itself, a res completa, something is left 
 out which is necessary to the determination of the world as a' 
 complete reality, and that it is just from this abstraction that 
 the difficulty arises. It is the former line of thought which 
 Kant follows when he declares that the mind, owinc to the 
 nature of its perception, cannot work out in the concrete the 
 
V- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 189 
 
 conception of reality whicH it yet must assert and assume in 
 the abstract ; and that " sensitive cogitata esse rerum representa- 
 tiones uti apparent, intellectualia autem sicuti sunt:" In other 
 words, Kant falls back by abstraction on that general concep- 
 tion of reality with which the idea of a world in space and 
 time was found not to agree, and through this general concep- 
 tion, taken by itself, he supposes that we have true knowledge 
 of things. At the same time that he does this, however, Kant 
 is forced to recognise that the knowledge thus attained is 
 imperfect, because merely general ; and that, therefore, the 
 element left out in it, the element of the particular, needs to 
 be restored in some other form. In other words, the abstract 
 conception taken by itself as the whole truth or reality would 
 be in contradiction with itself; it does not provide out of 
 itself the matter to which it shall be applied ; it does not limit, 
 or determine, or individualise itself. Just, therefore, as the 
 world could not be conceived as limited and determined by 
 space and time, since space and time require themselves to be 
 limited and determined by something else, so, on the other 
 hand, the limits or determinations by which phenomena are 
 brought to the unity of conception, cannot be regarded as 
 realities independent of the matter which they determine. 
 Hence a further abstraction becomes necessary. The reality 
 of .things, since it cannot be apprehended by our divided 
 intelligence either through perception or through conception, 
 either through sense or through understanding, must be appre- 
 hended by an intuitive intelligence, for which the division of 
 the two does not exist, or by which it is transcended ; i.e., it 
 can be apprehended only by an intelligence which is entirely 
 different from ours. 
 
 This may be put more clearly as follows. The process of Eesuitsof 
 
 Kant's method 
 
 knowledge, as Kant represents it, is a process in which limits "'abstraction. 
 and determinations of thought are applied to that which, in 
 itself, is unlimited and undetermined, because given as in space 
 and time. In this process, however, it is found that the intel- 
 
190 INTRODUCTION. chap. 
 
 ligence can never satisfy itself; for it can never attain to a 
 limit either in analysis or in synthesis : it can reach neither 
 simple unity nor totality in space ; it can reach neither begin- 
 ning nor end in time. Hence conception condemns perception. 
 But on the other hand, conception is insufficient in itself 
 because of its generality, and it is only to that which is other 
 than itself that it can be applied; it is only in bringing 
 that which is given in the difference of space and time to the 
 unity of thought, that the conception can find a meaning. 
 Hence perception equally condemns conception. Or rather, 
 criticism, which took its stand at first upon conception to 
 condemn perception, and then upon perception to condemn 
 conception, is now forced to make a still further regress, and 
 to condemn the dualism of perception and conception in view 
 of a unity of intellectual intuition in which that difference 
 altogether disappears. Such intellectual intuition, however, — 
 reached as it is by abstraction from the dualistic mode of 
 thought, in which conception and perception are essentially 
 opposed, — must immediately be recognised as an empty idea, 
 even the possibility of which we cannot imagine. For in 
 speaking of it, we are simply giving to an unknowable unity 
 a name, made up of two terms from the difference of which 
 abstraction is to be made, while yet we recognise that these 
 two terms are absolutely irreconcileable and that, therefore, 
 when we abstract from their differences, we leave nothing 
 behind. 
 His apparent In the above, we have been following Kant's logic by which 
 
 process of 
 
 atsto^tion i^e first abstracts from perception in its difference from concep- 
 pfSs." tion, then from conception in its difference from perception, and 
 finally ends in an abstract unity, of which logically nothing 
 can be said, but which he is obliged to name by combining the 
 two. If, however, we realise the meaning of this process, we 
 shall see that it is not correctly described by Kant himself, and 
 that the apparent abstraction hides a process which is really, if 
 we may use the term, one of concretion, or one by which the 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 191 
 
 abstraction of our first raode of thought is corrected. This is 
 a very important point ; for an understanding of it will enable 
 us to see just how far Kant reached a true idea of criticism, 
 and how far his view requires to be modified in order to 
 correspond to such a true idea. 
 
 A point of view from which criticism is possible cannot be critieiBm 
 
 possible only 
 
 one reached by mere abstraction. Tor abstraction does not from » pom* 
 
 " of view less 
 
 give us any universal which can cast a new light upon the thafwMoh™ 
 particulars ; it cannot raise us to any point of view really 
 higher than that which we have previously reached in our 
 determination of the particulars. It sitnply empties our ideas 
 of • their content, without really increasing their extension. It 
 is only in so far as the process of abstraction, while it lifts us 
 beyond the sphere of an opposition which seemed absolute, 
 brings into view a principle of unity which was hidden beneath 
 the difference of its terms, that it can enable us to criticise our 
 former conceptions. Now, it is just this point that Kant 
 seems to lose sight of in his first attempt to criticise the way 
 in which we represent the world in time and space. In that 
 representation he finds an antinomy between the requirements 
 of conception and the possibilities of perception And his 
 first solution of the antinomy is simply to abstract from the 
 latter and assert the former as absolute. But almost un- 
 consciously he is driven by the immanent dialectic of his 
 thought to recognise, what he afterwards stated explicitly in 
 the Critique,, that if perception without conception is blind, 
 conception without perception is empty. Hence, in sep- 
 arating conception from perception he has reached no higher 
 point of view, from which he can criticise the ordinary 
 empirical unity of the two. He is, therefore, forced to take 
 refuge in a still higher abstraction, which, however, as a mere 
 abstraction, cannot really supply him with any better point of 
 view for criticism. On the contrary, as it is an abstraction 
 not only from perception but also from conception, it neces- 
 siirily vanishes into a pure identity, which is unthinkable 
 
192 INTRODUCTION. chap. 
 
 as well as unimaginable. Kant's apparent process of abstrac- 
 tion, however, really has its value just in this, that it involves 
 more than it expresses. For while the attempt to sever 
 conception from perception shows that though opposed they 
 cannot be separated, or that their opposition is merely relative, 
 it reveals that there is an intelligible unity on which criticism 
 can take its stand, and in reference to which it can test our first 
 im/perfect view of things, and overcome the antagonism involved 
 in it. Or, to put the same thing in a different way, it shows 
 that the view of the world which gave rise to the difficulties 
 and antinomies in question was based on an abstraction ; and 
 that when we restore the element abstracted from, these 
 difficulties and antinomies are capable of solution. 
 How conre- This may be expressed more simply by saying that the real 
 obj^ctto'the source of the difficulties which perplexed Kant's mind lay in 
 beunderKd. the fact that the world in time and space was taken as a thing 
 in itself, which existed without reference to the thinking 
 subject ; and that the solution of them lay in the discovery 
 that that world is phenomenal and not noumenal, i.e., that 
 it cannot be conceived except as existing for a self. In saying 
 this, however, it is to be observed that we are giving a sense 
 to the contrast of phenomenal and noumenal which is not 
 recognised by Kant, or which we can find in Kant only when, 
 according to the reasoning of the last paragraph, we make 
 explicit what rather lies behind his thought than is directly 
 contained in it. To Kant the noumenon was, in the first 
 instance, the thing without relation to our perception of it, the 
 thing as it is to pure conception. In the second instance, 
 when the modes of conception also were condemned as in 
 themselves empty and, therefore, as deriving their whole value 
 from their application to some kind of perception, the noumenon 
 ceased for him to be the object of conception, and became 
 simply the object of an intuitive understanding. But this 
 idea, as it was reached in Kant by abstraction both from 
 conception and from perception, became equivalent to the 
 
V- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 193 
 
 conception of a thing in itself without reference to any 
 
 intelligence ; or, what is the same thing from this point of 
 
 view, a thing which is not distinguished from the intelligence 
 
 for which it is. For it is all the same whether we risgard the 
 
 thing as absolutely repelled from, or as absolutely identified 
 
 with, the intelligence that apprehends it. In both cases the 
 
 unity-in-difference of the knowable object and the knowing 
 
 subject is lost in a pure identity of which nothing can be said. 
 
 In both cases the real object or noumenon is reduced to the 
 
 thing as related to nothing but itself, while the phenomenon 
 
 is the thing as related to the intelligence. On the other hand, 
 
 if we follow the line of thought above indicated, the opposition 
 
 of reality to knowledge disappears. And the error of the 
 
 ordinary view of things is seen to lie in this, that it takes the 
 
 object as a thing in itself apart from, and unrelated to, the 
 
 intelligence. It is this error, the error of what we may call a 
 
 natural abstraction, which gives rise to all the difficulties and 
 
 antinomies previously spoken of; for they all originate in the 
 
 attempt to treat as a res oompleta what is not really a res 
 
 completa. In order, therefore, to solve these antinomies what 
 
 we require is, first of all, to recognise the abstraction which such 
 
 an attempt involves, i.e., the abstraction, in our determination 
 
 of the object, from the consciousness for which it is. From 
 
 this point of view, it may be said that what we at first take 
 
 for the thing in itself is a phenomenon or existence for another, 
 
 i.e., for the self; and that, on the other hand, when we have 
 
 recognised it to be a phenomenon or existence for another, we 
 
 have begun to apprehend it as a noumenon, i.e., as what it 
 
 really is in itself. The recognition that consciousness is a 
 
 necessary element in all that is for it, and that existence is 
 
 essentially existence-for-a-self, is at once the discovery that the 
 
 object of knowledge is phenomenal, and it is the discovery of 
 
 the noumenon of which it is phenomei^al. For to recognise 
 
 that all existence is existence for a self is to adopt a prin- 
 
 ciplig, the natural outcome or complement of which is the 
 VOL. I. N 
 
194 
 
 INTRODUCTIOlir. 
 
 CHAP. 
 
 Differences 
 between the 
 Dissertation 
 and the 
 Ontique of 
 Pure Re<i80n. 
 
 1772— Kant's 
 new difficulty 
 as to the pure 
 conceptions. 
 How can they 
 be objectively 
 valid 
 
 doctrine that all existence is the manifestation of a self 
 This, as above said, is in a sense to invert the use of the 
 conceptions of noumenon and phenomenon which we find in 
 Kant ; but it will be one of the objects of this book to show 
 that it is in this inversion that we discover the essential 
 meaning of Kant's work. 
 
 The course of the preceding argument is probably already 
 sufficient to show the main motive or inner necessity of the 
 course of thought by which Kant found his way from the ideas 
 of the Dissertation to those of the Critique,. There are two 
 great differences which separate the latter from the former. 
 The first is that in the Critique Kant ceases to regard the 
 a priori conceptions of the mind as determinations of things in 
 themselves, and regards them only as forms by which the data 
 of sense under the forms of sense are determined as phenomena 
 or objects of experience. The second, which is a necessary 
 consequence of this, is that, while in the Dissertation these con- 
 ceptions are regarded as expressing the ways in which God, as 
 the first cause of the existence of all finite substances, also 
 unites them into one world of which all parts are in mutuo 
 commercio ; in the Critique they are regarded as categories by 
 which the conscious self combines all the data of sense in 
 relation to itself into one experience. A few words of 
 explanation in regard to each of these points will enable us to 
 bring clearly before us the whole motive of the transition. 
 
 1. We have already seen how natural was the course of 
 thought by which Kant was led to give up his first view that 
 the a priori conceptions of the mind reveal to us the ultimate 
 reality of things. We have seen that even in the Dissertation 
 itself this transition is already made, in effect if not explicitly, 
 when Kant speaks of the ultimate reality as the object, not 
 of our understanding, but of a perceptive or intuitive under- 
 standing. For, as an intuitive understanding is one for which 
 the division of conception and perception does not exist, this 
 already implies that a priori conceptions have value only in 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 195 
 
 relation to perceptions given independently of them, and in our 
 case, therefore, only in reference to perceptions given under 
 the forms of space and time. This thought, however, did not 
 distinctly suggest itself till some time after the Dissertation was 
 written, and according to Kant's own statement, it arose in his 
 mind in connexion with the recollection of Hume's argument 
 in relation to causality. The first step in this direction, how- 
 ever, seeins to have been due to the development of Kant's own 
 thought, as appears from a letter to Herz, the date of which is 
 2 1st February, 1772, about two years after the Dissertation. In 
 that letter he announces that he has had before him for some 
 time the plan of a work to be called The Limits of Sense and 
 Reason, but that in attempting to write it, he had discovered 
 that in all his own previous metaphysical investigations, as 
 well as in those of others, an essential element was wanting,-^- 
 an element . which, if discovered, would supply the key to the 
 whole secret of metaphysic. " For," he goes on, " I put this 
 question to myself. On what ground rests the relation of that 
 in us which we call an idea to objects ? If all that the idea 
 contains is simply the mode in which the subject is affected by 
 an object, we may easily understand how it should correspond 
 to that object as an effect to its cause, and how, therefore, the 
 determination of our mind should be capable of giving us an 
 idea of something, or in other words, how it should be possible 
 for an idea of our mind to have an object. Thus it is 
 intelligible how our ideas, so far as they are sensuous affections 
 passively received, should have a relation to objects, and also 
 how the forms of sense, though borrowed from the nature of 
 our soul, should nevertheless apply to all things in so far as 
 they are presented in sense. On the other hand, if that in 
 us which is called an idea, were in relation to the object an 
 activity, that is to say, if the object itself were produced by the 
 idea, (as it is supposed that the ideas in the Divine Mind are 
 the archetypes of things,) then, in this way also, the conformity 
 ofWdeas with\objects might be understood. We can thus dis- 
 
196 INTRODUCTION. chap. 
 
 cern at least the possibility of two kinds of intelligence ; of 
 an intdlectus archetypus, whose perception itself should be the 
 ground for the existence of things, and of an intdlectus ectypus, 
 which should derive all the data of its thinking from the 
 impressions of sense. But our intelligence does not fall under 
 either description. It is not, if we leave moral ends out of 
 account, the cause of the objects it apprehends, nor are these 
 objects the causes of its conceptions {in sensu reali). Thus the 
 pure conceptions of the understanding cannot be abstracted 
 from the feelings of sense, nor are they merely the expression 
 of the character of our sensitive receptivity. They have their 
 sources, indeed, in the nature of the soul, but they are neither 
 the result of the action of the object upon it nor do they pro- 
 duce the object. In my Dissertation, I was content to explain 
 their nature in a negative way, and to say only that they are 
 not modifications of the soul produced by the object. But now I 
 must ask in what other way an idea is possible, which refers to 
 an object, without being the effect of an impression from that 
 object ? I ventured in the Dissertation to say, that the ideas of 
 sense represent things as they appear, while the conceptions of 
 the understanding represent things as they are. But how can the 
 ideas of these things be given to us, if not by the manner in 
 which they affect us ? Or, if the pure conceptions of them are 
 due to our own inner activity, whence comes the agreement 
 which they are supposed to have with objects, which yet are not 
 their products ? How can pure reason lay down axioms about 
 things without any experience of them ? In Mathematics, the 
 difficulty is not great, because objects are only capable of being, 
 or being represented as, quantities, in so far as we produce the 
 idea of them by repeating a unit for a number of times. The 
 mind is active in generating the conception of quantity, and, 
 therefore, we can see how the principles of quantity should be 
 determined a priori ; but when we ask how the understanding, 
 by a pure a priori process, can form to itself conceptions of 
 things in their qualitative determination, or lay down real 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 197 
 
 principles as to their possibility, principles which are not derived 
 from experience, hut to which experience must exaetly conform, — 
 we ask a question over which the greatest obscurity has 
 hitherto rested." 
 
 " Plato assumed a prenatal, Malebranche a present, intuition Dogmatic 
 
 solutions of the 
 
 of the Divine Being, as the source of the pure notions and difficulty. 
 principles of the understanding; and various moralists have 
 adopted a similar hypothesis to account for our primary ideas 
 of moral law. Crusius, on the other hand, spoke of certain 
 rules of judgment and certain conceptions, which God has 
 implanted in the human soul in exact conformity with the 
 constitution of things. The former writers, therefore, based 
 their systems on a supernatural influence, the latter on an 
 intellectual pre-established harmony. But such a Beus ex 
 machina is the most extravagant resource to which a philosopher 
 can betake himself in explaining the origin and value of our 
 knowledge : for it not only involves a palpable case of reasoning 
 in a circle, but also stimulates and encourages all the whims of 
 pious fancy and all the dreams of baseless speculation." ^^^.^^^''^ 
 " While in this way I was searching out the sources of Kant's first 
 
 •^ ■ ^ attempt to 
 
 intellectual knowledge, without which we cannot determine the of the"^ * ^*°* 
 nature and limits of Metaphysic, I succeeded in making a dis- °* ^^ones. 
 tinct division of the parts of this science ; and I sought further 
 to gather Transcendental Philosophy, or, in other words, all the 
 notions of pure understanding into a certain number of Cate- 
 gories. Nor did I follow the manner of Aristotle, who simply 
 set them down one after another as they occurred to him in his 
 ten Predicaments, but I aimed at a systematic classification, 
 determined by a few fundamental principles. Without, how- 
 ever, going into any further detail, I may say that the essential 
 part of my task is now done, and that I am in a position to lay 
 before the public a Criticism of Pure Reason, which explains 
 the nature of truth, both theoretical and practical, in so far as 
 
 'An allusiop to his own comparison of the metaphysic of Leibniz to the 
 dr^ms of Swedenborg. 
 
198 INTRODUCTION. chap. 
 
 it is derived purely from the intelligence : and I expect tq 
 complete and publish the first part of this system, containing 
 an account of the sources of Metaphysic, its methods and its 
 limits, within about three months." 
 Question of In this passage Kant draws attention to a defect of the pure 
 
 the objective jr o j. 
 
 purSJep-''^ conceptions, which is substantially identical with that which 
 he had already detected in the Dissertation. For there he had 
 observed that these conceptions are general, and that they can 
 be particularised and individualised only by the aid of percep- 
 tions drawn from sense ; and this had led him to set up the idea 
 of an intuitive understanding as that which alone apprehends 
 objects as they are. Here, in like manner, he points out that 
 pure conceptions as such, are subjective, and that they cannot 
 be determined as objective, unless objects are given which can 
 be brought under them ; and, further, that this difficulty would 
 not exist if our understanding were intuitive and therefore 
 creative ; for then there would be no division between the 
 general conception and the particulars to be subsumed under 
 it. In other words, the distinction of perception and concep- 
 tion made in the Dissertation, necessarily gives rise to a 
 difficulty as to their combination, a difficulty which now re- 
 quires to be dealt with. In the Dissertation, that difficulty 
 had not clearly emerged for two reasons ; first, because 
 Kant was too much occupied with establishing the dis- 
 tinction of perception and the forms of perception from con- 
 ception and its forms, to pay much attention to their relation ; 
 secondly, because he had not yet clearly seen that the pure 
 conceptions can find their realisation or objective meaning 
 only in their application to the perceptions of sense. Nay, he 
 had there spoken of them as determining things in themselves, 
 though he had implicitly retracted this view by setting up the 
 ideal of an intuitive understanding. At the time of the letter 
 to Herz, however, this new aspect of the case had dawned 
 upon him, and had brought with it the great problem of the 
 Critique, how conceptions, which are not due to experience, 
 
V- DEVELOPMENT OE THE CRITICAL PHILOSOPHY. 199. 
 
 should yet be conceptions to which experience must con- 
 form.^ 
 
 Now, it may be observed by anticipation that the Critique Distinction 
 
 and relation 
 
 from beginnins to end keeps its way between what we may »' perception 
 
 o o f J J and conception 
 
 call two opposite poles of thought, never losing sight of either, *^t'>ecw«i9«e- 
 though the attempt to steer by both often brings it into real 
 or apparent contradiction with itself. In the iirst place, the 
 Critiqvs holds by the thought of the Dissertation that concep- 
 tion and perception, understanding and sense, are essentially 
 disparate, and that indeed in their union they necessarily give 
 rise to certain antinomies, which we can solve only by point- 
 ing out their source. In the second place, it holds with equal 
 firmness by the thought that experience, i.e., all the knowledge 
 of objects that we possess, is based upon this very union of 
 perception with conception, or upon the determination of the 
 former by the latter. Experience is thus the combination of 
 two elements which can never be resolved into one, two , 
 elements for the differences of which we can never find a 
 principle of unity. It is the necessary synthesis of essentially 
 heterogeneous factors, which imply and demand each other as 
 ■factors of experience, but which, nevertheless, when brought 
 together, repel each other and will not be fused into a perfect 
 unity. In the end, therefore, when i^ is carried out to its 
 ultimate results, the combination breaks down, and its dispar- 
 ate elements give rise to antinomies ; which, however, do not 
 cast doubt on experience, but only show that the knowledge it 
 gives us is a knowledge of phenomena and not of things in 
 themselves. For, in contrast with this empirical union of 
 disparate elements, thought, which seeks unity in experience 
 and cannot find it, necessarily falls back upon an idea derived 
 from itself, in which it cannot but recognise the necessary 
 characteristics of the ultimate reality. This idea is that which 
 we have already spoken of as the idea of intuitive understanding, 
 or of its object, viz., the thing in itself or noumenon. It is for 
 • 1 Observe the sentence italicised in the letter to Herz. 
 
200 INTRODUCTION. chap. 
 
 Kant, as we have seen, the ultimate abstraction, which we reach 
 by negation of conception in its difference from perception, as 
 well as of perception in its difference from conception. Whether 
 this view of it can be maintained, or whether, as we have 
 already hinted, there is another principle which really guides 
 Kant's thought though it may not be consciously present to 
 himself, we need not here discuss. It is, however, necessary 
 so far to anticipate what follows as to mention that, in Kant's 
 view, the assurance which we have that the idea of reason, the 
 idea of the noumenon, is or represents the ultimate truth of 
 things, is due not to the speculative, but to the practical, reason. 
 Kant does not In the letter to Herz quoted above, the general line of 
 
 yet speak of 
 
 ^e^Kmitation investigation, upon which Kant had entered and which was to 
 SpTrience!" terminate only in the Critique, is clearly indicated. The 
 great question for him was to explain how pure conceptions 
 can have an objective meaning, and especially how they can 
 have an objective meaning in relation to objects given in 
 experience through sense. How, he asks, " can my under- 
 standing produce real principles as to the possibility of things 
 with which experience must agree, though they are indepen- 
 dent of experience " ? Kant as yet says nothing directly &i 
 to the limitation of these principles to experience or, in other 
 words, to phenomena given through sense, and we have 
 evideiice that for some time after the letter to Herz this 
 limitation was not recognised by Kant.^ In fact, this limita- 
 tion could not become clear to him till he had advanced some 
 way towards his subsequent deduction of the categories. For 
 it was the perception that a priori synthesis is possible only 
 through the determination of the forms of sense by the unity 
 of thought as expressed in the categories, which enabled Kant 
 at once to vindicate the a priori principles as principles of 
 experience and to limit them to experience. 
 
 ^ Cf. Dr. B. Erdmann's remarks in the Phil. Monatsheft, XIX. 141 seq., 
 and the quotations there given from Politz's edition of Kant's lectures on 
 Metaphyslo. Cf. also Phil. Monatsheft, XX. 89 seq. 
 
V- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 201 
 
 Now how was this step rendered possible ? The answer to g;J™ '^4^^ 
 this question will become easier, if we first ask ourselves i™iumbera™r 
 another question, viz., what was the point at which Kant re- 
 ceived that decisive influence from Hume, of which he so often 
 speaks ? ^ Let us look first at what we may call the locus 
 classicus on this subject : — 
 
 " I confess it freely," he declares in the introduction to the 
 Prolegomena, " the remembrance of David Hume was the first 
 thing which many years ago interrupted my dogmatic slumber, 
 and gave to my investigations in the field of speculative philo- 
 sophy a quite new direction. I was far, indeed, from following 
 him in the inferences which he drew, and which were so drawn 
 by him just because he did not take the problem he had to deal 
 with in all its generality, but only stumbled upon one aspect , 
 of it, which without reference to the whole could give him no 
 secure basis of reasoning. Now, when we start with a thought 
 that another has bequeathed to us, a thought well-grounded but 
 not fully developed, we may have good hope by persevering 
 reflexion to get further with it than the acute writer to whom 
 we owe the first spark of light. I, therefore, set myself to 
 inquire whether Hume's objections admitted of being univer- 
 salised ; and I was not long in finding that the conception of 
 the connexion of cause and effect is by no means the only one 
 under which the understanding represents to itself a priori the 
 connexions of objects, but that on the contrary, all metaphysic 
 consists of such connexions. I sought, therefore, to assure 
 myself of their number, and when I had succeeded to my wish 
 in doing this, i.e., when I had succeeded in determining them 
 all from one single principle, I then proceeded to the deduction 
 of them," i.e., to the justification of their use in relation to 
 objects ; " for I had already assured myself that they were not, 
 as Hume suspected, derived from experience but that they had 
 
 1 The view that it was at this period that the influence of Hume upon Kant 
 is to be dated, is maintained with great force by Dr. B. Erdmann in his intro- 
 duction to the Prolegomena and again in a, recent article in the ArchivfiXr 
 Phitbsophie. 
 
202 INTRODUCTION. ^H-*-^- 
 
 sprung from the pure understanding itself. This deduction, 
 which had appeared impossible to my acute predecessor, and 
 the very idea of which had never occurred to any one else, — 
 though every one confidently employed the conceptions in 
 question without once asking for the grounds of their objective 
 validity, — this deduction, I say, was the most difficult task 
 that could ever be undertaken for the behoof of metaphysic. 
 And what made it still harder was, that the metaphysic already 
 in existence could not give me the slightest help, since my aim 
 was nothing less than for the first time to determine the possi- 
 bility of metaphysic. Now, when I had succeeded in solving 
 Hume's problem, not only in a single case, but iu relation to 
 the whole faculty of pure reason, I could- proceed to take secure, 
 though as yet but slow steps towards what was my ultimate 
 end, viz., the complete determination, according to universal 
 principles, of the whole compass of pure reason in its limits as 
 well as in its content : which was all that metaphysic stood in 
 need of in order to develop its system on a safe and certain 
 plan." 
 ?iso"Son of ^^ *^^^ passage Kant tells us that the " remembrance " of 
 oauAiity? David Hume broke his dogmatic slumbers by suggesting a 
 thought, the sceptical consequences of which, as drawn by 
 Hume himself, he at once rejected. At the same time, in 
 rejecting those consequences, he generalised the objection from 
 which they flowed, and was therefore obliged to find a general 
 answer to it. He was thus led into a line of inquiry, the 
 result of which was to discover the list of a priori conceptions, 
 and at once to vindicate their objective validity and to limit 
 it to experience. IsTow if we turn to Hume's Essays}^ (the only 
 work of his with which Kant was acquainted,) we find that 
 Hume starts with the " proposition, which will not admit of 
 much dispute, that all our ideas are nothing but copies of our 
 impressions, or, in other words, that it is impossible for us to 
 
 ^ Hume's Essays, edited by Green and Grose, II. 50 seq. : Inquiry concern- 
 ing the Human Understanding, § 7. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 203 
 
 think of anything, which we have not antecedently /eZ^ either 
 by our external or internal senses." But, he goes on to argue, 
 there is nothing in the data of inner or outer sense which 
 suggests the idea of a necessary connexion between the par- 
 ticular phenomena which we regard as causes and effects of 
 each other. As regards outer experience, " solidity, extension, 
 motion, these qualities are all complete in themselves and 
 never point to any other event which may result from them." 
 " We know that as a fact, heat is a constant attendant of 
 flame: but what is the connexion between them, we have no 
 room so much as to conjecture or imagine." And as regards 
 inner experience, the case is not different. " Were we em- 
 powered by a secret .wish to remove mountains or to control 
 the planets in their orbit, this power would not be more extra- 
 ordinary or beyond our comprehension," than the fact that by 
 our will we can move our limbs. Hence there is nothing in 
 the first experience of any phenomenon to lead us to expect 
 that it should have a particular consequent ; and the experience 
 that it actually has such a consequent is not an " impression," 
 from which the idea of necessary connexion can be derived. 
 The only thing from which we can derive that idea is the 
 " new sentiment," which arises in consequence of the frequent 
 repetition of cases " in which the same object is followed by 
 the same event," i.e., " the customary connexion in thought 
 between an object and its usual attendant." 
 
 Now, in the last chapter I pointed out that in all this Hume Kant's repre- 
 
 ' -^ ■*■ aentation of 
 
 is not arguing against the idea of a synthetic power of thought ; ^g^e^t. 
 for he excludes that idea at the outset by the reduction of 
 ideas to copies of impressions. He speaks, indeed, of the rela- 
 tion of causes and effects as incomprehensible— in conformity 
 with his general view of experience as consisting of isolated 
 impressions, which have no relations except the external rela- 
 tions of time and space. But his object is, not to bring 
 arguments against the doctrine that" the intelligence possesses 
 a ^nthetic power, but simply to show that in the impressions as 
 
204 INTEODUCTION. <^'^^P- 
 
 given there is no hint or suggestion of " power or necessary 
 connexion." When, however, we turn to Kant's words, we 
 find that he connects these two ideas and reads Hume as if he 
 were arguing against the idea of an a priori synthetic power of 
 thought. " Hume," he declares, " started from a single but 
 very important conception of metaphysic, the conception of the 
 connexion of cause an d effect, afiihe required reason, which pre- 
 tends to have produced that conception in its own bosom, to 
 submit to question. He demanded, in short, that reason should 
 be forced explicitly to state what justification it has for 
 thinking that anything can be so constituted, that, if it is posited, 
 something else must necessarily be posited in consequence: for 
 that is what is expressed in the conception of cause. He thus 
 proved beyond contradiction that it is quite impossible for 
 reason, a 'priori and out of conceptions, to establish such a 
 connexion, a connexion which involves necessity : for it is 
 quite impossible to understand how, because one thing is, 
 something else must necessarily be, and consequently how the 
 conception of such a connexion can be attained a priori. Hence 
 he concluded that reason altogether deceives itself with this 
 idea, falsely regarding it as its own offspring, when it is 
 nothing but a bastard-child of imagination, begotten upon 
 it by experience. For the whole secret is that imagination, 
 having brought certain ideas under the law of association, tries 
 to pass off the subjective necessity arising from such associa- 
 tion, i.e., a mere habit of the mind itself, as if it were an 
 objective necessity of thought." ^ 
 
 Kant thus reads Hume's argument against the possibility of 
 deriving the idea of necessary connexion from experience, as if 
 "it were intended to disprove the existence of a synthetic power 
 in thought: a power which, as belonging to pwre thought, 
 must establish connexions of ideas which are necessary and 
 universal. Now, as already remarked, Hume did not seek to 
 prove that thought is not synthetic ; he assumed it to begin with. 
 ^ Prolegomena, Introduction. 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 205 
 
 It is the premise from which he starts, and indeed the presup- 
 position of the whole school of Locke to which he belongs, that 
 thought cannot legitimately add to the data given : and Kant 
 refutes this presupposition only in so far as his transcendental 
 deduction cuts away the ground from Hume's explanation of the 
 origin of the principle of causality, by proving that the particu- 
 lar experiences from which Hume derives that principle could 
 not exist without it. But, in the first instance, Kant's argument 
 is directed not against Hume's argument but against his presup- 
 position that the mind cannot add to the data given to it ; and 
 he sees now, what he did not see when he wrote the Essay on 
 Negative Quantity, that it is to add to these data to bring them 
 under the general principle of causality or to assert that that 
 principle has objective validity. Hume awakened Kant 
 from his dogmatic slumbers by making him realise that the 
 objective application of the conceptions of pure understanding, 
 the possibility of which is the problem of the letter to Herz, 
 involves a synthesis which, as a priori, must be universal and 
 necessary, and cannot therefore find its evidence in the parti- 
 cular experiences which fall under it. But, when this difficulty 
 was suggested, the point which by his own mental history 
 Kant was led first of all to consider was not, — what Hume 
 sought to prove, — the impossibility of deriving necessary 
 principles of synthesis from experience, but the difficulty of 
 conceiving how pure thought could become synthetic in rela- 
 tion to experience. For, up to the year 1766 at least, Kant 
 had been endeavouring, in opposition to the philosophy of 
 "Wolff, to exhibit the difference between the synthetic con- 
 nexion of reality and the analytic process of thought. And 
 though in the Dissertation, he had maintained that the 
 intelligence brings with it not only formal but also real 
 principles of knowledge, yet the question of the application 
 of these principles to the data of sense had not yet 
 suggested itself. N"ow, however, Kant is forced to consider 
 this question, and the result is that he recurs to the doc- 
 
206 INTEODUCTIOK chap. 
 
 trine of his earlier essays that thought is in itself analytic, 
 but at the same time maintains that it becomes synthetic in 
 relation to the forms and matter of sense. And by a natural 
 illusion he regards Hume as discussing the whole problem thus 
 suggested. Hence Kant seems to connect the name of Hume 
 with a question which really arose out of the Wolffian philo- 
 sophy—the question of the possibility of a synthesis of pure 
 thought ; and not merely with the problem which Hume 
 really considered — the problem of the possibility of extracting 
 necessary principles out of experience : the necessity and univer- 
 sality of the principles in question being for Kant a conse- 
 quence of the a priori character of the synthesis. It is, how- 
 ever, the transcendental deduction, in which the a priori 
 principles are shown to be presupposed in experience and 
 not derived from it, which alone contains Kant's direct 
 answer to the argument of Hume's Assays; while his ex- 
 planation of the way in which thought becomes capable of 
 an ampliative or synthetic movement is really an answer to a 
 question which had been suggested, not by Hume but by the 
 Wolffian philosophy.! 
 to flystema- Passiug, howevcr, from the question, as to the exact relation 
 categories. ° of Kant's thought to that of Hume, we can see that from the 
 point he had reached, there were two special problems which 
 opened up before him. On the one hand, he had to consider 
 how, consistently with its essential nature, pure thought could 
 become synthetic in relation to experience ; and on the other 
 hand, he had to show how experience, consistently with iis„„ 
 essential nature, could be proved to conform to the principles 
 thus developed. As a first step towards the solution of the 
 
 1 It may be said to be an answer to Hume in so far as Kant substitutes au a 
 prioi-i synthesis for the analysis of ideas derived from experience. But when 
 Kant says that Hume "required reason, which pretended to have produced 
 that conception," i.e., the conception of causality, "in its own bosom, to sub- 
 mit to question," he is not stating the problem as Hume would have stated it, 
 but as he himself was led to state it by his previous criticism of the Wolffian 
 philosophy. 
 
V- DEVELOPMENT OY THE CRITICAL PHILOSOPHY. 207 
 
 former of these problems, Kant sought to make a complete list 
 of- the a priori principles, or rather of the pure conceptions 
 underlying them. He, therefore, looked about for some 
 principle which should not only furnish a criterion of such 
 conceptions, but which should guide him to the discover)- 
 of them and enable him to make sure that the list so dis- 
 covered was complete. Now, the reference ,to Aristotle in 
 the letter to Herz shows where he first looked for guidance, 
 and this idea is confirmed by what he says in the Pfolegomena} 
 • As he there tells us, he took Aristotle's list of categories for an 
 attempt to realise what he wanted, and in the light of his own doc- 
 trine as to time and space, he was able at once to cast aside three 
 of the categories contained in the list^ — ubi, quando, and sit'us. 
 But even as thus purified, the list was of no use to him, for he 
 found in it " no principle according to which the understanding 
 could be measured out, and all the functions out of which 
 its pure conceptions arise could be precisely and completely 
 determined." He then goes on to say that " in order to find 
 such a principle, he looked for an act of the understanding, which 
 contains all its other acts, and which differentiates itself into a 
 series of different acts only in virtue of the different modifica- 
 tions or moments of the process, whereby the manifold which 
 is before the mind is brought under the unity of thought.'' 
 fThis act he found to be the act of judging. But in the 
 analysis of judgment, which was thus determined to be the 
 essential activity of the understanding, he had before him for 
 
 his guidance the work which was already done by the LogicianSi 
 I- 
 a work " perhaps not without defects, but sufficiently accurate 
 
 to enable him, on the basis of it, to set forth a complete table 
 of the pure functions of the understanding." These functions, 
 however, as treated by formal Logic, were without any deter- 
 mination in relation to objects, and he, therefore, proceeded 
 " to refer the functions for judgment to objects in general, or 
 rather to the conditions under which judgments are determined 
 • ^ Prolegomena, % 39. 
 
208 INTRODUCTION. chap. 
 
 as objectively valid " ; and this supplied him with the required 
 catalogue of the conceptions of the mind, a catalogue of which 
 " he could be certain that it contained all the elements of our 
 knowledge of things which are derived from the pure under- 
 standing, no more and no less." , The result of this method of 
 discovery of the categories was to determine for him at the 
 same time their true meaning and validity ; for it showed that 
 " in themselves they are mere logical functions involving no 
 conception of any object whatsoever," and that they get objective 
 significance only through the sense-perceptions, to which they 
 are applied and in application to which they serve only to 
 " give universal validity to empirical judgments, which in 
 themselves are undetermined in view of all the functions for 
 judgment." ^ Kant goes on further to say that the effect of 
 this systematic process was, in the first place, "to furnish a 
 systematic method, on which every object of pure reason could 
 be treated ;" and, in the second place, to separate from the 
 categories two other classes of conceptions," which were apt to 
 be confounded with them, viz., the " reflective conceptions," 
 and, what is more important, the Ideas of reason, for which a 
 quite different source has to be sought. 
 ?uch™ii°t' "' I* appears, then, that the inquiries out of which the Critique 
 oli™naiysis°of arosc procccded on this fashion. The separation of the pure 
 
 Judgment. . . 
 
 conceptions of the mmd from perceptions and especially from 
 the forms of perception, which was accomplished in the Disserta- 
 tion, led, so soon as it was seen that the conceptions could have 
 objective validity only as applied to those very perceptions 
 from which they were so distinguished, to an inquiry into the 
 possibility and the conditions of this application. This inquiry, 
 however, in Kant's view, could be satisfactorily carried on only 
 after a complete list of the pure conceptions, based upon the 
 nature of the synthetic activity which they subserve, had been 
 
 ' This, of course, refers to the distinction made in the Prolegomena between 
 judgments of perception and judgments of experience,— a distinction which 
 will be criticised in the sequel. 
 
 Mi 
 
y- DEVELOPMENT OP THE CRITICAL PHILOSOPHY. 209 
 
 discovered. Now, that activity must be conceived as a process 
 in which thought goes beyond itself, beyond its own pure 
 analytic unity, to unite with itself something not already com- 
 bined in that unity. In other words, it is a process oi judgment, 
 in which different data of sense are brought together under one 
 conception and so united with the consciousness of self. And 
 in the last resort the understanding itself must provide the con- 
 ceptions by which it gives unity to such differences. Hence, 
 if there are pure a priori conceptions which the mind derives 
 from itself and by the aid of which it unites empirical elements 
 with each other and with itself, they must be involved in, and 
 constitutive of, the activity of judging. They can, in short, only 
 be different "moments" or aspects which become distinguishable 
 in the unity of the understanding when we consider it in this 
 its essential and characteristic activity. But judgment had 
 been thoroughly examined by the Logicians, who had dealt 
 with it as a purely formal process of analysing ideas, and in 
 this view of^it had shown it to have certain aspects, summed 
 up under the heads of quantity, quality, relation, and modality, 
 with their subordinate species. This result, then, Kant took 
 for granted, and with some slight modifications he treated it as 
 supplying a complete table of logical functions. The list of 
 logical functions again, when judgment was regarded not as a 
 merely formal process of analysing, ideas but as a process of 
 determining objects, translated itself into a list of categories or 
 conceptions of objects in general, which was just the list of 
 pure conceptions Kant required. He could, therefore, now 
 proceed with his great task, the task of proving that we have 
 aright to apply these general conceptions to the objects of sense ; 
 or, to speak more accurately, to the appearances of sense, which, 
 as thus subsumed under the categories, are determined as the 
 objects of experience. 
 
 2. There are some very important results which may be Substitution 
 drawn from this account of the process whereby Kant arrived prf^clpw 
 atiiis list of categories ; but before attempting to state them, it knowledge. 
 
 VOL. I. O 
 
210 INTRODUCTION. chap; 
 
 seems desirable to look at the other aspect of the transition 
 from the Dissertation to the Critique to which reference has 
 been made. In the Dissertation, the pure conceptions are 
 regarded as ways in which God predetermines all the finite 
 and contingent substances with a view to their combination 
 into one world, all parts of which are in mutuo commercio: 
 while in the Critique, they are regarded as categories by which 
 the conscious self combines the data of sense in relation to 
 itself into one experience. In the Dissertation, the objective 
 connexion of substances as acting and reacting on each other 
 in one world is supposed to translate itself, — by reason of the 
 forms of sense through which the world is given to us m 
 experience, — into the subjective appearance of their existence 
 as in space and time ; and thus the unity of all things in their 
 dependence on God, appears to us as their existence in one 
 space [Omnipraesentia Phaenomenon) : and the reciprocal deter- 
 mination of their contingent existence through such dependence, 
 as their existence in one time (Aeternitas Phaenomenon). Finally, 
 if we ask for a ground for this belief in the coincidence or 
 parallelism of our subjective representation of the world as one 
 world in space and time, with the objective reality, Kant's 
 answer is that we must fall back on the idea that we, like all 
 other finite substances, are dependent on one absolute sub- 
 stance, whose unity makes the world one, or maintains all finite 
 substances in such relations, that they can influence each 
 other : and that, therefore, our perception of them which is 
 dependent on such external influence, presupposes and con- 
 forms to our conception of them which is not so dependent. 
 expifnK'" ^^i^ P^^* °f Kant's speculations is, as he indicates in 
 of knowledge. ^j.i^;ijig ^q Lambert, imperfectly developed in the Dissertation; 
 but it supplies the link which is needful to enable us to 
 connect his earlier and later speculations, and indirectly to 
 connect the earlier dogmatic philosophies of Spinoza, Leibniz, 
 and Wolff with the Criticism which Kant substituted for it. 
 As Kant at a later date said, the essential change from 
 
V- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 211 
 
 dogmatism to criticism was like the change from the Ptolemaic 
 to the Copernican system. Just as Copernicus taught 
 Astronomy to refer many of the apparent movements of the 
 heavens to the real movement of the earth on which the 
 spectator is placed, so criticism taught philosophy to regard 
 the objective world from the point of view of our connexion 
 with it, and to attribute the general characteristics of that 
 world, as it exists for consciousness, to the essential nature of 
 our consciousness itself. From this point of view the question 
 how all objects are combined into the unity of one world, turns 
 into the question how that world comes to be such a unity for ' 
 me. Now, while the former question might be answered by a 
 reference to the unity of God, on whom all finite existences alij?e 
 depend, and who therefore conditions their reciprocal influence ; 
 the latter was left, according to the principles of dogmatic 
 philosophy, to be answered by the theories of a pre-established 
 harmony, or of a continuous supernatural interference. Kant, 
 however, who had already explained the coincidence of things, 
 as we perceive them, with the a priori determinations of space 
 and time, by showing that time and space are the forms under 
 which alone we can perceive, was soon forced to find the 
 explanation also of the connexion of things as. we think them 
 under the pure principles of the understanding, in a subjective 
 necessity of the understanding itself. Whenever this idea 
 had been suggested, it became clear that the unity- of the 
 world, as determined by its dependence on one absolute 
 principle, will not of itself explain the unity of our knowledge 
 of it. On the other hand, the unity of our intelligence with I 
 itself will explain the necessary unity and interconnexion in its 
 parts of any world that exists for our intelligence. In other 
 words, if we take for granted the fact of knowledge and the 
 existence of a world which is an object of knowledge, we see 
 that that world cannot but conform to the conditions under 
 which alone it can become known : in other words, nothino- 
 cam be determined by the intelligence as an object that does 
 
212 INTRODUCTION. chap. 
 
 not conform to conditions under which alone it can become 
 such an object. Hence we do not need to say that the object 
 in itself is in pre-established harmony with the mind, or the 
 mind with the object in itself, — an assertion which contradicts 
 itself by breaking through the very limits of knowledge which 
 it at the same time sets up, — but we can say that the known 
 (pbjects as such, must conform to the conditions, of know- 
 ledge, and that the knowable object, as such, must he 
 in pre-established harmony, if we like so to express it, with 
 these conditions. 
 pSnatfonont When we have got so far, it begins to be clear that for the 
 unity which was before found in God, as an objective principle, 
 we must substitute the unity of the self If the world is one 
 world for us, it is not because it is one independently of the 
 action of our intelligence and solely in virtue of the unity of 
 its divine cause ; but because it cannot exist for us or be 
 known by us, except so far as it is brought in relation to one 
 self Formerly, Kant had said that the finite substances are 
 ill mutuo commercio and constitute one world, because, and 
 only because, they are all equally dependent on the one 
 absolute substance. Now he must say that they are neces- 
 sarily represented as acting and reacting on each other and so 
 constituting one world, only because they are all objects of one 
 conscious self and must, therefore conform to the conditions of 
 the unity of self-consciousness. There is here a complete 
 shifting of the centre to which all is referred, but in other 
 respects all that was previously said of being has now to be 
 said of knowing. Hence all the pure conceptions, and especi- 
 ally the conceptions of substance, causality, and reciprocity, — 
 by which in the Dissertation the world was determined as a 
 whole of parts essentially interrelated which in all its changes 
 maintains its identity and its unity, — cease to be regarded as 
 representing objective relations of things in themselves and are 
 taken as relations of phenomena or things as existing for us. 
 Instead of being treated as principles of relation by which 
 
 •d.^M 
 
V. DEVELOPMENT OP THE CRITICAL PHILOSOPHY. 216 
 
 objects are united into one real world in dependence on one 
 absolute Being, they are now represented as necessary relations 
 by which all objects must be determined, if they are to be 
 united as a known world in one consciousness, i.e., in the con- 
 sciousness of one self. The dogmatic presuppositions of the 
 unity of the world and of the interconnexion of its parts thus 
 become vitalised, the former being identified with the unity of 
 intelligence and the latter with the special functions of thought 
 in which that unity expresses itself. Thus the world of things 
 in themselves outside of consciousness is abandoned as the 
 unknown and the unknowable : and the whole interest of 
 knowledge is concentrated on the process within consciousness, 
 by which the data presented to the mind in sense are combined 
 and determined as the elements of one experience. The pure 
 intelligence by its a priori conceptions is still conceived as 
 giving objective determination to the data of sense, which are 
 regarded as in themselves merely subjective states bound 
 together, at best only in a subjective unity of apprehension. 
 But this contrast of subjective and objective has changed its 
 meaning ; for it now means that it is only as the self brings 
 the data of its perception under the forms of its conception, 
 that it can have consciouness of an objective world as such, or 
 can determine its subjective perceptions as perceptions of a 
 world of objects. 
 
 It is most important for the understanding of the Critique ^^^^ di'JSatio 
 that we should keep in mind this Copernican change of the ™''' 
 centre from which the intelligible world is regarded, and in 
 relation to which its varying aspects are explained. But in 
 order that we may do justice to previous philosophies, it is also 
 important that we should understand at once the value and 
 the limit of the alteration which it involves. Previous philo- 
 sophy had taken its stand on the Absolute. It had assumed 
 that we are able to place ourselves at the objective centre from 
 which all things are seen as they are, in the relations which , 
 th<ey really hold to each other. Thus Spinoza had maintained 
 
214 INTEODUCTTON. <^hap. 
 
 that science, the scientia intuitiva which alone deserves the 
 name of knowledge, necessarily requires us to start with that 
 idea on which all others depend and to deduce all other ideas 
 from it ; and in his Ethics he had attempted to realise this 
 conception by. placing the definition of God at the beginning of 
 his treatise, and making it the basis of everything else. But 
 he had not sufficiently explained. how the consciousness of man, 
 which according to him is a mode of one of the attributes, 
 should be able to reach this beginning, i.e., to rise to the appre- 
 hension, not only of the attribute of which it is a mode, but 
 even of the substance, which is beyond the special determina- 
 tions of the attributes and which embraces them all. And 
 though Leibniz had made this less unintelligible by treating 
 the individual as a microcosm, or by bestowing on the self- 
 conscious monad the exceptional gift -of a consciousness of God 
 and of general truth, (thus in spite of his own principles putting 
 
 ( a qualitative difference between the self-conscious monad and the 
 other monads,) he had by such expedients rather indicated than 
 solved the difficulty. The idea of a " pre-established harmony" 
 shows that its author was conscious that the division between 
 
 I knowing and being was still for him an absolute division, even 
 at the very moment when he was claiming the right to overpass 
 it. For if the self-conscious being had been conceived by him 
 as able to rise to the unity of being and knowing in God, he 
 would no longer have needed to bring in the thought of a 
 liarmony to bridge over the division. Kant, therefore, seemed 
 to be saying no more than was strictly justifiable, when he con- 
 demned all these theories as involving the introduction of a 
 Dei(,s ex. machina, and maintained that, besides involving the 
 fallacy of reasoning in a circle, such ways of explaining the 
 origin of knowledge, " stimulate and encourage aU the whims 
 of pious fancy and all the dreams of baseless speculation." 
 Element of At the Same time, while we recognise all this, and while we 
 
 truth in the 
 
 ^?^^*'<= , admit that such theories are inconsistent with themselves in 
 taking their stand at a universal point of view, which yet their 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 215 
 
 account of man's intelligence does not justify, we must remem- 
 ber that no criticism of our knowledge is possible Which does 
 not involve that in some way we can reach beyond the defect 
 which we criticise. If we recognise the relativity of know- 
 ledge, it must be in reference to some standard which is not 
 itself regarded as relative ; if we say that we know only 
 phenomena, it must be in reference to some consciousness of a 
 noumenon which we stiU possess. The assumption, if we can 
 call it an assumption, that we in some sense know what reality 
 is, is implied even in the extremest assertion of the limits of 
 knowledge. The consciousness that at some point thinking 
 and being meet together and coalesce in one, is necessarily in-' 
 volved even in the most dualistic view of their relations. We 
 must take our stand on something beyond tlie limit to discover 
 that there is a limit. In this sense, then, the theories which 
 Kant was criticising contain a principle, which cannot be re- 
 f garded as merely dogmatic except from the point of view of 
 \ absolute scepticism. Tor it is impossible that the individual 
 should be confined to the mere phenomena of his own con- 
 sciousness, and that at the same time he should be able to 
 recognise that he is so confined. Such a recognition involves 
 already that there is for him some point of view beyond the 
 limit which he asserts. The error of those theories, against 
 which Kant had a right to protest, was not that they took 
 their stand at a point of view which is beyond the opposition 
 , of subjective and objective, or that they started from the idea 
 of God, but rather that they took the idea of God as purely 
 [objective and distinct from the consciousness of self; or, in 
 other words, that they failed to recognise that in self-conscious- 
 ness there is contained not merely the consciousness of the 
 subject as opposed to the object, but also the consciousness of 
 a unity which, while it involves that opposition, at the same 
 time transcends it. In other words, they did not recognise 
 that in the consciousness of self is involved also the conscious- 
 l ness of the universal unity or centre which all knowledge 
 
216 INTRODUCTION. CHAP. 
 
 implies, and that in this sense, the consciousness of self and 
 the consciousness of God are essentially bound up with each 
 other. While, therefore, we are prepared to admit that they 
 were in error, in so far as they did not see that all objects are 
 phenomenal, i.e., that they are not things in themselves inde- 
 pendent of consciousness, we are not prepared to recognise 
 that the ultimate reality consists in such things : for this would 
 imply that the absolute reality is that which has no relation to 
 consciousness. Nor are we prepared to admit that it is a 
 defect of our knowledge, that we cannot apprehend such things. 
 On the contrary, it is in the apprehension of this relation of 
 things to the self that we gain the power at once of criticising 
 our knowledge, in so far as it abstracts from this relation or 
 leaves it out of account ; for in detecting the error of this 
 omission we. at the same time reach a consciousness of the 
 reality, in opposition to which the former object of our know- 
 ledge can be determined as phenomenal. 
 Kant admits Now, the kev to many difficulties of Kant's philosophy is 
 
 in a sense an 
 
 of^Siught""'' found, when we ask how far he did, and how far he did not, 
 and being. recognise this. I have already indicated in part what the 
 answer must be. Kant does see that there is a point of view 
 beyond the opposition of thought and being, a point at which 
 they fall together; and he sees also that it is because we are 
 able to take our stand at this point of view that we are able 
 to criticise our knowledge. Furthermore, he holds that this 
 point of view is revealed to us in or through the pure con- 
 sciousness of that self to which all phenomenal objects as such 
 are referred, but to which they are at the same time opposed. 
 For, with this consciousness of self, there comes necessarily the 
 idea of an intuitive understanding, which is not thus opposed 
 to its object, but creates as it apprehends, and apprehends as 
 it creates it. As, however, this idea is reached (or seems 
 to Kant to be reached) by negation and abstraction, — by an 
 abstraction from conception in its distinction from perception, . 
 as well as from perception in its distinction from conception,— 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 217 
 
 it shrinks, whenever Kant turns his attention to it, into a bare 
 identity, of which he can say that we " have no conception 
 whatsoever," though we have " a consciousness " of it, and 
 though this consciousness, through the moral law, is ultimately 
 
 ■determined to be a consciousness of the absolute reality. How 
 far this view is tenable will be considered hereafter ; what we 
 have here to observe is that while, as was indicated in the first 
 chapter of this Introduction, the opposition of critical to 
 dogmatic philosophy depends necessarily on the regress which 
 the former makes upon the presuppositions of the latter ; yet 
 this regress must find a point at which it stops, and from which, 
 as an ultimate point beyond the opposition of being and 
 knowledge, of noumena and phenomena, it is able to discover 
 the errors of the ordinary dualism. It follows, therefore, that 
 while Kant's immediate task is to criticise previous philosophies 
 
 ( for confusing thought with reality, yet indirectly and in the 
 end he has also to criticise them for not discovering the true 
 point in which these opposites are united. Unfortunately this 
 point is for him neither an object of conception nor of percep- 
 tion, though it is presupposed in both, and though he assumes 
 that we can so far bring it into consciousness as to take our 
 stand on it for the criticism- of both. 
 
 We are now prepared to take up our deferred examination Kant's use of 
 
 ■*- -^ formal Logic 
 
 of the method which Kant used in discovering the list of pure thre^*'*'"® 
 conceptions. Kant extracted these conceptions, as he tells us, 
 from the idea of the- judgment, as that was analysed by the 
 Logicians. In order, however, to understand this statement 
 we must have before us two points. In the first place, we 
 must be aware of Kant's confidence in formal Logic, which he 
 had received as a tradition of the schools, and supposed to 
 have come to him unaltered from- Aristotle. And, in the 
 second place, we must- keep in view the fact that in selecting 
 the analysis of judgment as his starting point, Kant did not 
 forget that judgment is only one of the operations which that 
 Logic; analyses, and that it always considers judgment in its 
 
218 INTRODtrCTION. <^hap. 
 
 relation to conception on the one side, and to syllogism on, the 
 other. In other words, he recognised that Logic is a system, 
 and that, if one of its doctrines be used as a principle of 
 discovery, this carries with it a similar use of all the 
 others. 
 Hisconfidenoe j^g to the first of thcsc points, w'c do not need to look far in 
 
 in formal ■*- 
 
 ^°^°- the Critique, to come upon the evidences of Kant's absolute 
 
 trust in formal Logic, as a sufficient analysis of the process of 
 thought, when that process is treated in itself without reference 
 to its objects. " Logic," he tells us in the preface to the 
 second edition, " already in the earliest times had got into 
 the secure path of science. This may be seen from the fact 
 that since Aristotle, it has never needed to retrace a single 
 step, except perhaps in getting rid of a few unnecessary 
 subtilties, or in more distinctly defining its subject matter, — 
 improvements which have rather to do with the appropriate- 
 ness of the form in which the science is taught than with its 
 scientific certitude. It is, moreover, noticeable that it has as 
 little been able to take a single step forward, and seems to all 
 appearance to have been once for all completed and exhausted. 
 Some moderns, indeed, have sought to enlarge it by intro- 
 ducing chapters from Psychology, about the nature of the 
 various faculties exercised in knowledge ; or from Metaphysic, 
 about the origin of knowledge and the differences of the certi- 
 tude of our knowledge of different objects; or from Anthropology, 
 about the causes and cures of prejudice, etc. But in making 
 such additions to it, they have only shown their ignorance of 
 the peculiar character of the science. It is not an extension, 
 it is only a distortion, of the sciences when we let them pass 
 their limits and run into each other ; and the limits of Logic 
 are very easily determined. It is the science which completely 
 exhibits and strictly demonstrates the formal rules of all 
 thinking (whether it be a priori or empirical, whatever be its 
 origin or object, and whatever accidental or natural hindrances 
 it may meet with in our mental constitution)." 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 219 
 
 Having this opinion of formal Logic, as a science complete Judgment md 
 
 or n ' r Syllogism as 
 
 and perfectly secure in its results, a science which has stood categoriesana 
 the proof of two thousand years without change, Kant can use respectively, 
 it as ain absolutely certain guide in founding his new Logic, 
 which is to consider not merely the rules for thinking, but the 
 rules for the knowing of objects, so far as that is possible a 
 priori. Now thought, as dealt with by formal Logic, starts 
 with the presupposition of a certain given content or idea, 
 which it analyses in judgment into a subject and predicate, and 
 which in syllogism it re-analyses till it finds the ultimate 
 condition of the judgment, i.e., of the assertion which attributes 
 that predicate to that subject. This play of pure thought in 
 itself, or rather (since analytic thought always supposes some- 
 thing to analyse) this play which it carries on without reference 
 to. the particular character of the matter in hand, is to be 
 explained as the mere assertion by the conscious subject of its 
 identity with itself in apprehending such matter. It is the 
 endless 'I am I ' of self-consciousness, which it repeats in 
 relation to every content brought within its scope. As the 
 mind is in identity with itself, so every content it receives into 
 its consciousness must be, in like manner, fixed in its self- 
 identity. In judgment, indeed, a division arises between subject 
 and predicate, but this division is stated only to be denied ; and 
 if it has not disappeared with this denial, syllogism by a further 
 analysis discovers a middle term by which the difference may 
 finally be brought back to identity. 
 
 Now, in all this analytic process thought deals with a matter How the 
 
 ' J r o synthetic, Is 
 
 which it has already appropriated, and in dealing with which l^^fi^ ^^" 
 it never needs to' go beyond itself. It is a purely subjective ■'" ^^^'^*- 
 process that has no reference to any object or to any thing 
 but ideas which, as ideas, are already combined with the ' I 
 think.' But how is any matter appropriated by our thought 
 at first ? The pure unity of the self produces nothing. Its 
 work is only to distinguish elements which it again reduces to 
 identity, just as in the 'I am I ' of self-consciousness it 
 
220 INTRODUCTION. "^^ap. 
 
 divides the subjective and the objective self only to identify 
 them again. What, however, we have to explain in the new 
 Logic is the mode in which, for the first time, a matter 
 which does not already belong to the self is brought into 
 relation to it ; or, what is the same thing from the opposite 
 side, the mode in which the pure seK goes beyond itself to 
 apprehend that which is not itself. And the only way in 
 which we can do so, is by supposing that, in relation to a 
 given manifold of perception, the pure unity expands into a 
 principle of connexion between the elements of this manifold. 
 It thus, as it were, supplies from itself the predicate for a 
 subject given in perception, and performs an act of judgment 
 in attaching that predicate to it. But this act of judgment is 
 synthetic, not analytic : the intelligence does not get the 
 predicate by analysis from the matter presented to it, but itself 
 evolves it out of its own unity. On this view, therefore, we 
 might say, that the intelligence determines the data of sense as 
 its object by making itself their predicate. 
 Derivation of Jq this Way we Can see how it is that Kant looks to the act 
 
 the categories. •^ 
 
 of judgment, when he is seeking for the pure a priori concep- 
 tions. These conceptions spring out of the understanding in 
 the activity of judgment, whereby it applies itself to the matter, 
 of perception, or, in other words, whereby it turns the matter of 
 perception into objects of knowledge to itself. Eesting in itself 
 thought is a pure identity, but in experience it becomes a 
 principle of synthesis in reference to a given manifold which 
 it appropriates. Its identity is differentiated in reference to a 
 ' manifold ' which it thus determines as its object ; and, on the 
 other hand, the manifold is integrated in reference to the self 
 of whose consciousness or experience it becomes a part. And 
 the different " moments " or modifications, which this differentia- 
 tion of thought presents, are to be discovered by considering 
 the different aspects of the formal judgment, i.e., the different 
 aspects in which the unity of the mind shows itself in dealing 
 with a content that has already been brought into unity with 
 
V. DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 221 
 
 it. If we watch what the mind does with a content which it i 
 has already appropriated, we shall be guided safely to thel 
 discovery of that which it does in appropriating a new content/ 
 Thus it is that Kant justifies his procedure in taking the 
 logical analysis of the formal judgment as a " guiding thread " 
 to the discovery of the pure conceptions of the understanding, 
 — a procedure the results of which we shall consider in detail 
 in a subsequent chapter. 
 
 But iudsment is. "made adequate" in syllogism; i.e., the How the 
 
 •> '^ ^ ./ & ; ; synthetic IS 
 
 identity of subject and predicate expressed in a judgment is ^nafyuo*," *^'' 
 reduced to its simplest form by bringing to light a middle ^^'^op™- 
 term which is the principle of unity between the extremes. 
 And this higher " function of unity " must also reappear in the 
 case when the unity has to be established, not between the 
 mind and itself, as already having a special content, but 
 between the mind and the new data which are presented to 
 it through sense. Hence the intelligence cannot be satisfied 
 with the act of judgment in which these data are determined 
 as objects in relation to it, but requires that this act of deter- 
 mination itself should find a mediation. It is obvious, how- 
 ever, that the middle term which is here required cannot be 
 found in pure thought or in the data of sense, but only in 
 something that contains both, i.e., in a prior act of determina- 
 tion of an object by a conception ; and this again points back 
 to a prior act and so on ad infinitum.^ What the mind is 
 
 1 As showing that the above is a correct analysis of Kant's way of thinking, 
 note what is said (A. 76 ; B. 101) of the apodictioal judgment as contrasted 
 with the assertorial and problematical judgments. "The apodictical judg- 
 ment involves that the assertorial is determined by the laws of the understand- 
 ing. It therefore makes an assertion a priori and expresses logical necessity. 
 Now as a thing asserted is only gradually incorporated with the understanding, 
 so that we begin by making a problematical judgment regarding it, then 
 accept it assertorially as true, and finally declare it to be inseparably bound up 
 with the understanding, i.e., as necessary and apodictically certain, we may 
 regard these three functions of judgment as so many moments in the process 
 of thought. " In this passage we have very clearly expressed that idea of a 
 synthesis with the understanding, which is the modal aspect of judgment 
 and which contrasts with the other aspects of it as synthesis for the 
 
222 INTRODUCTION. CHAP. 
 
 seeking, namely, a middle term between itself or its own con- 
 ceptive activity and the data of sense, can never be found ; for, 
 however far we go back, we find still the same dualism. The 
 movement of mind expressed, in syllogism can never be com- 
 pleted, because its completion would involve a unity of per- 
 ception and conception which had no middle term : i.e., an 
 " intellectual intuition." Such an intuition, therefore, is 
 suggested by reason as an unattainable ideal, although it can 
 never be either conceived or perceived. And if there are 
 different forms of syllogism, it is to be supposed that this 
 ideal will present itself in different forms corresponding 
 thereto. 
 Imperfection The full development of these ideas must be reserved for 
 
 of the synthe- ^ 
 
 ini/faita^rof t^e sequcl ; but enough has been said to show how the logical 
 syuo^^m.° '° system of judgment and syllogism, which Kant takes for his 
 guide, may expand into a system of categories and a system of 
 ideas, corresponding respectively to the acts of judging and 
 reasoning in which the mind expresses its unity, when that 
 unity is taken as a principle not of analysis but of synthesis. 
 The point, however, to which I wish here to call attention is 
 that Kant in all this development of his thought remains faith^ 
 ful to the primitive idea expressed in the Dissertation. For there, 
 as we saw, the double imperfection, on the one hand, of a per- 
 ception which is given to the mind and which, therefore, is not 
 one with the conception produced by the mind, — a perception 
 which is not conception, — and on the other hand, of a conception 
 which is only a general form under which the objects may be 
 brought, but which does not particularise itself or determine 
 itself as an individual object, — a conception which is not 
 perception, — suggested to Kant the idea of a unity in which 
 the dualism is removed. But this unity which is neither per- 
 ception nor conception, or which is both at once, could not, for 
 that reason, be either conceived or perceived. Kant, therefore, 
 
 uncjerstanding. These aspects, however, are essentially complementary of 
 each other. 
 
V. DEVELOPMENT 0¥ THE CRITICAL PHILOSOPHY. 223 
 
 though he speaks of it as a consciousness which is one with 
 the absolute reahty, is at the same time obliged to deny that 
 we can ever imagine what it is. The method of abstraction 
 by which he reaches it forces him to treat it as a movement of 
 thought, in which, nevertheless, thought never reaches beyond 
 itself: a process of self-determination in which it is forever 
 confined to itself. But this is just the analytic process of 
 thought with which formal Logic deals. Hence we arrive at the 
 conclusion that the ideal of truth or knowledge, which is 
 expressed one way as the intuitive understanding, is in another 
 way the ideal of a purely analytic thought which never moves 
 out of its identity with itself. It is, therefore, only a natural 
 development of the thought of the Dissertation, when formal 
 Logic, in its highest expression in the syllogism, is supposed to 
 give rise to an ideal of knowledge which is incapable of being 
 realised, because it represents only the pure identity of thought 
 with itself, irrespective of any matter. 
 
 This, however, only brings into more decisive prominence the source of the 
 great difficulty involved in Kant's supposition that the ultimate ?g^S|htTud 
 principle, on which we must take our stand to criticise know- °°^° ^'' 
 ledge, is reached by abstraction from both the elements 
 involved in knowledge. If it be so reached, the ideal act 
 of knowing will have to be conceived as a movement by pure 
 identity, a purely analytic movement, in which all differences 
 of subject and predicate are eliminated; i.e., it will have to 
 be conceived as the formal Logicians conceived the process 
 of thought. The question, however, wiU immediately recur, 
 whether this supposed analytic movement be not simply mie 
 aspect of the process of thought, a process which always in- 
 volves difference as well as identity ; and whether, therefore, it 
 be not a fundamental error to admit that formal Logic exhibits 
 any process of thought which could go on by itself. For, 
 if this be so, then it will become manifest that the criticism, 
 which Kant seems to base on the idea of an abstract identity, 
 fram which the difference of thought and perception is ex- 
 
224 INTKODUCTION. CHAP. 
 
 eluded, is l-eally based on the idea of a unity in which that 
 difference is reduced to a relative one. On this view, the 
 pure analytic or formal Logic wUl disappear ; and' the synthetic 
 Logic of experience and science will have to be contrasted, not 
 with it, but with a synthesis of a more perfect kind, "in which 
 the perception is not merely subsumed under the conception, 
 but in which both are seen to be essential counterparts of each 
 other. The intuitive or perceptive understanding will thus 
 present itself as an ideal of thought reached, not by abstraction 
 but rather by the opposite process, viz., by bringing before 
 consciousness the unity which underlies the difference of 
 conception and perception, and which manifests itself both 
 in producing that difference and in overcoming it. In such 
 a view of the process of thought, the syllogism will have a 
 meaning, which it has not either in formal Logic or in the 
 synthetic Logic of experience. For, in the former, there is no 
 explanation why the identity should ever have been left behind, 
 and, therefore, no need to return to it; and in the latter, the 
 difference introduced remains permanent and insoluble, and 
 there is no explanation of the fact that we ever desire to solve 
 it. According to the view now suggested, however, we can 
 find a real meaning for syllogism, as the process through 
 which the necessary differentiation of knowledge is brought 
 back to the unity of thought. These, however, are points 
 which we cannot yet fully discuss. For the present, let us 
 sum up the results arrived at in this chapter in relation to 
 the development of Kant's view of criticism. 
 rfttTdJveio^ Criticism necessarily involves a reconsideration of our ideas, 
 
 metit of the • jy j.* i j_» n t • j • n 
 
 critical idea. 1-&; 01 our Conceptions and perceptions oi objects, in reference 
 to the faculty of knowledge. With Kant, it begins in a recon- 
 sideration of the ideas of time and space, as ideas presupposed 
 in all experience. These ideas are determined as forms of 
 perception, and thus an escape is found from ' the difficulties 
 which arise when they are conceived as objective conditions of 
 things. Hence, in the Dissertation, things in themselves are re- 
 
v.- DEVELOPMENT OF THE CRITICAL PHILOSOPHY. 225 
 
 garded as defined by the pure conceptions only. But the pure 
 conceptions are general ; they are possible predicates for which 
 a subject must be given; in other words, they are subjective 
 forms which do not prove their own objectivity. And the only 
 m,£^tfcer to be subsumed under these conceptions is the matter 
 of sense given imder the forms of space and time. In spite, 
 therefore, of the opposition which has already been shown to 
 exist between the pure conceptions and the perceptions which 
 are given under the forms of time and space, it is only in the 
 latter that the former can be realised, or find individual objects 
 corresponding to them. Thus, in order to show that the pure 
 conceptions have any objective value at all, it becomes neces- 
 sary to prove that they have application to the form and matter 
 of sense. In this proof, however, the pure conceptions as well 
 as the forms of perception become limited to phenomena. 
 They become characterised as the forms by which the Ego 
 determines the appearances of sense as objects for itself. Hence, 
 it is necessary to find some other ground for the contrast of 
 noumena and phenomena than that which was first given. This 
 ground is found in the unity of pure self-consciousness, which, 
 though presupposed in the a priori conceptions of the under- 
 standing, is already in those conceptions differentiated in relation 
 to the manifold of sense. The pure analytic unity, on which 
 Kant makes his ultimate regress, thus yields him a point of view 
 from which he can criticise that consciousness of objects which 
 constitutes our experience, and oppose to it the idea of a higher 
 knowledge and a higher reality. And this at the same time 
 explains how the empirical consciousness itself is directed and 
 stimulated by an ideal which yet it can never realise, an ideal 
 suggested by the formal unity of thought in its contrast with 
 the synthetic unity of experience. "We have, however, only the 
 idea of such knowledge and such reality — as Kant, conscious of 
 the defect of the logic of analysis, himself suggests ; and this 
 result shows that his criticism rests, not on the basis on 
 whieh he himself places it, but on the conception of an intui- 
 
 VOL. I. P 
 
226 INTRODUCTION. chap; -v. 
 
 tive understanding, which shall solve the antagonism that he 
 declares to be insoluble. The fuller exposition of these ideas 
 must be deferred to the sequel : enough has been said to 
 suggest such a view of the Critique of Pure Reason as the 
 author might have had in his mind, before he set about the 
 final work of elaborating its different parts. 
 
227 
 
 BOOK I. 
 
 THE CRITIQUE OF PURE REASON. 
 
 CHAPTER I. 
 
 THE PROBLEM OF THE CRITIQUE OF PURE REASON. 
 
 nriHE subject of this chapter has been partly anticipated, but 
 there are some aspects of the problem of Criticism which 
 may be elucidated by a re-statement of it from a somewhat 
 different point of view. Moreover, it is necessary to bring 
 what has been said into closer relation with the language of 
 the Critique itself. 
 
 The object which Kant had. in view in the Critical Philo- ^^75\he 
 sophy in general, and especially in the Critique of Pure Reason, ™ °"'' 
 has been very variously stated by different commentators, 
 according as they have attached more or less importance to 
 particular aspects of it. 
 
 To adduce > only the most prominent of these differences : it 
 has been maintained that Kant's main object was to prove \/ 
 against Hume that there is an a priori knowledge of objects ; Y 
 and it has been maintained that it was to prove against Wolff / \ 
 and Leibniz that knowledge is limited to phenomena. Again, 
 it has been maintained that he was aiming at, and had nearly 
 reached, an Idealism, which involves the negation of all " things 
 in themselves," or the assertion that the self is the only ' thing 
 in itself; and it has been maintained that all he sought was to 
 
228 THE CRITIQUE .OF PURE REASON. book I. 
 
 fix once for all the opposition between knowledge and reality, 
 to confine sense to an empirical determination of phenomena, 
 and to relegate all noumenal reality to the region of the 
 unknown and the unknowable. And there can be no doubt 
 that passages may be easily quoted in support of each of these 
 views, seeing that the task of Kant is a very complex one, and 
 what appears as an end from one point of view, or in relation 
 to one part of his philosophy taken by itself, becomes in turn 
 a means when we contemplate that part in its relations to 
 other parts. The different sections of the Critique liave thus a 
 proximate purpose, different from that which is the aim of the 
 whole treatise ; and that treatise gets a different meaning 
 according as we contemplate it as a whole in itself, or as a 
 part of a wider plan which from the first was present to Kant's 
 mind, at least in its main outlines, though for convenience he 
 executed it in a series of successive treatises. 
 •Th&cHHgue The evidences of this last statement are easy to discover. 
 
 oi Pun Reason •' 
 
 ur^?s?heine. From the Critique itself we learn that in writing it, he was 
 contemplating and preparing the way for his other, and especi- 
 ally his ethical, works ; and a letter to Herz dated 7 th June, 
 1771, shows that in his original plan all the investigations; 
 which were afterwards spread over the three Critiques, were in- 
 tended to be gathered in one work, under the title of " The Limits 
 of Sense and Reason." That work he proposed to divide into "two 
 parts, one theoretical, and the other practical." The first part 
 was to contain two sections, one corresponding to the Critique 
 of Pure Reason and treating of Phenomenology in general ; the 
 other treating of Metaphysic, though only as regards its nature 
 and method. The second part was also to contain two sections, 
 one treating of the principles of feeling, of taste and sensuous 
 desire, and the other of the primary rational basis of morality, 
 Now the independent treatment of the different parts of this 
 general plan which Kant subsequently adopted, had the effect 
 of giving an appearance of finality to the results which were 
 arrived at in each of them, though they were really parts of 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 229 
 
 one whole,, and were originally conceived as such. This fact, 
 we may notice in passing, has a very important bearing on the 
 controversy as to the agreement or difference of the first and 
 second editions of the Critique of Pure Reason. For, as will 
 be shown more fully in the sequel, the alteration of Kant's 
 views which is supposed to be found in the second edition, 
 is, partly at least, the result of an effort on his part to remove 
 the misconceptions of certain of his readers who had regarded his 
 arguments with reference to their immediate results, and without 
 reference to the further results which he sought to reach through 
 them. These misconceptions Kant sought to meet by bringing 
 in anticipative statements of his ultimate purpose, — statements 
 which sometimes, it must be confessed, have the effect of intro- 
 ducing a new source of confusion into the immediate argument. 
 
 I may best indicate the view taken in this commentary by ^^^ ^jg^^t' 
 reference to the two poles of philosophy between which Kant eSisw.*''^ 
 tries to make his way; viz., Dogmatism, of which the main 
 representative is Wolff, and Scepticism, of which the main 
 representative is Hume. If we take the Aesthetic by itself, we The Aesthetic. 
 may view it as directed against Scepticism, in so far as it 
 shows that the determinations of space and time are not dis- 
 covered by analysis but developed by synthesis, and that they 
 are not pure " relations of ideas " but have objective reality. 
 At the same time, as it proves that this a priori synthesis has 
 reference to the forms of sense, it shows that the objective 
 reality in question can be regarded only as phenomenal, and 
 that, therefore, no inference can be drawn from the objective 
 value of mathematical science to support the idea of the 
 possibility of knowing things in themselves. The ultimate 
 result of the Aesthetic, therefore, seems to fall on the side of 
 Scepticism, in so far as it confines knowledge to phenomena, or 
 at least gives no encouragement to the idea of its extension 
 beyond phenomena ; and on the side of Dogmatism, in so far 
 as it proves that mathematical truth is due to a priori syn- 
 thesis, and yet objective. When we go on to the Analytic, ^he Anaii/tic. 
 
230 THE CEITIQUE OF PURE REASON. BOOK I, 
 
 we find that another step is made towards the exclusion of any 
 knowledge of things in themselves and the confirmation of the 
 reality of our a priori knowledge of phenomena. For it is 
 there shown that the a priori forms of conception cannot, any 
 more than the a priori forms of perception, supply an instru- ; 
 ment by which we may reach any knowledge of reality not 
 given in sense ; while they do supply, and are needed to supply, 
 j)rinciples for empirical knowledge, i.e., for knowledge of 
 phenoinenal reality. The result, therefore, is to fortify 
 still further the objective validity of a priori principles, 
 and to show indeed that without them no empirical 
 knowledge is possible : but at the same time, it is to 
 strengthen the position of Scepticism in relation to things 
 in themselves, by showing that such principles are objec- 
 tively valid only as principles for the determination of 
 rhc Dialectic, phenomena. The Dialectic, finally, falling back on that pure 
 unity of thought in itself, which is presupposed in all con- 
 ceptual synthesis, shows that, while that unity suggests ideas 
 of noumena or things in themselves, which are not, as such, 
 objects of experience, and so gives rise to certain problems 
 which experience cannot solve, it cannot enable us to make 
 mere ideas a basis of knowledge, but only to use them as an 
 ideal, by which experience may be stimulated and directed, 
 without the hope of ever reaching or verifying it. The conclu- 
 sion, therefore, to which the Critique as a whole brings us is 
 yone which coincides, in the main, with the results of Scepticism. 
 For while all the a priori possessions claimed for the mind are, 
 in a sense, vindicated ; while the a priori forms of perception 
 and the a priori conceptions of the understanding are both 
 proved to have objective validity; and even the ideas of reason 
 are shown to have a necessary function in relation to the 
 knowledge of objects: yet they are all conceived to expend 'I 
 their usefulness on experience, i.e., on the knowledge of objects 
 which are merely phenomenal : so that nothing seems to be left 
 to bring us into any relation to things in themselves. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 231 
 
 We must, however, remember that in relation to noumena Beiationojthe 
 
 ' Critique of 
 
 the verdict reached by the Critique in the last stage is really faTthe^cwajMe 
 an open one; for, while it shows that the ideas of reason haveiiccMon; 
 a use in relation to experience, it shows also that they stand in 
 an asymptotic relation to it, as giving rise to an ideal of know- 
 ledge which cannot be realised in experience. The Critique 
 thus leaves room for the possibility that the ideas of reason 
 may refer to realities which, because of the nature of our per- 
 ceptions, as well as of our a priori conceptions (which have 
 essential relation to these perceptions), cannot be determined 
 as objects of knowledge. And it is into this room that, 
 according to the Critique of Practical Meason, the moral con- 
 sciousness introduces itself — -giving assertorial value to the 
 ideas which the Critique of Pure Reason left problematical, and 
 changing the possibility of things in themselves, which corre- 
 spond to the ideas of reason, into a certainty, though a certainty 
 of faith and not of knowledge. The Critique of Practical Beason 
 thus, so to speak, puts in the keystone which was wanting to 
 the completeness of the Critique of Pure Meason, and gives the 
 final sanction to that dualism of phenomena and noumena 
 which was throughout presupposed. Por, though in the 
 Aesthetic and Analytic, the phenomenal character of the objects 
 of knowledge is proved, in one sense of the term phenomenal, 
 i.e., in the sense that they are essentially objects of our con- 
 sciousness, still the ultimate reason for separating noumena or 
 things in themselves from these objects is given only in the 
 Dialectic, where it is shown that there are ideas, which arise in 
 connexion with experience, and which even in a sense are 
 its presuppositions, but which yet are not realised in any of 
 the objects of experience. And this distinction only gives 
 rise to a doubt, or a consciousness of ignorance, until an 
 independent basis is found for the assertion of the reality 
 of the objects of these ideas. The Critique of Practical 
 Beason thus first enables us to give the right interpreta- 
 tion to the beginning of the Critique of Pure Beason. The 
 
232 THE CRITIQUE OF PURE REASON. book C 
 
 criii*« 0/ further interpretation of both in the Critique of Judgmmt, in 
 j'udgmmt. ^jjigji j^^jj^ attempts to mediate between the theoretical and 
 the practical consciousness, and in which he all but brings 
 them together in a higher unity, — all but turns the realistic 
 Bualism of the first two Critiques into an idealistic Monism, — 
 I need not here do more than mention.^ Enough has 
 been said to show how easily Kant is misinterpreted, if we 
 stop at any stage of his argument short of its final result. 
 Kant advances in a sort of alternation of movement between 
 scepticism and dogmatism ; but his ultimate aim and purpose 
 is to put the fundamental truths of Metaphysic on an im- 
 moveable basis by removing them from all appearance of 
 collision with the principles of empirical knowledge ; or, looking 
 at it from the opposite side, it is to show that the principles of 
 empirical knowledge imply a consciousness which is not limited 
 ' Jto experience, but rather itself limits experience ; and that that 
 consciousness, while incapable of giving us the kind of know- 
 ledge which we have of the objects of experience, is yet in itself 
 the source of a rational certitude, as to those things which can 
 neither be seen with the eye nor heard with the ear, and which it 
 is beyond the power of our imagination to picture or of our under- 
 standing to determine, i.e., to comprehend as objects of knowledge. 
 First form of Tuming now to the special question of the Critique of Pure 
 
 thoproblemof _„ ._. ^ ^ .^- 
 
 the orttique of BeosoTi, wc must Undoubtedly say that, taking that treatise as a 
 
 Pure Reason. -? «'-,-</ -....'_. o . _ _ 
 
 whole, and without reference to any of Kant's other works,^ it 
 
 is a proof of the limitation of a priori knowledge to experience, 
 based upon an examination of the conditions of the knowledge 
 which is thus limited. At the same time this stateinent is 
 too simple, if we do not add that, in showing this limita- 
 tion of knowledge, Kant at the same time shows the necessity 
 
 ^ This part of the system, it may be admitted, was not clearly within Kant's 
 view at the beginning, as is shown by the change made in the note to the 
 Aesthetic (A. 21) in the second edition of the Critique (B. 35). 
 
 ^Of course, the answer would be different, if we took the Aesthetic, or the 
 Analytic separately as wholes, which would be Ifeirly, though not quite, as 
 legitimate as to take the Critique of Pure Season itself as a whole. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 
 
 233: 
 
 of the thought of objects beyond experience, and leaves open 
 the question as tb their reality and as to the possibility of 
 proving it. Still, in the main and in the first instance, we have 
 a right to say that the Critique explains, by the investigation of 
 the conditions of the a priori knowledge which we actually 
 possess, that such knowledge is not possible in regard to the 
 objects of those inquiries, which we " cannot but recognise as 
 at once far superior in importance, and far more elevated in 
 purpose than all that the understanding can know in the sphere 
 of experience." Hence in the Introduction to the Critique, 
 where he formulates the general question, Kant does not ask 
 how experience, or knowledge of phenomenal objects in general, 
 is possible, (a question which rises upon him subsequently in 
 the course of thought into which he is led in answering the 
 first question,) but simply how a priori knowledge of such 
 objects is possible. For it seemed obvious to him that it is 
 only by an a priori synthesis that we can go heyond the region 
 of experience, and, primarily at least, it was with reference to 
 this " beyond " that the question of the conditions of knowledge 
 within experience interested him.^ In this point of view, then, 
 
 ' Of the distinction between Kant's first problem, "how a ptiori knowledge 
 is possible " and his second problem, how " experience is possible," more will 
 be said in the sequel. I owe much to Dr. Vaihinger's luminous critical account 
 of the different ways in which Kant stated his problem in his CommPMlary ore the 
 Critique. (Cf. especially I. 189, 357, 387 seg".) I think he sometimes goes beyond 
 what is conceivable in his representation of Kant's unconsciousness of the 
 change of his own point of view, and does not attribute enough importance to 
 the propaedeutic intention of Kant. We have evidence that Kant often 
 deliberately began with imperfect statements which he afterwards modified 
 and corrected. (Cf . the statement of Jaehmann quoted previously, p. 64. ) A 
 direct reference by Kant to the imperfection of his first statement we have in 
 the Critique (§ 26 ; B. 161) : and he must have been conscious that the whole 
 argument of the Aesthetic was more deeply modified by its abstraction from the 
 a priori conceptions than could be gathered from what is said at the beginning 
 (A. 21 ; B. .35). Again in the Protegomena, Kant speaks as if the question, 
 whether pure mathematics were possible, was completely answered in the 
 Aesthetic, but he could scarcely be supposed to have there forgotten his own 
 proof in the Critique that a priori conceptions are necessary to mathematics. 
 Further evidences will appear in the sequel. On the other hand, it must be 
 admitted that direct statements by Kant as to his method are not often forth- 
 <toming. Kant seldom looks backward ; and it is, therefore, hard to discover 
 
234 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 it is distinctly to be admitted that Kant, in the first instance, 
 raises no question as to the possibility of knowledge or of 
 a 'priori knowledge within the limits of experience. So far^ his 
 attitude of thought is shown clearly in the assertion that such 
 a priori knowledge of empirical objects exists, and that its 
 possibility is shown by its reality.^ Thus all that the critical 
 philosopher has to do is to explain the fact of the existence of 
 a priori knowledge of empirical objects, in order that he may 
 by this explanation determine the possibility of similar kno,w- 
 ledge of objects, which are not empirically given. In this point 
 of view, Kant, while he recognises with Hume, that the univer- 
 sality and necessity of the principle of causality cannot be 
 explained by experience, refuses to follow him in reducing 
 that principle to an effect of particular experiences (magnified 
 by custom) upon the imagination ; and he even suggests 
 that Hume himself would not have adopted such a view, if 
 he had seen what it involved, when universalised and 
 applied to the principles of mathematics. There is a pure 
 mathematical science, and there is a pure physical science, 
 which yet is objectively valid : so much Kant asserts as a fact, 
 and what he attempts to show is only that this fact does not 
 carry with itself the inference that a priori knowledge is 
 possible of any other than empirical objects. What the sceptical 
 suggestion of Hume does for Kant is to make him realise that 
 the fact of such knowledge is one that needs explanation, seeing 
 that it is a knowledge that goes greatly beyond anything that 
 could be verified by any number of particular experiences. 
 The fact itself is an interesting problem : for it involves an 
 anticipation of experience, nay even an anticipation which goes 
 beyond, not only all that has been, but all that could be, given 
 in experience. Thus, when we lay down the principle that 
 the angles of a triangle can never be either more or less than 
 
 how far he was aware o£ the extent to which Ms first statements needed modi- 
 fication. He was certainly not aware of the full extent to which this was 
 required. 
 1 B. 20. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 235 
 
 two right angles, or that space has three and only three dimen- 
 sions (while time has only one^), we anticipate an endless 
 series of experiences, which we have not yet had, and which, 
 indeed, we never can have. We, as it were, dispose by antici- 
 pation of all space and its contents. In like manner, when 
 we say that matter is indestructible and that its quantity 
 cannot be' either increased or diminished, we lay down a law 
 which" we cannot possibly verify, and yet which we do not 
 he.sitate to express without qualification. We are thus dispos- 
 ing by anticipation of all time and its contents. That we do 
 this without hesitation is an undoubted fact ; and it is probable 
 that we should never raise a question as to' our right to do it, 
 were it not that the principles used in determining objects in 
 time and space and used, as the history of science proves, with 
 great success, seem to lead us into many difficulties When we 
 apply them in relation to objects in general; and were it not 
 that the ideas of time and space themselves give rise to similar 
 difficulties, whenever we ask whether they are or are not 
 universal conditions of objective reality. Thus the very ideas 
 and perceptions, which up to a certain point form the securest 
 guides for our investigations, seem beyond that point to break 
 down, and leave us face to face with insoluble problems. 
 Mathematics by its a priori determination of space arrives at 
 conclusions which are capable of innumerable applications, so 
 long as we are dealing with external objects ; but if we take 
 space as a universal condition of all objective reality, we are 
 led directly to the denial of the existence of spiritual sub- 
 stances ; and even in regard to the material world itself, we are 
 entangled in hopeless dilemmas, as to its limited or unlimited 
 extent, and the finite or infinite divisibility of its parts. In 
 like manner, the principle of causality is a necessary guide to 
 all our investigations in relation to physical objects ; but if we 
 treat it as an absolute principle, we are obliged to deny the 
 existence of any free or self-determining being and of any first 
 • 'A. 30; B. 46. 
 
236 THE CRITIQUE OF PURE REASON. book I. 
 
 or uncaused cause, — a conclusion which comes into conflict 
 with the principle of causality itself. Such results show that 
 we cannot treat even the most certain principles of science as 
 absolute ; and if they are not absolute, the question immedi- 
 ately arises, what is the ground of their validity within the 
 sphere in which they are valid ? For if we can discover that 
 ground, we shall know also why they are invalid beyond the 
 sphere of experience. 
 doiAt^ofthe a H^nce, " reasou has this strange fate in one of the regions 
 pKf'Sw". which it seeks to bring within knowledge, that it finds itself 
 burdened with problems, which it is unable to set aside, 
 because they are presented to it for solution by its own very 
 nature ; but yet that it is unable to solve these problems, 
 because they transcend all its powers." 
 
 " Into this perplexity it falls without any fault of its own. 
 It starts with principles, the use of which in experience it 
 cannot avoid, and which, indeed, experience abundantly 
 justifies it in using. With these it rises gradually (its own 
 nature impelling it) to ever higher, ever more remote, con- 
 ditions. As, however, it becomes aware that in this way 
 its work must ever remain incomplete, because every answer 
 gives rise to a new question, it finds itself compelled to take 
 refuge in certain principles which transcend all possible 
 empirical application, and which yet seem to be so little 
 open to suspicion, that even the common intelligence readily 
 admits them. But in so acting it plunges at once into dark- 
 ness and contradiction : and though it may indeed gather 
 from this result that there is some latent error in its procedure, 
 yet what that error is it is unable to discover. For the 
 principles it has been using, just because they reach beyond 
 the limits of all experience, can be brought to no empirical 
 test.'" 
 thdr*be?^'^ ^® ^^^ paraphrase Kant's view of knowledge as follows : — 
 thI'toS™! Experience means knowledge of things given us in sense, and 
 
 experience. 
 
 ' Preface to the first edition of the Critique. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 237 
 
 it is admitted that, prior in time to the impressions of sense, 
 there is no such knowledge. But this is not to be taten to 
 imply that there is nothing in experience which is not due , to 
 sense. By experience'^ we mean knowledge of particular facts 
 as connected by general principles. Now, the knowledge of 
 such principles cannot be legitimately derived from the impres- 
 sions of sense, which, if they give us knowledge at all, give 
 knowledge only of particulars as such. If it has any legiti- 
 piate derivation, or, in other words, if it is knowledge at all, 
 it must be derived directly or indirectly from the mind itself, 
 or, in other words, it must be a priori or based on what 
 is a priori. If, however, we take the highest principles of' 
 this kind which we can find, the principles of mathematical 
 and physical science, and if we follow them out as if they 
 were absolute principles, we find ourselves entangled in in- 
 soluble difficulties. Thus the principle of causality carries 
 us back from experience to experience, till it finally carries 
 us beyond all possible experience, and forces us either to 
 extend the succession of phenomenal causes ad infinitum, or 
 to suppose the existence of an uncaused cause, — a new concep- 
 tion, altogether opposed to that with which we have hitherto 
 been working. 'Thus we are brought face to face with an 
 antinomy which we cannot solve, for whichever of these alterna- 
 tives we adopt, we come into collision with that demand for a 
 cause to explain every effect, which had been our unerring 
 guide within the limits of sense-experience. Now, this means 
 that reason is at variance with itself, if the principle of 
 causality is to be taken as absolute; or, if not, then reason 
 must have in itself a higher principle, which will at once 
 prescribe, and limit, the application of this law of causality, 
 prescribe it within, and limit it to, the sphere of sense- ex- 
 perience. And if this latter alternative be the true one, the 
 
 ' There is a certain ambiguity in Kant's use of the word experience which it 
 is difficult to avoid in reproducing his thought. It is here used in its highest 
 sense, but it often means merely the particulars as given in sense. 
 
238 THE CRITIQUE OF PURE REASON. book i. 
 
 only question remaining will be, whether the limiting principle 
 can itself be used to extend our knowledge into the region 
 beyond the limit. 
 ^ow£*^the Now, what has been exemplified in the case of causality, 
 the ySty of is to be applied universally to all the general principles, by 
 ^^esepnnci- ^^^ ^.^ ^^ which our empirical knowledge, whether ordinary 
 or scientific, is developed. For, as will be shown more fully 
 in the sequel, all such principles, when treated as absolute, 
 lead to difficulties similar to those just mentioned. We have, 
 \ therefore, no security for those principles or for anything which 
 
 j ^ we know by means of them, unless we can see how such know- 
 ledge is possible, and at the same time, what is its limit. But 
 it is to be observed, that the doubt so cast on the principles of 
 experience, as such, is reflected back on them from their applica- 
 tion beyond experience, and that, therefore, the determination 
 of the question how they are valid within experience, has its 
 main value in relation to the further question of our knowledge 
 of that which is beyond experience. We ask how a priori 
 knowledge of objects is possible in the sphere of sense-experi- 
 ence, (or in the sphere of physical and mathematical science,) 
 in order that we may discover whether and how it is possible 
 beyond that sphere. Por, if it is not so possible, it must be 
 because some element was present in the former case which 
 is wanting in the latter. On the other hand, if we ask 
 whether a priori knowledge is possible in the region of sense- 
 experience, it is only because those very principles, which are 
 our safest guides in relation to the physical world, seem, when 
 we carry them beyond it, to lead to a Materialism, which itself 
 again breaks down in Scepticism. 
 critS tolb? I have already spoken of the rationality of the attitude which 
 furttif?r Kant thus takes up. It has, indeed, been argued that in Kant's 
 assumption of the truth of the mathematical and dynamical 
 principles, there was still an element of '' dogmatic slumber" 
 which he had not yet thrown off, and that a thoroughly 
 critical philosophy would need to begin with a less limited 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 239 
 
 doubt. But the answer is that doubts as to a coherent body of 
 science can legitimately arise only, as they do with Kant, at 
 its boundary, i.e., only because there is something, — some fact 
 or generally accepted idea, — with which the principles of such- 
 science come into collision when they are universalised, or 
 because, when so universalised, they come into collision with 
 themselves. In other words, doubt arises because some other 
 element of reality-r-some element which the science in question 
 has neglected, but which yet is seen to be necessarily connected 
 with its objects as part of the same whole with them — is not 
 accounted for by such science. Or, what is the same thing in 
 another aspect, doubt arises, because the object of such science, 
 when taken as a res completa, as a whole in itself which is not 
 part of a greater whole, breaks down in self-contradiction. , 
 Doubt in any other sense, as was shown in the first chapter, 
 involves the self-contradiction of absolute scepticism. It is 
 virtually an attempt by means of the intelligence to reach 
 beyond the limit of the intelligible world. 
 
 Kant's denial that we can know what is not given in experi- ganwiuo 
 ence is, essentially and in the first instance, a limitation to k™owfedg^to 
 experience, i.e., to the physical world, of those universal g^jj(j^^P™™°®- 
 necessary principles, by the aid of which alone we can reach 
 beyond what is immediately given in sense. This purpose is 
 a,Iready indicated in the investigations out of which the Critique 
 arises ; for in the Dissertation, as was shown in the last 
 chapter, his main aim was to separate the forms of time and 
 space, — as forms of sense and therefore applicable only to 
 phenomena, — from the pure a priori conceptions, through 
 which he still believed that noumena or things in themselves 
 might be known. In the Critique, he extends the same pro- 
 cedure to the pure conceptions, which also he shows to be 
 essentially determinations of phenomenal objects as such. The 
 effect of this is to leave only the ideas of reason as means of 
 determining things in themselves ; and these have in his view a 
 value, not for knowledge but only for thought, as problematical 
 
■240 THE CRITIQUE OF PURE REASON. book I. 
 
 conceptions which are made assertorial only in the faith of 
 practical reason. 
 (2) to set free TMs, however, means for Kant not the negation of all certi- 
 
 what is ' 
 
 ence from^'"' ^^^^ ^.s to that which lic's beyond the limits of experience, but 
 cOTdMons. rather the opposite. It means the limitation of the a priori 
 principles, by means of which alone knowledge can be extended 
 beyond what is given, to experience ; but it means also the 
 freeing of that which is beyond sense-experience from the 
 determinations to which all empirical objects as such, i.e., all 
 the objects of Mathematics and Physics, are subjected. It 
 means the denial that nature and necessity include every- 
 thing ; nay, it means the assertion that nature and necessity 
 have a significance only in relation to a reality which is beyond 
 them. For what Kant conceives himself to prove is that the 
 necessity of nature with all its spatial and temporal con- 
 ditions, is necessarily referred to a self, which is not a part 
 of nature because it is that for which nature is : a self also 
 which has in its consciousness of itself a principle by which 
 it can determine its own activity independently of nature. 
 When, therefore, it is alleged, as it has been alleged 
 by some, that the Critique of Practical Reason is an after- 
 thought, the object of which is to undo the negative 
 results of the Critique of Pure Reason, this is not only a 
 mistake as to the way in which Kant conceived his own 
 system, but it involves the separation of two elements which in 
 it are essentially related, viz., the limitation of experience and 
 the assertion of that which limits experience as being itself 
 beyond the limit. In a passage in the Prolegomena, Kant 
 makes it the main distinction of his own philosophy from that 
 of Hume, that in the former Metaphysic has not merely a 
 negative but also a positive value. " Hume called his destruc- 
 tive philosophy Metaphysic and attached a great value to it : 
 indeed, he tells us that Metaphysics and Morals are the 
 weightiest branches of science and that Mathematics and 
 Physics are not half so important. That acute writer was 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 241 
 
 here, however, looking only to the negative advantages which 
 speculative philosophy would gain from the moderating of its 
 excessive pretensions, inasmuch as this would take away the 
 very ground for many interminable disputes which confuse the 
 human race : but he lost sight of the ^positive evil which must 
 accrue, if reason should be deprived of those most important of 
 all prospects or outlooks, which enable it to set the highest 
 good as a motive of endeavour before the will."^ And in the 
 same spirit, in the preface to the second edition of the Critique, 
 after dwelling upon the negative use of criticism as limiting 
 knowledge to objects of experience, he adds that this limitation 
 of knowledge is by no means a denial of the possibility of a 
 consciousness of objects beyond experience. "We must be 
 able if not to know, yet to think, things in themselves. For, 
 otherwise, there would follow from the limitation in question 
 the preposterous conclusion that phenomena exist without 
 anything of which they are the phenomena or appearances. 
 Now, if we were to admit that the distinction, which the 
 Critique makes necessary, between things as objects of experi- 
 ence and the same things as things in themselves, is an unreal 
 distinction, we should be obliged to extend the principle of 
 causality, and with it the mechanical order of nature, to all 
 things whatsoever. It would then be impossible for us to say 
 of the same being, e.g., the human soul, that its will is free and 
 yet that it is subjected to the necessity of nature, i.e., that it is 
 not free, without falling into manifest contradiction ; for we 
 should then be obliged to take the soul in the same sense in 
 both cases, and could not, therefore, attribute to it contradictory 
 predicates. If, however, the Critique has not erred in teaching 
 that the object should be taken in two senses, — as phenomenon 
 and as thing in itself; if its deduction of the conceptions of 
 understanding holds good, and if consequently the principle of 
 causality has application only to things in the former sense, i.e., 
 as objects of experience, while in the latter sense, or as things 
 
 • 1 R. III. 7 ; H. IV. 6. 
 
 VOL. I. Q 
 
242 THE CRITIQUE OF PURE REASON. book i. 
 
 in themselves, they are not subjected to it : then the same will, 
 which in its phenomenal appearance, its visible activity, is 
 conceived as necessarily in accord with the law of nature and 
 therefore not free, may without any contradiction be thought 
 of as free, or not subjected to that law, when it is regarded as 
 a thing in itself. . . . The same explanation of the positive 
 use of critical principles may be applied to our conceptions of 
 God and of the simple nature of the soul, though for the sake 
 of brevity I pass them over for the present. But the general 
 result is that I cannot postulate the reality of God, freedom 
 and immortality, for the behoof of the practical use of reason, 
 unless I am able to deprive speculative reason of all its claims 
 to a transcendent knowledge of the real being of things. And 
 this I can do only if I show that, in order to attain such 
 knowledge, it must make use of principles which really apply 
 only to the objects of possible experience, and must treat 
 things in themselves as if they were phenomena. ... I 
 must therefore put knowledge out of the way in order to make 
 room for faith; whereas the Dogmatism of Metaphysic, i.e., 
 the groundless presumption that we can make our way in 
 Metaphysic without criticism, is the true source of all that 
 counter-dogmatism of unbelief that comes into conflict with 
 morality."^ 
 ^^ect orth'e^ One more passage may be quoted from a later treatise of Kant, 
 fSpiS."* which shows how constantly he looked at the deduction of the 
 a priori principles of experience as the stepping-stone to a certi- 
 tude, though a certitude not of knowledge but of faith, in relation 
 to that which is beyond experience. Speaking of transcendental 
 philosophy, he says that, "in it shice the time of Aristotle 
 not much progress has been made," a slowness of advance 
 which is due to the nature of the subject : " for, just as 
 Grammar resolves the form of a language into its elementary 
 rules, and just as Logic in a similar way resolves the form of 
 thought into its elements, so Transcendental Philosophy 
 IB. XXVIII. -XXX. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 243 
 
 resolves knowledge into the conceptions which lie a priori in 
 the understanding and have their use. in experience. It is a 
 system, the toilsome elaboration of which we might well have 
 spared ourselves, if our aim were only to determine the rules 
 of the right use of those conceptions and of the principles 
 based on them, with a view to empirical knowledge. But the 
 case is very different, if our purpose is to advance from the 
 sensible to the supersensible. For then it becomes absolutely 
 necessary to measure out, in the most thorough and careful 
 way, the faculty and principles of the understanding, in order 
 that we may discover from what starting point and by ineans 
 of what aids and instruments, the reason may make its way 
 upwards, from the objects of experience to those objects which 
 are beyond experience."^ 
 
 ^ R. I. 489 ; cf. also I. 559 ; H. VIII. 520, 577. It may be said that 
 air these passages (from the Prolegomena, the second edition of the 
 Critique, and the Essay ore the Progress of Metaphysic. since Leibniz) are 
 taken from works written after Kant had been alarmed by the ' ' idealistic " or 
 Berkeleiau conclusions deduced from the first edition of the Critique. But the 
 doctrine that we are able to thinh noiimena though not to know them, belongs 
 also to the argument of the first edition. No doubt this ' thinJdng, ' as will be 
 shown more fully in the sequel, involves a really synthetic movement of pure 
 thought, which is inconsistent with Kant's view of it as essentially analytic. 
 But in any case it is an element necessarily involved in Kant's fundamental 
 conception of self-consciousness, which cannot be removed from his OiiigMe with- 
 out destroying it. All that can be admitted, therefore, is that Kant was led, by 
 the direction of the attack made upon him, to lay more emphasis on an aspect of 
 his argument to which he had hitherto given less attention. No doubt this gave 
 rise to a certain readjustment of the relation of his own to other philosophies ; 
 but even this was mainly a consequence of the immanent development of his 
 system ; of the necessity under whiohhe was of looking at its principle on a difi'er- 
 ent side when he came to apply it to morals ; and also of that clearer conscious- 
 ness of his own meaning, which he gained as he escaped from the labour of 
 detail and was able to regard his system as a whole. The extent to which 
 this development or change went will be considered more fully in the sequel. 
 In one important point only is there a change in the second edition, viz., as 
 to the relation of inner and outer sense. But I shall afterwards show that, in 
 this addition to the Critique and in the new Refutation of Idealism which 
 is connected with it, Kant was not really recoiling towards that common sense 
 realism, which, in the first edition, he had left behind, but advancing towards 
 that more complete and consequent application of the principle of his Transcen- 
 dental Deduction, which alone could clear it from the inconsistent psychological 
 elem'ent which had adhered to his first statement. 
 
244 THE CRITIQUE OF PURE REASON. book i. 
 
 Is there any It appears, then, that Kant's limitation of knowledge to ex- 
 change lu ^^ ' ' 
 
 ofTheMiaSon perience, and especially his denial of the possibility of any 
 to thought to a priori extension of knowledge beyond the limits of experi- 
 
 his later _ - . 
 
 works ence, may easily lead to a misunderstanding, unless we keep m 
 
 view two things : in the first place, that this limitation is based 
 on the reference of experience, (so far at least, as it is deter- 
 mined by a priori principles, and, therefore, so far as it goes 
 beyond the data of sense,) to a principle of unity in the conscious 
 self which is not an object of experience; and, in the second place, 
 that the ultimate end sought in this limitation is not merely 
 the refutation of dogmatism and of scepticism, but the eman- 
 cipation of the self and also of the other supersensible realities, 
 (the ideas of which arise out of, or in connexion with, the con- 
 sciousness of self,) from the conditions under which they must 
 be brought, if they were objects of experience. It is true that 
 the importance of this second object is more definitely insisted 
 on in the Prolegomena than in the first edition of the Critique 
 and in the second edition of the Critique than in the Prolego- 
 mena. This, however, may be explained without supposing 
 any actual change in Kant's view. It may be admitted that 
 Kant sometimes had his ideas drawn out of focus by the con- 
 centration of his thought on his immediate task, and that he 
 sometimes makes statements which are only to be explained 
 by supposing that, for the moment, he is not taking due account 
 of the relation of his immediate argument to the whole course 
 of reasoning of which it is a part. The mixture of the con- 
 scious and the unconscious in all original intellectual work is 
 very subtle, and a writer may often become aware that he has 
 not expressed, or not expressed exactly, all he meant, only 
 through the misunderstanding of others. Kant was perfectly 
 honest, if it is necessary to say so, in asserting that all the 
 changes introduced into the text of the second edition of the 
 Critique (with one or two specified exceptions) were merely 
 formal improvements in the manner of statement of a view 
 which he had all along been maintaining ; and undoubtedly, he 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 245 
 
 would have said the same thing of his latest restatement of 
 that view, in the essay on the Progress of Metaphysics since the 
 Time of Leibniz. Nevertheless, it may be admitted that through 
 all these formal changes Kant was progressing towards a clearer 
 apprehension of his own thought, such as was tantamount to 
 a development of it. With regard to the point to which we 
 have been specially referring, viz., the relation between the 
 negative and the positive uses of metaphysic, Kant's final 
 survey of the ground has a clearness and balanced fulness 
 which we do not find at any earlier point. But a fair criticism 
 will recognise that there is no external change, but .only a 
 development of ideas which are already present in the first 
 edition of the Critique. 
 
 In saying so much, however, we have to add the caution Theprobiem 
 that the problem of the Critique has an essentially dialectical K'JSfgoe^on 
 character ; or, in other words, that it changes its form as Kant 
 advances from one stage of its solution to another. Or, perhaps 
 it would be more accurate to say that, in dealing with the 
 special questions which need to be settled with a view to his 
 general problem, Kant comes upon answers which force him to 
 transform those very questions, and ultimately to modify the 
 general problem itself. The general problem was, as we have 
 seen, to discover the conditions of an a priori knowledge of 
 sensible objects which we are assumed and acknowledged to 
 possess, with a view to determine the possibility of a similar 
 knowledge of supersensible objects, of which we cannot be 
 assumed to be even capable. Thus Kant was not, in the first 
 instance, inquiring into the conditions of knowledge generally, 
 but simply into the conditions of a priori knowledge ; for, as it 
 is obvious that there can be no a posteriori knowledge of super- 
 sensible objects, it is only in as far as the knowledge of 
 empirical objects is a priori that it can cast light upon the 
 main problem which Kant has in view. Furthermore, his only 
 concern with the a priori knowledge of empirical objects is 
 simply to explain it, to exhibit the conditions which make it 
 
246 THE CRITIQUE OF PURE REASON. book i. 
 
 possible, and not to vindicate it or prove by any other prin- 
 ciples its acknowledged validity, 
 ofth"'' ^''™' But as Kant advances in his attempt to deal with the 
 prove\hat ° problem of the conditions of our a priori knowledge of empiri- 
 
 thereiaa .* i • 
 
 priori know- cal obiects, the nature and direction of his argument is gradually 
 
 ledge and that J > o o j 
 
 aionelxp'eri- changed, and that in two ways. I n the first place, instea d of an 
 possible explanation of the conditions of an a priori knowledge wh ich is 
 assumed to exist, we find Kant giving us a proof that it does 
 exist.; and in the second place, instead of an account of the 
 A conditions under which one kind of knowledge, viz., a priori 
 knowledge, of empirical objects is possible, we find him giving 
 us an explanation of the possibility of knowledge or experience 
 in general. These changes are at first very confusing, and they 
 have naturally led to the charge against Kant that his 
 argument is simply a " vicious circle " : yet a careful con- 
 sideration will show that they are the inevitable results of the 
 development of the problem itself. The difficulty alters as we 
 attempt to solve it ; and the only answer that can be given to 
 Kant's first question is one which shows that it is not the right 
 question, or at least that in it the problem does not take its 
 proper form. Hence what appears as a " vicious circle " is 
 really the result of the fact that the first question involves an 
 uncritical view of things : — or, in other words, that it involves 
 at once an unreasonable dogmatic assumption and an un- 
 reasonable doubt based on that assumption : and we can 
 only attack the doubt through the assumption. The steps 
 of this transition, as it is made by Kant, are as follows. 
 In the first place, the a priori principles of mathematics and of 
 pure physics, which are the sources of our most certain know- 
 ledge of empirical objects, and which were at first assumed as 
 unquestionable, — as, in fact, shown to be possible by the fact 
 of their existence, — become subjected to a doubt which is 
 reflected back upon them from their use beyond the limit of 
 sensible experience. As is pointed out in the first words of 
 the preface to the first edition of the Critique, it is in the 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 247 
 
 natural extension to what is beyond existence of the principles 
 which have guided us to so many valuable results within 
 experience, that we fall into the antinomical difficulties 
 of metaphysics.'- But these difficulties awake a doubt of the 
 principles that led us into them, which cannot be removed 
 until we are able to draw a line between the sphere in which 
 they are valid and the sphere in which they cease to be 
 valid. From this point of view Kant who at first, as 
 has been well said, was " led by Hume to call into question 
 not a priori science, but only the account of it given by 
 the Leibnizian school," ^ became conscious that that science 
 itself needed a vindication. Hence we find Kant saying that 
 Hume, because he did not distinguish the objects of experience 
 from things in themselves, was naturally led to regard the idea 
 of cause as itself illusive ;^ and that this view, when univer- 
 salised, must necessarily lead to a " universal scepticism," — a 
 scepticism not only in regard to all the knowledge which pure 
 reason claims for itself, i.e., not only in regard to metaphysics, 
 but also in regard to physics and mathematics, and even in 
 regard to the ordinary use of the intelligence. In other words, 
 the ai'gument from the reality of mathematical and physical 
 science to its possibility is disturbed by the fact that at a 
 
 1 Quoted above, p. 2.36. 
 
 ^ Cohen, Kant's Theorie der Erfahrung, p. 55. 
 
 3 E. VIII. 170 ; H. V. 50. Cf. A. 765 ; B. 793. Note that in these passages 
 Hume's Empiricism is conceived by Kant as in the first instance producing 
 only a scepticism as to the possibility of knowing things in themselves. But, 
 as Kant goes on to argue, the general repudiation of a priori principles involves 
 a denial also of the a priori elements of experience, and thus deprives experience 
 of that which is essential to it as a knowledge even of phenomenal objects. Hume, 
 therefore, oannotstop short of universal scepticism. Cf. A. 762; B. 790. "If we 
 cannot make intelligible the possibility of principles of the understanding which 
 anticipate experience, we may at first doubt whether we really have such a priori 
 
 possessions, but we cannot assert the impossibility of our having them 
 
 We can only say that if we could determine their origin and continuance, we 
 should be able to determine the compass and limits of our reason. " In this 
 way we have a right to express "a thoroughgoing scepticism as against all 
 dogmatic philosophy which proceeds on its way without previous criticism of 
 reason, but not to interdict such positive procedure when it has been duly 
 prepared for and secured against doubt by such criticism. " 
 
248 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 particular point the principles, on which such science is based, 
 break down and give rise to antinomies ; and so long as 
 the region of error and illusion is not marked off from the 
 region of truth, the doubt, which visits the former, must 
 invade the latter. It is only when the objectivity of empirical 
 phenomena is distinguished from the objectivity of things in 
 themselves, that scepticism can be prevented from passing from 
 the region beyond, to the region within, experience. Hence, 
 that a priori knowledge of the things of experience, which at 
 first was assumed as a fact that needed no proof but only an ex- 
 planation, comes to be regarded as itself requiring a "deduction" 
 or vindication. An account of the conditions of a knowledge 
 which we are supposed to have passes into a proof that it is 
 possible for us to have it. And this, of course, involves that 
 we are not allowed to reason from it as something certain 
 to something else which is proved by means of it, but that 
 we are obliged first to establish its truth on the ground of 
 the previously ascertained truth of something else. In other 
 words, we do not merely assume a priori principles as true of 
 the objects of experience and ask for an explanation of the 
 conditions on which such knowledge depends, but we assume 
 ordinary experience as our starting point and basis and argue 
 to the truth of the a priori principles, without which it could 
 not exist. 
 Mste^rihaa ^^^^ -^^^^ sentence, however, points to another change of the 
 apart froSThe problem wliich Kant's argument brings with it. The contrast ■ 
 with which Kant began was the contrast of a priori truth, 
 which as such is necessary and universal, with empirical truth, 
 which as such is particular and contingent. According to the 
 view thus presented, experience can tell us that an object has a 
 particular predicate at the moment when it is perceived, but 
 not that it must have it, or therefore, that it always has it : 
 for, unless we are able to recognise a necessary connexion 
 between subject and predicate, we cannot, from the perception 
 of their particular coexistence or succession, gather anything as 
 
 (I przon. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 249 
 
 to their relations at any other time or place. It follows, then, 
 that we can base universal propositions only on necessity of 
 connexion in thought. Thus what Hume states as to causality 
 must be universalised. Our sensible experience does not 
 authorise us to go beyond particular judgments of coexist- 
 ence or succession, or to say anything about the universal 
 relations of the predicates of an object to each other or to their 
 subject. But again such particular judgments tell us merely 
 what we individually experience at a particular time : they are 
 merely "judgments of perception," that indicate particular 
 states of our subjectivity ; and the question arises how we can 
 ever go beyond such judgments. As a matter of fact, we iind 
 ourselves continually making judgments which refer, not merely 
 to appearances of our subjectivity at a particular time, but to 
 objects which are conceived as permanently existing, even when 
 they are not perceived ; and we find ourselves, not merely 
 recognising successions of appearances of these objects for our 
 subjectivity, but recognising them as definite successions of 
 states in the objects themselves, which take place in them as 
 existences independent of our ' perceptions, and have their 
 ground in their nature as such existences. When we look 
 at it in this way, we begin to see that apart from a priori ' 
 principles which may enable us to go beyond our particular 
 experiences, these experiences would reduce themselves to mere 
 passing appearances of our own subjectivity. In other words, 
 even particular experiences of objects, as such, become im- 
 possible, unless we are able to transcend them. A priori 
 principles are necessary to experience as a consciousness of 
 objects, and the denial of a priori knowledge of objects means 
 the denial of all knowledge. Hence arises a necessity for a 
 further statement of the problem of the Critique. Kant began 
 by asking for the conditions of the a priori knowledge of 
 empirical objects^ taking that as one species of knowledge, 
 which can be set alongside of a posteriori or empirical know- 
 ledge : but now he finds that a posteriori is impossible without 
 
250 THE CRITIQUE OF PURE REASON. book'i. 
 
 a priori knowledge. From this point of view he declares that 
 "the highest problem, of transcendental philosophy is, How 
 experience is possible ? "^ In other words, particular experience 
 is taken as a fact, and the a priori principles as conditions which 
 are necessary to explain it, and which are " deduced " or vindi- 
 cated by showing that they are so necessary. The ultimate result 
 which we thus reach under Kant's guidance is a negative 
 argument against scepticism as to our a priori knowledge, 
 which shows that there is no foundation, no basis of proof, on 
 which scepticism can take its stand in order to direct an 
 attack against such knowledge. For that very consciousness 
 of the particular and contingent, which Hume had turned 
 against the consciousness of " necessary connexion," is itself 
 dependent upon the a priori it is used to condemn. This, I 
 say, is the ultimate result we reach under Kant's guidance ; but, 
 as we shall find, it is not clear that Kant himself ever fully 
 reSlised, (it is abundantly clear that he did not always realise,) 
 the degree to which his argument had altered the premises 
 from which at first it started. This point, however, will be 
 discussed more fully at a later stage in our inquiry.^ Here it 
 may be sufficient to remark that no one has the key to Kant's 
 logic, who does not see that this is the result to which it tends, 
 and no one does him justice who does not give him full credit 
 for it. 
 thought'™"* We have, however, to observe that there is another com- 
 knowiedge plcxity in Kant's thought, in which we may find an explana- 
 
 limits his 
 
 argument. tion of the fact that the above argument was not carried by 
 him to its ultimate result. While Kant argues that a priori 
 knowledge of the objects of experience is possible, because without 
 
 ^ E. I. 507 ; H. VIII. 536. It is a point worthy of notice that when Kant 
 thus brings the centre point of the Critique within the Analytic, he is on the 
 way to establish such a unity of the a priori and the a posteriori within experi- 
 ence, as makes intelligible the further step by which, in the Critiqun of 
 Judgment, he brings together the law of nature and necessity with the law of 
 freedom, using the idea of final cause or design as a middle term. 
 
 2 See below, Chap. IV. — The Transcendental Deduction of the Categories. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 251 
 
 it experience would not be possible ; on the other hand, he also 
 argues that such a prinri knowledge is not possible through 
 pure thought alone, but only through pure thought in relation 
 to the forms and matter of sense. For the pure a priori 
 principles of the understanding, when severed from the 
 forms and matter of sense, shrink into forms of analytic 
 judgment : i.e., into forms of a thought which never goes 
 beyond its own subjective self-identity so as to add to its 
 content or to apprehend an object. This view of pure thought 
 has already been referred to in the last chapter : but it requires 
 a fuller explanation. For the question of the Critique, it may 
 be observed, is stated in the Introduction in two alternative 
 forms : how is a priori knowledge of objects possible : and how 
 are a priori synthetic judgments possible ? Now the equival- 
 ence of these two questions is not at once obvious : yet a clear 
 view of their relation is essential to the right understanding of ♦ 
 the Critique. 
 
 The distinction of analytic and synthetic judgments in Kant anli*'ti?and°' 
 had a double root. On the one hand, it was connected with the judgments: (i) 
 Leibnizian distinction between the principle of identity and the by woiff: 
 principle of sufficient reason ; on the other with the Lockian 
 opposition of ' trifling ' and ' instructive ' propositions. It was 
 in the former aspect, as we have seen, that the germ of the dis- 
 tinction appeared in Kant's earlier works. Wolff, by reducing 
 the principle of sufficient reason to a form of the principle of 
 identity, had in effect denied that there is any principle of 
 a priori synthesis. And Kant only drew the necessary inference 
 when he proceeded (in his treatise On the Introdtwtion of the 
 Idea of Negative Quantity into Philosophy) to contrast the 
 logical relations of ideas with the real relations of things, 
 and to maintain that we needs must go beyond the former 
 to attain a knowledge of the latter. The logical ground 
 of the affirmation of a predicate is that it is analytically 
 contained in the subject; the logical ground of the denial 
 of a ♦predicate is that it is not so contained. But the real 
 
252 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 ground or cause why a thing comes into existence which 
 was not in existence before, or why a thing ceases to exist 
 which existed before ; or, putting it in another way, the real 
 ground of the existence or non-existence of anything, is not 
 thus given. Here the mere laws of identity and contradiction 
 will not help us, and we are thrown, according to Kant's view 
 at that time, altogether oh experience. For a dogmatic a priori 
 philosophy must develop itself purely by the aid of the laws 
 of thought, and if it cannot reach the real reasons or causes of 
 things by means of those laws, it cannot reach them at all. 
 Another consecpeuce necessarily follows, which is expressed by 
 Kant in the treatise on the Sole Rational Ground for a Proof of 
 the Being of God : viz., that existential propositions cannot be 
 due to pure thought. For thought, confined to analysis, can 
 only bring out clearly what is contained in a conception already 
 in the mind ; it can never enable us to go beyond our subjec- 
 tive conceptions so as to predicate existence of the object of 
 our thought. " Existential propositions are synthetical," and 
 synthesis is impossible to the pure intelligence, because it 
 cannot be achieved by aid of the principles of identity and 
 contradiction. At this time, therefore, Kant's view was that 
 to assert that an object exists for any conception, or to add to, 
 or take from, that conception, is impossible by any a priori 
 synthesis, and possible only by means of experience. The mind 
 is thus confined, so far as its own movement is concerned, to a 
 tautological judgment in which a conception is predicated of 
 itself, or in which, at most, the predicate only makes clear what 
 is already thought in the subject. It cannot of itself discover 
 any new connexion of predicates, nor can it go beyond its 
 own ideas so as to assert that any object of its thought exists. 
 The consciousness of the connexion of the predicates of the 
 object with each other as well as of the connexion of the 
 object with the subject that knows it, must be due to 
 impressions received from the object. How, through such 
 affections of our own, we could become conscious of an object 
 
CHAP. I. THE PROBLEM or THE CRITIQUE. 253 
 
 which is not such an affection, Kant did not yet attempt to 
 show. 
 
 The same distinction of analysis and synthesis, which grew ^^^^/^f' 
 out of Kant's criticism of Wolff, presented itself in a somewhat ^°°^^- 
 different aspect in the philosophy of Locke. Supposing know- 
 ledge of objects through bare thought to be impossible, is it any 
 more possible through bare perception or sensation ? In the 
 fourth book of his Essay, Locke assumes that the individual sub- 
 stance with all its qualities is given in perception, but that the 
 knowledge we receive of it through perception is strictly limited 
 to the moment in which we perceive it. We cannot learn from 
 it anything but a particular combination of predicates in a par- 
 ticular subject at a particular moment. We cannot say that 
 that subject will possess those predicates at any other time, 
 still less can we assume that the class of subjects to which we 
 give the same name will always possess them. If we make any 
 such statement, the subject of our predicates is only a " nominal 
 essence," i.e., we are speaking only of an idea in our own minds 
 to which nothing may correspond in reality ; and all that we can 
 do is to state what is contained in that idea, or in other words 
 what we mean by a particular name. Hence, general certainty 
 is never to be found but in our own ideas, and all general 
 propositions are " trifling," not " instructive " — i.e., they only 
 analyse, our ideas. And the counterpart of this is, that all 
 " instructive " propositions are singular ; in other words, we can 
 combine a new predicate with a subject only in so far as we have 
 the combination presented in perception ; though the " instruc- 
 tion " thus gained does not go very far, since we have no right 
 to generalise either the subject or its relation to the predicate. 
 All we learn is, e.g., that " this piece of gold is soluble in aqua 
 regia," or even strictly speaking, only that it, is so soluble at 
 this moment. It follows that " general propositions on matters 
 of fact " are impossible, because they would be instructive 
 or, in Kant's language, synthetic, which is possible only for 
 sii%ular propositions. On the other hand, if singular proposi- 
 
254 
 
 THE CRITIQUE OF PURE REASON. 
 
 BOOK I. 
 
 Influence of 
 Hume. 
 
 tions are instructive or synthetic, their synthesis does not 
 furnish a sufficient basis for a science of nature, seeing they 
 only state a particular combination of the subject and predicate 
 from which no general conclusions can be drawn. 
 
 Hume's criticism of the ordinary and scientific use of the 
 principle of causality was little more than a new application of 
 Locke's previous criticism of the principle of substance. The 
 latter pointed out that the coexistence of one quality along with 
 others can be known only as it is perceived, and that perception 
 does not give a ground for asserting any universal relation be- 
 tween those qualities such as is implied in asserting that they 
 belong to one substance. Hume only adds, that the succession 
 of two different qualities in a substance, in other words, a change, 
 can be known only as it is perceived, and that this perception 
 also affords no ground for asserting any universal relation be- 
 tween what precedes and what follows. The connexion, as 
 given in perception, is a singular one, from which no general 
 conclusion can legitimately be drawn. If we generalise such a 
 relation, we are no longer speaking of the facts of perception, 
 but of some ' nominal essence ' which exists in and for thought 
 alone.' 
 
 Now, if Kant abandoned the rationalistic philosophy of "Wolff, 
 explains ''^"^ bccausc he found that the pure activity of thought does 
 not explain synthesis, he soon began to learn from Locke and 
 Hume that the synthesis which can be explained by perception 
 is very narrowly limited ; that by it, in short, we can explain 
 only particular and not universal judgments ; i.e., only judg- 
 ments expressing the presence of a particular quality along 
 with others in a particular perception, and not a universal 
 relation of those qualities in an object which exists when it is 
 not perceived ; or, only judgments expressing the particular 
 sequence of the states of an object of perception which 
 corresponds to a particular succession of perceptions in the 
 
 1 Green's Introduction to Hume, § 223, seq., and Green's Philosophical Works, 
 II. p. 2, seq. 
 
 Neither 
 thought nor 
 
•^.t-UK' 
 
 CHAP. I. THE PROBLEM OE THE CRITIQUE. 255 
 
 subject, and not judgments expressing any general rule as to 
 the succession of such states under given conditions in an 
 object which exists independently of its being perceived. 
 
 When under the converging influence of the English and of'matoemati- 
 the German philosophy, Kant became aware of the difficulty as treated by 
 
 Locke. » , 
 
 of the question of a 'priori synthesis, his thoughts turned tj^ g.Jbt'Wc. ft****" 
 Mathematics, which both Locke and Hume had admitted to ^/" i. ,-, 
 stand on a different footing from all other knowledge, inasmuch '-- •' 
 
 as the generality of the propositions laid down in this science 
 was not conceived to preclude their being " instructive." Locke 
 had explained this on the ground that the ideas of Mathematics 
 are not ' ectypes ' but 'archetypes'; not derived from any 
 reality independent of the mind, and, therefore, not limited by 
 the perceptions in which alone such reality is supposed to be 
 given. The mind here constructs its own objects and can,! 
 therefore, determine their relations universally. The question 
 of Kant, how we can know that principles which are thus laid 
 down by the mind independently of experience, and which 
 confessedly go beyond anything that can be given in experi- 
 ence, have yet objective validity and are not mere constructions 
 of the imagination, had occurred to Locke ; but it was disposed 
 of so easily that he evidently did not comprehend its import- 
 ance. " Eeal things," he declares, " are no farther concerned, 
 nor intended to be meant, by any such propositions, than as 
 things really agree to those archetypes in his mind. Is it true 
 of the idea of a triangle, that its three angles are equal to two 
 right ones ? It is true also of a triangle wherever it really 
 exists. Whatever other figure exists that is not exactly 
 answerable to that idea of a triangle in his mind, is not at all 
 concerned in that proposition." ^ But the difficulty is that, in 
 this case we are supposed to be able to identify the present 
 object of perception with the ' nominal essence ' or general 
 idea of a triangle, and then to go on to predicate of that 
 particular object all the other properties which we have dis- 
 • ' Locke, Essay, IV. 4, 6. 
 
256 THE CRITIQUE OF PURE REASON. BOOK j. 
 
 covered in or developed from the idea ; we are supposed, in 
 fact, to be able to make general propositions which are not 
 trifling, and to apply them, not merely to a ' nominal essence ' 
 or idea, but to a real object. On this view, therefore, thought 
 without perception is synthetic as well as general, and its 
 general synthesis is objectively valid. 
 Hume's treat- Humc in his earlier treatise attempted to escape from this 
 
 ment of it. -^ ^ 
 
 difficulty, and to trace back the ' ideas ' of Mathematics to 
 ' impressions ' of space and time : but the attempt led him to 
 the denial of the objective validity of mathematical truth, in 
 so far as it goes beyond the possibility of empirical measure- 
 ment. In his Inquiry he rather avoids the discussion, and 
 uses language which is not essentially different from that of 
 Locke. " All the objects of human reason or inquiry," he says, 
 " may naturally be divided into two kinds, to wit, Eelations of 
 Ideas and Matters of Fact. Of the first kind are the sciences 
 of Geometry, Algebra, and Arithmetic ; and in short every 
 affirmation, which is either intuitively or demonstratively 
 certain. That the square of the hypothenuse is equal to the 
 squares of the two sides, is a proposition which expresses a 
 relation between these figures. That three times five is equal 
 to the half of thirty, expresses a relation between these num- 
 bers. Propositions of this kind are discoverable by the mere 
 operation of thought, without dependence on what is anywhere 
 existent in the universe. Though there were never a circle or 
 triangle in nature, the truths demonstrated by Euclid would 
 for ever retain their certainty and their evidence." 
 Seas and"' "Matters of fact, which are the second objects of human 
 
 matters of j_ j_ • i • j i 
 
 fact. reason, are not ascertamed m the same manner ; nor is our 
 
 evidence of their truth, however great, of a like nature with 
 the foregoing. The contrary of every matter of fact is still 
 possible ; because it can never imply a contradiction, and is 
 conceived by the mind with the same facility and distinctness, 
 as if ever so conformable to reality. That the sun will not 
 rise to-morrow, is no less intelligible a proposition, and implies 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 257 
 
 no more contradiction, than the affirmation, that it will rise. 
 We should in vain, therefore, attempt to demonstrate its false- 
 hood. Were it demonstratively false, it would imply a con- 
 tradiction, and could never be distinctly conceived by the 
 mind."^ Hume then goes on to say that, in his opinion, the 
 sole objects of the abstract sciences of demonstration are 
 quantity and number, and that all attempts to extend this 
 more perfect species of knowledge beyond these limits are 
 mere sophistry and illusion ; and he concludes with the well- 
 known saying — "When we run over the libraries, persuaded 
 of these principles, what havoc must we make ? If we take 
 in our hand any volume of divinity or school metaphysics, for 
 instance ; let us ask. Does it contain any abstract reasoning 
 concerning quantity or number ? JVo. Does it contain any 
 experimental reasoning concerning matter of fact and existence ? 
 No. Commit it then to the flames : for it can contain nothing 
 but sophistry and illusion." 
 
 Hume's assertion, that Mathematics deals with relations of Rant's under- 
 standing of 
 
 ideas, and that a denial of its propositions involves a con- ^"™*- 
 tradiction, was understood by Kant as meaning that these 
 propositions are merely analytic,^ and that, therefore, their 
 generality raises no question as to their objective validity. And, 
 in his ignorance of Hume's earlier treatise, he even thinks that 
 Hume would have been led to question his whole view of 
 knowledge, if he had seen that it involved a denial of the 
 truth of Mathematics. This view came the more easily to 
 Kant, because he himself held that mere ' relations of ideas ' as 
 such, relations established by the pure activity of thought, can 
 only be relations of identity or contradiction, such pure activity 
 being merely analytic. The point, therefore, to which he was 
 brought by his study of the English philosophy was to recog- 
 
 ^ Inquiry concerning Suman Understanding, Section IV. Part 1. 
 
 ^ An assertion which is accurate, in so far as it was Hume's first principle 
 that ideas cannot be anything but copies of impressions ; but not accurate in 
 the sense that Hume had considered the question of the possibility of synthetic 
 aj^ori judgment and tried to disprove it. See above, p. 203. 
 VOL. I. E 
 
258 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 nise that in mere perception as little as in mere conception is 
 there a ground for any synthesis which goes beyond the data 
 immediately given in sense, and that even Mathematics must 
 be explained by an empirical philosophy as a mere analysis of 
 certain general ideas got by abstraction from perception, but 
 which as general ideas must be merely ' nominal essences.' 
 Kant^s The case of Mathematics, however, as was obvious to Kant, 
 
 doctrine as to 
 
 ^nSs""*^ could not be satisfactorily treated in this way. For in Mathe- 
 matics there is a continual acquisition of new truths which are 
 all expressed generally, and which are at once applied with 
 perfect certitude to objects of perception, nay, which enable us 
 to prescribe beforehand the conditions under which these 
 objects must be perceived. Here, therefore, we have to provide 
 for the case of a synthesis which, as universal, cannot be 
 a posteriori, and which nevertheless has to do, not merely with 
 ' nominal essences ' or ' relations of ideas,' but with ' matters of 
 fact.' If we can explain this general synthetic movement of 
 thought in Mathematics in such a way as to justify its objec- 
 tive application, we may be able also to explain the other 
 instances, beyond the sphere of Mathematics, in which a 
 synthesis, which as universal must be a priori or at least 
 rest on an a priori basis, has an objective value. 
 
 A prion Kant, then, starts from this point of view. He does not 
 
 synthesis m ' ' ^ 
 
 ^a'^wS dispute, at least in the first instance, that there is an empirical 
 ence and ' synthcsis which enables us to make particular ' judgments of 
 
 agreement. 
 
 perception ' regarding phenomena or particular objects pre- 
 sented to us in perception. It may be the result of his criticism 
 to show that such judgments are impossible, seeing that it 
 ultimately leaves no ' object ' about which such ' judgments of 
 perception ' could be made. But, in the first instance, Kant 
 simply points out that, admitting that a particular synthesis 
 can be mediated by sense, which perceives an individual now 
 and here in its particular determinations, we cannot so explain 
 a universal synthesis, which yet is no mere arbitrary combina- 
 tion of ideas, but expresses a universal truth of fact. Nor, again, 
 
CHAP. I. THE PROBLEM OE THE CRITIQUE.' 259 
 
 can we thus explain how a particular synthesis of experience, 
 such as the proposition that A follows B, should be made the 
 ground for establishing a universal law of nature. Now the 
 former is what we see in Mathematics, the latter is what we 
 see in Physics. In Geometry, e.g., we determine space, as repre- 
 sented in perception or imagination, in certain arbitrary ways 
 so as to produce certain figures, and in the figures thus set 
 before us we demonstrate certain properties. These properties, 
 demonstrated in this particular ease, we then at once assume 
 to be demonstrated for all such figures, and we transfer the 
 results of this arbitrary process of construction in the form of 
 general laws to the objects which we perceive in experience. 
 In Physics we are not able thus to begin by an arbitrary deter- 
 mination of space or time, and by it to establish universal laws 
 for the synthesis of perceptions, but we are able to universahse 
 a connexion of perceptions which is given in a particular case. 
 Thus having perceived that A follows B, we assert universally 
 that A will always follow B in similar circumstances. Now 
 the question which Kant asks himself is. How is this possible ? 
 How are we able to make universal propositions as to matters 
 of fact prior to all actual experience, and how are we able to 
 universalise particular judgments of perception so as to base 
 general laws of nature upon them ? 
 
 The answer which Kant gives to this question is given in rte answer to 
 
 the question 
 
 successive stages, and the question itself somewhat alters as we ^^^o a prion 
 
 ° -'- . synthesia m 
 
 advance from one stage to the next. In the Aesthetic he *''° "*^"*'''''- 
 abstracts from any activity of thought in perception or in 
 judgments of perception. He defines sensibility as " a recep- 
 tivity for the perception of objects," or more fully as a capacity 
 of " being affected by objects so as to acquire through this 
 affection an immediate idea or representation of them." He 
 assumes, therefore, or he does not question the assumption, that 
 sense enables us to make judgments of perception, in which we 
 refer sensations to particular objects and so determine them as 
 having particular qualities. The only question here is, how in 
 
260 THE CRITIQUE OF PURE EEASON. book I. 
 
 Matliematics, or in the judgments of common sense out of 
 which Mathematics arises, we should be able to anticipate 
 the synthesis of empirical perception, and without its aid 
 to lay down universal laws as to its possibilities. And the 
 answer to this question is, that in order to make such a priori 
 determination of objects of perception possible, time and space 
 must be forms of our perception, i.e., the very conditions on 
 which affections give rise in us to perceptions of objects. For 
 if that be the case, then they will, of course, condition the 
 objects of perception ; and whatever determinations we are able 
 to discover as necessary for time and space will hold good for 
 all objects, which must be perceived as in time and space. 
 As presented in perception, all external objects are determined 
 as standing in spatial relations to other external objects, and 
 all states of objects whatsoever, even our own states, are deter- 
 mined as standing in relations of time to the states of these 
 and other objects. Just so far, therefore, as we can trace out 
 relations between different figures, or limitations of space, and 
 between different periods, or parts of time, we are able a priori 
 to determine relations which will beloijg to objects as occupying 
 space or time. Nay, we may go fiirther and say that it is 
 only when determined as occupying particular parts of space 
 and time that objects can be objects of perception for us at all. 
 Thus the universality of the synthesis by which we determine 
 figures in space merely means that this synthesis must neces- 
 sarily be transferred to the matter of sense, in the same process 
 by which it becomes, or gives rise to, the perception of an 
 object. The external object as such has this determination in 
 space as a prior condition of any further determinations, or 
 special qualities, which we apprehend by touch or sight or any' 
 of the other senses. " This is the theory that Space and Time' 
 are only subjective forms of our sensuous perception' and not 
 determinations belonging to objects themselves, but that, just 
 for that very reason, we can determine these our perceptions 
 a priori with a consciousness of the necessity of the judgments 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 261 
 
 which we thus make, as, e.g., in geometry. Now determining 
 means judging synthetically."^ 
 
 The argument of the Aesthetic, however, as has been already l^^^^^ the 
 suggested, abstracts throughout from an element which yet it ^"'J'^"'- 
 everywhere presupposes, viz., the activity of thought. For 
 synthesis or judgment obviously has two sides : it is an 
 addition of one thing to another, and it is the com- 
 bination of one thing with another: it is differentiation and 
 it is integration. Now, if we regard sense as a simple recep- 
 tivity, its sensations (or perceptions, if we could call them 
 perceptions,) simply resolve themselves into a series of states, 
 each without any connexion with the others ; and out of this 
 flux of feelings or images no consciousness of any definite 
 object could ever arise. No element could be retained to 
 qualify the others, no identity could be supposed to be present 
 through the differences as the object to which they were all 
 referred. Further, what is but another aspect of the same 
 thing, such a consciousness, if we could call it a consciousness! 
 could never gather itself up into self-consciousness. It would I 
 have no permanent unity in reference to which all that 
 presented itself successively in it was determined, and, there- •* 
 fore, it could -never become conscious of such a unity. Its life 
 would be like -a stream without banks to confine it to any 
 definite course. It could not knoio anything through the 
 impressions of its sensibility, it would be blind to all that 
 was happening in itself. ' Data of sense ' might be ' given ' 
 to it, but it would have no ideal hand, no intellectual grasp, to 
 receive or hold what was given, but would instantly let it slip 
 in the very moment in which it came. 
 
 These considerations show us that, for knowledge as well as Necessity of 
 
 the unity of 
 
 for peKseption in the proper sense of the terms, two things are f^^y^g j^™/' 
 necessary-- which we have not yet considered. In the first priSo^ie^of 
 
 " ■ synthesis. 
 
 place, it IS necessary that there should be a unity which is 
 permanent and into relation with which all the data of sense 
 » 1 R. I. 498 ; H. VIII. 529. 
 
262 
 
 THE CRITIQUE OF PURE REASON. 
 
 BOOK I. 
 
 The analytic 
 unity of 
 thought can- 
 not supply 
 such 
 principles ; 
 
 'are successively brought, so that through it they all come into 
 relation with each other ; and in the second place, it is neces- 
 
 I sary that this unity should be active in combining these data 
 in relation to itself. But, further, in order that this unity may 
 combine these data in relation to itself, it. must differentiate 
 itself in relation to them, or, in other words, it must supply 
 from itself forms of combination for that which is given as a 
 successive manifold. For the mind is not like a casket into 
 which things may be thrown and in which they may be 
 passivelj'' held together. The unity, which the manifold has 
 
 for it, must be bestowed on the manifold by its own activity: 
 and the same unity must be the means whereby that which 
 passes away as a sense- affection is retained as an object or an 
 element in an objective consciousness, so as to qualify, and 
 be qualified by, other elements given through other sense- 
 affections. 
 
 Now, the pure activity of thought, according to Kant, is 
 merely analytic ; or if we can call it synthetic, its synthesis is 
 based on analysis. If it brings many individuals to a unity 
 in one species, many species to a unity in one genius, it is 
 merely by aid of an abstraction, which extracts the common 
 element from those different species and individuals. It pre- 
 supposes a content as present to our consciousness, and merely 
 brings it to a higher unity for our consciousness by an external 
 process of generalisation. We say, " an external process " 
 because, though in this way an effort is made to satisfy 
 the mind's impulse to seek its own unity in its object, yet the 
 whole process does nothing to bring the content of the object " 
 any nearer to the mind than it was at first. We assume at 
 the outset of the process that the object has already been 
 brought into a relation to the mind, and made into one of its 
 ideas; but we do not ask how it has been brought into this 
 relation, nor do we ask how the relation may be improved or 
 the difference be dissolved in a more perfect unity. In the whole 
 process we have, therefore, only a subjective movement of 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 263 
 
 thoiight in relation to a matter which is assumed to be already 
 appropriated by it, without any explanation of the original 
 process of appropriation. We learn nothing of the way in 
 which that, which is not already determined as an element 
 in the mind's consciousness of itself, is to be brought into the 
 mind ; nor do we get any explanation of the method in which 
 the process of appropriation thus initiated may be carried still 
 farther. It is simply that the mind holds the content along 
 with its own identity, and, as it were, repeats the assertion of 
 its own identity upon it. Yet such a formal process of syn- 
 thesis by means of analysis is in contradiction with itself, 
 in so far as it is still a process of synthesis, a process in which 
 an attempt is made to reach a higher and ultimately the 
 highest universal. For the only meaning of this process is 
 that thought seeks to overcome the difierence between itself 
 and its content, or to find a point where that difference dis- 
 appears. The impulse of thought which makes it generalise is. | 
 thus its impulse to seek itself, i.e., to seek the transparent unity ■ 
 of self-consciousness, in the object. But in so far as it is 
 merely formal or analytic thought, this effort must be in vain : 
 since the highest universal stands opposed to it in an antagon- 
 ism which is as immediate and as irreconcileable, as was the 
 antagonism to it of the lowest particular with which it started. 
 
 It appears, then, that formal Logic, when it describes thought though it 
 
 as analytic, is describing it as confined to a process which is "^"^ *° *''™' 
 
 at once meaningless and self-contradictory, a process which is 
 
 the negation of all process. This idea will have to be further 
 
 developed when we come specially to deal with the Analytic. 
 
 Here, it is enough to point out that Kant uses the logical 
 
 account of the analytic movement of thought in judgment and 
 
 syllogism, as a guide to the discovery of the nature of the 
 
 synthetic process by which thought brings a foreign matter 
 
 into relation to itself, or by which it goes beyond itself to 
 
 apprehend a foreign matter. And he finds it possible so to 
 
 use it, simply because in the formal Logic of the schools there 
 
264 THE CEITIQTJE OP PUEE REASON. BOOK I. 
 
 was still a faint shadow of the real process of thought which he 
 was seeking to determine. As we have already seen, formal 
 Logic contradicts itself, in so far as it still represents thought 
 as a process of seeking for unity and for the highest unity, and 
 thus indicates in spite of itself the real nature of that synthe- 
 tic process by which thought unites itself with its object, or 
 finds itself in that object. Kant, therefore, could certainly find 
 in it a kind of sign-post to direct him to that truth of which it 
 was, we might almost say, the caricature. 
 Relation of In this point of view, then, formal Logic suggests the idea of 
 
 standing and *^° kiuds of Synth csis Corresponding to judgment and syllogism 
 o the reason, j-gspcctively. In the first kind of synthesis, a relation is 
 established between the analytic unity of pure thought or self- 
 consciousness and the manifold of perception ; or the manifold 
 of perception is subjected to a principle of unity derived from 
 the activity of thought. In the second kind of synthesis, the 
 relation so established, by reason of the imperfect corre- 
 spondence of its synthesis to the unity of thought, has again to 
 be subjected to another and higher form of synthesis, which 
 aims at altogether harmonising the first synthesis with that 
 unity. The former synthesis gives rise to experience, and the 
 latter to the ideal of reason and to the consciousness of the 
 limitation or phenomenal character of the objects of experience 
 as such. In other words, the former synthesis determines the 
 data of sense so that they present themselves as a system of 
 objects in one experience; and the latter, though failing to give 
 a further determination to these objects, yet even by its failure, 
 discloses that those objects and their system do not correspond 
 to the mind's highest idea of reality. 
 The deduction The possibility of this vindication and yet limitation of 
 
 oJthecate- r ^ J 
 
 th^ umL experience can only be understood if we recall the results of the 
 
 experience. Aesthctic, with such modifications as are necessary when we 
 
 regard them from this new point of view. These results were 
 
 that time and space are a priori forms of perception, under 
 
 .which all objects of perception as such must present themselves. 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 265 
 
 But now it appears that no objects can present themselves as 
 such to the conscious self except through its own synthetic' 
 activity, and that apart from that activity, what can be 
 presented is only a ' manifold ' without unity, which therefore 
 is not characterised even as a manifold. Not only the matter 
 of sense but even space and time, the pure forms of sense, 
 must wait for an activity of thought to gather them into 
 the unity of objects of perception. We cannot, therefore, say 
 that time and space are principles for the arrangement of the 
 matter of perception in the sense that their unity of itself 
 determines objects of perception in their relations to each 
 other ; but merely that they are forms in which it is possible 
 for the data of sense to he arranged in a certain order : for the 
 actual arranging of them in that order implies an activity of 
 thought. At the same time, in spite of this change, the 
 relation esbablished in the Aesthetic between these forms and 
 the matter that falls under them remains unaltered; in so 
 far as they are still a priori determinations of that matter, 
 which it necessarily receives in becoming the object of our 
 perception.. But instead of saying that the matter of sense as 
 it becomes object of perception takes on these forms, we must 
 now say that the matter of sense, when determined as object 
 for the conscious subject by the synthesis of the understanding, 
 is so determined through the forms of sense. Time and space 
 are, therefore, presupposed as unities in which the objects of ex- 
 perience are placed, in the very act by which they are determined 
 as objects of experience. In other words, the act of thought, by 
 which the data of sense are determined as objects of experience, 
 makes presupposition of another act by which time and space 
 are envisaged as unities in which all the objects of perception 
 are arranged.^ These two acts, however, are not separated from 
 each other: for we do not first perceive phenomena in space 
 and time, and then determine them as objects, but, in per- 
 
 1 This point cannot be fully explained till we have considered the part in 
 kft)wledge which Kant assigned to the imagination. 
 
266 THE CRITIQUE OF PDRE REASON. BOOK i. 
 
 ceiving and determining things as objects, we perceive and 
 determine them as in space and time. 
 Union of a j^q^ jf ^-j^^g jjg ^jig case, wc may understand how it is that 
 
 pnon percep- 3 j 
 
 coTepSonsin « priori synthstic judgments, i.e., a priori judgments in relation 
 experience. ^^ objects, are possible. They are possible in virtue of the fact 
 that there are both a priori conceptions and a priori percep- 
 tions, or at least a priori forms of perception. For so far as 
 this is the case, we have the necessary constituents for a priori 
 synthetic judgments, which yet will apply to all the objects of 
 experience. Mere a priori conceptions would not entitle us to 
 make such judgments : for they would express merely the general 
 combining activity of thought, by which it seeks to bring all 
 matter presented to it into relation to its own unity ; and such 
 combining activity is simply meaningless without anything to 
 combine. Even if we admitted that it could give rise to judg- 
 ments, they would not be judgments about anything. Nor 
 again would mere a priori forms of perception entitle us to 
 make such judgments : for, as forms of perception, they do not 
 entitle us to make any judgments whatever, either by them- 
 selves or in combination with a matter. But the a priori 
 forms of conception, when directed upon the matter of percep- 
 tion through its forms, will give rise to certain judgments which 
 are at once a priori and of objective validity, because they 
 express the conditions under which alone anything can be 
 determined as an object. 
 o?f plfie"''^ On the other hand, we cannot understand this use of the 
 thought. a priori conceptions without seeing that it carries with it a 
 limitation : for since they are mere forms of synthesis, they get 
 application and so acquire a real significance only in relation 
 to a given matter — -in the present case in relation to the 
 matter given under the forms of space and time. Their value, 
 therefore, as determinations for objects of experience so given, 
 affords no presumption in favour of the use of these or any 
 other kind of mental synthesis, where there is no such matter 
 given. If, therefore, the mind is not satisfied with the unity 
 
CHAP. 1. THE PROBLEM OF THE CRITIQUE. 267 
 
 of judgment, but goes on to seek the higher unity of reasoning, 
 it would need, in order to give objective value to the thought 
 thus conceived, another kind of intuition or perception, and in 
 the absence of such perception it cannot attain to any a priori 
 synthesis. The thought of such a unity may, indeed, be of use 
 to ns as setting up an ideal for the continued exercise of the 
 -understanding, and stimulating the mind to seek by pro- 
 syllogisms ' to complete its work and bring back the manifold 
 objects and events of experience to its own unity ; but such an 
 ideal is by the very nature of the case condemned to remain an 
 ideal, a thought which cannot be realised in knowledge. 
 
 We are now prepared to understand the importance attached in synthetic 
 
 ^ i: J- ■ judgment the 
 
 by Kant to the distinction of analytic and synthetic judgments; yondits°eUand 
 and we are prepared also to see how a synthetic judgment is ready knows. ' 
 conceived as an act of the mind in which it determines 
 objects, while an analytic judgment is conceived as an 
 act of the mind in which it abides in itself and merely 
 determines the relations of its own ideas. An analytic 
 judgment, in- short, deals with what is already determined as 
 an idea of the mind and so already united with the ' I think ' 
 of self-consciousness. But a synthetic judgment expresses the 
 process by which the mind, so to speak, makes a matter its own 
 or goes out of itself to apprehend a matter outside of itself, and 
 to unite that matter ^ for the first time with the ' I think ' ; or 
 it is an act by which new matter is combined with that which 
 has already been united with the ' I think.' Hence it is 
 synthetic in two ways : as it unites a certain matter of percep- 
 tion to self-consciousness, and as it unites a perceived matter 
 which has not yet been thought to a perceived matter which 
 already has been thought. As, then, thought can determine the 
 data of perception as objects only by combining them with 
 each other, and as, on the other hand, that it can determine no 
 object as such, except by combining these data, it is clear 
 that a synthetic act and the determination of the object as 
 . % ^ Of course matter here includes the forms of sense. 
 
268 THE CRITIQUE OF PURE REASON. book I. 
 
 such are, from Kant's point of view, one and the same 
 thing. 
 Is there such a Objection has been taken both to the distinction of analytic 
 
 thing as pure 
 
 ^ud'^'ent? ^^^ synthetic judgments in itself and to the examples by 
 which Kant illustrates it. The former objection has already 
 been shown, and will be further shown, to be just. It amounts 
 to this, that mere analysis is no process or act of thought at 
 all, and only appears to be one because the idea which underlies 
 it of a movement by bare identity, is not thoroughly carried 
 out. If this be true, an analytic judgment would really be no 
 judgment but only a meaningless tautology ; and it is another 
 aspect of the same statement to say that a purely subjective 
 movement of thought which did not determine an object would 
 not be a real movement of thought, but only a meaningless 
 resting of thought in its own self-identity. Such a judgment 
 would never be made, except as a logical experiment to illus- 
 trate an imaginary law of thought. If ever, in a sense, 
 analytical judgments are made, they are intended to communi- 
 cate to others a thought or synthesis which for us has been 
 already accomplished; or, in other words, to carry them through 
 a synthesis to a conception which we have already attained 
 and which we for their behoof analyse. From this it would 
 follow that a judgment which is analytic for one is synthetic 
 for another, and that we can tell which it is, only by consider- 
 ing the state of knowledge attained by the individual who 
 makes the judgment. Thus when Kant says that the judgment 
 ' Body is extended ' is analytic, but that the judgment ' Body 
 is heavy ' is synthetic, it might be answered that it is by a 
 synthesis that the idea of extended substance is formed, and 
 that it is only a further step in the same process that connects 
 the predicate of weight with that of extension. The sole ques- 
 tion is, therefore, whether the term ' body ' is taken by us as 
 representing the result of a synthesis which has reached the 
 second stage, or at most whether it is so taken in general 
 usage, — a question to which no definite answer could be given, 
 
CHAP. T. THE PROBLEM OP THE CRITIQUE. 269 
 
 for general usage is fluctuating and uncertain. If we overlook 
 this merely subjective distinction, all judgments are synthetic 
 in the making and analytic when made ; i.e., judgment is the 
 process whereby thought advances to a new conception of its 
 objects. The judgment is analytic in so far as it expresses an 
 identity, ,but the act of judging develops an identity to a (new) 
 difference, which it at once expresses and overcomes.^ 
 
 Now, although this is true, yet there is still an important importance of 
 
 the distinct! on 
 
 purpose served by the Kantian distinction, which becomes ^°'^ ^"*- 
 manifest when we have regard to the way in which Kant puts 
 the question of the possibility of synthesis. What, he asks, is 
 the tertium quid which enables us to get beyond a conception 
 we have so as to add new eleinents to it ? In the case of a 
 
 ^ There is a, synthesis in the process by which the mind goes beyond its pure 
 identity to any object (and, indeed, as has been indicated, there is a double 
 synthesis, since it is only as we combine the elements of the manifold 
 under a conception that we can combine the consciousness of the object 
 so attained with the 'I think'). Hence if we go back to the beginning, 
 there is, on Kant's view, nothing to analyse. If any analytic judgment 
 is possible at this stage, it is only the ' I am I ' of self-consciousness, 
 and that, as we shall find, Kant admits to be impossible without a synthesis. 
 Note, however, that the two ideas of bringing any manifold to a unity 
 under a conception and uniting it as an object with the consciousness 
 of the self, (or, what is the same thing, going beyond the self to it,) are 
 contrasted with the bare self-assertion of consciousness in its transparent 
 unity with self into which no difference has been introduced, and which 
 does not go beyond itself to anything other than itself. "When Hegel after- 
 wards spoke of the material or extended as that which is external or other 
 than itself, because external to, or other than, the self, he seemed to be making 
 a play upon words. But that idea springs very naturally out of the opposi- 
 tion of thought and perception as exhibited by Kant. I may observe that, if 
 we look at it from the point of view of thought, thought in taking into itself 
 the very form of sense s^nd so schematising its categories, is already synthetic. 
 Hence the determination of body as extended already implies synthesis. For 
 the pure category of substance, as Kant subsequently declares, has in itself no 
 reference to the form of space. On the other hand, if we look at it from the 
 point of view of sense, no manifold can be united into an image of perception 
 and brought under a conception except by a synthetic act, an act that carries 
 us beyond the data of sense. We shall afterwards have to consider more 
 closely the nature of this double synthesis. What is said above, is well ex- 
 pressed in a passage from Reuss (quoted by Vaihinger, I. 261) — "The act by 
 which the connexion of two ideas is established is a synthetic judgment : the 
 act by which the already established connexion of two ideas is expressed is an 
 analytic judgment.'' 
 
270 THE CRITIQUE OF PURE REASON. book I. 
 
 conception which is abstracted from experience, we can answer 
 that the tertium quid is to be found in new experience of the 
 same object. But in the case of a priori synthesis this answer 
 does not avail, and thought, therefore, it would seem, cannot 
 reach beyond itself. The answer Kant gives is that, though in 
 tjiis case there is no empirical perception of an object corre- 
 sponding to the conception, by which the conception might be 
 enriched, there is an a priori form of perception which may be 
 determined in accordance with the conception, and which, when 
 it is so determined, enables us to add to it. Thus, in the whole 
 science of Geometry, we are simply adding to our conceptions 
 by means of the a priori perception of space to which we apply 
 them. This is the only way in which we can deal with the 
 conception of a particular figure so as to discover anything 
 more about it than we already know from the conception itself. 
 " Give a philosopher the conception of a triangle, and let him 
 try in his own way to iind out how the sum of its angles is 
 related to a right angle. What he already has is nothing but 
 the conception of a figure which is enclosed by three straight 
 lines, together with the conception of its having three angles. 
 Let him reflect on this conception as long as he pleases, he wUl 
 get nothing more out of it. He may analyse and make quite 
 distinct the conception of a straight line, of an angle, or of the 
 number three ; but he will come upon no other properties than 
 those which already lie in these conceptions. But let the 
 geometer take the question in hand. He begins immediately 
 to construct a triangle. Knowing that two right angles are 
 equal to the sum of all the adjacent angles which can be made 
 by lines falling upon one point in a straight line upon one side 
 of it, he lengthens one side of the triangle and thus gets two 
 adjacent angles which together are equal to two right angles. 
 Then he divides the external angle of these two by drawing a 
 line parallel to the opposite side of the triangle, and perceives 
 that he has got an external adjacent angle which is equal to 
 one of the internal angles, — and so on through the rest of the 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 271 
 
 demonstration. In this way he arrives, by a chain of inferences 
 in which he is always guided by perception, at the perfectly 
 self-evidencing and at the same time universal solution of the 
 problem." ^ 
 
 In this reasoning it is assumed that we have alreadv the "n^e ultimate 
 
 ^ '' synthesis of 
 
 conception of a triangle and that it is a possible conception, pe?c^tion.'^ 
 i.e., a conception which can have an object. But this result 
 also must have been due to a previous synthesis in which we 
 constructed that conception in space ; for, as a mere matter of con- 
 ception, we can put together in thought any marks we please 
 so long as they do not contradict each other ;, but it is only by 
 actually constructing it in space that we know that one figure 
 is really possible and another not. The ultimate synthesis, 
 therefore, on which we fall back is the synthesis by which 
 space is determined in accordance with the a priori conception, 
 of quantity, or in which an object in space is represented as 
 an extensive quantum formed by adding spatial units to each 
 other. But in this, as in all the subsequent determinations of 
 space, we have the same process with its two sides, conception 
 and perception ; the former involving a law or principle accord- 
 ing to which the elements are to be combined, the latter sup- 
 plying a matter which, being combined in accordan6e with 
 this principle, gives rise to new conceptions or adds new 
 elements to conceptions formerly possessed by us. 
 
 When we have got thus far, we begin to see that what Difference of 
 
 mathematical 
 
 Kant is maintaining really amounts to this, that all progress in ^^of"™'"^'' 
 knowledge, or in other words, all synthesis, involves two com- 
 plementary processes of differentiation and of integration which 
 cannot be severed from each other ; but that there are certain 
 cases in which the mind can produce the difference as well as 
 the " functions of unity " by which it is integrated, while in 
 other cases it can only take up the elements which are given 
 to it through affections of sense. At the same time, as we 
 shall see, these two cases are only ideally separable by an act 
 • 1 A. 716 ; B. 744. 
 
272 THE CRITIQUE OF PURE KEASOIf. BOOK [. 
 
 of abstraction ; for the pure perceptions, which furnish the 
 matter for a priori synthesis, are after all only the forms 
 of other perceptions, and, on the other hand, the a posteriori 
 perceptions are integrated only through the forms of percep- 
 tion. Hence integration of the former kind has a side on 
 which it is dependent on the affection of sense, and integration 
 of the latter kind has a side on which it is a pure construction. 
 From this arises the distinction and the relation of the mathe- 
 matical and dynamical synthesis, or, to put it otherwise, of the 
 synthesis of pure mathematics and that of pure physics. 
 Perception That the mental process has these two aspects or sides, that 
 
 and concep- 
 tion as it is at once difierentiation and integration, it is not difficult 
 
 elements, not ° 
 
 knowie°dge. *° show, and Kant is nowhere more happy than in showing it. 
 Thus, when he argues for the impossibility of synthesis by mere 
 conceptions, what he is maintaining is that a mere ' function 
 of unity,' a mere integrating principle, cannot add to know- 
 ledge, but can at most only enable us to realise more clearly 
 the unity of the elements which for thought are already 
 united ; and when he argues for the impossibility of synthesis 
 by mere perceptions, what he is maintaining is the impossi- 
 bility that a mere series of feelings can construct itself into a 
 definite object of thought. The difficulty, however, which 
 embarrasses us most in following out Kant's reasoning, arises 
 from his preserving in relation to those ultimate functions of 
 differentiation and integration that independence, as of two 
 mental products externally fitted to each other, which is 
 intelligible enough when we speak of conception and perception 
 in the ordinary senge. Thus, to take an instance of his own, 
 we can easily understand that an object may be perceived in 
 the distance which on nearer approach we recognise as a man : 
 the conception and the perception, as it were, meeting together 
 and the latter being subsumed under the former. But when we 
 carry our analysis back to the primitive process by which the 
 mind puts its own unity into the ' manifold ' of feeling, which 
 it thus ' recognises ' as an object, we are troubled by the 
 
/ 
 
 CHAP. I. THE PROBLEM OF THE CRITIQUE. 273 
 
 difficulty that each part of the process involves the other parts. 
 The ' manifold ' must be there to be brought together as an 
 object by the mental unity ; yet it is only through the mental 
 unity, in relation to which it is successively taken up 
 and brought together, that it can be said to be there at all. 
 Kant seems to suppose, on the one side, a pure unity of 
 thought which in itself is merely analytic, and, on the other 
 side, a mere dispersed manifold without connection ; and then 
 he supposes that it is by reference to the manifold presented 
 in sense that the unity becomes synthetic, or supplies those con- 
 ceptions which may be used in synthesis ; while, on the other 
 hand, he also supposes that it is by reference to the unity that 
 the manifold becomes associated, so as to supply an image of 
 perception which can be brought under those conceptions. 
 Yet he himself, as we shall see, maintains that the analytic 
 unity of thought in us presupposes a synthesis of the manifold 
 of perceptions, i.e., that the unity of self-consciousness presup- 
 poses the unity of our objective consciousness ; and, on the 
 other hand, he is equally soKcitous to show that a mere 
 manifold without reference to the unity of self-consciousness is ■ 
 a mere chaos or Gevmhl, which " for us as thinking beings is 
 as good as nothing." Hence, if we hold Kant to the distinction 
 which he makes between perception and conception, it seems 
 impossible to relate them ; while if we hold him to the relation, 
 we must greatly modify what he says about their distinction. 
 Yet he was as earnest in upholding the distinction against 
 Leibniz as in upholding the relation against Hume, and to 
 leave out either would be to surrender one of the hinges upon 
 which his philosophy turns. 
 
 The full explanation of this collision of opposite tendencies How it was 
 
 . necessary for 
 
 or thought in Kant cannot yet be given. But when we J*^*"* '° ""'"- 
 consider the historical development of his philosophy, we fnaV^Saof 
 can see how it seemed necessary to Kant to keep firm MnceltioSf''"* 
 hold both of the distinction and the relation. To con- 
 fuse perception with conception was to fall back to the 
 VOL. I. s 
 
274 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 philosophical point of view of Leibniz or rather of Wolff, in 
 , which time and space were treated as " obscure conceptions " 
 and analysis was substituted for synthesis. It was thus to 
 build up a dogmatic philosophy of mere conceptions without 
 any restraint from experience. On the other hand, to deny 
 the necessary relation of perception to conception in know- 
 ledge was to fall into a pure empiricism, which extruded all 
 generality from experience, and which thus was fatal even to 
 empirical knowledge itself. Hence, the only theory which 
 could avoid the dangers seemed to him to be one which recog- 
 nised both conception and perception as elements necessary to 
 experience, but regarded them still as separate and only 
 externally combined with each other. The possibility that 
 they should be correlative elements which were neither to be 
 separated nor confused did not occur to him ; because, on the 
 one hand, he believed that there was an analytic movement of 
 thought in itself apart from perception, and because, on the 
 other hand, he conceived that sensation as such is not only 
 presupposed in our knowledge but actually contributes an 
 element to it — though he recognised that of that element 
 nothing can be said which does not involve its relation to the 
 ' I think ' of consciousness, and that all the terms we use, such 
 as the " manifold " or the " series of sensations," really involve 
 that reference. The value of the Kantian theory, however, is 
 just that it maintains a protest against a merely formal 
 philosophy on the one side, and a mere empiricism on the 
 other ; and that, while it shows the necessity of perception to 
 the development of conception, or, in other words, the necessity 
 of experience, it at the same time maintains that that percep- 
 tion is essentially related to the thought from which it is 
 distinguished. 
 Three points We sce, then, that the essential points of Kant's doctrine are 
 
 in Kaut's 
 
 knowMgl *ese:— >s<, that knowledge or synthetic judgment always 
 involves perception as well as conception ; secondly, that in the 
 case of mathematical synthesis both are a priori ; and thirihii 
 
CHAP. I. THE PROBLEM OF THE CRITIQUE. 275 
 
 that in the case of dynamical synthesis, though an a posteriori 
 element is involved, yet there is a determination of that 
 element through the forms of sense which gives universality 
 and therefore objectivity to that which for perception is merely 
 particular and subjective. 
 
 The first of these pbints has already been made sufficiently (i) synthesis 
 
 '■■ " " involves both 
 
 plain. It means simply that knowledge or judgment is ?o™|t^n™'' 
 always an apprehension of the many in the one, of the one in 
 the many, and that neither can be separated from the other 
 except by abstraction. There must be distinction yet unity in 
 the process of knowledge, a distinction which in Kant's view 
 is ultimately referable to the division of the perception of a 
 manifold in space and time from the thought which gathers 
 it into unity. 
 
 The second point depends on the fact that the a priori forms (2) Both are o 
 
 priori in 
 
 of space and time can by themselves be made objects of ™^^®™ig*:'°''^ 
 
 thought ; in other words, that objects may be regarded as they 
 
 are determined only by relations of space and time and by no 
 
 other relations. Hence, taking the conception of quantity as 
 
 our guide, we can make constructions in space and time and 
 
 the results we arrive at are principles universally valid for all 
 
 that is presented as in space and time.-^ 
 
 The case is different with the dynamical principles, such as (^) *>"* o"* ™ 
 
 '' 1. r 3 dynamical 
 
 substance and causality. For there we have to do, not with 'y°*''°^'^- 
 the mere determinations of spaces and times, which can be 
 reached by a construction in pure space and time, and which, 
 when so reached, can and must be transferred to all objects 
 occupying those spaces and times, but with determinations of 
 phenomena or perceptions, by which they are referred to 
 permanent objects existing in space and time and, as states of 
 such objects, have their relations of co-existence and succession 
 
 iTo avoid complexity of statement I do not here refer to the principle of The 
 Anticipations of EmpiricaZ Perception which have to do with the synthesis 
 whereby the matter of sensation must be united with the consciousness of 
 time, the form of inner sense, in order to make possible the qualitative deter- 
 min^ions of objects. 
 
276 . THE CEITIQTJE OF PURE REASON. book I. 
 
 fixed according to universal laws. In other words, these 
 principles refer to objects in so far as they are distinct 
 from the times and spaces they occupy, and yet are deter- 
 mined as occupying them and as in them standing in certain 
 permanent relations to each other. Hence, the dynamical 
 principles not only presuppose an empirical matter of per- 
 ception, but fix how that matter must be determined, in itself 
 and in relation to all other matters of perception, in being re- 
 ferred to objects known as objectively existent in space and time. 
 Thus, while the mathematical principles presuppose objects deter- 
 mined as in space and time, and authorise us to transfer to them 
 the quantitative determinations of pure space and time, i.e., 
 of the forms under which alone they can be represented, the 
 dynamical principles show us how the data of sense must be 
 determined in order to become referred to such objects. In other 
 words, the mathematical principles tell us how objects must be 
 conceived, if they are conceived as existing in space and time, 
 while the dynamical principles tell us how the matter of sense 
 must be determined, in order that it may be conceived as 
 referring to objects in space and^ time. 
 difflrenoTL In *^liis way the question how a priori synthesis is possible 
 of the two breaks up into two questions, how mathematical and how 
 
 forms of 
 
 synthesis. physical, « priori synthesis is possible. The answer to the 
 first of these questions is an explanation of mathematical 
 synthesis as a pure process by which experience is anticipated, 
 and a vindication of it on the ground that, when we have 
 separated the a priori forms of time and space from the 
 empirical objects which they condition, we have command over 
 those forms and can determine them at pleasure by a process 
 which does not need to be verified by experience, though 
 its results hold good for all experience. In the other case, 
 it is impossible to vindicate any such anticipation of ex- 
 perience in the particular qualitative determinations of it or 
 their particular coexistences and sequences; for we cannot say 
 prior to experience that we shall have particular perceptions of 
 
CHAP. T. THE PROBLEM OP THE CRITIQUE. 277 
 
 sight, taste, etc., or that we shall have them in any particular 
 order, as we can work out the particular properties of figures in 
 Geometry. We can, however, lay down certain general con- 
 ditions under which alone such particular perceptions and their 
 particular relations in time and space can be referred to objects. 
 We can say that if, and so far as, these qualitatively and 
 quantitatively determined perceptions are to be referred to 
 objects in space and time, they must fall under certain general 
 principles which fix their relations to each other. And in this 
 way we can lay down the general principle that all objects of 
 experience must be conceived as permanent substances, the 
 states or accidents of which follow upon each other according 
 to universal laws of causation, and coexist with each other 
 according to universal laws of reciprocity. 
 
 Now, the value of this distinction cannot yet be fully deter- Ground of tus 
 
 •' ■' distinction. 
 
 mined. But it may be seen at once that it rests on the idea 
 that our a priori determination of objects is limited to the 
 determination by the categories, first, of the pure forms of 
 perception, and secondly, of the matter of perception through 
 these forms. It is bound up with the idea of the absolute 
 a-posteriority of the matter of perception both as to its presence 
 or existence for us in perception, and as to its quality, i.e., 
 its non-quantitative character, (for its quantitative character is 
 given a priori through the necessity of its being presented 
 under the forms of perception, which again can be specified or 
 determined only in accordance with the category of quantity). 
 According to this view it is, so far as our understanding is 
 concerned, an accident that there is any matter of perception 
 given, which we can determine as objective or as referring to 
 objects ; and it is another accident that that matter should have 
 certain particular qualities rather than others. When, there- 
 fore, we go beyond the quantitative determinations which 
 belong to space and time, and proceed to determine anything 
 as existing in space and time, and as in space and time 
 stantiing in certain relations of coexistence or succession, we 
 
278 
 
 . THE CRITIQUE OF PURE REASON. 
 
 BOOK I.! 
 
 Belatious of 
 Aesthetic^ 
 Analytic, and 
 Dialectic. 
 
 Summary of 
 Kant's views 
 as to the 
 problem of 
 criticism. 
 
 are able only to anticipate those general principles whieh are 
 involved in all determination of objects as so existing and 
 related. Thus, e.ff., we cannot say when a special sequence of 
 phenomena will take place, or what phenomena will stand in 
 this relation in time ; but we can say that, if the sequence does 
 take place once, it will take place always, whenever all the 
 circumstances are similar ; for otherwise it could not be 
 referred to any object in an experience of which time is the 
 condition. 
 
 We have now seen how the problem of the Critique leads to 
 its main divisions. The Aesthetic and the Analytic deal respec- 
 tively with the two elements of knowledge, perception and 
 conception, which are at once distinguished and related; while 
 the Dialectic deals with that discordance between the ideal and 
 the reality of knowledge, which arises from the fact, that the 
 relation does not overcome, or is not coincident with, the 
 distinction, and that knowledge, therefore, means a combina- 
 tion of elements extraneous to each other. It is, therefore, a 
 combination which does not correspond with the unity of 
 thought or self-consciousness, which yet it presupposes as the 
 condition of its possibility. This dualism appears objectively 
 as the separation of phenomena from things in themselves, 
 which, therefore, have a problematic existence so far as the 
 Critique of Pure Reason is concerned. Meanwhile the ideas of 
 these problematic objects have a function in relation to that 
 very experience which they limit and for which they set up an 
 unattainable ideal. 
 
 If we state the general problem of philosophy in the form 
 in which Kant finally stated it, as the problem of " the possibility 
 of advancing by reason from the knowledge of the sensible to 
 that of the supersensible,"^ the answer of the Critical Philosophy 
 may be shortly summarised thus. If knowledge of the super- 
 sensible is possible, it must be rational or a priori knowledge,; 
 for only by an a priori process can we hope to deal with that 
 1 R. I. 488 ; H. VIII. 520. 
 
CHAP. I. THE PROBLEM Or THE CRITIQUE. 279 
 
 which is beyohd all sense^ But the insecuriby and contra- 
 dictions of metaphysic, the " ruinous disunion of philosophical 
 systems," forbids us to ask how it is possible, till we have first 
 settled whether it is possible. We turn, therefore, to the 
 question how a ^priori knowledge is possible, where it is admitted 
 to be actual, i.e., in relation to the objects of experience. On 
 the other hand, if by a recoil of the scepticism, which arises in 
 the region of metaphysics, upon the field of experience, it be 
 denied that the a priori principles are valid even in relation to 
 the latter, we are prepared- to show that experience is impossible 
 without them, and so to take away the ground from under the 
 feet of the sceptic. At the same time we have to explain that, 
 as pure thought in itself is ruled by the law of identity, its 
 going beyond itself at all must be mediated by perception, and 
 its going beyond itself a priori must be mediated by a priori 
 perceptions. Hence, where there is no such perception, there 
 can be no such synthesis,' and the character of the synthesis 
 will, be determined and limited by the character of the percep- 
 tions. But our a priori perceptions are essentially forms of 
 sense, i.e., they are forms of a matter which is essentially 
 a posteriori, and therefore external and alien to the pure 
 intelligence that apprehends it. Hence, neither they nor the 
 matter that falls under them can be brought into perfect unity 
 with the mind that knows them. The mind is never able to 
 consummate the synthesis of its object with itself, and the 
 forms of unity by which it determines sensible objects still 
 leave these objects inadequately determined, according to that 
 idea of knowledge which it carries with itself. Hence 
 it is led to make the distinction of the noumena it can 
 think from the phenomena it can know. But as the former 
 are presented to it in no perception or intuition, it is obliged 
 to recognise that it is incapable, so far at least as theoretical 
 reason is concerned, of rising beyond the problematical existence 
 of the noumena or of turning the thought of them into 
 knowledge. 
 
280 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 mt^mrte^f" Kant's proof that a 'priori knowledge is real, because without 
 meSt contra- it experience would not be possible, is thus one side of his 
 other? argument ; the other side is that the conditions under which 
 
 such knowledge is realised are such as are not found except in 
 sensuous experience. And the great question that arises out of 
 the whole process is, whether the effect of it is not to break 
 down that opposition of analysis and synthesis upon which the 
 opposition of phenomena and noumena ultimately rests ; or - 
 rather to show that that opposition is a relative, and not an 
 absolute, one. Or, to put all in a word, does not Kant, in - 
 showing the limitation of the empirical or phenomenal view of 
 the world, himself point out how that view is to be supple- 
 mented and corrected, so as tO' exhibit to us the world as it is 
 in its reality ? Is not the defect of the Critique what was 
 indicated in a previous chapter, that it conceals under the 
 appearance of a process of abstraction, by which we rise to a 
 mere analytic unity, a process of ' concretion,' which really is 
 an advance, through a relative unity of unreconciled elements, 
 to a unity of differences which is organic, and therefore, in a 
 true sense a priori and transparent ? ^' If so, we may repeat in 
 a serious sense what has been proposed as a kind of- reductio 
 ad dbsurdi(,m, that beyond the problem ^f the possibility of 
 a priori knowledge - is the problem of the possibility of the 
 Critique of Pure Reason^ For so long as there is an unexpressed 
 thought, which is latent beneath the expressed thought, and 
 which secretly governs it, so long we have a right to criticise 
 the expressed thought by bringing into prominence its implicit 
 presuppositions. 
 
281 
 
 CHAPTER II. 
 
 THE AESTHETIC. 
 
 TlsT a previous chapter, we have seen that the main ideas of ^iJrStothl 
 
 the Aesthetic, in gelation to time and space as the forms ^'^'*^'^- 
 of perception, were developed by Kant about ten or eleven 
 years before the publication of the Critique of Pure Reason, and 
 about two years before he even entered upon the investigation, 
 the results of which are given in the Analytic. It is, there- 
 fore, remarkable that after so long a period, the Aesthetic 
 repeats the corresponding sections of- the Dissertation with only 
 subordinate modifications. From this it might naturally be 
 inferred that Kant built his new edifice on a foundation which 
 he considered to be immovably established, and which was 
 left untouched by all the subsequent developments of his 
 speculation. This inference, however, would be erroneous. 
 The Aesthetic, it is true, remains almost unchanged in form : 
 but this is not because it was still regarded by Kant as in 
 itself satisfactory, but because in the first part of his book he 
 wished to treat one of the aspects or elements of the truth, 
 without being hampered by the necessity of considering its 
 relation to other elements. Thus, he leaves out of account all 
 consideration of the activity of thought, which is necessary ■ to 
 make the manifold of sense an object for the self-conscious 
 intelligence. Nay, in the corresponding section of the Pro- 
 legomena, he even goes the length of speaking of the science 
 
282 THE GEITIQUE OF PUBE REASON. book i. 
 
 of Mathematics, as if it were dependent only on the existence 
 of formal or pure a priori perceptions, and did not also imply 
 any activity of the understanding.^ Hence, also, he says little 
 or nothing in the Aesthetic of that process of a priori construc- 
 tion, that process of thought working through imagination, 
 which he afterwards shows to be necessary both to pure and 
 to empirical perception. As a consequence of this self-imposed 
 limit, he treats time and space and objects in time and space 
 - as if they were ' given ' to us directly and immediately through 
 sense, in such wise that we can make judgments regarding 
 them without any introduction of new elements by thought. 
 In short, he seems in the Aesthetic only to revive the view of 
 the Dissertation, according to which all that is necessary to 
 produce experience is that the understanding in its purely 
 formal use should generalise the ideas of perception.^ Sensi- 
 bility is represented as a receptivity, not merely of impressions 
 or sensations but of perceptions, as if perceptions could be 
 received without any activity of the consciousness that receives 
 them. And from these perceptions we are supposed to be able 
 to read off at once the characters of the individual objects 
 presented in them, provided we are careful not to make any 
 assertions which go beyond these individual objects. 
 Different This mcthod of proceeding upon assumptions afterwards 
 
 stages m x o x j. 
 
 recess.™''"''' modified was characteristic of Kant and it has undoubtedly 
 some advantages. It has especially the Socratic advantage of 
 meeting the ordinary consciousness on its own ground and 
 A leading it by gradual steps to refute itself, and so to discover 
 ; for itself a deeper basis of thought, the necessity of which it 
 
 might not otherwise have seen. It is a method which Kant 
 had learnt to use in his University teaching, and one which as 
 Jachmann tells us in a passage already quoted, often exposed 
 
 ^ In that section of the Prolegomena, he answers the question ' How pure 
 mathematics is possible, ' without reference to the necessity of any use of prin- 
 ciples of the understanding. 
 
 ^ He mentions at the outset tliat there are otter necessary elements (A. 21 ; 
 B. 35), but he does not allow this to affect his statements any further. 
 
CHAP. II. THE AESTHETIC. 28 a 
 
 Kant to serious misunderstanding from those who took his 
 first word for his last.^ It may be feared that the misunder- 
 standings to which the hearers of Kant's lectures were exposed, 
 have been fully paralleled among the readers of the Critique, who 
 have taken the statements of the Aesthetic as final, and have made 
 no allowance for the explicit as well as the tacit corrections sub- 
 , sequently introduced. At the same time, it has to be admitted 
 that Kant himself was not always careful to work out fully 
 that correction of his first presuppositions which was made 
 necessary by the new principles introduced in his subsequent 
 investigations. 
 
 The method of Kant has obvious advantages for any one who ^M^AetJc to*'"' 
 is prepared, not merely to receive passively the reasoning of the ""^ ^""'J"''- 
 Critique, but to rethink the whole process of reflexion by which 
 it was produced, and who, therefore, treats the Aesthetic only as 
 the first step in that transformation of ordinary conceptions, 
 which is to be continued throughout the whole work. What, 
 however, probably induced Kant to leave the Aesthetic in the 
 foriu which it has in the Critique was not merely his desire 
 to subject the reader to a process of Socratic dialectic and so 
 gradually to lead him on from his own ground to the point to 
 which it was desired to bring him. Nor was it simply the 
 historical fact, — though that may have had some influence, — 
 that Kant had thrown the principles of the Aesthetic into the 
 form which they take in the Critique, long before he saw the 
 necessity of the further step taken in the Analytic. (^The final 
 reason was that the Aesthetic is necessary to make intelligible -^ 
 the very statement of the problem of the Analytic^ For, as has 
 been shown in the previous chapters, what makes it necessary 
 to deduce, or justify, the objective use of the categories is just 
 that opposition between conception and perception, which it is 
 one main result of the Aesthetic to establish. Kant had 
 reached the question of the Analytic through a course of 
 investigation which was made necessary by the opposition of 
 ^ See above, p. 64. 
 
284 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 thought and sense established in the Dissertation, and he did 
 not see how to carry his readers to the same result by any 
 other road. 
 vhedistiiKium According to the principles of the Leibnizian philosophy 
 
 of sense and o ir ir -^ 
 
 ^Se"Sve3 from which Kant started, the distinction between a priori and 
 problem of a vosteriori coincides with the distinction between knowledge 
 
 their relation. 
 
 derived from conceptions and knowledge derived from percep- 
 tions ; and it is a distinction of degree and not of kind, 
 conception differing from perception merely in the distinctness 
 with which the elements in it are separated from each other.^ 
 Hence space and time were regarded as being merely confused 
 conceptions of the relations of objects, time corresponding to the 
 conception of dependence, and space to that of reciprocal 
 dependence. So long as this theory prevailed, no difficulty 
 could arise as to the relations of conception to perception, or, in 
 other words, as to the application of pure conceptions to 
 perceived objects ; for there could be no difficulty in bringing 
 together elements which were not essentially distinguished 
 from each other. But the argument of the Dissertation and of 
 the Aesthetic leads to the conclusion that the distinction of 
 conception and perception is not merely one of degree ; and 
 that it does not coincide with the distinction of a priori and 
 a posteriori ; but that on the contrary it rests on an essential 
 difference of kind l^ietween two faculties, each of which has its 
 own a priori : time and space constituting the a priori of sense 
 in distinction from the pure conceptions which constitute the ; 
 a priori of the understanding. It is at this point, that the 
 question as to the value or objective application of the pure 
 conceptions necessarily arises ; for if by their essential character 
 ^conceptions are severed from the perceptions through which 
 alone they can be individualised or realised in concreto, it 
 becomes necessary to show how, in spite of their opposition to, 
 and disparateness from, these perceptions, they can yet be -l 
 legitimately applied to them. If it be true that thought has 
 • Of. above, p. 90. 
 
CHAP. U. THE AESTHETIC. 285 
 
 laws of its own, which would lead it, if left to itself, to 
 determine objects in a particular way, — a way which is essen- 
 tially different from that which it must follow if it is to 
 determine objects through perception, — then we are forced to face 
 the.^jgroblfiju-€Hf-tha.j^naZy<ic, i.e., the problem whether thought 
 can be legitimately used to determine the data of perception at 
 all. And beyond this again arises the question of the Dialectic, 
 whether, besides the determination of the objects given in 
 perception, — a determination by the intelligence . of a matter 
 which is esseiitially disparate from it, and in which, therefore, 
 it cannot realise its own idea of knowledge, — the intelligence is 
 able to attain to any other determination of objects which is 
 adequate to that idea. 
 
 The Analytic and the Dialectic cannot, therefore, be understood ^"Se^s™" 
 except from the point of view reached through the Aesthetic. theySyu"' 
 
 and in tlie 
 
 The Aesthetic is the statement of a dualism, which is partly\DMfecKo. 
 overcome in the ATialytic by the (jeduction of seqse and /\ 
 thought to elements or factors in empirical knowledge/ But/ 
 in Kant's view, this dualism cannot be completely overcome ; 
 and the Dialectic, therefore, shows us thought recoiling from the 
 imperfect result of its work in experience, and setting up for 
 itself an ideal which with such materials can never be realised. 
 What hinders us from perceiving this is mainly the fact that 
 the Aesthetic fully states only one side of the opposition, 
 leaving the explanation of the other side till a later stage. 
 If, however, we connect with the Aesthetic the section of 
 the Analytic on the 'Amphiboly of the reflective conceptions,' in 
 which we have a direct criticism of that dogmatic philo- 
 sophy in opposition to which the theory of the Aesthetic 
 was developed, we are able to see clearly how the antagonism 
 there expressed is related to the combination of the opposed 
 elements attempted in the Analytic, and the partial re- 
 assertion of the antagonism in the Dialectic.^ 
 
 ^ This . point will be discussed more fully in the chapter on the Schematism 
 of the Categories. 
 
286 THE CRITIQTjE OF PURE REASON. J'OOK I. 
 
 The mete- What then is the main argument of the Aesthetic? We 
 
 physical and '-' 
 
 SuxpTsfttos have already seen how Kant in the Dissertation was led to 
 ofspaoeand gg^g^^^^j.^ ^^iQ forms of time and space from pure conceptions 
 on the one side, and from the special matter of sense on the 
 other. They are, he contends, a priori, but they are the a 
 priori not of thought but of sense, the forms of sensible per- 
 ception : and it is this view of them alone which is reconcile- 
 able with the existence and the nature of the science of 
 Mathematics. This is the general summary of the argument 
 which is divided into the two stages of what he calls the 
 inetaphysical and the transcendental expositions of space and 
 time. The ' metaphysical exposition ' shows that they are 
 a priori and what kind of « priori ^ they are: the--' tran- 
 scendental exposition ' shows that they are a source of a priori 
 synthetic judgments, and that this is possible only on the 
 supposition that they are the forms of perception. The meta- 
 physical exposition again is divided into two parts, as it proves 
 that time and space are a priori and that they are perceptions. 
 Let us take up successively these three points (1) that space 
 and time are a priori ; (2) that they are perceptions ; (3) that 
 only as. the a priori of perception, as the forms of perception, 
 can they explain, and that as such they can fully explain, our 
 a priori synthetic judgments in regard to space and time and 
 their development in the science of Mathematics. 
 (1) Space and 1 . Spacc, Kant argucs, is a priori, i.e., it is not a general S 
 
 time are n '- a > ± > > o ■ 
 
 prion. conception abstracted from particular objects previously given 
 
 in experience: for all reference of sensations to an external 
 object, i.e., to an object which is in a different place from that 
 which I occupy, or in a different place from other objects, pre- 
 supposes space. In other words, all determination of particular 
 objects as occupying a particular place and standing in a rela- 
 tion of externality to other objects, presupposes space as that in 
 which we place them. The idea of space, therefore, cannot be ', 
 got from objects as they are given in sense, for these could not 
 be determined as they are except on the presupposition of ' 
 
CHAP. 11. THE AESTHETIC. 287 
 
 space. External objects, in short, are primarily determinations 
 of space, and space, therefore, cannot be taken as a mere deter- 
 mination of thsm. An outer experience exists for us only 
 through the idea of space, as that which is necessary to 
 constitute it, and that idea cannot be abstracted from the 
 experience of which it is the condition. It may be said, 
 indeed, that we do not apprehend space before we apprehend 
 objects in it, and this is true. But the question is not of 
 priority in time. Though the spatial condition may not appear 
 except in the determination of external objects, it appears as 
 its logical presupposition. "We do not apprehend space and 
 then objects in it, but we apprehend objects on the presupposi- 
 tion of space, and we determine space by objects. Hence also 
 while we can, by abstraction, empty space of the objects which 
 determine it and thereby leave space as the presupposition of 
 an infinite number of particular determinations or objects in it, 
 we cannot invert this process and abstract space, while yet leav- 
 ing the objects or particular spatial determinations unchanged. 
 Neither in imagination nor in perception can we get rid of 
 space which is the universal condition of both. The same 
 reasoning may be repeated in relation to time, which is not 
 abstracted from the states of objects that are presented as 
 successive of coexistent, but is the presupposition of the per- 
 ception of phenomena in those relations, and which we can, 
 therefore, by abstraction empty of phenomena, though we are 
 unable to represent those phenomena as not in it. 
 
 Kant's argument is, on the one hand, directed against Leibniz, Kants argu- 
 
 o ' ' o ment directed 
 
 for whom space and time were confused ideas abstracted from Humf^t™t 
 an experience of things which were known in their pure nature Krton. 
 and relations independently of and prior to space and time, 
 and, on the other hand, against Newton, for whom space was an 
 objective reality independent of the mind. This explains cer^ 
 tain peculiarities in the form of the argument, which are apt to 
 be misunderstood by those who think of it as directed against 
 Huifie. It explains, e.g., why the priority of space and time is 
 
288 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 made the main argument for their a-priority: for it was the 
 adoption of the Newtonian view of space as prior to objects 
 which immediately led Kant in the Dissertation of 1770 to 
 infer that it is an a priori form of perception. At that time 
 Kant did not yet question the fact that individual objects as 
 such are given to us in sense through the affections which 
 they produce. But he saw that time and space are forms in 
 virtue of which these objects are arranged, or rather are capable 
 of' being arranged, in a certain order in relation to each other. 
 And, he argued that, if this order is prior to objects in our 
 consciousness of them, if they present themselves as special 
 determinations of it, it cannot be presented to us through the 
 affections in which they are given. It must be a form which 
 belongs to the receptivity of sense through which they are 
 perceived. ^ 
 
 wou*/ttS'° If Kant had been thinking primarily of Hume's reduction 
 have'teien if of time and spacc to copies of impressions,^ the argument would 
 
 directed 
 
 against probablv havc taken a somewhat different form. He would 
 
 Hume? r J 
 
 have dwelt more upon the impossibility of extracting the ideas 
 of space and time from sensation. He would have pointed out 
 that to say that the object which affects us is ' here ' and 
 ' now ' present to us will not explain how we should be con- 
 scious of it as ' here ' and ' now ' present. For we cannot be 
 conscious of the ' here ' and the ' now ' except as we relate it 
 to other ' heres ' and ' nows ' and, indeed, ultimately to all 
 ' heres ' and ' nows,' i.e., to all space and time ; and this im- 
 plies a synthetic act which brings together the manifold of 
 sense in relation to a permanent self. This line of argument, 
 however, would have involved an anticipation of the Analytic, 
 and would have at once forced us to consider the necessity of 
 the pure synthesis of the understanding for the apprehension 
 of space and time as well as of objects in them. But in the 
 Aesthetic, Kant still abstracts from all consideration of the 
 
 ^ As a matter of fact Hume's Treatise on Human Nature,, in which this reduc- 
 tion was effected, was not known to Kant. 
 
CHAP. II. THE AESTHETIC. 289 
 
 spontaneity of thought as distinguished from the receptivity of 
 sense. He, therefore, merely argues that, since the apprehen- 
 sion of space and time is presupposed in that apprehension of 
 objects which alone we can suppose to be given in perception, 
 they must be a priori forms with which the sensibility invests 
 its object in the very perception of it, and not themselves due 
 to the presentation of the object in sense.'^ 
 
 2. The second point which Kant seeks to prove is that time (2) space and 
 
 ^ ^ time are 
 
 and space are perceptions, not conceptions. His argument is i'"<'«p'io''s- 
 that all the marks by which a perception is distinguished from 
 a conception are applicable to space and time. The object of 
 a conception is universal, of a perception, individual. N"ow 
 there is only one space and one time : for all special spaces are 
 parts of the one space, and all times of the one time. " They 
 (the parts of space) cannot precede the one all-embracing space 
 as the constituents out of which it may be put together, but on 
 the contrary they can be thought of only as in it.'' To get a 
 general conception of space, we should ne&A, first to limit space 
 in different ways and then to abstract from all that is peculiar 
 to any special limitation of it. Such a general conception 
 
 ^ Cohen {Kant's Theorie der Ei-fahrung, p. 95) supposes Kant in his first argu- 
 ment to be opposing Hume, because Hume regarded space as an abstraction 
 from the images of perception, which were supposed to present us with 
 coloured points in a certain order, while Leibniz calls space an abstraction only 
 in order to exclude the idea of its independent existence as separated from the 
 monads, and regards the idea of it as an " eternal truth." But Kant was not 
 acquainted with Hume's Treatise, in which alone this derivation of space from 
 sensible impressions is given, and on the other hand, he certainly thought that 
 Leibniz, as he conceived space to be a relation of things which presupposed 
 their existence, was logically bound to regard it as dependent upon the experi- 
 ence of those things. This is clearly indicated in the passage where he con- 
 trasts the opinions of Newton and Leibniz (Aesthetic, § 7 ; A. 39 ; B. 57) and in 
 the corresponding passage of the Dissertation (R. I. 323 ; H. IL 407). Since 
 writing the above, I find that in the Eejkxionen Kant's (IL 347) it is directly 
 stated that "if the conception of space, as Leibniz thinks, were taken from 
 things, propositions relative to space, as being empirical judgments, would not 
 have any apodictic certainty. " In the note Dr. Erdmann points out that this 
 proves Cohen's opinion to be mistaken. It may also be pointed out that 
 Leibniz's assertion that the idea of space is an ' eternal truth ' is in conflict 
 with lais assertion that it is a confused idea. 
 
 VOL. I. T 
 
290 THE CRITIQUE OF PURE REASON. book i. 
 
 would, however, contrast directly with the idea of the one 
 space, which is presupposed in all limitations of it, i.e., in 
 all perceptions of special parts of space or of objects in space. 
 For a general conception, such as animal, plant, etc., is no 
 whole in which • all special animals and plants are parts, but 
 merely represents the common element present in each and all 
 of them. Space and time, further, are represented not only as 
 wholes but as infinite wholes, i.e., wholes which have an 
 infinite number of parts in them. At least we may go on 
 ad infinitum adding part to part without exhausting space, 
 (and we can see from the beginning that we shall never 
 exhaust it,) just as renewed or continued attention to the 
 perception of an object enables us to go on detecting new 
 qualities in it.^ But a conception is made up of a definite 
 number of marks, which cannot be increased without altering 
 the conception. It may contain an indefinite number of 
 individuals under it, but it cannot contain an infinite number 
 of possible determinations in it. 
 The argument The argument here given is directed against the view of 
 
 Is directed o o o 
 
 LeSz. Leibniz, who had denied the qualitative distinction of con- 
 ception and perception, and who regarded the difficulties 
 attending the conceptions of space and time as part of the 
 obscurity which necessarily attaches to perception as such. 
 Kant maintains, on the contrary, that whUe the supposed 
 obscurity does not exist, while perceptions are in their way 
 as distinct as conceptions, yet their distinctness is of a different 
 kind, of a kind which makes them incommensurable with con- 
 ceptions. It was this, in fact, that Leibniz really was aiming 
 at when he spoke of their obscurity.^ The obscurity consists 
 
 ^Cf. Beflexionen Kant's, II. 356. "The synthetic judgments regarding 
 space are not contained in the universal conception of it, any more than the 
 empirical judgments, as to the composition of gold are included in the general 
 conception of it. In both cases they are drawn out of, or found in, the 
 perception of the object." 
 
 2 The undividedness and infinite divisibility of the continuous is taken by 
 Leibniz for the confused representation of a divided aggregate of separate 
 monads. Cf. Introduction, Ch. III. p. 87. 
 
CHAP. II. THE AESTHETIC. 
 
 291 
 
 in this, that the object of a perception is individual and as an 
 individual is considered to have an infinity of determinations 
 which w^e can never think out to an end, or gather up in a 
 final conception. Hence perception, when it brings such an 
 object before us, offers an infinite problem to conception, a 
 problem which conception can never completely solve. For we 
 can never completely '' define the individual." Now such 
 inexhaustible individuality we find in time and space which 
 are " regr^entedas infinite given whole s " and, therefore, as 
 wholes which we can never completely " bring to conception." 
 
 When we put the matter in this way we provoke a natural ^l^^"^^^*/™!^ 
 objection, that the inexhaustibility of perception, or the impossi- ^""^3™//? 
 bility of defining the individual, does not mean that once for 
 all the individual in all its- determinations is given to us 
 in sense, and that conception then proceeds to analyse it in an 
 endless process. Space and time, no doubt, contain in them 
 the possibility of endless determinations, but only as a possi- 
 bility, the realisation of which must depend on the synthetic 
 activity of thought. . But to speak of these determinations as 
 existing in space and time prior to such activity is to suppose 
 the possibility realised independently of the only activity which ~ 
 for us is capable of realising it. Space and time are not for us 
 infinitely divided but infinitely divisible, not infinitely extended ■ 
 but infinitely extensible. All the determinations of quanta 
 which are discovered by the science of mathematics are not 
 determinations existing in pure space and time and given once 
 for all in the perception of them, but determinations which are 
 reached by the constructive effort of imagination guided by 
 thought. ( In like manner, the individual objects of experience 
 are not given once for all in perception as infinitely determined, 
 but they are continually being determined in the synthesis of 
 perceptions, which we bring together in thought and refer to 
 identical _ objects. It is true that in this process we seem to 
 ourselves to be discovering the properties of objects, which 
 exist independently of the process by which we discover them 
 
292 THE CRITIQUE OF PURE REASON. BOOK i. 
 
 and which in all that process we can never exhaust ; but it 
 would be absurd to say that all that we discover is already 
 given in the perception by wliich the object is first presented 
 to us. Eather, we must regard that perception as itself a step 
 or element in the process of determination of an object, which 
 is not perceived, any more than it is conceived, in its complete- 
 ness. And the facts of the case would be more truly expressed 
 by saying that we are guided in the whole process of experi- 
 ence by the idea of an object to the complete determination of 
 which, in the continuous synthesis of perceptions under con- 
 ceptions, we are continually approximating but can never 
 attain. The synthetic process does not; therefore, consist 
 in continually taking more of what is given at the first in 
 perception, but in continually bringing together new percep- 
 tions, in reference to an object thought as identical. 
 
 Now, it is certain that this view of the process of knowledge, 
 
 was not overlooked by Kant. It is, in fact, the very view 
 
 which later on in the Critique he himself sets before us. The 
 
 Kant's own -Aesthetic, indeed, tells us that space " is represented as an infinite 
 
 thr-riewofthe givcu wholc," and seems to imply that it is given as such in 
 
 Aesthetic, . i T^ • 7 • i 
 
 perception, but the Dialectic shows us that such a representa- 
 tion is simply an 'idea,'^ i.e., the thought of something which 
 can never be given in perception any more than it can be 
 brought to a unity in conception. In the Aesthetic, however, 
 Kant deliberately abstracts from the process of thought and 
 imagination, and treats of perception as if it were not a process, 
 but a passive reception of the result of the process. And he 
 even erases some expressions in the first edition in which the 
 former view had escaped from him : for while in the first 
 edition he had said that " if it were not that the process of 
 perception is limitless, no conception of relations could carry 
 with it a principle of their infinity," i.e., not that we have a 
 perception of the space containing all its determinations but 
 that we have in it a potentiality of them : in the second edition 
 ^ A. 427; B. 456 mte. 
 
CHAP. II. THE AESTHETIC. , 29 d 
 
 he says merely that " no conception as such can be thought as 
 containing an infinite multitude of ideas in it," as if space as a 
 perception were supposed by him to contain such an infinite 
 multitude. Kant, in fact, is determined to make us follow him 
 step by step, and, therefore, he treats for the moment the 
 infinite potentiality of determinations, which belongs to the 
 individual as an object of perception, — though really be- 
 longing to it merely as an object which we represent as the 
 same in the synthesis of many perceptions, — as if it were 
 actually realised in the first immediate perception of it ; or at 
 least as if in that perception all these determinations were at 
 once present, though the mind may need a long and even end- 
 less process to make them explicit or 'bring them to conceptions.' 
 And the reason why he does this seems to be, that he wishes, 
 in the first instance, to refute the Leibnizian view of space and 
 time, as conceptions, assuming provisionally the ordinary dis- 
 tinction made between perceptions and conceptions. A con- 
 ception, so to speak, is a certain definite amount drawn from 
 the inexhaustible bank of perception : it is the combination of 
 a certain ^dsfeite number of marks. Hence no conception will 
 afford matter for an indefinite process of analysis. If space 
 and time, were, as Leibniz said, ' obscure conceptions,' their 
 obscurity at a certain definite point must vanish before analysis : 
 for, at that point, we should get down to the simple elements of 
 the combination. But we can see that this is impossible in the 
 case of time and space in so far as they are " represented as 
 infinite given wholes '' ; which, therefore, cannot be analysed 
 into a finite number of parts, or regarded as compounded out of 
 these parts.i Por Kant's immediate purpose it was not neces- 
 sary to observe that the ideas of time and space as continuous 
 quanta contain only the potentiality, and not the actuality of 
 such infinite determination. That we can see that there is no 
 
 ^ "No conception as such can be thought of as if it contained an infinite 
 number of ideas in it. But space is so thought : for all parts of space ad in- 
 Jbnit^llln exist at once," B. 39. 
 
294 THE CRITIQUE OF PURE REASON. J^OOK I. 
 
 limit to the possibility of their determination is, from the point 
 of view of the Aesthetic, the same thing with their being actu^ 
 ally determined in all the infinity of their extent and division. 
 For in any case the determining process presupposes space and 
 time and does not produce them, any more than it produces 
 the individual things which are given in perception as in space 
 and time. In this point of view, there seems to be a special 
 force in the words " represented as given." The ordinary con- 
 sciousness represents both as given in all their infinite com- 
 pleteness of determination, and regards all the process of 
 thought by which the special determinations are discovered as 
 one of simple analysis of what is given. Now an ' infinite 
 whole ' cannot, as Kant afterwards shows, actually be ' given,' 
 any more than any other individual object in the completeness 
 of its individual determination can be given. Determination 
 of space and time and of individual objects in space and time 
 can take place only in a successive process of synthesis in 
 which the mind must be guided by a definite conception or, 
 rule. At the same time, in this successive process we always 
 presuppose the totality of determination of the object in itself, 
 as the goal towards which in our synthetic process of thought 
 we are tending. While, therefore, the progressive determination 
 of the object appears on the one hand as a synthesis, proceeding 
 a parte ad totum, which is never completed : on the other hand, 
 it is equally true that the idea of the whole, as that in, and in 
 reference to, which the part is determined, is implied in the 
 process from the first, and gives to it its interest and direction: 
 and in this point of view knowledge may be said to proceed a 
 toto ad partes. This idea of the whole, however, we do not de- 
 termine ; in Kantian language, we can never make it an object 
 of knowledge. But the thought of it is the presupposition of 
 all our efforts in determining such objects. The unity of all 
 experience, of all objects with each other and with the mind 
 that knows them, is, so to speak, the horizon within which all 
 special objects are determined, and \ye can see it as the limiting 
 
CHAP. 11. THE AESTHETIC. 295 
 
 condition of all knowledge but not as itself an object of know- 
 ledge. From this point of view all our progress in knowledge 
 may be regarded as a specification or differentiation of a pre- 
 supposed whole, and not as an integration of part to part with- 
 out any limit. It is this idea of a presupposed totality, within 
 which our experience grows, which appears in the Aesthetic as 
 the representation of space and time as ' infinite given wholes,' 
 by limitation of which particular places and times (and objects 
 as occupying these places and times) are determined. 
 
 The full explanation of these ideas cannot yet be given. 
 Enough, however, has been said to explain the fact that so 
 soon as, at the opening of his Analytic, Kant goes beyond the 
 common view of knowledge as the analytic recognition of the 
 different elements involved in the complete determination of 
 things as given in sense, — so soon as he begins to recognise that 
 a spontaneity of thought is necessary in the first instance 
 to connect the elements which are afterwards separated by 
 analysis, — at that moment the idea of a given totality dis- 
 appears and has to be replaced by the idea of a totality which 
 can be thought but can never be given, or even reached by the 
 immediate synthesis of that which is given. Something more 
 will be said on this subject at the end of this chapter.^ 
 
 The third step, which Kant takes in the " transcendental ex- (s) tms view 
 
 > of space and 
 
 position," is to show that time and space are principles of an *™^^^?{^jj.^ 
 
 of mathemati- 
 cal science.. 
 
 ' It is obvious that we have here the idea of a unity of experience pre- 
 supposed in all special determination of objects, an idea which in the Critique 
 takes the place h^ld by the idea of God in Kant's pre- critical philosophy. Ac- 
 cordingly, in the third section of the Dialectic, Kant attempts to show the 
 illogical nature of the process by which the Ideal of reason, which in reality is 
 merely the " regulative unity of experience," is first realised, i.e,, turned into 
 an object, then hypostasised, and finally in the natural progress of reason to 
 tlfe completion of its unity, personified (A. 583 ; B. 611 note). In the doctrine 
 of the Dissertation we have a link between the two ideas in the conception 
 of space and time as Omnipraesentia Phaenomenon and Aeternitas Phaenomenon. 
 Throughout Kant conceives the universal {i.e., not the abstract universal but 
 the universal as the principle of a systematic whole) as the presupposition of 
 the particular, though this is modified by the identification of the former with 
 the a priori element in the consciousness of the subject. „. , 
 
296 THE CRITIQUE OF PURE REASON. book I. 
 
 a priori synthesis which carries us far beyond the determina- 
 tion of time and space itself, and that this could not be possible 
 unless the conception of time and space as the a priori of 
 perception were true. A ' transcendental ' view of any idea is, 
 as Kant explains, always one which transcends or carries us 
 beyond the use of it in which we know, or suppose ourselves 
 to know, an object through it, and makes us consider how such 
 knowledge is made possible by the constitution of the faculty 
 of knowledge in the subject. It has, therefore, to do with the 
 a priori, the necessary and universal elements in knowledge, 
 or with the other elements only so far as the former elements 
 presuppose them. In the transcendental Deduction of the 
 categories, a priori conceptions are shown to be necessary to 
 determine the form and matter of perception, so as to produce 
 experience or knowledge of phenomenal objects. In the ' tran- 
 scendental exposition ' of the Aesthetic, Kant does not look 
 either so far back or so far forward : he does not show the 
 necessity of a priori conceptions to determine the forms of 
 space and time, nor does he show that this determination is 
 that which alone can constitute knowledge even of empirical 
 objects. All he does is to point out, e.g., that in Mathematics 
 we do " determine the properties of space and of time syntheti- 
 cally and a priori," and thus attain to a knowledge of all 
 objects in space and time — a knowledge which is not and 
 cannot be derived from an empirical consideration of those 
 objects, but anticipates it and lays down universal and neces- 
 sary laws for it. On the other hand, he argues that such 
 synthesis would not be possible unless space and time were 
 perceptions ; and it would not be a necessary synthesis, i.e., a 
 synthesis accompanied with the consciousness that the elements 
 which it combines are necessarily and universally combined, 
 unless they were a priori perceptions. For that must be a 
 perception which amplifies a conception or gives it meane to . 
 amplify itself And that perception must be a priori, or in- 
 volved in the very constitution of the faculty of perception, 
 
CHAP. II. THE AESTHETIC. 297 
 
 ' which enables us with apodictic certainty to anticipate actual 
 perception and lay down laws for it 
 
 In this argument, as has been already indicated, the pre- How 
 
 7 7 mathematical 
 
 ' ' critical theory of perception is ■ not yet essentially changed. ^^f/^?\^™ 
 Objects are supposed to affect us and produce a perception of ^^venenee. 
 themselves as individual objects. But from this perception we 
 cannot draw any conclusions which go beyond the particular 
 objects, or rather the particular phase in which the objects 
 present themselves to us in each particular case. So far as 
 we know the object throvigh the affections of sense, we know 
 merely that it has affected us in a particular way ; but whether 
 it will ever do so again we know not. We know nothing of 
 the object in itself, but only the appearance which it has 
 for us through a particular affection, and hence we cannot 
 say anything as to the constancy of that relation of the 
 object to our sensibility which is involved in its producing 
 that particular affection. The object is for us the objectified 
 sensation and not the thing that produces the sensation, 
 and the sensation is merely a particular state of our sensi- 
 bility. If we are able to say anything about such an 
 object which holds good universally and necessarily, it must 
 be, not because of the affection, but because of the nature 
 of- the sensibility into which the affection is received and 
 in which it becomes transformed into a perception of an 
 object. Necessity and Universality, therefore, cannot be 
 the effect of the object upon us ; they must be the result of 
 the form with which perception invests its objects. Hence we 
 are able to say not only that, if there is any universal and 
 necessary element in perception, it is derived from the form of 
 perception, but also that, if there is any knowledge derivable 
 from the form of our perception, it must be necessary and 
 universal, i.e., must equally affect all objects that are perceived 
 by -us. Geometry, e.g., as a synthetic a^priori science is possible, 
 if space is the form of all our perceptions of external objects : 
 anS if space is such a form, then Geometry, as the science 
 
298 THE CRITIQUE OP PURE REASOK. BOOK I. 
 
 in which principles of spatial determination in general are 
 developed, must consist of synthetic a priori propositions, in 
 which with apodictic certainty laws are laid down for all 
 objects in space. What is .universal and necessary in the 
 determination of objects is due to the nature of the subject for 
 which they are, seeing that the object, as Kant naively ex- 
 presses it, cannot ' pass over ' into our minds, but only affect 
 our sensibility ; and the particular affections which it produces 
 cannot warrant any conclusions beyond themselves. 
 Howfar does \i ig obvious that such a view does not in itself imply a 
 
 the Aesthetic 
 
 OTdtoi^ dig- complete awaking from " dogmatic slumber " ; for it does not 
 things ? yet raise the question what is meant by an ' object ' of percep- 
 tion, or how far an ' object ' can exist for a consciousness which 
 is merely receptive and which does not go beyond particular 
 affections so as to combine them in one conception. The 
 Dissertation, of which the Aesthetic is little more than a 
 repetition, took its start from the naive belief that individual 
 objects of perception, as such, may be given through the affec- 
 tions of sensibility which they produce ; and it modified that 
 belief only so far as to point out that those individual objects 
 are perceived as special determinations of time and space, and 
 that time and space, therefore, are prior to all other elements of 
 perception. A priori knowledge is knowledge which conditions, 
 as opposed to a posteriori knowledge which constitutes, the 
 perceptions of the individual objects through sense. Kant 
 does not yet attempt to show that it is just the subjection of 
 the particular element in sense to such universal and necessary 
 conditions which makes us refer them to objects. But when he 
 points out that they are in themselves merely particular and 
 contingent, he prepares the way for such a proof. And in the 
 Prolegomena, where he does not follow the same strict method 
 of evolving knowledge from its elements, he already anticipates 
 this kind of ' deduction ' of space and time.^ For there he 
 argues that his view of space and time is so far from reducing 
 ' Prolegomena, § 13, Remark 3rd. 
 
CHAP. II. THE AESTHETIC. 299 
 
 them to illusions that it is the only view which enables us to 
 prove their objective reality : seeing that objectivity and deter- 
 mination by universal and necessary principles are one and the 
 same thing. The full development of this thought, however, 
 following Kant's synthetic method in the. Critique, we must 
 postpone till we come to consider the deduction of the Cate- 
 gories in the Analytic. 
 
 Kant's discussion of the nature of time and space in the Empirical 
 
 '- reality and 
 
 metaphysical and transcendental expositions he now turns to tai'ideaut™"f 
 account to answer the question with which he had begun : " Are tFme! ™ 
 space and time real things ? Are they qualities or relations 
 that attach to such things apart from our perceptions of them ? 
 Or are they a priori and due to the subjective constitution of 
 the perceptive faculty ? " In saying that certain ideas are a 
 /priori, Kant might mean to oppose either the doctrine that they 
 represent objects and qualities of objects independently of the 
 manner of their apprehension by us, or the doctrine that our 
 apprehension of them is due to sensation. But it is the former 
 opposition which Kant has in view in the Aesthetic, as is 
 evident from the argument which we have already considered. 
 The priority of space and time to objects, i.e., the fact that 
 objects presuppose time and space and appear as special deter- 
 minations of time and space, was what first proved to Kant's 
 mind that the mere presentation of objects to sense could not 
 bring with it, or have consequent upon it, the consciousness of 
 their temporal and spatial relations, but that that consciousness 
 must be a contribution of the mind in perceiving them. 
 "Neither absolute nor relative determinations of things could 
 be perceived before the things to which they belong are 
 presented to us : hence they could not be perceived a priori."^ 
 Hence time and space, which are perceived a priori, are not 
 such determinations, and they must disappear •" whenever we 
 abstract from the subjective conditions of perception." If 
 space or time were taken as a thing which "subsisted for 
 * lA. 26; B. 42. 
 
300 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 itself," or, in the language of Spinoza, a res completa, — an 
 objective somewhat independent of our ways of perceiving, — 
 we should have in it a strange problematical existence (an 
 Unding), " a something which though not a real object was yet 
 real"; in other words, an order or form of arrangement for 
 things in relation to each other existing without any objects 
 to be arranged.^ If it were taken as au attribute or relation 
 attaching to such a thing, it and its properties could not be 
 determined a priori, but only empirically perceived in the thing 
 of which they were the determinations. Hence it is only 
 ' from the point of view of a man,' i.e., of a being who has such 
 a sensibility as ours, that we can talk of space and of extended 
 things, or of time and events happening in it. In both cases 
 what we have before us is not reality in any ultimate sense, 
 but reality as it appears to us under our forms of sense, i.e., 
 phenomenal reality. Space is the form of outer sense or " a 
 necessary condition of all relations in which objects are per- 
 ceived as external to us." Time is the " form of inner sense, 
 i.e., of our perceptions of ourselves and our inner states. It 
 cannot be a determination of external phenomena : it belongs 
 neither to figure nor position, etc., but determines merely the 
 relation of ideas in our inner state." " Yet as all ideas, whether 
 they have external things for their object or not, nevertheless 
 
 iCf. Kant's answer tp the objection of Schultz, who asked why space may 
 not be regarded as a true intellectual intuition or perception and as therefore 
 objective, (in letter to Herz dated 20th February, 1772. R. XI. 30 ; H. VIII. 
 692. ) In the Dissertation which Schultz was criticising, it will be remembered 
 that space is regarded as the form in which the reciprocal connexion of all 
 individual substances in this world, due to their common dependence on God, 
 Is represented in our sensible perception of them, i.e., it is Omnipraesentia 
 Phaenomenon. In the Critique the centre of unity in knowledge is changed from 
 God to the conscious self : and we might, therefore, say that space is the form 
 which in our sensible perception is taken by the connexion of all known 
 substances due to their common dependence on the unity of the self that 
 knows them. The same remarks mutatis mutandis might be made of time, 
 which in the Dissertation is Aeternitas Phaenomenon, and in the Critique comes 
 to be the form which in our sensible perception is taken by the causal depend- 
 ence of the changes of known substances due to their relation to the unity 
 of the knowing self. 
 
CHAP. II. THE AESTHETIC. 301 
 
 as determinations of the mind belong to our innet state ; and 
 as this inner state falls under the formal condition of inner 
 perception, i.e., of time, time is a condition of all phenomena- — 
 immediately of internal, and mediately also of external, pheno- 
 mena." Hence we cannot say that all things are in time or in 
 space, but we can say that all phenomena, i.e., all things as 
 they present themselves in perception to us, are in time, and 
 all external phenomena are also in space. This is expressed 
 more formally by saying that time and space are empirically 
 real and transcendentally ideal ; i.e., real in as far as they are the 
 forms under which objects are presented to us in sense, real 
 from the point of view of the ordinary consciousness or of 
 experience, but ideal when we look at them from the point of 
 view of that distinction between things in themselves and 
 things for us, which a critical view of the faculty of knowledge 
 forces us to make. 
 
 The " ideality " of space and time means their subjectivity. Distinction of 
 
 J r .1 J primary and 
 
 At the same time we must distinguish it from the subjectivity quamtes^ 
 of tastes, sounds, colours, the secondary qualities of matter Kant, 
 which were already, referred by Locke to the subjective consti- 
 tution of our special senses, in opposition to the primary 
 qualities which were conceived to belong to things in them- 
 selves. Kant admits the distinction between primary and 
 secondary qualities, but changes its meaning. It ceases to be 
 a distinction between qualities which belong to things and 
 qualities which indicate the " subjective nature of our sensi- 
 bility " : for in that sense both primary and secondary qualities 
 are subjective. But it remains as a distinction between that 
 which is universal and necessary in perception, because due to 
 the very form of perception, and that which is particular and 
 contingent, because due to the special affections we receive 
 from the objects. And Kant even suggests, — what, however, 
 if fully developed, would anticipate the argument of the 
 Analytic, — that these affections would not of themselves give 
 rise*to any consciousness of an object at all. They are " mere 
 
302 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 sensations and not perceptions, and therefore do not give rise 
 to any knowledge of an object, least of all to an a priori know- 
 ledge of it." " We must be careful, therefore, not to illustrate 
 the ideality of space (and time) by examples which are utterly 
 insufficient. Thus when it is said that colours, tastes, etc., are 
 not to be regarded as qualities of things, but merely as changes 
 in our subjectivity, that which is originally nothing but a 
 phenomenon, e.g., a rose, is taken for a thing in itself in an em- 
 pirical sense, though it may yet appear differently in respect of 
 its colour to different persons." Thus such ' a thing in itself in 
 an empirical sense ' -"^ must from a transcendental point of view 
 be regarded as merely an appearance, inasmuch as it is per- 
 ceived in space, 
 obtetimto Having thus expounded his doctrine, Kant proceeds to give 
 time-'^^"^'^ ° some further illustrations of it and to meet some objections 
 that had been taken to it. Lambert had objected to the 
 doctrine of Kant in relation to time, that it seemed to imply 
 that there is no such thing as change. Now, even if we reduce 
 external changes to appearances, we are directly conscious of 
 changes of our ideas, changes which are real, and involve the 
 reality of time as the presupposition of change. Kant answers 
 that the Critique does not imply a denial of the reality of the 
 changes which we recognise as taking place in our mental states 
 when we make ourself an object. But it points out that, because 
 that reality is apprehended by us under a special form, it cannot 
 be known by us except as a phenomenon. " If, however, I could 
 perceive myself or any other being could perceive me, without 
 the condition of sensibility, then those determinations which 
 we now represent as changes, would be known by us in a way 
 which did not involve the idea of time, nor, consequently, the 
 idea of change." The reason, however, why the doctrine of the 
 ideality of time had seemed a hard doctrine even to those who 
 could admit without difiGiculty the ideality of space, was that 
 the doctrine of the ideality of time affected inner experience, of 
 ^ A noticeable expression. 
 
CHAP. II. THE AESTHETIC. 303 
 
 which time is the form. Now, it was generally supposed that, 
 while external things are known by us only indirectly through 
 the inner states which they awake in us, these inner states 
 themselves are known directly and immediately. While, 
 therefore, we may admit a doubt in regard to the reality of the 
 former, we cannot admit any doubt in regard to the reality of 
 the latter. This, however, rests on a misconception as to the 
 nature of our knowledge of ourselves and our inner states. We 
 do not know ourselves as we absolutely are, but only as we 
 appear to ourselves under the form of time. In the case of 
 internal, as in the case of external, experience, we must draw 
 the distinction between the phenomenon and the thing in itself 
 or noumenon ; and we must remember that while we can think 
 the noumenon, we can know only the phenomenon. The value 
 of this answer will be considered hereafter in connexion with 
 the relations of inner to outer sense, a subject which re- 
 quires for its full discussion many of the conceptions of the 
 Analytic. Here it is sufficient to point out that, if outer sense 
 implies that we are affected by things in themselves, which are 
 not known to us except through the affections they produce, 
 inner sense implies that we are affected by ourselves ; and 
 that in neither case do we know that which produces the 
 affections except through the affections it produces, and under 
 the characteristic a priori forms which they receive in being 
 perceived. In both cases the form of arrangement of the data 
 of sense in relation to each other is presupposed and not itself 
 given through the affection : so that, even if we could suppose 
 the affections in themselves to correspond to the objects pro- 
 ducing them, (which involves an absurdity,) yet the form of 
 arrangement would itself introduce a discord between them. 
 There is, indeed, this difference between inner and outer per- 
 ception, that the object of the former is designated or marked 
 out for us by pure apperception or self-consciousness. But as 
 this apperception tells us only that we are and not what we are, 
 it dees not affect the cLuesti()n of knowledge. 
 
304 THE CRITIQUE OF PURE REASON. book i. 
 
 Relation of Kant then soes on to illustrate the doctrine of the empirical 
 
 Kant's ° 
 
 the*5ewsof reality and transcendental ideality of time and space by con- 
 Nlwton™'^"' sidering the light it casts upon the controversy as to the nature 
 respeo ive y. ^^ gp^ce and time, in which Clarke appeared as the representa- 
 tive of the Newtonians in opposition to Leibniz. That contro- 
 versy was not decisive in its result, because each side was able 
 on its own theory to explain a part of the characteristics of the 
 ideas in question, but neither was able to explain the whole of 
 them. The strong point of the Newtonians lay in their up- 
 holding the priority of space : for this, as we have seen, 
 enabled them to explain the possibility of mathematics as a 
 science which lays down a priori laws for all perceptions. As 
 all objects, according to this theory, are contained in space and 
 time, so we can understand that what is proved in relation to 
 space and time holds good for all objects. But then, in so far 
 as time and space are taken as existing in themselves, i.e., apart 
 from the relation of things to our faculty of perception, these 
 advantages are purchased at a great cost. For, in the first 
 place, what are we to make of the objective existence of what 
 after all are but forms of relation, prior to and independent of 
 any things related — " two eternal and infinite nonentities exist- 
 ing without anything actual being there, in order to embrace 
 all that is actual within them ? " How can we think of an . 
 infinite possibility of relations as an actuality subsisting by 
 itself ? In the second place, even if we set aside this difficulty, 
 we shall be obliged to treat time and space as conditions of the 
 ' supersensible as well as of the sensible, which is equivalent to 
 the denial of the existence of anything supersensible what- 
 soever. For then we fall into the " dangerous dilemma, 
 spoken of by Berkeley, of supposing " either that space is God, 
 or that there is something besides God which is eternal, un- 
 created, infinite, indivisible, immutable."^ Are we to " bind 
 God in matter or diffuse in space " ? It was in the attempt to 
 escape from such a result that Berkeley was led to reduce both 
 ' Berkeley, The Principles of Human Knowledge, % 117. 
 
CHAP. II. THE AESTHETIC. 305 
 
 space and extended substance to illusive appearances. Thus 
 the supposition of the transcendental reality of space which is 
 adopted by Newton, leads inevitably to the denial of its 
 empirical reality ; and the only way to save the latter is to 
 say that space is transcendentally ideal. And this leads us to 
 recognise the element of truth contained in the views of 
 LeibniZj who discovered the subjectivity of time and space and 
 thereby avoided the objections which have been just stated. 
 Leibniz, however, exposed himself to equally strong objections 
 on the other side, by misconceiving this subjectivity as a mere 
 " confusion " introduced by sense into our ideas of the relations 
 of things. For this view implies that space and time are not 
 a, 'priori, (because not prior to, or presupposed in, the percep- 
 tion of the things in question,) and makes it impossible to 
 explain how mathematics can derive from them universal laws 
 for perception. In other words, it explains the limitation of 
 mathematics to experience only on a principle that destroys its / 
 a 'priori character. 
 
 Let us, then, recapitulate Kant's results, remembering always summa^ 
 
 ' ' -"^ a ./ „g^ of Kant s 
 
 the assumption on which they are reached, viz., that individual '^o''*™^- 
 
 objects are presented to us in sense. On that assumption it is 
 
 argued that space and time are presupposed in the objects of 
 
 perception ; for such objects, if external, are represented as in 
 
 space and time, and, if mere modifications of mind, they are 
 
 represented as in time : and this means that all such objects /■ 
 
 appear as special determinations of space and time. . Hence 
 
 space and time cannot be given in the perceptions of these 
 
 objects, but are a ■priori conditions of such perceptions. Space 
 
 and time, further, are related to the special objects perceived, 
 
 not as general ideas under which they are subsumed, but as - > 
 
 individual wholes, of which their individuality is a limitation 
 
 or determination. They, in fact, exist for us as individual 
 
 objects only as we make a special construction in space and 
 
 time, which, as all-embracing, individual wholes, are presup- 
 
 poseti in that construction. And this enables us also to under- 
 VOL. I. u 
 
306 THE CRITIQUE OP PT7BE REASON, book i. 
 
 stand why a science, which anticipatively, i.e., without waiting 
 for objects to be given, determines or limits space and time in 
 various ways, or, in other words, makes arbitrary constructions 
 in them, should yet supply us with universal principles which 
 are necessarily applicable to all objects of perception. At the 
 same time it is obvious that such objects, — objects the deter- 
 mination of which is reached by such an a ^priori process, — can- 
 not be regarded as things in themselves, or objects independent 
 of all relation to our subjectivity. They can only be objects 
 as they are for our perception, not objects as they are in them- 
 selves ; they can be only phenomena and not noumena. 
 Trendeien- Xhis argument is not susceptible of attack, if we admit its 
 
 burg 8 attempt ° r > , 
 
 Kant's^^ presuppositions. Trendelenburg, indeed, seemed to make a 
 d- emma. plausible criticism on Kant's view when he objected that the 
 dilemma set up in the Aesthetic, between the objectivity and the 
 subjectivity of space and time, is not an exclusive one. For why, 
 he asked, should they not be both objective and subjective at 
 once ? Why should they not be a priori possessions of the 
 mind, and yet at the same time elements of knowledge, not 
 merely of phenomena, but of things in themselves ? Kant's 
 dilemma is thus defective : for it does not cover the whole field 
 of possibility. To this the simple answer is that the alterna- 
 tive supposed to be overlooked is excluded by the very way in 
 which the problem is stated by Kant. Kant, in fact, starts 
 with the supposition that individual objects are given through 
 affections of sense which they produce, and argues, in the 
 first place, that, because they are so given, the space and 
 time in which they are given is their presupposition, a pre- 
 supposition which must be due not to the affections but to the 
 nature of the receptivity of sense in which they are given ; and 
 in the second place, that this a priori subjectivity or ideality of 
 time and space explains how universal and necessary prin- 
 ciples applicable to the objects of sense may be developed by 
 mathematical science. Space and time and their a priori 
 determinations can be empirically real only because ' they are 
 
CHAP, II. THE AESTHETIC. 307 
 
 transcendentally ideal. If, on the other hand, they were 
 transcendentally real, if they belonged to the things in them- 
 selves which produce our sensible affections, they could be 
 known to us only through the affections of our sensibility, and, 
 therefore, only in their contingent individuality : in other 
 words, the knowledge we have of them would refer to nothing 
 but the individuals actually perceived in the moment when 
 they are perceived, and could have no value as a universal and 
 necessary principle of perception. Mathematics could not 
 anticipate empirical perception by a priori construction, nor 
 could time and space be the presuppositions of our perceptions 
 of objects. It appears, therefore, that, on Kantian principles, 
 the transcendental reality would necessarily imply the empirical 
 ideality of space and time, i.e., would imply that time and space 
 and all their a priori determinations are illusive ; and, on the 
 other hand, that their empirical reality can be based only on their 
 transcendental ideality. But Trendelenburg's supposition, that 
 space and time may be both 'subjective' and 'objective,' can mean 
 only that they may be both empirically and transcendentally 
 real at once; i.e., it involves a pre-established harmony between 
 the things as they are and the necessary subjective forms 
 under which they are perceived — a supposition which, as Kant 
 repeatedly points out, is absurd, because it supposes us to 
 transcend a distinction which at the same time we admit to be 
 for us absolute, and to know that which, ex hypothesi, is out of I 
 consciousness. " What," he asks, " should I need in order to 
 avoid the idealising of spa,ce ? It is obvious that I should need 
 to say, not only that the idea of space completely corresponds 
 with the relation which our sensibility has to objects, which is 
 what I have said, but that it is in all points like the objects 
 themselves. But this is an assertion to which I can attach no 
 meaning whatsoever, any more than I can attach a meaning to 
 the assertion that the sensation of redness is like that quality 
 of cinnabar which excites the sensation in me." ^ 
 ' Prolegomena, § 13, Remark 2nd. 
 
308 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 Impossibility The " third course " of Trendelenburg was, therefore, not 
 
 of escaping it ' • i» -l ' 
 
 sSppoSttonsf' overlooked by Kant ; it was excluded by the necessity of his 
 argument. The priority and a-priority of time and space are 
 essentially inconsistent with any view of them as conditions 
 or attributes of things in themselves. This is a conclusion 
 which Kant could not escape, except either by denying the 
 reality of things in themselves, or by asserting that the 
 objects of sense are such things. But, if the latter alterna- 
 tive were adopted, the priority of the knowledge of the con- 
 ditions or possible relations of such objects to our knowledge 
 of the objects themselves seemed to him inexplicable; while, if 
 the former were adopted, the a-posteriority of our knowledge of 
 the objects and its dependence upon sensible affections became 
 inexplicable. No doubt a step was taken by Kant to escape 
 the latter difficulty when he pointed out in the Analytic that 
 through these affections, apart from the synthesis of the under- 
 standing, objects are not presented to us as such. For, if the 
 reference of affections to objects is due to the spontaneity of 
 thought, that spontaneity must itself enable us to grasp the 
 objects to which it makes us refer the affections. But this 
 only extinguishes the opposition in one form to make it re- 
 appear in another : for the objects thus reached by the synthesis 
 of the given matter of sense under the form of sense, do not 
 correspond to the unity of thought from which they receive 
 their determination, a discord which necessarily gives rise to 
 the opposition of noumena (i.e., of the ideas of objects that do 
 so correspond) to the phenomenal objects actually apprehended 
 through the determination of sense by thought. It appears, 
 therefore, that there was no way for Kant to determine the 
 forms of space and time as at once empirically and transcen- 
 dentally real (which is what Trendelenburg must here mean by 
 objective and subjective), except one which involved a surrender 
 of his fixed distinction between conception and perception, 
 between the spontaneity and the receptivity of the mind. For 
 only by such a surrender would it have . been possible to reach 
 
CHAP. II. THE AESTHETIC. 309 
 
 the identification of the nonmenal objects of pure thought with 
 the phenomenal objects perceived in space and time, or even 
 the reduction of the distinction between them to a relative 
 distinction (so that phenomena should be seen to be noumena 
 imperfectly apprehended, and noumena to be only phenomena 
 perfectly understood). But such a solution of the opposition 
 between phenomena and noumena, as respectively ' subjective ' 
 and ' objective,' or in Kant's language, empirically and transcen- 
 dentally real, could not be realised by him even from the point 
 of view of the Analytic, and from the point of view of the 
 Aesthetic, the very idea of it lay beyond the horizon. For so 
 long as objects as such are supposed to be given to the, passive 
 mind through the affections of sense, it is an absurdity to 
 suppose that they can be known as they are in themselves, 
 and a double absurdity to suppose that they can be known in 
 themselves a priori} 
 
 We have now to consider how the point of view of the change of 
 
 ■*• point of view 
 
 Aesthetic changes as we pass to the Analytic, or, in other words, i^^eAnaiyUc. 
 to point out the defects in the theory of the Aesthetic, which 
 the Analytic seeks to supply. In order to do this, we have to / 
 a,nticipate the answer to two questions: (1) how, in Kant's / 
 view, do sensations give rise to perceptions? and (2) how do Jf 
 ■perceptions give rise to knowledge of objects ? / ^ 
 
 When it is said that " objects are given through sense," How do aensa- 
 
 J o ° ' tions become 
 
 or that they are " given in perception," what does this mean ? perceptions! 
 According to the ordinary view from which Kant' started, it 
 would mean that individual objects in their complete deter- 
 mination as individuals are presented to us through affections 
 of sense without any activity of thought, and that all that 
 thought can do is to dissect or analyse them. In the Aesthetic, 
 as we have seen, Kant does not disturb this view except to 
 point out that individual objects are presented to us in space 
 and time, and are thus invested with an a priori form in 
 
 ' Cf . E. Axnoldt, Kant's Transcendentcde IdeaXitat des Raumes und der Zeit, 
 whsre this subject is very thoroughly discussed. , 
 
310 THE CRITIQUE OF PURE REASON. book i. 
 
 becoming the objects of our perception. Sense is a mere re- 
 ceptivity of perceptions which, however, are subjected to its 
 own forms. At the same time, Kant does not overlook the 
 distinction between sensation and perception, and his words 
 are such as to show that the idea of passive reception could be 
 strictly applied only to the former. For while sensation is 
 described as " the effect of the object upon the mind in so far 
 as we are affected thereby," perception is said to " refer to the 
 object." The same distinction is implied in the description of 
 sensibility as " capacity of acquiring ideas of objects through 
 the manner in which we are affected by them" ; and again 
 when space is said to be the " formal constitution of the sub- 
 ject, in virtue of which it is affected by objects and acquires 
 thereby an immediate idea or perception of them." 
 Is It by means But it may be said that it is implied in such language that 
 
 of the forms "' ^ 7 
 
 alone™''*'™ *^^ forms of scusc of themsclvcs are sufficient to combine the 
 manifold elements of sensation so as to produce a perception of 
 objects. Sensations in themselves, it may be allowed, are 
 " perishing existences," which, for a being that merely feels 
 them, would have no relation, and certainly no necessary rela- 
 tion, to each other, and would therefore be incapable of being 
 referred to any object. But the forms of sense are forms of 
 synthesis. As the sensations are received, they arrange them- 
 selves, or are arranged by means of these forms, in an order of 
 coexistence and succession in space and time; and the combina- 
 tion thus produced is all that is necessary to turn them into 
 perceptions of objects. 
 
 No ; for these Xo this it must be answcred in the first place, that time and 
 
 forms require ^ 
 
 tho^h'tmcon- spacc as fovms of perception are not yet perceptions of time end 
 ofunderatMid- space, " represented as infinite given wholes " in which objects 
 
 ing to make 
 
 ^«™p«™p- are or may be arranged. Space and time do not exist, even as 
 " objects of perception," apart from the spontaneity of thought 
 by which their manifold elements are brought tb unity : for 
 " the combination of the many in one can never come into our 
 minds through sense, nor can it be contained in the pure form 
 
CHAP. It THE AESTHETIC. 311 
 
 of sensible perception and introduced along with it. It is an 
 act of mental spontaneity ; and, as we call this spontaneity 
 understanding to distinguish it from the receptivity we refer to 
 sense, so all combination, be it conscious or unconscious, be it 
 a combination of the manifold elements of perception or of 
 different conceptions, and in the former case, be it a combina- 
 tion of pure or of sensible perception, is an act of the 
 understanding. To this act we give the general name of 
 synthesis in order to indicate definitely that we, can rejpresent 
 nothing as united in the object unless we have first ourselves com- 
 hined it, and that of all ideas combination is the one which 
 cannot be given by objects but must be developed by the sub- 
 ject itself because it is an act of its self-activity."^ There is, 
 therefore, necessary to the genesis of perception as such, as 
 opposed to sensation on the one hand and to the mere forms of 
 sense on the other, an activity which Kant attributes to the 
 understanding: though not to the understanding in that con- 
 scious activity which is manifested in the application of the 
 categories, but in an ^mcQ9^sc^0MS activity, fajjyhich he gives the 
 name of i magination. And Kant warns us, therefore, that in 
 the Aesthetic there was an important omission, which requires 
 to be corrected, ere we can see the true relation of perception 
 to conception and the way in which they combine in know- 
 ledge. " Space represented as an object, as is necessary in the 
 geometrical determination of it, contains more than a mere 
 form of perception, to wit, the combination or integration of the 
 manifold given in the form of sense into a perceptive presenta- 
 tion ; so that, while in the form of perception is given only the 
 unconnected manifold, the formal perception is the conscious- 
 ness of the manifold as a unity. In the Aesthetic I referred 
 this unity to sense, but by doing so, I meant only to indicate 
 that it is a unity whidh precedes all conception. It is, in fact, 
 the product of a synthesis, which though it cannot be attri- 
 buted to thd senses, yet must be achieved before any 
 ♦ 1 B. 129, § 15. 
 
312 THE CRITIQUE OF PURE REASON. book I. ; 
 
 Y conceptions of space and time can become possible. lor it is 
 
 /\ by means of the unity of a priori perception, which is due to 
 
 '■ the direct determination of sensibility by understanding, that 
 
 space and time are at first given as perceptions. Hence it is 
 
 correct to refer it to space and time as perceptions, and not to 
 
 the conception of the understanding." ^ 
 
 and the The Unity of time and space as pure perceptions, as well 
 
 perceptions so ' 
 
 Sot^Tknow-^^ °^ ^^^ empirical perceptions as such, must, therefore, be 
 Srttorcon* accounted for by a direct action of understanding upon sense, 
 ofunderstand- which is prior to any application of conceptions, even of the 
 a priori conceptions, to these perceptions., But this is not all 
 This direct action of understanding upon sense, which produces 
 the images of perception, is an activity that transforms the 
 sensations, the proper data of sense upon which it acts, into per^ 
 ; ceptions. But it is a ' blind' or unconscious activity : and it does 
 I not, therefore, give rise to any intelligent consciousness of things. 
 It sets appearances before us, but does not enable us to refer them' ■ 
 to objects as such. We talk of ' objects of perception,' but per- 
 ception, as Kant says, in itself is ' blind,' and its images do not 
 for it represent anything. Apart from conception, it would be 
 " for us as thinking beings as good as nothing." What it would 
 set before us would be at best a mfere " Gewiihl " or chaos of 
 appearances, flitting before the mind's eye, but without our 
 being able to fix them to anything, a " blind play of images, i.e., 
 less than a dream." The same idea is expressed with great 
 emphasis in a letter of Kant to Herz (dated 26th May, 1789), 
 in which he tries to answer some of the objections of Maimon. 
 Without the application of the pure conceptions of the under- 
 standing, he there declares, " the data of sense would never set 
 any object before me, nay they would not enable me even to 
 attain that unity of consciousness- which is required for the 
 consciousness of myself, as an object of inner sense. I cquld < 
 not be capable of knowing even that these data of sense are 
 presented to me, and consequently for me as an intellioent 
 
 }B. 161 note. 
 
CHAP. II. THE AESTHETIC. . 313 
 
 being they would be absolutely nothing at all. It is true that, 
 if in thought I make myself into an animal, I can conceive • 
 sensible ideas to carry on their regular play in my soul, seeing 
 that they might still be bound together according to an empiri- 
 cal law of association and so have influence upon feeling or 
 desire. This I can conceive, if I suppose myself to be conscious 
 of every single idea, but not conscious, by means of the synthetic 
 unity of apperception, of the relation of those isolated ideas to * 
 the unity of the conception of their object : but then I should 
 not through these ideas have knowledge of anything, even of 
 that state of myself which the ideas imply." ' 
 
 The passage I have iust quoted introduces some conceptions conception 
 
 ■*■ ^ " ■*■ -*■ necessary for 
 
 which we are not yet in a position to appreciate fully. But knowledge, 
 we can see the general idea that underlies it. It is that mere 
 sense, or even mere perception, can give rise to no consciousness 
 •of an object as such, any more than it can give us a conscious- 
 ness of self as such. A mere animal, (if, without dogmatising 
 as to what kind of consciousness animals in fact possess, we 
 understand by an animal a purely sentient being,) cannot have 
 anything like what we call knowledge. Its inner life is a 
 series of fleeting states, which it does not combine according to 
 universal principles, and which, therefore, it does not refer to 
 permanent objects distinct from the passing states through 
 which they are known. And for the same reason, it cannot be 
 conscious of itself in opposition, yet in relation, to those 
 objects. It is true, indeed, as Kant admits, that the sensuous 
 consciousness of the animal is a unity, though it is not a unity 
 for itself. Its passing states may leave images behind them, 
 which are recalled by new sensations and recall each other 
 according to an empirical law of association, and which may 
 awaken appropriate impulses, as when a dog shrinks from the 
 stick that has beaten him ; but such a play of sensible ideas or 
 images will not. account for knowledge, which implies that the 
 different data of sense are distinguished yet bound together in 
 * iR. XI. 57; H. VIII. 714. 
 
lllg. 
 
 314 THE CEITIQUE OF PURE REASON. book i. 
 
 one consciousness and connected in definite ways as represent- 
 ing permanent objects. 
 Combination Jq y^^at has been iust said there are a number of ideas, 
 
 of the uncon- ^ 
 
 ofthTim'SSi^ which cannot yet be fuUy explained, but it may suffice to show 
 Mn^dous " that, as we pass from the Aesthetic to the Analytic, we come in 
 
 activity of the . , . „ 
 
 understand- sight of two Very important elements m Kant s view oi percep- 
 tion, of which in the Aesthetic there was scarcely any indication : ' 
 first, that in perception both pure and empirical, there is 
 involved an associative combination of the elementary data of 
 sense, which is to be traced to the imagination and through it 
 to the understanding ; and secondly, that, even after this action 
 of our mental spontaneity in perception has been recognised, 
 we have not yet fully accounted for the peculiarities of our 
 perception. For our perceptions are what they are, because of 
 a relation into which they are brought in our consciousness of 
 them, viz., the relation to the pure conceptions of our under- ; 
 standing. Apart from this relation they cannot give rise to a 
 consciousness of objects but merely to a series of images, which 
 pass before our mind's eye without being recognised as the 
 images of any objects, still less of an ordered world of objects. 
 The unconscious synthesis of imagination, by which the 
 appearances of sense are presented to us in an unbroken I 
 continuity of images in harmony with the forms of time and 
 space, if it explains how perceptions are made oub of sensations, 
 does not yet explain how that continuity should be broken by 
 the recognition of separate elements, which are distinguished 
 from each other and how at the same time these elements 
 should become connected together in certain definite ways. 
 Yet all this, in Kant's view, is involved in the reference of 
 these perceptions to objects. Hence the naive view of the 
 ordinary consciousness, according to which individual objects as 
 such are given to us in sense in all their complete determiua-a 
 tion as individuals, is once for all set aside. For even the 
 images of perception are not given to us apart from the activity ?, 
 of the. mind to which they are given; and, if they were, they | 
 
CHAP. II. THE AESTHETIC. 315 
 
 could not be recognised as images of objects, without a further 
 mental activity. If we reduce sense to what really belongs to 
 it, all that is left would be a series of transitory sensations 
 without relation to each other. Thus even for an associative 
 principle that could create a mere perceptive consciousness, we 
 require the unconscious working of a principle of unity 
 kindred in nature to that which is manifested in the self- 
 conscious activity of the understanding ; and in order that we 
 may be able to make any use of the perception, so as to know 
 an object by it, we reqxiire that conscious activity itself. 
 
 As we enter upon the study of the Analytic, therefore, we ^Sd'^^aSf "^ 
 have to strip perception of its borrowed attributes, and to 0™™!™!°!^^ 
 reduce it to an element in knowledge ; and we have to refer to synthesis 
 
 which never 
 
 the activity of thought much that we have hitherto regarded °™ ^f^ted 
 simply as given in sense. "We have to make a fresh analysis 
 of a result which hitherto has been represented as if it were 
 given once for all in its completed form. We have to see in 
 the making what in the Aesthetic we took for granted as made. 
 Eeducing the data of sense to a mere manifold, in which no 
 principle of unity is working, we have to ask what more is 
 wanted for a connected inner and outer experience, i.e., for the 
 consciousness of a world of objects in space and time, all 
 standing in relation to one self, which at the same time is 
 conscious of itself as one object among others. In other words, 
 we have to ask how a sentient subject, who is also intelligent, 
 must determine his feelings by thought, ere he can represent 
 himself as one individual in a world of individuals, all of which 
 are included in one space and one time, and have their co- 
 existent and successive phases definitely determined in relation 
 to each other. In the Aesthetic all this was supposed to be 
 given in sense. Perception was there regarded as setting us 
 face to face with the individual in his complete, not to say 
 infinite,. determination, which no conception could ever exhaust. 
 Now, we have to recognise that an individual object can be 
 det»rmined as such only by a Synthetic activity of tha under- 
 
316 THE CRITIQUE OF PURE REASON. ^OOK I. 
 
 standing in which the manifold of sense is brought to unity, 
 and that the infinite determination of the individual merely 
 means that this activity manifests itself in a continuous process; 
 of which we know that it can never be completed. We have 
 to realise, in fact, that experience is a process of continuous 
 synthesis which is stimulated and guided by an idea that can 
 never be completely realised. The beginning of this, process is 
 made, whenever different elements, given in sense are combined 
 in the unity of a conception and so referred to an object.^ 
 The end of it would be the determination of that object in re- 
 lation to an absolute whole beyond which no data can be given 
 for synthesis. In other words, the object in being determined 
 as such becomes the pundum stans of a progressive determina- 
 tion, which can find no limit unless we are able to connect it at 
 the other end with the absolute boundary of the universe. This 
 absolute goal we assume as a reality which exists objectively, 
 independent of the process of our knowledge, i.e., we assume 
 that the individual object is determined in relation to the 
 absolute whole of things ; and that we, in progressively deter- 
 mining it, are merely tracing out lines of connexion and 
 determination, which exist independently of the knowing sub- 
 ject and of all the activity of perception and thought whereby 
 he acquires a consciousness of it. Yet, on the other hand, 
 Kant points out that this idea by which we are guided comes 
 into contradiction with the very nature or essential conception 
 of the synthetic process by which we seek to realise it. For, 
 in the first place, in so far as the process is a process of 
 synthesis of a given manifold, ■ the mind that performs it can 
 never find in itself a ground' for the conviction that the 
 
 ^ Really, as we have seen and shall see more fully in the sequel, this 
 first process is more complex than is indicated in the words used above ; for 
 the synthesis of the understanding implies first the distinction of the elements 
 of perception (which have been combined in the blind synthesis of imagina- 
 tion), and then the recognition of their connexion as corresponding to a pure 
 conception. We have, therefore, to think of an unconscious associative process 
 of imagination as continually anticipating and forming the basis for the con- 
 scious process of knowledge. 
 
CHAP. II. THE AESTHETIC. 317 
 
 synthesis is completed : and, in the second place, if we look to 
 the forms under which the manifold is given, it is manifest 
 that they preclude the very idea of completion : for. as we 
 cannot find a limit tp extension in space or to duration in 
 time, so we cannot exhaust the possible determinations of a 
 world in space and time. It is a world which by the nature 
 of the case is without limit or boundary, which it would be 
 a contradiction to suppose limited or bounded. It appears, 
 therefore, that we have to regard experience or knowledge as 
 a process towards a goal which yet by the nature of the case 
 it can never reach ; so that the very idea of an objectively com- 
 pleted world, which in the process of knowledge we seek to 
 bring into consciousness, comes directly into collision with the 
 fact that we are conscious of that world as being in space and 
 . time, and. as therefore incapable pf being ever completely deter- 
 mined. Experience is thus an endless process, which yet 
 necessarily is assumed to have a fixed and definite end. Or it 
 is an endless effort to get beyond the subject to an object 
 which is supposed to be determined in itself; while yet the 
 very nature of the process shows that we cannot by means of 
 it reach that which is determined in itself, i.e., that we cannot 
 get beyond the subjective. 
 
 These considerations necessarily give rise to a new concep- tim ofThe^ 
 tion of the opposition between the phenomenon and the thing phenomeno^^ 
 
 . ■,, ftn p 1 .1 1*° ^^ thing in 
 
 m itself, a conception different from, or at least going beyond, itseu. 
 that in which the Aesthetic rested. There the thing in itself 
 was that object which was conceived as producing an affection 
 in the sensibility. This affection, with the aid of the forms of 
 sensibility, was further represented as directly giving rise to the 
 perception of an object, which, of course, could not be known 
 to agree in any characteristic with the object that affected the 
 sensibility. It might, however, be asked, — and from mere 
 perception taken by itself, and viewed as Kant views it in. the 
 Aesthetic, no answer to the question could be derived, — whence 
 coiiles this consciousness of an object different from the 
 
318 THE CRITIQUE OP PURE REASON. book i. 
 
 phenomenal object ;' or, what is the same thing, whence corned 
 this consciousness of sense as having an object, so to 
 speak, .thrust upon it from without ? For a passivity can be 
 known only in contrast with an activity, and in the Aesthetic 
 no activity is mentioned. The answer comes in the Analytic, 
 where we learn that the perception of an object does involve an 
 activity of synthesis, which, however, has reference to, or is 
 exerted upon, data given independently of it. Hence, as the 
 mind becomes conscious of its own activity, it sets up a goal 
 for itself in conformity with that activity, and by the deter- 
 mination of the matter of sense under its forms, it seeks to 
 reach that goal. But the effect of this effort is to bring into 
 prominence a collision or opposition between the end and the 
 means which the mind has to use in realizing it ; between the 
 combining activity of thought and the material with which 
 that activity has to work. The goal set by the mind before 
 itself, therefore, appears as a noumenon or thing in itself, which the 
 mind cannot grasp, because it is not purely active but has to 
 receive its matter as given. And the affection of the subjec- 
 tivity, which is thus the presupposition of its activity, must of 
 course, be referred to the noumenon or thing in itself, which 
 is the only thing that the mind is conscious of beyond its own 
 process in experience, since it is the goal which the mind sets 
 up for that process. In this way we can see how the Analytic 
 at once justifies the presupposition of a thing in itself made in 
 the Aesthetic and carries us beyond it. For it gives to the 
 thing in itself the character of a noumenon or ideal set up by 
 knowledge for itself, and just for that reason represents it as 
 the ultimate reality to which we must refer the affections, i.e., 
 the data on which the mind's activity is exerted, 
 summaryview The truth of this vicw wc do not vet seek to discuss. 
 
 of the pro- ^ 
 
 S^&and Indeed, the full discussion of it is not possible, until we have 
 e laectK. ^.q^^i^^-^q^ ho^N, in Kant's view, the character of th»»^ 
 activity of the pure intelligence limits it to synthesis of a 
 given matter, and at the same time makes it impossible for it 
 
CHAP. II. THE AESTHETIC, 319 
 
 to realise its ideal in a matter given under the forms of time 
 and space. At the point we have reached it is sufficient to 
 call attention to the way in which the problems of the 
 Analytic and the Dialectic open before us. Perhaps one of the 
 greatest hindrances which Kant has put in our way in compre- 
 hending the Critique, is the formalistic way in which he has 
 separated these two parts of it, setting understanding and 
 reason before us as two separate ' faculties,' without casting 
 sufficient light upon the identity of the intelligence in these 
 two different aspects of it. The preliminary view we have 
 now taken may be sufficient to indicate how Kant's reconsidera- 
 tion of the results of the Aesthetic, or rather of the Dissertation, 
 led him to recognize the necessity of dealing with the double 
 problem as to the process and the ideal of knowledge, and to 
 give a new interpretation of the contrasts between perception 
 and conception, and between the phenomenon and thing in 
 itself. In the Analytic and Dialectic respectively, our eyes 
 are directed first to one and then to the other of the terms 
 in these contrasts : in the Analytic to the process of experience 
 and the phenomenal object, in the Dialectic to the ideal of 
 experience and the noumenal object. But we cannot really 
 understand either, unless we keep in view that any result 
 reached by looking at one of the two complemeritary aspects 
 of the systeni is provisional, and that the bearing of the Critique 
 cannot be seen until we are able to combine them in one final 
 view. 
 
 With this caution we must now proceed to the exposition 
 and criticism of the Analytic. 
 
320 
 
 CHAPTEE III. 
 
 THE TWO LOGICS, AND THE DISCOVERY OK METAPHYSICAL 
 DEDUCTION OF THE CATEGOEIES. 
 
 The Aesthetic f\JJJi analvsis of the Aesthetic has shown that it is the first 
 
 is the first step II *' 
 
 SflSii)^ s*^P i"^ ''^** transformation of the ordinary consciousness 
 
 by going back upon its presuppositions, which is the essential 
 work of philosophical criticism. Taking the common view of 
 perception as the immediate consciousness of an individual 
 object through sense, the Aesthetic showed that such an object 
 always presents itself as a special determination of the pre- 
 supposed individual totality of space or time, and that only on 
 this theory is it possible to explain the way ia which mathe- 
 matical science anticipates sensible experience. The constructive 
 synthesis of mathematics would not be possible, if space and 
 tirne were not the a priori forms of perception. 
 ibe Analytic gut wheu wc pass ou to the Analytic, we find that this first 
 vohreanew stagc of Critical reflexion itself starts from presuppositions ■ 
 tL°pr™up?°° which are not strictly accurate. Por no totality, no individual 
 
 positions of 
 
 the Aesthetic, wholc Can cxist for US cxcept through a synthetic activity, 
 which brings together the data of sense in relation to the 
 unity of the conscious subject. If we abstract from such a 
 combining activity, what remains is merely a ' manifold,' i.e., a 
 'many,' which is not yet a 'many-in-one.' Further, even if 
 we attribute to sense, as immediately determined by the unity 
 of the conscious subject, a synthesis of the manifold which is 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. ' 321 
 
 sufficient to turn sensation into perception, we are obliged 
 to regard this synthesis as implicit and unconscious, i.e., as 
 a synthesis of elements which are not definitely -distinguished 
 and related ; and the unity of such perception still needs the 
 understanding to make it explicit. We are, therefore, obliged 
 to recognise that " perception without conception is blind," and 
 that it requires the activity of the understanding " to bring it 
 to conceptions,'"' even if it be also true that "conception 
 without perception is empty," i.e., that it reduces itself to 
 the barren unity of analytic thought, which has no relation 
 to any difference or ' manifold.' It is, therefore, the business 
 of the Analytic to show that the unity of apperceptien is 
 necessary to experience as well as the unity of perception; 
 and to explain how, in relation to the manifold, the former 
 unity gives rise to the principles of a priori synthesis which 
 are required for the determination of the objects of experience 
 as such. Lastly, it is the business of the Dialectic to show 
 how the same unity is also the source of a still higher kind of 
 principles, by which the determination of objects in experience 
 is connected with that idea of absolute totality which in the 
 Aesthetic had been " represented as given " in the case of time 
 and space. 
 
 Again, both in the Analytic and in the Dialectic we find that ^ya™|Jtnd 
 Kant divides each of his problems into two subordinate tei"deduoMons 
 
 , . of the categor- 
 
 mqumes. In the first place, he endeavours to analyse the ies and ideas, 
 understanding and the reason, and to discover the implements, 
 i.e., the forms of synthesis, with which they are armed for their 
 work : viz., in the one case, the. pure conceptions or categories 
 which the understanding uses to determine the objects of 
 experience as such, and in the other cape, the Ideas by means 
 of which the reason seeks to carry that synthesis on to the 
 unconditioned and to connect empirical knowledge with the 
 consciousness of things in themselves. Then, in the second 
 place, he seeks to justify both conceptions and ideas so far as 
 they.are capable of justification ; in other words, to show in 
 
 VOL. I. X 
 
322 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 the case of the pure conceptions that they are just those categories 
 which are required for, and presupposed in, our actual know- 
 ledge or experience ; and in the case of the Ideas, that, though 
 they are not capable of being used in the determination of 
 things in themselves, yet they have a regulative use in relation 
 to empirical- science, as setting before it the ideal aim after 
 which it must strive and by which its progress is guided. In 
 other words, Kant seeks to supply both a metwpTvysical and a 
 iranscendental deduction of his categories and ideas, as well as of 
 the a priori principles of understanding and reason which are 
 based on them. In this chapter we have to consider how 
 Kanf performed the former of these tasks, in relation to the 
 categories of the understanding, though it may also be neces- 
 sary to say a word about the ideas of reason in order to bring 
 out the unity of Kant's method in both cases, 
 formal iT^o ^^ ^ provious chapter^ we have referred to Kant's account 
 caiTeduotion. of the Way in which he was led to connect his lists of cate- 
 gories and of ideas with that analysis of the process of thought 
 which is supplied by formal Logic. Universalising the problem 
 suggested by Hume in relation to the objective value of the 
 I principle of causality, Kant asked how we are authorised to 
 say that an objective value attaches to any of the a priori 
 conceptions; and this again led him to ask what a priori 
 conceptions there are. In seeking to answer this question, 
 Kant found that the list of Aristotle's categories would not 
 serve his purpose : for it seemed to be constructed on no 
 definite principle, but merely by taking up any a priori con- 
 ception which happened to present itself in a general review 
 of experience. Besides, it contained several ideas which Kant 
 had alre^idy discovered to be determinations of the forms of; 
 sense. In this difficulty it suggested itself to Kant that formall 
 Logic had already analysed the process by which the mind | 
 manipulates the content of the ideas which it already possesses, 
 whatever be the source, a priori or a posteriori, from which it ; 
 ' See above, p. 207 seq. 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 323 
 
 has derived them ; and, though this is a different thing from 
 the process by which the mind goes beyond itself to apprehend 
 objects or to add new elements to those already combmed 
 in its ideas of objects, yet both processes belong to the 
 same mind, and may be expected to have an intimate relation 
 to each other. There will no doubt be an essential difference 
 in the two processes, if, and in so far as, in the latter 
 there is something which depends, not on the activity of 
 the subjective intelligence, but on the object or on the way in 
 which the subject is 'acted on by it (i.e., upon the affections of 
 sense, and the nature of the sense affected). But, in so far as 
 in both cases we have manifestations of the activity of the 
 intelligence itself, they will be closely akin to each o'ther ; and 
 what is ascertained of the one process may be expected to afford 
 a clue to discover what is true of the other. In other words, ' 
 the real process of intelligence, so far as it is a priori, may, 
 be expected to show characteristics identical with, or closely 
 analogous to, those of its formal process. Now, the character- i 
 istics of the latter process had been, as Kant held, sufBiciently 
 explained by formal Logic : a science which was all but com- 
 pleted at a stroke by Aristotle, and inherited from him without 
 substantial change by his successors. Hence this formal Logic 
 was for him a safe and secure point of departure for the 
 discovery of the transcendental Logic ; and he held that every- 
 one of the elements of the former might be treated as a 
 guiding thread in the search for some corresponding element 
 m the latter. 
 
 The process of thought as analysed by formal Logic is The fm-mai 
 
 processes of 
 
 divided into the three subordinate processes of conception, judg- ^3"^!™-°^^°^ 
 ment and reasoning, and each of these has its special character- ^'^onTaie 
 istics. A conception is defined in contrast with a perception ""* ^^^ ' 
 as a general idea, which, as general, is not in immediate relation 
 to an object, but constitutes a unity under which the percep- 
 tions of objects may be brought, either directly or indirectly, 
 through other conceptions less general than itself (the extension 
 
324 THE CRITIQUE OF PURE REASON. book i. 
 
 and comprehension of a conception being in inverse ratio to 
 
 each other). Judgment, again, is the process in which this 
 
 relation between a perception and a conception, or between two 
 
 conceptions, is established ; hence we may say that the one use 
 
 of conceptions is to make judgment possible, i.e., to subsume 
 
 under them perceptions or less general conceptions. Finally, 
 
 the reasoning or syllogistic process is a process in which the 
 
 mediation between the terms of the judgment is carried to the 
 
 furthest point, reaching its ultimate goal when no further 
 
 middle term can be found between the subject and the 
 
 predicate. 
 
 Theiv differ- Now in all this, the intelligence proceedsanEdytically; jn_Qther 
 
 proce^^sof words, it starts with an idea, i.e., a conception. Qrjtfirfiertinn_o£jan 
 
 jyuo^m *""* object, which it presupposes as given without .asking. how^,.is 
 
 based on a givcnj and it is by abstraction from that idea that the inteUi- 
 
 prvyri concep- 
 
 Ss'^"'* gence gets the conception which in the judgment it attaches 
 to it as a predicate, and also the conception which in the 
 syllogism it uses as a middle term to combine the subject and 
 predicate of the conclusion. When, however, we begin to 
 consider how this will apply to a kind of judgment and 
 syllogism in which the intelligence goes beyond itself to 
 determine an object, or goes beyond its conception of an object 
 to add new elements thereto, we find that something more is 
 necessary. Here we have to do not with the analysis of a com- 
 plex conception of an object which is already formed, but with 
 the first formation of the conception of the object, or with the 
 addition of new elements to it. At the point we have now 
 reached, the question how this is possible can no longer be 
 answered in the simple way which was considered sufficieiit 
 in the Aesthetic, i.e., by saying that it is in perception, pure 
 or empirical, that the object is presented to the mind, and 
 that it is from perception also that all new elements must be 
 drawn which are to be added to the conception of it. For an 
 object as such is now seen to be a 'many-in-one,' the conscious- 
 ness of which cannot be accounted for by the impressions of 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 325 
 
 sense taken singly, but only by their being combined on some 
 principle of union. And in like manner, the addition of new 
 elements to the conception of an object implies, not only that 
 these elements are given in sense, but also that the intelligence 
 is able to bind them up with the others already combined ; and 
 for this also there is required some principle of union. In 
 this case, the activity of thought cannot extract from the 
 objects with which it deals the more general conceptions under 
 which it subsumes them ; but it must bring these conceptions 
 with it to the ' given data ' in order to combine them in the 
 consciousness of an object, or to add new elements to the con- 
 ception of that object. .Further, as in analysis we cannot stop 
 till we find the simplest and most general idea under which a 
 class of objects can be brought, so in synthesis we cannot stop 
 till we find some first principle or all-embracing idea beyond 
 which our intelligence can seek for nothing further. Here, 
 therefore, for the analytic processes of judgment and reasoning, 
 in which the mind derives all the data it needs from the 
 conceptions with which it deals, we must substitute synthetic 
 processes, in which the intelligence must itself supply all the 
 principles of unity by means of which it determines its objects. 
 And, as' in the former case, we have in judgment and reasoning 
 analysis upon analysis up to the point where no further 
 analysis is possible, so here we shall have synthesis upon 
 synthesis, up to a first principle in which reason is finally 
 satisfied. But the synthetic judgment is impossible without 
 a priori conceptions to establish unity in the given manifold of 
 perception, and the synthetic syllogism is impossible without 
 a priori ideas to which we may carry back the entire synthetic 
 process of judgment, and by means of which we may give it 
 perfect completeness and unity. Or, to put the same thing in 
 another aspect, we require an a priori principle of understanding 
 to combine the elements of the manifold of perception in rela- 
 tion to the unity of the conscious subject ; while we require an 
 a p^ori principle of reason to enable us finally to complete the 
 
326 THE CRITIQUE 01" PURE REASON. BOOK I. 
 
 synthesis of objects, and so to overcome the division between 
 object and subject, and reach the determination of the thing in 
 itself. In the former case, therefore, the understanding, and in 
 the latter case the reason, has to supply the a priori principles 
 which are required for knowledge, and which are not to be got 
 by analysis from the data to which they are to be applied. 
 DifEerence According to this view, then, there are real functions of 
 
 between the ° ' 
 
 th?u?d«°^ °* understanding and reason in relation to knowledge, which 
 the^ide^of Correspond exactly with their formal functions in relation to 
 
 reason. , . i i 
 
 thought ; and both understanding and reason bring with them 
 a priori principles for the determination of objects. But there 
 is a very important difference. For the a priori principles of 
 understanding are necessary in order that anything may be 
 determined as an object at all ; and they must find application, 
 if any data of sense are given in such a way that they 
 can be brought in relation to consciousness : whereas the a 
 priori principles of reason are not necessary for the determina- 
 tion of objects in relation to the conscious subject, but only for 
 the absolute completion of such determination. The latter, 
 therefore, can find application, only if the manifold for synthesis 
 can be exhausted, and the synthetic process of the under- 
 standing brought to a conclusion. They are in fact principles, 
 the use of which can only be to bring the synthesis of the 
 understanding to a final unity. But, if it be the case that the 
 forms under which the manifold is given for synthesis, i.e., 
 time and space, are such as to preclude the completion of the 
 synthesis of the understanding, then the a priori principles of 
 reason can have no objective application. They can only 
 represent an ideal of reason which can never become actual. 
 On the other hand, the pure principles of the understanding, 
 i.e., the principles which are involved in the activity of 
 judgment, must needs, as we have said, find application, if there 
 is to be such a thing as knowledge ; for only through the 
 application of these principles can the manifold become united 
 in the conception of an object. We may therefore see, even 
 
CHAP. III. THB METAPHYSICAL DEDUCTION. 327 
 
 from this preliminary view of the matter, that, on Kant's 
 principles, the a, priori synthesis of understanding must be 
 possible, while the a priori synthesis of reason must be im- 
 possible ; or, what is the same thing, that the a priori 
 conceptions of understanding must be objectively valid in so 
 far as the manifold to which they have to be applied is given, 
 while the a priori ideas of reason cannot be valid, because the 
 synthesis of understanding to which they have to be applied, 
 cannot be completed. 
 
 Deferring, in the meantime, the consideration of the a priori wtat is 
 
 DecesBary f or 
 
 principles of reason, let us look more closely into Kant's ?^*^gnt, 
 " metaphysical deduction," i.e., his discovery of the categories i™ a^T^"*" 
 of the understanding ; — to which he was led by a consideration of synthesis. 
 the forms of judgment. The principle upon which he goes is, 
 as we have seen, that the analytic judgment, — in which our 
 understanding subsumes the perceptions or conceptions it 
 already possesses under higher conceptions got by analysis 
 from them, — has a close correspondence with the synthetic 
 judgment, in which the understanding goes beyond itself and 
 what it already knows in order to determine objects, (or to 
 determine them further than they are already determined for 
 us,) by bringing the manifold of perceptions under conceptions 
 derived from itself. There is, however, an important difference 
 in the two cases compared : for, in the former case, a subject is 
 brought under predicates that already form part of the idea of 
 that subject, while in the latter case, what is given in sense 
 is only a manifold, i.e., a number of elem'ents unconnected 
 with each other ; and it would seem as if these could become 
 connected together, so as to form a proper logical subject only 
 after the predicate had been applied to them. In other words, 
 it would seem as if in this case the predicate were needed to 
 give to the subject that unity in virtue of which alone it can 
 be an object of thought, or a subject to which any predicate 
 can be attributed. For the manifold must be combined with 
 thotight, if it is to be combined into an object for thought. 
 
328 THE CRITIQUE OP PURE REASON. 
 
 BOOK I. 
 
 Kant tries to get over this difficulty by distinguishing percep- 
 tion and conception as two steps in knowledge, in the first of 
 which we have synthesis, while in the second we have the 
 consciousness of the principle of the synthesis. Of the former 
 he says that it is the " work of imagination, a blind but neces- 
 sary function of the soul without which we should have no 
 knowledge, though we are seldom even conscious of it." But 
 though this synthesis is necessary for knowledge, it is not 
 sufficient for it. In addition to it there is required "for 
 knowledge, in the proper sense of the word," an activity of the 
 understanding, the business of which is to " bring the imagina- 
 tive synthesis to conceptions,'' i.e., to detect the principles of 
 unity implied in such synthesis, or to find some principle 
 under which it may be brought.^ Now, when we go back to 
 the beginning, the principles of unity, by which the manifold 
 is combined and is determined as a definite object for thought, 
 must be found in the unity which thought has with itself, and 
 which is expressed in the analytic judgment. Kant holds, 
 therefore, that the sy nthetic unity of an object, as a definite 
 object of perception which can be made the subje ct of analysis, 
 is due to the same principles which govern such analysis. 
 It is onIy~tEar the use to which they are put is somewhat 
 different. Thus the logical principles involved in judgment 
 become a guide to us in " bringing the primitive synthesis of 
 imagination to conceptions." In a pregnant passage, which 
 almost breaks down with the weight of meaning which he 
 makes it carry, Kant declares, that " the same function, which 
 gives unity to the different ideas in a judgment, gives unity 
 also to the bare synthesis of different ideas in a perception : 
 and this unity, when we express it in its generality, is the 
 conception of the understanding. Hence, the same understand-, 
 ing, by the same acts which enabled it, through the analytic 
 unity of thought with itself, to give to its conceptions the 
 
 1 In the next chapter we shall consider more particularly which of these 
 ways of expression is the more exact statement oi the truth. 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 329 
 
 logical form of the judgment, is enabled also, by means of 
 the synthetic unity of thought with itself in apprehending the 
 manifold of perception, to introduce into its ideas a transcen- 
 dental content. These acts, therefore, in the latter application, 
 are entitled the a priori conceptions of the understanding, and 
 they refer a priori to objects, a kind of reference which is 
 altogether beyond the scope of general Logic." ^ 
 
 The passage, from which the foregoing quotations are taken, 5,^^^g"°o('"' 
 contains some ideas which must wait for explanation till we ™d^nde^" 
 reach the Transcendental Deduction of the Categories. For the judgment, 
 the present purpose, it may be sufficient to exhibit its general 
 drift. Sense, whether pure or empirical, yields only a mani-l 
 fold, which requires to be brought to a unity ere by its means 
 we can have a perception of any object. Hence, a synthetic 
 judgment by which the manifold of perception is put together 
 in the unity of an object must precede all analytic judgments. 
 It might, indeed, seem that in the primitive synthesis judg- 
 ment was not required ; for synthesis, as a mere putting 
 together of the elements of perception, can be explained by 
 the imagination as a " blind " and unconscious manifestation of 
 the unity of thought. This at least might seem sufficient to 
 explain why we have distinct images of perception, in which 
 the manifold of sense is brought together into one idea which 
 may be made the subject of a judgment. But it is not so ; 
 for the consciousness of such a unity, as an object of thought 
 which may be made a subject of predication, is attained only 
 when the judgment is actually made; and the perception is 
 thus at once distinguished from, and referred to, a conception 
 which expresses the principle of its unity (or under which 
 its unity is brought). Such a judgment may be called analytic, 
 since what it expresses as the predicate is already involved in 
 the subject. But so to call it, would be to forget that it is as 
 distinguished from and related to the predicate that the subject 
 gets its determination as subject. In other words, the judgr 
 lA. 79; B. 105. 
 
330 THE CRITIQUE OF PURE EEASOK. book i. 
 
 ment appears to presuppose a movement of thought which is 
 only accomplished in it: and we may say from one point of/ 
 view, that the primitive judgment determines the perceived | 
 object synthetically in relation to a conception which is notj 
 contained in the idea of it as perceived ; while from another point j 
 of view we may say that the object as perceived is already deter- 
 mined by the conception and that the primitive judgment only 
 analyses it. In the passage which we are considering, Kant 
 uses language which is not quite distinct, and we shall have to 
 consider in the next chapter whether the difficulty suggested is 
 aT«iattono£ ^^^^ fairly met by him. For the present it is sufficient to 
 the^nttiedc" poiut out that the analytical process of judgment, ia which a 
 more general conception is abstracted from a less general and 
 then used as a predicate to determine it, is conceived by Kant 
 as analogous to, and in principle identical with, the synthetical 
 process, by which a conception derived from the understanding • 
 is used to determine a perception derived from sense, and so to 
 give objective value to the imaginative synthesis of the mani- 
 fold. In the former case, we have thought abiding with itself 
 and its possessions, and, so to speak, reasserting its own unity ; 
 with itself in the apprehension of all their differences ; in the 
 latter case, we have thought going out of itself to determine 
 by its own unity that which has not previously been brought 
 within its reach. But the processes are, as Kant thinks, 
 fundamentally identical. Hence, the account given by formal \ 
 Logic of the pure unity of thought with itself, which is ex- 
 hibited in the analytic judgment, may be used as a guide to ' 
 discover the categories by which perceptions are referred to 
 objects, or by which the synthesis of the manifold of percep- 
 tion made by imagination is so " brought to conceptions " that 
 it can be recognised as objective. 
 SS-amyofthe There are, however, some ambiguities in Kant's statement 
 the°unde™° ° which may require a word of explanation. Thus it is confusing 
 
 standing. , 
 
 that Kant at first speaks of the pure conception of the under- 
 standing in the singular, while at the end of the passage he 
 
CHAP. ra. THE METAPHYSICAL DEDUCTION. 331 
 
 uses the plural. This, however, we may account for by the 
 fact that what he is dealing with is the method in which the 
 pure unity of thought differentiates itself in relation to the 
 manifold of perception. The categories are " species of the 
 unity of apperception," which disappear in that unity when we 
 regard them only in relation to it, while they appear as a 
 number of separate conceptions when we view them in relation 
 to the manifold. Hence, in the former reference, it would be 
 natural to speak of the conception, rather than of the concep- 
 tions, qf the understanding ; only we must remember that, in 
 Kant's view, the pure unity reveals itself in various functions 
 of unity even in the pure analytic judgment, which for that 
 reason is used as a guide to the discovery of the categories. 
 
 Another and more important difficulty is to explain how toSiTuJe** 
 
 ■ 1 .. ■, . , . -T -, j_ j» j_i Kant makes 
 
 those pure conceptions, which primarily are only aspects oi the of the analytic 
 
 judgment. 
 
 unity of thought with itself independent of any matter, can 
 possibly yield principles of unity to determine the synthesis of 
 the manifold of sense. This difficulty cannot in the meantime 
 be fully considered. I have, however, already remarked that 
 it is one of the peculiarities of the synthetic act of judgment, 
 which Kant is describing, that it is only in reference to the 
 predicates, under which the understanding brings the manifold 
 of sense, that that manifold can itself be determined as a 
 subject. And now I may add that, according to Kant's 
 doctrine, it is only in reference to the manifold that the pure 
 function of unity gets its character as a predicate, under which 
 the manifold can be brought. There is thus a reciprocal pre- 
 supposition which it would be impossible to understand, if we 
 were 'here dealing with a process in which one step was com- 
 pleted before the other began, and not with the analysis of an 
 organic unity in which each element implies all the rest, and 
 can be separated from them only by abstraction. Kant, indeed, 
 holds that the pure process of thought can be separated, and 
 that it is separated by formal Logic, from any matter. But he 
 is otdkiged to recognise immediately that in this separation it is 
 
332 THE CRITIQUE OF PURE REASON. ^OOE I. 
 
 not what it is in the unity. And it is not difficult to show that 
 when we carry out such a separation to its necessary result, the 
 analytic unity itself disappears in identity. Kant himself 
 practically confesses as much when he says that analysis pre- 
 supposes synthesis, and the analytic unity the synthetic unity. 
 The main difficulty of this part of Kant's work, in fact, is just 
 this, that he thrusts in the pure analytic judgment as a middle 
 term between the unity of apperception and the categories, 
 and does not discern that his own reasoning is entirely fatal 
 to its existence. This, however, wiU be more clearly seen, 
 if we follow Kant closely in the process whereby he makes 
 the former a guiding thread to the latter. 
 ■thTtaMcfn*^ Kant, then, attempts to discover the pure a priori concep- 
 fh?t'i?S! °* tions by reference to the fourfold classification of judgments 
 according to Quantity, Quality, Eelation, and Modality, with 
 their subordinate modifications. Taking this classification as 
 an ascertained result of the Logic which analyses the formal 
 process of thought, he asks what will correspond to it in the real 
 process of knowledge, i.e., what is the " Transcendental System " 
 that corresponds to this " Logical System " ? Thus, if judgment 
 in analysing a conception must determine the Quantity of the \i, 
 subject in relation to the predicate, there must be some corre- < 
 spending " function of unity " by which the manifold of sense 
 is determined when it is brought together in the perception of 
 an object. And what that function must be, we may at once see 
 if we consider that as, according to the view of the formal 
 logicians, the analysis of ideas necessarily involves a reference 
 to Quantity, the idea of Quantity itself must be a principle on 
 which the manifold elements of the idea were put together. 
 We get, therefore, corresponding to the singular, particular, and 
 universal judgments, the categories or forms, of Unity, Plurality, ||| 
 and Totality. In like manner, if analysis necessarily has 
 respect to Quality, i.e., to the exclusion of the predicate from, 
 or inclusion of it in the subject, or finally to the inclusion of 
 the subject in a sphere defined only by the negation of the 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 333 
 
 predicate, synthesis must be negative, or positive, or limitative, 
 determination, which gives rise to the three categories of 
 EeaUty, Negation, and Limitation. Again, if analysis involves 
 that the ideas distinguished are at the same time related, — either 
 simply as a predicate to a subject in the categorical judgment, 
 or as reason and consequent in the hypothetical judgment, or 
 finally as genus to the species, which it includes but which 
 reciprocally exclude each other, in the disjunctive judgment, — 
 it follows, that in the synthesis by which the manifold is put 
 together in an objective consciousness there must be three 
 corresponding modes ; and these we may by anticipation recog- 
 nise as the relations of Substance and Accident, of Cause and 
 Effect, and of Eeciprocal Determination. Lastly, if the analytic 
 judgment implies that the predicate shall be conceived as 
 either assertorially, problematically, or apodictically connected 
 with the subject, it follows that in the synthesis whereby the 
 consciousness of an object is constituted, there is a principle of 
 determination of the manifold objects as possibly, actually, or 
 necessarily, united with the self-consciousness of the subject 
 for which they are. In short, as aU analytical judgments, by 
 which some element of an idea is separated from, and referred 
 as a predicate to, a subject, involve a determination of the subject 
 and predicate in relation to each other in Quantity, Quality, 
 Eelation, and Modality, according to one or other of three alter- 
 native forms of the categories, it would seem that the synthesis 
 by which the conception of an object is formed out of the mani- 
 fold, must involve a determining process according to corre- 
 sponding principles. For, as the separation of the elements of 
 an idea which the mind possesses only gives us a clear conscious- 
 ness of these elements as already determined in certain relations 
 to each other, it is obvious that in the process whereby the mind 
 first went beyond itself to form the idea of an object by uniting 
 the elements of the manifold together, there was involved the 
 determination of the manifold in all these various ways. Thus 
 the synthesis, by which the manifold must be brought together 
 
334 
 
 THE CRITIQUE OF PUEE REASON. 
 
 BOOK I. 
 
 Does it corre- 
 spond to 
 Kant's own 
 idea of a 
 metaptysij 
 deductioi 
 
 in order to be an object for the conscious subject, involves a 
 determination of the manifold by all these species of categories 
 according to one of the three forms of each species. 
 
 ?he first criticism to be made upon Kant's attempt thus to 
 Illicit the principles of determinant or synthetic judgment from ' 
 those of analytic or formal judgment, is that it does not quite 
 correspond to his own idea of a " metaphysical deduction " of 
 the categories. Objecting to Aristotle's list of categories as 
 empirically " picked up," he demands that the categories should 
 be evolved from " an idea of the whole of a priori knowledge," 
 which shall exhaustively determine all the parts and their 
 relations to each other. " Transcendental philosophy," he 
 says, " has the advantage that it can, but also the obligation 
 that it must, seek out its conceptions according to a principle ; 
 for its conceptions must spring pure and unmingled from the 
 understanding as an absolute unity, and, therefore, they must 
 be connected with each other according to one idea. But such 
 a connexion necessarily puts into our heads a rule, according 
 to which the place of each pure conception of the understand- 
 ing, and the completeness of the whole list, may be determined 
 a priori!' ^ Now, how does Kant realise this idea ? He 
 points out that the categories are forms of the a priori synthesis 
 by which objects are determined as such, and, as we shall see, 
 he carries them back to " pure apperception " as the unity out 
 of which they spring. But instead of showing directly how 
 they spring from that unity, he has taken the roundabout 
 'method of basing his list of the pure conceptions that rule the 
 synthetic judgment upon the aspects or modes of analytic 
 judgment, and he has simply adopted the list of these modes 
 from formal Logic. But, if he had realised his own ideal, he 
 would have been obliged, first of all, to show how it foUows 
 from the idea of the analytic judgment that the list should 
 contain just these and no other forms. And, even after he had 
 used the ' logical system ' so derived as a clue for the discovery 
 
 1 A. 67 ; B. 92. 
 
OHAP. III. THE METAPHYSICAL DEDUCTION. 335 
 
 ■of the categories, he would not have considered himself free from 
 theobligation of showing from the nature of the synthetic judgment 
 itself that they form a complete system of a priori conceptions. 
 
 Now, as to the first of these steps, the assumption that the Pure form, 
 
 taken 
 
 analytic judgment has all the forms mentioned and no others, ^^^^^ 
 or in other words, that these determinations belong to the form ippS^'the' 
 as distinguished from the matter of judgment, and therefore tity of subject 
 
 ° ^ and predicate. 
 
 fall within the scope of a Logic that deals with mere analysis, 
 Kant's trust in the finished work of the logician has obviously 
 misled him. For, if we strictly apply the idea of analysis, the 
 determinations of Quantity, Quality, Eolation, and Modality 
 must immediately disappear, like all other determinations, in the 
 bare identity of subject and predicate ; and with that the judg- 
 ment itself will lose all its meaning. To see this we need only 
 observe how the idea of limiting Logic to the form of thought 
 has worked itself out in the history of the science. The in- 
 evitable results of the search for bare form without matter has 
 been to eliminate one element after another, till the judgment 
 has disappeared in the expression of bare identity. First, 
 ModaUty was excluded, because analytic judgments are always 
 necessary, and with any other connexion of conceptions than 
 that which is indicated by analysis formal Logic has nothing 
 to do. In the next place, the various relations expressed in 
 the categorical, hypothetical, and disjunctive judgments were 
 reduced to the single relation of containing and contained 
 Quantity. Then Hamilton proposed that the predicate should 
 be quantified on the ground that what is " implicit in thought 
 should be made explicit in Logic " ; in other words, in order to 
 make manifest the identity, he would remove the quantity 
 from form into matter. The predicate being quantified and 
 the judgment thus reduced to an equation, the next step was 
 that the quantity of particular judgments should be made 
 definite by the introduction of number. Jevons went a step 
 further, and argued that if the predicate is to be quantified on 
 the ground that what is implicit in thought should be made 
 
336 THE CRITIQUE or PURE REASON. BOOK I. 
 
 explicit in Logic, then it should be also qualified by the sub- 
 ject, because in the judgment, it is implicitly limited thereby : 
 in other words, to make manifest the identity, Jevons removed 
 the quality from form into matter. Thus the formula of 
 judgment became, not A = B, or A = some B, but A = AB: not, 
 " All men are mortal," or, " All men are some mortals," but, 
 " AH men are mortal men." Thus there remains but a single 
 step to be taken to bring formal Logic to the euthanasia of 
 pure form : viz., that the subject also should be quahfied by 
 the predicate. The judgment would then take the form, 
 AB = AB, or, "Men mortal are mortal men," and would pass 
 into the tautological expression of an identity, 
 neve^naiytic. Ths truth, thcu, is that the elimination of matter from the 
 judgment is possible only when the form is reduced to a simple 
 identity. A tautological judgment is the only pure analytic 
 judgment, and a tautological judgment is, strictly speaking, no 
 judgment at all. It is a judgment that is never made except 
 as a logical exercise. Identity has no significance except in 
 relation to differences. Even where there is an appearance of 
 a simple identical judgment, there is always an implied nuame 
 of difference, left to be indicated by emphasis, between the '■ 
 subject and the predicate. " A man's a man for a' that " is 
 not a tautology. I n the purest analytic judgmen ts which w e 
 can make _^§,re_ is always a, synthesis of differgafie. Kant,, 
 indeed, says that the judgment, " All bodies are extended " is 
 analytic, because without this quality we cannot have the 
 conception of body at all. But no one would make that 
 judgment, unless there were other qualities or elements in the 
 conception of body with which extension was or might be 
 connected, or if it meant merely "The extended is the 
 extended." The separation of one quality from the complex 
 of qualities in an object is always made in view of establishing: 
 a new relation between the elements of that complex, or 
 between it and other objects.'' 
 
 1 We may express this otherwise by saying that the object of a judgment is 
 
CHAP, ni-^- THE METAPHYSICAL DEDUCTION. 337 
 
 Another consequence of what has been just said may be ^ uo°?'^f ° 
 mentioned, though it carries us beyond our immediate subject. ^y"°s'™- 
 If a merely analytic judgment is no judgment at all, then 
 a merely analytic syllogism is no reasoning at all. For 
 reasoning means the mediate combination of two ideas 
 which are not capable of being immediately united. But 
 in pure analysis there are no differences to dissolve, which are 
 not presupposed as already dissolved, or if there are, nothing 
 is done to dissolve them. A complex idea is supposed to be 
 present to the mind, but the whole analytic process does 
 nothing to bring its elements into closer relation either to 
 each other or to the mind. They are assumed to be already 
 united with each other, or, as far as they are not already 
 brought to unity, their reference to each other as subject and 
 predicate does nothing to help us to reunite them in a better 
 way : nor will any extension of the analytic process such as 
 gives rise to a " tree of Porphyry " and to a syllogistic process 
 of subsumption, make matters better. In like manner, the 
 idea which is to be subjected to the logical process is assumed 
 to be already united with thought, i.e., in Kant's language, with 
 the " I think " of Consciousness ; or, if its union is still 
 imperfect, the most thorough analysis cannot disclose any 
 link which will more perfectly combine the content with the 
 form of thought, the object with the mind which thinks it. 
 On this view of it the whole process of syllogistic reasoning 
 proves nothing, or proves only by a glaring petitio principii. 
 The mind simply revolves on itself, or does over again its own 
 finished work, and never can by any possibility gain one step 
 in advance. In this case, therefore, we have a reasoning that 
 contradicts the very idea of reasoning, just as in the former 
 case we had a judgment that contradicted the idea of 
 judgment. 
 
 to produce an identity, to reach & unity of conception. Judgment is concep- 
 tion in the making. Where the identity is reached, which it would be the 
 business of the analytic judgment to express, there is no longer any reason for 
 judging at all. (Cf. above, p. 269 seq.) 
 VOL. I. Y 
 
338 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 rfSSTaktog ^* appears, then, that formal Logic does not give us any 
 as"aM&Mof' account of a pure process of thought which might furnish a 
 clue for the discovery of the forms necessary to the a priori : 
 determination of objects. If its principle be carried out, it 
 empties the processes with which it deals of all the content 
 which they seem to have, and reduces their form to nothing:,: 
 indeed, it was only because that principle was not fully 
 developed in the " general Logic " of Kant, that it still 
 seemed to have a " form of thought " to speak about. In 
 other words, it was only in so far as formal Logic still 
 preserved a kind of ghost of that real process which Kant 
 was seeking to discover, that it seemed to supply him with 
 a clue to the discovery. And_KajnJ;'.s_aclY?£i25_toi^e new Logic.- 
 was simply_a disguised refutation of the old, a restoration to 
 Logic of the elements which, guided by a false principle, it had 
 been led to reject, and a correction of the false views wEIoh' 
 had arisen out of this rejection. To Kant, indeed, just because j 
 he holds to the idea that the pure process of thought is 
 analytic, the modifications which he has to introduce in order to 
 reach the synthetic process, appear to be accommodations of pure i 
 thought to the nature of our perceptions. Hence, in opposition i 
 to such accommodation, — which, by the nature of the case can 
 only give rise to a knowledge of phenomena, — pure thought is 
 conceived by him as setting up an ideal of knowledge, which, 
 if it could be realised, would be a knowledge of noumena 
 or things in themselves. If, however, we reject the idea of 
 analytic thought, these modifications wiU take an altogether 
 different aspect. They will cease to be accommodations of 
 thought to a foreign matter, with which somehow it has to . 
 deal, and will be seen to be the first steps toward a truer view 
 of the process of thought than that which was expressed in 
 formal Logic ; and the ideal of knowledge will cease to be 
 £omsS°" confused with that abstract identity in which all knowledge 
 
 Logical 
 
 system in and all thought disappear. 
 
 order to adapt 
 it to his 
 purpoae. 
 
 it to his rpj^g £j.g|- g^gpg ^^ ^j^^g transformation or modification of thei 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 339 
 
 analysis of judgment in formal Logic are already taken by Kant 
 in his account of that analysis. For Kant does not simply, 
 borrow his list of forms from the logicians, but modifies their 
 list with reference to the purpose for which he is going to use 
 it, remarking with some naivety, that certain distinctions are 
 of importance with reference to transcendental Logic, which 
 would not properly fall within the scope of formal Logic. The 
 "logical system" had thus to be somewhat modified ere it could 
 afford a clue to the " transcendental system " ; in other words, 
 the clue had to be carried to the point to which it was supposed 
 to lead. Thus, in order to reach a triple division of the sub- 
 classes of the categories, Kant adds, in quality, the infinite 
 judgment to the positive and the negative judgments, and, in 
 quantity, the singular judgment to the universal and particular 
 judgments. Formal Logic, as he points out, does not need to 
 distinguish the singular from the universal judgment, because 
 in both the subject in the whole of its extension is subsumed 
 un^er the predicate. Nor does it regard the infinite judgment 
 as distinct from the affirmative : for it has to do only with the 
 afiirmative or negative character of the judgment, and not with 
 the character of the predicates asserted or denied. But 
 transcendental Logic has to take account of these distinctions, 
 for it has to estimate the objective value of the judgments 
 in question. In truth, the schema of triplicity is out of place 
 in an analytical division of judgments ; — and where should 
 division be analytical if not in the vexj logic of analysis ? — 
 for, as Kant himself el^where tells us, analytical division 
 is always twofold, and only synthetical division threefold. " It 
 has been made a difiiculty that my divisions in pure philosophy 
 are almost always threefold. But this lies in the nature of the 
 case. For, if an a priori division is to be made, it must be 
 either analytical, according to the principle of contradiction, or 
 synthetical: and if, in the latter case, it is to be based on 
 a priori conceptions (and not as in mathematics on perceptions), 
 then, a'ccording to that which is required for synthetic unity 
 
340 THE CRITIQUE OP PURE REASON. BOOK i. 
 
 in general, the division must be a trichotomy : for it must be 
 a division which includes, first, a condition, then a conditioned, 
 and thirdly the conception which arises from the union of these 
 two."^ But if this be true, it follows that for formal Logic there 
 never can be any third category, and it can, therefore, give us 
 no real help in the discovery of a list, the essential feature 
 of which is triplicity. Or, to put the same idea in another 
 form, it is useless in a thought which proceeds by mere 
 identity, to seek for the same elements which are to be dis- 
 covered in a thought which proceeds by differentiation and 
 integration. The third category, expressing, as Kant himself 
 says, the combination of the other two, involves a function of 
 thought for which in mere formal Logic there is no room. 
 The categories This may be made clearer, if we refer anticipatively to the 
 
 m each class ^ j. ^ 
 
 aitenatives; ^se which Kant makcs of his list of categories. The mode of 
 their discovery would suggest the idea that, while every judg- 
 ment is necessarily determined in quantity, quality, relation, 
 and modality, yet that it can only be determined according to 
 one out of three alternatives under each head. And this, 
 indeed, is the usual way in which formal Logic presents i 
 the matter to us, though it offers the choice of two, rather 
 than of three, alternatives. Thus every judgment is either 
 affirmative or negative, and the affirmative and negative 
 judgments are regarded as two distinct and independent 
 acts which have no necessary relation to each other. If, 
 however, we insist, on this division, and treat it as absolute 
 i.e., if we hold to ihe idea that affirmative and negative judg-.^. 
 ments absolutely exclude each other, then both judgments will 
 lose all their signification ; for a mere affirmation which is not 
 determination and therefore not negation, and a pure negation 
 
 ^ Kr. der UrtlidUhraft, Introduction, § 9. We might ask also, how it is 
 possible consistently with Kant's principles that thought should find in itself 
 the tertium quid, which is necessary to combine the duality of its elements in a 
 third category, which as Kant tells us, unites the two others and is something 
 more than their mere external combination. (B. 110.) This subject will be 
 considered in the chapter on the schematism of the categories. , 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. ' 341 
 
 which is not determination and therefore not affirmation, are 
 equally unmeaning. It may, however, be safely said that we 
 never make a negative judgment which, in removing a predicate, 
 does not imply the position of some other predicate within 
 the general sphere of reality to which the judgment refers ; 
 nor an affirmative judgment, which does not further determine 
 the subject by excluding from it some predicate lying within the 
 sphere of reality to which it is already regarded as belonging. 
 When we say that, " The soul is not material," we are 
 thinking of the soul as belonging to a genus which is now 
 further divided or determined ; and, therefore, in denying 
 that it belongs to one species of the presupposed genus, we 
 are implying that it belongs to one of the other possible 
 species, which are more or less definitely marked off by the 
 exclusion. So, if we say that " This triangle is right-angled," 
 we are by this affirmation excluding other alternatives, which 
 are more or less definitely thought as left open by the 
 general definition 'of a triangle. There are not, therefore, 
 three alternative kinds of judgments, positive, negative, 
 and limitative (or infinite), but judgment, if we characterise 
 it fully, is always essentially limitative ; or, in other words, 
 judgment always implicitly involves what is expressed most 
 fully in the limitative form of judgment, i.e., affirmation through 
 negation. Thus the three forms do not stand beside each 
 other as forms between which a choice must be made, but 
 really express different " momenta " or phases in one process 
 of thought, to which indeed we may separately direct attention, 
 but which we cannot treat as independent processes. The pure 
 affirmative relation of a predicate to a subject would have no 
 meaning, if there were not also a difference which might take 
 the form of a negative judgment ; and the truth can be expressed 
 only by showing how through this difference or negation a posi- 
 tive unity stUl maintains itself. A similar line of argument 
 might be pursued in relation to the quantitative determination 
 of objects as one, many, or one-in-many. The functions of 
 
342 THE CRITIQUE OP PURE REASON. book I. 
 
 thought by which such determinations are made cannot be 
 separated, as singular, particular, and universal judgments 
 between which we have to choose ; but the determination of an 
 individual is always a determination of it as one particular 
 form of a universal, or as itself a universal in which many 
 particulars are reduced to unity. These judgments therefore 
 can only be distinguished from each other as special 
 "momenta" in the process of determination, which cannot 
 be separated from each other, though attention may be 
 especially directed to one or other of them, 
 nor does Now Kant, as we shall see more fully in the sequel, does 
 
 Kant treat J -L > 
 
 them as such, ^q^ Ixesd, thcsc Categories, or the judgments founded on them, 
 as alternatives, '^his is at once obvious in regard to the cate- 
 gories of Quantity and Quality, from each of which he derives 
 one, and only one a priori principle, for the determination of 
 experience. 'Thus under the head, of Quality he does not 
 attempt to find any application for the categories of reality or 
 negation, but only for that of limitation, which is the unity of 
 the two. Hence, the only principle which he derives from 
 this class of his categories is that the real necessarily has a 
 certain degree of intensity, and is determined as standing 
 somewhere between zero and absolute fulness. In the same 
 way the only principle which he derives from the categories of 
 Quantity is that every phenomenon must be represented as 
 extensive, i.e., a many-in-one, a number of homogeneous units. 
 There is no separate or alternative determination by Unity, 
 Plurality, and Totality, but every object is regarded as deter-, 
 mined by the last of. these (as including the other two). 
 Under Eelation, indeed, we have three separate principles, but 
 this does not really invalidate what, has been said, for the 
 three principles are not three alternatives between which we 
 have to choose, but three different steps in one process ; and, as 
 will be shown hereafter, the principle of causality presupposes, the 
 principle of substance and the principle of reciprocity implies 
 both the others. They are, however, in this case separately 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 343 
 
 stated, because each involves a complex movement of thought 
 in itself. Lastly, the categories of Modality are not alter- 
 natives, but, according to Kant's own showing, three stages in 
 the progressive determination of an object in relation to the 
 knowing subject ; for, as he says, " it is a gradual process by 
 which the understanding incorporates with itself that which is 
 presented to it ; first, it problematically judges something to 
 be possible ; next, assertorially accepts it as true, and finally 
 determines it as inseparably bound up with itself, i.e., as 
 necessary and apodictically certain." '■ 
 
 It appears, then, that we have not here a threefold division 
 of each class of categories, out of which one alternative may 
 be selected at one time, and another at another, for the deter- 
 mination of particular objects : we have only different aspects 
 or ' momenta ' of that determination of the manifold by thought 
 which is involved in the cousciousness of objects as such : and 
 the "judgment of experience" is not only a determination in 
 quality, quantity, relation, and modality, but in each of these 
 by all the categories. It is true, however, that Kant by his 
 way of taking up each category and principle by itself renders 
 this at first somewhat obscure, and that he scarcely gives us 
 sufficient indication that he is dealing with different aspects of 
 one synthesis, or ' momenta.' in one process, to counteract that 
 impression. 
 
 It may, however, be said that though Kant may not absolutely ^^^^/f^"" 
 separate the subordinate categories in each class from each other, ™dd™amicai 
 yet he does seem to make a decided division between the "^ ^^° 
 different classes of categories. Every logical judgment as such, 
 must be determined in quantity, quality, relation, and modality. 
 Biit, when Kant deals with the corresponding determination of 
 objects, he seems to draw a marked line of division between 
 their determination by quantity and quality, and their deter- 
 mination by relation and modality. "This table," he says, 
 " which contains four classes of conceptions of the understanding, 
 ♦ 1 A. 76 ; B. 101. 
 
344 THE CRITIQUE OP PURE REASON. book i. 
 
 has a primary division into two sections, of which the former 
 is directed to objects of ^perceiption (pure as well as empirical), 
 while the latter is directed to the existence, of these objects (in 
 relation to each other and to the understanding). The first 
 class I will call the mathematical, the second the dynamical 
 categories. The first class have, as we have seen, no correlates 
 such as are found in the second. This distinction must have some 
 ground in the nature of the understanding.''^;. The meaning 
 of this distinction will be more fully discussed in connexion 
 with the principles of pure understanding. For the present it is 
 sufficient to point out that in the mathematical categories and 
 principles we find an element supplied which was wanting to 
 the Aesthetic, in which Mathematics was shown to imply a pure 
 perception, and was taken as sufficiently explained by such per- 
 ception. Now, however, we begin to see that the synthesis of 
 Mathematics is possible only by a construction which is guided 
 and determined by the categories of quantity and quality. 
 Thus, understanding as well as sense is at work in the synthesis ;* 
 of pure space and time, and of matter as in space and time, by 
 which perception is anticipated and determined. But such 
 determination of the perception of objects is distinguishable from 
 the determination of objects for perception. For in this latter 
 case, we do not determine how the manifold must be put 
 together in space in our perception of it, or how it must be • 
 united with a consciousness which is subject to conditions of 
 time ; but we determine how it must exist for us as an object 
 distinct from and related to all other objects, and distinct from 
 and related to the mind that knows it. This kind of deter- | 
 mination is obviously different from the other, although it : 
 cannot be separated from it ; for in the former case we merely "* 
 lay down certain common conditions for all determination of 
 objects as apprehended under conditions of space and time, 
 while in the latter we show that in order to be known objects 
 for us, they must be regarded as permanent realities, each of 
 
 'B. 110. 
 
CHAP. III. THE METAPHYSICAL DEDUCTION. 345 
 
 which determines, and is determined by, all others, and 
 stands in certain definite relations to the knowing mind. 
 •We might, therefore, say that the mathematical categories 
 determine the object in itself, while the categories of relation 
 determine it as existing in one world with all other objects, and 
 the categories of modality determine it as existing in relation to 
 the mind. Yet, it is to be observed that the different kinds of 
 determination are not really separable from each other, except 
 in the sense that a determination by the categories of relation 
 and modality shows a deeper self-consciousness, a deeper con- 
 sciousness of that which was really present in the first judgments 
 of quantity and quality. In our first apprehension of objects Three stages 
 as in space and time, we take them as isolated from each other, 
 as each having quality and quantity, but not yet as having 
 necessary relations with each other ; for their mere presence in 
 one space and one time does not yet seem to be, or to involve, 
 such a relation, — rather it seems to be compatible with their 
 complete indifference to each other. In this stage of thought, 
 therefore, we treat things as having certain common qualities 
 because they occupy space and time, but we do not recognise their 
 necessary connexion with other objects. It is a further step of 
 reflexion when it is seen that such a consciousness of objects 
 involves that they have a permanent reality through all their 
 changing appearances, and that their phenomena are not merely 
 coexistent or successive but stand in essential relations to each 
 other. Lastly, it is only at a still more advanced stage of 
 reflexion that the objects so determined are seen to be neces- 
 sarily related ito the consciousness for which they are, and 
 that this relation also is regarded as an essential element 
 in the knowledge of them. These three stages roughly 
 correspond ' respectively with the ordinary consciousness, 
 with science, and with critical philosophy ; but they must 
 not be regarded as really separate from, or as independent 
 of, each othejy; for in the first "mathematical" determination 
 of perceptions, there is already involved the reference of them 
 
346 THE CRITIQUE OP PURE REASON. book I. 
 
 to permanent- objects, which, as such, stand in permanent 
 relations to each other in the whole of experience ; and in 
 this again is involved their relation as objects to a conscious 
 self. Thus 'the 'judgment of perception,' if we can properly 
 speak of a judgment of perception, is already implicitly the 
 'judgment of experience,' and the judgment of experience is 
 already implicitly the judgment of criticism or philosophic 
 reflexion. But if this be the case, then the classes of 
 categories cannot be treated as independent, any more than the 
 special categories under each head. Further, if this view be 
 accepted, we shall have really what Kant desired, viz., a complete 
 system of categories, which shows its completeness by the fact 
 that each category is a " moment " in a process which begins 
 in the first reference of an object to itself, goes on through 
 its determination in relation to all other objects, and ends with 
 its determination in relation to thought^ 
 Kanfs gradual How far did Kant approximate to such a view of the 
 
 advance to- 
 
 ^n^\yat°em Synthetic process of knowledge ? We have already seen how 
 ea^ones. j^^ began, even in the section devoted to the Metaphysical 
 Deduction, to transform and modify the assumptions of formal 
 Logic with which he seemed to start, in order to draw from the . 
 analysis of judgment a threefold classification of categories, 
 which he further departs from formal Logic in regarding, not as 
 alternatives, but as " momenta " in the process of the determina- 
 tion of objects : and we have seen that he fihally indicates 
 that the different classes of categories correspond respectively 
 to a determination of the object of perception in itself, in rela- 
 tion to other objects, and to the mind. How far he saw the 
 organic or systematic unity of this process, however, we cannot 
 discuss, till we have considered the Transcendental Deduction, 
 and the development of the principles of the pure under- 
 standing to which it leads. 
 
347 
 
 CHAPTEE IV. 
 
 THE TEAKSCENDENTAL DEDUCTION OF THE CATEGOEIES. 
 
 IN the last chapter it has been shown that Kant's meta- Heauitsofthe 
 metaphysical 
 
 ■ physical Deduction of the Categories is a failure, or that, Deduction, 
 in so far as it can be pronounced successful, it is in spite of the 
 supposed " guiding thread," rather than by means of it, that he 
 finds his way to the truth. It is only because he transforms the 
 conceptions derived from formal Logic by the aid of a principle 
 not so derived, that he gets from them anything lilie what he 
 wants : and even the conceptions from which he starts could 
 not have been found by him in such a Logic, if it had been 
 faithful to its principle of Identity. It follows that the claim 
 of the list to be an exhaustive enumeration of the categories, or 
 even of the primary categories, cannot be made good by a mere 
 reference to the functions of unity in the analytic judgment. 
 It must be considered on its own merits ; and it must be veri- 
 fied, if at all, by being traced back to the primary principle of 
 synthetic judgment. Such a verification of it, however, is 
 I impossible, until we have seen what that principle is, and how 
 I it manifests itself in experience or knowledge. And this it is 
 the business of the Transcendental Deduction to show. 
 
 The Transcendental Deduction undoubtedly contains the TheDeduotion 
 
 ., . i .™ the first 
 
 central thought of the Critique. It is, tnereiore, the point and in the 
 
 <-> -^ ^ second edition 
 
 on which Kant bestowed most labour, and with his exposition ""i^s cw'»9«- 
 of w^ich he found it most difficult to satisfy himself. In the 
 
348 THE CRITIQUE OF PURE REASON. book i. 
 
 second edition of the Critique, he entirely changed the form of 
 it, introducing into the reasoning an important link which had 
 been wanting in the first edition. On the other hand, he at 
 the same time deprived it of some of the illustrations and 
 developments of the first Deduction -^nd in the effort to free 
 it from PsychologyJ and to make it more distinctly a theory of 
 
 ' In the preface to the first edition of the Critique., Kant distinguishes the 
 •psychological question as to the faculties which make knowledge possible from 
 the proper question of the Deduction which relates to the theory of knowledge. 
 "This Deduction," he declares, " has two sides. On the one side it refers to 
 the objects of the pure understanding, and what it aims at is to prove and 
 make intelligible the objective validity of its ij,, priori conceptions. On the 
 other side, it considers the pure understanding itself in a subjective point of 
 view as to its possibility, and the faculties for knowledge which it implies. 
 Now, this latter inquiry has great importance in view of my main object, but 
 it is not essential to it : for the main question is always what, and how much, : 
 we can know by meaus of understanding and reason without any aid of experi- '}'• 
 ence, and not how the faculty of thinking itself is possible. In dealing with 
 the second question we may be said to be seeking a cause for a given eflfect, 
 and our answer might therefore be regarded as an hypothesis : and if I take 
 the liberty of having an opinion on such a question, why should not the reader , 
 be permitted to adopt a different opinion t In truth, as I am prepared to 
 show, there is really nothing hypothetical in my results : but for the present 
 I am content to preclude the objection by recalling to the reader that, even if 
 
 \ my subjective deduction may not have produced in his mind that assured 
 conviction which I expect, yet the objective deduction, which is my main 
 point, will retain all its force : and for this what is said in pp. 92-3 would alone 
 be sufficient." Kant elsewhere declares (Prolegomena, § 32) that the "ques- 
 
 ' tion is not of the origin of experience but of that which lies in it. " But this ■ 
 probably refers, as Dr. Vaihinger suggests, merely to Kant's rejection of 
 Locke's explanation of knowledge by Empirical Psychology. Dr. Vaihinger 
 (I. 324) distinguishes from this a " Transcendental Psychology,'" which inves- 
 tigates the "psychological possibility," as contrasted with the objective 
 validity, of the a priori principles. Now, it is true that, in order to show that, 
 and how, the pure a priori conceptions are objectively valid in relation to 
 experience and not beyond it, it is necessary to consider how they give rise to 
 a priori synthesis in relation to the forms and matter of sense : in other words, 
 Kant has to give a theory not of knowledge simply but of human knowledge. 
 And this may so far be called a psychological theory, as it deals with pheno- 
 mena which are relative to our faculties of perception and conception, as 
 contrasted with things in themselves, which would be the objects of an 
 intuitive understanding — an understanding whose thought was at the same 
 time perception. / Empirical Psychology, on the other hand, has, in Kant's 
 view of it, to do with the object of inner sense, i.e., with one of the objects of 
 experience, and not with the universal conditions of experience./ 
 
 The most important difference in the Deduction of the second edition lies, 
 as we shall see presently, in the way in which it is connected with the idea 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 349 
 
 knowledge, he discarded some of tliose middle terms which 
 helped to connect it with the ordinary views of the subject. 
 Hence, it is more than doubtful whether, on the whole, any 
 advantage was gained by the new method of treatment. It is 
 a subsidiary inconvenience that Kant neglected to remould the 
 later parts of the Critique, and especially his proof of the 
 principles of pure understanding, in conformity with the 
 changes which he had made in the Deduction, thereby 
 making it more difficult to understand the connexion of his 
 thought. In the following exposition I shall try to combine 
 the statements of both editions. 
 
 rSxQ main point which Kant seeks to prove is that the ^^^o??"^ 
 categories or forms of synthesis which belong to the pure proTCd^f be 
 understanding have an ■ objective value, because they are the consciousness 
 
 of objects as 
 
 necessary expression of the unity of thought in relation to the ™''^- 
 ' manifold ' of sense, without which the latter could not give 
 rise to the consciousness of a world of objects. This expression 
 Kant shows to be necessary for the explanation of knowledge, 
 because without synthesis the conscioiisness to which im- 
 pressions of sense could give rise would be only a scattered 
 and unconnected consciousness, and not the consciousness of 
 'Nature,' as a system of permanent objects acting in definite 
 ways on each other.') In fact, they could give rise to no 
 objective consciousness at all. On the other hand, Kant main- 
 tains that it is possible to explain that consciousness by the aid 
 of a synthetic activity of mind, guided by the categories. This, 
 indeed, seems to make objects dependent on the consciousness! 
 for which they exist. But then the materials with which the 
 
 of judgment. Instead of simply asserting that the categories, which have 
 been derived from the logical analysis of the act of judgment, are the prin- 
 ciples by which the understanding determines objects as such, as he does 
 in the first edition (A. Ill), Kant shows that the judgment as such is the 
 expression of the objective unity of apperception (B. 140, § 19). Unfortun- 
 ately this argument, w'hile it seems to authorise us to regard the categories 
 as determinations of objects, if they have been correctly deduced from the 
 idea of judgment, is fatal to the idea of analytic judgment from which they 
 have actually been deduced. 
 
350 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 understanding has to work, and out of which it has to make its 
 objects, are not things in themselves but perceptions which are 
 by them awakened in us : and these, as so awakened, already 
 stand in relation to the unity of consciousness, and thus are 
 prepared to receive that further determination which makes 
 them elements in knowledge as the consciousness of a ' Nature' 
 or objective world, 
 seif-oonsoious- Now, the Central point of this Deduction lies in what Kant 
 
 ness possible ■*■ 
 
 afonSef "alls 'The Transcendental Unity of Apperception.' What 
 of oiSe°ota?°^'' does this mean ? Apperception is a word borrowed from 
 Leibniz, who uses it to designate self-consciousness, and that 
 consciousness of objects which goes with self-consciousness, as 
 opposed to perception, a consciousness of objects which is not 
 ' reflective,' or does not return upon itself and become aware of 
 itself in distinction from its objects. Now, Kant seeks to 
 analyse the process involved in this return upon self, and what 
 he maintains is that it is the correlate of a determination of the 
 manifold present in ' perception,' a determination in virtue of 
 which the elements of that manifold are brought together in, 
 or referred to, a ' Nature ' or system of objects. Tin other words, 
 it is Kant's view that we can be conscious of one self in relation 
 to which all our ideas form a unity, only if, and in so far as, we 
 are able to bind together all the elements of our 'perception' in 
 the consciousness of one objective world. J Whatever matter, 
 therefore, there is in my consciousness, it can be present to me, 
 or recognised as part of my consciousness, only in so far as it 
 can be connected with the other parts of my consciousness as 
 an element in the consciousness of one objective world. If it 
 were not capable of being thus connected, it would be for me 
 ' as good as nothing,' i.e., it would remain a mere ' perception/ 
 a part of a merely individual and empirical consciousness ; but 
 it could not become an element in a consciousness of objects, as 
 such, or be combined with the consciousness of self. Here, 
 therefore, as Kant maintains, we have a compulsitur, under 
 which, as conscious subjects, we lay the world as an object of 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 351 
 
 knowledge to us. For we can always say this much of it 
 a priori, that nothing can be an object for us, unless it con- 
 forms to the conditions under which we can be conscious of 
 ourselves as self-identical subjects in apprehending it. In this 
 way all our consciousness of objects is determined a priori in 
 conformity with the conditions of a complete self-consciousness ; 
 and whatever predicates it is necessary to attach to objects in 
 order to fit the consciousness of them for being part of the one 
 consciousness of self, these we can attach to them a priori. 
 On this consciousness of self, therefore, we can ta,ke our stand, 
 and from it we can reason backward to all that is necessary for 
 it, or presupposed in it. We do not need to wait for the 
 realisation of knowledge to say that the conditions of it must 
 be there ; or, in other words, to say that anything that is to be 
 known must be known in such a way that self-consciousness, the 
 consciousness of a self which is the same in relation to it as to all 
 other objects, shall be possible. And if we suppose that, in the 
 individual, a percipient consciousness precedes an appercipient 
 consciousness, yet it will not be possible for us to account for the 
 latter by means of the former ; on the contrary, we shall be 
 obHged to treat the former as conformed a priori to the conditions 
 of the possibility of the latter. Hence we may, and indeed 
 must, start with the unity of intelligence with itself as the 
 precondition of all objects for the intelligence; seeing that it is 
 only through their conformity to that precondition that they 
 can become objects for us, or that we can be conscious of 
 ourselves in relation to them. 
 
 " The unity of the manifold in one subject is synthetic : ughed'*^' 
 therefore pure apperception puts into our hands a principle of pt™p?{^n* 
 synthetic unity for the manifold in all possible perception." ^ won. 
 What this means is that, as all the manifold of perception must 
 be present to one subject, so we can say a priori that it must 
 be combined together in such a way as to make tjie conscious- 
 ness of one self in relation to it possible. And this again 
 * lA. 117. 
 
352 THE CRITIQUE OF PURE REASOlSr. book i. 
 
 implies that, starting from the unity of self-consciousness, we 
 can lay down principles of unity to which perception in all its 
 manifoldness must conform. Even, therefore, if it be admitted 
 that perceptions are given prior to, and independent of the 
 consciousness of self, as well as of all the principles for the 
 determination of objects which are implied in the consciousness 
 of self, yet, as so given, they must be in ' pre-established 
 harmony ' with those principles. It is thus only that we can 
 combine the assertion, that the manifold of perception is given 
 prior to its being known, with the assertion, that it is for us 
 " as good as nothing," unless it be conformed to the conditions 
 under which it can be known. The ' blind perception,' and the 
 ' undetermined phenomenon ' which is said to be its object, 
 r are ideally separable from the determining conception, but 
 even in this separation we must conceive the former as standing ' 
 in such a relation to the latter as to be prepared for subsumption 
 under it. This may be illustrated by a passage from Kant's .{ 
 Logic. ■ 
 
 tion waite for " ^^^ cvcry cognition we .distinguish the matter, i.e,, the ;•, 
 theoKtoar™ objcct, and the form, i.e., the manner in which we know the 
 
 consciousness. , ,^^. -^^v 
 
 object. If, e.g., a savage sees a house m the distance, he has 
 the same object before him as a civilised man, who recognises 
 it as a habitation suited for human occupation. But, in form, 
 the consciousness of the one is quite different from the con- 
 sciousness of the other. In the case of the savage there is mere 
 perception, while in the case of the civilised man, there is both 
 perception and conception." ^ 
 
 This example may serve to illustrate how an image may be 
 present to us in perception, and yet waiting, as it were, for the 
 conception under which it is to be subsumed, and to which it :; 
 is a priori conformed. It is, however, necessary to keep in view 
 the distinction between a case like this, and the case in which .■ 
 we have, on the one side, a pure conception ^ which has to be 
 
 ^ Logic, Introduction, V.; R. III. 197 ; of. I. 439; H. VIII. 33; of. VI. 33. 
 2 A. 142; B. 181. 
 
 .^.^li 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 353 
 
 schematised in relation to perception, and on the other a per- 
 ception, i.e., an image, which has been formed out of the 
 manifold of sense by the synthesis of imagination, but has not 
 yet been determined by any conception, and therefore is not 
 yet referred to any object at all. Por in this latter case, it may 
 be a question whether the two elements which are ideally 
 distinguishable can be thought except in relation to each other. 
 This, however, is a point the discussion of which must be post- 
 poned, till we have more fully discussed the relation of both 
 conception and perception to the unity of self-conscjousness. 
 
 " The ' I think ' must be capable of accompanying all my ah our ideas 
 
 ■^ r J a J must be con- 
 
 ideas : otherwise there would be presented to my mind an idea th*™o*arfbUity 
 of something which could not be thought, and this means sciousnemin 
 that the idea would be impossible, or, at least, that it would them, 
 be nothing at all for me." ^ ( " The proposition that all the 
 various elements of our empirical consciousness must be bound 
 together in one self-consciousness, is absolutely the first 
 synthetic principle of all our thinking. \ We must, however, 
 observe that the mere idea ' I,' in reference to all other ideas, 
 (the collective unity of which it renders possible,) constitutes 
 the transcendental consciousness. This idea may be clear 
 or obscure — ^the question is not of its clearness or obscurity 
 nor even of the actual realisation of it in any shape — but the 
 possibility of the logical form of all knojvledge rests necessarily 
 on the relation of ideas to this apperception as a capacity " ^ 
 (i.e., to the possibility of self-consciousness). After what has 
 been said, such utterances can cause little difficulty. (Kant is 
 insisting on the possibility of all ' ideas ' being determined in 
 relation to the conscious self, as the one condition which we can 
 lay down for them a priori. \ Nothing can get into our mind, 
 nothing can exist for us, ii it is not capable of being referred 
 to the one self, or determined as an element in its consciousness 
 of itself Hence, we are authorised to lay down a priori for all 
 ideas whatever conditions are implied in this ultimate deter- 
 
 ♦ IB. 132. ^A.mn0te. 
 
 VOL. I. Z 
 
354 
 
 THE CRITIQUE OF PURE REASON. 
 
 BOOK I. 
 
 ( 
 
 But this 
 Implies a con- 
 formity of all 
 sense-percep- 
 tions to 
 
 mination, and we are authorised to exclude from them whatever 
 would prevent it. The data of sense are dependent on affec- 
 tions that come to the mind from without apart from any 
 activity of its own, and hence they may seem to be beyond the 
 reach of any direct determination by our thought. Neverthe- 
 less, we are in a position to make certain a 'priori assertions in 
 relation to these data, as taken up into our perceptions : for 
 all phenomena,' i.e., all that appears to us through sense, 
 " must so come into the mind and be apprehended by it as to 
 agree with the unity of apperception."^) Whatever charac- 
 teristics, therefore, they may have as determined by something 
 independent of self-consciousness, they can have none which are 
 inconsistent with their receiving the ultimate stamp of self- 
 consciousness : and all that is involved in this ultimate stamp 
 may be predicated of them a priori. 
 
 Now, what is so involved ? There is involved in this rela- 
 tion to self-consciousness, Kant answers, a conformity of the 
 tain VprioS scnse-perceptions both in form and matter to certain synthetic 
 
 synthetic 
 
 principles. principles, through which they are converted into the con- 
 sciousness of a ' Nature ' or world of objects. For it is " only 
 as I am able to bind together a multiplicity of given ideas m 
 owe consciousness that I can be aware of the identity of conscioiis- 
 ness in these ideas. In other words, the analytic unity of 
 apperception is possible only through a synthetic unity. The 
 thought that these ideas given in perception belong every one . 
 of them to me, means no more and no less than this, that I 
 unite them in one self-consciousness, or at least that I am able 
 so to unite them. If it is not itself the conseioiisness of the 
 synthesis of these ideas, it at least presupposes its possibility; ' 
 for only as I can comprehend all their manifold elements in 
 one consciousness, do I call them all my ideas : and, if I were 
 not able so to comprehend them, I should have a self split up 
 into as many parts and endowed with as varied qualities as 
 the ideas of which I am conscious." ^ In other words, the unity 
 'A. 122. 'B. 133, §16. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 355 
 
 which is given to all the perceptions, in so far as they are con- 
 nected together and related to each other as elements in the 
 consciousness of one world of objects, is the necessary correlate 
 of the consciousness of the identity of the self in apprehending 
 that world. If there were any break between perceptions, so 
 that they could not be connected with each other as referring 
 to one world, there would be a corresponding break in the 
 ■consciousness of self. As some have imagined that we are in 
 different worlds when we wake and when we sleep, so here 
 there would be . a hiatus in consciousness, which we might 
 describe as a difference of worlds ; this again would involve a 
 hiatus in the consciousness of self, which would be equivalent 
 to the existence in us of more than one self: in other 
 words, as more than one self is an absurdity, it would 
 involve the impossibility of any consciousness of self at 
 all. But, Kant argues, this unity of the world is not 
 given to us through sense : the world is one for us only 
 as we make it one. " Combination does not lie in the 
 objects and cannot be borrowed from them through sense- 
 perception, and so taken up into the passive understanding : it 
 is a thing achieved by the activity of the understanding itself, 
 as a faculty of a priori synthesis, which brings the manifold of 
 given ideas under the unity of apperception." ^ In bringing 
 the perceptions together with each other as perceptions of one 
 object or one world of objects, the understanding also brings 
 them together with the one self as its perceptions, and thus 
 only makes possible a consciousness of that self, as one with 
 itself in apprehending all these objects. 
 
 But if this is true, then all the forms of svnthesis which Kant mter- 
 
 ' " prets the pre- 
 
 are necessary to this consciousness of the unity of the world of hlmony rf 
 objects have an a priori justification for their application to owns^nse. 
 the manifold of perception. The manifold of perception, in 
 fact, must be in ' pre-established harmony ' with them. On 
 this point there is an interesting passage in Kant's answer to 
 * / 1 B. 134, § 16. 
 
356 THE CRITIQUE OF PTJEE REASON. BOOK I. 
 
 Eberhard's criticism of the Critique of Pure Reason, in which 
 he even goes the length of attributing his own view to 
 Leibniz. 
 
 " Is it possible to believe that Leibniz by his pre-established 
 harmony between soul and body can have meant an agreement 
 between two entirely independent beings, which by no powers 
 of their own can be brought into any relation with each other ? 
 That would have been to play into the hands of Idealism in' 
 the openest way : for why should we assume the existence of 
 body at all, if it is possible to explain all that goes on in the 
 soul as a working of its own powers, which it might carry on 
 quite as well in complete isolation from anything else. Now, 
 it is not to be denied that the soul and that (to us altogether 
 unknown) substratum of those phenomena which we call body, 
 are things quite different from each other. But these pheno- 
 mena themselves, as modes of perception, which are determined 
 by the subjective nature of the soul that perceives them, are 
 mere ideas. And as they are mere ideas, we can easily suppose 
 that between the sensibility through which they are presented, 
 and the understanding which is only another faculty of the 
 same subject, there is a community which is regulated a priori 
 by certain laws ; and this without denying the necessary 
 dependence of the sensibility for its affections upon outward 
 things, and without accepting the idealistic doctrine which 
 surrenders the existence of those things. This harmony between 
 the understanding and the sensibility (which makes possible 
 for us an a priori knowledge of the laws of nature) the Critique 
 of Pure Reason has asserted, on the ground that without it 
 experience would have been impossible. For, if the objects of 
 experience were not conformable, as regards their perception, 
 to the conditions of our sensibility, and also, as regards the 
 combination of their manifold, to the principles by which the 
 understanding unites its objects in one consciousness, they 
 could not be taken up into the unity of consciousness or form 
 part of one experience : and that is only another way of saying 
 
CHAP. IV. THE TRAISTSCENDENTAL DEDUCTION. 357 
 
 that for us they would be nothing at all. We cannot, however, 
 give any reason, why we -should have just such a sensibility, 
 and just such an understanding, as is needed in order that by 
 their joint action they may make experience possible. Still 
 less can we tell why such heterogeneous sources of knowledge 
 should conspire together, not only to secure the possibility of 
 experience in general, but also (as wiU be shown in the Critique 
 of Jvdgment) to make possible an experience of nature such as 
 we actually have, under those numerous, special, and merely 
 empirical, laws of which understanding tells us nothing a priori : 
 i.e., why a nature, of which we know nothing except through 
 experience, should yet be found to agree with our understand- 
 ing, just as if it had been purposely adapted to it. Leibniz, 
 speaking primarily in reference to our knowledge of bodies and 
 in particular of our own body as that through which we are 
 brought into relation with other bodies, gave as the ground for 
 this agreement a, pre-established harmony. By using this phrase 
 he did not really explain, and could not have intended to 
 explain, that agreement, but only to show that we must think 
 of it as a purposive adaptation, due to the ultimate- cause of 
 our own existence as well as of the existence of all things 
 without us ; which, however, we- are not obliged to regard as 
 an objective pre-adaptation of independent things to each 
 other, but only as a subjective pre-adaptation of our faculties 
 of sense and understanding, in virtue of which they have such 
 a relation to each other as is necessary for knowledge." ^ 
 / This " pre-established harmony " of the faculties, then, means/rhe general 
 
 ^ / purport of the 
 
 that perceptions must conform to the conditions of thair g^^gotjo" *" 
 determination by conception, so far as is necessary to the ^i^*™- 
 consciousness through them of a connected objective world, or)" 
 as Kant puts it, of a " Nature " ; which again is the correlate 
 of the consciousness of an identical self^ When, however, we 
 look a little more closely into Kant's account of the conditions 
 of knowledge, we find that his so-called " pre-established. 
 • 1 E. I. 480 ; H. VI 65 ; cf. K. XI. 57 ; H. VIII. 717. 
 
358 THE CRITIQUE OF PURE REASON. booki. 
 
 harmony" amounts to something more than a harmony of 
 independent faculties : that, in fact, he finds one principle at 
 work both in conception and perception, and that this identity 
 of principle is his ultimate reason for their agreement with each 
 other. This becomes manifest, when under the guidance of the 
 Deduction as given in the first edition of the Critique, we follow 
 Kant's account of the different elements required for knowledge 
 from the manifold of sense up to the unity of thought. For there 
 we find him showing that, if we take the data of sense as mere 
 passive impressions, and if we ask how we can derive knowledge 
 from them, we are gradually driven on to add new qualifica- 
 tions to these data, until finally we reach the conclusion that 
 they must be determined by the categories, if they are to give 
 rise to that consciousness of objects of which the consciousness 
 of self is the correlate. But this he is able to show only by 
 , taking the perceptive consciousness as already determined by 
 the " blind " or unconscious operation of the imagination, the 
 same faculty which, when it - ceases to be blind and becomes | 
 self-conscious, is called the understanding. 
 TiieaynthesiE Kant begins by observing that the impressions of sense,, if 
 tlon.'"^^'^°^^'we take them as mere passive affections of the subject, can by 
 that subject be neither distinguished nor related. For, though 
 sense presents to us a " manifold," it cannot present it as a 
 manifold, and so as held together in one idea, without a 
 synthetic process. " As contained in one moment, an idea 
 cannot be anything but an absolutely simple unit. In order, j 
 therefore, to make a unity of perception (such e.g., as we have 
 in the idea of space) arise out of this manifold, we must first 
 run through the manifold, and then we must gather it together ,j 
 into one." And this process is necessary, not only in relation j 
 to the manifold of sense, but also to the time and space in 
 which we put the elements of that manifold together : for the 
 pure synthesis of time and space is presupposed in the syn- 
 thesis of any perception of things as in time and space. To / 
 have any image of sense, therefore, we require, in addition to 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 359 
 
 the reception of impressions, a synthesis of the imagination, 
 which " takes up impressions into its activity " and without 
 confusing them brings them together in one image ; or, in 
 other words, enables us to " envisage " them as one. 
 
 But again, when we consider this process more minutely, we The synthesis 
 see that it involves a continual reproduction of elements corre- liatim.''"*^' 
 spending to impressions that have ceased : for, as already 
 stated, the immediate impression can have the value of an 
 element in a perception only when brought into connexion 
 with the other elements, with which it is held together in one 
 consciousness ; and mere impressions as such are a flux in which 
 each drives out its predecessor. It appears, therefore, that the 
 successive " apprehension of the elements of the manifold 
 by itself would give rise to no image or connection of impres- 
 sions, unless there were a subjective ground to call up a 
 perception, from which the mind has passed to another per- 
 ception, in connexion with those that follow it, so as to 
 exhibit a whole series of perceptions."^ But, again, "if ideas 
 reproduced each other without distinction as they happened to 
 come together, no definite connexion of them would be pro- 
 duced, but mere confused aggregations, which could not be the 
 basis of a knowledge of anything. Eeproduction, therefore,; 
 must be subjected to some rule, according to which an idea in) 
 our imagination comes into connexion, or associates itself, withj ^ 
 this, rather than with that, attendant idea." But, then, this? 
 means that association itself must have an objective ground, 
 unless we are prepared to say that it is a " mere accident that 
 phenomena fit themselves into such a connexion as is required 
 for human knowledge." In other words, "if phenomena or 
 perceptions were not really subjected to a rule, according to 
 which the elements of the manifold accompany or foUow each 
 other in a certain definite order, our empiric imagination would 
 never get any work to do in conformity with its capacities, but 
 would remain hidden in the depths of the mind as a faculty 
 ♦ lA. 121. 
 
360 THE CEITTQUE OP PURE REASON. BOOK I. 
 
 inert and unknown even to ourselves. If copper were now 
 red, now black ; if a man changed now into this, now into that 
 animal form ; if on the longest day the earth were now covered 
 with fruits, now with ice and snow, my empiric imagination 
 would never get the opportunity of forming such an association, 
 as when the idea of red colour awakens in us the thought 
 of the heavy copper." ^ 
 This synthesis Now, what is the obiective ground in question? It cannot 
 
 IS conformed ' d o J. 
 
 poSmty*o¥be said, Kant argues, that the ground of connexion lies in 
 ness, and so to things in thcmselves independent of the mind, and that it is 
 
 the a 'p^-iori 
 
 conceptions, introduced from them into the mind through the affections of 
 sense. For the affections as given are isolated, and cannot 
 bring into the mind along with themselves a principle of unity 
 among them. Of the unity of the qualities in the thing in 
 itself, if there be such unity, we know and can know nothing. 
 The thing in itself is that which, by the very idea of it, is out 
 of our reach. "We have to do only with the play of ideas in 
 our own minds, and can find no explanation of their connexion 
 outside of the mind. But, starting (from the other side, we can 
 say that perceptions miist come to us in such a way as to be 
 conformable to the conditions under which alone they can form 
 part of our knowledge or consciousness of objects, and so can 
 be brought into relation to the consciousness of self. Now, no 
 perceptions can be so conformed, unless the manifold elements 
 in them have such a tendency to adhere or group themselves 
 together, that it is possible to reproduce each element in con- 
 nexion with the other elements to which it gives place. For 
 " when in thought I draw a line, or represent to myself the 
 time from the noon of one day to that of another day ; or even 
 when I merely seek to set before my thought a certain number, 
 I must obviously, in the first place, apprehend one of the ideas 
 combined in this complex idea after another. But if I always 
 let the elements which come first, escape out of my thought '' 
 (the first parts of the line, the earlier periods of the time, or 
 
 lA. 100. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 361 
 
 the units of the number as they are represented one after 
 another), and if I did not reproduce them as I advanced to 
 those that succeed, my mind could never by any possibility 
 attain to any whole idea, not even to the purest and most 
 primary of all ideas, i.e., the perceptions of space and 
 time."^ The association of our ideas, therefore, rests uponj 
 what Kant calls their affinity or associableness, which itself isl 
 presupposed as the condition necessary to their being brought / 
 in relation to the transcendental unity. of apperception. But 
 this implies that, as received into the mind, ideas are subjected 
 to a synthesis of imagination, which is conformed a -priori to 
 the possibility of self-consciousness. The imagination, indeed, 
 operates " blindly " or unconsciously, but it operates according 
 to certain rules which make its products capable of being 
 united with the consciousness of self; and when that con- 
 sciousness comes, all that it does is " to make the function 
 of imagination intellectual," or, in other words, "to bring to 
 conceptions" the synthesis of the imagination. " Por, in itself, 
 the synthesis of imagination, though exercised a -priori, is yet. 
 always sensuous. For it combines the manifold only in such 
 wise as it appears in perception, e.g., in the figure of a triangle. 
 But in virtue of the relation of the manifold to the unity of 
 apperception, conceptions of objects can be attained by the 
 understanding, though only through the mediation of the 
 imagination, which brings that unity to bear on the per- 
 ceptions of sense." ^ 
 
 ^hus, sensuous perception, in its very genesis, is conformed ^^^meunit 
 « priori to the unity which afterwards in relation to it pi^'^esTtaeif 
 expresses itself in the form of a conception. And if we ask andicTconoS- 
 
 tion. 
 
 tor any explanation of this, the only answer is that the unity 
 which underlies the perceptive consciousness is the same as 
 that which underlies the thinking consciousness, only that in 
 the latter case it is not " blind " but self-conscious.^ " The 
 synthesis of imagination, which is empirical, must necessarily 
 lA. 102. 'A. 124. 
 
362 THE CRITIQUE OP PURE REASOK. book I. 
 
 be conformed to the synthesis of apperception which is intel- 
 lectual, and which in all its pure a priori character is expressed 
 in the category : for it is one and the same spontaneity which, 
 in the one case under the name of imagination, and in the 
 other case under the name of understanding, brings connexion 
 into the manifold of perception." ^ Thus " we are able, and, 
 indeed, we are compelled, to lay it down as a rule extending 
 to all phenomena that, as data of sense, they must be regarded 
 as in themselves associable, and as subjected in their reproduc- 
 tion to the laws of a thoroughgoing cohhexion ;" ^ for "they 
 must come into the mind and be apprehended by it in such a 
 way, that they will agree with the unity of apperception." 
 
 It is necessary here to take note of Kant's exact point of 
 view. He is asking where to look for the source of that con- 
 straint which is laid upon us to combine the elements of 
 perception so as to form complete images,^ which are repro- 
 ducible just because their elements are put together in a 
 definite way according to a rule. We cannot, he maintains, 
 refer this constraint to the thing in itself: we cannot say that 
 these elements are so connected in it, and that we perceive 
 their connexion ; for such connexion, even if we could conceive 
 it to exist in the thing in itself, would not exist there for us. 
 But, starting from the other side, we are able to say that 
 perceptions must so group themselves in being perceived as to 
 harmonise with the possibility of the consciousness of self, if 
 they are ever to be recognised as ours, or to form elements in 
 that objective consciousness which we call knowledge. And 
 this means that these perceptions must already, as images of 
 sense, have their elements related according to definite rules, 
 which we may discern to be exemplified in them. But this 
 connexion of the elements of perception, which is prior to con- 
 ception and yet conformable to it, must be referred to an 
 activity of mind ; and it can be referred to no other activity 
 than the imagination, which acts according to rule, though it is 
 
 1 B. 162 note. " A. 122. ^ a_ io2, ' Eine gauze Vorstelluug.' 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 363 
 
 not conscious of any rule. Hence, although we are obliged to 
 admit that, from the point of view of the thing in itself as 
 apart from our consciousness, we can see no reason why it 
 should afi'ect us in any special way, and though we cannot, 
 therefore, from this point of view, deny that it might affect us 
 in such a way that we could never bind the sense-data into the 
 unity of distinct perceptions, made up of definite elements stand- 
 ing in definite relations to each other; yet we can indirectly 
 exclude this possibility from the point of view of the knowing 
 mind, when we consider what is necessary for knowledge. For 
 if we had such perceptions, we could never determine them as 
 perceptions of anything, and so unite them with the conscious- 
 ness of self It is, however, the same subject which thinks 
 and perceives ; and, if it is to come to self-consciousness in 
 regard to all its perceptions, its perceptions must be accom- 
 modated a priori to the unity of self-consciousness, and, 
 therefore, to conceptions. In other words, a rule of com- 
 bination must be capable of being found in them. 
 
 But this already suggests that for knowledge something ofEe^S^tion 
 more is wanted than reproduction, or even reproduction accord- ™ ™°°^p '"°- 
 ing to a universal rule. The rule itself must in some form be 
 present to consciousness. For " without the consciousness that 
 what we now think is the same with that which we thought a 
 moment before, all reproduction in the series of ideas would be 
 in vain." ^ In other words, knowledge implies not only that I 
 take up the elements of a perception successively, and, retaining 
 them, bring them together in an image, but also that I ' recog- 
 nise ' the unity of the object so constructed : and this means 
 that I become conscious of the unity of thought in the process 
 of its construction. " In order to know an object in space, 
 e.g., a Hne, I must in thought drawl it, ^^'^ ^° ^7 synthesis 
 bring about a definite combination of the given manifold, and 
 that in such a way that the unity of this act is, at the same 
 time, the unity of consciousness (in the conception of a line) : 
 • 1 A. 103. 
 
364 THE CRITIQUE OF PURE REASON, BOOK I. 
 
 it is in this way alone that an object, in this case a particular 
 space, can be known." ' In other words, I am conscious of the 
 object as one through all the manifold of perception united in 
 it, because, and in so far as, I recognise the identity of the 
 conception which has guided me in the whole process whereby 
 I put the elements of perception together. And the same 
 principle must be extended to the whole content of conscious- 
 ness ; for my consciousness, that it is one objective world which 
 is represented in all my perceptions, is the same thing with 
 my consciousness that it is one thought which has guided me 
 in putting all these perceptions together. Or, as Kant else- 
 where puts it, " the original and necessary consciousness of the 
 identity of the self is at the same time a consciousness of an 
 equally necessary unity of the synthesis of all phenomena 
 according to conceptions, i.e., according to rules, which not only 
 make the perceptions necessarily reproducible, but in doing so, 
 also determine an object for such perceptions, or, in othe!^ 
 words, fix for us a conception of something in which they are 
 necessarily combined : for the mind could not possibly think, 
 and think a priori, its own identity through all the manifold- 
 ness of its ideas, if it had not before its eyes the identity of 
 the action by which it subsumes all empirical synthesis of 
 apprehension under a transcendental unity." ^ 
 General reia- Kaut's thought, then, mav be thus expressed: — The con- 
 
 tlonofthe _ o > ' J r 
 
 of oWectrto^ sciousness of an object means the recognition that the imag- 
 nees?""^"""^' iuativc Synthesis, whereby the elements of a perception are 
 put together, agrees with a certain conception, and so can be 
 subsumed under it. /But when the mind thus carries on 
 its synthesis according to conception, and recognises that it 
 does so, it is recognising that its thought maintains identity 
 with itself through the synthetic process. Hence, what on the 
 one side is the consciousness of the object, is on the other side 
 the consciousness of the identity of the self that knows it. 
 The mind in apprehending the object as such, apprehends 
 1 B. 138. ■■ A. 108. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 365 
 
 really the identity of its own action in the synthesis whereby 
 the object is constituted. And thus in correlation, the con- 
 sciousness of self and the consciousness of the not-self spring 
 out of the same synthetic act. Or the consciousness of the 
 conception, by which, as its rule or guiding principle, the 
 imagination is determined, so that it follows one order rather 
 than any other, contains in it at once the consciousness that 
 the manifold thus combined is the manifold of a definite object, 
 
 , and also the consciousness of the unity or identity of the self, 
 which by this synthesis apprehends the object. Further, if we 
 carry out this view to its necessary result, we can see that the 
 consciousness of the identity of the self in all its knowledge of 
 objects must depend on all the manifold of perception being 
 brought under what is really one systematic conception, a con- 
 ception of Nature as a system, all parts of which are neces- 
 sarily determined in relation to each other by general laws. 
 
 n^n other words, it is the consciousness that the mind, in all the 
 synthesis of its perceptions, is guided by one conception, which 
 underlies and explains, on the one hand, our consciousness of 
 one objective world, and on the other hand, our consciousness, 
 in distinction from and relation to that world, of the identical 
 self which knows it^ 
 
 We may get a firmer grasp of this view, if we consider what Eeterence to 
 is implied when we state that our ideas are ideas of something, "f^^X^^ 
 or that they have an object which " is distinguished from them ""^^"^ * ™'^' 
 and yet corresponds to them." How can there be anything of 
 which we have an idea and which yet is not itself an idea ? 
 How can we become conscious of that of which the one charac- 
 teristic is that it is not in consciousness ? It is obvious that, 
 when we look at it in this way, the object reduces itself to a 
 mere X,^an unknown and unknowable something, the very 
 idea of which involves a contradiction : " for outside of our 
 knowledge we have nothing which we could set over against it 
 as corresponding to it." ^ There is, however, something more in 
 * ' A. 104. 
 
366 THE CRITIQUE OF PURE BEASON. BOOK I. 
 
 our thought of the object than this ; for we regard the object 
 as the source of a necessity, which attaches to the mode of 
 combination of the ideas through which we know it. " We 
 view the object as that which hinders our ideas from coming 
 in upon us pell-mell or at haphazard, and which determines 
 them a 'priori in a certain definite way : for we consider that 
 ideas, which are to be referred to an object, must necessarily 
 agree with each other in reference to it, i.e.., that they must 
 have that unity which constitutes the conception of an object." ^ 
 Correspondence to an object is, therefore, not really correspon- 
 dence to an unknown X, which would mean nothing: it is 
 correspondence to a conception or known rule of relation 
 between the elements of the manifold of perception, which we 
 combine as referring to that object. This, however, is im- 
 possible, unless the perception is such a connected whole that 
 it " could have been produced by a function of synthesis 
 according to a rule, and unless the rule was of such a nature 
 as to fix the manner of the reproduction of the manifold by an 
 a priori necessity, and so to make possible the conception of 
 an object in which all the elements of the manifold were 
 combined." ^ In other words, the synthesis of imagiaation 
 involved in perception must be of the same character as it 
 would have been, if imagination, instead of acting on the data 
 of sense priar to the synthesis of the understanding, had from 
 the beginning, been guided by the understanding to produce an 
 image of an object according to a definite conception or rule 
 supplied by itself. If it were not so, if the unconscious 
 synthesis of imagination did not thus anticipate the conscious 
 synthesis of the understanding, it would be impossible for the 
 understanding to form aconception of an object under which 
 the particular image might be subsumed. " So we think a 
 triangle when we are conscious of the combination of three 
 lines according to a rule by which such an image of perception 
 could always be constructed. This unity of rule determines 
 ^ A. 104. « A. 105. 
 
, consciousness. 
 
 CHAP. IV. THE TRANSCENDENT A.L DEDUCTION. 367 
 
 all the manifold, and limits it to conditions which make the 
 unity of apperception possible, and the conception of this 
 unity is just that consciousness of the object ( = X) which I 
 think through the predicates contained in the definition of a 
 triangle." ^ The conception of anything as an object is, in 
 fact, nothing but the recognition of a rule according to which 
 the manifold of perception is put together in it.^ Hence, it is 
 a mere tautology to say that we must always proceed according 
 to the same rule of construction, when we are dealing with the 
 same object. The universality of the rule and the objectivity 
 of the conception are different words for the same thing. 
 
 We may put Kant's reasoning in the following way. Percep- tms is the 
 tions must come to me in such a way that I can unite them ^^i^Hhe""* 
 with the consciousness of self. But I can unite them with upon our * 
 the consciousness of self only if I can recognise the identity of 
 the act whereby I combine their manifold, i.e., if I can recog- 
 nise that it is one conception which throughout guides me in 
 putting together the successively given elements of the manifold. 
 Now, it is just conformity to such a conception that makes me 
 refer the manifold, to an object. Why ? Because in a mere 
 manifold of separate impressions there seems to be no necessity 
 of their coming together according to any rule, and when I 
 discern a rule in their relation, I think of something as con- 
 straining them to this definite connexion. But this constraint 
 is not laid upon the perceptions from without but from within, 
 as it is the necessary condition under which alone I can be 
 conscious of them as mine. I cannot be conscious of them at 
 
 'A. 105. 
 
 ^ To meet an objection which may probably occur, observe what Kant says 
 in the immediate context. " It is this one consciousness, which combines 
 the manifold, successively perceived and then reproduced, into one idea. 
 The consciousness in question may often be weak, so that it is observed 
 only in its effect and not in the activity itself (»'. e. , we do not connect it imme- 
 diately with the process by which the idea is produced). Nevertheless, such a 
 consciousness must always be present, whatever it may want in vividness ; for 
 without it conceptions, and therefore a knowledge of objects, would be quite 
 impoMible." A. 103. , . 
 
368 THE CEITIQUE OF PURE BEASON. BOOK I. 
 
 all as isolated elements of perception, but only as I combine 
 them, and recognise the principle of their combination as 
 necessitating that they should be combined in just this definite 
 way. Such a principle alone fills up, and gives meaning to, 
 the empty conception of an object as something which is 
 different from our ideas or perceptions, but which nevertheless 
 determines them to stand in a particular relation to each other. 
 For we are forced to refer our ideas to objects, just because we 
 are forced to combine them according to definite and un- 
 changeable rules, in order to be conscious of them as ours. 
 And the consciousness of them as ours is, therefore, at the 
 same time the consciousness of these rules as determining them 
 a priori. Hence, it would appear that we are conscious of an 
 idea as ours only when we have referred it to an object ; for 
 only as it is an idea of something can it be recognised as our 
 idea. 
 The subjective The difficulty of Understanding this is that the "recogni- 
 
 in one sense ./ t;> ^ 
 
 aShl?^ ™ tion " in question seems to consist in referring to something 
 the'objeotive. clsc ihafh the sclf what for the very first time by this very 
 recognition we become capable of appropriating to the self. 
 Kant, in fact, speaks, in the first instance, as if he were ex- 
 plaining how that which consciously is subjective ('our ideas') 
 becomes referred to objects ; and, therefore, he seems to con- 
 tradict himself when he goes on to say that our ideas are 
 capable of being recognised as our ideas only when they are 
 thus referred. But by the purely perceptive consciousness the 
 matter of sense is not really referred to the self at all : its ideas 
 are not for it determined as its own ideas, as something con- 
 sciously subjective which requires to be objectified. Eather, 
 as Kant shows, they must be objectified, in' order that they 
 may be determined in relation to the subject. The being, for 
 whom there are no objects, is not for itself a subject : and our 
 ideas are consciously ours only when they are ideas of objects. 
 Language, however, " bewrayeth " us, when we talk of the 
 elements of our conscious experience apart from each other: 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 369 
 
 for we are thus obliged to speak of them as if, thus taken 
 apart, they had that determination which they gain only in 
 becoming such elements. What we call subjective, merely in 
 view of its not being determined in relation to objects, is not 
 consciously subjective ; and it can become consciously subjective 
 only in being so determined. In like manner, Kant allows 
 himself to speak of perceptions in general as determinations of 
 inner sense, which they are not for the subject of them, if we 
 mean, " consciously inner " : for they can be consciously inner 
 only as distinguished from, and related to, the objects of outer 
 sense. And, in fact, according to the Kantian doctrine just 
 explained, it is just the reference of perceptions to objects, — 
 a reference by which they in one sense cease to be merely 
 subjective, — which makes it possible, in another sense, to refer 
 these perceptions to the self as its inner or subjective ex- 
 perience. 
 
 When this explanation is made, a new light begins to be ^1™™'™ 
 thrown on some parts of the Kantian analysis of knowledge, sumptiSi'rf'' 
 especially on the priority which Kant ascribes to the synthesis of under ooncep- 
 imagination. It was his view, as we have seen, that imagination 
 h&a first to combine the manifold of sense in certain ways in 
 perception, and that then the understanding goes on to " bring 
 this synthesis to conceptions." Now, two interpretations might 
 be given to such language. It might mean merely that the 
 imagination acts on certain principles, which the understanding 
 then brings to self-consciousness. Imagination, on this view, 
 is the same faculty working blindly and unconsciously, which 
 we call understanding, when the principle of its action is 
 brought to light and consciously applied. There are many 
 passages in Kant which seem to favour this view. On the 
 other hand, it may be urged that Kant often speaks of the 
 understanding, not merely as bringing us to a consciousness of 
 rules which imagination and perception obey, but as itself the 
 source of the rules, under which it then proceeds to subsume 
 
 the perceptions (as if perceptions in themselves had a nature 
 VOL. I. 2 a 
 
370 THE CRITIQUE OF PURE REASON. book i. 
 
 altogether independent of the conceptions under which they 
 are subsumed). Now, whichever of these interpretations we 
 adopt, we have to encounter serious difficulties, — difficulties, I 
 may add, which lie not merely in Kant's language but in the 
 nature of the subject. For, if we take the former view, that 
 understanding simply analyses the unities produced by the 
 synthesis of perception, we shall have to ask what can be 
 meant by an implicit or unconscious synthesis ; and how, if 
 there be such a synthesis, its products can be brought before 
 the understanding to be analysed by it. And supposing this 
 to be possible, we shall have further to ask what good can 
 come out of a merely analytic process, and how it can be said 
 to add to our knowledge ; a fortiori, how such a process can 
 be said to make all the difference between a mere ' imdeter- 
 mined ' perception which is not knowledge, and a perception 
 determined by conception which is knowledge. On the other 
 hand, if we adopt the latter view, we have to face the opposite 
 difficulty that the perceptions, being given or produced alto- 
 gether apart from the action of the understanding and its 
 conceptions, can have no necessary correspondence with these 
 conceptions. Nay, we might rather conclude that they cannot 
 possibly have any such correspondence, since, ex hypothesi, it is 
 the application of these conceptions which has for the first time 
 to give to perceptions the determination in virtue of which they 
 are referred to objects. It would seem, therefore, that we are 
 between the horns of a dilemma. For either there is nothing 
 in the perception akin to the conception under which it is to 
 be subsumed, in which case there is no possibility of so sub- 
 suming it ; or, if there is such an affinity and such a possibility 
 of subsumption, then the perception as such must have already 
 belonging to it every qualification which it could get from the 
 conception so subsumed. 
 The appear- In stating the matter thus, we have revived the ghost of a 
 
 ance of a simi- *^ ^ 
 
 totoSk""^ controversy as old at least as the time of Plato — a controversy 
 p losop y. ^jjjgjj almost inevitably arises in some form or other whenever 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 371 
 
 an attempt is made to explain the development of knowledge. 
 It was one of the favourite weapons of that dialectical scepti- 
 cism, that so-called " Eristic," which sprang out of the conflict 
 of the earliest schools of Greek philosophy, to ask how learning 
 is possible, whether by an (analytic) progress from knowledge 
 to knowledge or by a (synthetic) progress from ignorance to 
 knowledge. In the former case, it would seem to be useless ; 
 for if we know, why should we learn ? In the latter, it would 
 seem to be impossible ; for how is consciousness to go beyond 
 itself and connect the new matter with its previous contents ? 
 This difficulty was considered by Plato particularly in relation 
 to that consciousness of the universal to which alone he gave 
 the name of scientific knowledge. If we say that we rise from 
 the consciousness of the particular to the consciousness of 
 the universal, it may be answered that the particular is 
 incommensurable with the universal, and that, therefore, 
 knowledge of the former cannot be made a stepping stone 
 to the knowledge of the latter. If, on the other hand, we 
 say that the consciousness of the universal is presupposed 
 from the first in the consciousness of the particular, then it 
 may be objected that we need not seek a knowledge which 
 we already possess. 
 
 Plato's solution of the difficulty was that the alternatives of piato's solu- 
 tion of it. 
 knowledge and ignorance are not really exhaustive. There is ^^l^'^g?'*^*^ 
 
 a middle term which may be called opinion, and which psycho- '^nloTLd 
 logically is explained as a combination of sense with intelligence, " 
 while metaphysically its object is defined as a combination of 
 TO ov and to fxrj ov. An opinion, in short, is a name for a judg- 
 ment in which a genei-al idea is used to determine a particular, 
 without having itself been made an object of thought. The 
 general idea so used is not, strictly speaking, known or defined, 
 and therefore the use of it is uncertain. Opinion may be true 
 or false, but in any case it has in itself no criterion of its own 
 truth : it is right, if it is right, by a kind of madness or inspira- 
 tion#yhich can give ,no account of itself. We are able to bring 
 
 science. 
 
372 THE CRITIQUE OF PURE REASON. book I. 
 
 it to book only by calling attention to the general ideas it uses, 
 and by demanding that those general ideas or universals should 
 be defined. Such definition, indeed, can be got only out of a 
 reflexion upon the particular cases of its application ; but it 
 will not result from a mere determination of the common 
 element in all these cases ; for the immediate application of 
 the universal in opinion may be erroneous. But the separation 
 of the universal from the particulars itself throws a new light 
 upon them, or is at least a step to clear up the misunder- 
 standings upon which its erroneous application rested. The 
 movement towards definition is, therefore, not only an analytic 
 but a synthetic process, which transforms our view of the very 
 data with which we start, and which seem to be the sole 
 premises of our reasoning. But this process can be easily under- 
 stood, if we observe that our progress toward definition of the 
 universal is a progress toward self-consciousness, and that, in 
 examining the applications of the universal, we are really 
 bringing to light certain presuppositions upon which we go in 
 our particular judgments, but which it takes no little efforb of 
 reflexion to discover. In analysing an idea with a view to its 
 definition, we are, therefore, developing that idea. The idea, 
 after it has been defined, has acquired a new meaning which it 
 had not before ; and, therefore, as so defined, it brings a new 
 qualification to the particulars subsumed under it. Here, then, 
 we have an example of a process which may be regarded eibher 
 as a transition from ignorance to knowledge, i.e., as a synthetic 
 movement of thought, or from knowledge to knowledge, i.e., as 
 an analytic movement of thought, according to the point of 
 view from which we regard it. For the universal must be in 
 a sense known when it is used in a judgment about a particular; 
 yet the first aim of the Socratic elenchus is to call forth a con- 
 fession of ignorance. But this consciousness of ignorance at 
 the same time carries in it a consciousness of potential know- 
 ledge ; for it is a consciousness that the universal, for which the 
 definition is wanted, is a principle which we have been always 
 
CHAP. IV. THE THANSCENDENTAL DEDUCTION. 373 
 
 using by virtue of our rational nature, so that to know it we 
 
 have only to know ourselves. 
 
 The point, then, to which Plato brings us, is that knowledge, ^gl^Z/fn^hioh 
 
 i.e., consciousness of particular objects as determined by univer- 'a^kn^wn""'" 
 
 sals, is not reached merely by rising from the particular to the univeraai be- 
 fore, and after, 
 universal — which is impossible, for we cannot know the particu- reflexion. 
 
 lar except through the universal ; nor, on the other hand, do 
 we possess the distinct consciousness of the universal from the 
 iirst, so as to be able to apply it with complete insight. But, 
 as rational beings, we apply to particulars a universal which 
 yet we have never made the object of thought. Hence, the 
 induction from particulars, by which we rise to the definition 
 of the universal, is at the same time a process by which we 
 come to be definitely conscious of the assumptions which 
 underlie our consciousness of these particulars from the first. 
 This is the idea which Plato expresses in the myth of reminis- 
 cence. When something is recalled to our memory, we cannot 
 be said either to acquire knowledge of it, or to have possessed 
 such knowledge previously ; but our latent or potential know- 
 ledge of it is made actual. In like manner, all our progress in 
 knowledge of universal principles is, in a sense, only a dis- 
 covery of that which we already possess, of that which is bound 
 up with our rational nature and presupposed in all our deter- 
 mination of particular objects. On the other hand, if we had 
 ever apprehended the particular without the universalj it would 
 be impossible for us ever from such apprehension to rise to the 
 consciousness of any universal, by aid of which we could throw 
 new light upon the particulars. Or, to express the same thing 
 in the terms of a later philosophy, if the a posteriori were ever 
 given apart from the a priori, it would be impossible, starting 
 from the former, ever to come within reach of the latter. 
 
 Now, Kant's answer to the problem of the possibility of Kant's first 
 
 ^ X ./ solution of the 
 
 knowledge is mutatis mutandis very similar to that of Plato. oonoepMm is 
 But, as he starts from a more definitely dualistic basis, he has gmlraa^e ° 
 
 j./vi 1. 1. . n-P'Tii perception; 
 
 mojte aiinculty ui reachmg a consistent result, it indeed he ever 
 
374 THE CRITIQuJi; UJ i-unji j^jijio^/xi. 
 
 reaches it. How, he asks, is the empirical consciousness of 
 particulars as given in sense to be brought under the pure 
 conceptions of the understanding ? To this question Kant, 
 like many other writers, had at one period of his philosophical 
 development given the simple answer, that all general concep- 
 tions are derived from the particular objects of sense to which 
 afterwards they are applied/ In other words, the synthetic 
 judgments, by which the conceptions of objects are formed, are 
 empirical " judgments of perception." But the inadequacy of 
 this answer was already shown by the criticism of Locke, who, 
 though he assumed that the individual object in all its deter- 
 mination is given in sense, yet denied that the particular 
 judgments which sense authorises us to make, can be legiti- 
 mately universalised. " General knowledge of matters of fact " 
 seemed to him impossible ; and, therefore, all general proposi- 
 tions must necessarily be " trifling," or, in other words, they 
 must express only nominal essences. It is at this point that 
 Kant takes up the question in the Critique? In the Introduc- 
 tion, indeed, he still seems to regard experience {i.e., perception) 
 as the source of synthetic judgments, which are the ground of 
 our first conceptions of objects, and the only means whereby 
 these conceptions are to be made more complete — the under- 
 standing being for the most part confined to the task of 
 analysing the data which it thus receives from the hands of 
 sense. But in one point, the " empirical synthesis " thus 
 realised, is regarded as falling short of the ideal of knowledge ; 
 for it only enables us to make particular judgments, i.e., judg- 
 ments about the individual object presented to us, and not to 
 rise to necessary or universal truth. For no accumulation of 
 particulars will enable us to reach the universal: no mere 
 
 1 This was Kant's view in the period of Critical Empiricism (1763-66). It 
 will be remembered that, at the end of his Essay towards the Introduction of the 
 Idea of Negative Quantity into Philosophy (of. Introduction, Ch. IV., p. 129), he 
 implies that experience supplies us with conceptions iu which all synthesis is 
 already done to our hands. 
 
 '^ Cf. above, p. 258. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 375 
 
 inductive process will tell us what must be, and, therefore, 
 what will always be. Hence, the problem of knowledge, in 
 the strict scientific sense, is the problem of the possibility of 
 an a priori synthesis which shall anticipate experience and 
 not wait for it. Experience, therefore, is regarded as need- 
 ing a priori principles to extend it to universality. But it is 
 not as yet denied that it enables us to determine particular 
 objects as such, or to add new predicates to them : on the con- 
 trary, it is assumed that the difficulty is not in understanding 
 how experience enables us to add to our conceptions of objects, 
 but only how the pure intelligence should enable us to do so. 
 But the point of view thus expressed could not long hold its 
 ground before the advance of criticism ; and we soon find Kant 
 making a transition to another view of the relation of percep- 
 tion to conception, for which the way had been prepared by 
 Berkeley. The judgment of perception is merely, e.g., that 
 " this rose is red " ; it does not authorise us to lay down any 
 propositions about roses in general, or even about the particular 
 rose when it has ceased to be perceived. If, however, we press and, therefore, 
 
 also to objec- 
 
 home this contrast of universal and particular, it immediately ^^y *'• 
 passes into another contrast, viz., that of objective and subjec- 
 tive. Perceptions, if they are merely isolated perceptions, are 
 perceptions of something that is (or is known to be) only at the 
 moment that it is perceived, and cannot have the character of 
 objectivity ; for it is just the idea that the object persists or 
 maintains itself and its properties beyond the moment of 
 perception that makes us regard it as an object at all, i.e., as 
 something distinct from the perception itself. If we are not 
 entitled to say that what we have perceived exists as we have 
 perceived it for one moment beyond the perception, then there 
 is no ground for distinguishing it from the perception. The 
 esse of things is their per dpi, or the object and the perception 
 
 ' Note the double force of the word synthesis. Experience is synthetic, as it 
 enables us to add to our conceptions or to particularise them. On the other 
 hand, the pure intelligence'ia synthetic, as it enables us to add to our percep- 
 tions or to universalise them. 
 
376 THE CKITIQUE OF PUEE REASON. BOOK I. 
 
 fall together: and in the former as in the latter, we have 
 merely a modification of the perceiving consciousness. From 
 this point of view, the transition from perception to experience 
 must involve not merely a communication of universality and 
 necessity to judgments abmit objects, which in themselves are 
 particular and therefore contingent, but a reference to objects ' 
 of perceptions, which, as sensible perceptions, are mere ideas 
 or subjective appearances. It is this conception of the relations 
 of sensible or empirical perception to the a priori principles 
 of understanding, which Kant sets before us when, in the 
 Prolegomena, he makes his well-known distinction between 
 "judgments of perception" and "judgments of experience." 
 " All our judgments," he declares, " are at first mere judgments 
 of perception ( Wahrnehmungsurtheile) : they hold good only for 
 us, i.e., for our subjectivity, and it is only afterwards that we 
 give them a new reference, namely to an object, with the under- 
 standing that they shall hold good for us not only at the moment 
 of perception but at all other times, and not only for us but for 
 every subject ; for if a judgment agrees with its object, all judg- 
 ments about the same object must agree with each other. And so 
 the objective validity of a judgment of experience means just that 
 it is necessarily and universally valid," i.e., valid for all subjects 
 at all times. " But conversely," he goes on, "if we have reason 
 to regard a judgment as necessary and universal, (which can 
 never be due to the perception, but only to the pure conception 
 under which it is subsumed,) we must regard it also as objective, 
 i.e., we must look upon it as expressing, not merely a reference 
 of the perception to a subject, but of an attribute to an object; for 
 there could be no reason why the judgment of others should 
 necessarily agree with mine, except the unity of the object, to 
 which they all relate, with which they agree, and which, there- 
 fore, makes them. all agree with each other." ^ In a judgment 
 
 ^ Prolegomena, § 18. From the point of view here stated, it would be more 
 natural to regard the judgments of perception as analytic. Kant, indeed, says 
 in the Introduction to the Critique, that judgments of experience (the same 
 which he here calls judgments of perception) are always synthetic. But when 
 
*HAP. IV. THE TRANSCENDENTAL DEDUCTION. 377 
 
 of perception, therefore, I " unite perceptions in one conscious- 
 ness of my state," while in a judgment of experience, I " unite 
 them in one consciousness in general"; i.e., without reference to 
 me as an individual subject, or to the special circumstances 
 which condition my perception at a particular time. 
 
 The view thus expressed involves at the outset an entire objectivity i» 
 
 ^ thus reached 
 
 inversion of the usual way of combining the two ideas of u„™"fg'JjiHy 
 universality and objectivity. For, what Kant here teaches categories give 
 
 to particular 
 
 is that we cannot reach universality through obiectivity, but juasments of 
 
 ■^ o (J ./ ' perception. 
 
 must reach objectivity through universality ; or rather, we 
 
 must say that the only true objectivity is universality. Or, 
 
 to bring the question more closely into relation with Kant's 
 
 mode of expression, the categories can determine perceptions 
 
 in relation tO; objects only in so far as they give to the 
 
 judgments of perception that generality which implies their 
 
 independence of the particular perceptions, and even of the 
 
 individual mind as perceiving. We must no longer say that 
 
 our perceptions agree with each other because they agree with 
 
 the same object, but we must say that they agree with the) 
 
 same object because they agree with each other.^ Now, what 
 
 does such agreement mean ? It means that we can assert that 
 
 a judgment of perception which has been once made, is always * 
 
 he says this, he is looking at the subject from the point of view of the under- 
 standing and its doaoeptions ; and he therefore regards synthetic judgments as 
 those in which the understanding goes beyond its own subjective unity, and 
 beyond the conceptions which arealready joined with that unity, to determine 
 objects or to add new elements to the content of the conceptions of them which 
 it already possesses. But in the Prolegomena, Kant has been brought by the 
 turning movement of hia dialectic to a new point of view, from which the 
 starting point lies not in conceptions but in perceptions. Hence, from this 
 point of view, perceptions are supposed of themselves to enable us to make 
 judgments, which of course are merely analytic of the perceptions. Here, 
 therefore, synthesis appears only in so far as the u, priori conceptions of the 
 understanding enable us to go beyond the particular judgments of perception, 
 and to turn them into universal judgments of experience. Of course, the 
 synthetic nature of the judgments of perception will reappear whenever, by 
 another turn of Kant's dialectic, it is recognised that we must go beyond the 
 data of sense in order to make any judgment whatsoever. 
 ^ TWs reasoning is contained in A. 104, which has been already quoted, p. 365. 
 
378 THE CRITIQUE OF PURE REASON. book i. 
 
 capable of being made under the same conditions, because it 
 expresses or involves a relation which is universal. It means 
 that that judgment is recognised as expressing a general law, 
 according to which the elements combined in it are always to 
 be combined whenever they appear. What, therefore, from this 
 point of view the category must add to the judgment of percep- 
 tion to make it a judgment of experience is just the recognition 
 that it expresses a relation, which must always be expressed ia 
 the same way ; a relation, therefore, which is independent of 
 the particular perception and the particular perceptive sub- 
 ject who makes it, and holds good for ' consciousness in 
 general.' And this is all that is meant by an objective 
 judgment. 
 But is even a When, howcvcr, we have reached this point, a new difficulty 
 iosfSio^with- presents itself ; for it would seem impossible that even a par- 
 oategorieB? ticulttr judgment should be made without the aid of the 
 category which is supposed only to be required to universalise 
 it. In a judgment the particular always presupposes the 
 universal, and the moment we assert, say, that an object has a 
 quality (e.g., This ball is elastic) we express a universal relation. 
 For, as has been just said, the attribution of a quality to a 
 substance has no meaning, if the substance be not something 
 which exists beyond the moment of perception, and if there be 
 no definite relation between it and the quality, except that it 
 with the other qualities attributed to the same substance forms 
 at a particular moment a complex image of sense. It is just 
 the consciousness that the elements of the idea of perception 
 are combined in a way which remains the same for the thinking 
 subject when the moment of perception is past, or which at 
 least points to a relation that holds good beyond that moment, 
 that makes it possible to express that unity in a judgment at 
 all. This seems to be recognised by Kant himself, when, in 
 the Deduction of the second edition, he points out that the 
 " is," which is the copula of judgment, is the expression of the 
 objective unity of apperception. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 379 
 
 " I have never been satisfied," Kant tliere declares, " with k™* fi^a^y 
 
 declares that 
 
 the explanation of judgment given by the Logicians, that it is IsUoh^in-™* 
 the idea of a relation between two conceptions. I will not j'e''ctiTCuifityof 
 
 ... T , . , . apperception. 
 
 now raise the objection that this explanation is defective, as 
 being adapted only to the categorical, and not to hypothetical, 
 or disjunctive, judgments (which latter express a relation not 
 of conceptions but of judgments) ; though this oversight leads 
 to many awkward consequences. But I must point out that it 
 tells us nothing about the nature of the relation in question. 
 When, however, I investigate more accurately the relations of 
 the elements of knowledge brought together in every judgment, 
 and when I distinguish it, as belonging to the understanding, 
 from a relation of association, established according to the laws 
 of reproductive imagination (a relation which has only subjective 
 vaHdity), I find that judgment is nothing but the process by 
 which we bring given ideas to the objective unity of appercep- 
 tion. This is just what is implied in the use of the verb 
 of existence as the copula. The use of this verb points to the 
 fact that we are dealing with the objective unity of given ideas, 
 as distinguished from their subjective association in our minds ; 
 for it indicates that they are brought into relation to the 
 original apperception and its necessary unity, and that even 
 where the judgment itself is empirical and, therefore, accidental. 
 Thus in the judgment : " Bodies are heavy " I do not mean to 
 assert that' "body" and "weight" as empirically perceived are 
 necessarily comUned with each other : but I do mean to assert^ 
 that they are combined with' each other by means of the neces- \ 
 sary unity of apperception in the synthesis of perceptions, i.e., 
 thfey are combined according to the principles of objective 
 determination, which are deduced from the one ultimate prin- 
 ciple of the unity of apperception, under which all ideas must 
 be brought, if knowledge is to be derived from them. Only by 
 subsumption under such principles can the relation of two 
 ideas become one which admits of being expressed in a judg- 
 ment,* i.e., a relation which is objectively valid; whereas a 
 
380 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 relation of ideas determined by the laws of association could 
 only have subjective validity. For principles of empirical 
 association would not authorise me to say more than that 
 " when I carry a body, I feel a pressure of weight " ; they could 
 never authorise me to say that " it, the body, is heavy." For 
 the latter statement is equivalent to the assertion that these 
 two ideas are bound together in the object (i.e., altogether apart 
 from the state of the subject perceiving them) and not merely 
 that in our sense-perception they have been present together, 
 however often it might be."^ 
 menfS'mere" Thcrc is a Certain inconsequence of statement in this passage, 
 lossifie.™^^ in so far as Kant seems to admit that we are able to make a 
 judgment as to the succession of our own sensitive states under 
 conditions in which, as he at the same time contends, we are 
 not able to make any judgment at all.^ For, on Kant's piin- 
 
 'B. 141, §19. 
 
 2 There is a farther inconsequence in Kant's statement, to which reference has 
 already been made (p. 348 note). In the first edition, Kant omits to connect 
 the transcendental, with the metaphysical, Deduction of the categories ; for he 
 does not show that the principles derived from the analysis of judgment are 
 necessarily those by which objects are determined as such. He simply shows 
 that, unless there were a connexion of perceptions according to universal and 
 necessary laws, these perceptions could not be referred to objects ; and then 
 he asserts that the categories supply the principles required for this purpose. 
 " I assert that the categories are the conditions of thought for a possible 
 experience, just as space and time are the conditions of the perceptions which 
 supply the matter for the same experience." (A. 111.) No attempt is here 
 made to show from the nature of the synthesis whereby experience is con- 
 stituted that his categories are just those required. The Deduction of the 
 second edition, as we have seen, supplies this want, by showing that judgment 
 is essentially the expression of the objective unity of apperception. For this, 
 as Kant argues, entitles us to take the list of logical "functions of unity" in 
 judgment, as a complete index to the metaphysical system of categories. 
 Unfortunately, the new piece heals the rent in the old garment only to 
 produce a worse rent in another place. For the definition of judgment, as 
 being essentially the expression of the objective unity of knowledge, cuts away 
 the ground from a metaphysical deduction based on the very definition of 
 judgment which Kant here rejects, viz., that it is an "idea of the relation 
 between two conceptions." Kant's argument, if it shows that the forms of 
 judgment may be a guide to us in determining the categories, at the same time 
 throws more than a suspicion upon a list of categories derived from what now 
 appears to be a wrong view of judgment. I^ judgment is essentially synthetic, 
 
CHAP. IV. THE TRANSCENDENTAL DEDTJOTION. 381 
 
 ciple, the sequence of one feeling on another, though it may 
 tend to associate them, will not enable us to make any judgment 
 about these sensations, even that they are sequent in our 
 experience; any more than it wiU enable us to make the same 
 sequence the ground of a judgment as to any other object. 
 But putting aside this inconsequence, the purport of the 
 passage is that perception makes no judgments, but that what 
 is caUed a 'judgment of perception,' (or in the Introduction 
 to the Cfitique, a 'judgment of experience,') is really due to 
 the subsumption of the manifold of perception under the unity 
 of apperception, as expressed in the categories. As confirming 
 this view, we may observe that the judgment, "Bodies are 
 heavy," which is here declared to be based upon an a priori 
 principle of synthesis, is one which in the Introduction was 
 treated as an a posteriori synthetic judgment. Either, there- 
 , fore, the Introduction must be taken as a merely provisional 
 statement, the inaccuracy of which is purposely admitted by 
 Kant in order not to embarrass the reader too soon with the 
 ultimate results of his thought ; or at least, it must be admitted 
 that Kant afterwards recognised it as such a statement. For 
 in the manuscript notes of Kant upon the first edition of the 
 Oritique^ which have been lately published, we find him saying 
 that, when in the Introduction to the Critique he had asserted 
 that experience " consists of synthetic judgments," he had 
 merely been referring to the fact that there are such a 
 posteriori synthetic judgments, and he had not, therefore, found 
 it necessary to raise the question, how they are possible. 
 " Now, however," he goes on, " we may raise the question how 
 this fact is possible. Experience consists of judgments, but the 
 
 it is absurd to suppose that the conceptions which were reached by those who 
 took judgment as an analytic process, must be the categories, or even that they 
 supply a guiding thread to find them. 
 
 * Erdmann, NacHrage zu Kant's Kritih, p. 21, § 33. This passage is quoted 
 by Vaihinger (I. 357), where alsd the above mentioned change in Kant's view 
 of the judgment, "Bodies are heavy," is noted ; and the other passages bearing 
 on the subject are collected and discussed. 
 
382 
 
 THE CRITIQUE OF PURE REASON- 
 
 BOOI 
 
 Is even the 
 consciousness 
 of particular 
 images of 
 perception 
 
 ■without the 
 synthesis of 
 the under- 
 standing? 
 
 question is whether these empirical judgments do not final 
 presuppose a _pr'iori judgments." And he immediately proeee 
 to sketch out a proof that they do, which seems to be 
 anticipation of the ideas expressed in the Deduction of t 
 second edition. 
 
 But if we cannot derive from sense, as separated from t 
 understanding, any " judgment of perception " in which defini 
 objects are determined by particular qualities as their prec 
 cates, can we derive from it even an image of perception su( 
 as might supply the subject for such a judgment ? Such 
 view seems to be suggested in various passages, in whic 
 Kant says that sense presents to us images or appearance 
 which in themselves are only modifications of our own sul 
 jective state; and that so long as we confine ourselves to thai 
 appearances as subjective phenomena and make no assertion i 
 to their objective meaning, no a /priori principles of unde: 
 standing are required. Such principles are needed only whe 
 we proceed to refer those appearances to objects which are n( 
 states of our consciousness. Now, if we set aside for tl 
 moment the question how " mere appearances " come to I 
 recognised as such, i.e., how they come to be recognised by x 
 as modes of our own consciousness which have no objecti^ 
 validity, it seems reasonable to say that there can be no doul 
 of the subjective reality of the phenomena that are presents 
 to us by sense, whatever doubt there may be about the: 
 objective reality. There can be no doubt that appearancf 
 appear, that mental representations are in our minds, or ai 
 presented to us in sense. So long as we conceive such phenc 
 mena as a mere phantasmagoria passing before our menfc 
 vision, and do not ask any question, or make any assertion, £ 
 to their correspondence with any object beyond themselves, s 
 long, it would seem, we cannot be deceived. From this poii 
 of view Kant says that " the senses set the planets before u 
 now as moving onward, and now as reversing their course 
 and in this there is neither truth nor falsehood, so lor 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 383 
 
 as we are content to regard all we see as mere appearances, and 
 to make no judgment in regard to the objective movements." ^ 
 The question of truth or reality arises only when we go beyond 
 the appearances, and make a judgment in which they are 
 referred to an object. So long as the mind passively appre- 
 hends that which is presented to it, so long it cannot err ; for 
 so long there exists for it no distinction between appearance and 
 reality, and therefore no possibility of mistaking the one for 
 the other. To render such mistake possible, the mind must 
 be active ; it must go beyond what is immediately given in 
 sense and refer it to some object, which perception may 
 represent but which it does not exhaust, and with which, 
 therefore, it is not immediately identical. This object we do 
 not need here to determine, except by saying that it is some- 
 thing which is independent of the particular perception of it, 
 something which exists apart from its appearance. The 
 immediate perception must be referred to something not 
 perceived, must be compared with it, asserted to agree or 
 not to agree with it, ere we can pronounce that it is either true 
 or false, that it gives or does not give knowledge. 
 
 When, however, we consider the matter more carefully, we ^° ■ ^'"" *''® 
 
 J ' consciousness 
 
 see that the statement just made is not strictly accurate, or at ano°aTS" 
 least is Hable to mislead. For it might naturally be understood judgment, 
 to imply that we first have the consciousness of phenomena as 
 mere appearances to us, and then on certain principles refer 
 them, or some of them, to an objective reality. But this is so 
 far from being true that, on the contrary, we must first have a 
 consciousness of reality as a connected system of objects, ere, in 
 opposition to this system, we can think of a particular pheno- 
 menon as a mere appearance or mode of our own consciousness. 
 And even when that thought arises, it cannot go the length of 
 an absolute denial of the objective reality of the appearance in 
 question. All it can mean is that a particular appearance has 
 been wrongly interpreted, or put in a wrong place in the con- 
 % iR. in. 48; H. IV. 39. 
 
384 THE CRITIQUE OP PURE REASON. book I. 
 
 text of experience. In the example given, the question 
 supposed to be left undecided is, whether certain apparent 
 movements of the planets are real. But if this be taken to 
 mean that we are hesitating as to their reality, it is obvious 
 that such a consciousness could arise in us only in the form of 
 a doubt, whether a certain interpretation, which had already 
 been given to the sensations, were the right interpretation; 
 whether, e.g., the movements in question had not wrongly been 
 referred to the planets, instead of being referred to the earth on 
 which the spectator is placed. Or a larger doubt might arise, 
 and it might be questioned whether the whole phenomena 
 present to us were not to be attributed to an " optical illu- 
 sion," i.e., to some abnormal action of the organs of sense 
 excited by the sensuous imagination of the individual; such as 
 we may suppose in the case of a ghost-seer, like Swedenborg. 
 In all such doubts, however, what is iu question is, not the 
 objective consciousness in general, but only the place of certain 
 phenomena in it. We find that we cannot combine certain 
 phenomena with the general context of experience, unless we 
 alter our first interpretation of them ; and, therefore, we say 
 that they are illusory. But what was illusory was- the inter- 
 pretation, i.e., a mental synthesis which we now find to come 
 into collision with other results of mental synthesis which are 
 more general or better established. The question is one 
 touching the particular, not the universal : it relates, strictly 
 speaking, not to the reality of the facts, but to their position in 
 the context of experience. For an illusion merely means a reality 
 referred to the wrong place in that context ; and the assertion 
 that it is a mere appearance does not mean that it is purely 
 subjective, but merely that the manner of its objectivity is as 
 yet for us undetermined. To suppose, however, that we can 
 go back to certain perceptions or appearances to which no 
 objective interpretation has yet been given, is to suppose that 
 we can have knowledge of that which is " for us as thinking 
 beings as good as nothing." It is " to make oneself in thought 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 385 
 
 into an animal," and yet to attribute to that animal a conscious- 
 ness of self in opposition to objects. 
 
 We have, therefore, here again to guard against the mistake Perception is 
 
 then only the 
 
 of attributing to sense, or to what we caL. its object, those F^^=™°^°?^^° 
 characteristics which it can have only for a subject which is subfeotail'^* 
 not merely sentient. It may be true to say that sense has a thing, 
 content which is merely appearance, i.e., which consists of 
 elements that have not yet been so determined by thought as 
 to yield a consciousness of objects ; but it is not true that such 
 a content is determined for us by sense as an appearance, or, 
 indeed, as anything whatsoever. Now, the former is obviously 
 Kant's meaning in the beginning of the passage which we have 
 quoted, where he says that "appearance rests on the senses, the 
 judgment regarding the appearance on the understanding," and 
 that " the distinction between reality and dream lies not in the 
 character of the ideas presented (for these are in both cases 
 the same), but in their combination according to the rules 
 which determine the connexion of ideas in the conception of an 
 object, and which therefore form our guides in discovering how 
 far these ideas can be put together in an experience. It is, 
 therefore, not the fault of the appearances when we take 
 illusion for truth, i.e., when the perception, through which an 
 object is given, is confused with the conception of the object 
 or of its existence, which can only be thought by the under- 
 standing." ^ Yet Kant's example, as well as his expressions 
 about an " object being given," and even the word " appear- 
 ance," are apt to suggest another form of the same error 
 which he is combating ; namely, that we are conscious of the 
 appearance as merely an appearance, before we are conscious 
 of any reality. 
 
 When we have got thus far, we begin to see how Kant, But even an 
 
 (■(.••II -I- T 1. image, as a 
 
 siartmg with the common empirical view, according to whicn nnity ofmani- 
 
 jT ' o foj^ elements, 
 
 all our conceptions are derived directly from experience and can tynthSof 
 only be analysed, and not amplified or extended by thought, aten(Sn|r 
 
 • 1 R. III. 47 ; H. IV. 39. 
 
 VOL. I. 2 B 
 
386 THE CRITIQUE OF PURE REASON. book I. 
 
 was led, in the first place, to admit an exception to this rule 
 in the case of certain a priori principles, from which all the 
 truths of mathematics and the primary laws of physical science 
 are derived ; how, then, in the second place, he was brought 
 gradually to limit the range of the a posteriori synthesis of 
 perception — first, to judgments about the particular and contin- 
 gent, then to judgments which express only the staiiis representari 
 tivus of the subject, and then to images of sense which do not 
 in themselves imply any act of judgment at all. The last step 
 which it remained for Kant to take, was to recognise that even 
 the images of sense imply a synthesis of pure or empirical data, 
 which is possible only to the understanding, and that in itself 
 sense can only supply a " manifold." And this is the view 
 to which he brings us when he declares that " combination 
 is the one thing which cannot be given through the object, but 
 must be realised by the subject for itself." Hence " analysis 
 presupposes synthesis," and "the understanding cannot analyse 
 (or resolve into simpler elements) any combination which it has 
 not itself made ; for no complex whole could be given to the 
 mind, except as combined by the understanding itself" But 
 if this be so, and if we are compelled to reduce the data of 
 sense to an unconnected manifold, which has not even the con- 
 nexion necessary for an image of sense or imagination, we are 
 thrown upon the other horn of the dilemma with which we 
 started. The activity of the understanding which, in the purely 
 a,nalytic view of it, was useless, has gradually grown in 
 importance, till it is seen that without it neither a singular and 
 subjective judgment of perception, nor even an image of per- 
 ception, is possible. But as the synthesis of the understanding 
 becomes indispensable, it seems also to become impossible ; for 
 we have gradually removed from the perception all those 
 elements in virtue of which it was possible to subsume it under ' 
 a conception : and between a mere " manifold " as such and the 
 unity of thought, (supposing them once to, be separated,) it is 
 impossible to see how any link of connexion could be established. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 387 \ 
 
 This dif&culty becomes still greater when we look at the 4i*'"j °° *''" 
 
 •' ° side of ooncop- 
 
 metaphysical deduction of the categories, and take into account principle cm!*'° 
 the way in which in that deduction Kant accepts the doctrine from pure 
 
 thought. 
 
 of formal Logic — that the movement of thought in itself is purely 
 analytic, and that it becomes synthetic only in relation to per- 
 ception. Enough has already been said on the impossibility of 
 deriving the categories as principles of synthesis from the 
 "functions of unity" in the analytic judgment. And it is 
 only necessary here to refer to the way in which Kant tries to 
 make this derivation more intelligible in the Prolegomena and 
 the second edition of the Critique. In, the latter he had, as we 
 have seen, practically surrendered the idea of an analytic judg- 
 ment, when he declared that judgment is " that action of the 
 understanding whereby the manifold of given ideas (be they 
 perceptions or conceptions) is brought under an apperception." 
 But from this he goes on to argue, not that the idea of an 
 analytic judgment is illusory, but merely that in a synthetic 
 judgment, " all the manifold, so far as it is given in an empirical 
 perception, must be determined in view of one of the logical 
 functions for judgment." The meaning of this expression is 
 explained by a previous passage, in which Kant says that the 
 categories are conceptions of an object in general, by which its 
 perception is determined in view of one of the logical functions 
 for judgment. Thus the function of the categorical judgment is 
 to express a relation of one idea to another as that of subject to 
 predicate. But in so far as the logical use of the understand- 
 ing is concerned, it is indifferent which of the two ideas should 
 take the place of subject and which of predicate. I may, e.g., 
 say either that " all bodies are divisible," or that " some divisible 
 things are bodies," as I please. " But by the category of sub- 
 stance, when I bring the conception of body under it, it is 
 determined that the empirical perception of a body must always 
 be viewed in experience as a subject, and never as merely a 
 predicate. And so it is with all the other categories." ^ In 
 ♦ IB. 129. 
 
388 THE CRITIQUE OF PURE REASON. book i. 
 
 other words, Kant, notwithstanding all he has said of the judg- 
 ment as the expression of the objective unity of apperception, 
 still supposes that we can take it, according to the conceptions 
 of formal Logic, as the expression of a relation of mere ideas in 
 our minds: in which case we can abstract from all relations of 
 subject and predicate, except that which they have as analyti- 
 cally contained, or not contained, in each other. I have, 
 however, already shown that siich a view of the judgment 
 cannot stop short of emptying it of all meaning whatever ; or, 
 in other words, that such a view must end in an identity in 
 which there is no room for judgment at all. And Kant himself 
 practically acknowledges that purely analytic thought is not 
 thought at all, when he says that analysis has a meaning only ' 
 in relation to a presupposed synthesis. In the above passage, 
 Kant seems to maintain that conversion is permissible from 
 the point of view of Logic, i.e., when we view subject and predi- 
 cate as mere thoughts, (for then we abstract from all except the 
 question whether the one idea is analytically contained in the 
 other, and, therefore, it does not matter which idea occupies! 
 the place of the subject) ; but that it is not permissible when 
 the judgment is taken as expressing an objective relation ; for 
 then the position of subject and predicate corresponds to, and ; 
 expresses, the relation of substance and accident. To this it , 
 might be answered that the position of subject and predicate 
 is not indifferent even logically, unless we quantify both, and 
 say " some divisibles are all bodies," or, " all bodies are some 
 divisibles '' ; and this, as we have seen, is a step toward that 
 extinction of the judgment, which is the nemesis of formal 
 Logic. On Kant's own showing, a judgment of pure thought; 
 is a judgment in which nothing is said ; for pure thought can' 
 say nothing but that "A is A." Only such a judgment can 
 be converted without change of subject or predicate, but that 
 is just because it is no judgment at all. On the other handi- 
 if we admit any difference between subject and predicate in 
 the judgment, we necessarily take it as the expression of a 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 389 
 
 synthesis, and, therefore, as the assertion not merely of a 
 subjective relation of thought, but of an objective relation of 
 fact. 
 
 In this way we are brought to the conclusion that the H™<=e ooncep- 
 
 •' ° tion and per- 
 
 abstraction of an analytic intelligence, which becomes synthetic to^beleit 
 only in relation to the manifold of sense, is as unreal as the dSfnton. 
 abstraction of a manifold of sense, which is in itself uncon- 
 nected, and which derives all its connexion from the action of 
 the understanding. In addition to the difficulty of bringing 
 the manifold of sense together in such a way that it may be 
 subsumed under the pure conceptions, Kant's conception of the 
 analytic nature of thought loads him with the further difficulty 
 of making the barren unity of the intelligence develop into 
 conceptions of relations under which the associated manifold 
 may be brought ; and thus he becomes entangled in a problem . 
 the exact reverse of that with which, in the first instance, he 
 had been dealing. The gradual process of abstraction by which, 
 on the one side, he has emptied the data of sense of all prin- 
 ciples of connexion, and, on the other side, has deprived pure 
 thought of all claim to synthetic movement, has enabled him 
 to refute both the rationalistic and empiricist views of the 
 origin of knowledge, and to prove that both factors are neces- 
 sary to experience. But, while he has shown that both factors 
 are indispensable, he has made it harder than ever to explain 
 how they can be brought into that relation, which is required 
 for knowledge. Are we then to say that in the energy of his 
 counter-attack upon the enemies who were threatening him 
 upon either wing, Kant has broken his army in pieces, and left 
 the centre undefended ? In showing that neither Eationalists 
 nor Empiricists can explain knowledge, has he left room for 
 himself to explain it by uniting the elements which they 
 separated ; or has he left nothing but the disjecta membra, to 
 which it is impossible to restore the unity of life ? In any 
 case, it is this recombination which he attempts in the Trans- 
 cend^tal Deduction. 
 
390 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 ttL^noir^^ Now, up to a certain point, it is possible to explain the 
 dXtog^that emphasis which Kant lays on the opposition of perception and 
 tionOTMncTp- conception, sense and understanding, without supposing that he 
 
 tion exist out 
 
 °t*'?6uinty^of admits anything inconsistent with their necessary unity, as 
 factors in one experience. For, on the one hand, it is only as 
 separated from perception and at the same time referred to it, 
 that conception gets its power as a principle of knowledge ; '■. 
 and in like manner, on the other hand, it is only as separated; 
 from, and at the same time referred to, conception that percep- 
 tion becomes a datum to which conception can be applied. 
 Knowledge, Kant rightly maintains, is judgment, and judgment 
 implies the determination of perception by conception. It seems, 
 however, as if this determination could not take place unless 
 perception had already in it the unity expressed in the con- 
 ception ; i.e., unless perception already had the unity which it 
 cannot have till it is determined by the conception. And it 
 seems also as if conception, in order to determine perception 
 must have already in it a diversity, or a relation to diversity, 
 which it cannot have till it determines the perception. The 
 act of judgment would thus involve, from one point of view, a 
 combination of perception and conception which presupposes 
 their differences ; and, from another point of view, a distinction 
 of conception and perception which presupposes their unity ; 
 for each of its elements can be related to the other only as it 
 has the other implicit in it. Or, to put it otherwise, conception 
 and perception both appear to presuppose the judgment, yet 
 they are the very elements brought together in the judgment. 
 a^tSmedT ^his difficulty, as has been already indicated,^ Kant tries 
 them.* ^°'^° to get over from both sides by the interposition of a third 
 element, which appears on the stage without notice, viz., the 
 imagination. On the side of perception, imagination is the 
 power which combines the manifold into an image, which is 
 thus made capable of being brought under the unity of under- 1 
 standing : while on the side of understanding, it schematises 
 
 ^ Above, p. 327. 
 
CHAP. IV. THE TRAJSrSCENDEJSTTAL DEDUCTION. 391 
 
 the categories and so brings them into potential relation with 
 the manifold of sense. In virtue of the former synthesis, time, 
 space, and all the " manifold " elements given under both these 
 forms, are combined into definite individualised images. In 
 virtue of the latter, the forms of analytic judgment become 
 principles of unity which can be applied to those images, and 
 in relation to which, therefore, those images may be " brought 
 to conceptions," or recognised as containing a unity that cor- 
 responds to conceptions. Further, if we ask how it is that the 
 principles of unity in perception and in conception correspond 
 with each other, Kant points to the fact that both are derived 
 from the same self, which as sentient perceives, and as intel- 
 ligent conceives. 
 
 Now, what is indicated by this complex system of mediation? Kant's way of 
 
 ''■*■*' reaching the 
 
 It points to that circle of presuppositions which I have already preSppo^d 
 expressed by saying that, while it is only as referred to each ' ^'^^''^■ 
 other in the judgment that perception and conception exhibit 
 those characteristics by which we distinguish them, yet, on the 
 other hand, the judgment presupposes both perception and con- 
 ception as elements which it refers to each other. This circle, 
 however, is simply the result of the attempt to trace back the 
 unity of the judgment to two elements which yet, as such 
 elements, have no existence apart from that unity. To get over 
 this difficulty, Kant first recognises that the unity which ex- 
 presses itself in conception is present also in perception, as it is 
 the same self which at once conceives and perceives. Yet, 
 secondly, he is compelled to add that this unity is not' present 
 in perception and in conception in the same way ; for in the 
 former it is only a principle of association, the highest product 
 of which is an individual image, while in the latter it is a self- 
 conscious unity, which, therefore, expresses itself in " concep- 
 tions of objects in general." Lastly, he conceives the integration 
 of the two elements as a subsumption of the image under the 
 conception, which is schematised for that purpose by the 
 prodflictive imagination. But '' it is one and the same spon- 
 
392 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 taneity" which, "under the name of imagination," produces 
 the empirical synthesis of apprehension, and, " under the name 
 of understanding," subsumes the perception so produced under 
 the conception. We have, therefore, a ' pre-established har- 
 mony ' of the faculties and their products, based on the unity 
 of the self which underlies each of them and which brings 
 together the products of both in the unity of knowledge. For 
 the self, in this double manifestation of its activity, must always 
 remain in harmony with itself, as otherwise it could not be 
 conscious of itself as a self at all. And it is only' as the image 
 is recognised as conformable to the categories, by which the 
 understanding determines objects in relation to the self, that it 
 can " be united with the ' I think ' of consciousness." 
 Shed'^hw-**''" ^^® ™°^* important point to notice in this theory is the way 
 b^k'to?"'* in which Kant, on the one hand, distinguishes the unity of per- 
 ception from the unity of conception ; yet, on the other hand, 
 maintains the ultimate identity of the principle manifested in 
 both. It is, he says, due to the same spontaneity which, how- 
 ever, acts in different ways in the two cases ; in the one case, 
 by an unconscious activity, gathering the data of sense into one 
 image, and ultimately into one imaged continuity of a world in 
 space and time': in the other case, by a conscious activity, 
 producing (in relation to this pictorial unity of perception) 
 a number of pure conceptions, which in their combination 
 constitute the idea of Nature as a system. Now Kant speaks 
 of the latter, the conceptions that spring out of the unity, of 
 the apperception, as an 'Epigenesis' upon the former,^ which 
 must mean that they are not a simple extension of what 
 existed before in the perceptive consciousness, nor an external 
 addition to it, but a new development in which the same 
 principle reaches what Schelling called a 'higher potency.' 
 If apperception is thus a development of the same activity 
 present in perception, it is not difficult to understand how the 
 images of the latter should be adapted a priori to conceptions 
 
 1 B. 166. 
 
<!HAP. IV. THE TRANSCENDENTAL DEDUCTION. 393 
 
 -which are ' the species of apperception.' What prevents this 
 view from being perfectly intelligible is only the thought, 
 which stni holds its ground, of an external application of the 
 •conceptions to the perceptions. But a 'pre-established har- 
 mony ' of perception to conception could only exist for a 
 consciousness which, while it distinguishes these two forms, 
 transcends the opposition between them. And though the 
 understanding generally appears as one of the two terms 
 opposed, yet there are many passages which imply that it also 
 over-reaches, " under the name of imagination," the gulf which 
 seems to be fixed between it and sense. For its application of 
 the conception to the perception is at the same time declared 
 to be a "recognition" that, as the result of its own unconscious 
 action, the conception is already involved in the perception. 
 
 But the key to the whole difficulty lies in the fact, not Perception 
 
 not prior to 
 
 <;learly recognised by Kant,i that sense, as it can exist in a thefense it'" 
 
 has as sub- 
 sumed under 
 ^That it is not clearly recognised is involved in the fact that, though Kant *!i^ concep- 
 says that perception apart from conception is for us ' as good as nothing, ' he 
 yet draws a line of distinction between what is actually united with self- 
 consciousness, and what is capable of being so united ; and seems to suppose 
 that the latter, though actual as a perception in us, is yet still in the condition 
 in which it would be in » consciousness which did not think but merely 
 perceived. Thus, he asserts that conformity to the unity of self-consciousness, 
 and so to the categories, is the condition without which nothing can be brought 
 within the sphere of an intelligible experience : but he does not seem to 
 recognise the change that is made iu perception, so soon as it has entered this 
 sphere. In other words, he does not distinctly realise that, as distinguished 
 from, and related to, conception, perception is not what it was, or could 
 possibly be, in a being that merely perceived. Yet, it is the most important 
 effect of the transcendental Deduction, and of Kant's whole method of dealing 
 with the subject, that it enables us to realise the truth, that the development 
 from consciousness to self -consciousness is not merely the addition of the latter 
 to the former, but at the same time the transformation of the former in 
 relation to the latter. Something like this indeed is essentially involved in the 
 idea of an organic development in all cases, and a fortiori in the case of the 
 development of intelligence. In what is here said, it is not meant to deny that 
 sensations exist in us as beings who are developing out of the sensitive into 
 the self-conscious life, and my contention is only that sensations as part of our 
 self-conscious life have already become perceptions, and that as perceptions 
 they are already determined in relation to conceptions, which may, as Kant 
 says (A. 103), be 'wanting in clearness,' but are never entirely absent in 
 any odfcsoiousness of objects as such. 
 
394 THE CRITIQUE OF PURE REASON. BOOK i. 
 
 being which is not self-conscious, does not correspond to 
 perception, as it exists in beings who ar& self-conscious, any 
 more than it corresponds to conception. In one point of view 
 a merely sensitive consciousness is capable neither of conception 
 nor of perception, as these appear in our consciousness: in 
 another point of view, it possesses the germ or beginnings of 
 hoth. In the sensitive being, in fact, the two elements, which 
 in us are distinguished as conception and perception, are not 
 yet separated : it has a consciousness therefore in which we can- 
 not find either in its distinguishing characteristics. For the 
 severance of the two elements from one another gives to each 
 of them a new qualification. The sensitive consciousness 
 maintains its unity with itself and we may add, feds its unity 
 with itself, through all the changes of its states ; but it does not 
 distinguish that unity from, and relate it to, those states. 
 When, however, the 'lightning of subjectivity' (to use a 
 phrase of Hegel) breaks this charmed sleep of sense, it brings 
 with it a distinction of the self from its sensitive states which 
 at the same time become perceptions of objects. But this 
 is possible only if the manifold of perception j.s combined 
 according to a conception which is at once distinguished from 
 it and predicated of it. Hence, as Kant points out, the 
 recognition of the perception as falling under a rule, and the 
 reference of it to an object, is one and the same thing; and 
 this reference again makes possible a consciousness of the self, 
 as having the idea of an object through the perception. Thus 
 the same change by which sensations become perceptions gives 
 rise to the distinction between conceptions and the perceptions 
 of which we predicate them, i.e., the distinction of the activity 
 of thought from the matter which it combines, or recognises as 
 combined, in the idea of an object : a distinction which finally 
 is presupposed in the consciousness of the self as having these 
 ideas. Thus the self is not first conscious of having ideas 
 which it then refers to objects ; but, on the contrary, it is only 
 as it refers its ' ideas of perception ' to the objective world 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 395 
 
 which is known through them, that it can be conscious of 
 them as its own ideas. The ego can claim as its own its 
 sensitive states only in so far as it separates itself from them, 
 and refers them to the objective connexion of the world it 
 knows. They are its own, or capable of being known as its 
 own, just as it rises above them and refers them to objects. 
 For, as mere sensations they would not be ideas which we could 
 refer to objects and so unite with self-consciousness : they 
 would be only events taking place unknown to us, in our 
 sensitive organism. On the other hand, the objective world, 
 to the consciousness of which the consciousness of self is 
 correlative, cannot be presented to us as a mass of isolated 
 particulars. To put it in Kant's phraseology, it is necessarily 
 subjected to the categories as the condition of its being united 
 with self-consciousness. Thus it comes to be determined as a 
 system, all the elements of which have their character fixed in 
 relation to the other elements according to 'universal principles. 
 For there can be nothing for the self, which is not fixed as 
 an element which is permanent with the permanence of the 
 self, and which is not in this permanence determined as an 
 element in the one world to which the consciousness of self 
 is correlative. 
 
 When, in view of what has been said, we make the neces- ?f .■"^** ^fn^e 
 
 ' ' it IS prior to 
 
 sary corrections in Kant's doctrine of the relation of perception """"^pti""- 
 to apperception, the result may be thus stated : — The advance 
 from consciousness to self-consciousness, from perception to 
 apperception, is in one point of view analytic, in another 
 synthetic. There is a sense in which it may be said that the 
 matter remains the same, not only in the consciousness of the 
 savage and the civilised man (as Kant says in a passage already 
 quoted ^), but even in the consciousness of the animal and the 
 man ; and that the change is only in the form. Yet, in truth, 
 this formal change is equivalent to a complete transformation 
 
 of the matter. The life of an animal (supposing an animal not 
 
 % 
 
 1 See above, p. 352. 
 
396 THE CRITIQUE OF PURE REASON. book I. 
 
 to be self-conscious) is in unity with itself through all the 
 difference of its states. It is not conscious of itself in distinc- 
 tion from its states, but through all the various feelings and 
 images that may be present to it, it preserves the continuity 
 of its feeling of itself unbroken. On the other hand, the 
 life of a self-conscious being is a life in which the unity is 
 constantly broken, yet continually restoring itself. Self-con- 
 sciousness is always a return iuto the identity of the self from 
 the difference of its objects. This return, indeed, presupposes 
 the unity to which the return is made : yet with some truth it 
 might be said that in its highest form that unity does not exist 
 except in and through the return. The unity of a merely con- 
 scious or merely sensitive life does not fully deserve the name 
 of a self, since it does not exigt/or itself as an object, and does 
 not, therefore, really recover itself out. of the difference in which 
 it is immersed. It is a universal in all its particularity, in so 
 far as the unity of its life maintains itself in all the change of 
 states ; but it is not a universal for itself, as it does not dis- 
 tinguish itself from, and relate itself to, objects through them. 
 But the self-conscious being exists for itself as this universal 
 individual. It is for itself — not merely for another — as this 
 permanent identical self through all differences and changes of 
 its states. This, however, it can be only as it separates itself 
 from them and so becomes conscious of them in their separa- 
 tion from each other, while, at the same time, it binds them 
 together as elements in one objective consciousness. We can 
 ideally distinguish the two movements in the progress from the 
 sensitive to the self-conscious life : first, the withdrawal of the 
 unity of the sensitive life by which the sensations or images 
 become an unconnected manifold ; and, secondly, the recombina- 
 tion of them as elements in the consciousness of one objective 
 world, which is the correlate of the consciousness of self. But 
 these two movements cannot be really separated ; for the with- 
 drawal of the unity from the sensitive life is, at the same time, 
 the dawning consciousness of the objective world, to the con- 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 397 
 
 nexion of which its elements are referred : and the subject for 
 which the manifold can exist as a manifold, must already have 
 brought that manifold to a unity in one consciousness. At the 
 same time, it is natural to start with the manifold as given 
 prior to the synthesis whereby it is united in one conscious- 
 ness ; for self-consciousness, as contrasted with the simple unity 
 of the life of a sensitive subject, is, in its idea, a restored unity, 
 or rather a unity which restores itself out of the dualism and 
 opposition in which it begins. That dualism, however, cannot 
 be absolute even at the beginning of a conscious life, — i.e., it 
 cannot be such as to leave a mere manifold on one side and a 
 bare unity on the other, — otherwise it would be a division 
 which could never be healed. Even in our earliest apprehen- 
 sion of the world, it is one world in one space and one time, in 
 spite of the infinite and apparently unrelated manifoldness of 
 its changing phenomena; and the self which is opposed to it, is 
 yet related to it in such' a way as to make its difference a per- 
 plexing problem. From the earliest dawn of intelligence the 
 world as perceived is a mystery, perplexing man with its 
 ■opposition and unlikeness to himself, yet, on tlie other hand, 
 stimulating him, by its relationship and likeness, to vague 
 anthropomorphic explanations of its phenomena. For even in 
 his simplest perception of it, the world, in spite of its mani- 
 foldness, is represented as a unity; and the growing consciousness 
 of the unity of the self as opposed to it, is a stimulus to 
 develop the conception of that unity. While, therefore, in one 
 point of view, we may say that the mind seeks its own unity 
 in the world, in another point of view what it seeks is only to 
 discover what the given unity of perception is, or, in other 
 words, to ' bring to conceptions ' the blind synthesis of percep- 
 tions, i.e., to recognise the principle which already is in action 
 in that synthesis. But in using such language we have to 
 remember (what has been already indicated) that, as self-con- 
 sciousness is a return to an identity which comes to exist as 
 such*identity only in the very process we call a return, so the 
 
398 THE CRITIQUE OF PURE REASON. book i. 
 
 recognition of the conception as the principle of the synthesis 
 of imagination in perception, is the recognition of that which, 
 as such principle, has never been known before. Further, we 
 have to observe that the perception from which the return is 
 made gets its qualification as an object, which can be subsumed 
 under the conception, only in relation to the conception which 
 is thus separated from, and related to, it. It is Kant's imper- 
 fect apprehension of the organic nature of consciousness and of 
 the consequent dialectical character of its development, in 
 which each movement implies all the others, that here em- 
 barrasses his statement, leading him to speak as if one operation I I 
 were completed before the other begins, and to overlook the --(->-,* 
 way in which all the different momenta reciprocally imply each 
 other. This especially is what entangles him in difficulty in 
 dealing with the relations of the conscious synthesis of concep- 
 tion to the blind synthesis of perception. For, though the 
 latter, without the former, could at the most give rise only to 
 the unity of an image, yet Kant, in view of the necessary 
 subsumption of the image under a conception, goes so far 
 as even to attribute to it a judgment.^ In truth, to the 
 conscious self, perception necessarily takes the form of a 
 judgment; but when it takes this form it, at the same time, 
 becomes possible to separate from it the principle of its unity 
 and to use that principle as its criterion. The determination 
 of things as objects of perception, therefore, is not only a 
 separation of perception from conception, but at the same time 
 it involves the possibility of a further advance, in which 
 conception is our guide in reconstructing the perceptive con- 
 sciousness. In this way the opposition between consciousness 
 and self-consciousness, between perception and apperception, 
 becomes the stimulus to a progressive movement by which 
 that opposition may be removed. But this whole process would 
 be impossible, if the opposition had been absolute; if the terms 
 so opposed were not essentially related to each other; if the 
 ^ In the Prolegomena, § 18. Cf. above, p. 376. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 399 
 
 division with which self-conscious life begins did not imply 
 in the very statement of it the unity of intelligence. To 
 find the key to Kant's alternate analysis and synthesis, — to his 
 alternate opposition and reconciliation of the elements of the 
 intellectual life and the intelligible world, which it is as fatal 
 to confuse as to isolate, — we need only to keep before us the 
 idea of that organic unity in difference which he never fully 
 expresses, yet to which it was his main merit to direct the 
 attention of philosophy. 
 
 The simplest statement that can be given of Kant's proced- ^^^^J'^^^'^' 
 ure in the Deduction is that, while he masks the idea of an OTganio*Mity 
 organic unity of the intelligence under the form of a reciprocity ence, and ws 
 
 rr ° "^ . " r J negative view 
 
 of action and reaction of its different faculties, he is obliged to of theretum 
 
 ' ^ of conscious- 
 
 bring in the unity of these elements as a kind of deus w-'Sf"^™ 
 maehina to supply a link between them. Thus he starts with 
 a thing in itself, which affects the subject, and so produces a 
 series of changes in its sensibility. As against this differentia- 
 tion introduced into it from without, Kant conceives the self, 
 in virtue of its unity, as showing a kind of self-defensive 
 power of synthesis, by which it takes up and combines the im- 
 pressions into one image, and then becomes conscious of its 
 own activity in so combining them. In this way the manifold, 
 as manifold, is separated from, and opposed to, the self as a 
 self-identical unity ; while yet, as combined into the conscious- 
 ness of one objective world, it is related to that self and capable 
 of being made a part of its self-consciousness. In other words, 
 Kant thinks of that return of the subject to itself, whereby it 
 becomes self-conscious, as the result of the process whereby it 
 unites the manifold into one consciousness of objects. Thus 
 the analytic judgment ' I am I ' (through all the difference of 
 my ideas) is dependent on the synthetic judgment which, com- 
 bining all the manifold elements of perception into one con- 
 sciousness, makes it possible for the unity of the self to become 
 a self-conscious unity. If, however, we view the whole process 
 in thjs way, starting from the dualistic hypothesis of two things 
 
400 THE CRITIQUE OF PURE REASON. book L 
 
 in themselves, an object in itself and a subject in itself, — the 
 former of which gives a repeated ' Anstoss ' to the latter from 
 without, while the latter reacts on the affections thus produced, 
 and by synthesis of them produces a correlative consciousness 
 of the object and of the subject as phenomena, — we are obliged 
 immediately to recognise that both of these things in themselves 
 lie outside of the unity of knowledge. It can, therefore, only 
 be some defect in the unity of this correlative consciousness of 
 subject and object, which suggests that its terms are pheno- 
 menal, and that they must be referred to two independent 
 things in themselves. In other words, the dualism of a 
 noumenal subject and object out of our knowledge is made 
 necessary for Kant by the imperfection of the unity of the 
 phenomenal subject and object in knowledge, which again, as a 
 conscious imperfection, implies that knowledge does not corre- 
 spond to its idea. But what is that idea ? It is, as Kant 
 frequently tells us, the idea of an intuitive or perceptive 
 understanding — an understanding in which the difference of 
 perception and apperception, of consciousness and self-con- ' 
 sciousness, either does not exist or is transcended. In our 
 consciousness, as Kant maintains, the forms of sense, like its 
 matter, stand unrelated to the conceptions under which they 
 are brought ; and these again, though derived from the forms of 
 judgment, and so from the analytic unity of self-consciousness, 
 are yet so far dependent on the unity of the self being 
 brought in relation to an extraneously given manifold, that 
 their number and nature cannot be explained from that unity 
 alone.^ It appears, then, that the subject reacting on the 
 manifold, which is forced on it from without, according to 
 peculiar forms that belong to its peculiar sensitive constitution, 
 synthetically combines that manifold ; and that in opposition, ' \ 
 though in relation, to the objects so determined it becomes con- 
 scious of its own self-identity. But this return upon its own 
 pure identity, upon the ' I am I ' of pure self-consciousness is, 
 
 'B. 146. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 401 
 
 in Kant's view, a negative return, i.e., it is the recovery of the 
 bare identity of self out of the foreign element in which it has 
 become involved through its connexion with a sensibility. 
 The final effect of this synthetic action of thought and of the 
 return from it upon the unity of the self is, therefore, not to 
 appropriate to the self the matter of sense taken up into 
 consciousness, (which is impossible owing to the alien nature 
 of that matter) but to repel it, and with it the phenomenal 
 world of objects, (the consciousness of which has been attained 
 through synthesis) from the self. Nay, the recoil even points 
 beyond the analytic judgment ' I am- 1 ' to a simpler identity ; 
 since it is recognised that it is only by means of the synthetic 
 unity that the analytic unity exists as a judgment, in which 
 subject and predicate are distinguished. But when the intel- 
 ligence thus draws back upon its analytic unity, or upon an 
 abstract identity .which is even simpler than that analytic 
 unity, the movement of abstraction necessarily carries with it 
 the condemnation of all our empirical knowledge as imperfect 
 and inadequate to its idea. For the intelligence, conscious of 
 its pure unity with itself, repels experience from it as involving 
 a synthetic unity in which difference is never completely over- 
 come. Thus, the dualism of the subjective thing in itself and 
 objective thing in itself out of knowledge ultimately rests on the 
 discordance of empirical knowledge with the ideal which the 
 mind carries with itself in self-consciousness. 
 
 Now, in an earlier chapter it has already been pointed out is tis ideal of 
 
 knowledge au 
 
 what this means. It means that Kant, recognising the opposi- f^|*yty ^^^^^ 
 tion of perception and conception, and the impossibility of option n"°"' 
 taking either in its abstraction as absolute truth, falls back by an intuitive 
 
 understand- 
 
 a still further abstraction upon an ideal of truth which is most {JJ.th?^"''*' 
 nearly represented by the analytic judgment of pure self-con- 
 sciousness. The ideal of knowledge, according to this view, 
 would be, that the object should be one with the self that 
 knows it, as the object self is one with the subject self in pure 
 
 self-cwnsciousness. As, however, even in pure or analytic self- 
 VOL. I. 2 c 
 
402 THE CRITIQUE OF PURE REASON. BOOE 
 
 consciousness there is a dualism, which is due to its relation 
 the consciousness of objects, the ideal carries us still furth 
 back, to the negation even of this difference in a unity: 
 which there is not even the distinction implied in the judgmei 
 ' I am I.' But, as I have already shown, this regress upon 
 unity, which is neither perception nor conception, but both : 
 one, has implied in it an unconscious logic, which is exactly tl 
 reverse of its explicit reasoning.^ It is stated by Kant as s 
 abstraction from both perception and conception, and tli 
 would logically bring us to a bare empty identity ; but real 
 it is a recognition that the truth is to be found only in a unil 
 which includes both. Hence, Kant silently substitutes for th 
 idea of a pure unity that is neither perception nor coneeptio 
 the idea of an intuitive understanding which is ioth. And 
 is only as he does so that he can get from it that ideal < 
 knowledge which he opposes to experience. For it is impo 
 sible that by mere abstraction we can reach a point of vie 
 from which we can see the limitation of that from which \i 
 abstract. Such a point of view we can find only in a unity i 
 relation to which the opposition of conception and perceptic 
 sinks into an opposition of elements which imply each othe 
 Only as we are able to rise above the relative or imperfect tmit 
 of perception and conception in experience to a princip 
 which is capable of turning it into an dbsohde unity, can it 1 
 possible for us to see its relativity and imperfection. Or if, i 
 relation to the unity of self-consciousness, we can see tl 
 phenomenal character of the objects of experience, self-coi 
 sciousness must be itself a principle which will ultimate] 
 enable us to turn knowledge of the phenomenon into kno\ 
 ledge of the noumenon. 
 
 "ure°houg&t ^° ®^°^ ^^^^ ^^^^ ^^ ^°' ^* would be necessary to get rid 
 Sytia° Kant's idea that the movement of thought in itself is pura 
 analytic; and that, therefore, synthesis must always be tl 
 combination of a manifold which is external to the unii 
 1 Cf. above, p. 187 seq. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 403 
 
 implied in such synthesis, and capable, therefore, only of an 
 external or mechanical combination. In opposition to this 
 view, we should have to show that even the pure unity of 
 thought, the pure consciousness of self, is essentially synthetic ; 
 and that, therefore, while it is true that the comparison of it 
 with the relative unity of experience may bring to light an 
 ideal which is not realised in ordinary experience, or even in 
 science, yet the ideal so disclosed does not stand in irreconcil- 
 able opposition to that knowledge which it thus enables us to 
 criticise. On the contrary, it contains a principle for the 
 correction of the defect which it discloses. 
 
 Now, the first step in the proof of this thesis is taken by ^™f^„g^f"; 
 Kant himself, when he points out that the analytic unity of aaut'u''"''^ 
 apperception presupposes the synthetic unity ; for this is ^^° 
 practically ec[uivalent to saying that there is no purely analytic 
 unity. In the judgment, ' I am I,' subject and predicate are 
 the same, and in view of this identity it might be said to be an 
 analytic judgment ; but a consciousness which returns upon 
 itself, or reduplicates itself in the manner expressed in that 
 judgment, involves a synthesis, seeing that it is in this return 
 that the ego, strictly speaking, comes into existence ; for only 
 that being is truly to be called an ' I,' which calls itself so. 
 Self-consciousness is not merely the consciousness of the unity 
 presupposed iu the consciousness of objects ; or at least this 
 unity acquires a new qualification in the process whereby 
 it becomes conscious of itself. In other words, the dual unity 
 reached by the process is not the same with the undifferentiated 
 unity which that process presupposes. Hence, it is not merely 
 that, as Kant says, the analytic unity implies the synthetic, 
 not merely that the consciousness of the identity of its own 
 action in the synthesis whereby all perceptions are made 
 elements of an intelligible experience, contains in it as an 
 element the consciousness of the self as identical. The judgment 
 of self-consciousness is itself synthetic, in so far as the ' I ' 
 which seems to be presupposed, is really the result of the judg- 
 
404 THE CRITIQUE OF PURE REASON. BOO 
 
 ment ; or, at least, gets in the very act of judgment a hig 
 
 meaning than it has as the mere presupposition of that act.' 
 
 How the On the other hand, it is also true that this synthesis 
 
 recognition of ' 
 
 moveSent*n self-consciousncss implies a previous synthesis with which 
 ness trans- transparent unity is contrasted ; and that the judgment, 
 
 forms our , p n 
 
 view of . am I,' correlates with a consciousness of the unity of all t 
 
 objects and of 
 
 the self. manifold of perception in one objective world. For it is or 
 because I am able to put all objects into one ' context 
 experience ' that, in spite of the variety of that experien 
 I can be conscious of one self; and an empty 'I am I,'- 
 judgmeut of self-consciousness which did not involve at t 
 same time a consciousness of a world of objects,— would 
 impossible. But the transcendental criticism which reves 
 this correlation obviously leads to a change both in o 
 consciousness of objects and in our consciousness of se' 
 for it teaches us that the determinations which we ga 
 to objects, when we considered them merely in themseh 
 or in their relations to each other, are not sufficient, 
 showing us that objects are phenomenal, i.e., that they a 
 existences for a self, it at the same time shows what th( 
 reality is ; i.e., it shows that they are not taken rightly, t 
 they are regarded as elements in a process of existence whi 
 must finally reveal itself as a spiritual process.^ On the otl 
 
 ^ We find that Kant himself reduces the ego to an indeterminate someth; 
 (" =X, he or it that thinks"), when he considers it as the mere subject, 
 abstraction from its consciousness of other objects and of itself as an obj 
 (A. 346 ; B. 404). But the judgment ' I am I ' is analytic only in the se 
 that its synthesis is perfect, i.e., that the difference in it has become trs 
 parent. And it is really this (as we shall see hereafter) that makes it 
 source of an ideal of knowledge which is not satisfied by the imperfect s; 
 thesis of experience. But for Kant it is analytic, because he regards a 
 consciousness simply as the recognition of a unity which is prior to 
 consciousness of objects, and not as the ultimate form into which that ui 
 develops through the consciousness of objects : a form, it may be added, 
 which it for the first time reveals itself distinctly o-s an ego or self ; for, if 
 are to speak with absolute precision, we must say, only a self-conscious sub; 
 can be called a self. 
 
 - It shows this, I mean, in so far as it can be shown by general reason 
 without being actually realised in a complete systematic view of the univei 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. -405 
 
 hand, it also changes our view of the self ; for, while in our 
 first consciousness of it the self is taken as exclusive of objects 
 and independent of them, now we see that this exclusive or 
 negative relation, presupposes a positive relation and so a unity 
 beyond the difference of object and subject. We are, indeed, 
 conscious of the self only as we return upon it from, and 
 oppose it to, objects ; but this opposition must be taken 
 as relative and not absolute, else it would not be a conscious 
 relation at all. Hence, when we are conscious of the self 
 merely in opposition to objects and therefore as a pure self- 
 identical unity, in what Kant calls the analytic judgment of 
 self-consciousness, we are not fully aware of what is involved 
 in our own thought. In being conscious of its opposition 
 to objects, the self has transcended that opposition ; and it is 
 unaware of what itself does and is, so long as it considers that 
 opposition to be absolute. Hence, just as the consciousness of 
 the object was imperfectly developed, when the object was 
 regarded as a thing in itself independent of the subject that 
 knows it; and as that consciousness and its idea of its object 
 required to be corrected by a ' transcendental reflexion,' which 
 called attention to its relation to the subject, so it is here with 
 self-consciousness. The consciousness of self cannot be con- 
 sidered to be perfectly developed, or the self to be fully aware 
 of what it really is and does, so long as it rests in the idea of 
 itself as a simple self-identical subject apart from objects. In 
 other words, it is not completely self-conscious, till it recognises 
 that, (as it can exist for itself only when the object exists for 
 it,) the consciousness of the object is an essential element of 
 the consciousness of self; and the consciousness of self is 
 therefore not merely the consciousness of a self-identical 
 subject opposed to the object, but it is also the consciousness 
 of a principle, which underlies the consciousness of the object. 
 
 which, of course, is impossible. If, in this sense, it is called a reasonable faith 
 and not knowledge, I should have nothing to object. But in this sense all 
 knowledge of the universal is a faith, and all knowledge of the particular is 
 also a faith so far as it is built upon that knowledge. 
 
 ..A^iJji 
 
406 THE CRITIQUE OF PURE REASON- ^OOK 
 
 While, therefore, it is only through its relation to the self tb 
 the object comes to be known as it really is, or, we might eve 
 say, comes to the consciousness of itself in us : on the othf 
 hand, it is also true tha.t the self is not known as it really i 
 so long as it is conceived as an abstract identity, (as we ai 
 disposed at first to conceive it,) but only when it is taken as 
 principle of unity in difference, a principle which cannc 
 realise itself without going beyond itself ; in short, a princip] 
 which is essentially synthetic. In this way, the 'transcei 
 dental reflexion ' lifts us above dualism on both sides, and leac 
 us to regard the consciousness of the object and the conscioui 
 ness of the self as issuing in their difference from a commo 
 source, and pointing in their final form to a unity in whic 
 that difference is resolved. 
 The dualism How far docs Kant's transcendental reflexion answer to thi 
 
 of Descartes. 
 
 lint jrttiTMs' description ? The answer is partly contained in what has bee 
 mind'aiid ° already said ; but some further light may be thrown upon i 
 
 matter. , t i • r> i 
 
 by considering m what way the opposition and relation of tl 
 subjective and the objective consciousness appeared in writei 
 before Kant. Now, it is in the philosophy of Descartes ths 
 we have the most simple and naive transcription of the dualisi 
 of object and subject, matter and mind, as it .appears in tl 
 ordinary consciousness. The subject, one with itself and ii 
 divisible, stands on the one side, and the material world i 
 infinite diversity on the other ; and there is no room for 
 mediation except by a detis ex machina. Thus the pure unit 
 of the self is just the opposite counterpart of the extended sul 
 stance, i.e., of a world which is infinitely self-external or divide 
 into partes extra partes ad infinitum. But in showing thi 
 Descartes showed also, though he was never aware of tl 
 result of his logic, that these opposites imply each other; ar 
 that the consciousness which apprehends each in relation i 
 the other, has already potentially transcended their oppositic 
 and grasped a unity, in view of which it ceases to be a contr 
 diction, and reduces itself into an opposition of elements whi( 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 407 
 
 are necessary correlates of each other. From this point of 
 view, the opposition of the consciousness of the object to 
 self-consciousness is, as Kant expresses it, an opposition of 
 ' different branches which spring from one and the same stock.' 
 And we are obliged to think of the unity of self as a principle 
 which is the source of all that difference in the objective world, 
 in opposition yet in relation to which it becomes self-conscious. 
 Hence, the opposition between the conscious self and the ex- 
 ternal world is not to be regarded, as Descartes regarded it, 
 as an opposition between the intelligence and that which is 
 essentially other than the intelligence, but as an opposition of 
 correlative elements in the intelligible world ; and a relation of 
 things as in space is a relation which can exist only for an 
 inteUigenee, which, therefore, in apprehending such things, is 
 not going beyond itself ; for the external object is not external 
 to the self for which it is. 
 
 Now, in Kant we undoubtedly find the first definite expres- Kant over- 
 comes this 
 sion of this new idealistic view of things, though in him it is dualism, 
 
 P ' ° though only so 
 
 still embarrassed in its expression by elements derived from an tophenomena. 
 earlier dualism. For Kant, the absolute opposition of the in- 
 telligence to a world in space has disappeared. He holds that 
 " space itself with all its phenomena, as ideas, are only in me, 
 and it is impossible that in this space anything should be 
 external to me in a transcendental sense, because space itself 
 is nothing external to our sensibility '' ; and that, " if anything 
 were in a transcendental sense external to us, it could not be 
 represented or perceived as external in the sense of being in 
 space." 1 Furthermore, he maintains that the consciousness of 
 the object in space, and of the self in contrast with it, does not 
 arise except by the activity of the self, which out of the data 
 of sense constructs for itself the objective world, and in dis- 
 tinction therefrom becomes conscious of itself. At the same 
 time, the activity of the self is always conceived by Kant to 
 have relation to certain affections of its passivity which have 
 ' 1 A. 372 seq. 
 
408 THE CRITIQUE OE PURE REASON. book r. 
 
 to be referred to things in themselves ; i.e., to that which is in 
 a transcendental sense, external to the self. " The understand- 
 ing makes Nature, but does not create it," for it makes it out 
 of a matter passively received. And as little as Descartes, 
 does Kant observe any correlation between the simplicity of 
 the pure ego and the form of space under which its object 
 appears to it. Hence, Kant only changes the old dualism 
 between the extended substance and the mind into a subjective 
 dualism between perception and conception, and ultimately 
 between the given data, together with the forms under which 
 they are received, and the activity of the understanding in 
 regard to these data. Thus the spectre of the old objective 
 dualism reappears in the form of the opposition between 
 the ego in itself that manifests itself in the activity of 
 thought,, and the thing in itself which affects it, and so 
 provides material for this activity. Further, according to this 
 view, the ego knows itself in its noumenal reality as little 
 as it knows the object in itself ; for, if all that it knows of the 
 object in itself are the affections it receives from it, and if 
 through these it knows only that the object as a thing in itself 
 is but not what it is, on the other hand, all that it knows of 
 the ego in itself is the activities by which it gives to these 
 data the form of experience, and this only enables it to say 
 that the ego is but not what it is. It cannot^ therefore, deter- 
 mine itself any more than it can determine the object as a 
 thing in itself. It knows each only in its relation to the other ; — 
 on the objective side, phenomena or the data of sense as deter- 
 mined by the unity of the understanding ; and on the subjective 
 side, the unity of the understanding in the acts whereby it 
 determines the data of sense. As has been well shown by 
 Dr. Staudinger,^ Kant keeps us steadily within the relation 
 " I — object," and will not let us determine either term of the 
 relation in itself. But, nevertheless, he still keeps up the 
 dualism between the terms of this relation and regards them 
 ^ Noumena, p. 10 seg. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 409 
 
 as related, not directly or in themselves, but only as each 
 supplies a factor v^hich is necessary to our experience. Con- 
 sequently he maintains that we have a consciousness and an 
 assurance of the existence of both, but not a knowledge of 
 either in itself. And the transcendental reflexion which detects 
 the existence of the factors, and shows how the consciousness 
 of the object and of the self in relation to each other arise out 
 of their interaction, is able only to detect this relativity ; but 
 not to escape from it. 
 
 After what has been already said, it is not difficult to ^^'that' th""' 
 show the defect of this view. In fact Kant himself has the^SSty'of' 
 
 -,. 'I'l. p • ■• 11 knowledge in 
 
 already shown the impossibility of resting m it and the seu-conscious- 
 
 ■ " ness is also a 
 
 way out of it when he says that the analytic unity of p™^'''^^- 
 apperception presupposes the synthetic unity : for this can- 
 not mean that the former is simply the abstraction of one 
 element in the latter, but only that self- consciousness is 
 consciousness of objects and something more ; or, it is con- 
 sciousness of objects brought to the knowledge of what it 
 implicitly but really is. Kant's mistake on this point is not 
 altogether to be explained by the fact that self-consciousness 
 appears at first as the consciousness of a self from which the 
 object is excluded. Indeed, he detected the insufficiency of 
 this view when he showed that we are conscious of ourselves 
 as knowing subjects, only as we are conscious of the activity 
 whereby we determine objects as such. But while he thus 
 apprehended that in self-consciousness we make a regress upon 
 the unity which underlies our determination of objects, and so 
 are able to go back, as it were, to the beginning of that 
 consciousness, and view it in the making, he did not see with 
 equal clearness that this regress is at the same time a progress, 
 by which our whole view of the objects is raised into a higher 
 form. For self-consciousness in this sense is the consciousness 
 of a principle, which is presupposed in our consciousness of 
 the objective world, and without which the data of sense would 
 be fof lis nothing at all. But it is simply a consciousness of 
 
410 THE CRITIQUE OF PURE REASOK. book i. 
 
 this principle in its abstraction, and the consciousness of the 
 principle is not supposed by him to add anything to the 
 principle itself. It is simply the analytic recognition that it is 
 what it is. As Spinoza said that desire is simply " appetite 
 with the consciousness of it," as if the consciousness of appetite 
 did not change its character, so Kant says that self-conscious- 
 ness is simply the consciousness of the unity that is pre- 
 supposed in all consciousness of objects, and in all the 
 synthesis whereby that consciousness is developed. It is 
 thus an analytic judgment which brings to light "the X, 
 the ' he ' or ' it ' that thinks " ; nor does Kant recognise that in 
 becoming conscious of itself, this " he or it " for the first time 
 becomes, or manifests itself, as an ' I.' " This ego, or he, or it, 
 (the thing) which thinks, means nothing for us except a trans- 
 cendental subject (=X) which can be known only through the 
 thoughts which are its predicates, and of which, as apart from 
 them, we never can have the smallest conception. Hence, in 
 attempting to fix it as an object, we turn in a continual circle 
 round about it, for we are obliged always to use it in order to 
 make any judgment about it."^ We are able, therefore, only 
 to mark it out as the subject of all thought and knowledge, but 
 never to determine it as an object ; and our return to it is 
 a negative return, in which we abstract from part of the 
 conditions under which objects are known. 
 tency°of thil' "^^ ^his the couclusive answer is that which is already 
 Twn account suggcsted by Kant's words. For if the unity that underlies 
 experience is revealed as an ego only in its return upon itself 
 in self-consciousness, and if further, it is only as a self- 
 conscious ego that it can be conceived as the source of the 
 categories by which objects are determined, then we must 
 regard that return not .merely as a regress by which the 
 constituents of our actual consciousness of objects are dis- 
 covered, but at the same time as the beginning of a new 
 consciousness of objects. Kant's view implies that in going 
 1 A. 345 ; B. 403. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 411 
 
 beyond itself to determine objects the subject is going away 
 from itself; hence there is to him a contradiction in the very 
 conception of its determining or knowing itself as an object. 
 But if it can be conscious of itself in its identity only through 
 the synthesis by which it determines objects, and if, prior to 
 that synthesis, it is not revealed as an ego or self at all, it is 
 absurd to say tha.t the synthesis by which it becomes conscious 
 of itself as an object, at the same time hides it from itself ; or, 
 what is the same thing, to say that the only true consciousness 
 we have of the self is as in an analytic judgment, in which we 
 simply recognise the unity presupposed in experience as a 
 simple unity which can have no difference in it. 
 
 To do iustice to the movement of thought from consciousness consequences 
 
 " ^ of conceiving 
 
 of objects to self-consciousness, we must regard it at once as a *g bX^™™* 
 regress and a progress, at once an analytic and a synthetic prl^essiv™ 
 
 . . . ... analytic and 
 
 movement. It is an analytic regress, in so far as it brings to synthetic. 
 light the unity of the self which is presupposed in the con- 
 sciousness of objects and makes that unity an object to itself, 
 at the same time that it abstracts from all the objects 
 determined by it and in relation to it. It is a synthetic 
 progress, in so far as it is only as thus made object to itself 
 that the unity for the first time reveals what it is. Hence, 
 before this process, we call it an ego only by anticipation, i.e., 
 in view of its capacity of thus making itself an object. 
 It is, indeed, just because of this synthetic movement 
 implied in the consciousness of self that in our ordinary 
 consciousness of self we oppose it to all objects. It is this 
 natural dualism, as it has been called,— in which the conscious- 
 ness of the subject and of the object are directly opposed 
 to each other as the consciousness of two separate objects,- — 
 which is expressed in its most abstract form in the philo- 
 sophy of Descartes ; and the transcendental reflexion of 
 Kant has its value mainly in bringing us into a new 
 attitude of thought in which these opposites are brought 
 togetlftr. Thus self-consciousness is not isolated from the 
 
412 THE CRITIQUE OF PURE REASON. book i. 
 
 consciousness of objects, but the latter is seen to be necessarily 
 involved in the former ; while at the same time the former is 
 regarded as the consciousness of that principle which is mani- 
 fested in the latter. Bat if in this way self-consciousness be 
 recognised as the consciousness of the principle which underlies 
 the consciousness of objects, another step becomes necessary. 
 For on the one hand, a fact seen in the light of its principle is 
 no longer the same fact to us, and thus experience or the con- 
 sciousness of objects, when carried back to the unity revealed 
 in self-consciousness, must change its character ; and on the 
 other hand, the consciousness of self, when regarded as con- 
 sciousness of the principle of unity in objects, must also change 
 Us character. It must cease to be regarded as an analytic 
 unity which rests in simple identity with itself, and it must be 
 seen to be a synthetic principle, a principle of difference which 
 goes out of itself to objects in order through them to realise its 
 unity with itself. 
 How this view If from this point of view we reconsider Kant's idea of 
 
 throws light 
 
 phy^oaT"'"' judgment, we get a new light as to his deducbion of the cate- 
 the categories, gories by meaus of it. In the judgment of experience, accorduig 
 to Kant, perceptions are brought under a priori conceptions, 
 and so at once determined as perceptions of objects and brought 
 into relation to the unity of the self which is the source of 
 these conceptions. In the process of judgment, therefore, the 
 unity of thought is regarded as, so to speak, differentiating 
 itself into the various forms which supply principles of unity 
 for the data of sense, or, putting it in the other way, which 
 (inable us to recognise perceptions as perceptions of objects and 
 so to unite them with the " I think." It seems, therefore, that 
 we are entitled to regard judgment as the process in which 
 thought goes out of itself to determine objects, that through 
 such determination it may reach a consciousness of itself. Or, 
 as we might otherwise express it, the judgments by which 
 we determine objects are steps in the synthetic process by 
 which we finally reach the judgment of self-consciousness. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 413 
 
 Further, according to Kant, the consciousness of self is primarily 
 a consciousness of the activity whereby we bind all the data of 
 sense into the consciousness of one objective world ; or, in other 
 words, it is the return of that consciousness upon its principle, 
 or the revelation in its purest and most abstract form of the 
 unity of thought which underlay all its movements. We may, 
 therefore, conclude that all the categories, as steps in the process 
 of the determination of objects, are involved in the judgment of 
 self-consciousness ; or that self-consciousness is a process which 
 includes all the categories, and brings them back to the unity 
 involved in them. Hence, Kant's assertion, that from the 
 analytic judgment all the categories may be derived, points to a 
 truth. Self-consciousness, in its transparent unity-in-difference, 
 contains all the keys by which we are to unlock the secrets of 
 the world : it is the brief abstract of the whole process of 
 knowledge and so of all knowable reality ; for, as it is the con- 
 sciousness of the first unity out of which all the principles 
 of knowledge must be developed, so it is itself the final unity in 
 which they are summed up and brought to completion. Hence, 
 if the synthetic judgment can be taken as the expression of the 
 objective unity of apperception, and the analytic judgment as 
 the expression of the subjective unity which we reach when we 
 become conscious of the principle of the objective unity, it is 
 easy to see that the latter may well afford a " guiding thread " 
 to the discovery of the principles of the former. Unfortunately, 
 Kant's conception of the latter unity as purely analytic destroys 
 the value of his deduction of the categories from it ; and, as 
 usual, the want of a perception of the organic nature of know- 
 ledge, — as based on a principle of unity which manifests itself 
 in difference and through difference returns to itself, — has to 
 be made up for in a mechanical way, by the reflexion of the 
 unity of thought upon an externally given difference. This 
 method of Kant's, as we have already seen, is exemplified in his 
 metaphysical deduction of the categories from the forms of 
 judg&ent, regarded in relation to the determination of the 
 
414 THE CRITIQUE OP PURE REASON. book i. 
 
 manifold of perception. But this vague general reflexion Kant 
 finds to be insufficient, without a definite reference to the mani- 
 fold as given in our inner sense under the forms of time. 
 Hence, he is obliged to introduce the schematism of the cate- 
 gories in order to give them a more definite synthetic value, 
 and prepare the way for their application to objects. In the 
 following chapter accordingly we shall consider Kant's attempt 
 there to enrich these empty forms with content, and so to give 
 them objective value ; or, in other words, to turn them into 
 principles of unity in difference : and we shall indicate how 
 Kant's failure might be turned into a success by reference to 
 the true view of judgment which that very failure suggests, 
 f °Tie^ toe Before doing so, however, it is necessary to consider how it 
 
 oM?perc^p-° is that Kant comes to consider time as the form of inner sense 
 under the in reference to which the categories must be schematised, in 
 
 forms of space 
 
 and time. order that they may be applied to objects as perceived in time 
 and space. The answer to this question has already been in- 
 dicated in the account of the Deduction of the first edition 
 given in the beginning of this chapter ; but the statement of 
 this relation becomes more definite in the second edition, where 
 he separates the Deduction into two stages, in the first of which 
 he shows that the categories are necessary to bring any given 
 manifold of sense into unity with the " I think " of self- 
 consciousness, by determining it as a connected consciousness 
 of an objective world ; while, in the second part, he takes into 
 account that the manifold of our perception is given under the 
 forms of space and time, and shows the same thing with refer- 
 ence to it. Here, therefore, he has to explain more fully the 
 relations of inner and outer sense, of which, in the Aesthetic, 
 he had declared time and space to be the forms ; he has to 
 show how the categories come into action in determining the 
 inner and the outer object in relation to each other, and both 
 in relation to the " I think " of self-consciousness. The circle 
 of experience, in which, through determination of the manifold 
 in the unity of an objective world, we return upon the identity 
 
chap.it. the transcendental debtjction. 415 
 
 of the self, is now enlarged, by considering the two forms of 
 perception and the difference of inner and outer sense. In the 
 tirst part, it was argued that any manifold perceived as a mere 
 manifold of impressions of one subject, would give rise to a 
 broken and unconnected consciousness of self, i.e.., it would not 
 give rise to a consciousness of self at all. All such manifold 
 must be united into one consciousness of objects in order to 
 make possible the consciousness of one self in relation to it. 
 We have now to apply the same thought with this additional 
 qualification, that our objective consciousness, — so far as it is 
 a consciousness of objects other than ourselves,- — is a conscious- 
 ness of objects in space and time ; and that the self of which 
 we are conscious in distinction from these objects is present 
 to us as an object only imder conditions of time. We have 
 already seen that, according to Kant, the consciousness of self 
 presupposes the consciousness of objects, which are objects only 
 so far as the manifold of sense is bound together by an uncon- 
 scious synthesis in images of perception conformably to the 
 categories, and as these images are recognised to conform to 
 the categories as conceptions of objects in general, in applying 
 which, therefore, the intelligence is conscious of its own activity. 
 Sow, we have to realise that the circle of the elements of our 
 experience is wider than this. For our consciousness of self is 
 a consciousness of the subject as one with itself through all the 
 succession of its inner experience in time, in distinction from, 
 yet in relation to, external objects in space and time. But 
 these objects, again, can be objects for us only as the manifold 
 data of outer sense are taken up by a successive synthesis of 
 the imagination in conformity with the categories, and combined 
 into the images of objects in space and time. We have, there- 
 fore, to suppose that the impressions which we receive from 
 without (in a transcendental sense, i.e., from things in them- 
 selves) are met by an unconscious activity from within, which 
 combines them iato images of perception in conformity with 
 the unity of the self; and the consciousness of self as one with 
 
416 THE CRITIQUE OF PURE REASON. book i. 
 
 itself in apprehending these images is explicable only as a 
 conscious repetition of the unconscious activity by which they 
 were formed. But this means a consciousness that the pro- 
 ducts of the unconscious activity are in conformity with the 
 categories, as determining the way in which the manifold of 
 outer sense is taken up under conditions of time and combiaed 
 into one consciousness. 
 What is meant jt jg then Kant's view that, as in the jowre consciousness of self 
 
 by the deter- ' ^ 
 
 to™CTs°ens°J we go back upon the unity which is presupposed in all con- 
 the°categori°es? sciousness of objccts, SO in the consciousness of the inner life 
 of the self in which it maintains its unity through all the 
 succession of its ideas, we go back upon the process by which 
 that unity determines the sensitive subjectivity, through all 
 the succession of its states, in conformity with itself. The con- 
 sciousness of the self as one with itself is the consciousness of the 
 activity of the subject as remaining identical with itself through 
 all its passivity, that is, through all the succession of its 
 impressions ; which implies that these impressions are combined 
 into images in conformity with the categories. When, there- 
 fore, we oppose inner to outer experience, we are simply 
 opposing the process whereby outer experience is constructed — 
 the process in which " the sensitive ego is determined by the 
 intellectual ego," so that the affections of the former may be 
 capable of being " received into consciousness " — -to the outer 
 experience which is its result. This idea is very clearly 
 expressed in the chapter of the Deduction in which the rela- 
 tion of the consciousness of inner to outer experience is touched 
 upon. " We cannot," Kant declares, " think of a line without 
 in thought drawing it, nor of a circle without in thought 
 describing it : we cannot represent to ourselves the three 
 dimensions of space without setting before our mind's eye 
 three lines drawn from one point at right angles to one 
 another: we cannot think of time itself without drawing 
 a straight line (which is for us the necessary outward re- 
 presentation of time), and, while we draw it, attending merely 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 417 
 
 to the activity of synthesis by vyhich we successively de- 
 termine the inner sense. Motion, as an act of the subject 
 (not as a determination of the object), implying as it does 
 a synthesis of the manifold in space, when we abstract from 
 it and attend merely to the act whereby we determine the 
 inner sense in its form, produces the conception of succession. 
 The understanding, therefore, does not find such a combination 
 of the manifold in inner sense but produces it by affecting inner 
 sense." Kant then proceeds to speak of the difficulty of 
 conceiving how the self that thinks should be identical with 
 the self that perceives itself, which is, he says, simply the 
 difficulty of conceiving how I could be an object to myself 
 in inner perception. " But," he goes on, " however difficult 
 it may be to understand this, it may be shown to be the fact, 
 if it be admitted that space is the pure form of external percep- 
 tion. For we are to ourselves objects in time; and though time is 
 not presented to us in external perception, we cannot represent 
 it except under the image of a line drawn in space. Without 
 this image we should not be able to recognise that time is 
 of one dimension, any more than we could determine the 
 duration and relative place in time of our inner perceptions 
 without reference to the changes in outward things. Thus we 
 are obliged to arrange the determinations of inner sense or 
 phenomena in time in the same way in which we arrange the 
 phenomena of external sense in space. If then it is admitted 
 that we know objects of external sense only as we are ex- 
 ternally affected, we must also acknowledge that we perceive 
 ourselves through inner sense only as we are affected by 
 ourselves ; and therefore that in inner perception we know our 
 own subjectivity only as a phenomenon and not as it is 
 in itself." 1 
 
 ^ B. 156 ; § 24. It must be observed that the development of the view of the 
 relation of outer sense in relation to inner sense, which is here given, belongs 
 to the second edition of Kant's Critique. In a subsequent chapter I shall 
 attempt to show more fully that this development is not, as is maintained by 
 Sohoplhhauer and others, to be regarded as a sort of reaction towards a 
 VOL. I. 2d 
 
418 THE CRITIQUE OF PURE REASON. book I. 
 
 enoe as=Mn- From this passage it is evident that, in Kant's view, the 
 
 sciousness of . t* t n i ■ j ^ • • • •^ 
 
 the process consciousness 01 Self as an object of inner sense is primarily a 
 
 whereby 
 
 external consciousness of that process in which the impressions of 
 
 objects are ■*- '• 
 
 such™"'^'^ "^ outer sense are taken up and combined into images accord^ 
 ing to the categories : i.e., a consciousness of the activity 
 by which the ego determines its own passivity, or reacts 
 upon its determination from without. But as the first mani- 
 festation of this activity is unconscious, — as it is an activity 
 of the imagination which is, so to speak, merged in its 
 passivity, or in the process whereby the affections of that 
 passivity become perceptions, — so the consciousness of it takes . 
 the form of a consciousness of the self in distinction from 
 and in relation to an object, or to a world of objects, which it 
 determines. 
 S'treate"^'' Now, as wc have already seen in considering the earlier part 
 eneei^amme of the Dcduction, Kant here gives expression to an impor- 
 
 abstract than 
 
 outer experi- taut truth as to the real nature of knowledge : but he expresses 
 
 ence, enables ° ' ^ 
 
 tteoppositeu t^^t truth in a somewhat misleading way, because he regards 
 the regressive process by which inner experience is constituted 
 as a process of abstraction. The consciousness of the self 
 as one with itself in all the succession of its inner experience, 
 is taken as an analytic judgment, in which we inetely become 
 conscious of the unity in difference which was presupposed 
 in the consciousness of objects, without adding any new 
 elements thereto ; just as the pure consciousness of self was 
 taken as an analytic judgment, which added nothing to the 
 unity of the self. In this view, the consciousness of self in time 
 is a consciousness of an inner object which excludes all the 
 determinations that belong to the outer objects as in space ; 
 
 common-sense dualism, but as an advance towards a more consistent applica- 
 tion of the main idea of the transcendental deduction and the elimination of 
 certain psychological elements which at first intruded themselves upon him. 
 In what remains of this chapter I confine myself to the indication of the con- 
 flicting lines of thought in Kant, which arise from his regarding the movement 
 of reflexion by which the universal principle of experience is discovered as 
 merely a movement of abstraction. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 419 
 
 while the consciousness of the outer world as such includes all 
 the content of inner experience and something more. On the 
 other hand, Kant here also provides the means for the correction 
 of his own error, when he maintains that inner experience pre- 
 supposes outer experience. Por if that be the case, then the 
 former cannot be . taken as more simple than the latter, 
 but on the contrary, the latter must be taken as containing 
 all that is in the former plus an additional reflexion.^ For in 
 the first place, it is shown that, just as the consciousness o,f the 
 unity of the subject gives a new character to that unity, 
 which may be said- even to become for the first time a self 
 or ego when it thus returns upon itself ; so, the consciousness 
 of the inner succession of our affections, as determined in con- 
 formity with the unity of self, is more than the mere fact 
 of such determination. And, in the second place, if, as Kant 
 admits, such consciousness in both cases presupposes thei 
 consciousness of objects, it must contain that consciousness 
 as a necessary element in itself, or it must itself be that 
 consciousness and something more. And as pure self-con- 
 sciousness includes all the categories by which objects are 
 determined as such and carries them back to their principle, 
 so the consciousness of a self as -one with itself through all 
 the changes of its inner experience, must contain all the 
 variety of an outer experience, with the further qualification 
 that such outer experience is at the same time the history 
 of a mind, — a mind whose consciousness of itself is developed 
 by the same process whereby its knowledge of objects is 
 increased. 
 
 Now, a clear apprehension of this truth, — that the regress The error in 
 
 ' ':r o Kant s view is 
 
 by which self-consciousness is developed, is at the same time a '^l^^^oKhe 
 progress, — will save us from that subjective conception of of BerSy, ° 
 knowledge which was initiated by Berkeley, and from the 
 
 ' It should, however, always be remembered that, as has been shown above, 
 m an organic development every new element involves a reoonstitution in 
 relation to it of all the other elements. 
 
420 THE CRITIQUE OF PURE REASON. book i. 
 
 illusions of which Kant did not altogether escape. For 
 Berkeley, reflecting on the relativity of the consciousness 
 of the object to the self, maintained that the percvpi of 
 things is their esse, in the sense that the real and only 
 
 -■&" 
 
 objects of our consciousness are our own " ideas " or sen- 
 sations, as states of our own subjectivity. He did not see 
 that a reflexion which would resolve our knowledge into the 
 affections of an individual subjectivity, is in contradiction with 
 itself. For the subject which is conscious of- its ideas as its 
 own, and refers them to objects, is not the individual sensitive 
 subject as such, but an ego which, as it is conscious of itself only 
 in distinction from, and in relation to, objects, cannot reject the 
 consciousness of objects as unreal. If the object be reduced 
 to a state of the subject, the subject ceases ipso facto to be 
 an ego ; and a self which knows nothing but its own states 
 is an absurdity, a cross between a sensitive subject which does 
 not know but merely feels, and a self-conscious subject which 
 can be conscious of itself only as it is conscious of objects. 
 If Berkeley had realised this.^ he would have seen that the 
 true meaning of the reflexion that objects exist only for a 
 subject is, not that objects are reducible to the sensations 
 through which we know them, but that we know no objects 
 except those which are relative to a self, which therefore 
 require to be contemplated in that relation in order that their 
 true nature may be seen. But this implies, not that the 
 objective consciousness must be reduced to the merely sub- 
 jective, but that we must retract the abstraction in which we 
 regard it as merely objective, and correct errors into which we 
 fall in so regarding it. Of course, at the same time, we must 
 equally retract the abstraction in which we regard the subject 
 as having a pure inward life of its own, or a consciousness of 
 its ideas as mere states of itself apart from their reference to 
 objects. 
 
 1 Of course there are elements in Berkeley's doctrine, especially in his later 
 doctrine, which may be regarded as anticipations of Kant's view. 
 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 421 
 
 Now, with Kant we partly escape from this tangle, in so far ^^^i^g'^^to 
 as the regress upon the unity of the self with itself in all the thoughhedoes 
 
 „.,.,. . . „ not altogether 
 
 successions of its ideas is seen to presuppose a consciousness of escape from it 
 external objects. But as I have already suggested, Kant did 
 not recognise all that is involved in this dependence of inner 
 upon outer experience. Just as he took the pure consciousness ' 
 of self to be merely an analytic consciousness of the abstract 
 identity of the ego, without observing that the consciousness of 
 that identity changes its character ; so he took the conscious- 
 ness of the unity of the ego with itself in the succession of its 
 inner experiences as merely a consciousness of the determinate 
 succession of its subjective states, without observing that the 
 consciousness of such a succession changes the character of the 
 states which form its parts. As in the former case, the mere 
 identity of a sensational subject is changed into an "I " or self, 
 which is its own object in distinction from, yet in relation to, 
 other objects ; so, in the latter case, the successive sensational 
 states are changed into the connected parts of an intelligible con- 
 sciousness of the objective world, as it becomes known in time 
 to one self In other words, we cannot think of the succession 
 of sensations as states of a self ; for as parts of the experience 
 of a self, they are no longer its sensational states; and, 
 conversely, the states of a sensitive subject are not parts of 
 the experience of a self. Hence, the transcendental regress 
 cannot be a mere return upon the process whereby the sensa- 
 tional states of a sensitive subject are changed by the activity 
 of an ego into a consciousness of objects ; it must be the. 
 discovery of what is latent in such a consciousness of objects 
 as belonging to a self It must, in short, be a reflexion 
 upon the distinction of consciousness from self-consciousness, 
 of outer from inner experience, which discloses the unity pre- 
 supposed in that distinction as it exists for the ordinary 
 consciousness. Now, Kant goes so far in this direction as to 
 show that self-consciousness is dependent upon a consciousness 
 of objects, and to explain the latter as the recognition that the 
 
422 THE CRITIQUE OF PURE REASON. book I. 
 
 unity of perception, which is the result of the unconscious 
 working of the understanding under the name of imagination, 
 corresponds to the unity of conception expressed in the cate- 
 gories. But he falls back into the subjective attitude of 
 Berkeley in so far as he speaks of an ego armed with categories 
 by means of which it has to turn certain data of sense, given 
 to it from without, into a consciousness of objects. Such data, 
 however, as he is immediately obliged to acknowledge, cannot be 
 given to the self-conscious ego as such, which presupposes the 
 consciousness of objects ; and to say that they are given to a 
 subject which ultimately becomes conscious of itself as an ego 
 when it has determined them in relation to objects, is not to 
 the purpose, for such a subject cannot be conceived as the 
 source of categories by which the object is determined as such. 
 As has already been pointed out, the transition from a sensitive 
 to a knowing and thinking consciousness cannot be explained, 
 as Kant attempts to explain it, by the " determination of the 
 sensitive by the intellectual ego " ;^ for when the intellectual 
 ego appears, the merely sensitive ego has disappeared ; and the 
 same process which has turned the sensitive subject into a 
 thinking consciousness, has turned sensation into perception. 
 But this means that we cannot explain the consciousness of the 
 object or of the outer world, through a determination of the 
 subject by itself: i.e., a determination of the supposed passive 
 subject, as affected by the thing in itself, by an active subject 
 which does not yet exist. What we can do in the way of 
 explanation of the objective consciousness is, not to go back 
 upon something more simple, upon a unity of the self and a 
 manifold of the sensations which combine in some, way to 
 produce it, but to go forward and to correct our first view of 
 the object as a thing in itself by bringing out its relation to 
 the self : for the object is truly seen only when it is viewed as 
 a factor necessary to the completion of self-consciousness. 
 Thus only does transcendental reflexion enable us to correct 
 iR. I. 501 ; H. VIII. 5.31. 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 423 
 
 our primary ideas of the external object as external to the self, 
 and of the self as conscious of itself in an inner life from which 
 the object is excluded. It makes us regard the external world 
 as not only revealed to mind but as coming to self-conscious- 
 ness in it, and the mind as coming to the consciousness of 
 itself only as it goes out of itself to determine objects. 
 
 It appears, then, that the supposition that inner and outer The deveiop- 
 
 rt^ ' ' rr mentofthe 
 
 sense are essentially different, or present different objects to "^"b'Strem- 
 us, really arises from a confusion of the successive states of a ?e°paratedfrom 
 
 •11 11 ni 11 • ^^® develop- 
 
 sensitive subject with the development oi Icnowledge m a ment of seu- 
 
 consciousness, 
 
 thinking consciousness such as ours. For the fact that our 
 consciousness of the objective world is a growing and changing 
 thing, seems at first to imply that that consciousness is 
 subjective, though it may be continually advancing towards 
 objectivity, i.e., continually approximating to the comprehen- 
 sion of that which exists independently of our knowing it : and 
 this again involves the conception of an object independent of 
 thought, which by successive efforts on our part is brought 
 within its' compass. Self-consciousness, therefore, appears as a 
 narrower circle, which gradually takes in matter from without, 
 until finally its object becomes identical with the object as it 
 is ia itself When, however, we consider the matter more 
 closely, we begin to see that as within and without, subject 
 and object, are strictly correlative, so the presence or absence 
 of a knowledge of the one cannot be separated from the 
 presence or absence of the knowledge of the other. 'Sor is the 
 growth of the knowledge of objects a process in which , the 
 mind can be sfeid to go out of itself, as distinct from the process 
 whereby it goes into itself or comes to itself. All ignorance of 
 the object is ignorance of the self, all development of con- 
 sciousness is also a development of self-consciousness. An 
 object that we may, but do not, know has for its counterpart a 
 potentiality in us of perceptions which we have not realised, a 
 ' faculty which we have never used.' The consciousness of defect 
 in oi»r knowledge of the world is a consciousness of disunion 
 
424 THE CRITIQUE OF PURE REASON. book i. 
 
 in ourselves ; or, what is the same thing, it is a consciousness 
 of union with, and at the same time of separation from, a 
 perfect intelligence for which the process of development is 
 completed. From this point of view the Leibniziau and the 
 Lockian views of knowledge are necessary complements of each 
 other : for while, on the one hand, nothing new can come to 
 existence for us, except by the same process whereby we 
 acquire a new consciousness of ourselves ; yet, on the other 
 hand, in so far as the process is one in which new knowledge 
 of objects and of ourselves is actually acquired, we must regard 
 our knowledge in both aspects of it, and therefore our existence 
 as conscious subjects, as derivative. To say that we know 
 nothing purely a priori, but only gradually come to know the 
 world as it reveals itself to us, is another way of describing the 
 same fact, which is expressed when we say that our conscious 
 life is the realisation in us of a perfect intelligence ; i.e.., of an 
 intelligence which knows all that as self-conscious subjects we 
 have the possibility of knowing, and, therefore, is all that we 
 can become. The history of a self-conscious being is thus in 
 one sense a time-process, in so far as in time he advances from 
 knowledge to greater knowledge : but in another sense it is 
 not in time, in so far as the consciousness it implies remains 
 identical with itself in all its acquisitions, and can make no 
 acquisition of any knowledge of other objects, except that 
 which gives it a deeper understanding of itself and satisfies a 
 need present to it from the first and involved in its ideal' of 
 knowledge, 
 though The only modification which we can make in this statement 
 
 logically self- 
 
 oonsciousnesB gf \^q corrclativitv of Self and not-self, of inner and outer, is 
 
 IS posterior. " i t~ ^ 
 
 one which is just the reverse of that suggested by Berkeley 
 and Kant ;■ viz., that in self-consciousness we have a movement 
 of thought which is logically posterior to the consciousness 
 of objects, in inner experience a movement which is logically 
 posterior to outer experience. It is, indeed, only in this move- 
 ment of return upon itself that the object gets its distinct 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 42'5 
 
 qualification as an object ; it is only in contrast with inner expe- 
 rience that outer experience is determined as such, and therefore 
 the statement just made remains correct. Still, it is necessary 
 to call attention to the fact that self-consciousness is logically 
 .posterior to the consciousness of objects and includes it, in 
 order to make way for the transcendental reflexion, which sees 
 in the former at once the principle presupposed in the latter 
 and the goal to which it tends. The final interpretation of the 
 world must be idealistic or speculative ; it must correct, not 
 only the materialism which springs out of our natural abstrac- 
 tion from the subject, but also the dualism which treats subject 
 and object as co-equal factors, by showing that the correlativity 
 of the object and subject is a correlativity for the subject. 
 Thus, it must " raise consciousness to the form of selfrconscious- 
 ness" and show outer experience to be an element in inner 
 experience ; or, what is the same thing in other words, it must 
 explain the world as the self-manifestation of a spiritual prin- 
 ciple, which, therefore, must be a manifestation not only to, 
 but in, spiritual or self-conscious beings. 
 
 Taking this view of the process of knowledge, we may now summary of 
 sum up the result of the transcendental Deduction and show tiieDeduction. 
 its relation to the parts of the Analytic for which it prepares 
 the way, especially for the discussion of the Schematism 
 and the Principles of Pure Understanding. The transcen- 
 dental Deduction is intended to prove that a list of cate- 
 gories, which has been based upon the logical analysis of 
 the judgment, has objective application ; because it contains 
 the conceptions, without the application of which to the 
 data of sense there would be for us no experience, and hence 
 no objects of experience. In order to show this, it calls 
 attention to the fact that objects of experience are objects for 
 a self. But the manifold data of sense can be brought into 
 relation to the 'J think,' only in so far as they are synthetically 
 combined into one consciousness according to certain general 
 principles, and in so far as they are recognised as so combined. 
 
426 THE CRITIQUE Oi' PUEE REASON. book i. 
 
 Hence the necessity for an imaginative synthesis of the mani- 
 fold in perception, and for a recognition of this synthesis as 
 conforming to certain a 'priori principles- But as such recog- 
 nition is involved in every act of judgment in which we 
 determine an object, the conceptions involved in the act of 
 judgment are those by which the unity of objective consciousness 
 is realised, and realised in such a way as to make it possible that 
 the consciousness of objects may be united with the consciousness 
 of self. The result then is that we are obliged to regard the 
 whole of our intellectual life as a process to self-consciousness — 
 a kind of circular movement by which the mind goes out of 
 itself to determine objects, in order that through these it may 
 come to a consciousness of itself. The possibility of self-con- 
 sciousness . is the ultimate term to which the question of the 
 possibility of experience points ; for the question of knowledge 
 is the question how an object can exist for me, or be united 
 with the ' I think.' 
 Its main defect IsTow, the main defect which we have discovered in this 
 
 lies in the con- 
 ception of the reasoning of Kant was that, while he spoke of judgment as 
 
 unity of self- o ' r J o 
 
 asanaS^c.^^^ essentially that determination of objects through which they 
 are at the same time brought into relation to the self, and 
 while it was just for this reason that he regarded the " func- 
 tions of unity " in judgment as supplying the categories by 
 which objects as such are determined, he yet based his list of 
 the categories upon a different view of judgment as the expres- 
 sion of the analytic, and therefore the merely subjective, unity 
 of consciousness. To trace back this defect to its source, we 
 needed only to observe how Kant regarded the unity of self- 
 consciousness, the judgment ' I am I,' as analytic ; for this 
 necessarily involved that the categories, as ' species of appercep- 
 tion,' i.e.., as aspects of the unity of consciousness with itself in 
 all its apprehension of objects, cannot be synthetic (= unities; 
 of differences) ; or at least that they cannot be synthetic in 
 their own right, but only, if at all, in virtue of a relation into 
 which they are brought to an externally given difference. We 
 
CHAP. IV. THE Transcendental deduction. 427 
 
 have seen, however, that Kant, in effect, refutes himself when 
 he points out that, except in relation to the synthetic unity of 
 experience, the analytic unity of self-consciousness does not 
 exist at all. Self-consciousness is, in fact, the synthetic unity 
 returning to its principle. Hence it is quite true that the so- 
 called analytic judgment of self-consciousness reveals the 
 principle of all objective synthesis ; but this is because it is 
 not really an analytic judgment, but rather a judgment in 
 which the difference has become transparent. We thus reach 
 the idea of a metaphysical Deduction of the categories, which 
 would be more satisfactory than that of Kant, and which would 
 account both for the successive steps in the. process of know- 
 ledge and for the ideal of knowledge to which it points as its 
 culmination.^ 
 
 The second stage of Kant's deduction involves a similar Tte same en-or 
 
 appears in the 
 
 movement of thought. In it Kant brings into view the fact f?'=™<i part in 
 
 ^ ° tl^e conception 
 
 that our experience is an experience of external objects in mo?e"ab6tract 
 space and time, and of an object-self which is known under experience, 
 conditions of time ; and he seeks to prove the validity of the 
 categories in relation to such an experience. Here, therefore, 
 we have to consider that the determinations of outer sense are 
 taken up into consciousness by a successive synthesis, the 
 elements so taken up being combined according to certain 
 principles into images of external objects under the forms of 
 space and time ; while our consciousness of an inner life 
 belonging to the self arises only when we recognise the 
 successive process involved in such synthesis, which is 
 really a process of ' determining the sensitive self in us 
 by the intellectual self.' Kant, therefore, supposes that we 
 are conscious of ourselves as empirical objects, only as we are 
 conscious of the determination of our sensitive subjectivity 
 to that definite successive synthesis of the manifold by which 
 the objective consciousness is produced. The recognition of 
 
 ^In the two following chapters we shall have to discuss more in detail 
 the possibility and nature of such a deduction. 
 
428 THE CRITIQUE OF PURE REASON. book: 
 
 this fact, therefore, is supposed to enable us to make a transcen 
 dental regress upon the development of our experience, and t 
 detect the manner in which the unity of the self, which i 
 expressed in the categories, becomes a principle of determina 
 tion for an inner sense through its form of time. Ii 
 other words, it enables us to detect that schematism of th 
 categories in relation to time, by which our experience i 
 mediated. For, as in Kant's view the categories are merel; 
 the conceptions of the laws which must control and regulati 
 all finite minds in the formation of images of sensible perceptior 
 in so far as these are to be capable of being brought into relatioi 
 to a self ; so the schematised categories are merely the form 
 of these conceptions which must guide us as men — i.e., as being 
 whose inner sense is conditioned by time — in combining ou: 
 perceptions with our consciousness of self. Now, the criticisn 
 to be made upon this view of inner experience is quite paralle 
 to that made upon Kant's view of pure self -consciousness. Kan 
 acknowledges the consciousness of the succession of our experi 
 ences to be possible, only through the determination of externa 
 objects ; but yet he supposes the former to be a simple analyti 
 return from the latter, instead of being, as it really is, a syn 
 thetic advance to a new qualification of the former by refer 
 ence to an element which, though present, is at first only laten 
 in it. Hence, here again we find that what Kant really teachei 
 us is to correct our first inadequate view of the objects of oute 
 experience, by showing that outer experience is ultimately ti 
 be explained as an element in inner experience. For, as tb 
 existence of objects is essentially an existence for a self, si 
 they must ultimately be regarded as factors — real factors 
 indeed, but still merely factors- — in the process of a spiritua 
 life. In this way, it is possible to recognise in Kant' 
 schematism of the categories, and in his justification of thi 
 schematised categories as applied to objects of experience, 
 higher meaning than is at first obvious ; and also to see that w 
 can do justice to Kant only when we recognise this. ■ For wha 
 
CHAP. IV. THE TRANSCENDENTAL DEDUCTION. 429 
 
 Kant really proves is that the categories, so far as they are 
 'species of apperception/ or expressions of the different 
 momenta in the pure consciousness of self, must necessarily 
 be at the same time the guiding principles in all the different 
 stages of our consciousness of a world, the knowledge of which 
 can be completed only when it is brought into the form of 
 self-consciousness, ie., when the external or material world is 
 recognised as in reality a spiritual world, as the phenomenon of 
 which the spiritual world is the noumenon. 
 
 Such reinterpretation of Kant would be arbitrary, if we how far did 
 
 Kant realise 
 
 attributed to Kant himself a consciousness of the result to ^iHrgument? 
 which his reasoning points ; but without it we cannot possibly 
 do justice to the truth that is in him, and which is ever and 
 anon breaking through the mask of his immediate statements. 
 Without it, e.g., we cannot understand how he should say that 
 the transcendental deduction does not seek to exhibit the 
 genesis of experience, but to " show what is in it," i.e., to bring 
 the ^ordinary, and also, it may be added, the scientific, con- 
 sciousness to a clear understanding of itself. For this means 
 that I in our theory of knowledge we cannot go back upon 
 exp^ience, i.e., either upon the ordinary or the scientific con- 
 sciousness of objects, in the sense of reducing it to something 
 morel simple, out of which it was formed : we can only make 
 it more complex by bringing it to self-consciousness, or enabling 
 it to discover in itself latent elements of which it was not 
 wont to take any notice. "What Kant did not fully see is that 
 such a view of the true nature of experience cannot stop short 
 at the mere recognition in it of elements which were not formerly 
 recognised, and the consequent admission of the imperfect truth 
 of a knowledge in which they were left out of account, — which is 
 what Kant means when he says that our experience is confined 
 to phenomena. It must necessarily lead to a transformation of 
 that knowledge in view of those factors in it which were formerly 
 neglected, but are now brought to light. It must be, not 
 merelfi a regress by which the foundations of our experience 
 
430 THE CRITIQUE OP PURE REASON. book 
 
 are detected, but a progress by which it is reconstituted, 
 must not merely show that objects, as we know them in t] 
 common consciousness and even in science, are not things i 
 they are in themselves ; but, in showing us this, it mu, 
 supply, at the same time, the elements which were wantii 
 to enable us to apprehend-^ese objects in their truth < 
 complete reality. 
 
431 
 
 CHAPTEE V. 
 
 THE SCHEMATISM OF THE CATEGORIES. 
 
 THE Deduction of the Catesjories is based on the necessity The categories 
 ° ^ as prmciples 
 
 of the combination of all perceptions with the ' I think,' mi^umlrak 
 in order to experience or knowledge. For such combination is time! 
 possible only as perceptions are determined in themselves or in 
 their relations by certain principles, and recognised to be so 
 determined. Only so can they be referred to any object, hold- 
 ing a place in the one world of objects, which, as such, is 
 distinguished from, and related to, one conscious self. Again, 
 these principles were regarded by Kant, as derived from the 
 _ conceptions, or rather, we might say, the forms of conception, 
 involved in the process of judgment, as that is analysed by 
 formal Logic. As so derived, however, they are pure deter- 
 minations of analytic thought, determinations of the subject or 
 of the predicate or of their relations in the analytic judgment : 
 hence they require to receive concreteness in relation to time, 
 as the form of inner, and therefore of all, sense, ere they can 
 be applied to any object presented in time. As so " concreted I' 
 with time," they are supposed to be taken out of the simple | 
 tautology of the analytic judgment, and made into principles / 
 for the determination of all objects which are given under the The judgment 
 forms of time, i.e., of all objects of sense whatever. "s tS'sub"''^ 
 
 -.J. sumption of an 
 
 As Kant puts the matter, then, the judgment of experience 'mage of per- 
 
 ^ ' ' •• ^ J- oeption under 
 
 is to be analysed into a conception of a relation of, or in, time, ^tego™*'*** 
 
432 THE CRITIQUE OF PURE REASON. book i. 
 
 which is got by determining the category in relation to time ; 
 and, on the other hand, into, a perception or appearance 
 presented to us under conditions of time as the form of inner 
 sense. TKe judgment is the synthesis of these two elements, or 
 the subsumption of the latter under the former ; and this, as 
 we have seen, is equivalent to the determination of that as an 
 object, which, in the first instance, was merely an appearance 
 of that object for us. The general type of the judgment of 
 experience, therefore, is the assertion that the perception of an 
 object falls under the conception of it ; or, if we like so to 
 express it, that the subjective unity of a perception (which, if 
 expressed in a judgment, would be only the assertion that the 
 appearance here before me, qua appearance, has certain predi^ 
 cates,) is conformable to a relation, which, as it exists only 
 for thought, is not limited to present perception, but is the 
 determination of something that is objectively true, i.e., true 
 independently of my present consciousness of it. For, so soon 
 as we have realised that we cannot know things in themselves, 
 and that, therefore, ' object ' can only mean that which has 
 existence for a conscious self, we must also recognisp that 
 objectivity is nothing but universal validity or vahdi.ty for • 
 " consciousness in general." The reference of the perception 
 to the conception in judgment may be expressed as the asser- 
 tion that the perception is -.a case of the conceptioii ;. an 
 assertion which involves that the object which I perceive, has 
 a reality independent of my particular perception, because the 
 rela,tions that define it for me, as it is now perceived, are 
 themselves independent of time. In other words, the pure 
 conceptions determine perceptions as having their elements 
 united in time with each other or with other percepti!0,ns, 
 according to certain general types of relation. The proposition, \ 
 therefore, whicli predicates the conception of the .§er_ception, is 
 synthetic and synthetic a 'priori; for it takes the isolated percep- ' 
 tion, which, as such, is confined to a moment of time, and so de- 
 termines it as to draw from it a rule that holds good for all time. 
 
CHAP. V. THE SCHEMATISM OP THE CATEGOE.IES, 433 
 
 Now, we have already seen how this should be understood. ^^J'g'fJgJ^'to 
 It is not true if taken to mean that we have conceptions here ^^ rapawe^' 
 and perceptions there, prior to the act of judgment, yet each percepSms.^ 
 determined as it is in the judgment ; and that the act of judg- 
 ment merely brings the latter under the former. For this 
 would make the act of judgment purely analytic ; and it would 
 imply that both perception and conception were all that they 
 could be, apart from their union in the judgment. But it is 
 true in the sense that it is just in and through the act of 
 judgment, that the perception and the conception get their 
 qualification. Or, to put it otherwise, it is the same act of '^ 
 
 : consciousness whereby the conception is distinguished from, 
 and related to, the perception, which also gives such deter- 
 mination to the latter that it can be recognised as a case of 
 the former, and to the former such determination that it can 
 be, recognised as the principle of unity in the latter. Now, 
 this is just the point to which Kant calls our attention, when 
 he introduces the schema between conception and perception. 
 The unschematised conception, as derived from the analytic '' 
 judgment, shrinks into a mere form for possible conceptions ; 
 just as the mere perception, as given in sense and unrelated to 
 
 . the unity of conception, shrinks into a mere manifold : but by 
 the interposition of imagination with its schema, Kant seeks to 
 realise the conception, and to idealise the perception, and so to ^ 
 make them meet in the judgment of experience. The necessity 
 of the conception being schematised before it is applied, corre- 
 sponds, on, the other side, with the necessity of a reproductive 
 synthesis to give rise to the individual image, which, 'when 
 brought into relation to the schema, is recognised as a case to 
 which it can be applied. But the final qualification of each, 
 whereby -they are so distinguished and related, must be re- 
 garded- as, that which is realised in the judgment, rather than 
 as that which- is presupposed by it. 
 
 Keeping this in view, and remembering also the opposition Distinction of 
 
 and reiation of consciousness and self-consciousness with which ™ ™*^°- 
 VOL. I. 2 b 
 
434 THE CRITIQUE OF PURE REASON. book I. 
 
 Kant connects it, we shall be less likely to misunderstand his 
 procedure in the chapter on the Schematism of the Categories, 
 in which, as is his uniform practice, he starts on the level of 
 ordinary psychological or logical distinctions, in order to make 
 his way up to the transcendental theory of knowledge. The 
 idea of a schema, accordingly, is first illustrated by that in- 
 choate activity of imagination which is implied in recognisiag 
 an object as one of a certain species : for, as this involves the 
 subsuming of a particular under a universal, it implies that we 
 realise to ourselves the rule of specification for the universal. 
 In thinking such a rule, the general conception is not, indeed, 
 envisaged in an individual object; but the conditions of its 
 envisagement in such an object are realised, in such a way that 
 the particular image of conception can be brought under it. 
 Thus the judgment, as the synthesis of conception and percep- 
 tion, is mediated by an activity which plays between the two ; 
 which schematises the conception that it may be applied to the 
 perception, and gathers the data of sense into a definite image, 
 so that they may be brought into relation to the conception. 
 How i* ^^ts^Y Now, the schematising, which is the first part of this process, , 
 conception. ^^^ ^^ generally described as a kind of realisation in relation 
 to space and time of conceptions which, in pure thought, are 
 apprehended without regard to space and time. In a mere ■ 
 conception, we have a simple unity of abstract elements, which 
 must not logically exclude each other, but which otherwise 
 may be what you please. But in the schema we bring such a 
 conception into relation to the conditions of time and space ; 
 and these may affect the possibility of the objects conceived in 
 various ways. For, on the one hand, it may be possible, by the 
 aid of the additional conditions introduced into the schema, to 
 combine in one subject different elements which the conception 
 alone would not enable us to bring together ; and, on the other ; 
 hand, what we can combine in the conception we may be incap- 
 able of exhibiting as united in the schema. An instance,of the 
 former is the idea of change; for, the object that changes being j] 
 
CHAP. V. THE SCHEMATISM OF THE CATEGOKIES. 435 
 
 represented as in time, we are able to combine in it contradic- 
 tpry predicates, i.6., as succeeding each other. As an example 
 of the latter, we may take the idea of a triangle, the angles of 
 which are less or more- than two right angles ; an idea which 
 is not excluded by the law of identity, but by conditions 
 revealed only^ in the construction of the conception in space.^ 
 
 Now, when a conception is generalised from many individual the pure*"' 
 images, there seems to be no difficulty in reversing the process """"^p ^'"^"' 
 of abstraction, so as to restore to the abstract idea some of the 
 characteristics of which it has been robbed. But the pure^ 
 conceptions of the understanding, as Kant maintains, are not ,' 
 produced by abstraction from any perception; and, moreover, as i 
 pure conceptions, they are not homogeneous with the percep- | 
 lions to which they have to be applied. They are pure_i 
 functions of unity in thought, different determinations of the 
 analytic judgment vSrhich expresses only the unity of thought 
 with itself ; while the manifold data of sense are combined in 
 relations of space or time, — relations which imply no necessity 
 of connexion in these data, and therefore no objective unity 
 corresponding to the imity of thought. If, however, we can 
 charge the former with difference, so as to make them express 
 universal relations of different elements as in time, then, as 
 time is the necessary form of all perception, perception may be 
 recognised as conformable to conception. And then, moreover, 
 we can go on to argue that, since such recognition is necessary 
 to self-consciousness, it is not only possible but necessary in the 
 determination of objects, which must be capable of being united 
 with self-consciousness. 
 
 It is important here to observe that the Schematism is made The schemata 
 
 supply what is 
 
 necessary simply and solely by Kant's view of self-conscious- ^^IV'l^ 
 ness. On his view, self-consciousness, though presupposing the *£ pe^SpUon 
 consciousness of the objective world, is yet not regarded as tkm."™"^^' 
 containing it in itself. Self-consciousness is an analytic judg- 
 ment, because it is simply a regress upon the unity which is 
 * ' A. 716 ; B. 744. Cf. Dissertation, § 5. 
 
436 
 
 THE CRITIQUE OF PURE REASON. 
 
 BOOK 
 
 presupposed in the consciousness of objects. Hence, no speci 
 mediation is necessary to explain how the pure unity of thoug] 
 as expressed in that judgment, should differentiate itself and ! 
 bring itself into relation to the data of sense. If Kant hi 
 seen that the unity of the self correlates with the differen( 
 and manifoldness of the object, he would have had no need i 
 a middle term to bring together factors which cannot be sepa 
 ated. If there is no purely analytic unity of thought wit 
 itself, as opposed to its synthetic unity in knowledge, thei 
 cannot be any separate forms of thought as a subjecti\ 
 process, apart from the forms by which it determines object 
 But Kant, having put the analytic judgment mid-way betwee 
 the pure self and the determination of the object, and haviu 
 thus deprived thought of its synthetic character, in virtue ( 
 which it has relation to objects, is obliged to restore the loi 
 synthetic movement by an external reflexion of the forms ( 
 thought upon the forms of perception. Thus the categorie 
 understood as the relations implied in pure thought, relatioi 
 which have no reference to time, became interpreted as leh 
 tions of what is presented as in time, which maintain themselve 
 irrespective of the change of times. On the other hand, timi 
 as a form of sense, is a form under which we can have pn 
 sented to us only individual objects as such, in the narrowes 
 limitation to the time in which they are presented. But whe 
 we bring them both together in the judgment, which indi 
 vidualises the conception and generalises the perception, we a 
 the same time recognise that what is presented to us here an 
 now, in a particular perception, is an instance corresponding t 
 a relation which holds good at all times ; or, what is the sam 
 thing, we are conscious that we have before us an objec 
 which exists independently of its presentation in this particula 
 case. 
 
 If we thus work out the idea of the unity of the universi; 
 and the particular, of conception and perception, in the judg 
 his premises, ment, we See that Kant's mediation of each moment by th 
 
 Kant's 
 argument, 
 therefore, 
 necessarily 
 leads to an 
 
CHAP. y. THE SHEMATISM OF THE CATEGORIES. 437 
 
 others must necessarily reduce them to relative elements which 
 exist only in this unity. The reciprocity of determination 
 between the two terms, which is thus disclosed, reduces their 
 difference into a difference of correlative elements ; and, at the 
 same time, it makes unnecessary the interposition of any middle 
 term to connect them. In this way the apparent circle which 
 is involved in Kant's assertion, that the principles of pure 
 understanding are proved through an experience which they 
 make possible, becomes intelligible ; whereas it cannot be made 
 intelhgible so long as any of the elements are supposed to have a 
 character independent of their relation to the others. There is, 
 however, great advantage in following Kant in his analytic 
 method of exposition, if only we understand that his conclusion 
 involves the recasting of his premises. For, it is just his merit 
 that he shows us the very process in which, starting from the 
 ordinary dualism, we reach a higher conception of knowledge. 
 To state this more definitely : — Kant's argument involves msviewof the 
 
 *' ^ connexion of 
 
 the isolation of the act of judgment ex parte nostra from the and the''*'" 
 determination of the object, which is just the same act viewed judgment, 
 from the other side. As a consequence, he has to look for a 
 medium of qualification, through which the subjective con- 
 sciousness of combination of ideas (in our status representativus) 
 may be determined as a consciousness of the combination of 
 qualities in objects. And this he finds in the idea that pure 
 consciousness, which expresses itself in the analytic judgment, 
 can determine the empirical consciousness through its form, in 
 such a way as to fix the combination of its elements in time 
 by universal rules of relation ; and that thus it can give 
 objective reality to such combination, i.e., can determine it as 
 independent of the special time-relations in which these elements 
 are given to the individual subject. In this process, therefore, 
 the predicate, which is the conception, becomes so determined 
 in relation to time, that it can be applied to the subject, which 
 is the perception, in such a way as to fix its constitutive rela- 
 tions ffor all time. 
 
438 THE CRITIQUE OF PURE REASON. book I. 
 
 SadLS'toa ^^*' ^^ *^^^ ^^y °^ putting it, Kant seems to take for 
 the°rSlHo°n of granted that we have presented to us in time, altogether apart 
 
 perceptionand - . • t i i ■ j r» 
 
 conception in trom any synthetic act, a particular phenomenon or ' object of 
 of knowledge, perception,' in which manifold elements are brought together as 
 constituting an individual object, while yet that object is only 
 characterised as an appearance which may never recur again. 
 Ifow, Kant himself recognises that an individual object as such 
 cannot be presented to us in sensation, but only by means of a 
 synthesis of imagination, which retains the elements that have 
 been given in sense and combines them into one whole ; a 
 whole, therefore, in which all the parts are already taken out 
 of their existence as merely successive feelings, and qualified 
 by their relation to each other. If I say that I perceive such 
 an object now, I am saying that I now stand in a relation to 
 something, which is what it is for me irrespective of the time- 
 succession of my sensations. In other words; the object of 
 perception I have before me, is already determined by the 
 conceptions under which, as it is supposed, it needs to be 
 brought in order to determine it as an object of experience. 
 All that remains to be done, therefore, when we bring together 
 the conception and the perception, is to recognise that the latter 
 is already qualified by that universal relation which is ab- 
 stractly expressed in the former; to recognise, in short, that 
 " the synthesis of imagination universally expressed is the con- 
 ception of the understanding." It is true that the fact that I 
 have a sensation is included in the conception of the object, 
 which I call a perceived object. But the inclusion of this does 
 not make it less a conception ; still less does it reduce it to a 
 §ensation. If it did, Kant is ready to tell us that such a 
 perception would be for us as good as nothing. Already, 
 therefore, in a perception which can be subsumed under the 
 schematised conception, that qualification by conception must 
 have taken place which the theory supposes the judgment to 
 give by so subsuming it. For, as perceived, the object has 
 already been taken out of the abstract singularity or singleness 
 
CHAP. V. THE SCHEMATISM OF THE CA.TEGOEIES. 439 
 
 in time, which belongs to the sensation. And this is what 
 Kant really shows when he points out that objectivity and 
 universal validity for consciousness are the same thing. For if 
 we have once recognised that a sensation has any reference to 
 an object more permanent than itself, (and this is already 
 involved in everything we can call a perception,) we are no 
 longer dealing with that which has individuality in such a 
 sense as to exclude determination by relation to other sensa- 
 tions, or to require that such determination should be brought 
 to it from without. On the other hand, if such determination 
 be excluded from the sensation, it ceases to be an object which 
 we can subsume under a conception. Kant, however, taking 
 his start, as usual, from those very conceptions which it is the 
 result of his argument to transform, seems here to admit that 
 we have an inner sense in which individual objects or images 
 are successively presented to us in time, with a view to their 
 subsumption under the categories, i.e., he speaks as if inner 
 sense, undetermined by the categories, were already perception 
 of a succession of particular facts or objects ; while yet, at the 
 same time, he regards those perceptions as having the single- 
 ness of mere sensations. Hence, he supposes them, in the 
 latter character, to need the application of the categories to 
 make them objective, and, in the former character, to be capable 
 of this application. But if we take it in the former sense, 
 mere perception is already the experience which was supposed 
 to be derived from it by the application of the categories ; and 
 if we take it in the latter sense, it is only a series of sensations, 
 which cannot even be conscious of itself as a series, still less 
 as a series of perceptions of particular objects. We can, indeed, 
 vindicate Kant to some extent by referring to what he else- 
 where says, to the effect (1) that the synthesis of imagination, 
 by which perception is brought about, is conformable to the 
 categories, and (2) that the consciousness of self in inner ex- 
 perience is possible only in relation to outer experience. But 
 when<Bffe make the correction necessitated by these two admis- 
 
440 THE CRITIQUE OF PUEE REASON. book i. 
 
 sions, there is no longer any need to schematise the conception, 
 with reference to its use as a predicate for perceptions given 
 independently. And what we must substitute is the idea that 
 the perception, in being determined as a perception, i.e., as the 
 consciousness of a present object, has already been brought 
 under a conception. This, indeed, does not affect the truth of 
 the doctrine that knowledge is a judgment in which perception 
 and conception are determined and referred to each other. For, 
 as was sufficiently shown in the last chapter, the synthetic act 
 of knowledge has two aspects or factors : it is at once a 
 synthesis of the manifold according to a conception, and a 
 recognition of the conception as the principle of that synthesis ; 
 and this duality in unity is what is involved in calHng it a 
 judgment. But the fundamental error of Kant's statement lies 
 in this, that it suggests that we can find the conception com- 
 plete here and the perception complete there, prior to the act 
 by which they are fitted together, — a view which leads to the 
 result that the process of so fitting them is either unnecessary 
 or impossible. Eeally it would be truer to say that judgment 
 is the differentiation of conception and perception in order to 
 their reunion. For, on the one hand, if we go back beyond the 
 judgment, we arrive at a unity in which the difference of con- 
 ception and perception has not yet shown itself. And, on the 
 other hand, the difiference of conception and perception, which 
 shows itself in the judgment, is necessary to that identification 
 of the two, which is also implied in the judgment in so far as 
 it determines the latter by the former. The particular and the 
 universal must be separated in order that either may gain that 
 qualification which it has only as referred to the other. If, 
 however, we adopt this view, we must regard the dualistic 
 beginning of Kant, — whether it be in the form of the opposi- 
 tion of the thing in itself to the ego in itself, or in the form 
 of the opposition of the analytic unity of self-consciousness to 
 the manifold of sense, or finally in the form of the opposition 
 of the unity of the schematised conception to the perception of 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 441 
 
 the individual as determined in time and space, — as a mere 
 abstraction, which is made in order to show that these elements 
 cannot be separated. For if, in any way, we begin with a 
 dualism, it is impossible for us to get beyond it except by re- 
 ducing it from an absolute to a relative dualism ; or, in other 
 words, by disclosing a unity which always underlies and is 
 presupposed in the opposition of the two terms, though not 
 consciously present in our first apprehension of it. In other 
 words, the development of knowledge is impossible, unless in 
 each factor all are involved, or unless in the difference of the 
 factors the unity is involved. As a consequence, we are obliged 
 to conceive the judgment, not as a combination of what was 
 previously distinct, but as a separation, in order to a more per- 
 fect unity, of elements which were previously undistinguished, 
 and which, even in being distinguished, are never wholly 
 separated. In this way, we may regard it as a process which 
 is at once analytic and synthetic : which produces the difference 
 it overcomes, and which it is on the way to overcome even in 
 the act of producing it. For the unity has begun to emerge, 
 even in that first dualistic consciousness in which the terms 
 are characterised only in opposition to each other. 
 
 In the light of what has been said, we may now proceed to Beaimeanmgr 
 
 o ' ./ IT of the asser 
 
 consider the Schematism of Kant, as a translation of pure o'^egoriraare 
 ' functions of unity ' (supposed to be involved in the analytic apperception, 
 judgment) into terms of time, i.e., into definite ' conceptions of 
 objects in general,' by which data of sense given under condi- 
 tions of time are to be fixed as elements of an objective 
 consciousness which is independent of time. It has been 
 pointed out in a previous chapter that Kant in his meta- 
 physical deduction already qualifies the pure unity of thought 
 with itself by relating it to the manifold : for the addition of 
 the third category in quantity and quality, (which, as Kant 
 states, involves the two others,) is confessedly due to a 
 consideration of the judgment as relating to objects ; and the 
 very ideas of relation and Tnodality, so far as they go beyond 
 
442 THE CRITIQUE OP PUBE REASON. book i. 
 
 abstract relations of comprehension and extension, are illegiti- 
 mate additions to a list of categories derived from the analytic 
 judgment.^ If there is a pure self-assertion of thought, with- 
 out reference to the manifold of objects, we cannot get the 
 categories from it. On the other hand, if in that self-assertion 
 the relation to the manifold is involved, then the whole concep- 
 tion of thought or judgment which is set before us in formal 
 Logic must be rejected. In truth, what we get by abstraction 
 of the subjective from the objective, of the pure relation of 
 thought to itself from all matter, is, as Kant himself shows, the 
 extinction of all distinction or relation in a bare " X that 
 thinks " and of which we cannot even say that it is an ' I ' ; 
 since there can be no ' I ' which is not object to itself (so that 
 " I am " = " I am I "), and " the analytic unity of apperception 
 is possible only in relation to the synthetic unity." This 
 being so, the derivation of the categories from the unity of self- 
 consciousness must take another aspect. They must present 
 themselves as different stages in the synthetic process of self- 
 consciousness, the process whereby through the determination 
 of the objects, there is developed in us a consciousness of self 
 as that to which all objects are relative. And this, as we have 
 seen, is involved in the reasoning of the Transcendental Deduc- 
 Eeiationof tion. For, the fundamental conception of the Transcendental 
 
 the Kantian 
 
 and Hegelian Deduction is that all determination of obiects is predetermined 
 
 Logic. "' -"^ 
 
 in relation to the possibility of self-consciousness ; or, in other 
 words, that every step in the process whereby the conscious self 
 goes beyond itself to determine objects is a step towards the 
 final movement by which it returns to itself. It is, therefore, 
 no mere confusion of different points of view, but simply a 
 recognition of the ultimate meaning of Kant's thought, when 
 we assert that the principle of the Hegelian Logic lies in germ 
 in the Transcendental Deduction of the Categories. For what 
 Hegel attempts to show is just that the categories by- which 
 thought must determine its object, are stages in a process that, 
 1 See above, p. 339. 
 
CHAP, V. THE SCHEMATISM OE THE CATEGORIES. 443 
 
 beginning with the idea of ' Being,' the simplest of all deter- 
 minations, is driven on by its own dialectic till it reaches the 
 idea of self-consciousness. In other words, the intelligence, 
 when it once begins to define an object for itself, finds itself 
 launched upon a movement of self-asserting synthesis, in which 
 it cannot stop till it has recognised that the unity of the object 
 with itself involves its unity with all other objects and with 
 the mind that knows it. Hence whatever we begin by saying, 
 we must ultimately say ' mind.' The idea of self-consciousness 
 may be truly said to contain all the categories which the self- 
 conscious subject can apply to any object ; or these categories 
 may be regarded simply as different steps in the movement by 
 which thought, through determination of its object, comes to a 
 consciousness of itself. 
 
 Kant, it need not be said, does not so develop his Transcen- t'ansfflOTi™*^ 
 dental Deduction. He was precluded from doing so by his aiSfytic^tothe 
 
 . synttietic 
 
 conception that the pure judgment of self-consciousness is unity, 
 merely analytic, instead of being, as it is, the most perfect of 
 synthetic judgments. Hence, he could find in its forms only 
 pale spectres of abstraction, which could not of themselves do 
 duty as categories. When, however, in spite of the conception 
 of these functions of unity in the analytic judgment, Kant still 
 tries to derive the categories from them, simply by considering 
 them as determinations of objects as such, or as principles of 
 unity in a ' manifold in general,' he is already breaking down 
 the division he has established between pure thought in its 
 unity with itself and pure thought in its relation to perception ; 
 and he is showing that the analytic unity already contains the 
 synthetic. He proceeds, indeed, in a somewhat mechanical 
 method in his advance toward a recognition of the synthetic 
 movement of thought, rather externally adding one idea to 
 another than showing that the one cannot exist without the 
 other, or that taken in its abstraction it breaks down into con- 
 tradiction. What he shows is,- not that one element necessarily 
 involves another, but only that it cannot exist without the 
 
444 THE CRITIQUE OF PURE REASON. BOOKi. 
 
 other, if a certain result is to be achieved. Still, after all, the 
 result to be achieved turns out to be simply the unity of the 
 elements through which it is to be achieved ; and under the 
 guise of a mechanical combination of independent things, we do 
 not find it difficult to detect the vital or organic connection, 
 which Kant gradually reveals to our eyes. If we foUow 
 Kant's process to the last, the externality of his method takes 
 the appearance of a scaffolding, which falls away when it has 
 done its work. 
 It is in the First, then, as we have seen, Kant gets the categories, as 
 
 schematism 
 
 tha^ tiie third ' couceptions of objects in general, by a reference of the pure 
 comes'into fuuctious of Unity in analytic judgment to a ' manifold in 
 exis ence. general.' Then he proceeds farther to schematise these by 
 reference to the form of time, under which our manifold is 
 actually given. Then, going on to describe the schematic 
 process, he makes still farther advances in the same direction. 
 Thus in the case of the first two classes of categories, (which 
 alone it can seem plausible to derive from the analytic judg- 
 ment), it is the third category alone that is schematised ; 
 and the third category, as Kant himself teUs us, " demands 
 a peculiar act of thought," not involved in the first two 
 categories, an act of thought which, as he elsewhere says, 
 is essentially synthetic. Thus, ' Number ' is the schema of 
 Quantity : but Number is a form of the category of Totality^ 
 which cannot possibly be extracted out of Unity and Plurality; 
 in other words, it is a conception, not of the one or of the 
 many as opposites, but of the one in the many. In like 
 manner. Degree is the schema of Quality : but Degree, as 
 treated by Kant, is a form of the category of Limitation, which 
 unites Eeality and Negation. In truth, on Kant's own showing,, 
 the third category in each of these cases does not belong to 
 pure thought as reflected in the analytic judgment at all, but 
 makes its appearance only in that application of pure thought 
 to the form' of perception which Kant calls Schematism. This 
 
 ^B. 111. 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 445 
 
 is shown conclusively in the chapter of the Critique upon the 
 'Amphiboly of the Eeflective Conceptions ' : for what Kant is 
 there maintaining is that, while unity and plurality, positive 
 reality and negation, are absolutely separated in pure thought 
 and its objects, i.e., in things in themselves,^ they can be, and 
 indeed necessarily are, brought together in empirical knowledge 
 and its objects. In opposing Leibniz, therefore, Kant accuses 
 him not of inaccurate thinking, but simply of treating the 
 objects of our knowledge as if they were objects of pure thought. 
 If pure thought were capable of determining objects as Leibniz 
 took it to be, Kant admits that it must necessarily regard 
 unity and plurality, reality and negation, as pairs of absolute 
 opposites which could not be combined or reconciled in these 
 objects. Thus, to take first the category of Quantity, the joteiit"*'*'' 
 Leibnizian principle of the ' Identity of Indiscernibles ' is ifS".*^ 
 attacked by Kant, not on the ground that it involves an 
 irrational conception of quantity,^ but on the ground that it 
 involves a confusion of phenomena with things in themselves, 
 i.e., with things as determined by pure thought. Of the latter, 
 Kant holds that the principle would be true, which obviously 
 excludes the idea that things in themselves, i.e., things as con- 
 ceived in pure thought, should be ' numerable quanta.' Hence 
 the idea of extended matter as a ' numerable quantum ' can 
 arise only as we ' generate time ' in the apprehension of the 
 elements of perception, i.e., as we distinguish and relate the 
 elements successively apprehended (which may be parts of 
 space, as well as of time, but in both cases involve, according 
 
 ' Things in themselves, it is presupposed, loould be the objects of pure 
 thought, if pure thought could have any objects. The dogmatic Metaphysic 
 of Leibniz, as Kant often points out, treats pure thought as capable of deter- 
 mining objects in spite of its analytic character ; it turns propositions which 
 are true of mere conceptions into determinations of objects. 
 
 ^ Rather it involves the rejection of the idea of Quantity altogether as a 
 'synthesis of homogeneous units.' If the units cannot be regarded as homo- 
 geneous, if they are altogether incommensurable (like Plato's ideal numbers) 
 there can be no quantitative determination of their sum, i.e., there can be no 
 roomlEcr the idea of Quantity. 
 
446 THE CKITIQUE OF PURE REASON. book i. 
 
 to Kant, a distinction and a synthesis of times in which they 
 are apprehended). Only as we add together elements which 
 are homogeneous units, can we combine unity and plurality in 
 the conception of a whole number ; and the homogeneous units 
 so combined must primarily be times (or spaces). On the 
 other hand, no object can be conceived as existing in time or 
 space, which is one with itself in such a sense that it is not 
 capable of being divided and regarded as a number of units. 
 Thus it is the reflexion of thought upon the forms of percep- 
 tion which makes it possible and necessary that quantity 
 should be represented as number, and so as at once unity and 
 plurality.^ 
 idmitatfon '^^® ^^^ ^^ regards the categories of Beality and Negation is 
 i?el?-ee!*^ a little more difficult to disentangle, but the result is similar. 
 The analytic Logic rests on the idea that affirmation and 
 negation are absolutely exclusive, i.e., exclusive in such a way 
 that the exclusion does not leave room for any relation be- 
 tween them. This view is accepted by Kant when he attacks 
 Leibniz for his reduction of evil to mere nothingness- or ab- 
 sence of good. " When we represent what is real by the pure 
 understanding alone, (as realitas nouvienon,) we find ourselves 
 obliged to exclude as unthinkable any opposition between 
 realities, i.e., any relation in which they cancel each other's 
 effects, so that 3 — 3=0. Contrariwise, things phenomenally real 
 may conflict with each other, and as united in one subject, they 
 may wholly or partly annihilate each other's effects, as is the 
 ' case with two moving forces on the same straight line, which 
 push or pull in opposite directions, or a pleasure which holds 
 
 1 There is a certain difficulty in bringing together the two passages on 
 Number in the chapter on the ' Amphiboly ' and in that on the ' Schematism!' 
 We have to observe, Ijowever, that in the latter chapter Kant is maintaining 
 against Leibniz the possibility of a multitude of homogeneous units as objects of 
 experience : while in the former, he is showing the necessity that the category 
 of quantity should be schematised as number, with a view to its application to 
 objects of experience. From this necessity it follows, that no object can be 
 presented to us in experience which is not, or does not admit of being, regarded 
 as a multitude of homogeneous units. 
 
CHAP. V. THE SCHEMATISM OF THE CATEG-OKIES. 447 
 
 the balance against a pain."^ In other words, the conditions of 
 perception, time and space, make possible a union of opposites 
 in one phenomenal object, which would not be possible in an 
 object of pure thought. The chapter on Schematism supplies 
 the complement to this when it maintains that the category of 
 Quality must be schematised as Degree or Intensive Quantity ; 
 which is represented as a mean between complete vacuity 
 and absolute fulness, between pure negation and pure position. 
 Thus Kant lays the basis for the doctrine, not only that real 
 opposites maij be envisaged as existing in objects known under 
 the conditions of space and time, but that objects known under 
 these conditions must be envisaged as realities limited by other 
 realities, the affirmation of which therefore is qualified by 
 negation and so determined as degree of reality.^ 
 
 Before going on to consider the other schemata, let us imperfection 
 
 ° ° ' of the form In 
 
 examine what is involved in the view here taken of pure th^dcategory 
 thought in its relation to the forms of sense. It involves, in "^'^p^'^"- 
 the first place, that a ' one ' which is also a ' many,' and in like 
 manner a ' reality ' which includes ' negation,' are untliinhahle 
 or cannot be regarded as possible determinations of objects, if 
 objects are to be determined by pure thought alone. But, in 
 the second place, both become intelligible as determinations of 
 phenomenal objects, that is, objects in determining which the 
 unity of thought is applied to the manifold of sense. This 
 agrees with the doctrine that in pure self-consciousness, i.e., in 
 the consciousness of the thinking subject of itself as thinking, all 
 difference and negation is excluded. Yet, on the other hand, it 
 is equally Kant's doctrine that it is only in relation to the syn- 
 thetic unity of the consciousness of the object that the analytic 
 unity of self-consciousness is possible. Hence the necessity for 
 that curious see-saw already noticed in the last chapter. For, if 
 the thinking subject can be conscious of itself as an object, i.e., 
 
 ' A. 264 ; B. 320. 
 
 ^ The full explanation of this conception must be reserved for the chapter on 
 the Mathematical principles. 
 
448 THE CRITIQUE OF PURE REASON. book i. 
 
 can affirm itself and recognise its unity with itself, only in relation 
 to that which has difference in itself; and if in that relation it 
 canThot affirm itself or he conscious of itself in the pure self- 
 iderdity of its being, (seeing that this would involve a removal 
 of all such relation,) it follows that the process of knowledge is 
 necessarily a progress toward a self-consciousness which can 
 never he attained : a progress which is necessary, hecause 
 thought must apply itself to that which is not itself in order to 
 become self-conscious, and endless, hecause in relation to what 
 is not itself, it cannot become conscious of the pure self. As 
 Schelling puts it, the process of experience is a continual 
 " struggle between the incapacity of uniting factors absolutely 
 opposed, on the one side ; and the necessity of uniting them, if 
 the identity of self-consciousness is not to be cancelled, on the 
 other." ^ Or it is a process in which the object is alternately 
 taken into, and expelled from, the consciousness of the self, 
 because of the supposed necessity of consciousness of the object 
 to self-consciousness, while yet, as purely analytic, the latter 
 seems to exclude the former. 
 Contradiction Such a thcorv of knowledge, however, involves an altema- 
 
 involvedin ^ •' ° 
 
 synthesis"' *^°^ °^ *^° points of view which are never really brought 
 tMrddategorj! together. It is impossible to conceive that the consciousness of 
 self as occlusive, of the consciousness of the object, and the con- 
 sciousness of self as inclusive of the consciousness of the object, 
 should be identified. The unity must be regarded as prior to 
 the difference and overreaching it, if through the difference 
 that unity returns upon itself, and so becomes conscious of 
 itself. And if the unity thus returns upon itself through the 
 difference, it cannot but, in its full development, reduce the 
 difference to an element in, or a necessary expression of itself 
 If this be so, we are obliged to admit a different view both of 
 the relations of pure thought in itself and of its relation to the 
 manifold in space and time, which in knowledge it has to 
 subsume, i.e., of the conceivable in itself and in relation to the 
 'Schelling's Werhe, III. 394. 
 
CHAP. V. THE SCHEMATISM OP THE CATEGORIES. 449 
 
 perceivable. Instead of saying that for pure thought unity is 
 absolutely opposed to plurality and plurality to unity, we shall 
 be obliged to say that for pure thought the one is nothing 
 except in relation to the many : and instead of saying that for 
 pure thought affirmation is absolutely separated from negation, 
 we shall be obliged to say that determinatio est negatio, i.e., that 
 affirmation can be conceived of only in relation to negation. 
 Further, in both cases the relation will not be a relation of 
 things externally combined, but of things each of which exists 
 only through its relation to the other, or, as we may even 
 express it, is only its relation to the other. In other words, 
 what really exists is the unity in which each is one element, 
 the one in the many, the reality which returns upon itself 
 through its negation. Hence thought does not require to be 
 drawn out of itself by a shock from something foreign, ere it 
 can differentiate itself, and by synthesis of differences attain 
 consciousness of its own unity. It has the principle of 
 differentiation in itself — -i.e., in itself as pure thought, without 
 relation to anything externally given, even to the pure mani- 
 fold of space and time. On the other hand, if we can conceive 
 thought as coming into any relation to such a manifold, it must 
 be because the consciousness of such manifold is only a further 
 stage in the evolution of itself, and not the application of it to 
 a foreign matter. Tor the same reason the synthesis of such 
 matter cannot be limited to an external synthesis — as of that 
 which cannot be brought into unity with the form of thought 
 applied to it— but must go on to the recognition by thought of 
 its own unity in the matter. "We may, indeed, admit that 
 there is a stage" in the process of knowledge in which thought 
 deals with the manifold, in space and time, as if with a foreign 
 matter, in which it seeks but in which it cannot find itself ; and 
 that in this stage only a relative synthesis is possible, i.e., a 
 synthesis of objects with each other ly thought, which is not 
 their synthesis with thought. But this imperfect unity of 
 
 thought with its object will be. possible only in so far as 
 VOL. I. 2 F 
 
450 THE CRITIQUE Or PURE REASON. book l 
 
 thought does not bring to bear on its objects its own highest 
 categories, i.e., that idea of its own unity with itself which is 
 at the same time the idea of its unity with all that is an 
 object for it. 
 
 Eeturning again to the categories before us, the application 
 of the remarks is obvious. Kant's reasoning practically in- 
 volves that the category of the one in the many, i.e., of Totality, is 
 possible to us only in the form of Number or Extensive Quantity, 
 which is generated by the schematism of the pure category of 
 Quantity in relation to time. Hence, Totality must for us 
 take the form of Extensive Quantity or Number, i.e., the 
 synthesis of homogeneous units, and the thought of a com- 
 pleted totality, a totality brought back to unity, is a mere idea, 
 which is contradictory as a pure conception, and unrealisable 
 in a perception. In like manner, the category- of Limitation or 
 Determination can take for us only the imperfect form of the 
 conception of one reality as limited by another, and not that 
 of a being, which is self-limited and self-determined. The 
 third category is, therefore, only an idea, except in the form of 
 a synthesis of elements which remain external to each other. 
 A new pro- So far, howcver, we have been dealing only with the cate- 
 
 Dlem arises in o ./ 
 
 ^tegories'of" gories of Quantity and Quality. The categories of ' Eelation,' 
 ■^yhich we have next to consider, stand on a somewhat different 
 footing, for they are categories of reflexion. In other words, as 
 Kant puts it, they are categories which go in pairs,^ substance 
 and accident, cause and effect, etc. : they express a relation of 
 two elements, which are at once necessarily distinguished and 
 necessarily connected. This difference between the categories 
 of Relation and those of Quantity and Quality may seem to be 
 cancelled by what has been said of the Schematism of the two 
 latter ; for, as we have just seen, it is Kant's view that in re- 
 lation to time, unity and plurality, reality and negation, are 
 brought together as correlated factors in the ideas of number 
 and determinate reality (or degree of reality). 
 
 B. 110. 
 
CHAP. V. THE SCHEMATISM OF THE OATEGOEIES, 451 
 
 The difference, however, is twofold : for, in the first place, poS™o*n*'' 
 the categories of Eelation involve a more perfect kind of syn- Sp^V?^"^"' 
 thesis of opposites than is involved in number and degree, in 
 which the different elements combined are conceived as ex- 
 ternally limiting each other ; and, in the second place, this 
 synthesis is regarded as already completed in pure thought; 
 for the categories of Eelation, already as mere conceptions, 
 express that duality in unity or unity in duality which, 
 as we have seen, the categories of Quality and Quantity 
 gain only by being schematised. On Kant's principles, it 
 is difficult to see how the analytic unity of thought 
 should give rise to any conception of a unity of differ- 
 ences apart from the schematism ; and still more difficult 
 to conceive how, even with the aid of the schematism, it should 
 give rise to a reflective conception in which the difference is 
 one of correlative elements. For, in the case of the Categories 
 of Quantity and Quality, it was the schema which was supposed 
 to enable us to bring together opposites which for pure thought 
 were absolutely separated : and, as we have seen, it did not 
 enable us to bring them, together as correlates, but only as 
 limiting each other. To allow pure thought to differentiate 
 itself without relation to any manifold, or, when differentiated, 
 to recover its unity with itself by reducing the difference to a 
 difference of elements that essentially imply each other, was for 
 Kant to retract all that he had said of the analytic character 
 of the movement of pure thought. All that can be done to 
 explain this inconsistency is to point out that, as Kant derives 
 the categories from the Judgment, in which, however formally 
 it may be conceived, the unity of thought appears as containing 
 a duality of elements, — a subject and' a predicate which have 
 at least a relative difference, — he took this duality for granted. 
 If, however, it be admitted that such a duality of aspects is 
 necessary or even possible for pure thought as such, we must 
 cease to treat the judgment in which pure thought expresses 
 itself .as> a mere movement by identity, and the synthetic judg- 
 
452 THE CEITIQTJE OP PURE REASON. book 
 
 ment as an external bringing together of elements previous] 
 unrelated. We must substitute for both a view of thej'udgmei 
 as a movement to the differentiation of elements, which, on] 
 as differentiated, can he held together in unity. Kant was ni 
 prepared thus to fill. up the gulf between the analytic and tl 
 synthetic judgment. But he seeks, as it were, to bridge 
 over, first by taking the duality of thought as already give 
 even in the analytic judgment, which enables him to deri-\ 
 from the latter the categories of relation in a somewhat atteni 
 ated form ; and, secondly, by using the schema as a meai 
 of turning this analytic duality into a synthetic duality, s 
 that, e.g., reason and consequent become translated into cauf 
 and effect. 
 The peculiar It appcars, then, that the categories of Eelation involve' 
 
 character of rr ' ' a 
 
 of RdlSi™ kind of synthesis of conceptions or elements of conceptio; 
 ^oe'iJahe'^' which, if we hold to the principle of Kant's own treatment" i 
 ofKanfs the Categories of Quantity and Quality, cannot be brougl 
 
 philosophy. ° 
 
 together except through ' a third something,' to wit, the fori 
 of time ; and they involve a more perfect synthesis than can I 
 achieved even with the aid of that form. For, in the refleetii 
 synthesis of substance and accident, cause and effect, action ar 
 reaction, each term has meaning only in relation to the othe 
 which at the same time it excludes. Here, therefore, we ha^ 
 at once negation and af&rmation, unity and difference ; for eac 
 term is at once negatively and positively related to the othe 
 and the whole thought is of a unity of differences which 
 once repel and attract each otherj If, therefore, it be tn 
 that even in attributing to pure thought the origination of t] 
 categories of Number and Degree, Kant is giving to it 
 synthetic movement of which, on the formal view of it accept 
 by him, it is not capable ; still more obvious is the inco 
 sistency of attributing to it the origination of categories 
 Eelation. Accordingly, we find that it was one of the 
 categories, viz., the category of causality, that first gave r 
 to Kant's doubts in relation to the dogmatic metaphysic 
 
CHAP. V. THE SCHEMATISM OF THE CATEaORIES. 453 
 
 Wolff, to which his invariable objection was that it treated 
 the analytic movement of thought as if it were a synthetic 
 movement, and pretended to get difference out of identity^ 
 And Hume's scepticism had such influence upon him, mainly 
 because it suggested the counter problem, how to teduce to 
 identity things given as different. Kant's first problem in 
 1763, was how conception can 'fertilise itself or add new 
 matter to its content ; and at that time he had no solution to 
 give except that thought must wait on- experience, and accept 
 from experience whatever connexions of phenomena it may 
 present, as the presupposition of its own analytic action. But 
 Hume showed this answer to be unsatisfactory, when he main- 
 tained that experience can give us only the post hoc and not 
 the propter hoc, one thing after another and not one thing as 
 necessarily connected with another. Given one thing, how can 
 you get another from it 1 was the first question of Kant, which 
 forced him to seek in experience for that synthesis which the 
 intelligence seemed unable to explain. Given two different 
 things in succession, how can you establish a necessary relation 
 between them ? was the question of Hume,^ which forced Kant 
 again to look to the intelligence itself for a priori principles of 
 synthesis, which, it was obvious, could not be given in sense 
 any more than they could be accounted for by the movement 
 of pure thought. Hence arose the double-edged problem of 
 the Critique, which is hidden under the ambiguity of the word 
 synthesis : the problem how, on the one hand, by perception we 
 are able to add new matter to conception, and how, on the 
 other hand, by conception we are able to connect the new 
 matter added with the old. This double problem is, Kant 
 holds, to be solved by the relation of pure thought to the 
 manifold of sense through its a priori forms ; for, as he con- 
 tends, it is in this relation that the unity of thought develops 
 into conceptions, which supply principles of synthesis whereby 
 
 'I have, however, pointed out above, p. 203 seq., that Kant sometimes 
 confuses the two points when he is speaking of his relation to Hume. 
 
454 THE CEITIQTJE OF PURE REASON. BOOK I. 
 
 that manifold may be brought back to the unity of thought. 
 Such a development has been shown to be necessary even in 
 the case of the categories of Quality and Quantity, which are 
 schematised as Number and Degree. But these schemata do 
 not yet explain how the unity of thought should ever suggest 
 the idea of a necessary connexion among the elements of the 
 manifold, which was the special form of synthesis attacked by 
 Hume. For Hume took the synthesis of external relation in 
 time and space for granted, as given in experience, and merely 
 raised the question as to the necessity of the connexion of 
 things so related, and as to the origin of the very idea of such 
 a necessity. Kant universalises the problem, and asks for the 
 origin of the syntheses of quantity and quality implied in the 
 determination of objects in time and space, as well as of the 
 synthesis of relatioir ; and he further distinguishes the two 
 questions of the origin of the conceptions, and of their applica- 
 tio7h to the manifold. Now, it is the former question we have 
 been treating of in this chapter; and we have already seen how he 
 attempts to show that through schematism the analytic unity 
 of thought, in the pure conceptions of Quantity and Quality, 
 turns into the synthetic unity of the conceptions of Number 
 Origin of the and Degree. But then he goes on to contend that, as that 
 
 jiecemiy of the 
 
 expressed in ^'^^^J expresscs itsclf in a judgment in which subject and 
 categories, predicate are analytically and therefore necessarily connected, 
 so the different elements of the manifold brought into relation 
 to the unity must be determined as necessarily connected with 
 each other. Further, he maintains that as the judgment has 
 three forms which may determine the analytic unity of subject 
 and predicate, so the synthetic unity of necessary relation in 
 the elements of the manifold appears in the th^ee forms of 
 substance and accident, cause and effect, and reciprocity, which 
 again, as referred to time, are translated into the three relations 
 of the permanent to the changing, of antecedent to consequent, 
 and of coexistent to coexistent. The dif&culty here lies in the 
 assumption that thought without being synthetic should yet 
 
CHAP. V. THE SCHEMATISM OP THE CATEGORIES, 455 
 
 develop into an analytic judgment, which again by an ex- 
 ternal reflexion upon the manifold as given in tinie, becomes 
 a synthetic judgment, or a conception of elements that are 
 distinguished and at the same time necessarily connected. And 
 this difficulty is heightened by the way in which Kant reaches 
 to the synthetic conceptions of Number and Degree as relative 
 combinations of momenta which pure thought cannot combine, 
 but the combination of which is made possible by the ideas 
 of space and time. If, however, we recognise in the so-called 
 analytic judgment of self-consciousness, not the abstraction 
 from aU synthesis, but rather the perfect type of synthesis 
 as the return of thought upon its own unity, we shall 
 not find it difficult to see how the intelligence should 
 give rise to the idea of a necessary connexion in all objects in 
 space and time which it brings into relation to that unity. 
 In fact, the determination of objects as standing in qualitative 
 and quantitative relations to each other, and the determination 
 of them as necessarily connected with each other, would seem to 
 be just steps in the process of knowledge by which these objects 
 are brought back to the unity of thought, — a process which can 
 end only in the recognition of the unity of the intelligence with 
 itself as maintaining itself in all its consciousness of objects, 
 and, therefore, of the unity of all objects not only with each 
 other, but with the mind that knows them. Kant's account 
 of knowledge as an act of judgment really involves this. Por 
 in the act of judgment we have not, as formal Logic supposes, 
 the mere conception of an object as self-identical, but the 
 determination of it' by distinction and relation, and also the 
 recognition that in and through this distinction there is a unity 
 which maintains itself or returns upon itself ; and consequently 
 that the object is ultimately one with the thought that appre- 
 hends it — a truth which is partially expressed by Kant himself, 
 when he says that judgment is the expression of the objective 
 unity of apperception J Peouii.rities of 
 
 This last remark may fairly introduce the consideration of ^^oaXyT 
 
456 THE CRITIQUE OP PURE REASON. BOOK I. 
 
 i™p°^srbmty the schematism as applied to -the categories of Modality. 
 Ixpfination of These Categories, according to Kant, express relations, not of 
 Kantian objccts to cach other, but of objects to the mind which knows 
 
 principles. ' 
 
 them. Fow, it was the principle of the Transcendental Deduc- 
 tion that all our ideas must be capable of being united with 
 the consciousness of self, and that they can be so united only 
 as they are combined into one objective consciousness. Thus, 
 the unity of all objects with the conscious subject goes along 
 with the unity of all objects with each other as elements in one 
 experience. The mind that is to be capable of knowledge must 
 be capable of the double synthesis of objects with each other 
 and with the self for which they are objects ; or, to put it 
 otherwise, must be capable of producing by its own activity 
 the categories, or forms of thought, necessary to establish such 
 relations. On the other hand, it seems impossible to explain 
 the existence of such functions of thought, so long as thought 
 is regarded as ruled by the law of identity. How can a con- 
 sciousness, gifted only with analytic powers, be capable of 
 synthetically combining the objects which it apprehends ? 
 Still more, — how can such a consciousness go beyond itself to 
 apprehend objects which are other than itself, so as to bind up 
 the consciousness of them in the one consciousness of itself? 
 Now, we have seen already that Kant was involved in great 
 difficulties in answering the first of these questions, and that 
 his introduction of the categories of relation as necessary cate- 
 gories of thought could not be justified so long as thought 
 was taken as purely analytic : (though it might be explained by 
 the fact that this analytic thought was yet conceived by Kant 
 as expressing itself in the form of judgment.) For, if such 
 categories are legitimate, then thought does not necessarily 
 determine its object as abstract identity apart from all rela- 
 tion ; but, on the contrary, it necessarily relates the object 
 which it determines to other objects in the very process of 
 determining it : and we can meet Kant's question, " How are 
 we to understand that because one thing is asserted, something 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 457 
 
 different should be also asserted as flowing from it ? " with the 
 
 answer, that we can understand nothing else ; or, in other 
 
 words, that we cannot understand the assertion of a thing 
 
 from which nothing else should follow. But if this be 
 
 admitted, the second question is less difificult to answer. If 
 
 thought is necessarily determination by relation, i.e., if it 
 
 necessarily goes beyond the object it apprehends, to unite it 
 
 with other objects, it becomes intelligible how it should go 
 
 beyond its own bare identity to apprehend objects, and also 
 
 how it is just by this going beyond itself that it becomes 
 
 conscious of itself, or returns to itself; so that, in Kant's 
 
 language, the analytic unity of apperception is dependent on 
 
 the synthetic unity. For, if thought is essentially and in 
 
 itself synthetic, we can understand how it maintains its 
 
 identity in the differences it produces or apprehends, and how 
 
 through them it becomes conscious of itself 
 
 We have now seen the relation of the different schemata to The schema- 
 tism is virtu- 
 
 the categories and the forms of judgment from which they are t|on of the"'*' 
 derived, and we are in a position to review the whole process of'thraghtf " 
 and understand its meaning. The Schematism is a middle 
 term which is supposed to unite perception and conception 
 with a view to knowledge. Such a middle term is made 
 necessary by Kant's analytic view of pure thought, which 
 requires that the intellectual ' functions of unity ' should be 
 "concreted with time," in order that they may be brought into 
 relation with perception. On consideration, however, it appears 
 that the scaffolding which Kant thus erects really hinders us 
 from seeing the nature of the building he is constructing. 
 For what he really gives us is a disguised refutation of the 
 analytic view of thought with which he started. In the first 
 place, thought, taken as a movement by identity, is yet sup- 
 posed to be essentially judgment, and that in spite of the 
 admission that the analytic unity is possible only through the 
 synthetic. Being taken in the form of judgment, thought 
 alreadf involves the three functions of thesis, antithesis, and 
 
458 THE CRITIQUE OF PURE BEASON, book i. 
 
 synthesis : i.e., it involves the idea of a unity which through 
 the difference of subject and predicate returns upon itself. In 
 this way, it becomes easy to extract from it a number of 
 categories, each expressing one of the different phases through 
 which thought moves in its affirmation of itself. In the next 
 place, this shadowy differentiation, which Kant is never weary 
 of declaring to be in itself empty of meaning, is, so to speak, 
 substantiated, by reflecting it upon the contents of inner sense 
 through its form of time — a process which would be nugatory, 
 unless we supposed all the categories in their difference from 
 each other as species of apperception to be already derived from 
 pure thought ; for all it can possibly explain is a qualification 
 of the pure relations of thought by the externality of time, 
 which, e.g., translates the relation of cause and effect into 
 invariable sequence. In truth, if we held to the idea of 
 thought as pure identity, there would be only one, or at most 
 two, ' functions of unity ' to qualify ; and the schematic quali- 
 fication of such functions could not produce a multitude of 
 categories. Further, if thought needed the aid of the idea of 
 time to attain even such an imperfect unity of negative and 
 positive, of unity and plurality, as is implied in the ideas of 
 Number and Degree, we could not suppose it to be capable of 
 developing the categories of Eelation and Modality (which 
 involve higher forms of such a unity) either with or without 
 the aid of the idea of time. This, as we have already seen and 
 as will be shown more fully in the sequel, is admitted by Kant 
 himself in regard to the categories of modality, which express 
 the unity of thought with the object from which it yet dis- 
 tinguishes itself. Hence, for him this unity shrinks into an 
 unrealisable idea. But it is equally true of the categories of 
 relation which express the essentially reflective character of 
 It Is an uncon. thought, a character which it is vain to suppose that it can 
 
 scious refuta- ,, ^ . .-.-. ..-„ 
 
 tionoftha acquirc by .reference to time, if it has it not in itself. 
 
 abstract ideas 
 
 andcSpSon ^^ ^^® obliged, therefore, to regard this whole process, in 
 SSiItarts. which thought is gifted with new powers by reference to time, 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 459 
 
 as simply indicating the steps by which Kant gets out of the 
 false position in which he was placed by his original view of 
 thought as in itself purely analytic. It is a refutation of the 
 principles of formal Logic in relation to thought, disguising 
 itself as a correction of that Logic in relation to knowledge ; 
 and its ultimate result is to bring into view the antithetic and 
 synthetic movement of intelligence, which formal Logic con- 
 ceals. The supposed opposition of pure thought to schematised 
 thought is an illusion. It is true that the unity -in-difference of 
 the intelHgence in itself is not equivalent to the unity-in- 
 difference of the intelligence and the intelligible world ; or, in 
 other words, that pure self-consciousness, or the unity of 
 objective and subjective self, is not the same thing with the 
 unity of self-consciousness as including and presupposing a 
 consciousness of the external world. But Kant's way of 
 taking thought as a bare identity and perception as a bare 
 manifold, or the former as purely universal and the latter as 
 purely particular, and then looking for a middle term to con- 
 nect them, involves an abstraction, which, if he had strictly 
 adhered to it, would have made such a middle term impossible. 
 On the other hand, when he goes on to substitute for these 
 abstractions, the idea of thought as judgment and as therefore 
 including in itself a whole series of categories, and the 
 idea of perception as the apprehension of individual things 
 as such, the middle term becomes unnecessary ; for in such 
 thought the universal has already particularised itself, and 
 in such perception the individual is already the particularised 
 universal. The great value of Kant's argument here as else- 
 where lies in this, that it refutes the abstract ways of thinking 
 in which it begins. It is an argument which cancels and 
 reconstitutes the imperfect premises from which it starts. Nor 
 is the value of the instruction it gives to us, when viewed in 
 this way, essentially lessened by the fact that Kant himself 
 never quite realised the full bearing, of his own work, or the 
 inconlistency of the end of it with the beginning. What 
 
460 THE CRITIQUE OF PURE BEASON. book I. 
 
 Jacobi says of the thing in itself, that without it we cannot 
 get into the Critique, and with it we cannot come out of the 
 Critique, is true of almost all Kant's provisional conceptions. 
 And to appreciate the real meaning and value of his method 
 is mainly this — to give him due credit for the process by 
 which he leads us to correct our first notions, while not con- 
 cealing his imperfect consciousness of what he was doing, 
 tavo^ng* *^ In the present case, however, there is something more to be 
 which are ' Said. For, as was suggested in a previous chapter, Kant him- 
 
 expressed in 
 
 of wTtf oriee ^^^' ^^ ^^^^ general remarks on the categories, especially ,in 
 what he says of the relation of the third category in each 
 class to the other two, and again of the relation of th^ mathe- 
 matical to the dynamical categories and principles, does much 
 to help us to a better ' metaphysical deduction of the cate- 
 gories ' from the idea of judgment, than that which he himself 
 has given. The categories are not, Kant tells us, to be taken 
 as unconnected functions of thought, but as a difference of 
 elements in the system, as which the understanding reveals 
 itself in its characteristic act of judgment. Now, judg- 
 ment, as we have seen, is a process which involves both 
 distinction and relation ; we might even say, in Kantian 
 terms, both analysis and synthesis. It is a process in which 
 we can ideally distinguish three stages or movements : first, the 
 thesis, the simple position or assertion of a thing in which it 
 is referred to itself or set before the mind in its self-identity;,.: 
 secondly, the antithesis or determination of the thing by dis- 
 tinction from, and relation to, other things ; and lastly, the 
 synthesis or re-integration of the elements thus differentiated 
 and related — a synthetic movement of which we become con- 
 scious when we bring into view the unity of thought which 
 underlies all such distinction and relation. Or, to put it other- 
 wise, in judgment the mind posits its object, determines it by 
 relation, and recognises its unity with itself in this determina- 
 tion ; or, what is the same thing in another point of view, it 
 recognises the unity of the object with the thought for which 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 461 
 
 it isj Hence, taking judgment as the determination of the real 
 as such, or, in Kantian language, the process whereby " given 
 ideas are brought to the objective unity of apperception," we 
 can trace in it a movement of thought, whereby the real is first 
 determined as if it were a thing in itself, with no necessary rela- 
 tion to anything but itself; then, as a law of relation between 
 phenomena or things which manifest their nature only in their 
 relations; and finally, as a circle of distinguished and related 
 elements, which, therefore, no longer stand in indifferent opposi- 
 tion to the thought which distinguishes and relates them, but 
 find in it their centre. 'We may, therefore, define judgment as 
 at once the distinction of objects from, and their relation to 
 each other, and their distinction from, and relation to, the 
 thought for which they are,— a definition which is nearly 
 eq[uivalent to that given by Kant when he speaks of it as 
 a,' synthesis of the manifold in relation to the unity of the 
 self, and when he points out that the ' is ' of the copula is the 
 expression of the objective unity of apperception. In fact, it 
 is just because the ' is ' of the copula thus indicates the return 
 of thought into identity with itself through distinction and 
 relation, that we have in the judgment, what we have still 
 more explicitly in the syllogism, the essential unity of thought, 
 by which it determines its object and itselfj 
 
 Now, to say that thought is judgment, and that judgment is Judgmental- 
 
 ■^ o J o ' JO wa,ya involves 
 
 always synthetic, is to say (1) that thought always involves g^Jes^^Jt it 
 these three movements : and (2) that they can never be thepredioa™ 
 
 which of them 
 
 separated. In other words, thought, when it becomes fully ^ made 
 self-conscious, cannot but recognise in itself these three phases, 
 which it is impossible to separate, yet which it is obliged to 
 distinguish. It is obliged to distinguish them ; for its synthetic, 
 teeessarily presupposes its thetic, and antithetic, movements : 
 yet it cannot separate them ; for each movement is a step in a 
 process not of mere change, but of development, i.e., a process 
 in which each stage potentially contains the others, and cannot 
 be fully expressed except in relation to the others. Looking 
 
462 THE CRITIQUE OF PURE REASON. book i. 
 
 at it in the former light, and attending only to the necessary 
 distinction of the phases of thought, we may forget that we are 
 dealing with the elements which exist only in a process, and 
 thus we may naturally be led to treat the three attitudes of 
 thought as three orders of categories : first, categories of simple 
 position, i.e., categories which determine an object without ex- 
 plicit reference to anything but itself; secondly, categories of 
 reflexion, by which an object is determined as finding its ex- 
 planation, its essence or principle, in something else than itself; 
 and lastly, categories of ideal unity, by which an object is 
 again referred to itself, but only as it is regarded as a monad or 
 world of relations which has its ideal centre in itself, or, to put 
 it otherwise, only iii so far as it is in unity with the thought 
 for which it is. On the other hand, when we regard these 
 categories as factors or phases in the process of judgment, we 
 have to recognise that they cannot be fixed in isolation from 
 each other, or, in other words, that we cannot understand the 
 categories of simple position and reflexion except in reference 
 to the highest category which is always involved in the unity 
 of the judgment. And the abstraction which hardens itself 
 against this transitional character of the categories and fixes 
 them against each other, must end in reducing them either to 
 tautology or contradiction^ As judgment is synthesis, it cannot 
 be fully understood by us except as expressing the unity of 
 thought with itself through all the differences it states, or, in 
 other words, the unity of the object with itself by negation of 
 all its differences from other things and from thought. But 
 while the judgment is this /or us, or, in other words, while it 
 cannot be fully understood except in this point of view, it is 
 not necessarily this for the individual who judges. For, while 
 judgment always involves, and even expresses, the highest cate- 
 gory, it expresses it only abstractly and imperfectly in the ' is ' 
 of the copula, which does not attract attention for itself, but is 
 regarded only as a connecting link between the subject and 
 predicate. Nay, even the separation of the ' is ' itself is an 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 463 
 
 abstraction which is not often made in actual predication. In 
 
 the judgment the unity of thought remains undeveloped, and 
 
 is, so to speak, subdued to the element it works in. The 
 
 explicit meaning of the judgment is determined by the nature 
 
 of the terms it binds together, and these may or may not 
 
 furnish a matter which is adequate to the form of judgment, 
 
 Thus in a judgment in which a simple quality is asserted to (i) categories 
 
 belong to a given object (' This rose is red ') there is, no doubt, »o««»<m. 
 
 a qualification of that object by negation and relation ; for 
 
 ' determination is negation,' and ' qualities are relations in 
 
 disguise ' : but the negation is not made explicit or attended 
 
 to, and the disguise is not thrown off. The judgment, therefore, 
 
 appears to determine the object only as an isolated unit, which 
 
 has no commercium with anything but itself, and to determine 
 
 it only by a quality which also has no necessary connexion 
 
 with any other quality. It is thus explicitly nothing more 
 
 than the reference of a thing to itself apart from all relation to 
 
 other things. Eelation, indeed, cannot be quite excluded, 
 
 but it appears only in the indifferent form of quantity, in so 
 
 far as the subject in reference to the predicate is thought of as 
 
 one of many actual or possible individuals participating in the 
 
 same quality, and the predicate in reference to the subject as a 
 
 special degree or quantitatively determined modification of a 
 
 given quality. But a quantitative relation is, as Kant points 
 
 out, a relation of homogeneous units which do not require each 
 
 other, i.e., it is a relation which seems not essentially to belong 
 
 to the things related. On the other hand, that which is im- (2) categories 
 
 ° ot Relatiou or 
 
 phcit in sudi judgment, becomes explicit in judgments where deflexion, 
 the predicate expresses an essential relation of the subject to 
 something else than itself (' This ball is elastic ').^ Here the 
 abstract bopula seems to be elevated into an expression of unity 
 
 ' Note that in the Prolegomena (§ 19 note), Kant denies that judgments of 
 experience can be based on judgments of perception, the predicate of which is 
 a secondary quality. This is equivalent to saying that judgments of experi- 
 ence imf ly not only the mathematical but the dynamical principles. 
 
464 THE CRITIQUE OP PURE REASON. BOOK i, 
 
 in difference, though as yet only an external or reilective unity, 
 i.e., a unity of things which are primarily determined as different. 
 The subject is determined by relation to that which is other 
 than itself, and which yet, as so determined, it presupposes. 
 The act of judgment, therefore, does not, as in the judgment of 
 simple position, appear to be a gratuitous or arbitrary division 
 for our thought of that which is essentially one, but rather to 
 express a negative and relative aspect, which belongs to the 
 object as such. At the same time, if in this way the difference 
 of the judgment of Eelation is explained, its unity seems to 
 break down in contradiction ; or the unity of the object with 
 itself which is presupposed, seems to disappear in the relativity 
 asserted of it. It may be added that the thought which appre- 
 hends the relativity seems to move backward and forward, 
 between fixed points which it presupposes as given indepen- 
 dently. In truth, the form of judgment combines position and 
 negation, unity and difference, in a contradictory way, because 
 the relativity of the predicate negates the self-reference of the 
 object. To make the matter of the judgment adequate to the 
 form, it would be necessary that the predicate should express 
 the conformity of the subject in its differences and relations 
 (3) Categories with the ideal principle of unity implied in it. Hence, Hegel 
 
 of Modality or r r J r > a 
 
 Ideal Unity, gj^yg ^jjg^j; ^]^g ^^^gg, Qf judgment is realised only in such judg- 
 ments as " This man is good," which refers to an ideal principle 
 of humanity as manifesting itself in all the special character- 
 istics and relations of the individual man, and thus giving him 
 the most perfect harmony and unity with himself In other 
 words, such a predicate alone gives to the copula tts full com- 
 plement of meaning, as expressing the unity of thought with the 
 object which it determines, or, looking at it in the other way, the 
 unity of the particular object with its own universal. For the 
 reflective or antithetic movement by which thought determines 
 its object in relation to other objects, is now, as it were, bent back 
 into itself, and reconciled with the immediate reference of the 
 object to itself, which is expressed in judgment of mere thesis 
 
CHAP. V. THE SCHEMATISM 0¥ THE CATEGORIES. 465 
 
 or position (i.e., in Kant's judgments of Quality and Quantity). 
 And, in this reinstated unity of the ohject with itself, is also 
 expressed its unity with the thought for which it is ; for such 
 thought no longer presupposes given objects as points between 
 which it moves in apprehending their relations, but the objec- 
 tive difference of the correlated elements or objects is viewed 
 as itself the expression of an underlying ideal unity. The 
 object is now recognised as that in apprehending which thought 
 is conscious of its own unity, or as an object in which thought 
 can find itself. To put it in Kant's phrase, the transcendental 
 unity of apperception is in such judgments recognised to be 
 constitutive of the object. The abstraction, so natural to the 
 common consciousness and even to science, by which objects 
 are taken as if they existed for themselves without relation to 
 the subject, is in such judgments expressly annulled in the 
 unity of thought with its object. 
 This view of judgment may be made a little clearer by ivhen is the 
 
 matter of 
 
 a shghtly different mode of statement. If we look at judgment j,ae^a™toits 
 
 as the act in which perceptions are determined by conceptions, '^°™' 
 
 it is obvious that we may make any one of the a priori 
 
 conceptions or of their schemata, or any conception that falls 
 
 under them, into the predicate of a judgment. This is done 
 
 by Kant when he lays down in the principles of the pure 
 
 understanding that all objects of experience have extensive 
 
 and intensive quantity; that they are determined in their 
 
 relations to each other in time (and space) by the categories of 
 
 substance and accident, cause and effect, and reciprocity ; and 
 
 that in relation to the consciousness for which they exist, they 
 
 ■ have a certain modality (as hypothetically necessary). But, 
 
 though all such determinations are expressible in judgments, it 
 
 cannot be said that the judgment has a matter adequate to 
 
 its form, except when the predicate expresses an idea, which 
 
 explains at once the distinction and the unity of the elements 
 
 of thought implied in that form ; for only then have we an 
 
 idea foi*which the judgment is the necessary expression. Now, 
 VOL. I. 2 G 
 
466 THE CRITIQUE OP PUEE REASON. book i. 
 
 ideas of quantity and quality do not, as such, explain the division 
 of subject and predicate in the judgment; and ideas of relation 
 do not adequately explain the identity which the judgment 
 asserts between them through the copula. Hence judgments 
 which have not modal predicates, i.e., judgments which do not 
 express the unity of the object with the thought for which it 
 is, express less than they imply. They are judgments which 
 do not explain their own form as judgments ; or they are 
 judgments made by anticipation. Thus when we say that 
 objects of experience are extensive quanta, the reference of 
 quantity as a predicate to, and its distinction from, those 
 objects is not explained by their nature as quanta. Again, 
 when we say that all objects of experience are necessarily 
 determined as substances which reveal themselves in accidents, 
 or accidents which are referred to substances, we show the 
 reason for a reflexion which passes from one side to the other 
 of this dual existence, but not yet a reason for the identity of 
 the two correlatives, each of which refers us to the other. But 
 it is just this identity which is asserted in the judgment. 
 Hence even in asserting such a predicate, the form of judgment 
 goes beyond its matter, or the matter does not contain in 
 itself a reason for this form. From this point of view, there- 
 fore, we can justify Kant's assertion that what judgment 
 expresses is, that the relation of its subject and predicate is 
 determined by the transcendental unity of apperception. For 
 this means that, until we make explicit the relation of the 
 object to the conscious self for which it is, we have not fully 
 or adequately explained what the judgment asserts.^ In other 
 words, so long as by the predicate the subject of the judgment 
 is referred to itself, as in judgments of quality and quantity ; 
 or, so long as it is merely referred to other objects, as in 
 judgments of relation, the judgment does not express all 
 it implies. It expresses all it implies only when by the 
 predicate the subject is referred to an ideal principle, which 
 ^ I do not, of course, say that it meant all this for Kant. 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 467 
 
 is regarded as realising itself in the particular existence of the 
 subject and constituting its individuality ; for then, the deter- 
 mination of the subject by the ideal principle is a determination 
 of that subject by itself. 
 
 Now, it is at once obvious that the distinction which Kant ^ow the 
 
 ' categories oi 
 
 makes between the mathematical and the dynamical categories SeirobfectiTe 
 (those with and those without correlatives) corresponds generally and come to be 
 
 regarded as 
 
 to the distinction between thesis and antithesis, or position and ' ™«"^« ^^^^^^■ 
 reflexion, which has just been given. " For that distinction 
 rests," as he tells us, " on the fact that the dynamical cate- 
 gories have correlates and the mathematical have not ; " or, as 
 he elsewhere puts it, the former express a "synthesis of 
 different elements which are not homogeneous," but which, 
 for that very reason, " require each other," so that the one 
 cannot be posited without the other, while the latter express a 
 "synthesis of different elements which are homogeneous," and 
 which, therefore, " do not require each other."^ It is true that 
 under the dynamical categories he includes the categories of 
 Modality as well as those of Eolation ; while, on the principles 
 just stated, the modal categories are rather to be regarded as 
 expressing the conception of ideal or organic unity, the unity 
 of an object with itself in all its difference, or, what is the 
 same thing, the unity of the object with the thought for which 
 it is. But this directly points to that defect of Kant's theory, 
 of knowledge, which has been already characterised, viz., that 
 he does not admit that the mind can find its own unity in its 
 object, or bring its object into perfect unity with the thought 
 for which it is. Hence, though he holds that objects must be 
 capable of being brought into relation with the unity of self, as 
 otherwise experience could not exist, he equally holds that as 
 given in sense, they are always externally related to that unity. 
 Thus the unity of the self is the source of all the principles of 
 
 This is said of the mathematical and dynamical principles, but for our 
 present purpose it is not necessary to take note of the distinction of principles 
 from oaflgories. (B. 201.) 
 
468 THE CRITIQUE OF PURE REASON. BOOKI. 
 
 that synthesis by which alone objects can be determined as 
 such, but yet the consciousness of self is negatively related to 
 the consciousness of objects. This being the case, it would 
 seem inevitable that knowledge itself, as being the relation of 
 the object to the subject, should be brought under the category 
 of reciprocity. Yet, as being their conscious relation, it neces- 
 sarily involves a higher category ; for to be conscious of the 
 self and the object in distinction from and relation to each 
 other, — and only so can we be conscious of either, — ^is to see 
 them from the point of view of a unity which transcends their 
 distinction. This difficulty is partly disguised when the recip- 
 rocity in question is taken as a reciprocity prior to knowledge, 
 a reciprocity between the subject in itself and the object in 
 itself — the latter being the ground of the affections that furnish 
 the matter for the activity of the former ; but it necessarily 
 reappears as a reciprocity in knowledge between perception 
 and conception, consciousness and self-consciousness, unless the 
 latter of each of these pairs be taken to include the former as 
 an element in itself. Now Kant, as we have seen, goes so far 
 in this direction that he makes thought or self-consciousness 
 reach beyond the consciousness of objects, and exercise a 
 spontaneity which is not a mere reaction upon the affections 
 that it receives from without. But he insists that in this 
 unlimited spontaneity, just because it meets with no obstacle 
 which reacts upon it, thought determines no object. It is a 
 free movement, but a free movement in vacuo, where there 
 is no atmosphere to sustain the wings of thought, and where 
 therefore no progress is made.^ The result, therefore, is merely 
 the production of ideas of objects, which have no objective 
 value; ideas which are ideas and nothing more. This opposi- 
 tion of thought to knowledge, as transcending it, but not 
 transcending it to any purpose, or only to the purpose of 
 setting up an ideal of knowledge with which our actual know- 
 ledge is incommensurable, makes it difficult for Kant to deal 
 
 lA. 4; B. 8. 
 
CHAP. V. THE SCHEMATISM OF THE CATEGORIES. 469 
 
 with the categories of modality which express the relation of 
 the object, as known to the mind. And we can only express 
 the result he arrives at by saying that that relation is for Kant 
 more than reciprocity, yet less than organic unity : more than 
 reciprocity or the relation of independent things, because it is 
 a conscious relation ; i.e., a relation of factors which are bbth in 
 consciousness ; and less than organic unity, because the factors 
 in consciousness are supposed to be incapable of being reduced 
 to identity. Kant's classification of the modal categories as 
 one species of dynamical categories, is, therefore, just one 
 indication of his tendency to regard the relation of the mind 
 to its object under the same point of view in which he regards 
 the relation of one object to another ; though, of course, the 
 very idea of his Transcendental Deduction involves that the 
 relations of the ego to the object can Twt be brought under the 
 categories by which it determines the relation of objects to 
 each other. 
 If, however, we view the consciousness of self as involving seasons for 
 
 regarding 
 
 not an analytic but a synthetic unity, and as, indeed, the very t''?'" ^^ *'^''. 
 
 */ .J t/ ' ' ' J only adequate 
 
 type of such a unity, we shall be led to correct Kant's view of reaiwy"°^ °' 
 the categories as derived from it ; we shall be led to recognise, 
 not only that, without any schematism or reflexion of the unity 
 of thought upon the manifold, we can derive from it categories 
 both of position and reflexion, of thesis and antithesis, but also 
 that these must be recognised as imperfect categories, i.e., as 
 categories which imply a higher unity than they express. 
 And this cannot mean anything else than that which is shown 
 by Hegel, viz., that the idea of organic or ideal unity neces- 
 sarily arises, whenever we make clear what is involved in the 
 unity of thought with itself in any determination of its objects. 
 Every other category is an imperfect representation of this 
 category, or a stage in its development ; and to take any other 
 category as a final explanation, or a.g the principle of a final 
 explanation, of anything is to fossilise thought at one stage of 
 its nibvement. Every other category implies something it 
 
470 THE CRITIQUE OF PURE REASON. book l 
 
 does not explain, and therefore must break down in contra- 
 diction, if it is taken as absolute, i.e., as the ultimate truth ol 
 things. For the intelligence cannot rest in its progressive 
 definition or determination of any object, till it finds its owi 
 unity in that object. 
 The schema- When WB reject the idea that the judgment of self-con- 
 
 tism may be 
 
 towSthe^"^ sciousness is analytic, Kant's schematism of the categories is 
 thfcategories. no longer needed to give them synthetic value. Its value now 
 is that it points to the conditions of the application of the 
 categories as principles of knowledge. For, as I have already 
 said, the unity-in-difference of pure self-consciousness in itseli 
 is different from the unity-in-difference of self-consciousness 
 and the consciousness of the world of objects in space anc 
 time. The categories may, therefore, be regarded as the pre- 
 dicates in the primary judgments of knowledge or experience 
 by which objects are determined in relation to the self, and sc 
 as the judgments that are implied in all other judgments 
 But this means that the determination of these objects by th( 
 categories is presupposed in all other determination of them 
 though it may not need to be explicitly recognised in sucl 
 determination. Hence, Kant has no sooner got his categoric! 
 schematised, and so made available for the determination o: 
 objects, than he proceeds to consider the judgments of know 
 ledge from the other side, and to show that the objects of ou: 
 sensible perceptions as known in space and time must neces 
 sarily be subsumed under these categories, in order to bi 
 determined as objects. Or, to put it in Kant's language, thi 
 synthesis of imagination, by which images of perception ar^ 
 formed, must necessarily be conformable to the whole system c 
 the categories, and must be recognised as so conformable, i 
 objects are to be known ag such. The discussion of this thesi 
 will form the subject of the following chapters. 
 
471 
 
 CHAPTER VI. 
 
 THE SYSTEM OF THE PRINCIPLES OF PURE UNDERSTANDING. 
 
 rpHE principles of the pure understanding are the universal Beiuction of 
 judgments in relation to objects of experience which it of Pureuider- 
 
 standing. 
 
 becomes possible to make, if we are authorised to apply the 
 schematised categories to the data of perception. Here, there- 
 fore, Kant starts from the point of view of perception, as in 
 dealing with the schemata he started from the point of view 
 of conception. There he had to consider how the categories, 
 as conceptions of objects in general, could be translated into 
 conceptions of objects in time. Here he has to consider how 
 it is possible actually to apply these conceptions to perceptions. 
 And the general answer given is, that it is possible because it 
 is necessary, i.e., because it is only through the application of Ij 
 these conceptions that perceptions can give rise .to experience,/' 
 in the sense of a knowledge of objects. 
 
 As usual, Kant begins by going back to Logic, and again ^*"J'^''™- 
 pointing the antithesis between thinking and knowing. Think- * thSkilig'^^ 
 ing, as a purely analytic process of judging, is tied down to the principles of 
 
 1 n • ^ .1 . knowing. 
 
 law of identity ; and its highest principle is that no conception 
 taken as a subject can have a predicate which is its negation, 
 or which is the negative of any element contained in it. This 
 is a principle which has nothing to do with the matter of 
 thought, or with the way in which that matter has come into 
 our ffiinds. It is therefore no positive, but only a negative. 
 
472 THE CRITIQUE OF PUEE BBASON. book i. 
 
 criterion of truth ; for what is self-consistent is not, therefore, 
 true, though what is true must be self-consistent. Kant 
 points out further that in the statement of this principle, it is 
 not right to introduce a reference to time. We ought not to 
 say that according to the principle of contradiction no man can 
 be learned and unlearned at the same time ; for this would 
 imply a consciousness of the possibility of contradictory predi- 
 cates succeeding each other in one subject, a consciousness 
 which cannot be derived from pure thought without reference 
 to the conditions under which objects are given in time. By 
 the law of pure thought, we can exclude from the subject only 
 the negative of any predicate included in the conception of that 
 subject, as when we say " no learned man is unlearned." So 
 much as to the rule of thinking, but it is quite different as to 
 the rule of knowing ; for we know that all objects of perception 
 as such come under conditions of time, and therefore we can 
 apply to them all the principles which arise out of the applica- 
 tion of the categories to time, or to matter as given under time- 
 conditions, (and also, if they are outward objects, under 
 space -conditions). In so far, therefore, as the intelligence in 
 relation to time, as the form of inner sense, gives rise to cate- 
 gories which correspond to time-relations, i.e., which are 
 capable of being translated into terms of relations of or in 
 time, in so far we can lay down general principles under 
 which all objects empirically known must come ; seeing that, 
 if our perceptions do not come under these principles, they 
 will not " furnish us with knowledge, but only with a rhapsody 
 of impressions which would not fit into a context . determiued 
 by the laws wliich a connected consciousness involves, and 
 would not therefore be in harmony with the transcendental and 
 necessary unity of apperception." It is thus "the highest 
 principle of all synthetic judgments that every object stands 
 under the necessary conditions of the synthetic unity of the 
 manifold of perception in a possible experience." In other 
 words, we can lay down general principles which carry us 
 
CHAP. VI. THE PRINCIPLES OP PURE UNDERSTANDING. 473 
 
 beyond particular perceptions, just in so far as such general 
 principles express the conditions under which objects are 
 known through these perceptions. " In this way a priori 
 synthetic judgments are possible, if we bring the formal condi- 
 tions of perception a ■priori, the synthesis of imagination and 
 the necessary unity of such synthesis in a transcendental 
 apperception, into relation with the possibility of empirical 
 knowledge. For we are entitled to'say that the conditions of 
 the possibility of experience are conditions also of the possibility 
 of the objects of experience, and that the synthetic a priori 
 judgments which express those conditions have objective 
 vahdity."! 
 
 Of the distinction here made between analysis and synthesis Relation of the 
 
 Schematism to ' 
 
 enough has been already said. It is Kant's great error that, the i)edu(rtion 
 failing to apprehend the essential relativity of thought, as undlrsLnd-^ 
 imphed in the categories of Quantity and Quality and ex- '°^' 
 pressed in the categories of Eolation, he sharply separates 
 analysis from synthesis, thought that abides with its object 
 from thought that goes beyond it. Hence relativity is sup- 
 posed to be due only to the reflexion of thought upon the forms 
 or matter of sense; and this finally involves the consequence 
 that the distinction of the relative elements can never be over- 
 come. Holding to this view, thought is conceived as by its 
 unity (inasmuch as that unity takes the form of judgment) 
 supplying forms of synthesis for possible perceptions — which, 
 however, apart from actual perceptions, have no objective 
 vahdity. In this way the necessary relation of the pure 
 consciousness of self to the consciousness of the world, takes 
 the appearance of the relation of an empty unity, which has, 
 however, a variety of forms, to a blind or unintelligible matter, 
 which, however, is determined by the forms of space and time. 
 At this point we have to bring in Kant's view of imagination, 
 as a blind synthetic activity which is implied in perception, and 
 in virtue of which perception is in pre-established harmony 
 ♦ ' A. 154, 158 ; B. 193, 197. 
 
474 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 with the conscious activity of conception. But the reason 
 why these two activities are in pre-established harmony is, as 
 we have seen, that it is the same activity which is implied in 
 both, though in the ohe ease it acts blindly and in the other 
 consciously. Hence the relative difference of perception and 
 conception, and at the same time their necessary reference to 
 each other ; and hence also the possibility of escaping the 
 dilemma in which we are involved when we represent the 
 principles of the understanding either as already present in 
 perception, (in which case they would be useless) or as not so 
 present (in which case they would be impossible of application). 
 On this enough has been already said. Here we have only to 
 recall the imperfection of Kant's statement of the organic unity 
 of consciousness with self-consciousness, an imperfection which 
 arises from his treating both perception and conception as if 
 they could be determined in themselves, apart from the 
 distinction and relation into which they are brought in the 
 judgment. As so taken, they necessarily tend to lose their 
 meaning, and to shrink into mere abstractions of unity and 
 difference ; and this leads to an attempt to mediate externally 
 between the two terms. In this mediation, on the one hand, 
 it is shown how the categories get schematised in relation to 
 time, in order to be applied to objects ; and, on the other hand, 
 it is shown that the manifold of perception must be combined 
 into images, which are capable of being brought under the 
 principles of the pure understanding and so determined in 
 relation to objects. The first of the two points we have 
 already discussed, and we have seen that the Schematism 
 conceals, under the appearance of an external determination of 
 the categories in relation to time, the truth that pure thought 
 loses its meaning unless it be taken as a movement of deter- 
 mination — a movement of differentiation and integration, 
 culminating in the transparent unity in difference of self- 
 consciousness, which therefore contains in itself a whole system 
 of categories for the determination of objects. Here we have 
 
CHAP. VI. THE PRINCIPLES OP PURE UNDERSTANDING. 475 
 
 to deal with the other aspect of the question, that the manifold 
 of perception must be brought together in a synthetic unity 
 which is conformable to the system of categories ; since other- 
 wise it cannot be made to yield a consciousness of objects 
 which can be united with the consciousness of self 
 
 Now, in proving this thesis, Kant speaks in the iirst instance Nature of 
 
 ' r o X- Kant's Deduo- 
 
 as if perceptions of objects were presented to us altogether "^^^^^^^^^j 
 apart from that conscious process in which they are determined ^Mch mcod- 
 by the categories. But the effect of his argument is to show ownpremiaes. 
 that they are not presented as perceptions of objects except 
 through this very determination. The essential difficulty of 
 following his argument is, however, just this ; that the former 
 mode of speech, as it necessarily elevates that element which it 
 isolates into a res completa, seems to give to the perception by 
 itself all that afterwards can come to it from the conception, 
 and to make the subsumption of the former under the latter 
 unnecessary. This characteristic of the proof according to 
 which it, so to speak, transforms and reconstitutes its own 
 premises, is often referred to by Kant, though perhaps he does 
 not keep it so steadily before him as might be desired. 
 
 The following passage is one of those in which this " turning 
 movement " of Kant's dialectic is most clearly expressed. 
 
 " Eeason, taken as a whole, in its merely speculative use, 
 does not contain a single directly synthetic judgment based 
 upon conceptions. For, from Ideas of reason, as we have 
 shown " {i.e., in the Dialectic), " we cannot derive any synthetic 
 judgments which have objective validity : and though it is 
 different with the conceptions of the understanding, by aid of 
 which we can establish secure principles for the determination 
 of objects, yet we cannot derive these principles from the con- 
 ceptions directly, but only indirectly through the reference of 
 these conceptions to something which is quite contingent, 
 namely, posBihle experience. If it be presupposed that some- 
 thing is ■ given as object of possible experience, then these 
 princijiles can be seen to be apodictically certain ; but in them- 
 
476 THE CRITIQUE OF PUKE REASON. book I. 
 
 selves by a direct a •priori process of thought, they cannot be 
 known at all. Thus, it is impossible to prove the proposition, 
 that everything that happens has its cause, directly out of these 
 given conceptions. Hence that principle is no dogma, though 
 in another point of view, viz., when we consider it in reference 
 to the one sphere in which it can possibly be applied, i.e., 
 experience, it can be shown to be apodictically certain. If, 
 therefore, we call it a principle [Grundsatz] and not a dogma 
 [Lehrsatz], and that in spite of the fact that it requires proof, 
 this title finds its justification in the strange pecuharity 
 attaching to it, viz., that it makes possible the very experience 
 which furnishes the basis for its own proof, and that in such 
 experience it must always be presupposed."^ 
 S'pTrtoce be The fuU explanation of this method must be postponed till 
 deduction of wc reach the Analogies of Experience, for it is in special refer- 
 
 the principles 
 
 of "?.^^? ence to Hume's criticism of the principle of causahty that 
 Ambiguity m j^^^nt generally, as in the above passage, explains his own 
 argumen . procedure. Here, however, it is necessary to point out the 
 main ambiguity which is apt to trouble us throughout. How, 
 we are apt to say, can experience prove the principle of its own 
 possibility ? Does this mean that, if we analyse our particular 
 experience, we find certain principles contained in it ? Is it 
 possible that particular experiences should yield principles 
 more general than the experiences from which they were 
 derived ? Kant answers, no ; our particular experiences cannot 
 directly yield universal principles. They can yield such prin- 
 ciples only indirectly, when we regard them " in a certain 
 point of view," i.e., when we consider how it is possible for us 
 to have such particular experiences. So to consider them, is 
 to bring them into relation to the conscious self which has 
 them, and which can have them o^nly if they are capable of 
 being combined with the consciousness of self. This combina- 
 tion, however, is possible only as they are determined by the 
 principles of the pure understanding. It appears, therefore, 
 1 A. 737 ; B. 765. 
 
CHAP. VI. THE PRINCIPLES OF PURE UNDERSTANDING. 477 
 
 that, when he says that our particular experiences do not- 
 themselves imply the universal principles in question, Kant is 
 speaking of those particular experiences not as they really are, 
 but as they would be if they were not determined by the prin- 
 ciples of the pure understanding. Eeally, as the experience of 
 a conscious self, the perception of the particular is an experi- 
 ence which implies the determination of the particular by the 
 universal; though it does not necessarily imply that the 
 universal as such, in distinction from the particular, has been 
 made the object of thought. It is, indeed, a confusion of these 
 two things, viz., a confusion of the presence of the universal 
 principle in our determination of the particulars, with the explicit 
 consciousness of the universal principle as such, that most often 
 leads to the denial of the former. But there is no necessary 
 connexion between the two ; for, just as, in our ordinary 
 consciousness of objects, we overlook their relation to the 
 conscious self, so also we overlook the determination of the 
 objects by universal principles which is implied in that relation. 
 Hence, our first view of objects is abstract and imperfect ; 
 because it leaves out of account an important factor, which in 
 reality is always present in our determination of these objects. 
 This, however, does not mean that we really can empty our 
 objective consciousness of all relation to the self, and of the 
 principles involved in such relation, but merely that we can, 
 and do, omit to attend to them. But it is not less true that 
 thought has been at work in the constitution of an intelligible 
 experience, although we may ignore its activity. In this point ■ 
 of view, Kant's deduction of the principles of pure under- 
 standing as conditions of possible experience means simply that 
 he calls our attention to the elements presupposed in such an 
 experience. In doing so,-*owever, he shows that our previous 
 experience was not what we had supposed it to be, a conscious- 
 ness of the particular as given in sense without any determina- 
 tion by the universal. He shows, in fact, that an experience • 
 which» takes such a view of itself, is imperfectly self-conscious, 
 
478 THE CRITIQUE OF PURE KBASON. book I. 
 
 and that it could not have been even that consciousness of the 
 particular which it knows itself to be, if it had not been more. 
 But this implies that, if we reduce ordinary experience to the 
 level of its own view of itself, we must take away from it much 
 that it seems to have. If it supposes itself to be a bare con- 
 sciousness of the particular apart from the universal, it has to 
 be pointed out that the bare particular as such is no object at 
 all, but is merely an unrelated something of which nothing can 
 be said. If it supposes itself to be a mere consciousness of the 
 data of sense, it has to be pointed out that the data of sense 
 are sensations and not perceptions. In this way what, 
 regarded from one side, is the process of clothing or investing 
 the sensible particular with determinations which are not seen 
 to belong to it till it is viewed as an object for a conscious 
 subject, regarded from the other side is the unclothing or 
 divesting the same sensible particular of determinations which 
 it has usurped, but to which it had no title, i.e., no title 
 in its own right or except as being such an object. Now, it is 
 this last aspect of the process which Kant seems most often to 
 leave out of view, or at least it is the aspect to which he is less 
 careful of giving prominence. And it is his omission so to 
 insist upon it that seems to deprive his argument of due weight 
 and to expose him to the dilemma above mentioned. For the 
 moment we regard the perception as having a character of its 
 own, apart from any determination by the conceptions under 
 which it is to be subsumed, we seem to make that subsumption 
 either useless or impossible. 
 This ambigu- Wc may bcst cxplaiu the defect of Kant's statement by 
 
 '*■" '8 partly 
 
 ^^jj'y*^. saying that he does not always keep separate and distinct the 
 
 ity is partly 
 caused by 1 " 
 confusion c 
 
 prLdpiIs^r^ two things which we have just been distinguishing ; the use of 
 
 ordinary ex- , . ...... . , ^, 
 
 perionce with Certain a prion prmciples in common experience, and the con- 
 
 the conscious 
 
 use 
 
 science. 
 
 of them in scious use of them in science, which presupposes that they have 
 been separated by reflexion from the particulars to which they 
 have been applied, and that they have thus been made objects 
 of special attention. Yet the ground for the distinction is 
 
CHAP. VI. THE PRINCIPLES OF PURE UNDERSTANDING. 479 
 
 clearly stated by Kant himself. According to the principles of 
 the Transcendental Deduction of the Categories, the pure con- 
 ceptions may be regarded as supplying the primary predicates 
 in the judgments of experience, predicates which are pre- 
 supposed in all further determination of its objects. But this, 
 of course, does not mean that in our first consciousness of 
 objects we clearly recognise them as qualified and quantified 
 substances, which in all their changing states are determined 
 by laws of causality and reciprocity. Kant's principles of the 
 pure understanding are not present to the ordinary empirical 
 consciousness, any more than the principles of grammar are 
 present to everyone who can give expression to his ideas in 
 language'. The kind of consciousness to which such principles 
 are present in their abstract form, and in which they are 
 deliberately used as guides in the scientific investigation of 
 phenomena, is a result of reflexion. When, therefore, Kant 
 says that " our intelligence is itself a system, but in its pure 
 use in which it operates with mere conceptions, it is only a 
 system of inquiry according to principles of unity for which 
 experience must furnish all the matter," ^ he is speaking of the 
 pure conceptions as supplying guiding principles for the 
 method of science. And he is contending simply that the con- 
 sciousness that separates the categories from experience, will 
 find in them and in the conceptions based on them, not means 
 for the discovery of another world of things in themselves, but 
 only principles by which the experience from which they have 
 been abstracted may be tested, corrected, and raised into the 
 form of science. Science, in fact, differs from the ordinary 
 consciousness just in this, that it uses the principles pre- 
 supposed in experience to transform and reconstitute experience. 
 The ordinary consciousness involves these principles, and uses 
 them constantly to determine and relate its objects, but without 
 ever bringing them distinctly before it as objects of thought, or 
 even becoming aware of their existence as general principles. 
 * ' A. 738 ; B. 766. 
 
480 THE CRITIQUE OF PURE REASON. BOOK i. 
 
 Locke could easily demonstrate that the child knows nothing 
 of a principle of causality, though it constantly explains 
 particular changes by reference to preceding conditions, and 
 even begins to reject its first explanations as inadequate and to 
 look for others. "When, therefore, Kant treats the categories 
 as the primary predicates in the judgments of experience, lie 
 does not mean that we actually use such abstract categories as 
 predicates in our first judgments ; any more than in saying that 
 all experience of objects implies the conscious identity of the 
 self for which they are objects, he means that in our ordinary 
 experience we think of that identity as the principle which 
 determines objects as such. It is a " transcendental reflexion " 
 which teaches us that apart from the conscious unity of the self, 
 there could be no consciousness of a world of objects ; and it is 
 a " transcendental reflexion " equally which teaches us that every 
 judgment of experience is, in Kant's language, determined in 
 relation to all the " functions of unity " which we recognise 
 as the categories. Though holding by the a priori, Kant 
 altogether rejects the theory of innate ideas ; i nor does he ever 
 suppose that in our first consciousness we have the universal 
 brought before us as a special object of attention, but only that 
 it is presupposed in the consciousness of the particular. Thus, 
 we do not begin with a conception of space in its unity and 
 infinity and then localise objects in it, but in localising objects 
 with reference to each other we presuppose space as one and 
 unlimited. In like manner we do not first predicate of our 
 perceptions those conceptions in virtue of which they become 
 for us perceptions of definite intelligible objects, but we pre- 
 suppose these predicates ^in all our particular judgments about 
 
 ^ See especially E. I. 444 ; H. VI. 37 ; where, in answer to Eberhard, Kant 
 contends that all our ideas are acquired, but that there ia, in the language of 
 law, an acquidtio oritjinaria in the case of the ideas of space and time and the 
 pure conceptions of the understanding : in so far as " the formal ground of the 
 possibility " of these ideas lies in our sensibility and our understanding. But 
 this implies that these ideas must be used in experience before they are 
 reflected on, or made the special objects of attention, and therefore before they 
 are stated in their generality. 
 
CHAP. VI. THE PEINCIPLES OF PUEE UNDERSTANDING. 481 
 
 objects. Hence, the reflective consciousness, which goes back 
 upon the primary principles for the determination of objects and 
 in them finds the means of correcting our first judgments, is a 
 thing quite different from that consciousness of these principles 
 which is involved in the fact that objects exist for us as such; yet 
 the former consciousness is only the recognition of what is pre- 
 supposed in the latter ; and Kant could fairly say that we find 
 the principle of causality in our experience only because we 
 have put it there in the primary exercise of our intelligence.^ 
 We must, therefore, recognise that, when Kant is showing the 
 necessity of the determination of perceptions by conceptions 
 with a view to experience, i.e., to that consciousness of objects 
 which relates them to the conscious self, he is not directly 
 justifying the man of science in the use of principles of the 
 understanding; he is justifying an application of these prin- 
 ciples which is already implied in the empirical matter with 
 which the man of science deals. Hence, the problem of experi- 
 ence will need a different solution, according as we regard it 
 as referring to the former or to the latter application of the 
 categories. If it be asked what authorises the" man of science 
 to apply his principles; what is the suggestion or guiding thread 
 which he can find in experience for such application, Kant's 
 answer must be that the matter of experience, as a conscious- 
 ness of objects, already involves these principles, and that in 
 using them to put the question to experience, we are only 
 testing experience by its own fundamental constitutive ideas. 
 But if it be asked, what then authorises that first unreflective 
 application of conceptions to perceptions, which is involved in 
 all experience, his answer is to be found in the doctrine of the 
 pre-established harmony of perception and conception, and 
 ultimately in the doctrine of the identity of the principles of 
 unity in both. In fact, as we have seen, the ideal priority of 
 perception to conception is not to be understood as if the percep- 
 tion were present, in the character which it has for the conscious 
 
 1 A. 196 ; B. 241. 
 VOL. I. ■ 2 H 
 
482 THE CRITIQUE OP PURE REASON. book 
 
 self who determines it by conceptions, before the process 
 such determination begins. A datum, of sense can be a pt 
 ception, the consciousness of an object, only for the self th 
 thus determines it ; and though the consciousness of tl 
 determination " may be often very weak, so that we recogni 
 it only in the result and not in the activity itself," ^ yet witho 
 it a consciousness of objects is impossible. Hence, if we tra 
 back experience to its earliest form, we shall still find it to be 
 consciousness of an objective world which, as such consciou 
 ness, is determined by universal principles, principles whi( 
 reflexion detects to be the categories. But our first synthes 
 by means of the categories, by no means implies such reflexio 
 though the imperfection of the result of our first synthes 
 gradually awakens it. In this way, the scientific consciousne 
 arises out of the ordinary empirical consciousness, when tl 
 mistakes and contradictions of the latter force it to go ba( 
 upon the principles of its constitution, and so, by definition 
 these principles, to strike into what Kant calls " the secu 
 path of science." Still, though the correction of ordinary e 
 perience by science with its clear consciousness of its ov 
 principles, is only a correction of experience by itself, we mu 
 be careful to distinguish the sense in which it is said th 
 Ordinary experience rests upon these principles, from the sen 
 in which scientific knowledge rests upon them. 
 imguage Now, does Kant always keep this distinction in viev\ 
 
 betrays this That hc saw it, is obvious enough from passages that ha' 
 
 ambiguity. 
 
 already been quoted ; but there are other passages in which 1 
 seems to lose sight of it. Thus, he recognises that an id' 
 which cannot be united with the " I think " is for us as go( 
 as nothing, and he demands conformity to the unity of sa 
 consciousness, and so to the categories, as the condition witho 
 which nothing can enter into the sphere of intelligence ; but 1 
 does not seem always to recognise that perception, as havii 
 entered that sphere, is already determined by conception. I 
 
 lA. 101 
 
CHAP. VI. THE PRINCIPLES OF PURE UNDERSTANDING. 483 
 
 seems to draw a line of distinction between what is capable of 
 being united to the "I think," and what is so united; as if the 
 former, though an actual perception, were still in the condition 
 in which it would be for a subject that did not think or bring 
 its perceptions under conceptions. And this error leads 
 almost necessarily to another, viz., the error of supposing that 
 the determination of perception by conception comes into play 
 only when the conception, or the principle based upon it, is 
 consciously used, as it is used by the scientific man, as a 
 guiding principle for discovery. It is thus that we may 
 explain how in the Prolegomena Kant could speak of judg- 
 ments of perception, as if perception could apprehend particular 
 facts, or make judgments in relation to particular objects, 
 altogether apart from the determination of sense by concep- 
 tions ; and as if the latter came in only when the judgment 
 was " viewed as determined " by one of the categories, i.e., 
 when the a priori principle involved in the judgment was 
 reflected on. If this is not to be explained by the popular 
 character of the Prolegomena, and by its special regressive 
 method,^ it involves a confusion between that conscious use of 
 the principles of pure understanding, which is necessary if 
 ordinary experience is to be raised to the form of science, and 
 that use of them which is necessary to thfe constitution of 
 ordinary experience itself, i.e., to any consciousness of objects 
 as such. 
 If we set aside this confusion, it becomes evident that science 
 
 already in- 
 
 seience is distinguished from ordinary experience just by the ™na|„tai™" 
 fact that it has a definite method ; and this means that it has Sweve^is"^' 
 made a " transcendental regress " upon the principles that are fav back as 
 
 .... Kant'sregreas. 
 
 presupposed in ordinary experience, a regress which is similar 
 in kind to that which Kant himself makes in the Critique, 
 though not carried so far. For the process of reflexion, which 
 enables us to recognise that the principles of the understand- 
 
 ' Which starts with the principles of science and works backward to their 
 conditions. Of., however, what is said above, p. 376 seq. 
 
484 THE CRITIQUE OF PCTRE REASON. BOOl 
 
 ing are presupposed in our ordinary consciousness of objec 
 and events, only needs to be carried a little further to rev( 
 that the unity of the self is presupposed in all our conscioi 
 ness of the world through these principles. And if it is 
 legitimate inference from the former regress that these pri 
 ciples may be used to remould our ordinary experience ai 
 elevate it to the form of science, it is equally a legitima 
 inference from the latter that the idea of the unity of the se 
 as underlying the consciousness of an objective world, may 
 used to re-interpret that consciousness even in its scienti 
 form, and raise it to the form of philosophy. And this, 
 fact, is what Kant does when he shows that the objects 
 experience are phenomena or existences for a self; though ] 
 finally destroys the value of this demonstration : first, by mai 
 taining the old dualism in relation to the subject in itself ai 
 the thing in itself, and, secondly, by refusing to complete tl 
 philosophical transformation of science which' is the necessa: 
 consequence of the new view of its principle of unity, 
 fusionof the The ultimate explanation of the defect of Kant's stateme: 
 
 transcenden- ... 
 
 tai regress is one With which wc are already familiar, viz., his suppositit 
 
 with a psycho- •' -^ ^ 
 
 of^he'lenesrs' ^^^^ *^^ rcgress by which the fundamental principles of knoi 
 outo?in'de-°° ledge are discovered, takes us back to subjective data given 
 
 pendent . i» i i . . p i 
 
 iactors. scuse, and to a subjective process of the determination of the 
 
 data. Now, as we have seen, no such regress upon the elemen 
 out of which objective experience was constituted, is possih 
 If we attempt to go back to sensation without thought, ^ 
 extinguish perception as well as conception, i.e., we abstra 
 from the matter determined in an intelligible experience 
 well as from the form that determines it ; for the unity of t 
 sensitive life cannot be said to contain the one any more thi 
 the other. On the other hand, if the transcendental regress 
 a process in which by reflexion we become conscious of t 
 principles involved in an intelligible experience, we are not 
 it going back upon independent factors out of which expe: 
 ence was constructed, so as to catch these factors in a simp] 
 
CHAP. VI. THE PRINCIPLES OP PURE UNDERSTANDING. 485 
 
 form than that in which they appear in experience. "We are 
 simply showing that that experience could not be what it 
 seems to itself unless it were more, and are thus bringing it to 
 recognise factors in itself which it did not formerly recognise. 
 But Kant seems to fluctuate between these two conceptions of 
 the transcendental regress; and while, according to the last 
 mentioned view of it, he recognises that our perceptions are 
 what they are for us through their determination by concep- 
 tions, he seems to himself, according to the first-mentioned 
 view of it, to be discussing the process by which perceptions, 
 as mere data of sense, are for the first time brought under 
 determination by conceptions. But, as mere data of sense are 
 not even elements in an intelligible experience, Kant is driven 
 to attribute to them, as the matter to which the pure concep- 
 tions have to be applied, a determination which they can have 
 only in a consciousness determined by conceptions, and thus to 
 stultify his own arguments. For when they are thus con- 
 ceived, they have already the determination which he would 
 prove them to need ; and by making this determination explicit, 
 what he does is to show, not how experience is possible, but only 
 how experience is to be raised into the form of science. If, 
 however, we set aside this misunderstanding and take the 
 transcendental deduction in its true sense, its whole purport is 
 to show that common experience is more than it knows; and 
 that, therefore, if we take its own view of itself as true, we 
 must strip it of much that it claims, and must, in fact, reduce 
 it to a mere flux of sensation out of which no intelligible con- 
 sciousness could be made. If we allow it what it " hath," we 
 must give it niore ; if we admit that it " hath not " more than 
 it supposes, we must take from it even that which it " seemeth 
 
 to have." ' n we take 
 
 away the 
 
 It appears, then, that the system of pure conceptions or ?^„*ex^eri^. 
 schematised conceptions which Kant presents to us, is simply a o?srasat°oM? 
 
 . .„ . if we reflect on 
 
 systematic account of the principles of the scientific conscious- them, we raise 
 
 *• ^ experience to 
 
 ness as he found it exemplified especially in the works of ISTewton; *^? *°™ °* 
 
486 THE CRITIQUE OF PTJ.RB REASON. book i. 
 
 and Kant's proof of these principles is that when we regard 
 ordinary experience from the point of view of the self for 
 which it is, we cannot explain it except by aid of the whole 
 system of these principles, On the one hand, therefore, if we 
 remove the principles, and take ordinary experience as what it 
 is conscious of being, it falls together lilce a house of cards. 
 And, on the other hand, if we admit the principles, ordinary ex- 
 perience must be remoulded by them, and the world of experience 
 will thus become for us what it is to the scientific conscious- 
 ness ; or, in other words, all particular experiences wiU. be 
 regarded as elements which have to find their place in relation 
 to other experiences according to the scientific idea of the 
 order of nature. 
 Themathema- The elements of the scientific idea of nature are expressed 
 
 tical and -*■ 
 
 prficipies^ in the mathematical and dynamical principles. The former 
 making a ° includc the principles of the Axioms of perception and of the 
 
 further dis- . . , , . . , ^ . ^ . 
 
 tinction Anticipations of sense-perception ; the latter include the prin- 
 
 D6tiW66n tllG 
 
 The Anaiories' ciplcs of the Analogies of experience and the Postulates of 
 empirical thinking^ If, however, we examine closely Kant's 
 explanations of this classification, we find it admits of simpli- 
 fication. For the Postulates of empirical thought are said 
 not to contain any determination of the objects of experience 
 as such, but only of their relation to our faculties of knowledge. 
 But this relation does not form part of the scientific view of 
 objects at all. Science, like the ordinary consciousness, still 
 abstracts from the relation of objects as such to the mind, and 
 it is only the transcendental regress itself which brings that 
 relation into view. Whether the consideration of this relation 
 does, or, as Kant maintains, does not, bring any new qualifica- 
 tions of the object, we need not at present consider. At any 
 rate it is a qualification With which it is the business of philo- 
 sophy, and not of science, to deal. When, therefore, Kant says 
 that the mathematical principles exhibit " the conditions of 
 
 ^ I translate Anschaung hj " perception" and Wahmehmung by " sense-per- 
 ception," wherever Inhere is any danger of confusion. 
 
CHAP. VI. THE PRINCIPLES OF PURE UNDERSTANDING. 487 
 
 perception in view of possible experience," while the dynamical 
 principles exhibit the "conditions of the existence of the 
 objects of a possible empirical perception," he does, not express 
 his thought precisely. For the question of the existence of 
 objects in the sense of their actuality as distinguished from, yet 
 related to, thought, is a question of modality; and it arises only 
 when we ask whether the objects, which we have hitherto 
 treated as things in themselves, have, or have not, an existence 
 independent of, or in any way distinguishable from, our con- 
 sciousness of them ; and how they can have such independence; 
 or be thus distinguished, consistently with their existence being 
 an existence for a thinking self?. In other words, Kant has 
 under this head to consider the possibility of the existence, cts 
 objects, of phenomena which, as phenomena, are essentially 
 objects for us. As, however, the answer to this question is, 
 that we can determine objects as existing only so far as 
 we combine their phenomena, with each other in one general 
 context of experience by means of the Analogies, the 
 Analogies also are regarded as principles which condition 
 the existence of objects of experience. Immediately, however, 
 the function of the Analogies is only to determine phenomena 
 as essentially related to each other in one context of experience, 
 so that one of them cannot be posited without the others ; 
 while the mathematical principles merely determine the quanti- 
 tative conditions under which one must be added to another, 
 */ they are so added. As regards each of these determinations, 
 we can ask the transcendental question how the object comes to 
 be so determined /or us, and in each case the answer must be that 
 it is so determined only because the synthetic process of deter- 
 mination by which the object comes to exist for us, is controlled 
 by certain principles, and that the recognition that it is so 
 controlled is involved even in our simplest consciousness of the 
 object. 
 
 17-i -1 -1 --I i-i-i- The relation of 
 
 Jlant treats the three classes of principles which m their the ti>rea 
 
 ■*■ classes of pnn- 
 
 combination make up the 'scientific idea of nature almost as if dS^SLd. 
 
488 THE CRITIQUE OF PURE REASON. book I. 
 
 they were co-ordinate elements of that idea. In so doing he 
 is sufficiently in accord with the idea of science as it was 
 exemplified in the work of Newton and his successors. It 
 has, however, been pointed out in the first chapter of the 
 Introduction, that the transcendental regress in which science 
 begins did not, in the first instance, bring into view the 
 dynamical, but only the mathematical synthesis ; and that 
 there was a time in the history of science when the principles 
 of such synthesis were regarded as in themselves a sufficient 
 guide to scientific investigation. As, in our first consciousness 
 of the world, we direct our attention mainly to the external 
 relations of things as outside of each other in space, and coexistent 
 or successive in time, and pay little or no regard to any deeper 
 reason for such relations or any causal or reciprocal determina- 
 tion which may be implied in them, so the earliest science 
 brings to light only the mathematical or quantitative synthesis 
 by which things are added to each other, and does not reflect 
 on any necessity of their connexion. The transition by which 
 the idea of physical causation has in modern times become the 
 guiding principle of science, was, therefore, an important step 
 in advance, involving, as it did, a new consciousness of the 
 relativity and connexion of the changing phenomena, the 
 succession or coexistence of which could no longer be regarded 
 as ultimate facts, or as sufficiently explained by the mathe- 
 matical synthesis. Kant recognises this step, in so far as he 
 discovers that the dynamical synthesis is below and beyond the 
 mathematical, and that it is necessary to explain the deter-., 
 mination of objects in space and time as such. And he 
 himself shows the way to a still further synthesis, in so far as 
 he makes a transcendental regress upon the unity of the 
 conscious self, as a principle which underlies both the mathe- 
 matical and the dynamical synthesis; though, as has been said, 
 he uses this principle only to determine the objects of 
 experience as phenomenal. 
 
489 
 
 CHAPTEK VIT. 
 
 THE MATHEMATICAL PRINCIPLES. 
 
 WE have seen, in the fifth chapter of this Book, in what Eecapituiaaon 
 way Kant was led to schematise the categories ofof^^e^^'^™ 
 Quantity and Quality, as respectively Number and Degree. Number and 
 
 Degi'ee. 
 
 SucK schematism was necessary because he started from the 
 presupposition that unity and plurality, position and negation, 
 could not be combined in pure thought. He had, indeed, 
 spoken in the metaphysical Deduction of the categories, as if 
 the third category, in which these pairs of opposites were 
 imited — the category of Totality in Quantity and of Limitation 
 in Quality,— ^were in some sense the product of pure thought, 
 and therefore derivable from the analytic judgment. But as 
 he comes closer to the problem, he finds that such categories 
 imply a synthesis, which is possible for us only through the 
 intermediation of the idea of time. It is this idea which enables 
 us to bring together terms which in pure thought start asunder 
 as absolute opposites. On the other hand, though the idea of 
 Time enables us to bring these opposites together, it is only 
 under certain limitations that it does so. It does not enable 
 us finally to reconcile them or resolve them into elements of 
 one conception, but only to produce a relative synthesis, a com- 
 bination in which the elements still repel each other, and so 
 give rise to an infinite series. Accordingly, Totality, as the 
 synthaeis of Unity and Plurality, takes the form of Number, 
 
490 THE CRITIQUE OF PURE REASON. book r. 
 
 i.e., a synthesis of separate units in a whole which is never 
 complete ; and Limitation, as the unity of Eeality and Nega- 
 tion, takes the form of Degree, i.e., a synthesis by which different 
 elements of reality, which are negatively related to each other, 
 are combined in a real being, which yet always is ex parte negafio, 
 i.e., always defined by reference to a reality beyond it, and 
 which never reaches a position that excludes negation, and 
 a determination which is self-determination. In fact, the 
 conceptions of Totality and Limitation, as the perfect syntheses 
 of those opposed elements, are now seen to arise out of the 
 contrast between the analytic unity of pure thought, which is 
 one with itself because in it the opposing elements are not 
 brought together at all, and the synthetic unity of schematised 
 thought, which contains an unsolved contradiction. These 
 conceptions are really Ideas, which are unrealisable by the 
 understanding, either in thought or in knowledge : Ideas which 
 serve to point out the phenomenal character of the objects of 
 knowledge, without enabling us to determine the noumenal 
 reality we oppose to them, 
 portotthe'' This last statement, however, takes us beyond the point 
 
 Deduction of... , ,t.,.^,. 
 
 the Mathema- which we have at present reached, viz., the Deduction of the 
 
 tical ^ ' 
 
 Principles. Mathematical Principles, i.e., the justification of the appK- 
 cation of the schematised categories of Quality and Quantity as 
 universal predicates of all the objects we know. How are we 
 to prove that all such objects are extensive quanta, and again 
 that, as qualitatively determined, they are intensive quanta,^ 
 In making such universal statements, we are obviously going 
 beyond what can be given in particular sensation or perception, 
 and stating certain conditions under which alone such percep- 
 tions or sensations can be combined with self-consciousness. 
 Now, the Transcendental Deduction was a proof of the general 
 proposition that nothiug can be so combined except in so far as 
 its elements are put together into an image of perception in 
 conformity with a priori conceptions, and in so far as the per- 
 ceptions so constituted are recognised as conforming to such 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES., 491 
 
 conceptions. The Deductioii of each special principle of pure 
 understanding, accordingly, can lie only in the proof that it in 
 particular is one of the functions essential to this process, in 
 view of some special characteristic of our experience. In other 
 words, taking the different characteristics v^hich belong to our 
 experience of objects in time and space, we have to ask how 
 such experience can be my experience, i.e., can be united with 
 my consciousness of self For, while the conceptions of the 
 objects of such experience, taken in abstraction from this unity 
 (i.e., from the fact that they are known objects), might not give 
 ground for the assertion of any principle that goes beyond the 
 particular experience itself, it is Kant's contention that, when 
 we correct this abstraction, and observe that they are known 
 objects, or objects for a conscious self, we are obliged to go 
 beyond, the particular, and lay down a principle which is 
 absolutely universal, i.e., which holds good for all objects of 
 possible experience. 
 
 How are we to apply this view to the principle that all Deduction of 
 
 '^'^ •' r r tlig Principle 
 
 objects of esyperience are extensive quanta ? The answer which °| perctptim! 
 Kant gives is simply that we can reach the proof of this 
 principle by considering that all such objects are objects of 
 perception, and that, as such, they are apprehended as existing 
 in space and time. Now, the perception of an object as in 
 space and time, as here and now, is possible only in so far as 
 the " here " is determined in relation to other " heres " and the 
 " now " in relation to other " nows." The particular time can- 
 not be taken as a moment cut off from all other times, or the 
 particular place as a point cut off from all other places ; in 
 fact, such points and moments are merely limits in the con- 
 tinuity of space and time, while every perceivable space or 
 time is a quantum, exclusive of other spaces and times, but con- 
 tinuous with them. It is a " here," which may be divided into 
 many " heres " ; a " now," which may be divided into many 
 " nows " ; and, on the other hand, it is necessarily taken as 
 itself • included, if we may so express it, in the one great 
 
492 THE CRITIQUE OF PURE REASON. book i. 
 
 " Here " and " Now " of Space and Time. But how then can 
 a particular space or time be perceived ? In the Aesthetic the 
 answer was given, that it can be perceived only as a special 
 limitation of Space or Time. Space and Time are thus sup- 
 posed to be " infinite given wholes," which are received into 
 the passive mind through perception, and in which objects 
 are given as occupying definite places and times. But this 
 answer was merely provisional ; for space and time are not 
 given wholes with reference to which objects could be defined, 
 or in which they could be perceived as parts. Even in the 
 Aesthetic they are declared to be mere forms of the relations of 
 possible perceptions, a view of them quite inconsistent with the 
 idea that they could be given as complete wholes determined 
 in themselves, apart from any synthetic activity of the mind 
 to which they are presented. And, indeed, if we try to think 
 of them in that way, our mind is driven into an infinite series 
 by the alternation between its effort to represent its object as 
 a whole and the nature of the object represented. " The con- 
 ception of Totality in this case is nothing more than the idea 
 of the completed synthesis of the parts of space : for, as we 
 cannot get the conception by abstraction from the perception of 
 the whole (a perception which in this case is impossible), we 
 can apprehend it only through the synthesis of the parts up to 
 the completion of the infinite, which we think at least as an 
 Idea."'^ 
 Objects can be If howcver, a particular space and time cannot be perceived 
 
 known as in ■*■■*■ *^ 
 
 tto?oSyby without relation to other spaces and times, and indeed to all 
 Snd^rthe^ spacc and time; and if, further, space and time cannot be given 
 
 schema of 
 
 Number, and in perception, as infinite individual wholes in which aU the 
 
 by the con- ^ ^ 
 
 that"™hlma! P^^^s are, or can be, fixed with reference to the whole, how can 
 any object be known by us as in space and time ? Only, it is 
 obvious, in so far as space and time are " generated " for us as 
 perceptions in the same synthetic process in which the mind 
 combines the data of sense into definite images, i.e., into 
 lA. 428; B. 456. 
 
CHAP. VII. THE MATHEMATICAL PEINCIPLES. 493 
 
 images the elements of which are put together in a definite 
 way. Further, in order that we may have spaces and times 
 before us as objects, the mind must not only thus " generate " 
 space and time, but it must become conscious of the unity of 
 its own synthetic action in this process ; i.e., it must become con- 
 scious of the principle of relation upon which it has been acting. 
 Now this principle, stated generally, is the conception of a, 
 synthesis of homogeneous units, which, as homogeneous, form a 
 continuity in which there is no smallest part and no largest 
 whole. Hence, in regard to all times and spaces, we can lay 
 down the principle that they exist for us as objects only by 
 means of a continuous synthesis of homogeneous units in per- 
 ception, and also by means of a consciousness on our part of 
 the unity of our .thought with itself in such synthesis, as being 
 throughout determined by the conception of extensive quantity. 
 Now, it is this consciousness of the unity of our thought with 
 itself in determining the object which is expressed in the 
 judgment, "This space or time is an extensive quantum." 
 And, as all sensible objects are perceived as in space and time, 
 it follows that what has been said of space and time must be 
 said of all objects of experience. " They cannot be appre- 
 hended or received into our empirical consciousness, except by 
 that synthesis of the manifold whereby the ideas of a definite 
 space or time are produced, i.e., by the combination of the 
 homogeneous, and the consciousness of the synthetic unity of 
 its manifold elements (as homogeneous)." ^ Hence, phenomenal 
 objects are all quanta, and indeed extensive quanta; since, as 
 perceptions in space or time, they must be set before our 
 minds by the same synthesis whereby space and time are 
 determined. 
 To do iustice to this " deduction," we must observe its Positive and 
 
 " neg:ative 
 
 double meaning. On the one hand, it demands that we deduction!"'^ 
 should consider the object in a transcendental point of view, 
 as an object known through perception to a conscious self; 
 • 1 B. 202. 
 
494 THE CRITIQUE OP PURE REASON. BOOK I. 
 
 and it shows us that, from this point of view, the object has 
 necessarily certain characteristics which do not belong to it, 
 when this its relation to knowledge is abstracted from. On 
 the other hand, it makes us recognise that, when we take away 
 from the object what it has in virtue of this relation, we must 
 deprive it also of everything that at first it seemed to have 
 belonging to it as an object in time and space. In the first 
 part of this argument, we take the object as if it were indepen- 
 dent of consciousness, and we ask how it can become such an 
 object /or us. And the answer we get is, that it cannot even 
 be given to us in perception as an object in space and time 
 unless the data of sense are bound together in continuous 
 synthesis, as an extensive quantum, i.e., as a many-in-one of 
 homogeneous units ; and that, even when so given, it cannot, so 
 to speak, be taken by us, or united with the " I think," unless 
 we become conscious of the unity of thought with itself in this 
 synthesis, i.e., conscious of the object as an extensive quantum. 
 However we may suppose the object to be qualified in itself, 
 synthesis according to the schematised conception, and con- 
 sciousness of the unity that determines the synthesis, i.e., of 
 the schematised conception itself, are necessary, ere the object 
 can exist for me as an object in space or time. But then, (and 
 this is the second turn of the argument) to suppose the object 
 in itself as having any qualification apart from, or prior to, this 
 its determination for us, is an absurdity. For, if we start with 
 the consciousness of the object, as related to the " I think," in 
 the way just described, and ask what would be left of such an 
 object, if we took away from it all that is due to the synthetic 
 process by which it is generated in perception in conformity 
 with the category of extensive quantity, and all that is due to 
 the consciousness of it as determined by that category, the 
 answer must be that without the process of synthesis there 
 could be no perception at all ; and that without the conscious- 
 ness of the unity of thought with itself in that process, i.e., 
 without the consciousness more or less clear of -the category. 
 
CHAP. VII. THE MATHEMATICAL PEINCIPLES. 495 
 
 the perception would reduce itself to an image on which no 
 judgment was made, and which, therefore, was no image of 
 anything for the perceiving subject. And we must always 
 remember that an image so distinguished from, and related to, 
 the conception which is its principle of unity, is a very 
 different thing from' a mere image not so distinguished and 
 related. It is a mistake, as has been repeatedly pointed out, 
 to take the perceptive half of our consciousness and suppose an 
 animal to be gifted with it ; for the perception which is the 1 
 subject of the judgment of knowledge, cannot, 'prior to thej 
 judgment, have the form which it gets m the judgment. It is notsuffloTenti 
 this point which Kant seems most often to forget; but if it ^n?.'^*'^*"''^ 
 be forgotten, the judgment at once becomes either unmeaning 
 or impossible ; for if the particular be taken to be something 
 apart from the universal, either it will have nothing in virtue 
 of which it is capable of being subsumed under the conception, 
 or, if it has, it wiU have no need of the subsumption. In the 
 present case, it appears to be a simple analytic judgment to say 
 that objects, since they can be objects for us only as in space 
 and time, are extensive quanta ; for the analysis of space and 
 time at once supplies us with this predicate. But this is 
 simply because in the " generation " of space and time, we are 
 guided by a principle of which we become aware in the very 
 act by which we determine a particular space or time, or any 
 phenomena in it, as an object. On the other hand, untU this 
 consciousness arises, — involving, as it does, the distinction and 
 relation of perception and conception, as particular and univer- 
 sal, subject and predicate, in the judgment, — we have no 
 perception of a space and time such as could supply a subject 
 to be judged about. The synthetic movement is not one in 
 which a ready-made subject and predicate are brought together 
 (in which case we should be obliged to take the judgment as 
 analytic), but one by which they are for the first time dis- 
 tinguished from and referred to each other, and so constituted 
 as a illiity in«difference, such as finds its full expression in a 
 
496 THE CRITIQUE OF PURE BE ASON. book I. 
 
 judgment. A particular object in time or space can be repre- 
 sented as such object, only in so far as the universal principles 
 of synthesis in conformity with which its manifold, as a 
 manifold of space and time, is combined into one image, are 
 separated from and referred to it. On the other hand, its 
 manifold can be determined as the maiiifold of a particular 
 object in time and space, only in so far as the elements of that 
 manifold are combined with each other in the unity which is 
 thus recognised as conformed to the principle. And hence, also, 
 we see how it is possible, from the determination of particular 
 given spaces and time, to derive rules which hold good for all 
 space and time; for we know them as particular spaces and 
 times, only in recognising a universal principle of determination 
 which holds good for all spaces and times, and of course for 
 all objects determined as in space and time. This principle, 
 therefore, justifies the application of Geometry to all objects of 
 external sense, and of Arithmetic to all objects of sense what- 
 ever. For it shows us that the " ostensive " construction of 
 figures which we make in Geometry, as well as the " symboli- 
 cal " construction of numbers which we make in Arithmetic, 
 are identical witti the constructions which we are obliged to 
 make in determiniug particular objects of perception in time 
 and space as objects for us. It is for this reason that Kant 
 entitles this principle the " Principle of the Axioms of Per- 
 ception." 
 oftoTdeSr -^^ phenomenal objects, as present to us in space and time, 
 S>^cte°as°' are thus necessarily determined as extensive quanta. It should 
 quanta'™ perhaps be added, to complete Kant's thought, that they are 
 determined only as extensive quanta, i.e., that the conception of 
 quantity can be applied to them only according to the schema 
 of number. Hence, as such objects, they are represented as 
 parts of a whole, which yet is no whole, for its synthesis can 
 never be completed ; and again, as wholes of parts which yet 
 are not ultimate parts or units, for its division cannot be com- 
 pleted. Hence, the principle that all phenomenal objects are 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 497 
 
 extensive quanta, if it is a principle which is necessary to the 
 determination of the manifold of the elements of perception as 
 objects for a conscious self, is, at the same time, a principle by 
 which they can never be adequately determined for that self. 
 For, in order that they may be so determined, it is necessary, 
 on the one hand, that they should be brought under a unity 
 beyond which there is no further difference, and, on the other 
 hand, that they should be reduced to elements which are indi- 
 visible units. While, therefore, Kant insists that all pheno- 
 mena must be determined by this principle, if they are to be 
 known as objects, he, at the same time, has laid the basis for 
 the counter assertion of the Dialectic, that all objects, so far 
 as they are known through this principle, are determined as 
 mere phenomena. 
 
 The second mathematical principle is that all objects of Deduction of 
 
 ■*-■*■ "^ the principle 
 
 experience, as objects corresponding to sensations, have inten- tio'nsS'St- 
 sive quantity ; or, as Kant puts it, that " in all phenomena, the ^^^"^^ '°°- 
 real, which is the object of sensation, has intensive quantity or 
 degree." The meaning of this will be clearer if we keep be- 
 fore us its contrast with the previous principle. Under the 
 principle of extensive quantity, Kant had maintained that, in- 
 asmuch as every object of experience is known as an object in 
 time and space, and inasmuch as times and spaces can become 
 objects for us only by a synthesis in conformity with the 
 category of extensive quantity, and by the recognition of the 
 category of extensive quantity as the principle of unity in such 
 synthesis, we can lay it down as a universal principle that all 
 objects of experience are extensive quanta. Now, he goes on to 
 say that objects of experience are not merely spaces and times, 
 though, for the purpose of mathematics, we may regard them 
 as if they were. Time and space are forms of relation between 
 objects the nature of which is not completely expressed in 
 these relations. Now, the question is how such a nature can exist 
 for MS, and what determinations we are obliged to attach to it as so 
 
 existing. The first answer naturally would be that the qualitative 
 VOL. I. 2 I 
 
498 THE CRITIQUE OF PURE REASON. BOOK r. 
 
 nature of things is given in sensation, just as their quantitative 
 nature was given in the forms of sense. But as, in the latter 
 case, we had to point out that the forms of sense could not be 
 united with the " I think," so that a particular object might be 
 known by us as an object in space and time, without a synthesis 
 of its manifold, and without a recognition of the unity of thought 
 in that synthesis, so here we have to ask whether there is not 
 necessary a similar synthesis and a similar recognition, in order 
 that the qualitatively determined object may be known as such. 
 Now, the qualitative determination of an object in pure thought 
 would not, as Kant in criticisiug Leibniz admits, imply any nega- 
 tive determination of it ; it would be the absolute affirmation of a 
 reality without respect to any other limiting reality. But in our 
 consciousness of objects through perception, determination is 
 always ex parte negaiio ; it is the assertion of a quality in a thing, 
 not absolutely, but in relation to something which limits that 
 assertion. This, however, is for Kant a consequence of the fact 
 that our determination of objects takes place under the form of 
 time, which is supposed, on the one hand, to make it necessary that 
 the qualitative determination of an object should be by limita- 
 tion, i.e., by affirmation in relation to negation, and on the other 
 hand, so to condition this determination by limitation that it 
 cannot get beyond a determination in degree or intensive 
 quantity ; i.e., an assertion which is ex parte negatio, an assertion 
 of a certain degree of reality which lies between zero and 
 infinity, (iafinity being = an assertion that would exclude all 
 negation).! Now, sensation cannot of itself furnish us with the 
 qualitative determination of an object ; for, even if it be re- 
 garded as furnishing the matter for such determination, yet 
 such matter cannot become the consciousness of an object as 
 qualitatively determined, unless, by a continuous synthesis, we 
 determine it as against its negative, and unless we become 
 conscious of the unity of thought with itself in this synthesis, 
 i.e., of the category of intensive quantity as giving unity to it. 
 1 Cf. above, p. 446. 
 
CHAP. vir. THE MATHEMATICAL PRINCIPLES. 499 
 
 Hence, while we cannot anticipate any of the special qualities 
 which belong to sensible objects as such, (for this would be to 
 anticipate our special sensations) we can lay down by anticipa- 
 tion the general principle, that all objects of sense can be 
 affirmatively determined by us as having a certain quality, only 
 by a continuous synthesis of affirmation as against negation, 
 and by the recognition of such synthesis as having its unity in 
 the principle of intensive quantity. 
 
 There is a good deal of obscurity in Kant's explanation of J^^'^|«^?™*f 
 his principle owing to his not fully stating in connexion with ^nVs'* '° 
 it certain aspects of his thought, which are presented elsewhere, 
 particularly in the chapter on the " Amphiboly of the Eeflective 
 Conceptions." It will, therefore, be necessary to show more 
 definitely, (1) that Kant conceives of Limitation as a category ^ 
 brought into play only in relation to phenomena as known 
 under conditions of time ; (2) that in this relation. Limitation 
 is regarded by him as necessarily taking the form, not of the 
 simple assertion of a quality as opposed to its absolute 
 negation, but of the partial assertion and partial negation of that 
 quality ; and this again is the assertion of it as having a certain 
 degree of reality, which is limited by another reality that in 
 relation to it is negative ; and lastly (3) that it is only as we 
 determine sensation according to this principle, that it can 
 furnish the material for the determination of objects as 
 .qualified. 
 
 The first of these points was the necessary consequence of Position and 
 
 ^ 1/ J. negation' are 
 
 ideas which were present to Kant as early as his JEssay towards exXs"™ in 
 the Introduction of the Idea of Negative Quantity into Philosophy ; but not°in ' 
 
 n . „- ^ , .IIP! knowledge as 
 
 lor, as it Will be remembered, the mam thought oi that essay mediated by 
 
 perception. 
 
 is the contrast between real and logical opposition. According 
 to the laws of thought, we must treat negation as the mere 
 absence of affirmation ; but the negation of any positive quality 
 or state of a thing, requires, or at least may require, a positive 
 ground. It may require to be accounted for as the result of 
 the neultalisation of the ground for one position by the presence 
 
500 THE CRITIQUE OF PURE REASON. book 
 
 of the ground for another.^ In Kant's critical period th 
 principle is maintained, but is limited to phenomenal realitj 
 and it leads to the following criticism of Leibniz, which wit 
 slight verbal alteration, is frequently repeated : — " Leibniz cou] 
 not see his way to bring in a 'priori perception as an elemei 
 or factor in his principle of sufficient reason, but reduced thi 
 principle to mere a priori conception. The consequence i 
 this was, that, like Democritus, who conceived the universe 1 
 be made up out of atoms and the void, he reduced all things, i 
 a metaphysical point of view, to compounds of reality an 
 negation, being and not-being ; so that no reason could be give 
 for a negation except only the absence of any reason, for 
 positive determination. Hence, out of so-called metaphysici 
 evil in union with metaphysical good, he produced a world ( 
 mere light and shadow; nor did he stop to consider that, i 
 order to put a space in shadow, there must be a material bod; 
 i.e., something real which prevents the light from penetratir 
 into it. In his view, pain had its reason only in want ( 
 pleasure, vice only in the absence of virtuous impulses, and tl 
 rest of a material body only in the want of moving force. Fc 
 he argued that according to mere conceptions, reality = A cai 
 not be opposed to reality = B, but only to want of reality =( 
 Nor did he reflect that the same thing does not hold good i 
 perception. Thus in external perception, i.e., under the a prio: 
 conditions of space, there is possible an opposition of one reaUf 
 (a moving force) to another, to wit, a moving force in tl 
 opposite direction ; and in like manner, in internal perceptio 
 opposite realmotives may be combined in one subject in such a we 
 that a priori the result of this conflict of realities may be known ' 
 be negation. It is true that in order to see this, Leibniz wou' 
 hav6 been obliged to introduce the idea of opposite directior 
 a thing which can be perceived or envisaged, but not repr 
 sented in mere conception. But as he was not aware of th 
 oversight, he was landed in the absurd and even immoT 
 ' Cf. above, p. 126 seij. 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 501 
 
 principle, that all evil, regarded as a reason, is = 0, i.e., is a 
 mere limitation or, as the metaphysicians say, merely formal. 
 Hence, his principle of sufficient reason, taken as a mere law 
 of thought, did not help him in the least to get beyond 
 the principle of analytic judgment, i.e., the principle of 
 contradiction, or to widen his knowledge by an a priori 
 synthesis." ^ 
 
 The bearing of this criticism of Leibniz is to show that, Hence they 
 
 are reciprocal- 
 
 in the phenomenal world, an opposition between realities is i^fn'^iin^^^" 
 possible, so that the negation of one may be its neutralisation butTn pheno- 
 
 1 -T7- 1 1 r» T T •! ■ J ■ mena they even 
 
 by another. Xant does not, therefore, dispute Leibniz s view imply each 
 that, according to conceptions, (and therefore in things in them- 
 selves as determined by conceptions only), the negative and the 
 positive are absolutely exclusive of each other ; but he contends 
 only that in objects as represented under conditions of time 
 and space, it becomes possible to conceive an opposition of 
 realities, and a negation which is the resultant of such opposition. 
 Thus, he does not admit the principle that " determination is 
 negation," or that negation and position cannot be separated, 
 and that the truth of both is to be found in limitation. Still 
 less does he go on to admit a conception of reality as a unity 
 of factors, which are negatively related as excluding each other, 
 yet positively related as each implying the other. Hence, the 
 category of Limitation, in so far as it involves such a unity of 
 affirmation and negation, is for him irrational, and what takes 
 its place is the idea of the determination of an object by an 
 absolute affirmation w'hich excludes all negation. But this 
 remains a mere idea ; for the determination of phenomena as 
 .objects, (and these are the only objects we can determine) is 
 always by an affirmation which is relative to an opposite 
 limiting affirmation. This, however, involves a step, beyond 
 what is stated in the above criticism of Leibniz ; for, whereas 
 that criticism showed that in the phenomenal world negation 
 is possible as the determination by each other of opposites 
 • 1 E. I. 517 ; H. VIIL 544. 
 
502 THE CRITIQUE OF PURE REASON. BOOK i. 
 
 which are equally positive, Kant now goes on to assert that it 
 is possible only as such determination ; and, conversely, that 
 position is possible only in relation to opposite position, i.e., a 
 ' position which is ex parte negatio. In other words, he main- 
 tains that a reality cannot exist for us under conditions of 
 time and space, except in so far as we represent it as limited 
 by an opposite reality. For this (as was pointed out before) ^ is 
 what is really meant by saying that reality is schematised for 
 us as degree, and so as standing between zero and absolute 
 reality ; i.e., that we are not capable of determining an object posi- 
 tively, except as we at the same time determine it negatively ; 
 or our af&rmation of it must be at the same time an affirma- 
 tion of its limit, and so a partial assertion and partial negation 
 of it — an assertion of it as so far real, and so far not real. 
 But this incapacity of ours to affirm any object as absolutely 
 so and so qualified without reference to its negation, is, Kant 
 holds, to be explained by the conditions of time under which 
 we apprehend it, and determine it as an object. For, to say 
 that it is given in sensation as so and so qualified explains 
 This is due to nothing, unless we can also see how it is determined as so 
 
 the form of 
 
 time under and SO qualified for us. Now, such determination under con- 
 
 which reality 
 
 E^suoh™"^'^ ditions of time iijiplies a continuous process, by which we take 
 up and unite together the elements which go to the positive 
 determination of the object as so and so qualified ; and this at 
 the same time involves the negative determination of it by 
 exclusion of an opposite qualification. Hence, Kant says that, 
 " the schema of a reality, as the quantity of something so far 
 as it fills time, is just this continuous and uniform production 
 of it in time, which we realise when we gradually go down from 
 the sensation which has a certain degree till it vanishes alto- 
 gether ; or gradually rise from negation to the intensity which 
 it has for us." In other words, the determination of an object 
 as so and so qualified is possible, not by the immediate presenta- 
 tion of it in sensation, but only by a synthetic process which, as 
 ^ Of. above, p. 446 seq. 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 503 
 
 in time, is coutinuous and infinitely divisible; a process in which 
 the sensation is represented as filling the moment of time more 
 and more completely to the exclusion of its negative ; and in 
 which we become conscious of the unity of the perception with 
 itself, as thus produced. Our qualitative determination of an 
 object is, therefore, at the same time a determination of it as 
 having a certain degree of its quality, which is above zero, but 
 short of an absolute reality that excludes all other reality. For 
 the absolute affirmation of a quality would involve that it 
 should fill consciousness so as to exclude its negative altogether; 
 and with this exclusion the affirmation itself, which is possible 
 only in relation to its negative, would disappear. When, there- 
 fore, we become conscious through sensation of an object as 
 qualitatively this or that, we are necessarily conscious of it in 
 relation to its negative, which divides the sphere of conscious- 
 ness with it, and which we conceive as varying in inverse ratio 
 with it. And this quantitative variation of quality moves 
 between limits determined by the fact that pure affirmation 
 and pure negation are equally impossible, as neither is possible 
 except in relation to the other. This, however, as we have 
 already seen, does not, in Kant's view, result from any law of 
 pure thought, according to which affirmation and negation are 
 tied together; for in pure thought they would exclude each 
 other. But a sensation, if it is to be the ground for a con- 
 sciousness of an object as qualified, must be conceived as 
 occupying the moment of time with a certain degree of in- 
 tensity. And that it may do so, its matter must be "joined 
 with time" by a continuous synthesis in which the quality 
 gains grouhd against its negation. In other words, we can 
 become conscious of an object as qualified, only as we deter- 
 mine it as so qualified in a synthetic process, which continuously 
 blends the elements of sensation into an intensive unity, and as 
 we become conscious of the unity of thought with itself in this 
 process, a unity of thought which is defined as the category of 
 intensive quantity. 
 
504 THE CRITIQUE OF PURE REASON. book I. 
 
 Sfon^T^is This view of Kant as to the way in which quality becomes 
 mattoin'" known as against its negation, but as standing in inverse 
 quantitative ratio to it, becomes more explicit in Kant's treatise 
 on the Metaphysical Rudiments of Physics, where he apphes it 
 to external experience. In that treatise he attempts to dis- 
 cover what is involved in the determination of matter as 
 occupying space, and he finds that this affirmative determina- 
 tion of it is possible only if we conceive it, at the same time, 
 as standing in relation to other matter to which it is negatively 
 related. In other words, he finds that matter can be conceived 
 only as the subject of two opposite forces of repulsion and 
 attraction, which are both positive, but which are negatively 
 related to each other ; for if its parts were conceived as merely 
 attracting each other, it would disappear in a point, {i.e., would 
 lose that self-externality which is one of its essential charac- 
 teristics,) and if they were conceived as simply repelling each- 
 other, it would disappear by diffusion into the infinite ; (i.e., it 
 would lose that connexion of the reciprocally exclusive parts 
 which is its other essential characteristic). In both cases, the 
 qualitative nature of matter as occupying space would dis- 
 appear, and we should be left with empty space, which by 
 itself is no possible object of perception. This view is ex- 
 pressed in the words with which Kant sums up his chapter on 
 the Qualitative Determination of Matter : " If we look back 
 on the course of our argument, we shall see that ft first con.- 
 siders the Real in space (otherwise called the solid) as filling 
 it by repulsive force ; next, it proceeds to consider that force 
 which in reference to the repulsive force, as the proper object of 
 external perception, may be called Negative, viz., the attractive 
 force, by which, if it were to act by itself, all space would be 
 penetrated, and the solid would be entirely cancelled ; thirdly 
 it takes note of the Limitation of the former force by the latter 
 and of the resulting determination of the degree of the filling 
 of space. Thus we have a complete treatment of the Qualitf' 
 of Matter under the titles of Reality, Negation, and Limitation, 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 505 
 
 SO far as it is required for the purposes of Metaphysical 
 Bynamics." ^ / 
 
 In this passage, Kant transfers to the determination oi'^^^^^°^ 
 
 objects in space what he had before said in relation to the^Lefand ^ 
 
 « 1 . -r^ time. 
 
 determination of objects as presented to us in time. Deter-, 
 mination is by limitation ; but the possibility and the necessity 
 of this is regarded as arising, not from anything in the nature 
 of thought by which positive and negative are bound together, 
 but from the fact that all objects are regarded as known under 
 the conditions of space and time. Space makes opposite 
 directions possible ; nay, it makes it impossible to determine 
 any direction of movement, except by relation to that which 
 has an opposite direction of movement, — which of course involves 
 that it is impossible absolutely to deterrhine anything as in a state 
 of rest or of movement. And if anything be represented as 
 occupying space, it must occupy it by a repulsive force which 
 can be conceived only as exerted against an attractive force 
 that hmits it. In like manner, time makes it possible to con- 
 ceive of the growth of quality through an infinite number of 
 degrees of intensity, in which there is no absolutely lowest or 
 absolutely highest point : nay, we can represent such a quality 
 as a determination of an object only in so far as we conceive it 
 as produced by a synthesis, in which we may either say that it 
 grows up for us from nothing by continuous additions, or that, 
 by such additions, it continually encroaches upon, and dimin- 
 ishes, its negative. In this way, it becomes possible to represent 
 in perception a unity of opposites, which for pure thought is 
 impossible. Nay, it becomes necessary to represent such a 
 unity, as the condition under which alone objects can be deter- 
 mined for us as qualified objects in space and time. 
 In the argument we have just tried to reproduce, there is a Difficulty in 
 
 ' combining 
 
 certain difficulty in connecting Kant's different statements. Kant's differ- 
 
 ■^ *^ ent statements 
 
 This difficulty arises from the fact, that in the chapter on ^Ln and"**" 
 Schematism and also in that which treats of the principle of the ^^'^^''' 
 1 R. V. 378 ; H. IV. 416. 
 
506 THE CRITIQUE OF PURE REASON. BOOK i. 
 
 Anticipations of Sense-perception, (Wahmehmung) Kant is seek- 
 ing, as his main object, to show that the process by which the 
 object is determined as a qualified object, involves a synthesis 
 of the elementary data required for such perception in accordance 
 with the conception of intensive quantity. Hence, he does not 
 dwell upon the fact that the process of determination is at the 
 same time a process of limitation, and that, therefore, it is a 
 determination of that which limits and of that which is Hmited 
 in opposition to each other. This, indeed, is implied in the 
 account of such determination as ex parte negatio, but it is 
 thrown into the background by the other aspect of the process. 
 In other words, Kant shows more clearly that the process is a 
 continuous synthesis, by which a sensation is taken up into 
 our consciousness under conditions of time, than that it is a 
 determination by negation. Hence we are apt to understand 
 him as admitting that the determination of a thing as qualified 
 is given in sensation, and as bringing in the principle of the 
 Anticipations of Sense-perception only to determine the quality 
 as having a certain degree. In truth, however, his argument 
 is not merely that quality must be determined as having degree, 
 but that quality cannot be known as such except by a synthesis 
 which involves at once positive and negative determination. 
 What follows from this, that sensation is not even a part of our 
 consciousness of objects, except as already determined by a 
 conception, is just the counterpart of what was said in relation 
 to time and space as the forms of perception. As it is only 
 through the conception of extensive quantity that times and- 
 spaces, or objects as in time and space, become for us definite 
 objects of perception of which judgments may be made; so it is 
 only as referred to the conception of limitation in the form of 
 intensive quantity, and as determined by that conception, that 
 a sensation becomes a definite " object of sensation," a qualita- 
 tively determined thing of which anything can be said. But 
 in such consciousness the sensation is not a state of a sensitive 
 subject, but the perception of an object such as exists only for 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 507 
 
 a conscious self. And, so taken, it implies the category by 
 which in the judgment of experience it is determined. Hence, 
 here also, the ultimate effect of Kant's argument is to show, not 
 only that sensation must be determined according to the 
 principle of the Anticipations of Sense-perception, but that, as 
 mere sensation, apart from such determination, it has no place 
 even as an element in the experience of a conscious self. This 
 point, however (in which Kant's argument as it were returns 
 upon itself), is always the point where he seems to recoil from 
 the result of his own logic, and to make reservation of some- 
 thing given, which is independent of the determination by which 
 alone, as it is allowed, it can be brought into relation to self- 
 consciousness. 
 
 We are now prepared to gather up the results of Kant's view JJ^^^^^j .^^ 
 of the Mathematical Principles, and to point out the fundamental InXttoe bT"'' 
 misconception which prevents it from being completely satis- otTotauty'and 
 
 Limitation. 
 
 factory. In both cases, Kant supposes that the schematising 
 of the categories in relation to time (and space) makes possible 
 a combination of factors which in pure thought would repel 
 each other, a combination, that is, of unity and multiplicity, of 
 reality and negation ; and in both cases he thinks that this 
 synthesis is confined by the nature of time and space to an 
 imperfect, external, and therefore quantitative form. Now, we 
 may admit the latter of these propositions, but not the former. 
 Space and time, as We have already seen, are not required to 
 make possible the combination of the conceptions of the one 
 -and of the many; for these conceptions are nothing at all 
 apart from their unity. Nor, again, are they required to make 
 it possible to unite position and negation, which are abstract 
 elements of thought that mean nothing except as correlated. 
 On the other hand, it may be admitted, that our first concep- 
 tions of things receive their colour from the external, pictorial 
 representation of them as in space and time. It may also 
 he admitted that in such conception, the combination of 
 reality ♦and negation can appear only as the external and 
 
508 THE CRITIQUE OF PURE REASON. booki. 
 
 therefore quantitative limitation of one reality by another, 
 and the combination of unity and multiplicity can appear 
 only as extensive quantity or number. A merely quantitative 
 determination of things is a determination in which they 
 may be taken as one or as many indifferently, according to 
 the aspect in which we view them, but in which we can 
 never conclusively determine them as totalities ; for quantity, 
 as such, has no definite limits of division or of combination 
 except such as are determined for it from without. And so 
 space and time, as quantities, are without limits of extension 
 or division. In like manner, in the merely qualitative 
 determination of things, their essential relativity is hidden 
 from us ; and consequently, the affirmation of them as so and so 
 qualified seems to be independent of the determination of any 
 limit to their existence as so qualified. Such determination of 
 things is not explicitly a relative determination of them, still 
 less is it a determination in which positive and negative are 
 essentially correlated with each other. It is a determination 
 in which one thing limits another, but seems at the same time 
 to be indifferent to it. Hence the limit itself is conceived as 
 external and merely quantitative. In both cases, therefore, we 
 have a synthesis which can never be final, because of the merely 
 external relation into which the terms are brought ; and it is 
 easy to see how, under these conditions, the effort to combine 
 the terms should give rise to an antinomy and a progresms in 
 infinitum. This antinomy and this progressus, however, are due 
 to an abstraction, or to put it more exactly, to the neglect of an 
 element which, though already present in our thought, when we 
 use the categories of quality and quantity, is not itself made 
 the object of attention. For the truth is, not that unity and 
 multiplicity, reality and negation, cannot be finally united in 
 ■ one consciousness, but that, when we so unite them, we go 
 beyond the determination of things by the simple categories of 
 quantity and quality. In fact, when we use these categories, 
 we are already beyond them, i.e., we have in our consciousness 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 509 
 
 the means of correcting their inadeq^uacy. For if we think out 
 the category of quantity, we see that there is no meaning in a 
 multiplicity which is not the differentiation of a unity, or in a 
 unity which is not the principle of a determinate difference ; 
 and that, therefore, quantity, as a mere external synthesis of 
 units, is the abstraction of one aspect of things which cannot 
 be taken by itself. In like manner, the idea of an affirmative 
 or positive determination of a thing which has not its negative 
 in itself, but only meets it as an external limit, is an abstrac- 
 tion which breaks down when we see what it really means. 
 Now, the fact that we bring things under such inadequate 
 categories is immediately bound up with the other fact, that we 
 represent them as in space and time, and do not yet recognise 
 that they have any relations except as coexistent or successive, 
 or as externally limiting each other in the occupation of space 
 and time. When, however, we have reached the point of- 
 seeing that such categories are, in themselves and in their 
 application, antinomical, we have already begun to look beyond 
 the antinomy to the unity of factors opposed. When we have 
 discovered that unity generates multiplicity as its counterpart, 
 and that this again must be apprehended as unity, which again 
 generates multiplicity, we are near seeing that it is the very 
 effort to take an object as one and indivisible which forces us 
 to go beyond it, and regard it as one of many, and con- 
 versely; and that the only valid conception is that of a totality, 
 in which each element is determined through the others. Again, 
 when we have seen that a reality, determined as such, is by 
 that very determination referred to another than itself, we are 
 near to a conception of the unity of position and negation, not 
 as a quantitative limitation of one thing by another, but as a 
 reciprocity of determination. It is such a conception that Kant 
 presents to us in his construction of matter, in which each 
 factor is the negation of the other, but in which each is at the 
 same time bound to the other, and would disappear if it were 
 not tllere. But the conception of matter as the subject of two 
 
510 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 opposite forces must be essentially altered whenever the two 
 forces are taken as correlative. For, so taken, they cannot be 
 conceived as externally working against each other and 
 externally limiting each other, or as gaining and losing at each 
 other's expense. Eather, we are obliged to think of both as 
 the manifestations of one principle, which maintains itself 
 through the division and relative opposition of its factors, and 
 controls that opposition by reference to the unity to which 
 they both belong. Even, however, abstracting from this unity, 
 and looking only to the correlation which Kant estabhshes 
 between the two factors, we can see that, in the Metaphysical 
 Budiments of Physics, the mere attempt to understand what is 
 meant by the qualitative determination of things subject to the 
 schema of Degree, carries Kant beyond all the mathematical 
 principles, and forces him to bring in the highest dynamical 
 principle of Reciprocity. 
 Kantwewof There is another point of view suggested by Kant, from 
 the ideaa'o? °' which the two mathematical principles may be .viewed in 
 intensive relation to each other. Kant separates extension from inten- 
 
 quantity. 
 
 sion and seems even to speak sometimes as if there were no 
 necessary relation between them. This separation is, however, 
 parallel with and dependent on another, viz., the separation 
 between the form and the matter of perception. Both, as 
 given, have to wait for a mental synthesis and for the recogni- 
 tion of the principle of that synthesis. Take away the matter 
 of sense, and suppose the synthesis exercised only on the pure 
 manifold of space and time, and we should have as our object 
 pure extension without intension. Take away the form of 
 sense, and suppose a synthesis exerted on the matter, and we 
 should have pure intension without extension. Such a separa- 
 tion of the two forms of synthesis, however, is, on Kant's own 
 showing, impossible. For, on the one hand, time and space 
 are mere forms of relation which cannot be perceived except by 
 a synthesis which, at the same time, combines the manifold of 
 sense as in time or space. On the other hand, though a 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 511. 
 
 sensation as occupying a moment in my sensitive life may have 
 none but intensive quantity, yet it cannot be determined as 
 thus occupying it, except by a synthesis which combines it with 
 time in accordance with the schema of degree, and by a 
 recognition of the unity of thought in such synthesis. If, 
 therefore, Kant says that " apprehension by means of sensation, 
 if we look to it only, fills but a moment," he is yet obliged to 
 add that " intensive quantity belongs to sensation through the 
 apprehension of it, in which the empirical consciousness in a 
 certain time rises from nothing to a given measure"; in other 
 words, that in order to represent a sensation as real with a 
 certain intensity, in virtue of which it occupies a certain 
 moment of time, we have to represent it as generated by a 
 successive synthesis in the previous moments. It would 
 appear, therefore, that the consciousness of intensive quantity 
 or of occupation of time, implies the consciousness of a process 
 ia time; as, on the other hand, a consciousness of any part 
 of time impKes the consciousness of something that occupies it. 
 
 This becomes intelligible if we consider for a moment the intensive 
 
 quantity 
 
 nature of intensive quantity or degree. A degree is simply a }|fted to'erfen- 
 quantum which is contemplated as in itself continuous, so that ^'™ 9"™'"y- 
 its manifoldness is regarded as merely the determination of its 
 limit. In this way, however, it loses all its determination as a 
 quantum, except in relation to other quanta outside of it. Thus 
 if we consider twenty degrees of heat as the twentieth degree, 
 we take it as a unity which has no multiplicity in itself, but 
 which derives its determination as manifold, and so as quantum, 
 from its place in a scale in which the other degrees are external 
 to it. Thus a sensation can be represented as having a par- 
 ticular degree of intensity only in so far as we conceive it as 
 generated in a process which passes through all inferior degrees 
 up to the given amount. And this means that, if it is deter- 
 mined as occupying a moment of time with a certain degTce of 
 intensity, it must be by a synthesis which has refetence to 
 other mements, and so to time as an extensive quantity. The 
 
512 THE CRITIQUE OF PURE REASON. book I. 
 
 same correlation of intensive and extensive is illustrated by 
 Kant's construction of matter out of the two opposite forces of 
 repulsion and attraction ; for the intensity of each force is 
 measured by the extent of its influence, and a mere intensive 
 force is just one which moves a given mass through a given 
 space in a given time. 
 Sm ofae" Kant uses the principle of intensive quantity to disprove the 
 degreebymore doctrinc of the mathematical physicists, who hold that the 
 
 or less of 
 
 voAmum. different weights of different kinds of matter when taken in the 
 same volume, can be explained by the existence of more or less 
 of vacuum between their material particles. This inference is, 
 he thinks, illogical ; for we have no reason to deny the possi- 
 bility of a variation in the intensive occupation of space by 
 different kinds of matter, while nevertheless each of them 
 is extended through the whole of the space so occupied. And 
 he supports this objection by a reference to the impossibility 
 of our having an experience of empty space. In truth, the 
 inference flows from the fact that the repulsive force by which 
 matter fills space to the exclusion of other matter, is deter- 
 mined only in relation to the attractive force which holds them 
 together. Hence what is not occupied by the one object, must 
 be occupied by the other. In the Critique, where Kant speaks 
 of the limitation of the positive by the negative as the degree 
 of a quality, it is not so clear that the negative, in relation to 
 which the positive is determined, is another positive. Hence, 
 it would seem at first as if the true conclusion were, that 
 exj)erience can never be either of absolutely full or of abso- 
 lutely empty time or space, but only of relative fulness 
 or emptiness. It is only when we see that the negative, 
 in relation to which the positive is externally determined, 
 must be conceived as the limit of another positive, that 
 the force of Kant's reasoning becomes evident. The same 
 ambiguity, it is to be observed, appears in Kant's Metaphysical 
 Buclimcnts of Physics, where it is said that an object fills space 
 by repulsive force ; if it did so fill it, attractive force must be 
 
CHAP. VII. THE MATHEMATICAL PRINCIPLES. 513 
 
 conceived as emptying space, and the balance of the two 
 forces vrould imply that space was neither absolutely full nor 
 empty. In fact, however, as Kant himself points out, either 
 force would empty space, if it were not conceived as limited by 
 the other.^ The conclusion, therefore, seems to be that the 
 true positive or real — in this case, the fulness of space — can- 
 not be found in that which is positively determined apart from 
 its negative, but only in the unity of opposites, which implies 
 at once their negative and their positive relation to each other. 
 
 The principle of the Axioms of Perception showed that "^^^ principle 
 
 r r J M Qf ^jjQ Antici- 
 
 Geometry and Arithmetic are applicable to objects of experi- genSperLp- 
 
 , ,T ii_'i.i? • 1 j_i tion justifies 
 
 ence, because they can be objects oi experience only as they are theappuoation 
 
 , . , , . , . - o£ the higher 
 
 ,, subjected to the same synthesis, by which times and spaces are mathematical 
 determined ; i.e., to a synthesis according to the principle of ^^P'^noe. 
 extensive quantity. The principle of the Anticiipations of 
 Sense-perception [Wahrnehmung'] shows that the higher mathe- 
 matical Calculus is applicable to objects of experience, because 
 they can be objects of experience only as qualitatively deter- 
 mined ; and because they can be qualitatively determined only 
 as they are subjected to a synthesis in which they are generated 
 out of nothing according to the principle of intensive quantity. 
 Kant does not, however, anywhere distinctly refer to this 
 principle as a vindication of the objective validity of the 
 Differential and Integral Calculus, as corresponding to the 
 process by which objects are qualitatively determined in our 
 experience, although it is an extension of his thought which is 
 quite in his spirit, and which is favoured by some of his 
 expressions.^ 
 
 ^R. V. 358; H. IV. 400. 
 
 ^Itis, however, implied in several passages in the Anfangsgriinde, e.g., R. V. 
 375 seq. ; H. IV. 413, and more detinitely stated in some of the Reflexionen. 
 This thought is due to Prof. Cohen (Kant's Theorie der Erfahrung, p. 422 seq., 
 and his separately published Essay, " Das Princip der Infinilesimal Meihode in 
 seiner Geschkhte "). My knowledge of Mathematics is not sufficient to enable 
 me to discuss this subject with any confidence. The passages in Kant's works 
 and in the Refle,xicmen Kant's which bear on this interpretation are discussed 
 by Prof. CiJhen in his Theorie der Erfahrimg, p. 430 seq. 
 
 VOL. I. 2 K 
 
514 
 
 CHAPTEE VIII. 
 
 THE ANALOGIES OF EXPEEIENCE. 
 
 ^roMem was T^HE Analogies of experience occupy a central place in tl 
 caSitytbut Critique for many reasons. They were the principle 
 
 solve it, he of knowledge to which Kant's attention was directed even i 
 
 found it 
 
 "reyiousT to ^^^ earliest of his purely metaphysical works, the Dilucidat'i 
 m^hematloai Nova / and it was to them that he returned in the Dissertatio 
 of 1770 as the " principles of the intelligible world." ^ It we 
 Hume's treatment of one of those principles which made hii 
 finally shake off his " dogmatic slumber," and opened the we 
 for those investigations which issued in the Critique, and moi 
 especially in the metaphysical and transcendental deduction > 
 the categories : and when he has to give a specimen of h 
 method, he almost invariably takes the principle of causality i 
 his example. The other principles were brought into view by tl 
 effort of Kant to do what he charges Hume with omitting to d 
 i.e., to universalise his problem. For, no sooner had Kant sef 
 that a doubt could be thrown on the universality and objecti' 
 validity of the category of causality, than he began to ai 
 whether there were not other conceptions of the pure unde 
 standing which were exposed to a similar challenge, and to lo( 
 for that guiding thread to their discovery which he found in t 
 finished work of Logic. Hume's argument rested upon t^ 
 fundamental positions, each of which Kant attempted to tui 
 ^See above, p. 183. i, 
 
 :M: :. ■ 
 
CHAP. VIII. • THE ANALOGIES OF EXPERIENCE. 515 
 
 On the one hand, it rested on the idea that things are given in 
 their qualitative and quantitative determination as objects in ■ 
 space and time apart from all mental synthesis ; and, on the / 
 other hand, on the idea that out of this given determination of 
 things, by the aid of custom and imagination, the determination 
 of them by causality might be explained. N"ovr, as I have 
 already suggested, Kant turns the first of these positions by 
 pointing out that the mathematical determination of things 
 involves a synthesis of the matter of sense according to its "^ 
 form, and the recognition of this synthesis as in harmony with 
 the schematised categories of number and degree, or of extensive 
 and intensive quantity. Hence, any doubt thrown on the prin- 
 ciple of cause as carrying us beyond the simple data of percep- ■ 
 tion, must equally be thrown upon the science of mathematics ; 
 and also upon all that determination of objects of experience 
 as definitely quantified and qualified objects in space and time, 
 by reason of which the science of mathematics has objective 
 validity instead of being a mere systematic dream. Even, there- 
 fore,'if we admitted that there could be an experience of objects 
 apart from the use of the principle of causality or any similar 
 principle, and that from it we might, by the aid of imagination 
 and custom, derive that principle, we should not have escaped 
 the necessity for a synthesis according to a priori conceptions. 
 The deduction of the Mathematical Principles is the proof that 
 such a synthesis is necessary to the determination of objects as 
 quantified and qualified, a determination which they must have 
 if they are to be known as in space and time. 
 
 But is this enough ? Can we say that these principles are why Kant re- 
 
 , quireaaspeclal 
 
 suflacient for the determination of objects of experience as such ? deduction of 
 
 "" ^ the Analogies. 
 
 " The two preceding principles," answers Kant, " which I have 
 caUed mathematical in view of the fact that they authorise us 
 to apply mathematics to phenomena, had to do with phenomena 
 as respects their mere possibility. Their object was to show 
 how such phenomena could be produced according to the laws 
 of mathematical synthesis, both as respects the pure element in 
 
516 THE CRITIQUE OF PURE REASON. book i. 
 
 their perceptive form and as respects the real element in their 
 empirical perception." ^ But " experience is empirical know- 
 ledge; i.e., it is a knowledge which through sense-perception 
 determines an object. It is, therefore, a synthesis of sense- 
 perceptions, a synthesis which is not itself contained in sense- 
 perception as such, but which involves that its manifold is 
 brought to synthetic unity in one consciousness. For this unity 
 is what is essential to the constitution of a knowledge of the 
 objects of sense, i.e., to experience, as distinct from mere percep- 
 tion or sensation." ^ It is true that we have a kind of synthetic 
 unity established by the mathematical principles. But "it is a 
 synthesis of the homogeneous, and so of elements which do not 
 necessarily belong to, or require, each other, as, e.g., the two 
 triangles into which a square is divided by the diagonal do not, 
 when taken severally, require each other." But the dynamical 
 synthesis " is the sj'nthesis of the manifold in so far as its 
 elements necessarily belong to each other, as, e.g., the accident 
 belongs to some particular substance, or the effect to its cause: 
 a synthesis in which different elements are represented as 
 heterogeneous and yet as united together a priori." Such 
 synthesis " is not arbitrary '' (like the other), for it is " a 
 connexion which concerns the existence of the manifold." ^ Now, 
 " experience is a consciousness of objects through sense-per- 
 ceptions, and consequently it claims to be a consciousness of 
 the relation of the manifold in existence, i.e., not as it is put 
 together in time, but as it objectively is in time : and, as 
 time itself cannot be perceived," (so that objects could be 
 determined in relation to it,) " the determination of the 
 existence of objects in time is possible only through their 
 combination in time generally ; which means their combination 
 by means of conceptions which establish an a priori connexion. 
 But as such conceptions always carry necessity with them,- 
 experience is possible only through a consciousness of the 
 necessary connexion of sense-perceptions." * 
 
 1 A. 178; B. 221. ^g, 2I8. ^B. 201 note. ^B. 219. 
 
CHAP. vill. THE ANALOGIES OF EXPERIENCE. 517 
 
 I have already partly explained the meaning of these Bxpiauatton 
 
 •i i- •' -^ o of Kant 8 view , 
 
 passages. Kant holds that the determination of objects as gi^sdSfemTne' 
 quantified and qualified, presupposes in its turn the deter- °xi^ingT 
 mination of them as objects in distinction from and in 
 relation to the self. The transcendental regress, which first 
 brought into view the necessity of recognising the action of 
 the categories of quantity and quality in experience, needs only 
 to be carried a little farther to bring into view the necessity 
 of this further determination of them, or in other words, of 
 recognising that objects cannot merely be given as such. It 
 is not enough to say they are objects and so distinct for the 
 self that knows them ; for let them be objects ever so much, 
 we have to ask how they come to be objects for us. Now, 
 Kant answers that it is just the function of the Analogies of 
 Experience to determine objects as such, or so to connect the 
 data of sense with each other that for us they determine objects. 
 On this I have already remarked that the determination of objects 
 as existing, and not merely thought or imagined, would naturally 
 be expected to be the function of the categories of modality; as 
 it is the modality of the judgment which alone brings into view 
 the relation of that determination of the object which the 
 judgment expresses to the subject who makes it. But it is 
 Kant's view that the consciousness of the self that judges is 
 the consciousness of that synthetic activity by which the data 
 of sense are .combined into the unity of experience. It goes 
 along with,' and is dependent upon, the recognition of the 
 synthesis of phenomena as conforming to certain a priori prin- 
 ciples. Hence, the determination of phenomena as objects 
 existing for the self is conceived as resulting from their deter- 
 mination as necessarily connected elements in one experi- 
 ence; or we might even say, the former determination is 
 regarded as essentially one with the latter. In Kantian 
 language, it is only as the data of sense are combined by 
 necessary laws into one context of experience, and recognised as \ 
 so combined, that they can be known as objects, and so united J 
 
518 THE CRITIQUE OF PURE REASON. book i. 
 
 with the consciousness of self. Eeiiexion upon the modality of 
 objects or their relation to the conscious self, therefore, only 
 brings into view the fact that they are determined as objects 
 because of their necessary relation to each other, as elements in 
 one experience. Such reflexion, accordingly, destroys that 
 natural dualism in which objects are regarded as things in 
 themselves independent of consciousness, and reduces them to 
 phenomena, at the same time that it explains how, by means of 
 the Analogies of Experience, they are determined as objects. 
 Of the defects of this view of the modal determination of 
 objects, ■ a good deal has been said already, and more will be 
 said in the next chapter. Here it is sufficient for us to 
 recognise that, as was already indicated in the first chapter of 
 the Introduction, the reflexion which brings into view the 
 connexion of things with each other according to the Analogies 
 of Experience, a reflexion which is implied in the modem 
 scientific view of nature, is a real advance toward a more 
 accurate and complete view of things. And it is an advance 
 gained in the only way in which a new development of human 
 knowledge is possible ; i.e., by a transcendental regress, which 
 makes us aware of elements previously working in our con- 
 sciousness, but not previously attended to. And it is of no 
 little interest to examine how Kant proves that such a regress 
 is at once necessary and legitimate. 
 ?hT?eiation of ^^ begins by speaking of the connexion of phenomena by which 
 'experimoe*^" they are determined as objects as a connexion of them in time, 
 '™ ^ ■ which is the form of inner sense and therefore of all perception. 
 This view is afterwards supplemented by a somewhat imperfectly 
 developed theory of the relation of inner to outer experience ; 
 and the fact that, in the first instance, Kant pays regard only 
 to connexion in time, must be considered as a result of the 
 lingering psychological view of the Deduction — as if its object 
 were to explain the process whereby a succession of sensations 
 or perceptions is turned into an experience of objects which, as 
 such, are in space. Kant, however, more and more worked him- 
 
.CHAP. vm. THE ANALOGIES OE EXPERIENCE. 519 
 
 self free of this idea, and substituted for it the true conception 
 of a transcendental regress as, not an account of the genesis of an 
 intelligible experience out of elements that existed before it, but 
 an attempt to realise all that experience is, and to bring to 
 light all the elements implied in it. Taking the argument as 
 it stands, however, we may paraphrase it as follows : 
 
 Experience, as the objective determination of phenomena, what Kant 
 
 seeks to prove. 
 
 really means their determination as necessarily connected in 
 time, i.e., as connected in time according to certain rules which 
 are a priori, and which, therefore, are not dependent upon the 
 sequence in which these phenomena, as mere determinations of 
 sense, are presented in the subjective experience of the indivi- 
 dual, or in which he may arbitrarily choose to combine them 
 in his imagination. In other words, we recognise a perception 
 as referring to an object, only in so far as we recognise it as 
 necessarily related in time to our other perceptions according 
 to certain general principles, which are not derived from the 
 actual succession of our perceptions, but which determine that 
 succession (as they determine all other successions). But to 
 recognise such a connexion is not to recognise something which 
 could be given to us from without ; it is simply to discover a | 
 connexion which our intelligence necessarily introduces into 
 the data of sense, in the very process of synthesis by which it 
 constructs them into perceptions. 
 
 Now, what is Kant's way of proving this thesis ? As usual ^fe ^Sio^es 
 with him, we have a dialectic which turns upon itself and thedcterminl* 
 
 . ^ , tions of 
 
 reconstitutes the premises from which it starts. In other phenomena 
 
 -*- as in time. 
 
 words, he begins with the supposition of that, as known inde- 
 pendently of the categories, which in the end he shows to be a 
 possible experience only by means of these very categories. 
 For his purpose is to show that it is only through the applica- 
 tion of the categories that objects can be determined as existing 
 in time, and as in time exhibiting phenomena which succeed or 
 coexist with each other. But at first he seems to speak as if 
 such tfene-relations of phenomenal objects could be given apart 
 
520 THE CRITIQUE OF PURE REASON. book I. 
 
 from the categories, and as if the categories were needed merely 
 to universalise them. Thus, he tells us that the conception of 
 objects as existing, changing, or coexisting in time, does not 
 directly involve any necessity of connexion between them. If, 
 then, we confine ourselves to what we apprehend, what is to 
 become of substance, causality, and reciprocity ? What right 
 can we possibly have to add to the perceived facts what is not 
 given in them ? If all I perceive is the existence of an object 
 at a particular time, what can possibly give me a right to say 
 that the object remains in existence and maintains its identity 
 with itself beyond the moment when I am perceiving it : still 
 more, as the principle would seem to involve, to say that it 
 exists throughout all time ? If all I perceive is a particular 
 change, how can I ever be authorised to say that the succession 
 which I see now is determined according to a universal rule 
 which holds good for all time ? If all I perceive is a particu- 
 lar coexistence, how can I have a right to say that the co- 
 existing objects are determined as such according to laws of 
 reciprocal dependence, which hold good not only in this case 
 but always and everywhere ? Obviously in each of these cases, 
 I am bringing a particular phenomenon under a category or 
 universal form of relation, and so giving to the determination 
 of a particula,r fact a universal value. But what gives me the 
 right or the power to do so ? How can I thus subsume the 
 particular case under a law which carries me so far beyond it ? 
 Kant's answer is, that I must so subsume the particular case 
 under the universal rule, because otherwise I could not bring it 
 into relation with the possibility of knowledge ; and it must be 
 capable of being so subsuined, otherwise it would be as good as 
 nothing for me. 
 Ambigmtyin Kow, what docs tliis mean ? Does it mean that, the parti- 
 
 the statement ' ^ 
 
 neSty. cukr case being given in its complete determination as a sub- 
 jective appearance, I then, in order to determine it as an 
 objective fact, bring it under the category, for which it is 
 marked out by something in its character as an appearance ? 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 521 
 
 Or does it mean simply that, if I take a particular object 
 merely as particular, without thinking of the conditions under 
 which it can be known as such an object, I need not think of 
 it as determined by the category ; but that the moment I begin 
 to think how it is possible that such an object should be 
 known to me, I see that such a category is necessarily implied 
 in it ? I have already said a good deal on this subject ; but, as 
 Kant generally discusses it with reference to the Analogies of 
 Experience, it seemed best to reserve the full consideration of it 
 for this place. First, then, let us look at one or two of the 
 most definite passages on the subject.^ 
 
 " Axioms are synthetic a 'priori principles which are 
 immediately certain. Now, it is impossible to combine one 
 conception with another synthetically, and yet immediately ; 
 for we need a third mediating principle to enable us to reach 
 beyond the first to the second. Now, as philosophy is nothing 
 but rational knowledge according to conceptions, no principle 
 can be found in it which deserves the name of an axiom. We 
 can have axioms in mathematics, just because we can give 
 intuitive construction to our conceptions, and so ' envisage ' 
 their objects. For thus we are able immediately and a priori 
 to connect the predicates of such objects with them. Thus we 
 can see at once, e.g., that three points always lie in one plane. 
 On the other hand, a synthetic proposition based on mere con- 
 ceptions, cannot be immediately certain, e.g., the proposition 
 that ' everything that happens has its cause ' ; for here 1 must 
 look about for a tertium quid, which I can find only in the 
 conditions of the determination of the time of phenomena in an 
 experience, and I could not immediately and directly derive' 
 such a principle from conceptions alone. Thus, we see that 
 discursive principles are quite different from axioms or intui- 
 tive principles." ^ 
 
 " This proof does not show that the given conception {e.g., of 
 
 * One of the clearest passages, A. 737 ; B. 765, has been already quoted, p. 475. 
 ^A. 736; B. 760. 
 
522 THE CRITIQUE OF PURE REASON. book i. 
 
 that which happens) leads directly to another conception (that 
 of cause) ; for such a transition would be a leap in the dark 
 which could not be justified. What it does show is that 
 experience itself, and so the object of experience, would be 
 impossible without such a combination." ^ 
 
 " In the Transcendental Analytic, we have deduced the 
 principle that all that happens has a cause, by considering 
 wliat is the one condition of the objective possibility of a con- 
 ception of that which happens. We have proved, in short, that 
 the determination of an event in time, and therefore the event 
 itself, as a fact of experience, is impossible, if it does not stand 
 under such a dynamic rule. And this is the sole ground of 
 proof possible in such a case ; for it is only as an object is 
 determined for it, according to the law of causality, that the 
 idea of the event has objective validity or truth." ^ 
 
 " That all that happens has a cause cannot be inferred from 
 the conception of that which happens ; rather, it is the prin- 
 ciple in question which shows how alone we can attain a 
 definite empirical conception of that which happens." ^ 
 Meaning of Thcse passagcs show us clearly what Kant is seeking to 
 
 Kant's state- r n j o 
 
 princi*ie*of'^° cxpress. He starts with Hume's difBculty, that what we call a 
 cannot'be cause and an effect are two phenomena which we apprehend 
 derived from Only as Standing in a time-relation to each other. He admits, 
 
 the idea of an 
 
 event. therefore, that in itself the conception of an event or change 
 
 does not immediately involve the conception of causality, does 
 not immediately imply a universal relation of succession 
 between the two phenomena in question. For, he goes on to 
 ask : where can we find such a proof ? The only attempt to 
 ■give one, is the argument that what happens is accidental, and 
 what is accidental has necessarily a cause in something else. 
 But what do we mean by a thing being accidental ? We do 
 not mean merely that we can think the opposite, but rather 
 that, as it did not exist before, it does not now exist by any 
 
 lA. 783 ; B. 811. ^A. 787 ; B. 815. 
 
 '' A. 301 ; B. 357 ; cf. also A. 722 ; B. 750. 
 
succession of 
 impressions. 
 
 CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 523 
 
 inherent necessity, but only because something else causes it to 
 exist. But this is just the principle of causality. Tor, to say 
 that which comes into existence has a cause, is just to assume 
 the principle of causality. If, then, we simply take the con- 
 ception of an event in this way, abstracting from the conditions 
 under which it can he known to us as an objective fact, we 
 must allow to Hume that the principle of causality is expli- 
 cable only as an illegitimate conversion of the mere repeated 
 succession of like impressions into a belief in the necessary 
 sec[uence of different objective phenomena. 
 
 When, however, we look at it in this way, we begin to see it implies the 
 
 reduction of 
 
 that a further criticism of Hume's doctrine is necessary. To *° ^™"* *° * 
 
 */ SU( 
 
 say that an event or change occurs, means for Hume only that 
 one particular impression follows another in consciousness. It 
 is subject to this reduction, or on the tacit assumption of it, — 
 i.e., on the ground that in perception, we have given only one 
 impression and then another after it, — that Hume denies the 
 principle of causality. "What he really shows is that the bare 
 consciousness of perception after perception, or rather of impres- 
 sion after impression, does not contain in it any necessity of 
 relation. Or, in other words, reducing an event to terms of mere 
 sense-perception, all that we find in it is simply the conscious- 
 ness of impression after impression connected by the relation 
 of time. It is of such a sequence, and not of a change in an 
 object, that Hume is speaking when he says that it does not 
 involve any universal relation of causality ; for he himself 
 admits that we can reach the objectivity of a relation only 
 through its supposed universality. When Kant, therefore, in 
 answering Hume, admits that the principle " that all that 
 happens, has a cause,'' involves a synthesis, in which we go 
 beyond the conception of what happens, he is speaking from 
 Hume's point of view, and reducing the sequence to that 
 which alone, from that point of view, it can be, i.e., to a mere 
 consciousness of sequence of impressions, a perception of a 
 sequen'ce of perceptions. He withdraws, in fact, from the 
 
524 
 
 THE CRITIQUE OF PQRB REASON. 
 
 BOOK I. 
 
 Of the three 
 determina- 
 tions of 
 things as in 
 time only 
 sequence can 
 plausibly be 
 derived from 
 our impres- 
 sions. 
 
 Hume's treat- 
 ment of the 
 Analogies in 
 his earlier 
 Treatise. 
 
 perception of what happens all determination by conception, 
 and then asks simply whether the residuum would give us 
 ■causality ; and he answers that it would not.^ 
 
 Now, it is well to follow out the consequence of this a little 
 further. The possible determination of things as in time are 
 three : they may last through time, and they may succeed each 
 other, or coexist with each other in time. But in such a con- 
 sciousness as we have described, there is no room for co- 
 existence, since its states are always successive ; and there is 
 really as little room for permanence, since the impressions of 
 no two moments are exactly alike, and even if they were, this 
 would constitute no identity between them. To say that we 
 recognise a perception as identical with another can mean only 
 that it is a perception of the same object ; and of this objective 
 reference we empty our thoughts just in so far as we confine 
 ourselves to our perceptions or impressions as such. In fact, 
 sequence is the only relation in time with reference to which it 
 is possible even to suppose that it could be given in perception; 
 and we shall soon find reason to question whether even this 
 could be so given. 
 
 N"ow Hume, in his earlier treatise, had maintained, with 
 some inconsistency^ that we can through impressions be con- 
 scious of the relations of coexistence in space and succession 
 in time, and then he had proceeded by means of these relations 
 to explain away the conceptions of identity and causality. 
 
 In dealing with the first of these principles he had made 
 matters easy for himself by assuming at the outset that identity 
 is actually given to us in perception, in so far as the impressions 
 which we have in successive moments are not different in con- 
 tent.^ For, on this assumption, he had only to explain why like 
 impressions, received after an interval, are referred to the same 
 object — a difficulty which he solved by supposing that we 
 
 ^ The question whether it can he even so much as this, is not here con- 
 sidered. 
 ''■ Cf. Green's Introduction to ffume's Treatise, § 236 seq. 
 
CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 525 
 
 confuse like impressions, before and after the interruption, with 
 identical impressions, and that we invent the fiction of a con- 
 tinuous existence of an object, which is not an impression, to 
 explain the recurrence. Causality, again, he explained as a 
 conception generated by the recurrence of two different im- 
 pressions in succession (which are, of course, identified at every 
 recurrence as the same or of the same class), and by the in- 
 creasing tendency of the mind, as affected by this recurrence, to 
 pass from the one impression to the " idea " of the other. 
 Eeciprocity he did not deal with, but it is easy to see that, on 
 the same method, he would have been able to generate it out 
 of the repeated coexistence of different impressions ; nor would 
 any great difficulty have stood in his way in such an explana- 
 tion, seeing that he had already prepared the way for it by 
 treating the relations of things in space as given directly in 
 our impressions, in spite of the successive character of the 
 latter. 
 
 In Hume's later work, the discussion upon identity dis- H"me makes 
 
 ■*■ *' the problem 
 
 appears, and with it the difficulty of generating permanence, ^^li^hy^ 
 
 . , , . j_ ^ j_i J • confining it to 
 
 coexistence and succession out of the same perceptive con- causality, 
 seiousness, which is, as Hume describes it, a " bundle or 
 collection of impressions and ideas, succeeding each other in 
 perpetual flux or movement." For, he has now to deal merely 
 with the one perceived relation of succession, . and the parallel 
 conceived relation of causality. Further, the omission of the 
 discussion upon identity has made it easy for him to treat the 
 consciousness of an objective change as nothing more than the 
 perceived presence in our consciousness of one impression after 
 another. Hence, in order to the reduction of the knowledge 
 of objective changes to impressions, nothing more is needed 
 than the confusion of such a consciousness of impressions as 
 succeeding each other in time with the successive impressions uant, on the 
 
 .1 . other hand, 
 
 tnemselves. deals with the 
 
 three Analo- 
 
 Now, there seems no ground to believe that Kant ever read gf^^^^n^cted 
 the Treatise. But his argument carries us back to the con- ^h^rf"°^ 
 
526 THE CRITIQUE OF PTJEE REASON. book i. 
 
 ceptions of the Treatise, in so far as he adds coexistence to 
 succession, and permanence to both, as " modes '' of Time, and 
 thus raises the question not as to causality alone, but also as 
 to substance and reciprocity. And it is to be noted that he 
 dwells on the impossibility of getting either permanence or co- 
 existence out of a consciousness which is merely successive, 
 and that he clinches his argument by pointing out that neither 
 the succession nor the coexistence of different phenomena can 
 be perceived by us, except as relations of the states of the 
 permanent. In other words, he throws on the perceiving con- 
 sciousness the weight of all the different relations of its matter 
 which it will have to bear, if all the determinations of objects 
 in time are to be explained out of the time-relations of impres- 
 sions ; while he also points out that these time-determinations 
 are not isolated from each other, so that one of them could be 
 explained without the others. And, in particular, he shows 
 that the other time-relations of phenomena have for their basis 
 the consciousness of a permanent . self-identical object, which 
 does not change with the change of its successive states. 
 
 " The three Analogies of Experience are nothing but prin- 
 ciples for the determination of the existence of phenomena in 
 time, according to all its three modes : — their relation to time 
 itself as a quantum, (the quantum of existence, i.e., duration); 
 their relation in time, regarded as a series or succession of 
 moments one after another ; and, finally, their relation in time, 
 regarded as having a certain breadth or compass in which all 
 existence is included at once. Such a unity of the determina- 
 tion of time is completely dynamical, i.e., time in it is not 
 viewed as that in which experience immediately determines its 
 place for every existence. That indeed is impossible, seeing 
 that absolute time is no object of perception, with which 
 phenomena could be brought into relation. But the rule of the 
 understanding, through which alone the existence of phenomena 
 can get synthetic unity in accordance with the relations made 
 possible by time, determines for every phenomenon its place in 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 527 
 
 time, and so determines it a priori for all time, and for every 
 time." " Our Analogies, therefore, exhibit the unity of nature, 
 as it manifests itself in the connexion of all phenomena under 
 certain exponents, which can express nothing but the relation 
 of time (in so far as it embraces all existence within itself) to 
 the unity of apperception, which can realise itself only in a 
 synthesis according to rules. Taken together, what they express 
 is, that all phenomena lie, and must lie, in one Nature ; since 
 without this a priori unity there could be no unity of ex-, 
 perience, and, consequently, no determination of objects in 
 experience. ^ ' 
 
 " In regard to the manner of proving these transcendental of the°iJa'n-"^ 
 
 scendental 
 
 laws or nature to which we have recourse, and the charac- Deduction of 
 
 the Analogies. 
 
 teristics by which it is distinguished from other ways of 
 demonstration, I shall make one remark, which must be of the 
 highest importance as a prescription which we can lay down 
 beforehand for every attempt to prove intellectual principles 
 which contain an a 'priori synthesis. If we had endeavoured 
 to prove these Analogies dogmatically, i.e., from conceptions, — 
 if, in other words, we had attempted in this way to prove that 
 nothing exists except what is found in a permanent object, 
 that every event presupposes something in the previous state 
 of things on which it follows according to a rule, and, finally, 
 that the states of things that coexist are determined in relation 
 to each other according to a universal rule of coexistence, — all 
 our endeavours would have been in vain. For, it is impossible 
 to pass from one object and its existence to the existence of 
 another or its manner of existence merely by means of the 
 conceptions of them, however carefully we may analyse these 
 conceptions. What, then, was left for us to do ? All that 
 was left for us, was to take our stand upon the possibility of 
 experience, as a form of knowledge in which all objects must 
 finally be capable of being, given to us, if the ideas of them are 
 to have objective reality for us. In this tertium quid, the 
 essenMal form of which consists in the synthetic unity of the 
 
528 THE CRITIQUE OF PURE REASON. book i. 
 
 apperception of all phenomena, we found certain a priori con- 
 ditions of the complete and necessary determination of all 
 phenomenal existence in time, without which even the empiri- 
 cal determination of time would be impossible. In these, 
 therefore, we discovered rules of the a priori synthetic unity 
 of experience, by means of which we are enabled to anticipate 
 its facts." 1 
 Isto^itiouT It i^ worth while to notice how Kant here states the principles 
 objectsas"' which are to be proved by means of the idea of the possibility of 
 
 existing in 
 
 time and the experience. The first is the principle that all that exists must 
 
 sequence and ■*■ 
 
 Sthei™tates ^"'s fouud in that which is permanent, i.e., that all existence is 
 Siown tous. summed up in permanent substances and their states. Now, 
 Kant maintains that the mere conception of an object does 
 not imply that it is a permanent substance. It is seen to do 
 so, only when we reflect that we cannot know any object as an 
 existence in time, unless we refer our perception to a substance, 
 which we conceive to remain the same through all changes of 
 its states or determinations. Again, the conception of an event, 
 if we think of it simply as the coming into existence of some- 
 thing that did not exist before, may be analysed as we please, 
 and we cannot get out of it any necessity of connecting it with 
 any other event which preceded it in time : but when we ask 
 how we come to know that something happens, or begins to be, 
 we have to recognise, that for us nothing can begin to be except 
 the state of a permanent object, and that that state can begin 
 to be only as determined according to a universal rule in rela- 
 tion to some previous change. Finally, we cannot, from the 
 mere conception of coexisting things, draw the inference that 
 they reciprocally influence each other's states, but this inference 
 becomes necessary whenever we reflect on the conditions under 
 wlaieh such a relation of existence can become known by us ; 
 for then we find that we cannot connect the states of sub- 
 stances coexisting, except on the presupposition of such influ- 
 ence. 
 
 > A. 215-17 ; B. 262-3. 
 
CHAP. vill. THE ANALOGIES OF EXPERIENCE. 529 
 
 Now, I have already indicated the point of view which we ^"^IX*^^"^ 
 have to take up in order to appreciate Kant's argument. It is 
 confused somewhat by the fact that, following Hume, Kant 
 does not, in the first instance, divest the perceptions of that 
 qualification by conception, which he afterwards shows to be 
 necessary to them in order that they may become elements of 
 experience. No doubt he corrects this when he speaks of the 
 perceptions as a mere manifold, which imagination takes up in 
 succession as it is given. But, if we take as our starting-point 
 the consciousness of objects as existing in time, — of events as 
 occurring in time, and of the states of different things as co- 
 existing in time, — -we already presuppose that determination of 
 perception by the necessary relations expressed in the cate- 
 gories which Kant seeks to prove. The objects of such a 
 consciousness are already conceived objects, and not merely 
 sense-perceptions. Hence, in the process of showing the neces- 
 sity of the determination of perceptions by the categories, Kant 
 is obliged to withdraw from objects perceived the attributes he 
 had lent them. He is, in fact, simply showing that objects 
 cannot have for us the predicates which we attribute to them 
 in our ordinary consciousness, except by the aid of a principle 
 which in that consciousness we usually leave out of account ; 
 and so is teaching us to rise above the natural abstraction in 
 which we take objects as if they were things in themselves, 
 given as objects in space and time, but otherwise unrelated 
 either to each other or to the self 
 
 We may put it thus: — the point which Kant would prove itreaUyiathe 
 
 '^ ^ ^ ^ correction of 
 
 is, that in relation to possible experience, the perceptions absS view 
 become qualified by conceptions, and hereby are determined ° ° ''^° ^' 
 as cases of universal laws, while in themselves, as perceptions, 
 they express' mere particular facts. Now, the real meaning of 
 this process lies in this, that the ordinary consciousness in its 
 determination of objects abstracts from, or neglects, the sub- 
 ject for which they are. Thus, things are conceived as exist- 
 ing, and" events as taking place, without reference to the fact 
 VOL. I. 2 L 
 
530 THE CRITIQDE OF PURE REASON. book I. 
 
 that they are objects for a subject, or to the conditions under 
 which alone they can become such objects. Hence, when we 
 correct our abstraction and restore the conditions which were 
 left out of account, we seem to be attaching new conditions to 
 those objects, whereas all that we do is to call attention to 
 conditions hitherto overlooked. From this point of view, then, 
 it may fairly be said that, when we consider things in relation 
 to the possibility of knowing them, we find a new light thrown 
 upon them which was not there so long as we took them as things 
 in themselves, determined to be what they are independently of 
 any such relation. It is in this sense that Kant's assertion of 
 the phenomenal character of the objects of experience reaUy 
 throws so much light on them. And it is open to us, from 
 this point of view, to say that objects, considered without refer- 
 ence to the possibility of knowing them, have not the predicates 
 that belong to them when considered in that reference, 
 objects to -"-f '^^ ^'^ ^^"^ °^^ ordinary conception of objects needs to be 
 
 Sme-reiations remouldcd by this reference to the subject for which they are, 
 
 are known . _ ^ _ . , . 
 
 through uni- the answcr IS that, taken apart from their relation to the 
 
 veraal prin- 
 ciples, subject and the conditions under which that relation can be 
 
 established, objects appear as merely particular, i.e., not as 
 
 instances of any principle that goes beyond themselves. Thus, 
 
 when I say that I am conscious of an object as such, I appear 
 
 to be expressing merely that which is present to me now and 
 
 here, but which may never be presented again. When I say 
 
 that I am conscious of an event as occurring, I appear 
 
 to be saying merely that one phenomenon makes room for 
 
 another in my perception, without any implication of a 
 
 universal relation between that which comes to be and that 
 
 which disappears. And when I say that I am conscious of 
 
 two objects as coexisting, this does not seem to involve any 
 
 correlation that goes beyond the moment of time in which they 
 
 are so determined. But, when I reconsider these propositions 
 
 in the light of the idea that the objects I am speaking of are 
 
 ' objects known, — objects, therefore, which imply a subject, and 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 531 
 
 which are capable of being united with the self-consciousness 
 of that subject, as one subject through all the variety of its 
 experiences, — I begin to see that this determination of objects 
 impHes conditions which I have not yet taken into account. 
 And these conditions, according to Kant's argument, are just 
 the dynamical relations of substance and accident, cause and 
 effect, and reciprocity. N"o object can exist for us through an 
 impression of .sense, except in so far as the content of such 
 impression is referred as a quality or state to some object, 
 which is permanent and so independent of the impression. 
 No events can be conceived as occurring simply because 
 one impression makes room for another, unless the con- 
 tent of the impressions so referred to a permanent objects- 
 are conceived as necessarily connected by a law which is 
 independent of the particular impressions or states in my 
 subjectivity. And no coexistence can be established, unless 
 our successive experiences have their content referred to 
 different objects which are permanent, and which are fixed as 
 in one time by a law of necessary dependence — a la"^ which 
 determines their succession in our consciousness and is not 
 , determined by it. 
 
 Now, it is to be observed that this process may, in one How tho o<m- 
 
 «ciowsjie«« of 
 
 point of view, be described as adding to perceived objects p^lf^g^^"^^" 
 predicates which are not analytically contained in fhem, but knowtedgTot 
 
 1-1 , . n 1 n . t ■ 1 -1 -T these objects. 
 
 which are combmed with them only m relation to the possibility 
 of knowledge. For, taken apart from this possibility, the objects 
 are considered to present themselves as mere particulars, and 
 not as particular instances of universal principles : — particular 
 objects in a particular time and space, with respect to which 
 we cannot know whether they will ever again present them- 
 selves in time and space ; particular events, as to which we 
 can never know that they are connected with any general 
 conditions under which they must occur ; particular co-exist- 
 ences, of which we cannot know that they are more than 
 accidenfal coincidences. On the other hand, nothing is really 
 
532 THE CRITIQUE OF PURE REASON. book i, 
 
 added to the conception of the objects by this new synthesis, so 
 far as these particular objects or events are really known 
 objects and events ; since, as such, they must contain all 
 the conditions that belong to known objects, however little 
 these may be present to the consciousness of the individual 
 who knows them. Hence, from this point of view, the move- 
 ment of thought which enables us to recognise the conditions 
 under which the particular objects are known as they are 
 known, leads us to withdraw from them, as apart from these 
 conditions, the qualifications which they seemed to have when 
 they were taken as things in themselves, i.e., as objective facts 
 given independently of any relation to a subject. 
 Positive and The meaning of Kant's argument then, in its first aspect, is 
 
 negative ° = r ' 
 
 Deductionr^ that it calls attention to the conditions under which objects or 
 facts are known, and shows that they are known as particular 
 objects or facts, only in relation to a self-consciousness, which 
 always determines its objects through universal principles. 
 As relation to the subject is part of the idea of the object, so it 
 is an object only as determined by principles which make it 
 no longer a mere particular. From the point of view of the 
 possibility of experience, therefore, we may enrich our empirical 
 knowledge by recognising, in each empirical fact as such, a 
 determination by universal principles. Dn the other hand, 
 from the point of view of the perceived object, if we take that 
 object as what it was originally supposed to be — as an object 
 merely given to the understanding and not determined by it — 
 we find that we must divest it of the qualification it seemed 
 to have, and reduce it to bare perception or sensation. Like 
 the will-o'-the-wisps in Goethe's allegory, we suck out the veins 
 of gold that made the clay image stand, and it falls together 
 in a confused heap, as a mere ' manifold of sense.' Now, ad- 
 mitting for the moment that this manifold could be apprehended 
 by sen.se as a series of subjective states succeeding one an- 
 other in time, it can be shown that, out of this mere succession, 
 we could not get any objective consciousness whatsoever ; and 
 
CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 533 
 
 that the categories must come in to add to perceptions something 
 not contained in them, ere we can reach such a consciousness. 
 But the addition, now takes a slightly different appearance. It 
 is not that the object or fact, taken as a particular object 
 or fact, carries with it no consequences beyond its own imme- 
 diate existence or occurrence in this moment and at this spot ; 
 for wa now discern that its determination as a particular object 
 for us involves universal principles, which carry us far beyond 
 its particular existence. But it is that, taking the mere per- 
 ception or impression of sense as it really is, we find that 
 it does not contain the determination in virtue of which it 
 represents to us particular objects. Now there is in Kant 
 a certain want of adjustment in these two aspects, which makes 
 it difficult for us to see whether he is speaking of what is 
 given in sense as it really is, — ^in which case he is speaking of 
 that which in itself cannot be an object of knowledge, and 
 which, therefore, requires to be combined with a conception . 
 not in it, ere it can become such an object ; or whether he is 
 
 h speaking of the particular object of knowledge, which is not 
 recognised as being determined by universal principles but 
 which really is so determined, — in which case he is speaking of 
 an object which is not only perceived but conceived, and which 
 we need only to understand in oraer to see that it is not 
 a mere particular, but an instance of a law. 
 
 After this general explanation of the method of Kant's 
 
 , deduction, we may now go on to consider how he applies it to 
 each of the Analogies of Experience : — to the principle of Sub- 
 stance, to the principle of Causahty, and to the principle of 
 
 pl^ciprocity. 
 
 '■' The principle of the permanence of substance, as stated in tS^pSndpie^ 
 the first edition of the Critiqim, is that " all phenomena contain "n tie first 
 
 and second 
 
 the permanent (substance) as the object itself, and the change- editions, 
 able as its mere determination, i.e., the manner in which the 
 object exists." In the second edition, the formula is — " In all 
 change*' of phenomena the substance is permanent, and the 
 
534 THE CRITIQUE OF PCTRE REASON. book I. 
 
 quantity of it in Nature is nevei- either increased or dim- 
 inished." The formula of the first edition brings out more 
 definitely the twofold aspect of the principle, as involving 
 a correlation of substance and accident, each of which is 
 determined in relation to the other ; while that of the second 
 edition starts with the assumption of the fact of change, and 
 goes on to refer the changing to the permanent as its necessary 
 presupposition. 
 The latter Now, the former mode of statement corresponds most 
 
 statement 
 
 exaotiywith ^xactly with what we should expect from the general account 
 Ka^nt"™" of Kant's method. For, according to that method, we should 
 
 deduction. i i -i i . c 
 
 expect him to prove that, while the mere conception of an 
 object does not necessarily imply that it must be conceived as 
 a permanent substance in a changeable phase of its existence, 
 we are obliged so to conceive it whenever we consider the con- 
 ditions under which alone we can know objects. In other 
 words, we should expect him to show that the distinction of 
 an object from the perception of it, which is implied in its 
 determination as an object, cannot be made except by the 
 application to it of a category which determines it as, or 
 changes it into, the consciousness of a substance manifesting 
 itself in a special accident ; or, passing from the category to the 
 schema, the consciousness of an absolutely permanent object in 
 a special changeable phase of its existence. The deduction 
 would, therefore, consist of two parts ; first, of the proof that 
 the mere conception of an object in dbstracto does not imply 
 that it is a permanent substance which manifests itself in 
 changing states ; and secondly, of the proof that such a deter- 
 mination of the object is necessary whenever we consider how 
 it has become known as an object to us. As a matter of fact, 
 however, Kant's argument takes a somewhat different course, 
 — viz., the course indicated for it by the formula of the second 
 edition ; and what he attempts to show is, not that any object 
 for us must be determined as a permanent subject of change- 
 able states, but rather that no consciousness of change is 
 
CHAP. vni. THE ANALOGIES OF EXPERIENCE. 535 
 
 possible which is not a consciousness of the sequence of states 
 in a permanent substance, and that no consciousness of the co- 
 existence of objects is possible which is not a consciousness of 
 the coexistence of such changing or changeable states in different 
 permanent substances. In other words, he does not directly 
 attempt to connect the idea of an object as known by us under 
 conditions of time, with the idea of permanence in change; but, 
 taking his stand on our consciousness of objective sequence and 
 coexistence, he endeavours to prove that such a consciousness 
 presupposes the consciousness of permanent substances, in 
 which, and by relation to which, such sequence and coexistence 
 are determined. And his language is in agreement with this view 
 when he speaks of this category as "put under the head of Eela- 
 tion, rather as being a condition of all the other relations -than 
 as itself being one of them.''^ The principle of substance is thus 
 proved with a double indirectness ; i.e., not as the principle 
 involved in the determination of an object as existing for us in 
 time, but as a principle presupposed in those other principles, 
 by which objects are determined for us as changing or co- 
 existing. 
 
 Now, one reason for this anomaly might be found in the ^ reason for 
 
 '' ^ thia treatment 
 
 fact that Kant was following Hume in his Inquiry. For in ^ay be^ound 
 that work, which was the only source of Kant's knowledge of statement of 
 
 the problem. 
 
 Hume,^ che idea that the only fact underlying the consciousness 
 of necessary relation is sequence of impressions, is applied only 
 to the causal principle, and not to the principles of identity (or 
 substance) and reciprocity. And, as I have already indicated, 
 half the plausibility of Hume's argument is due to this limita- 
 . tion. For, it is only the principle of causality which could 
 even be supposed to be generated from impressions which are 
 
 'A. 187; B. 230. 
 
 2 Or at least the only direct source. It has heen maintained that Kant's 
 view of Hume was partly derived from Beattie, whose Essay on the Natwre of 
 Truth was translated into German in 1772 (Vaihinger, I. 347). The know- 
 ledge to be got from this source would not, however, affect the point we are 
 here considering. 
 
536 THE CRITIQUE OF PURE REASON. book I. 
 
 always successive. Indeed, when Hume had once identified 
 the sequence of impressions with the consciousness of such 
 sequence, and had assumed that the consciousness of such, 
 sequence can be given apart from any other consciousness, he 
 could easily explain how custom should generate necessary 
 relation out of repeated consciousness of sequence. Now, Kant 
 does not, in this connexion, directly attack Hume's assumption 
 that the sequence of impressions involves a consciousness of 
 such sequence. But he points out, first, that such a conscious- 
 ness of inner sequence cannot explain the consciousness of the 
 permanent, which is implied in all determination of objects in 
 time ; any more than it can explain the consciousness of a co- 
 existence of objects, the impressions of which are for us neces- 
 sarily successive. Hence, secondly, it cannot explain even the 
 consciousness of objective sequence or change, which we can 
 have only on the presupposition of the former, and in contra- 
 distinction from the latter. "We cannot get the permanence, 
 coexistence, and succession of objects all out of the same sub- 
 jective sequence of impressions, even supposing we were to be 
 conscious of that sequence. And as these relations of things 
 as in time are known only in connexion with, and in distinction 
 from each other, we cannot get from it any one of them in the 
 sense required, unless the relations of our sequent impressions 
 are determined by principles which are not given in the 
 impressions themselves. 
 £°the"viewot ^^"^ there is another and more important reason for the 
 th™!(S™ c!" indirect way in which Kant deals with the principle of 
 causality, which, however, cannot be fully stated till we have 
 examined more closely the nature of his Deduction of it. In 
 that Deduction, Kant seeks to show that the determination of 
 things as permanent substances is the presupposition of all 
 determination of them in time, and especially that it is pre- 
 supposed in the determination of succession and change in 
 them. To understand this proof, however, we have to recall 
 the " metaphysical exposition " of time in the Aesthetic, and 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 537 
 
 especially Kant's assertion that time is not a general conception 
 which expresses the common element in many particular^, but 
 that, on the contrary, particular times are limitations of the 
 one time which is their presupposition. Thus all time- 
 relations, whether of coexistence or of succession, presuppose 
 the unity of time. N"ow, the Analogies are " principles for the 
 determination of the existence of phenomena in time according 
 to its three modes, — according to their relation to time itself as 
 a quantum (the quantum of existence, i.e., duration) ; according 
 to their relation in time, as a series of moments or times (one 
 after another) ; and again, according to their relation in time, 
 as having a certain breadth or compass in which all existence 
 is included (at once)." ^ But the first -of these relations is, 
 according to Kant, the basis and presupposition of the other 
 two; in other words, we cannot determine phenomena as 
 sequent or coexistent, except as we refer them as states to 
 objects which persist or endure throughout all time. To prove 
 this, we must consider, on the one hand, the idea of time and 
 the different relations which it makes possible ; and, on the 
 other hand, the conditions under which objects may be deter- 
 mined, according to their different possible relations. Now, 
 while, as we have seen, the consciousness of time, in its unity 
 or identity through all difference of times, is the presupposition 
 of the relation of coexistence and succession, yet time in its 
 unity is nothing but the presupposition of these relations. 
 Time, indeed, is the mere form of the relations of phenomena 
 which are determined as coexistent or successive, and therefore 
 it might be argued that we cannot perceive it by itself either 
 in its unity or in its differences. We cannot take time as a 
 permanent unity and date phenomena in their succession or 
 coexistence in relation to it, any more than we can take time 
 in its changing difference and, in relation to it, recognise 
 objects as permanent. But, in determining phenomena as suc- 
 cessive or coexistent, we presuppose the unity of time, and 
 ' Ala einem Inbegrifif alles Daseins (zugleich). 
 
538 THE CRITIQUE OP PURE REASON. book i. 
 
 therefore we are obliged to refer them to objects which are 
 permanent, and which as such " represent " the unity of time. 
 " Change can be apprehended only through the permanent and 
 as the alteration of it. For the distinction of the times in 
 which things are, can be perceived in them only as parts of one 
 and the same time. All change is only division of time; 
 therefore, there must be something which exists through the 
 whole time, because the whole lies always at the basis of the 
 division. Therefore, the substance is the substratum and the 
 changing is only the mode of its existence." ^ We can, in fact, 
 connect successive phenomena as in time, only as we determiae 
 them as states of a permanent object. Hence Kant declares 
 that " only the permanent changes," and that which has no 
 permanence — that which passes away absolutely and is not 
 contained in anything else which abides — cannot be determined 
 as changing ; for if anything were thus successive, its different 
 parts could not be connected as in one time. We might be 
 inclined to ask whether, on the same principle, we must not 
 say that " only that which changes is permanent " ; for it is 
 obvious that we cannot determine anything in relation to the 
 unity of time, except in so far as we also determine it in rela- 
 tion to the difference of times. But Kant does not put it in that 
 way ; because he conceives that, while we know particular times 
 as parts of one time, yet this does not mean that the one time 
 is presented to us first, and that then we determine particular 
 times by limitation of it ; but merely that, in determining 
 particular times in relation to each other, we presuppose the 
 unity of time. But, as we are not conscious of the difference 
 any more than of the unity of time by itself, this means that 
 we do not at first determine the object as permanent, and then 
 connect its successive phenomena as its changing states, but 
 that, in connecting phenomena with each other as successive 
 changing states, we presuppose the permanent identity of the 
 substance. This way of looking at the matter explains why 
 1 Erdmaim's NachtrtUje zu Kant's Kritik, § 79. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 539 
 
 the principle of substance is regarded by Kant merely as a pre- 
 supposition of the other analogies, and especially why Kant 
 starts, not with the idea of an object existing through time, 
 but with the idea of change, and why he proves the principle of 
 substance as well as that of causality as necessary to the know- 
 ledge of change. Change, as we shall find, is with Kant the 
 empirical fact, which is at the basis of all other empirical facts, 
 in such a sense that we might even say that the main work of 
 Kant was to make the consciousness of objective change intelli- 
 gible ; or, in other words, to explain how changing phenomena 
 could be brought into relation to the unity of thought and 
 subsumed under it. Hence, we find him, on the one hand, 
 insisting that change can never be other than an empirical fact 
 which belongs to the objects as given, and which is not capable 
 of being anticipated by pure thought ; for the pure unity of 
 thought — which, even when it becomes self-conscious, is con- 
 ceived by Kant to be purely analytic — could not, of course, give 
 rise to the idea of change. On the other hand, he does not, 
 indeed, insist upon the idea that change is the fundamental fact 
 in all experience ; but he constantly assumes that it is so, and 
 he directly states that motion is the central fact in all external 
 experience.^ And, if we may here anticipate the result of 
 
 ' In the Introduction to the Metaphysical Rudiments of Physics, E.. V. 317 ; 
 H. IV. 366. "The fundamental determination of anything that is to be au 
 object of external sense must be motion ; for by motion alone can that sense 
 be afifeoted. To this also the understanding reduces all the other predicates 
 of matter which belong to its nature. Hence natural science is nothing but a 
 pure or applied doctrine of motion." As, according to Kant's doctrine (especi- 
 ally as it is developed in the second edition of the Critique), inner sense is 
 simply a consciousness of the successive process of determination from within, 
 by which the determinations of sense from without are taken up, and by 
 which the consciousness of external objects is thus generated, we may say that 
 motion is the fundamental fact of experience. But this subject will be dis- 
 cussed in the next chapter. When Kant says that " by motion alone can 
 external sense be affected," he cannot be taken to mean that the thing in 
 itself moves and so stimulates our senses, (which would involve the absurdity 
 that the thing in itself which, ex hypothesi, we do not know, is known as 
 material substance,) but merely that external objects can be known only as 
 moving or movable substances. 
 
540 THE CEITIQTJE OF PURE REASON. book l. 
 
 Kant's whole discussion of the Analogies of Experience, we find 
 that the way in which he answers the above question as to the 
 possibility of change as a fact of experience, i.e., the possibility 
 of its being brought under the unity of thought, is to point out : 
 — first, that it must be conceived as the change of a permanent 
 object, which through all changes remains one with itself; but, 
 secondly, that all the successive changes of substances must 
 be conceived as necessarily connected ; and lastly, that .they 
 must be conceived as the reciprocally determined changes of 
 the parts of an unchangeable whole. In this way the facts of 
 the changing phenomenal world are, as it were, brought back 
 under the unity of thought, though not even thus perfectly 
 1/ the'tenn °' hamionised with its pure self-identity. For, it is to be observed 
 that what we have in the laws of substance, causality, and 
 reciprocity, is only " Analogies " of experience ; i.e., something 
 in experience which, though not identical, is parallel with the 
 relations of pure thought to itself shown in the judgments of 
 relation, as analysed by formal logic. The identity of the 
 subject with the predicate in the pure categorical judgment 
 thus appears, in relation to a changing matter given under con- 
 ditions of time, as the permanence of substance under the 
 change of its states ; the analytical connexion of reason and 
 consequent in the pure hypothetical judgment, when applied to 
 successive events or changes, appears as their necessary 
 sequence ; and the analytical unity of the genus with itself 
 in all its species, as the necessarily contemporaneous changes 
 of the parts of a permanent whole. Thus conceived, the 
 changing world of phenomena is subsumed under the schema- 
 tised unity-in-difPerence of pure thought, in which, even as 
 judgment, the difference is transparent. Or, to put it other- 
 wise, the world of objects in all its differences in time and 
 space, is conceived as an analogon of the pure unity of apper- 
 ception ; and is so made capable of being joined with the con- 
 sciousness of that unity, though it ca.n never be identified 
 therewith. 
 
CHAP. vili. THE ANALOGIES OP EXPERIENCE. 541 
 
 Before criticising this view, it is, however, necessary to refer J'^^J^^tion*^"^ 
 to a modiiication of it which Kant himself suggests in the ^efefo*"* 
 second edition of the Critique. Kant's argument is that, as introducing 
 
 thedetermina- 
 
 special times are limitations or determinations of the one time *'"? "' objects 
 
 ^ as 111 space. 
 
 which remains identical with itself through all the change of 
 times, so all the special determinations in which an object in 
 time manifests itself are to be regarded as modifications of an 
 object which remains one with itself through all their change. 
 The object in time must endure while it changes, because all 
 times are determined as successive in the one time ; and this, 
 as time is not perceived by itself, means that the connexion of 
 successive states must be determined as a succession in the one 
 substance. To this it may be objected that to say that "time 
 itself does not change,'' is like saying that passing away does 
 not itself pass away. So far the endurance of time and the 
 permanence of the changing might even seem to mean only 
 that the moments of time never cease to pass away, and the 
 changing never ceases to change. A perpetual flux would 
 therefore sufficiently " represent" all the permanence that is in 
 time.^ To try to extract, from the necessity of determining 
 change as in time, the idea that the " quantity of substance that 
 changes can never be increased or diminished," seems to be 
 absurd. For time has no quantity except continuance, and a 
 continuity of change would therefore seem to be all that is, on 
 this ground, required in the object. In the interval between 
 the first and the second edition, Kant seems to have become 
 aware of this, and to have come to the conclusion that time 
 itself can be represented only in relation to space, and that^ 
 therefore the principle of substaaee can be applied only to an 
 object which is not merely in time, but also in space. Thus, 
 in one of the manuscript notes on the first edition which have 
 been recently published, we find him declaring that " the proof 
 (of the principle of substance) must be so given that it applies 
 
 only to substances as phenomena of external sense. Conse- 
 
 % 
 
 1 For a perpetual flux never changes its relation to time. 
 
542 THE CRITIQUE OF PURE REASON. book i. 
 
 queiitly, it must be drawn from space, which, with its 
 determinations, exists at all times. Now, in space the only 
 possible change is motion,"^ i.e., a change in the relative 
 position of objects which remain otherwise unchanged. Un- 
 fortunately this new view of the matter is not introduced 
 into the deduction of the principle of substance in the second 
 edition, but only brought in, in certain general supplementary 
 remarks upon the principles of the pure understanding. Thus 
 Kant declares that " in order to supply something permanent 
 in perception which corresponds to the conception of substance, 
 we need a perception (of matter) in space ; for space alone is 
 determined as permanent, while time and all that is in inner 
 seiise is in constant flux." ^ Now, this would seem to involve 
 that in treating permanence and coexistence as well as sequence 
 as modes of time, Kant was referring to things as in time 
 determinations which they have only as being also in space. 
 Time and space by this reciprocal qualification give, as it were, 
 new dimensions to each other ; for it is the relation of time to 
 space which, in the first instance at least, enables us to think 
 of coexistence as a mode of time ; time in itself being merely 
 a relation of transitory movements as after each other, until we 
 regard what Newton calls the " ubiquitas teniporis" \^ while it 
 
 lErdmann, NacMrage zu Kant's Kritih, § SO ; of. §§ 77, 79, 81. Erdmann 
 suggests that the reason why the Derluction was not remoulded, was that this 
 would have involved also a complete remoulding of the chapter on the 
 Schematism, in which the categories were schematised only in relation to 
 time. In fact, in the first edition, inner and outer experience seem often to be 
 taken as oo-ordiuate, and, therefore, the principles of pure understanding have 
 to be stated generally, and in a way that makes them applicable to both. In 
 the second edition, inner experience is regarded as dependent upon outer 
 experience, or as arising in the rotUrn upon self from it ; and, as a conse- 
 quence, outer experience takes the place of experience in general. If this 
 thought had been fully worked out, the "Metaphysical Rudiments of Physios " 
 would have been absorbed into the Critique. This subject has been referred to 
 above in the chapter on the Deduction of the Categories and will be more fully 
 considered in the next chapter. 
 
 aB. 291. 
 
 3 Kant's Dissertation, § 14 : R. I. 31fl ; H. II. 408. " Though time is of one 
 dimension, yet the ubiquity of time (if I may use an expression of Newton's), 
 
CHAP. viii. . THE ANALOGIES OP EXPERIENCE. 543 
 
 is the relation of time to space which enables us to think of 
 space (in itself merely a relation of parts which exist together 
 with each other) as enduring through different times. Thus, it 
 is as schematised in reference to space as well as to time, that 
 the self-identity of the object, (which remains one with itself 
 through all its differences, in correspondence with the unity for 
 which it is,) becomes translated into the maintenance of its 
 quantum or the multitude of its external parts, which makes it 
 capable of occupying space. And in this connexion, Kant is 
 led to what seems a direct contradiction of his assertion that 
 the existent in time must be permanent. 
 
 "What essentially characterises the substance, which is possible o^irf^o? 
 only in space, and under spatial conditions, and, therefore, only as can be 
 
 the object oi external sense,is that its quantum cannot be increased ^"^"tenTe 
 or diminished without substance coming into existence or being 
 annihilated ; for the quantity of an object, which is possible 
 only in space, must consist of parts which are external to each 
 other, and these, therefore, if they are real (something moveable) 
 must necessarily be substances. On the other hand, that which 
 is regarded as an object of inner sense may, as substance, have 
 a quantum which does not consist of parts outside of each 
 other, and its parts are, therefore, not substances, nor should 
 their coming into existence or their going out of existence be 
 regarded as the creation or annihilation of substance. Their 
 increase or diminution is, therefore, possible without prejudice 
 to the principle of the permanence of substance." ^ 
 
 by which all things sensible are nomewhere, adds to the quantity of real things 
 another dimension, in so far as they, as it were, hang upon one moment of 
 time. For if you picture time as a line produced in infinitum, and coexistents 
 by lines applied at right angles in any point of time, the superficies which is 
 thus generated will represent the Muridus Phaenomenon both in its substance 
 and its accidents." 
 
 1 Metapk. AnfangsgrHnde, ch. III. ; R. V. 405 ; H. IV. 437. In the next 
 chapter we shall have to deal with Kant's argument against Idealism, in which 
 he contends that inner experience derives its time-determination by reference, 
 not to a substance underlying its changes, but to the external siihstantia 
 pliaenomtiion in space. 
 
544 
 
 THE CRITIQUE OF PURE REASON. 
 
 BOOK I. 
 
 thTdfductiou '^^^^ proof of the principle of substance in relation to changing 
 Principle of objects determined in space would, as appears from the indica- 
 which this tions which Kant has furnished, have followed the same course 
 
 would 
 
 necessitate, f^^ (;j|a,t which is now given in relation to time. It would have 
 rested on the idea that any change other than the motion of 
 the parts of matter relatively to each other, any change which 
 involved the diminution or increase of the substance moved, 
 would make the consciousness of such motion impossible. For, 
 as we do not perceive space itself, we cannot determine changes 
 of place in it except as movements in a substance which repre- 
 sents space, i.e., which is as unchangeable as space itself; in 
 other words, we must determine them as changes of the position 
 of the parts of a substance which, as a whole, remains unchange- 
 able. • This is well expressed in another manuscript note of 
 Kant. " If substance endures while accidents change, but 
 substance apart from accidents is the empty abstraction of the 
 substantial, what is it that endures ? All that in experience 
 can be distinguished from the changing is the quantity" (i.e., 
 of matter), " and that can be estimated only by the greatness of 
 the merely relative effect " (of different matters) " in similar 
 external relations. The principle of substance, therefore, can 
 be applied only to material bodies." ^ The last words mean 
 that the mass or quantity of a substance in any material object 
 can be estimated, only by considering its relative effect in pro- 
 ducing motion as compared with other bodies moving with the 
 same velocity. Of course, no such estimate can be applied to 
 the whole world, but only to the parts of substance in relation' 
 to each other. Hence the assertion that the quantum of 
 substance in nature is capable neither of increase nor of 
 diminution, does not mean that we know the world in space 
 as a definite and limited guantuvi, of which each particular 
 substance is a part. We have no experience of the world as a 
 whole, or of its relation to empty space ; and, indeed, the 
 attempt to determine the world as either limited or unlimited, 
 ' Erdmann, Ncichtrage, % 81. 
 
CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 545 
 
 gives rise to an antinomy, which Kant solves only by showing 
 that, as we can never complete the synthesis of objects deter- 
 mined as in space, we can never represent the world as a whole. 
 The only meaning, therefore, which can be given to the assertion 
 that the quantum of substance in nature is constant, is that in 
 the relations of different substances to each other as moving, 
 there is something which presupposes a constant relation of all 
 substances to the space they occupy. In other words, we can- 
 not determine them as changing place in relation to each 
 other, unless we presuppose that there is no change in their 
 character as occupying space. And, therefore, if we say with 
 Kant that they occupy space in virtue of the repulsive and 
 attractive forces they exert towards each other, we must say that 
 relatively to each other, the parts of matter are determined by 
 forces which vary in inverse ratio to some function of their 
 nearness to, or distance from each other, according to an 
 .unchangeable law. 
 
 As to this general line of proof, we may remark that it The Principle 
 
 ° r ' J of Substance 
 
 starts with the conceptions of the Aesthetic, according to which j^^reiy ^^he 
 space and time are represented as infinite given wholes, and all won^ThT 
 
 . . other 
 
 particular spaces and times are reached by the determination Analogies, 
 of these given wholes : but it immediately corrects this repre- 
 sentation, by pointing out that these infinite wholes are mere 
 forms of relation, which are not apprehended for themselves, 
 but only as the presupposition of the determination of phe- 
 nomena in space and time in relation to each other. For, 
 while we cannot directly relate phenomena to space and time, 
 it is only on the presupposition of the permanence of their 
 relation to the time and space which they occupy, that we can 
 relate them to each other. As all times are in one time, which, 
 however, is not perceived as one, but is simply a permanent 
 form of relation ; so, Kant argues, all changes are in one sub- 
 stance, which does not pass away, since, if it did, it could not 
 be determined even as passing away. And as all spaces are in 
 one spac8, of the continuity of which they are determinations, 
 VOL. I. 2 m 
 
546 THE CRITIQUE OF PURE REASON. book i. 
 
 and as space also is not perceived, but is simply a permanent 
 form of relation: so all determination of objects as changing 
 their relations in space, i.e., as moving, presupposes a permanent 
 substance in which such relative changes take place. But (since 
 all our knowledge of external objects is based on their change 
 or motion) this means that all our determinations of objects as 
 in space and time, are determinations of their states as changing 
 or changeable in relation to each other ; determinations which 
 would be impossible unless we presupposed a substance which 
 remains identical with itself through all the change of its parts. 
 Hence, we cannot determine phenomena in relation to each 
 other as in space and time, unless we regard each phenomenon 
 as a changeable state of a permanent substance. Only as so 
 determined can a phenomenon form a term in a relation of 
 succession or coexistence which can be referred to objects. 
 This, indeed, looks like saying that the phenomena must be 
 presupposed as objective in the very acts by which they are 
 referred to objects. But this apparent circle really means that 
 the reference of a phenomenon to a permanent object, and its 
 connexion with other phenomena according to laws which 
 determine its relations to them in time and space, are different 
 aspects of the same process which are only ideally separable. 
 Or, to put it in a slightly different way, it is only as they 
 relatively determine each other's place and time according to 
 universal laws, that they are referred to a world of objects 
 which remains one with itself through all the differences of 
 spaces and times ; and, conversely, it is only on the presup- 
 position of such unchangeable identity of the whole, that the 
 relative determination of the parts becomes possible. Kant's 
 expression of this is somewhat obscured by the way in which 
 he separates the proofs of the different Analogies, showing first 
 that change implies permanence ; then, that change can only be 
 known as determined by previous change ; and, finally, that the 
 succession of states in different substances implies a reciprocal 
 determination of each by the others. But he partly corrects 
 
CHAP. Vlll. THE ANALOGIES OF EXPERIENCE. 547 
 
 tbis imperfection of statement by. treating the first analogy as 
 the mere presiipposition of the second; and, as we shall see, he 
 shows also, though less distinctly, that the second may be re- 
 garded simply as the presupposition of the third. For each 
 partial substance can be defined only as a " complex of mere 
 relations " ^ to the others, and the phenomenal identity of the 
 whole is known only as the necessary presupposition of all 
 relative determination of the parts by each other. 
 ' The relation of the principle of substance to the other ^t ^e 
 
 substance is 
 
 analogies as merely their presupposition, is expressed most S.o*m?te''^^''' 
 definitely in the following passage : — accidents. 
 
 " The determinations of a substance, which are nothing but 
 its special modes of existence, are called its accidents. They 
 are always real, because they relate to the existence of the 
 substance. (Negations are only determinations which express 
 the non-existence of some determination in the substance.) 
 Now, when we attribute to such real or positive determinations 
 a special existence in the substance (as when we regard motion 
 as an accident of matter), we call this existence inherence, in 
 flistinction from the existence of the substance, which we call 
 sithistence. But this way of speaking gives rise to many mis- 
 conceptions, and it would be at once more definite and more 
 accarate to say that the accident is only the mode in which 
 the existence of a substance is positively determined. Yet we 
 must acknowledge that, by the conditions of the logical use of 
 our understanding, we are constrained, as it were, to isolate 
 that in the existence of a substance which can change while 
 the substance remains, and to set it, as an accident, over 
 against that in the substance which is permanent and funda- 
 mental; hence also we put this category under the head of 
 Relation, rather as being the condition of all relations than as 
 
 ' ° This limita- 
 
 being itself one of them." ^ sUn^e'Snt 
 
 In this paragraph, Kant emphasises the truth that it is nouo^th?iaws 
 
 , . , . _ of thought, but 
 
 through the accidents and their relations that the substance tothesensuous 
 
 ° ■' conditions of 
 
 *i A. 265 ; B. 321. , ' A. 186 ; B. 229. ledge."'"'" 
 
548 THE CEITIQTJE OF PURE REASON. book I. 
 
 exists for us. The substance, in fact, is nothing in itself apart 
 from its accidents and their relations. Yet we iind ourselves 
 obliged " by the conditions of the logical use of our under- 
 standing " to distinguish two aspects of the real, the permanent 
 and the changeable, and to refer the latter to the former as 
 predicate to subject. And Kant fears that this duality in our 
 determination of objects may easily lead to the idea of a deter- 
 mination of the object in itself, as a substance apart from the 
 determination of its accidents and their relations. It is, how- 
 ever, only in reference to the determination of the reciprocal 
 relations of the phenomena which we call the accidents of 
 substances, that the determination of them as substances has 
 any value. Now, Kant holds this to be just one of the cases 
 in which the determination of things in our experience does 
 not conform to their determination in pure thought. " In an 
 object of pure understanding, only that constitutes its inner being 
 which has no reference (as regards its existence) to anything 
 different from itself. But contrariwise, the inner determina- 
 tions of a substantia phenomenon in space are nothing but 
 relations, and it is itself nothing but a complex of mere relations. 
 yor we know it only by forces which are active in it, either to 
 attract other substances or to prevent other substances from 
 penetrating it." ^ In other words, Kant opposes the merely re- 
 lative determination of phenomenal substances, to that absolute 
 determination of them which is demanded by pure thought; 
 and he finds in the form of space, under which phenomena are 
 given to us, the reason why thought can determine objects only 
 in relation to each other, so that the permanence which we 
 attribute to them as substances, is only the permanence of 
 certain relations. Conversely, if we could determine objects in 
 pure thought, (which would be necessary if we were to deter- 
 mine them as things in themselves,) we should determine them 
 purely by predicates " which involve no reference of them as 
 regards their existence to anything different from themselves." 
 ' A. 265 ; B. 321. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 549 
 
 Here, therefore, we have another reappearance of the analytic 
 idea of thought in contrast with the synthetic idea of knowledge, 
 and further the condemnation of the latter as phenomenal, just 
 because it is synthetic. Kant thinks that we do not know ' 
 objects as things in themselves, because we do not find an ^ 
 inner being in them, apart from their actual determination in 
 relation to other things. And this means that it is only the 
 connexion of phenomena, i.e. of our ideas, with each, other 
 according to necessary laws, which makes us refer them to 
 permanent objects distinct from our perceptions. We have no 
 consciousness of the object as a permanent substance apart 
 from our consciousness of the causal, or reciprocally causal, 
 ooimexions of phenomena. And it is just this fact which we 
 discover when we ask what is meant by the existence of an 
 object for us, i.e., when transcendental reflexion forces us, in the 
 deterniination of objects, to take account of the self for which 
 they are. In other words, when reflecting on the conditions 
 under which permanent objects come to exist for us, we see 
 that they can do so only as the phenomena referred to them 
 form part of a context of experience determined by universal 
 laws, and that the " object itself," the " transcendental object," 
 is nothing apart from that context- — though, prior to the trans- 
 cendental reflexion, we naturally regard it as something which 
 exists in itself, and is revealed to us as so existing.^ If, 
 however, it were so revealed to us, we should be able to 
 determine it in itself, by predicates which did not involve 
 its relations to other things and their predicates. But, justl 
 because we know things only in their relations to each other, ^ 
 we must say that the things we know are only phenomenal, 
 i only existences for the self, and not real things independent 
 of it.^ 
 
 ' A. 104. 
 
 ^B. 67. " Now, through mere relations a thing in itself is not known ; 
 wherefore we may fairly judge that, as by our external sense nothing but ideas 
 of relation are given to us, that sense cannot contain anything but the relation 
 of an object to the subject, and not any inner determinations which belong to 
 
550 THE CRITIQUE OF PURE REASON. book i. 
 
 ^|feotsonhi3 jq-Q^^ i^Qy^ should all this be interpreted from the point of 
 SSofthTught view which has been taken in our previous criticism of Kant, 
 analytic. i.e., if we reject his view of the analytic movement of thought, 
 and with it the conception of the regress on self-consciousness 
 as being a regress upon that which is more abstract and simple 
 than the experience it explains ? The ordinary consciousness, 
 no doubt, takes objects as things in themselves, and that in two 
 senses. It takes them as things which are determined in 
 themselves, apart from relations to other things and from the 
 changing predicates which they receive from such relations; 
 and it also takes them as things which have their determina- 
 tion apart altogether from their relation to the conscious self. 
 Hence to it the demonstration, that the object has no deter- 
 mination for us apart from its relations to other things, and 
 that these relations are continually changing, would seem to be 
 
 the object in itself." A. 284 ; B. 240. " Through mere conceptions, it is trne, 
 I cannot think of anything external without something internal, just because 
 conceptions of relations necessarily presuppose given things, and are not 
 possible without them. But in our perception of objects something is con- 
 tained which does not lie in the mere conception of a thing, and it is this 
 which supplies the substratum which could not be known through mere con- 
 ceptions, to wit, a space which, with all it contains, consists of mere relations 
 — formal, or it may be also real, relations. It is, therefore, impossible for me 
 to argue that, because nothing can be represented by mere conceptions without 
 something absolutely internal, the same must be the case with the things 
 which fall under such conceptions, and that in the perception of these things 
 there can be nothing external without something absolutely internal beneath 
 it.'' Here we have, on the one hand, the necessity of pure thought, which 
 arises out of the nature of judgment in so far as any object is supposed to be 
 determined by it, i.e. the necessity of separating the object of thought from its 
 predicates and regarding it as something simple and primary, something deter- 
 mined in itself, to which they are attached ; and, on the other hand, the idea 
 that as perception with us takes the place of the subject in the judgment of 
 knowledge, and as the form of perception is essentially a form of relation between 
 elements which are diflferent and external to each other, so our object can only 
 be a substratum or permanent possibility of certain relations. But, it might 
 be asked, how should pure thought take the form of judgment, and so give 
 rise to the duality of subject and predicate, which, again, interpreted ontologi- 
 eally, passes into the duality of substance and accident here referred to? 
 This, as we have seen, is a difficulty for which Kant's view of pure thought 
 provides no solution (see above, p. 451). No answer can be given, except that 
 pure thought is taken by Kant as already judgment, which it could not be if it 
 were not in some sense synthetic. 
 
CHAP. vill. THE ANALOGIES OF EXPERIENCE. 551 
 
 the denial that we have any real knowledge of the object at all. 
 Now, Kant at once vindicates and gives a new sense to the 
 ordinary consciousness of phenomena as objective, when he 
 shows that the permanence attributed to them as objects is not 
 the permanence of things as determined by predicates which 
 are independent of their relations, but the permanence of the 
 laws of relation which maintain themselves in all the changes 
 of the special determinations of objects ; a permanence which 
 itself presupposes the permanence of the objects related, though 
 only as terms in such relations. But, while Kant thus 
 advances to the idea of objects as determined only in relation 
 to each other, or as subjects of relations, he treats the demand 
 of the ordinary consciousness for a non-relative determination 
 of things as, from the point of view of thought, a legitimate ; 
 demand, a demand that arises from the very nature of the 
 intelligence ; and he regards the fact that we are confined to a 
 relative determination of things as a result of the mediation of 
 , our knowledge by sense and its forms. Because objects can 
 exist for us only as we combine or relate to each other the per- 
 ceptions or phenomena through which we know them, it is 
 maintained that we cannot know them as they are: though the 
 funiversality of the forms of synthesis explains why we suppose 
 that we do know them as they are, independently of our per- 
 ceptions. Hence, the relativity of the objects to each other is 
 immediately connected with their relativity to the subject, and 
 this relativity to the subject is opposed to their reality as things 
 in themselves. 
 
 Now, it is undoubtedly true that our consciousness of objects But tte idea 
 
 ' ^ '^otSL tning m 
 
 cannot be logically separated from a consciousness of their rela- £*„„ Sions 
 tions ; and that a consciousness of the permanence of the abstract 
 
 nature of our 
 
 obiects in the change of their relations, cannot logically be first oonscious- 
 
 •' o JO,/ ness of things. 
 
 separated from a consciousness of the permanence of the 
 laws of relation in all the change of the determinations of 
 the objects. But, to suppose that the recognition of this 
 relativity of the objects of our knowledge is. equivalent 
 
552 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 to the denial of the real objectivity of the things so 
 determined, would imply that we have some idea of real objects 
 or things in themselves as divested of such relations. It would 
 imply that we can at least tUinh, if we cannot know, an object 
 absolutely out of all relation. But where are we to get such 
 an idea ? ^ That we ever think we have it, is a result of the 
 imperfection of that self-consciousness which goes along with 
 our first consciousness of objects. But, in fact, so far .is it 
 from being true that we can or must think any object as in itself 
 unrelated, that on the contrary we are unable to think of it 
 except as related. And our recognition of relations as im- 
 plied in our thought of an object is, therefore, merely a step 
 to correct that thought by completing it. Again, it is equally 
 true that the consciousness of the object as essentially related 
 to other objects, — or, in other words, the consciousness that it is 
 knowable only as its phenomena are connected in one context 
 of experience with the phenomena of all other objects, — implies 
 and leads to the consciousness that it is essentially an existence 
 for a self. But, when Kant regards this recognition of the 
 relativity of objects to the self as reducing them to mere 
 phenomena, behind which things in themselves are impenetrably 
 concealed, we are obliged to point out that the relation of 
 objects to the self, — the very relation implied in speaking of 
 things as objects, — cannot involve the negation of their claim 
 to be objects in distinction from the subject. This Kant him- 
 self shows us, when he points out that the consciousness of self 
 (is possible only as the self is distinguished from, and related to, 
 the object. For we need only to remember that a self, strictly 
 speaking, does not exist except as self-conscious, in order to see 
 that a denial of the reality of the object is also a denial of the 
 reality of the self for which it is. To say that the permanent . 
 object exists merely in the determination of the connexion of 
 
 ' Thought, according to Kant himself, is essentially judgment ; and, as such, 
 it is at once the reference of an object to itself and its relative determination 
 by a predicate which distinguishes it from, and connects it witb, other objects. 
 
CHAP. vnr. THE ANALOGIES OF EXPERIENCE. 553 
 
 our experiences by universal laws of causality and reciprocity, 
 may convey a truth, if it teaches us not to isolate one element 
 in the determination of the object from another, or the deter- 
 mination of the object from its relation to the self. But it is 
 misleading, in so far as it suggests the idea that the object is a 
 determination of a merely subjective consciousness which knows 
 only its own states. For, if we can have no consciousness of 
 the self except in distinction from and relation to the object, 
 we may, indeed, say that the consciousness of self includes the 
 consciousness of the object and goes beyond it, but we are not at 
 liberty to speak of the latter as a mere state of the former. 
 Otherwise we fall into the old confusion of the return upon the 
 self, which implies and presupposes its distinction from the 
 object, with a regress upon a subject for which there is as yet 
 no object, and which therefore is not for itself a subject of 
 knowledge, i.e., is not, properly speaking, a self at all. 
 
 The result, then, is that Kant is right in saying that the But, instead 
 
 ' ' o ■' ° of going back 
 
 recognition of the permanent identity of the object through all ™b*te,nT™^ 
 jits changes is necessary, as the first determination of an object in ?Sio™we 
 
 j ^ should correct 
 
 irelation to, and in distinction from, which we can be conscious "'^'^ first oon- 
 
 ' ception by 
 
 of self. But, as it is in relation to the subject that objects in thi'mSf °"* 
 space and time are determined as objects, we must, in a relation, 
 sense, regard their permanence as phenomenal ; i.e., we must 
 regard the permanence of matter in its occupation of space as 
 not belonging to it, so to speak, in its own right, but as the 
 necessary attribute of the object of consciousness as such. But 
 as the self-externality of the object is not its externality to 
 consciousness, and its occupation of space is not exclusive of 
 consciousness, so when we realise the relation of the object to 
 consciousness, we see that those attributes given to the object 
 in abstraction from the subject, must be reinterpreted by that 
 relation. In this sense, we may say that the permanence of 
 the material world is the analogon of the self-identity of thought, 
 which in every object presented to it must seek and find its 
 own product or manifestation. The defect, however, in the 
 
554 THE CRITIQUE OF PURE REASON. book i. 
 
 scientific conception of the objective world, as a permanent 
 matter which in all its relative motions never ceases to 
 maintain its occupation of space, lies in this, that it takes the 
 object in its externality and difference as ultimate, and only so 
 far makes it intelligible as it represents the same identical 
 whole as maintaining itself in all changes of its parts. But, in 
 this there is a double contradiction ; for (1) that is taken as a 
 whole which cannot be conceived as a whole, because no prin- 
 ciple of unity in its difference is seen ; and therefore also, (2) 
 the relative change of the parts is taken as something which is 
 inexplicable from the identity of the whole, which it presup- 
 poses. In other words, matter, conceived as a thing in itself, 
 wants at once a principle from which the differences of the 
 object could be conceived as flowing, and a principle to which 
 they could be conceived as returning. The same difficulty, 
 mutatis mutandis, is to be found in the category of substance 
 itself, in so far as the application of that category implies the 
 reference of phenomena as transitory accidents to a permanent 
 substratum, which yet has no determination apart from these 
 transitory accidents which are its correlate. While, therefore, 
 that category carries our view beyond the transitory, and while 
 in the substance we suppose ourselves to find something more-than 
 1 in the accidents, yet we do find in it only a unity which turns into 
 the totality of the accidents. This dialectic we find in the system 
 of Spinoza, whose substance, in which all finite parts are lost, 
 immediately converts itself into a sum of finite parts or modes. 
 Kant's use of the category does not betray the same contradic- 
 tion so clearly ; for while he tells us that " only the permanent 
 changes," he does not go on to say that only the changing is 
 permanent, but takes change as the fact from which we may 
 go back to the permanent. And the necessary correction of 
 the contradiction involved in the category with which he begins, 
 appears only in the addition of other categories to it. We 
 can, however, understand the progress of his thought only if 
 we regard the first of the Analogies as an imperfect expression 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 555 
 
 of a truth which is restated in a corrected form in the second 
 and third Analogies ; and, as we have seen, Kant himself indi- 
 cates this, when he speaks of the iirst Analogy as merely the 
 presupposition of the others.^ This result, however, we shall 
 be able to reach in a satisfactory way only when we have con- 
 sidered Kant's method of proving these Analogies. 
 
 The second Analogy, in the first edition of the Gritigue, is ^ca^^Sity as 
 entitled " The Principle of Production," and the formula in flrat^nd 
 
 Becond 
 
 which it is expressed is : — " All that happens (begins to be) editions, 
 presupposes something upon which it follows according to a 
 rule." In the second edition the title givefn is, " The Principle 
 of'iEime-sequence according to the Law of Causality," and the 
 
 • :• formula given is: — "All changes take place according to the 
 law of the connexion of cause and effect." The second formula 
 has the advantage of separating the category from the fact to 
 which it is applied, and bringing, therefore, the idea of the 
 subsumption of the latter under the former more distinctly 
 , before us. Kant hegins his proof by calling our attention to 
 
 t s the fact that the previous principle has' already excluded the 
 
 ; , ■ idea of absolute change, i.e., of any change which involves the 
 extinction of one substance, and the beginning to be of another. 
 Changes can be known only as alterations in the special deter- 
 minations or accidents of a permanent substance. In fact, they 
 ean, according to Kant's principles, only be relative alterations 
 of the parts of a whole which is presupposed to be unchange- \ 
 able — as unchangeable as the self for which it is.^ Presupposing 
 then such unchangeable reality, Kant goes on to argue that the 
 successive phases of it must be determined by the law of 
 
 ,»»» eausality, on the ground that a change can be known as such 
 only if it be connected according to the law of causality with 
 a previous change. 
 
 ' Also when he speaks of the third Category in each class as including the 
 other two. (B. 111.) 
 
 ' The self is unchangeable as the subject of knowledge. This does not, of 
 course, affect the question of the permanence of the soul — ^^a question which Kant 
 afterwards discusses in the chapter on the Paralogisms of Rational Psychology. 
 
556 THE CRITIQUE OP PURE REASON. book I. 
 
 Relation of The Statement of the question in this way, as we have seen, 
 
 this statement 
 
 *?gHume's already involves a partial answer to Hume ; as it brings to light 
 the distinction between the consciousness of a real or objective 
 succession, and the consciousness of our perceptions as succes- 
 sive — the former being a consciousness of a succession in the 
 permanent object, the latter merely of the successive synthesis 
 of apprehension involved in all consciousness of the object 
 whatsoever, whether the manifold of the object be determined 
 as coexistent or successive. In both cases equally there is a 
 subjective succession of ideas, and a synthesis of these ideas in 
 which they are successively taken up and combined in the con- 
 sciousness of an object. Now, even if we could suppose that 
 we were conscious of such successive synthesis apart from any 
 determination of an object by it, it would not be of itself the 
 consciousness of a succession in the objects. This we see if we 
 consider, (1) that succession in the object is only one of the 
 determinations of the object, and that it must be distinguished 
 from all the other determinations of it, though these equally 
 imply a successive synthesis ; and (2) that; it immediately pre- 
 supposes the determination of the objects as permanent, a ■ 
 determination which could not be transferred to it from the 
 synthesis of apprehension, which is always successive, 
 perceptions"^ Now, Kant goes ou, there can be no doubt that we do make 
 obfect^e'^'^ the distinction between real succession and coexistence. We 
 
 aiiccesaion any • i i . • i • /» i 
 
 more than it recognisc that the succession in our apprehension of the mani- 
 
 yields " '■ ^ 
 
 existence or ^'~'^^ °^ ^ ®^^P Aoatiug dowu a stream, corresponds to a succession 
 in the object ; while we as certainly recognise that the succession 
 in the apprehension of the manifold of a house, does not corre- 
 spond to any such succession. If, however, the determination 
 of the succession in the object were derived simply from the 
 I consciousness of the succession of our perceptions, such a 
 distinction would be impossible. Either the manifold would 
 in all cases be combined as successive in the object, or it 
 would not be referred to the object at all : i.e., we should be 
 conscious merely of a succession in our apprehensions. To 
 
 permanence. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 557 
 
 generalise. Hume's theory is, therefore, fatal to it ; for he 
 seeks to derive one particular determination of objects from a 
 consciousness of that succession of our apprehensions which is 
 involved in all determination of objects. On the other hand, 
 if in Hume's method we are to derive from the successive 
 synthesis of apprehension the determination of objects as per- 
 manent, as changing, and as coexistent, we must get them all 
 equally out of the subjective succession of our states ; and we 
 must be able to show how an association of perceptions, that 
 we are conscious of only as continually changing, can generate 
 a consciousness of permanent objects, which (1) do not pass 
 away with the perceptions, but (2) pass through different 
 states, which states again are (3) determined both as successive 
 to, and as contemporaneous with certain other states previously 
 presented to us. Now, for Hume as for Kant, the way to 
 objectivity lay through necessity ; for both had given up the 
 idea that we are directly conscious of things in themselves and 
 can from them read off their qualities and relations. What we 
 are immediately conscious of is only, as Hume contends, the 
 succession of perceptions which change upon us every moment. 
 Hence, it is only in so far as we come to recognise between 
 such perceptions a necessity of relation, which determines their 
 succession in our consciousness and is not determined by it, 
 that we refer them to objects. But Hume regards this neces- 
 sity as itself generated by custom, which establishes an in- 
 separable connexion between perceptions that have frequently 
 followed each other. "When such connexion has been estab- 
 lished, so that the presence of the one perception necessarily 
 brings the idea of the other as its consequent, we refer the 
 connexion to the object and regard our perceptions as corre- 
 sponding to something which is independent of them. Now, 
 Kant goes along with Hume, in so far as he holds that the 
 determination of a sequence as objective means the determina- 
 tion of it as connected with a perception which precedes it 
 accordiag to a universal rule, — means, in short, that the con- 
 
558 THE CRITIQUE OF PURE REASON. book i. 
 
 sequent is necessarily connected with the antecedent. But he 
 rejects Hume's explanation of the way in which their connexion 
 is established for us ; and he proves the inadequacy of that 
 explanation by showing that it will not bear the weight which 
 must be thrown upon it, if it is to be used as a general 
 explanation of objectivity. It looks plausible to say that a 
 necessary connexion of perceptions as successive can be estab- 
 lished between them when they are repeatedly given as succes- 
 sive, and that when this consciousness of them as necessarily 
 sequent is established, we regard them as objectively sequent. 
 But, as the perceptions are always successive, it is obvious that 
 causality or necessary sequence, on this hypothesis, should be 
 the only principle of objective connexion, and that we should 
 ' , never connect objects or their states as coexistent : nor, again, 
 should we refer perceptions to objects which are permanent while 
 the perceptions change. The plausibility of Hume's whole explan- 
 ation lay in the fact that, while he appears to get objectivity out 
 of necessity, he really assumes, to begin with, that our successive 
 perceptions are perceptions of changes in permanent objects, and 
 not merely successive perceptions, of which one disappears 
 when the other comes. Such perception, (even if we admit the 
 possibility of the genesis of a consciousness of necessity of suc- 
 cession out of a consciousness of repeated succession) could 
 produce only the consciousness of a connexion of different 
 J objects or perceptions, as necessarily presented in succession. 
 But if Hume had assumed, to begin with, the consciousness of 
 changes in o,^ects, he had already assumed that very objective 
 connexion of sequence which he had to explain as an in- 
 separable association generated by custom. 
 Objects are Kaut's proof , then, begins by the rejection of any explanation 
 
 not perceived ^ .j u « .. j. 
 
 chang?ng °^ ^^^ consciousncss of real or objective sequence which assumes 
 tSghthe that changing objects are perceived by us. The object exists 
 
 synthesis of ,,^ • t n i • t 
 
 imagination for US Only throuoh our own ideas and %n them : we may 
 
 and the recog- 
 
 principfe"" ^^®^ ^^^ ^^^^ ^^ ^^ °^'^'^ ^^ i&QSi. But, nevcrthcless, it is quite 
 true that " our thought of the reference of knowledge to its 
 
CHAP. vill. THE ANALOGIES OF EXPERIENCE. 559 
 
 |)bject carries with it something of necessity ; for the object is 
 regarded as that which hinders the elements of our knowledge 
 of it from coming upon us peU mell and at haphazard, and 
 causes them to be determined a priori in certain ways. For, 
 just in so far as our ideas are to refer to an object, they must 
 necessarily agree with each other in reference to it, i.e., they must 
 have that unity which constitutes the conception of an object." ^ 
 Now, when we are asking for the origin of this necessity, this 
 constraint upon our ideas, it is no answer to say that we deter- 
 mine the object so, because it is so determined in itself. With 
 the thing in itself we have nothing to do ; nor can it constrain 
 us to combine the manifold of our perceptions in any particular 
 way. The thing in itself is, no doubt, the ground to which we 
 refer the affections of sense, " in order that we may have some- 
 thing which corresponds to the sensibility as a receptivity." ^ 
 But these affections become part of a consciousness of objects, 
 not as they are given, but in virtue of a synthesis of imagina- 
 tion, which takes up the given matter so as to combine it into 
 images. Now, no doubt, this process is successive ; but it does 
 not explain a consciousness of succession in the object so 
 imaged or apprehended, any more than it explains a conscious- 
 ness of its permanence, or of the coexistence of its parts with 
 each other. In all cases, therefore, we have to look for a 
 principle that determines the mind to put the manifold 
 together in a particular way in the objects, a principle which 
 is not given in the succession in which the elements of per- 
 ception are taken up by imagination ; and this principle is not 
 less required in the case where the states of objects are deter- 
 mined as successive, in correspondence with that succession, 
 than in the case where they are determined as coexistent and 
 therefore not in correspondence therewith. For, "we take that, 
 which lies in our successive apprehension to be mere ideas, 
 while we regard the phenomenon which is given to us through 
 them as the object of these ideas, with which the conception 
 * lA. 104. ! A. 494; B. 522. 
 
560 THE CRITIQUE OF PURE REASON. ' book i. 
 
 we draw from the ideas of apprehension is required to agree : 
 though in truth the object in question is nothing but those very- 
 ideas as a complex unity." ^ Now, Kant goes on to say, this 
 twofold consciousness of our ideas, as successive states of our 
 subjectivity, and as representing states of objects which may or 
 may not be successive, can be accounted for only in one way. 
 " A phenomenon can be set over against the ideas of appre- 
 hension as corresponding to them, and yet contrasted with them 
 as their object, only if it stands under a rule which separates 
 it from every other apprehension and which makes a special 
 way of connecting its manifold necessary." What, then, is the 
 rule to which we must subject the perceptions in this case, in 
 order that we may conceive them not merely as subjectively 
 successive (which they always are), but as representing a 
 succession of states in an object ? 
 and that N"ow, Kaut couteuds that no answer can be given to this 
 
 lirinciple is 
 
 causality. qucstion whicli does not involve the principle of causality. I 
 can say that a subjective succession of perceptions corresponds 
 to an objective succession of events, or, in other words, I can 
 determine the subjective succession in my apprehension as 
 representing a change in the states of an object, only in so far 
 as I presuppose or discover that, in the moment when the 
 change began, some phenomenon had come into existence which 
 is the " condition of a rule " according to which the second 
 perception must always follow upon the first. Or, in other 
 words, I regard the succession of perceptions as representing an 
 event or change in a substance, in so far as I regard such event 
 as referring back to some other event as its necessary correlate. 
 And, conversely, in so far as we do not refer back a succession 
 in our perceptions to an event or change in the previous time 
 as its necessary correlate, we do not regard it as itself repre- 
 senting an objective succession or event. In this latter case, 
 the "coming to be" of the new perception for me is not 
 regarded as the objective " coming to be " of the state of a 
 lA. 191; B. 236. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 561 
 
 substance whicli it represents ; i.e., I regard the sequence as 
 merely subjective, or as not representing an objective sequence, 
 though the synthesis of phenomena or sequents in my percep- 
 tion may represent some other objective relation. The new 
 perception always drives out the old in my apprehension ; but 
 I can fix the succession as necessary and objective only so far 
 as I regard it as determined in relation to a phenomenon in the 
 previous time, which may or may not actually have been 
 perceived, but which in any case I regard as making the order 
 of my perceptions necessary. And if I did not regard the 
 change of my perceptions as having such a necessary correlate, 
 I should not regard it as representing a change in an object. 
 
 Kant expresses this thought as follows :■ — " In all empirical Ig^/'jg ^j^g 
 knowledge there is a synthesis of the manifold by imagination, ™uIS'?J'a°'^ 
 
 and this synthesis is always successive." In other words, in in the percep- 
 tion of an 
 
 whatever way I determine the manifold so as to unite it in the ^™°* *= ™'=''- ■ 
 conception of an object, I must always take up one element of 
 it after another. But this, Kant goes on to say, does not 
 determine how I shall conceive of the order in the object, 
 which imagination might (as far as this sequence is concerned) 
 make what it pleases. For the elements which I put together 
 one after another, I may take as representing what stands in 
 the reverse order in the object. Suppose, however, that the 
 synthesis is a synthesis in which I apprehend an event or 
 objective succession, then "the order is determined in the object ; 
 or, to put it more accurately, there is in it an order of successive 
 synthesis which determines an object, so that something must 
 necessarily precede, and when this is posited, something else 
 must necessarily follow. The perception of an event, therefore, ] 
 involves an empiric judgment in which we think of a sequence 
 as determined, i.e., as presupposing another phenomenon as 
 previous in time, on which it must necessarily follow according 
 to a rule. And if, on the contrary, I were to posit the 
 antecedent, and the event were not to follow necessarily upon 
 it, I sholild be obliged to regard it as a subjective play of my 
 VOL. I. 2 N 
 
562 THE CRITIQUE OF PURE REASON. book r. 
 
 imagination, and if I still held it to represent anything 
 objective, I should call it a mere dream."^ In other words, the 
 condition on which I regard a particular succession of any 
 perceptions as representing a change in the objective world, is 
 j that I regard it as a link in a chain of successive changes, each 
 of which contains in it the necessity of the subsequent link, 
 just as in time itself each moment contains the necessity of the 
 following moment.^ And if, on this principle, I had objectified 
 a succession of my perceptions by connecting it with some 
 antecedent on which I regarded it as a necessary consequent, 
 and then found that when the antecedent was posited, the 
 consequent did not follow, I should immediately think myself 
 to have been dreaming. For it is just this thoroughgoing 
 connexion of phenomena according to rules, which constitutes 
 that objective reality in relation to which alone we can be 
 conscious of ourselves. Kant's statement is only so far 
 imperfect as the recognition of an event as a dream is only the 
 reference of it to a different place in the context of experience ; 
 whereas in the clause, " if I still held it to represent anythuig 
 objective," he seems to mean that apart from the reference 
 of successions of phenomena to definite correlates in the 
 preceding time, we should not be conscious of them as events 
 at all. In other words, just as an object would not be 
 distinguished from a perception, unless it were conceived to be 
 permanent while the perception changes, so a change would not 
 be distinguished from a succession of perceptions, unless it were 
 regarded as determined in relation to previous changes. Kant 
 would have expressed his meaning better, if he had said that 
 a change not regarded as a link in the chain of connected 
 changes, would be to us as a dream ; and that if antecedents 
 did not bring with them definite consequents, experience as a 
 consciousness of objects in time could not exist; for "it is 
 
 1 A. 201 ; B. 243. 
 
 ^ These, indeed, are not two separate things, for it is, indeed, through the 
 objective succession of events that the succession of times is determined for us. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 563 
 
 only as a certain order in the time-relations of our ideas is 
 il'ecessary, that objective significance can be attributed to them," 
 and apart from such determination, " we should have a play of 
 ideas which referred itself to no object, i.e., our perception 
 would not enable us to distinguish one phenomenon from 
 another as respects its time-relation." 
 
 Now, as such a determination of phenomena cannot be given The idea of 
 
 causality can 
 
 to us in sense, nor in the mere imaginative synthesis of sense- ^rom pheno- 
 data, it must be derived from the understanding, as one of the Secause'the 
 
 understand- 
 
 conditions under which we bring the data of sense in that ing has 
 
 previously 
 
 determination of them as objects, which is necessary to their t^embyitf 
 being brought into relation to the " I think." This, indeed, 
 " seems to contradict all the observations usually made as to the 
 conduct of our understanding. For the view usually taken is 
 that it is the sequence of many phenomena which have been 
 observed, compared and found to agree, which first guides us to 
 the discovery of a rule according to which certain events always 
 follow upon- certain phenomena, and that it is only as a result 
 of this experience that we are led to form the conception of 
 a cause. If that were the case, the idea of cause would be 
 empirical, and the rule of causality as accidental as the 
 experiences from which it is derived ; and its necessity and 
 universality would be altogether fictitious." In truth, however, 
 " it is in' this case exactly as in the case of other pure a priori 
 ideas (e:g., space and time), which we can extract as clear con- 
 ceptions out of experience only because we have ourselves put 
 them into experience, and thereby brought experience into exist- 
 ence.",i It is true that it is only after we have made use of 
 the conception in experience that we can give it "logical 
 clearness," i.e., that we can be conscious of it for itself as a 
 general principle ; but " a reference to it as a condition of the 
 synthetic unity of phenomena in time, is the ground of experi- 
 ence itself, and anticipates it as an a priori condition." 
 
 These two points, then, are to be firmly kept in view : first, U^SIduction. 
 1 A. 196 ; B. 241. 
 
564 THE CRITIQUE OF PURE REASON. book i. 
 
 that " I never refer sequence to the object and distinguish it 
 from the subjective sequence of our apprehension, except on 
 the basis of a rule which compels me to observe that order of 
 perceptions rather than another, and that indeed it is this 
 necessity which first makes the idea of succession in the object 
 possible " ; and, secondly, that this rule is not given in the 
 sense, nor in the imaginative synthesis of sense-data, except in 
 so far as the latter is determined in relation to the unity of 
 apperception, which expresses itself in the category of causality. 
 But then the imaginative synthesis musi be so determined iu 
 order that the consciousness of events as happening may be 
 possible; or, what is the same thing, in order that the percep- 
 tion of such events may be capable of being united with the 
 consciousness of self 
 S'itl'a) Caul We must remember always the conditions of proof or deduc- 
 bogoAy' ° tion according to the transcendental method. We have to 
 
 analysis from 
 
 eTen?^*°'*° dcducc this, like the other principles of piire understanding,- 
 " not directly from conceptions, but indirectly through the re- 
 ference of these conceptions to something which is itself quite 
 contingent, viz., possible experience. For if this (i.e., something, 
 as an object of possible experience) be presupposed, the prin- 
 ciples of pure understanding can be shown to be apodietically 
 certain ; whereas if we take them by themselves (directly) it 
 is impossible a priori to know anything about them. Thus, no 
 one can discover a basis for the principle that everything that 
 happens has a cause, by means of those conceptions alone. 
 . . . But, though it needs proof, it is not to be called a 
 dogma, but rather a principle ; because it has the strange 
 peculiarity that it makes possible the very experience which is 
 its own ground of proof and in such experience requires always 
 to be presupposed." ^ In other words, if we take the conception 
 (2) But an of an cvcut by itself, no analysis of that conception will enable 
 beknownTo US to discovcr the uccessity of relating it to a previous event 
 
 us except 
 
 through the as its causc. If, however, we consider how it is possible 'for 
 
 TJnncinle of ^ 
 
 principle of 
 causality. 
 
 1 A. 737 ; B. 765. 
 
CHAP. vili. THE ANALOGIES OF EXPERIENCE. 565 
 
 us to come to the knowledge of it, or, in other words, how it is 
 to be brought into relation to the " I think," we see that the 
 mere fact that the event happens will not enable me to be 
 conscious of it as happening. An event, as a change in an 
 object, can be known to me through my perceptions, which in 
 themselves are merely successive states of my consciousness, 
 only in so far as my imagination retains and connects them 
 according to universal rules of determination ; and in this 
 case the rules in question must express, (1) the reference of 
 these states as successive to a permanent object ; and (2) the 
 conception of such succession as a link in a series of changes, 
 each of which is determined in relation to a previous change. 
 Only in so far as imagination is determined in its synthesis 
 by such rules, and in so far' as we become conscious of it as 
 so determined, can we " recognise " the successively appre- 
 hended data of sense as representing to us an objective change 
 and not as a mere sequence in our perceptions. 
 
 The difficulty of understanding this view arises, as usual, f^^'^^''f^°^^^ 
 from the twofold Aspect of the judgment of knowledge, in which, ?°l5Sp°*Sy' 
 as we must always remember, subject and predicate are not 
 elements given apart and brought together without change, but 
 in which each gets the character it has in the judgment only 
 through the relation to the other into which the judgment 
 brings it. If we forget this, the judgment becomes unneces- 
 sary or impossible ; unnecessary, if the subject already has the 
 character which it gets as subsumed under the predicate, and 
 impossible if it has not. In truth, what Kant shows is not that 
 a change, perceived as such, can be brought under the principle 
 of causality, but that, if we take from the perceived change 
 the characteristics which we give to it in recognising it as an 
 effect, we reduce it to a mere sequence of perceptions which 
 we could not recognise as representing a change, or, indeed, as 
 representing anything. It is in relation to the unity of ap- 
 perception which expresses itself in the category of causality 
 that a "succession of perceptions can alone be determined as a 
 
566 THE CRITIQUE OF PURE REASON. book I. 
 
 change. On the other hand, it is only in relation to the deter- 
 mination of the object as changing, that the category, which 
 corresponds to the relation of reason and consequent in the 
 hypothetical judgment, becomes schematised as a relation of 
 invariable succession. Apart from their relation in the judg- 
 ment of experience, in fact, the subject would be a mere 
 succession of perceptions, and the predicate would be a mere 
 relation of the factors in an analytical judgment of pure 
 thought. The main ambiguity that remains is that which 
 arises from the way in which Kant, as it were, substan- 
 tiates perception and conception in their independence, and 
 refuses to recognise that they are merely abstract elements 
 which are nothing apart from their unity. Hence, espe- 
 cially, he seems here to admit that the sequence of percep' 
 tions as combined by imagination is something of which we 
 can be conscious for itself, apart from any determination of it 
 by conceptions. . It is, however, obvious that, on Kant's own 
 principles, the consciousness of nature as an ordered system, in 
 which all the elements are determined as in necessary relation 
 to each other, is the necessary correlate of the consciousness of 
 the unity of the self ; and that the consciousness of the succes- 
 sive process, by which this consciousness of an objective world 
 is developed ex parte nostra, implies a reflexion which cannot 
 precede the process itself.^ 
 cuftyas'to^the ^^ ^^' ^owever, nccessary at this point to notice a certain 
 oaimiity." Want of adjustment between the category and the perception 
 subsumed under it, which makes the argument of Kant more 
 difficult to interpret in the case of the category of causality 
 than in the case of the category of substance. According to 
 Kant, the (analytic) relation of reason and consequent is 
 schematised in relation to time as the (synthetic) relation of 
 essentially different phenomena which, though quite distinct, 
 are yet necessarily connected in our experience. In this way 
 Kant meets the double objection to the category of causality, 
 
 1 B. 155. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 567 
 
 the objection (which he had encountered as early as the Essay 
 tmeards the Introduction of the Idea of Negative Quantity into 
 Philosophy), that pure thought cannot explain change, and 
 the objection suggested by Hume, that sense when it presents 
 us with different successive phenomena does not enable us to 
 establish any necessary relation between them. Kant seeks to 
 meet the difficulty by the idea that thought in relation to sense 
 is brought into contact with the changing series of phenomena, 
 and that between the different successive links it establishes 
 a necessary relation, which is analogous to the identical rela- 
 tion of reason and consequent in pure thought. This necessity 
 of relation between phenomena (events) that are different, is 
 the nearest analogon to the pure analytic relation of thought 
 in the hypothetical judgment. 
 
 Now, it is easy to see that this way of stating the matter "'Kanrl'ate^ 
 gives rise to great difficulty, just because Kant conceives that, between™" 
 
 analytic aud 
 
 on the one side, there is the mere difference of sense, and on synthetic 
 
 judgment, 
 
 the : other side, the transparent unity of thought ; or, to put it 
 more definitely, that on the one side, we have an analytic con- 
 nexion of thought, such as is expressed in the proposition, " If 
 man is ipimortal, he cannot die," and on the other side, a 
 succession of events like the appearance of a flame and the 
 sensation of heat, between which there seems to be no more 
 necessity of connexion than between day and night. And 
 Kant's answer — that the identity of thought broken in rela- 
 tion to difference of phenomena in time, yet so far recovers 
 itself as to establish necessary relations between them, and so 
 alone determines their succession as objective — seems forced 
 and artificial ; because, in this connexion at least, he seems to 
 look for no identity continuing itself through the- different 
 events that are thus brought together as necessarily related ; 
 and therefore their necessary relation appears to be a kind of 
 miracle operated upon them by thought. We have, however, 
 to observe, that Kant is led thus to omit reference to the iden- 
 tity that maintains itself through change, partly because he has 
 
568 THE CRITIQUE OP PURE REASON. book i. 
 
 made, as he thinks, sufficient provision for it in the principle 
 his^sep'amto" of substancc. The imperfection of Kant's statement lies mainly 
 thTprinoipies in the sharp way in which he has distinguished the identity 
 
 of substance 
 
 and oausauty. of the substaucc and the necessity of connexion between the 
 changing states, as if the latter were not just the expression of 
 the former. For in this way, the relation of cause and effect 
 seems to be externally superadded to the relation of substance 
 and accident, the former being the relation of the permanent 
 to its successive states, while the latter is the relation of these 
 states to each other. But when the identity is thus left on 
 one side, the relation of the successive states can appear only 
 as a succession of qualitatively distinguished phenomena, between 
 which no intelligible connexion can be discerned. Kant, how- 
 ever, when he declares that the permanence of substance is 
 merely the presupposition of the other analogies, points to the 
 truth that the category of causality is not simply an addition 
 to the category of substance, but that it is a higher form of the 
 same idea. In fact, science passes from the category of sub- 
 stance to that of causality whenever reflexion has been carried 
 so far as to see that what is wanted to explain phenomena is 
 to discover not merely an identity under their difference, but 
 an identity in their difference. And Kant again suggests the 
 same idea when he insists on the continuity of change ; for such 
 continuity is inconsistent with the thought of any qualitative 
 break between cause and effect, and makes us rather think of 
 a process of transformation, in which an identity manifests 
 itself and maintains itself in continual change. 
 
 railtaiity to ^^ truth, as the categories are different phases of one 
 thought, it is impossible to work with more than one category 
 at a time ; and the point of view expressed in the category of 
 causality excludes the point of view expressed in the category 
 of substance, just because it contains and goes beyond it. It 
 is the product of a more advanced reflexion, which brings the 
 identity of substance into unity with the difference of the 
 accidents, and views it as an identity which continually goes 
 
 substance. 
 
CHAP. Vlll. THE ANALOGIES OF EXPERIENCE. 569 
 
 beyond itself ; in fact, it is just the notion which Kant found 
 so paradoxical in his early essay, that the position or assertion 
 of one thing is at the same time the position or assertion of 
 something different from it. A cause is just an identity which 
 is regarded as revealing itself in transition to another, which, 
 however, is "not another " ; for the identity is regarded as still 
 expressing itself in it. From this point of view, we can under- 
 stand why such opposite and apparently contradictory things 
 should be capable of being said of the relation of cause and 
 effect; why, from one point of view, the cause should be regarded 
 as necessarily other than its effect, while, from another point of 
 view, it is thought of as identical with it. For that which is 
 thought of as a cause, is thought of as manifesting itself only as 
 it goes beyond itself, while yet it is regarded as passing into and 
 reproducing itself in the effect. Hence, we are never satisfied 
 till we have resolved the effect into a cause which is the unity 
 of all its conditions, a unity which means just the effect 
 analysed. Yet, on the other hand, the difference of cause and 
 effect returns upon us, in so far as we are analysing a process 
 in which the elements of the cause are disengaged from one 
 connexion to reappear in another ; and the difficulty thus 
 suggested finds its proximate solution in the category of reci- 
 procity. 
 
 Leaving this last point for the present, we may observe that ^^^^^^ ;„ 
 the fundamental error of Kant lay in his partial acceptance of Su^iSy ^ 
 the principle of Hume, " that no idea or object considered in ws view of 
 
 pure thought. 
 
 itself, can give a reason for drawing a conclusion beyond it;" 
 ie., his acceptance of this principle as true for pure thought, 
 though not true for thought as schematised. In fact, as we 
 have seen, if this view of thought were true, judgment, even 
 analytical judgment, would be impossible ; or, admitting its 
 possibility, it would be confined to simple position and nega- 
 tion — position as ,the assertion of a thing as identical with 
 itself, and negation as the exclusion of all difference from it- — - 
 and frdin this we could at the utmost derive only the categories 
 
570 THE CRITIQUE OF PURE KEASON. book i. 
 
 of quality and quantity. For, in the categories of relation the 
 ideas of identity and difference, position and negation, are 
 brought together and viewed as correlative aspects of the same 
 thing. The simplest form of such relative conception is the 
 idea of substance, i.e., the idea of a unity in which the difference 
 of the accidents continually disappears, but which yet is deter- 
 mined only in relation to its accidents. Hence, we necessarily 
 think of it as continually reproducing the accidents, which as 
 continually disappear in it. This idea of the reproduction of 
 the differences out of the unity is, however, only latent in the 
 category of substance. It is a second thought which comes to 
 us when we realise that the substance, which we at first con- 
 trast with its accidents as reality with appearance, is after aU 
 nothing except in relation to its accidents. When this second 
 thought is emphasised or becomes explicit, we pass from the 
 category of substance to that of causality ; i.e., we get the idea 
 of an identity which is negatively related to itself, or which 
 manifests itself in going beyond itself and positing something 
 other than itself; we get, in fact, the very idea which caused 
 Kant so much difficulty in his early essay. And the other 
 diflBiculty specially suggested by Hume, i.e., the difficulty of 
 establishing a connexion between phenomena given as different, 
 is just the counterpart of this ; for a thought which cannot go 
 beyond its own identity, cannot bring back into its identity the 
 phenomena which are given in difference. On the other hand, 
 if we say that there is no possibility of isolating a thought so 
 that it is cut off from relation, (for even in our utmost abstrac- 
 tion we cannot destroy the relativity of a conception to that 
 from which we abstract, and indeed the effort of abstraction 
 only makes the relativity more manifest,) we take away the 
 ground for the opposition of thought and perception, and the 
 necessity for the elaborate mediation between them, which 
 Kant has contrived. Thus, the principle of reason and conse- 
 quent ceases to be distinguished from the principle of causality; 
 and the latter takes its place, not, indeed, as an absolute prin- 
 
CHAP. vm. XHE ANALOGIES OF EXPERIENCE. 571 
 
 ciple, but as one of the principles by which the objective world 
 .must be determined, in the process of bringing it into relation 
 to the unity of apperception. 
 ! I have already said that the proximate solution of the con- sohopen- 
 
 ^ hauer's objec- 
 
 tradietion involved in the principle of causality is to be found ^™ *°f ^^"''^ 
 in the principle of reciprocity, into which it passes, when that aumession to 
 
 causality. 
 
 which is latent in it is brought to light. The principle of sub- 
 stance emphasises the reality of the identity of the object as 
 maintaining itself in change ; the principle of causality empha- 
 sises the reality of the difference of the forms in which the 
 identity shows itself; but the category of reciprocity brings 
 ' both together, in the idea of a changing relation of substances 
 which through it maintain their unchanging identity. N'ow, 
 we may best understand the necessity of this transition, and 
 the way in which Kant expresses it, if we consider a special 
 difficulty which has been raised, in connexion with Kant's 
 deduction of the principle of causality. Schopenhauer, in his 
 criticism of the Critique, argues that Kant's deduction would 
 reaUy prove too much ; for it would involve that all objective 
 sequence is causal, and that, therefore, any event must be the 
 cause of any other perceived after it, e.g., night and day must be 
 causes of each other. Now, to this a good answer can be made 
 entirely from the point of view of the deduction itself. Kant 
 is showing, not that objective succession is always causal, but 
 that the determination of a succession of perceptions as referring 
 to a succession of states in an object, involves the principle of 
 causality. If Kant had meant to assert that all succession is 
 necessarily causal, he would have said what is obviously 
 absurd, and what could easily be refuted by his own illustration ; 
 for the successive positions of the ship going down the stream 
 cannot he said to be causes and effects of each other; though, 
 no doubt, if we talce all the conditions under which the ship is 
 placed in its successive positions, we can discover a complete 
 cham of causality between them, and Kant would be prepared 
 to argiife that such, a succession of positions cannot be referred 
 
572 THE CRITIQUE OF PURE REASON. book I. 
 
 to the boat as an object, without aid from the principle of 
 causality. For what he contends is, that there is no recogni- 
 tion of a succession of states as in an object, which does not 
 imply a recognition that such succession is determined by a rule 
 which is independent of any perception. Day follows night 
 and night follows day without either being the cause of the 
 other ; but if we consider what are the changes in the relations 
 of the earth to the sun, we can find in the circumstances of the 
 earth at each point of time the cause of the transition towards 
 light or darkness through which it is passing. And the sup- 
 position that we can do so is bound up with the recognition of 
 the successive perception of light and darkness as an objective 
 change of the earth's state. In other words, the recognition of 
 a sequence of ideas as representing an objective change is the 
 recognition of that sequence as a link in a chain of changes 
 which is determined by a general rule of possible perception, 
 and not merely a sequence of states in the experience of an 
 individual, 
 answer'to this " When Schopenhaucr adduces the sequence of musical notes 
 objection. ^^ ^^ ^^^ ^^^ night, as objective sequences which can be known 
 without the causal law, we need only meet him with the ques- 
 tion. Where in these cases is the substance that changes ? So 
 soon as he is forced to put his objection into the form required 
 to bring it into relation to the question of the possibility of 
 knowledge, his error becomes obvious. His instances must 
 then be expressed thus : — The instrument passes from one 
 state of sound into another ; the earth changes from the 
 measure of enlightenment which makes day, to that which 
 makes night. Of such changes no one will say that they are 
 not referred to a cause. And we may quote in this reference 
 the appropriate saying of Kant himself, ' Days are, as it were, 
 the children of Time, since the following day with that which 
 it contains is the product of the previous day.' " 
 
 How casual 
 
 sequence is to " Anotlicr obicction of Schopcnhauer is not unimportant and 
 
 be explained <j i. jr 
 
 ?riTOipi'e3.° deserves to be mentioned. He points out that that which we 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 573 
 
 call chance is just a sequence of events which do not stand in 
 causal connexion. ' I come out of the house and a tile falls 
 from the roof which strikes me ; in such a case there is no 
 causal connexion between the falling of the tile and my 
 coming out of the house, yet the succession of these two 
 events is objectively determined in my apprehension of them.' 
 How have we to criticise this case from the transcendental 
 point of view ? We know that successions become necessary, 
 i.e., objective, for our consciousness, when we regard them as 
 changes of a substance which are determined by a cause. But 
 it is shown here that there are successions in which the single 
 members are changes of different substances. If substance S 
 changes its state A into B on account of the cause X, and 
 substance S' changes its state A' into B' on account of the 
 cause X', and if 1 call the first change V and the second V, 
 the question arises how the objectivity of the succession V V 
 is related to the law of causality. Sequences such as W are 
 very frequent, and our consciousness of the objectivity is 
 certain. Do we owe this consciousness to the same rule as 
 holds good in the other cases ? " 
 
 "Certainly. The distinction is not qualitative, but rests 
 only on the greater complication of the change in question. 
 The sequence VV can become objective only if I think it as a 
 . necessary connexion. It must be so determined that V can only 
 follow V in ' consciousness in general ' ; there must be a U, the 
 introduction of which is the cause that V foUows V. To be con- 
 vinced of this, I do not need actually to know U. I know that 
 on every occasion U causes the succession VV. Of course, this 
 presupposes that all data of the states considered, A and A', 
 remain identical. But whether these data are very simple or end- 
 lessly complex, whether they are likely to combine to the given 
 result frequently or seldom, is indifferent for the objectifying of 
 the event ; ^ it is not the perception of IT, but the presupposi- 
 
 'Stadler, Die Grundadtze der reinen Erkenntniss-Theorie, p. 152. The 
 subject is very clearly Ulustrated In this treatise. 
 
574 THE CRITIQUE OF PURE REASON. book i. 
 
 tion of it, which makes the change necessary and so objective 
 
 for us." 
 Won to^Kant^s ^ 8°°'^ answex may thus be given to the objection of 
 statemmrJf Schopcnhauer from the point of view of Kant. At the same 
 it to be tkken time, it should be admitted that there is an ambiguity in Kant's 
 
 as directly 
 
 applying to proof of the principle of causality, which naturally led to mis- 
 expenence? ^iKjerstanding. For, in attempting to show that objective 
 sequence or change cannot be determined as such except when 
 it is conceived as a link in a series of changes, he at least 
 suggests the idea that states of the same substance may be 
 regarded as causal in relation to each other. It seems, indeed, 
 to have been Kant's view, or at least his earlier view, that we 
 can prove the principle of causality in general in reference to 
 all substances, whether given in external or internal experi- 
 ence ; and whether they be determined to their changes by 
 their own previous states or by the states of other substances. 
 Accordingly, when in the Metaphysical Rudiments of Physics, 
 he proceeds to limit the law of causality in regard to objects 
 given in external experience, he lays down the law that for 
 them the cause of change must always lie in another substance. 
 Yet it is clear that by this time Kant had begun to suspect that 
 the principle of causality, like the principle of substance which it 
 presupposes, can be applied, in the first instance at least, only to 
 objects of external experience, and to them only as determined 
 by external causes. For if, as was pointed out above, the idea of 
 existence in time taken by itself, does not involve permanence 
 except in the sense of a continued flux, neither can we say that 
 it involves a perma.nent relation between the elements that thus 
 pass into each other. Hence, in a remark added in the second 
 edition of the Critique, Kant maintains the necessity, not only 
 of perception, but of external perception to supply the subject 
 to which the predicate of causality is to be applied. " In order 
 to exhibit change as the perception corresponding to the con- 
 ception of causality, we must take motion, as change in space, 
 for our example. This, indeed, is the only way in which we 
 
CHAP. vill. THE ANALOGIES OP EXPERIENCE. 575 
 
 can bring change before us in perception, pure understanding 
 being totally unable to grasp the possibility of change. Change 
 is the combination of contradictorily opposed predicates in the 
 existence of one and the same thing. But it is impossible for 
 any one by mere thinking, without an example, to comprehend 
 how, out of a given state of a thing, an opposite state of 
 the same thing should follow ; nay, he cannot attach any 
 meaning to such an idea without a perception. And the per- 
 ception required is that of the motion of a point in space, the 
 existence of which in different places (as a consequence of 
 opposite determinations) alone makes it possible for us to 
 realise change to ourselves. Por, in order subsequently to 
 make even inner changes intelligible to ourselves, we need to 
 figure time, as the form of inner sense, by a line, and the inner 
 change by the drawing of this line (motion) : thus using ex- 
 ternal perception as a means to the understanding of our own 
 successive existence in different states. And the reason of this 
 is, that all change presupposes in the perception of it something 
 permanent ere it can be perceived as change, but that in inner 
 sense no permanent perception can be found." ^ 
 
 To this it need only be added that, as all motion is relative, ^^ot'o*""^ 
 there is no possible consciousness of an object moving directly ™uBaiitymust 
 in relation to space, or of the world as a whole moving in ultimately as 
 
 reciprocity. 
 
 space, but only of the motion of two or more substances 
 which change their position in relation to each other. And 
 it has already been pointed out, that the principle of the 
 pennaneiice of substance in all change means that the objects 
 in changing their relation to each other in space, cannot change 
 their relation to space as occupying it. It follows, then, that 
 change can be conceived only as an alteration of substances in 
 their relation to other substances, and that all causation is ex- 
 ternal. And, in a single substance conceived as existing by 
 itself, or in the world as a whole, we cannot conceive of any 
 
 ' Kritik, B. 292. We postpone till the next chapter the consideration of the 
 necondary application of the Categories, here suggested, to inner experience. 
 
576 THE CRITIQUE OP PURE REASON. book I. 
 
 change as taking place. Or rather, as we should say, from 
 Kant's point of view, we have not in these cases the con- 
 ditions under which alone we can be conscious of a change. 
 The " second law of mechanics," that every object maintains its 
 state of rest or motion unless it be determined to change it from 
 without, i.e., by another object, is therefore the true expression 
 of the law of causality. It also shows its necessary relation to 
 the law of reciprocity, which simply means that as all change 
 is relative, it must be equally attributed to both terms in the 
 relation. If we regard the change as only in one substance, 
 we are simply attending to one term and abstracting from the 
 other. It might indeed be said that, if we conceive the world 
 as one substance passing through a series of states, we must 
 conceive the successive states as necessarily determined in 
 relation to each other ; but, if we thus take the world as one 
 substance, we cannot account for any change taking place in it ; 
 at least, we cannot do so consistently with Kant's proof of the 
 principle of substance, which, as we have seen, involves that 
 there can be no absolute change of substance, but only a 
 relative change within its identity, i.e., a change of its parts in 
 relation to each other. The principle of substance, therefore, 
 when we combine it with the principle of causality, necessarily 
 expands into the principle of reciprocity, 
 men* and*"" Kaut's dcduction of the principle of reciprocity follows the 
 the Principle Same coursc as the previous deduction. In the first edition it 
 
 ofReciprocity. 
 
 stands thus : " All substances, so far as they are coexistent, 
 are in thoroughgoing community or reciprocity of action with 
 each other." There is a somewhat improved statement of this 
 principle in the second edition, which, (1) gives prominence to 
 the transcendental aspect of it, and (2) limits the determination 
 of coexistence to the determination of substances as in space. 
 " All substances, in so far as they can he perceived in space as 
 coexistent, are in complete reciprocity." Why should it be so ? 
 The mere conception of things coexistent does not imply their 
 , reciprocal determination by each other. " For, supposing 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 577 
 
 several substances to exist, how can we think it possible that 
 something should follow as effect from the existence of each of 
 them in the existence of the others reciprocally ; and, therefore, 
 that in the existence of each, there should be something which 
 cannot be explained from itseK without reference to the exist- 
 ence of the others?"^ This question was answered by Kant 
 himself in the Dissertation, by reference to the unity of God as 
 the cause of the existence of all the dependent substances, 
 who at the same time binds them to a unity with each 
 other. Uhitas in conjundione suhstantiarum est consectarimn de- 
 pendentiae omnium ah Uno. For, as Kant argued, " the 
 possibility of a commereium of substances is not necessarily 
 involved in their mere existence," seeing that as substances, 
 they are rather conceived as independent of each other. The 
 individualism of Leibniz, in fact, started with the isolation of 
 individual substances as such, and hence it had to treat of their 
 union in one world as a harmony, which was externally estab- 
 lished between them. In the Critique the problem changes, and 
 for the unity of the existent world as dependent on one God, we 
 have the unity of the known world as present to the conscious- 
 ness of one self. The commereium of substances is a necessity, 
 because otherwise they could not be known as coexistent in one 
 space. For, Kant argues, space by itself is not perceived so 
 that the coexistence of objects might be ascertained by their 
 relation to it. On the contrary, it is the reciprocal determina- 
 tion of objects by each other, or in other words, their 
 "dynamical community," which is the ground of the deter- 
 mination of them as coexistent in space. " Without 
 community every perception (of a phenomena in space) would 
 be broken off from the others, and the chain of empirical ideas, 
 "4.6., experience, would have to begin anew with every new 
 object, without its being possible to connect the previous ex- 
 perience with it, or to place it in a relation of time thereto." ^ 
 
 », . 1 B. 292. General remark on the system of Principles. 
 
 • 2 A. 213; B. 260. 
 VOL. I. - 2 
 
578 THE CRITIQUE OF PURE REASON. book i. 
 
 topiuerta' " ^® cannot directly perceive coexistence, for the synthesis of 
 mTm»».™ imagination by which the data of sense are taken up and com- 
 bined into images, is always a successive process. Hence, if we 
 were conscious of that synthesis, and of that synthesis only, 
 all we should have before us would be a mere succession of 
 ideas, some of which recurred at different times, one being 
 present when another was absent. But, as a consciousness of a 
 succession of ideas in itself does not give us a consciousness of 
 change, unless that succession is referred to an object which is 
 permanent, and regarded as a link in the chain of changes of 
 state to which such objects are subjected : so the mere con- 
 sciousness of a recurrence of similar ideas in successive times 
 would not enable us to interpret that recurrence as showing 
 the coexistence of their objects, unless we regarded these ideas 
 as representing states of different objects, each of which recipro- 
 cally was a condition implied in the corresponding state of the 
 other. Thus, when I observe first the moon, then the earth, 
 then the moon again, and determine them as coexistent, there 
 is something more in my consciousness than a mere succession 
 of perceptions, first in one order and then in the opposite 
 order. And this something more is not the perception of time 
 or space, as something objective to which each perception is 
 referred, and through which they are dated and placed in 
 reference to each other ; for time and space are mere forms of 
 relation. The consciousness of coexistence can, therefore, be 
 attained only if, in the synthesis of my perceptions, I recognise 
 them as representing states of different objects and as con- 
 nected in such a v^ay, that each of them refers back to the 
 other as its condition. " Now, a relation of substances, 
 wherein one contains determinations of which the ground lies 
 in the other, is a relation of influence ; and when reciprocally 
 the substance so determined in itself contains the ground of 
 determination in the substance that determines it, it is a 
 relation of community or reciprocity. Hence, the coexistence 
 of substances in space canuot be empirically known otherwise 
 
CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 579 
 
 than under the presupposition of a reciprocity between them ; 
 and this reciprocity is, therefore, the condition of the possibility 
 of the things themselves as objects of experience." ^ It is the 
 double refraction, as it might be called, of . objects upon each 
 other that determines them for us as coexistent. 
 
 In all this we have just another specimen of the transcen- positive and 
 dental argument, which in this case starts with the supposed ^^"ot this 
 
 . ^ ^ . Deduction. 
 
 fact of coexistence, and asks under what conditions it can 
 become a fact for me ? In answering this question, Kant 
 endeavours to show that, in order to a consciousness of the 
 coexistence of objects, it is required that the imaginative 
 synthesis by which the elements of the sense manifold are put 
 together, should be determined by the category of reciprocity, 
 and that we should in some way recognise this determination. 
 For, space not being perceived in itself, so that the phenomena 
 might be relatively placed in reference to it, the unity of the 
 space in which objects coexist can be apprehended only as the 
 presupposition of the reciprocal determination of the parts of 
 space. But even these are not perceived by themselves, but 
 only through the reciprocal determination of the phenomena 
 which are perceived in space. These phenomena, therefore, 
 must be apprehended by us as the states of substances, which 
 reciprocally determine each other : the substance being con- 
 ' ceived as permanently occupying space and so " representing " 
 its identity, while the changing states in their reciprocal deter- 
 mination, fix at the same time the different substances as 
 coexistent in different parts of the same space. Thus alone 
 can different objects be apprehended as present to us in the 
 unity of one world, a world the consciousness of which can be 
 united with the consciousness of one self. " For the unity of 
 the world, the whole in which all phenomena are supposed to 
 be combined, is manifestly a mere consequence of the tacitly 
 assumed principle of the community of all substances which 
 coexist ; for, if they were isolated, they would not constitute 
 
 1 B. 258. 
 
580 THE CRITIQUE OF PURE REASON. book l. 
 
 parts of one whole ; and if their connexion (the reciprocity of 
 the manifold) were not necessary as the presupposition of the 
 coexistence, we could never argue from the latter, which is a 
 merely ideal relation, to the former, which is a real relation of 
 them. We have, however, shown that community is the ground 
 of the possibility of any empirical knowledge of coexistence, 
 and, therefore, we can quite legitimately conclude from the 
 latter to the formfer as its necessary precondition."^ 
 
 Of course, this argument — that in order that objects may 
 be determined for us as coexistent, they must be determined as 
 in community- — has as its complement the argument, that, if 
 we take away community as not given in perception, we must 
 take away coexistence with it. To put it formally, in the 
 judgment of knowledge coexistent phenomena may be taken as 
 the subject, and reciprocal determination of these phenomena 
 by each other as the predicate ; but it is only in relation to 
 this conception as a predicate, that that perception can be 
 given as the subject. And with the withdrawal of the concep- 
 tion, the perception shrinks into a sensation, of which nothing 
 can be said. 
 SfoinXgonot The difficulty of following this deduction is similar to that 
 of dfijunctive which WO have found in the case of causality, viz., that the 
 
 judgment. 
 
 unity of thought and the difference of perception seem to be 
 immediately brought together ; and the latter subsumed under 
 the former. Thus, the relation of reciprocity, i.e., the action 
 and reaction of quite different substances of which each deter- 
 mines the other's state, is the analoc/on of the analytic unity of 
 the genus with the sum of its species, which shows itself in 
 the fact that they reciprocally exclude each other, while yet 
 each requires the other as its complement in the totality of 
 the genus.2 The defect in the analogy, however, seems to be 
 
 J A. 219 ; B. 265. 
 
 ^ A. 73 ; B. 99. An analytic division of the species would necessarily be 
 dichotomons, though even for such a division we should need to give a sense 
 to the negative, which formal logic does not admit. 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 581 
 
 that the different substances are here taken as primarily un- 
 related to' each other, and that the reciprocal determination 
 of their states is regarded as simply the necessary contem- 
 poraneity of a state or a change of state in one of them, with 
 a state or change of state in the other. Just, therefore, as 
 Kant took cause and effect as two disparate events or changes 
 which follow each other, like the feeling of heat and the percep- 
 tion of flame, so here he takes the action and the reaction as 
 two contemporaneous events or changes in the two substances ; 
 yet, just as in the former case he conceives that the successive 
 phenomena must be subsumed under the analytic unity of 
 reason and consequent, that so their unrelated difference may 
 be turned into a necessary connexion of events, so here the 
 different contemporaneous phenomena, in order that we may 
 know them as contemporaneous, have to be subsumed under 
 the relation of species in one genus, (which, while they exclude, 
 yet imply each other,) in order that their unrelated difference 
 may also be turned into a necessary relation of the states of 
 different substances. Now, we have already seen that, in the 
 case of the causal relation, the defect of Kant's deduction lies 
 in his omission to look for the identity which persists through 
 the difference of the events related as cause and effect, an 
 omission which made their connexion appear as a kind of 
 miracle operated on them by thought; but this defect was 
 partly rectified by the way in which he connected the prin- 
 ciple of causality with that of substance, and also by the way 
 in which he insisted on the continuity of change. Now, there f '°^^|^^y °' 
 is a similar error and a similar correction of it in Kant's treat- affflrent 
 
 ment of the principle of reciprocity, which also he conceives as related ele- 
 ments corre- 
 |)resupposing the principle of substance. For, this means that |P°j!4^p*°„*^^ 
 
 in determining the changeable states of substances as contem- j'ui^nt. 
 poraneous, we must not only unite them as reciprocally 
 determining each other, but we must presuppose one all-em- 
 bracing substance, which " represents " space as an infinite 
 given *whole, and which remains unchanged in all relative 
 
582 THE CRITIQUE OF PURE REASON. book I. 
 
 changes of its parts. But such reciprocity of changes within 
 an identical substance, cannot be conceived as a relation of 
 phenomena which are qualitatively indifferent to each other, 
 but only as a relation of polar opposites which at once imply 
 and exclude each other. It is such reciprocity that Kant m 
 his Metaphysical Hudiments of Physics seeks to discover in 
 matter (as determined by repulsive and attractive forces), and 
 it is such reciprocity that physical and chemical science is 
 always striving to reach in all the different classes of phenomena 
 with which it deals. As in the case of causality, the scientific 
 man is not satisfied without the reduction of cause and effect. to 
 identity, so in the case of reciprocity (into which causality 
 passes when fully understood), he is not satisfied till he has 
 resolved the different constitutive elements of a complex result, 
 into correlated factors which have no meaning except as factors 
 of this relation. In both cases, as has been already shown, the 
 categories with which he operates are categories of reflexion, 
 categories which involve that double movement of differentia- 
 tion and integration which gave rise to the two problems of 
 Kant. For, as we have seen, it was his first difficulty to 
 understand how it is possible to proceed from the position or 
 assertion of one thing to the position or assertion of something 
 different from it ; and it was his second difficulty, suggested 
 mainly by Hume, to understand how, different phenomena 
 being given, it should be possible to discern necessary relation 
 between them, and so to return from difference to identity. But 
 Kant's formal view of thought, rendering the former impossible, 
 made it necessary for him to suppose that in reference to space 
 and time, the pure identity of thought with itself, as expressed 
 in the various analytic judgments, gives rise to various 
 analogues of itself; i.e., conceptions of necessary relations 
 between factors given as different. If, however, we reject 
 this formal view of thought, and regard it as essentially reflec- 
 tive, i.e., as essentially the position of one thing through 
 the position of another different from it, we find no difficulty 
 
CHAP. VIII. THE ANALOGIES OF EXPERIENCE. 583 
 
 in the counter problem, how thought can enable us to com- 
 bine as necessarily related, the elements which we at first 
 apprehend as different and external to each other. Now, the 
 category of reciprocity is the highest category in which this 
 reflective movement expresses itself; and it includes, therefore, 
 the elements involved in the categories of substance and 
 causality, in which the emphasis is laid upon the identity and 
 the difference respectively. It may be said, therefore, to solve 
 the difficulties which they leave unexplained, and to bring to 
 light what in them is latent. But, though it thus carries us 
 beyond the other categories of reflexion, it does not remove the 
 essential defect of reflective thought itself. This becomes 
 manifest if we consider that in reciprocity, the identity is pre- 
 supposed and latent ; or, what is the same thing, we seem, in 
 thinking things under this category, to start with a difference 
 of substances which externally determine each other, though 
 ultimately we are driven by the latent logic of the category to 
 look for such a thorough-going reciprocity in the determina- 
 tion of these substances, that their independence disappears 
 and gives way to a correlation of factors, neither of which can 
 exist without the other. And the only problem that remains 
 is the problem how the unity or identity which we have thus 
 reached should have expressed itself in such a duality of forms ; 
 a problem, however, which is not difficult to solve, if we con- 
 sider the correlativity of these forms, and the fact that the 
 unity or identity is a mere abstraction when we do not regard 
 it as expressing itself in their duality. 
 
 Another way of indicating the defect of these categories, is to T^^^J^^^{^ 
 point out that in them all Kant is seeking to make intelligible SSerlnof ™" 
 
 , , . -i ■ i. which is to be 
 
 the consciousness of change, or, m other words, to bring it mto reduced to 
 
 ° identity. 
 
 relation with the unity of thought. Yet, ultimately, the ex- 
 planation of change given by means of these categories seems 
 to ehminate the fact to be explained by reducing cause and 
 effect to identity, and resolving all difference of the elements in 
 this identity into a difference of factors which are correlative 
 
584 THE CRITIQUE OP PURE BEASON. book I. 
 
 with each other, and only ideally separable. In this way we 
 carry back the difference to unity, but we do not explain how, 
 out of the unity, the difference should ever arise. For, as the 
 category of causality loses its meaning whenever we cease to 
 hold to the idea of disparate phenomena, of which one disap- 
 pears when the other comes ; so the category of reciprocity 
 loses its meaning when we cease to hold to the idea of 
 independent substances which maintain their independence 
 even while they act and react on each other. The category of 
 reciprocity, therefore, still contains a contradiction, as it starts 
 with a difference which it explains away or idealises, yet 
 without enabling us to conceive of the identity to which it 
 brings us as itself the source of the difference, or as an identity 
 which determines or differentiates itself. It is, therefore, a 
 category in which we cannot ultimately rest satisfied ; for, 
 though it brings together the identity of substance and the 
 difference of causality, it still brings them together in an 
 imperfect and self-contradictory way, and it also must find its 
 explanation in a higher category. 
 This contra- What that higher category is, we may see if we consider 
 
 diction can ° a J > J 
 
 ?n 'aWgher'"'^ that, uudcr the category of reciprocity, objects are considered in 
 ca egory. themsclves and in their relations to each other, without being 
 also considered iu relation to the thought for which they are. 
 For, so long as this is the case, to the ideal unity and identity 
 to which in their difference they point, must always be opposed 
 their real difference and change, and they cannot be conceived 
 as factors in an organic whole, which in all its change remains 
 one with itself. Now, Kant at least points the way to this 
 higher view of things when he maintains, not merely that the 
 conscious self can draw from itself categories under which the 
 manifold of sense may be determined as a world of coexisting 
 objects with changing states, but that the consciousness of self 
 presupposes such a consciousness of objects. For this means, 
 not only that the unity of the self can, as it were, lend its own 
 nature to the manifold to such an extent as to bind its different 
 
CHAP. VIII. THE ANALOGIES OP EXPERIENCE. 585 
 
 elements together as parts of one whole, but that it Twwsi do so, 
 , in order to bring that manifold into relation to itself. It means, 
 therefore, that the determination of a manifold under such 
 categories is a necessary step in the development of the 
 consciousness of self; or, in other words, that that conscious- 
 ness is not possible except in correlation with, though in 
 distinction from, a world in space and time. And, if this is 
 the case, then the manifold which is supposed to be externally 
 given, as the material out of which the spontaneity of the 
 understanding may manufacture a world, cannot really be an 
 external or alien datum for the mind which thus operates on it. 
 That which the subject needs, if it were merely as the 'Anstoss ' 
 in reaction upon which it becomes conscious of itself, cannot 
 really be a foreign or alien matter ; nor can the process by which 
 it is " given," be regarded as the intrusion of something external 
 upon the unity of pure thought. Now, if this is so, the categories 
 of necessary relation, (which presuppose the manifold as given, 
 and only determine it so far as to bring its, elements together 
 by an external necessity, and therefore in the consciousness of 
 an objective world which is opposed to the self,) will not be the 
 highest categories by which we can determine objects. They 
 will appear to be so, only so long as attention is directed mainly 
 to the connexion of objects with each other, and not to their 
 necessary relation to the unity of the self. For, when this last 
 relation is brought into view, it appears at once that, as the 
 consciousness of self presupposes the consciousness of objects, it 
 presupposes the manifold to be connected as well as a synthesis 
 of connexion established in that manifold. And the complete de- 
 termination of the object in relation to the conscious self, will not 
 be reached, so long as the manifold is connected together merely 
 as constituting a world of objects in necessary external relations 
 to each other. For such complete determination, it will be 
 required also that the world of objects should be recognised as 
 the essential counterpart of the conscious self, and therefore as 
 in all 'its difference involving the unity of self-consciousness, i.e., 
 
586 THE CRITIQUE OF PURE BEASON. book i. 
 
 a unity which in self-consciousness becomes clearly revealed or 
 manifested. The stage of thought in which we see the world 
 as a whole of separate parts, which externally determine each 
 other according to necessary laws, will not, therefore, on this 
 view, be the highest stage of knowledge. It will be only 
 the preparation for a still higher stage, in which the external 
 y/ connexion of things with each other, is regarded as itself the 
 manifestation of an essential unity which they have with each 
 other in virtue of their common relation to the self. Now, it 
 was Kant's great work to show that the determination of objects 
 as such, and as necessarily connected parts of one objective 
 world, is a determination which they have only in relation to 
 the unity of the self And it would seem as if, having taken 
 this step to correct the abstraetness of the view of the world as 
 a whole, independent of the intelligence, he must go on to view 
 that world as an organic whole in which the unity of the self 
 manifests itself. But, for reasons which we shall have to dis- 
 cuss in the next chapter, Kant stops at this point, and having 
 recognised that the world is a system of necessity only as an 
 object for our intelligence, and therefore that as such a system 
 it is merely a phenomenal world, he does not go' on to modify 
 this view of it, by the idea that the unity of intelligence under- 
 lies it. Hence, he does not see that " freedom is the ultimate 
 truth of necessity." On the contrary, he regards the intelligence 
 as conscious of itself in relation, but only in negative relation 
 to the phenomenal world of necessity, and therefore as rising to 
 a consciousness of freedom only as it abstracts from that world. 
 While, therefore, in virtue of the opposition between the unity 
 of the self and the diversity of the world, the world is regarded 
 as phenomenal, Kant does not see that the very principle which 
 enables us to recognise it as phenomenal, enables us to pene- 
 trate to the noumenon which is contrasted therewith. The 
 necessary relation of the object to the self casts for Kant no 
 new light on the object, which remains in opposition to the self, 
 though necessarily related to it. It is this dubious position in 
 
CHAP. VIII. THE ANALOGIES OF EXPEEIENCE. 587 
 
 which the object is left in relation to the mind as essentially, 
 yet merely negatively related to it, which occasions the main 
 difficulties that appear in connexion with the Principles of 
 Modality. And it is this also which finally gives rise to the 
 Antinomies of Eeason ; for the contradictions which arise in 
 relation to the phenomenal world, are due to the fact that we 
 are obliged to regard it as a whole in itself, apart from the self, 
 with a view to its determination as an object of knowledge, 
 while yet we are unable to regard it as such a whole, because 
 of its essential relation to the self. 
 
588 
 
 CHAPTER IX. 
 
 THE POSTULATES OF EMPIRICAL THOUGHT. KANT'S VIEW OF 
 IDEALISM, AND OF THE EELATION OF INNER AND OUTER 
 
 SENSE. 
 
 Import of the 
 categories of 
 Modality. 
 
 The Postulates 
 of ISmpirical 
 Thought as 
 summing up 
 the result of 
 the Analytic. 
 
 " rpHE categories of modality have this peculiarity, that they 
 do not in the smallest degree add to the conception to 
 which they are attached as predicates, but merely express its 
 relation to the faculty of knowledge, . . . Hence, the 
 principles of modality are nothing but explanations of the con- 
 ceptions of possibility, actuality, and necessity in their empiri- 
 cal application ; and therewith, at the same time, restrictions of 
 all the categories to their purely empirical use, so as to exclude 
 any employment of them to determine things in themselves. 
 For, if we are not to take them in a merely logical sense, in 
 which case they would only express analytically the form of 
 thought, but to apply them to the possibility, actuality, and 
 necessity of things, then it must be in reference to possible 
 experience and its synthetic unity that we use them, as being 
 that in which alone objects of knowledge can be given." ^ 
 
 The principles of modality, in fact, gather up in themselves 
 the results of the Criticism of knowledge, which has been going 
 on in all the previous pages of the Critique; for the main 
 business of the deduction of the categories and of the principles 
 of pure understanding, was just to vindicate these conceptions 
 1 A. 219 ; B. 266. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 589 
 
 and principles in relation to possible experience and to confine 
 them to such experience. At every step, therefore, we had to 
 turn the ontological question as to the possibility of things into 
 the critical question of their possibility as objects of knowledge : 
 or, in other words, we had at every step to set aside the 
 ordinary dogmatic method of determining known or knowable 
 objects without considering their relation to the knowing sub- 
 ject. For, in answer to the ontological question, Kant maintained 
 that we can say nothing, except so far as it is one with the 
 logical question. In short, we can assert the possibility of that 
 which is not, and the impossibility of that which is, contradic- 
 tory with itself. The unintelligible cannot exist ; for the 
 assertion of its existence has no meaning. We can go beyond 
 this only by considering the relation of objects to our faculty 
 of knowledge. When we take this relation into account, we 
 see that for the consciousness of objects as such, it is necessary 
 that we should bring together conception and perception, 
 determining the " blind '' perceptions by pure conceptions, and 
 giving a real content to the " empty " categories by means of 
 the forms and matter of sense. Knowledge was thus shown 
 to involve at once the " idealising " or universalising of the 
 particular matter of sense by thought, and the " realising " or 
 particularising of the universals of thought by sense. The 
 possibility, actuality, and necessity of objects of knowledge 
 may, therefore, be determined by reference to the conditions of 
 knowledge, as well as in relation to the conditions of pure 
 thought ; and we can say, not only that objects, as conceived 
 must not be self-contradictory, but that, as known, they must be 
 conformed to the a priori conditions of perception and thought. 
 Being so conformed, they will be possible ; if given in percep- 
 tion, they will be actual ; while they will be necessary only if 
 they are connected with what is already determined as actual 
 by means of th6 universal conditions of experience. 
 
 '' Difficulty as to 
 
 We found, however, that there was a special difficulty Kant^^^^ew of 
 running through all the Critig^iC as to the connexion of the two KLut'y?"""* 
 
590 THE CEITIQTJE OF PURE BEASON. book i. 
 
 elements which have to be distinguished from and related to 
 each other in experience ; and we may expect that this difficulty 
 will recur in the discussion of the Postulates of Empirical 
 Thought, which gathers up its general result. This difficulty 
 comes into view the moment we look closely at Kant's defini- 
 tions of possibility and actuality. " That which agrees with 
 the formal conditions of experience (of perception and concep- 
 tion) is possible." " That which agrees with the material condi- 
 tions of experience (of sensation) is actual." Does this mean 
 that these two sets of conditions may be separated — that, on the 
 one hand, we can determine a thing as possible which we have 
 not presented to us as actual, and that, on the other hand, we can 
 have presented to us as actual that which we have not deter- 
 mined as possible, i.e., which is not determined as such by the 
 principles of pure understanding ? Or, does it mean simply to 
 call attention to two aspects in which we must regard every 
 object of experience as perceived and as conceived ? The 
 answer to these questions cannot be given by a simple " yes '' 
 or " no," without a consideration of the different forms of 
 language and the different ways of representing the connexion 
 of the two elements to which Kant is led, according as he 
 approaches conception from the point of view of perception, or 
 perception from the point of view of conception. 
 Is it peicep- In the Aesthetic, Kant deals with the ordinary view of per- 
 
 tionthat •' '^ 
 
 retoou"^'" ceptions as immediately given ideas of particular objects. 
 
 conceptions to xx i i r j.' t'jt r» ij*j. 
 
 objects, or Heucc, hc spcaks 01 perception as directly referred to its 
 
 conception in • .7.7 
 
 that enables obiect, and of Conception as indirectly referred to its obiect 
 
 us to refer our •> ' J^ ■? •> 
 
 obj^cte ?°°° *° through the perception. The effect, however, of the Aesthetic 
 was to take over to the side of the subject not only the sensa- 
 tions but also the forms of time and space by which the object 
 was supposed to be. determined in perception, and to leave the 
 object in itself utterly indeterminate, as a thing from which all 
 perceptive predicates are withdrawn. But the next step, taken 
 by Kant in the Analytic, involves the further reduction of the 
 perception to a mere manifold, which has no unity, and is, 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 591 
 
 therefore, no idea of anything. To give it such unity, and 
 therefore to refer it to an object, the conception must be 
 predicated of it. We are, therefore, required to conceive our . 
 knowledge of an object as such as due to a process of judgment, 
 
 . in which the perception is brought under the conception of it ; 
 yet, at the same time, we are taught that it is only through this 
 process that there can exist for us any object about which it is 
 possible to judge. This, however, is made more intelligible by 
 the interposition, between sense and thought, of a synthesis of 
 the imagination which is conformable to the conception, 
 or of which the conception is the principle " represented 
 in its universality." Imagination, acting in conformity with 
 the conception upon the manifold of sense, is supposed to pro- 
 duce an image, to which the conception can then be attached as 
 a predicate. And we can see that the judgment so made is not 
 a process of analysis, if we observe that the detachment of the 
 conception from the perception, and the reference of the per- 
 
 ;; caption to the conception, are not two independent processes, 
 hut different aspects of one process of differentiation and 
 integration, which is involved in the transition from a sensitive 
 to a thinking consciousness. 
 
 If we try to work out this view on the lines of Kant we find why both 
 
 •^ these modes or 
 
 the old modes of expression taking new meanings. To say that be^rtiflTd""" 
 conception is attributed to the object through perception ^ 
 remains true, only if we do not forget that it is just in relation 
 to the conception which is thus separated from it, and as de- 
 termined by that conception, that a sensation becomes the per- 
 ception of an object. Hence, we have two apparently con- 
 tradictory forms of expression — (1) that the understanding by 
 means of its conceptions refers our perceptions to objects, and 
 (2) that conceptions are referred to objects only indirectly 
 through perceptions. The former mode of expression is pre- 
 ferred whenever Kant has to show that "perceptions with- 
 out conceptions are blind " ; the latter when he has to show 
 that " (jonceptions without perceptions are empty." For, in so 
 
592 THE CRITIQUE OF PURE REASON. book I. 
 
 far as it is the recognition of the conformity of the synthesis 
 of perception with a rule derived from a conception which 
 ^ enables us to objectify our perceptions or refer them to objects, 
 it may be said that the object is not perceived but only con- 
 ceived ; while, in so far as the images of perception are supposed 
 to be formed by synthesis according to a rule in order that they 
 may be brought under the conception, it may be said that 
 it is through the perception that the conception gets objective 
 meaning.-" If, however, we do not carefully note the two 
 aspects of the process of determining objects as such — as at 
 once a determination of perception by conception, and of con- 
 ception by perception— we may fail to appreciate either the 
 necessity of the separation of the two, or the way in which 
 they are brought back to unity in the judgment of experi- 
 ence. 
 ^owpeTcep-*" This difficulty of co-ordinating the two sides of Kant's 
 determinedas thought shows itself cspccially in the different views of 
 
 perceptions of • n . 
 
 objects and so reality or actuality, which we get when we are considering 
 
 ™e"an'sworf°* ^ow conccptions, and how perceptions, are to be determined 
 
 as objectively valid or real. In order to the possibility of 
 
 ^ The transcendental object is the correlate of the transcendental subject, 
 which is " neither a perception nor a conception, but merely the form of con- 
 sciousness which is implied in all perception and conception, and enables us to 
 turn them into knowledge." (A. 382.) But the same thing is said of the 
 transcendental object itself, which also is neither a perception nor a conception. 
 It is "a mere X, of which we know nothing, and of which, according to the 
 present constitution of our understanding, we can never know anything, but 
 which comes in only to serve as the correlate of the unity of apperception, and to 
 supply that unity to the manifold of sensible perception by means of which the 
 understanding unites that manifold in the conception of an object." (A. 251.) 
 Kant goes on to say that, " the transcendental object cannot be detached from 
 the data of sense, for otherwise nothing would remain through which it could 
 be thought," and that it is not to be taken as " a particular object given to the 
 understanding alone," and determined for it by the categories ; for " the cate- 
 gories only serve to determine the transcendental object (the conception of 
 something in general) by that which is given in sense. " This must be taken 
 to mean that, while the perception is referred to an object only as brought 
 under the conception, the conception is the conception of an object only as 
 determining the perception. Each, therefore, apart from its relation to the 
 other in the judgment, loses all its meaning. 
 
CHAP. IX. POSTULATES OP EMPIRICAL THOUGHT. 593 
 
 experience our perceptions must be determined by the schema- 
 tised categories. But this means that a sensuous image cannot 
 be recognised as the image of an object, unless it be brought' 
 under the principles of the understanding and determined as 
 an element in the "context of experience." Dreams and mere 
 imaginations refuse to enter into that context, and it is by 
 this that we distinguish them from realities. Hence, we find 
 Kant saying in the Prolegomena that "space itself, with all 
 the phenomena which it contains, are ideas, the objective 
 validity of which is shown by their connexion according 
 to the law of experience, just as the actual existence of 
 my soul is shown by the connexion of the phenomena 
 of inner sense " ; and a few lines after he adds that " the 
 doubt whether experience furnishes secure criteria to dis- 
 tinguish it from imagination"^ is easily removed, as, in fact, 
 it is always removed in common life, by " investigating the 
 connexion of the phenomena of inner and outer sense, accord- 
 ing to the universal laws of experience." It is thus the 
 principles of understanding, and especially the Analogies of 
 Experience, which determine for us what is real and what 
 unreal, or, we should rather say, enable us to distinguish 
 in what sense any particular appearance is real ; for even 
 a dream, recognised as such, is referred to a particular place 
 in the context of experience. In this sense, it is only for 
 L .a thinking consciousness that there can be any question as 
 to the reality and unreality of its perceptions ; and such a 
 consciousness can answer the question only by showing that 
 they are, or are not, really possible, i.e., that they can, or 
 cannot, be connected with other perceptions according to the 
 pure principles of the understanding. 
 
 So far, however, we have been looking at experience from the when we ask 
 
 how concep- 
 
 side of perception. Prom the side of conception we seem to reach iet°^fn™as 
 a somewhat different result. For conceptions have all their objeotelTnd so 
 validity in relation to perceptions; they are, therefore, regarded actual, we get 
 
 •'':'■ ^ a different 
 
 • 1 Prolegomena, § 49, K. IIL 106 ; H. IV. 84. ^''^^^'■ 
 
 VOL. L 2 P 
 
594 THE CRITIQUE OP PURE REASON. book I. 
 
 merely as possible predicates of objects, which become actual- 
 ised only in so' far as their objects are given in sense. Prom 
 this point of view, Kant maintains that no conceptions can 
 be shown a priori to be conceptions of actual or even possible 
 objects except so far as they are necessary to the possibility 
 of experience. Even these are in themselves merely possible; 
 for they can be shown to be necessary " only in relation to 
 something accidental, viz., possible experience " ; and they 
 become actual only when, and because, data of sense are 
 given which can be subsumed under them. From this point of 
 view, then, conceptions as such cannot go beyond possibility; 
 and they cannot go even so far, unless they are conceptions 
 necessary to experience. And it is perception alone which 
 can carry us over from possibility to actuality. Nay, we 
 find Kant even speaking as if perception supplied in itself 
 a sufficient criterion of reality, apart from all determination 
 of it by conception, and as if such determination were neces- 
 sary only to establish the reality of that which is not given 
 in perception. To this view at least such language as the 
 following would be most conformable : " That the conception 
 of a thing precedes its perception, signifies merely its possi- 
 bility. But the perception, which furnishes the matter for 
 the conception, is the one thing that gives it the character 
 of reality. We may, however, recognise the existence of a 
 thing before the perception of it, and therefore comparatively 
 a priori, if only it be connected with certain of our perceptions 
 according to the principles of their empirical combination, 
 i.6., the analogies of experience." ^ Here, therefore, Kant 
 speaks as if the difficulty were only as to the reality of 
 what is not directly perceived, and as if, therefore, nothing 
 were needed to certify for us the reality of that which is 
 perceived or to " distinguish it from an imagination." And it 
 is in harmony with this that in the immediate context we 
 find him asserting that "in the mere conception of a thing 
 1 A. 226 ; B. 273. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 595 
 
 no character which determines it as existing can be in- 
 cluded."^ 
 
 We can understand the possibility of Kant's looking at ^^y ^ant 
 
 ^ 1/ D never com- 
 
 the subject in these two opposite ways only if we remember fit'contradk- 
 
 ,1.1 .i.. J. , . T , . tioii between 
 
 the reciprocal presupposition oi perception and conception these two 
 
 answers. 
 
 in the judgment of knowledge, and the way in which Kant 
 tries to explain it, now from the point of view of perception, 
 and now from the point of view of conception. The effect 
 of this is, no doubt, a formal contradiction which Kant himself 
 never disentangles, but which we must endeavour to dis- 
 entangle, if we would do justice to him. The important 
 point, as 1 have already indicated, is to distinguish the oSoi 
 avui and the oSw Korrw, i.e., the regressive method in which 
 Kant follows experience back to its a priori conditions, and 
 the progressive method in which he advances from the a 
 priori conditions to the experience which depends upon them. 
 In the former point of view, as we have seen, Kant begins 
 with a supposed fact, say, the fact that " something happens," 
 and asks how it can become a fact for me. In this way he 
 tries to show the necessity of the forms of sense, of the con-. 
 
 ' In the Rejlexionen Kant's (ii. § 1095) we find Kant saying that " Possibility- 
 is thought without being given, while actuality is given without being thought, 
 and necessity is given through being thought." This statement (as Erdmann 
 points out in his note) finds its parallel in the idea of the Prolegomena, accord- 
 ing to which sense-perception enables us to make judgments which become 
 judgments of experience by the aid of the categories ; and also in the language 
 of the Critique (A. 90 ; B. 122), which, however, may be regarded as a pre- 
 liminary statement. There are other passages where the same ideas are 
 traceable. On the other hand, it is the prevailing view of the Analytic 
 ■ that perceptions (not conceptions) are in themselves the subjective deter- 
 minations of the mind, which are referred to objects as they are bound 
 together in accordance with the principles of the pure understanding, and 
 that, apart from this, they are 'for us as good as nothing.' We cannot bring 
 these two statements into line with each other, unless we regard them as 
 an imperfect expression of the truth that neither perception nor conception 
 has any meaning for us except as combined in the judgment. In the unity 
 of the judgment, the perception may be regarded as expressing the reality, 
 and the conception the possibility ; but they have that force only as factors 
 in the jij|(igment of necessity, which, as Kant says, is "existence as given 
 through its very possibility." 
 
596 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 ceptions of the understanding, and finally of the unity of 
 apperception, — all in order that the supposed facts of nature 
 may become facts for me. Kant thus vindicates the a ^priori 
 generally as necessary to the possibility of experience. What 
 he does not recognise is that, as these conditions are con- 
 ditions without which the facts would not exist for us, i.e., 
 as without these conditions the facts would lose for us all 
 the determination which they seemed at first to have, we can 
 no longer speak as if there were an element in them which 
 is given apart from the process whereby they become known. 
 What we have been doing in the deduction of the a priori prin- 
 ciples has been simply to show that the object has no meaning 
 except as an existence for a self, and that, therefore, we are 
 obliged to add on the new qualifications which are involved in 
 this relation, if we would not take away from it all the quali- 
 fications which it has as an object. But, having followed 
 out this argument to its result, we cannot be allowed to turn 
 round, and speak as if there were something given — if 
 not an object, a perception, or if not a perception, a manifold 
 of senfee — apart from its being known. This residuum of the 
 original fact, however, Kant always preserves ; and, consequently, 
 the actuality or reality seems to him to be something addi- 
 tional to the complete thought or real possibility of the object. 
 Hence, while it is said that being is "for us as good as 
 nothing," if it is not being-for-thought, yet leing as given is 
 still conceived to be something over and above 'being-for-ihought. 
 The idea of a logical subsumption of perception under con- 
 ception holds its ground, even when it is confessed that 
 jxrception is nothing for us except as subsumed, and it is 
 all but confessed that the thought is nothing for us (it is 
 said to be empty) except in relation to the perception it is 
 supposed to subsume, i.e., that thought has no meaning except 
 as referring to being and including it. In other words, as 
 there is no being except being-/or-thought, so there is no 
 thought which is not the thought of being, — being manifest- 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 597 
 
 ing itself in consciousness to another, or, in self-consciousness, 
 to itself. 
 
 If the 6S69 avw stops short of its eoal, the 656? Karoo conception 
 
 -*■ ^ ' presupposes 
 
 hecomes impossible. This is what is meant by Kant's asser- scetarperoep- 
 tion that, even if we have the complete thought of an object, consciousness 
 
 presupposes 
 
 even though we fully understand the conditions of its possi- •™<* '™- 
 
 ° *' ^ scends the 
 
 bility, we are not able apart from perception to say that it is of objMta™^^ 
 real. If this meant merely that speculation has no meaning 
 except as an interpretation of facts, or that all it can do is to 
 discover what is really meant by these facts, the statement 
 would be quite unexceptionable. It would in that case be 
 only an explanation of the true meaning that underlies the 
 saying, nihil est in intellectu quod non prius in sensu. It would 
 be another way of stating that in our experience self-conscious- 
 ness presupposes the consciousness of objects, and cannot be 
 other than a return upon the unity involved in that conscious- 
 ness.-^ If we could, therefore, conceive of a thought which was 
 not relative to perception, or did not include it, such thought 
 could never by its own movement make its way to perception. 
 If self-consciousness could be divorced from the consciousness 
 of objects, the breach could never be healed. But no such breach 
 ever exists. No doubt, in the first appearance of self-conscious- 
 ness, the self is abstractly opposed to the object ; and in the 
 advance of reflexion, this opposition reappears as the distinction 
 of conception from perception — conception in which the mind 
 seems to abide with itself, from perception in which it seems to 
 come into contact with something different from itself. Kant's 
 
 1 We must, however, always remember that the consciousness of objects 
 does not exist apart from this return. Apperception, from one point of view, 
 is only the interpretation of perception ; but this interpretation exists always, 
 though it may be as yet in what Kant calls a, " weak " form, in all that can 
 be properly called perception. " This consciousness may be but weak, so that 
 we are aware of it only in the eflfect and not in the act itself, i.e., that we do not 
 connect it immediately with the genesis of the idea ; but, in spite of this dis- 
 tinction, one consciousness must always be discoverable," (when a manifold is 
 bound together in the idea of an object,) " however it may want outstanding 
 clearness," i.e., the clearness which belongs to the conception as at once dis- 
 tinguilhed from, and referred to, the perception (A. 104). 
 
598 THE CBITIQUE OF PURE REASON. book i. 
 
 separation of the analytic movement of thought from the 
 syntlietic movement of knowledge is another form of the same 
 fundamental contrast. But, if we ask how the analytic unity 
 of thought becomes possible, we are obliged to say with Kant 
 that it is mediated by the synthetic unity of the consciousness 
 of objects.^ And it is difficult to see how, consistently with 
 this view of analysis as being synthesis and something more 
 Kant could still speak of it as something less. In any case, 
 the absolute opposition of possibility to reality, of conception 
 to perception, stands or falls with the opposition of analysis to 
 synthesis ; and the same considerations that are fatal to the 
 latter must be equally fatal to the former. It is true, indeed, 
 that we can think of something as possible which we do not 
 know to be real, and which we cannot know to be real or 
 unreal till new perceptions have enabled us to verify or reject 
 it. Every scientific hypothesis furnishes an instance of such 
 conceptions which are merely possible. But this means only 
 that conception develops with, and in relation to, percep- 
 tion ; which is of course necessary, if conception is but 
 perception returning upon the unity implied in it. If we do 
 not know all the conditions of the possiljility of an object, we 
 do not know its reality ; and if a new perception enables us 
 to discover such reality, it is by enabling us to complete our 
 thought of its possibility. Kant, however, speaks as if that ' 
 thought might be completed, and yet the object be conceived 
 as merely possible so long as it is not presented in sense : and 
 on this it is a sufficient criticism to point out that in sense 
 apart from thought no object can be presented, and that, if we 
 take sense as meaning perception, the presentation of the. object 
 in perception necessarily implies an addition to our thought of 
 it, and, — in the case where a perception enables us to verify a 
 hypothesis, — this addition must be just that which is needed to 
 correct and complete our hypothesis, or our conception of the 
 real possibility of the object in question. If we could know 
 
 iR. 1.33?iofc 
 
CHAP. IX. POSTULATES OJb' EMPIRICAL THOUGHT. 599 
 
 the whole ■ conditions of an object apart from perception, we 
 should know its reality ; and that we cannot do so, merely 
 means that there is no such thing as thought apart from 
 perception, no thought which is not the return of perception 
 upon itself. On the other hand, through the presence of an 
 object in sense, we shoiild not know its reality if such presence 
 were anything externally added to thought ; for, in that case it 
 would be at most the presence of a sensuous image, which 
 could not tell us anything about the possibility of any object 
 as such.^ 
 
 1 In the Works of T. H. Green, II. 63 seq. , this distinction is fully 
 discussed, and Kant's position in relation to it clearly stated. The fol- 
 lowing passage may be quoted : — " It is not that there are three sorts of 
 objeots, the possible, the real, the necessary, but that the real world is 
 known to us through a succession of experiences, which the unity of the under- 
 
 '* standing renders a whole of mutually qualifying elements, and that thus to us 
 in any stage of experience, there are many possibilities of which we cannot yet 
 say whether they are real ; they are possibilities, as not being inconsistent, 
 according to the formal conditions of experience, with our hitherto experience, 
 but possibilities of which we cannot say that they are real, because our hitherto 
 experience is only a part of possible experience." 
 
 " Kant's error (I think) lies in treating such possibility as ' objective possi- 
 bility.' It is not objective possibility, unless consistent with the whole order 
 of the world as it is, and whatever is possible in this sense is also real. In 
 this latter (the true) sense of the 'objectively possible,' it is quite true that 
 the object, when from possible it becomes real, is 'not further determined,' 
 but only so because in this sense the possible and the real are the same. To 
 the objectively ■possible in the above sense, the occurrence of a sensation (a new 
 
 ) perception) on our part makes no difference. To the svhjectively possible it may 
 
 make a. great difference. It may verify or falsify an hypothesis. A ' subjec- 
 
 ! tively possible ' conception must precede every experiment. The experiment 
 
 shows whether a relation of phenomena, supposed to be possible, is real or not. 
 Through it nothing becomes real which was not real before. ' Is it not the 
 case, however,' it may be said, ' that through it what was conceived as possible 
 comes to be conceived as real, and that without any change in the content of 
 
 K • conception ? ' No, because the experiment always involves the analysis of some 
 phenomena not analysed before ; it enables you to judge that a really always 
 accompanies 6, whereas before you only guessed it, because, after a cnicial 
 experiment, you are able to set aside all conditions in the complex phenomena, 
 which included b, and which a had been found to follow, except b itself." 
 
 " Thus, taking the ' possible object ' in one sense, it is quite true that the 
 occurrence of a perception corresponding to it makes no difference to its con- 
 tent; but of such an object it is unmeaning to say that, through the occurrence 
 of perceptions, from being possible it becomes real. Taking ' possible object ' 
 in another sense, it is quite true that the occurrence of a perception converts 
 
600 THE CRITIQUE OF PURE REASON. book i. 
 
 Smperieot While, then, it must, in a sense, be admitted that perception 
 of tws^"™' is logically prior to thought, and consciousness of objects to 
 Kant. self-consciousness, it is not allowable for us to treat the former 
 
 as having any residuum in them which is separate from, and 
 inaccessible to, the latter. We can, however, see how Kant 
 should hold that it was so, when we consider how he regarded 
 pure self-consciousness as an analytic judgment, and how con- 
 sequently he conceived thought as confined to mere analysis of 
 all matter given to it. It is true that he gives us the means 
 of correcting his error when he speaks of the analytic unity of 
 apperception as implying or presupposing the synthetic unity 
 (of the consciousness of objects). But, as I have often pointed 
 out already, he regards the relation of the former to the latter 
 as negative, just as if self-consciousness were reached merely by 
 abstraction from some part of the contents of the consciousness 
 of objects. And, as the method of mere abstraction cannot be 
 reversed, he holds that the necessity of the oSos koltw, i.e., of 
 the reference of perception to thought, does not involve the 
 necessity of the oSos avoo, i.e., of the reference of thought to 
 perception. He is obliged, indeed, to regard thought as taking 
 on a synthetic function in relation to the data of sense ; but he 
 holds (1) that this synthetic function is to be explained by the 
 reflexion of the pure or formal unity of thought (in the analytic 
 judgment) upon the forms of sense, and (2) that it cannot 
 yield any results higher than the mathematical principles and 
 the analogies of experience, i.e.. it cannot do more than enable 
 us to establish necessary relations between elements given as 
 disparate from thought and from each other. The categories of 
 external relation are thus its highest products, the furthest point 
 to which pure thought can go in introducing its own unity into 
 the data of sense. And in these categories, after all, we have 
 only analogues of the unity of thought. To go further than 
 this, and establish an organic unity between the difi'erent 
 
 its possibility into reality, but, in doing so, it further determines the con- 
 ception of the object. " 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 601 
 
 elements thus taken up, is a step from which we are pre- 
 cluded by the nature of space and time, the conditions of 
 the given matter which thought tries to absorb. And it 
 follows, of course, that we are not able to establish such a 
 unity between thought itself and the matter it takes xip. In 
 dealing with the matter of sense, thought is, as it were, an 
 episcapus in partibus infidelium, and cannot thoroughly convert 
 that matter to the likeness of its own transparent unity : 
 rather, it is itself subdued to the element it works in, and 
 produces only a synthesis or external combination which is 
 |;>|^i analogous to that unity. And it is this that explains the final 
 recoil of thought upon ideas set up by itself as a goal for its 
 endeavour, as an aim which it must follow in all its empirical 
 work, but with which the result of that work remains always 
 incommensurable. 
 
 If we reject this view, on the ground that the seeming con- of'^ji^'^'""'''^^ 
 tingency of objects in relation to the conscious self, and the experience "^ 
 
 ... . . . cannot be 
 
 ; : consequent opposition between perception and conception, is regarded as 
 , simply due to the ordinary abstract view of things as unrelated po'siwe. 
 . to the subject for which they are, we must also reject Kant's 
 idea that conceptions in general, and especially the conceptions 
 which are the conditions under which objects are determined as 
 such, are themselves merely possible. The universal conditions 
 of all experience cannot themselves be regarded as particular 
 conditions, which require to be united with other conditions in 
 order to constitute knowledge. It is true that there is no 
 meaning in the universal except in relation to the particular, 
 any more than in the particular except in relation to the 
 universal ; but this only proves that we cannot trace experi- 
 ence back to a thought and a perception, which exist independ- 
 ently of each other. The transcendental deduction, while it 
 shows that particular facts or objects exist for us only through 
 the conceptions of the pure understanding, forbids us to turn 
 our hand and speak of these conceptions as hypothetical till 
 real objects are given in sense. For, on the one hand, it is 
 
602 THE CRITIQUE OF PURE REASON. book L 
 
 only as already determined by those principles that sense can 
 supply any particulars that could be supposed to verify them, 
 and such verification, therefore, would come too late or would 
 involve a circle in reasoning. And, on the other hand, the 
 transcendental reflexion which calls attention to the universal 
 as presupposed in the particular, as it reaches the universal 
 only through the particular, does not need to look to the 
 particular for any further verification of the reality of the 
 universal. What appears in particular cases as such verifica- 
 tion is always, as has been already shown, the completion and 
 correction of our conceptions by new conceptions ; and if the 
 new perceptions which are said to verify our conceptions, did 
 not also bring new conceptions, i.e., if they did not add to 
 our conception, they could not bring to it any verification which 
 it has not got already simply in being thought. 
 The modal jf howcver, this view be adopted, a great change will be 
 
 principles ' ' jr ' o o 
 
 regarded as produccd in the Kantian conception of the principles of 
 organic unity modality. For these principles are supposed to call attention 
 
 of objects 
 
 with each to the nature of known obiects as determined by their relation 
 
 other and the •' -^ 
 
 intelligence, j.^ ^.j^g subjcct, and especially to the limitation of knowledge to 
 phenomena, which arises out of this relation. But in what 
 sense is this limitation to be understood ? It does not mean, 
 we have seen, that we are limited to the knowledge of objects 
 given as such in sense, for no objects are so given. It means 
 that objects exist for us only through a synthesis of the 
 manifold of sense in relation to the unity of the conscious self, 
 and that, therefore, the determination of these objects in them- 
 selves and in their relations, as constituting one world of 
 experience, cannot give us a complete knowledge of them, till 
 we have realised that it is a determination for the self and by 
 the self. When, however, we thus regard all objects as objects 
 for-a-self, our conception of these objects and their relation 
 becomes essentially altered. Thus, the perception that the 
 action and reaction of objects upon each other only exists for 
 a conscious self and cannot be separated from their relation to 
 
CHAP. IX. POSTULATES OF EMPIKICAL THOUGHT. 603 
 
 such a self, should lead us to recognise that, beneath the appear- 
 ance of external action and reaction, there is a deeper community 
 or unity of objects, of which such apparent external determina- 
 tion by each other is the manifestation. It should prepare us, 
 in fact, for the discovery that the external relations of objects 
 are the result of a deeper internal or organic relation of them, 
 which is hid from us only by the abstract way in which the 
 ordinary consciousness and also science deal with objects, with- 
 out regard to their relation to the self This, Kant, of course, 
 does not recognise ; but he at least recognises that the relation 
 of objects to the self cannot be brought under the same cate- 
 gories as those which determine the relations of objects to each 
 other fmr the self. And it is one of the chief objects of his 
 critical work, to show that the consciousness of this relation of 
 objects to the subject makes a complete change in our conception 
 of these objects, authorising us to bring them under determina- 
 tions to which they could not be subjected if they were regarded 
 as things in themselves. But he uses this idea only as a justi- 
 fication for the application of the Mathematical Principles 
 and the Analogies of Experience. That is, he attempts to 
 show only that objects are necessarily determined in relation 
 to each other by these principles, because they are deter- 
 mined in relation to the self. But he does not see that the 
 perception of the necessary relation to the self of the objects 
 as so determined, throws a new light on the objects themselves, 
 and alters our view of their relations. It is, however, obvious 
 that, if we say that objects are determined, e.g., by the Analogies 
 of Experience, only in relation to the self, we are forced to 
 ■ change our views of these very Analogies, and to recognise that 
 the objects which, under these Analogies are regarded as ex- 
 ternally determining each other, have really a deeper community 
 or unity with each other. 
 
 Of this point I shall speak further at the end of this chapter. ^«^^^°° j^^y 
 For the present, it is sufficient to repeat that the reason why ^^^ "^''• 
 Kant fefuses to advance to such a conception, which yet he was 
 
604 
 
 THE CRITIQUE OF PURE REASOK. 
 
 BOOK I. 
 
 General and 
 particular 
 contingency 
 of objects in 
 relation to the 
 intelligeucc. 
 
 the first to suggest, lies in his rooted prejudice as to the 
 analytic character of thought in itself. In spite of this analytic 
 character, Kant regards thought as furnishing in relation to the 
 forms and matter of sense, certain principles of connexion, by 
 which the manifold of sense may be determined so as to give 
 rise to a consciousness of objects ; but here he stops. Having 
 brought the extraneously given manifold into unity under the 
 conception of an object, the thinking subject, according to Kant, 
 is conscious of itself in relation, but only in negative relation, to 
 that object. And it still appears as an accident that an object 
 should be present to it at all ; or, to put it more definitely, that a 
 manifold of sense should be present to a subject (which is in itself 
 a pure unity) as the material out of which an objective conscious- 
 ness can be developed. This mere contingency of objects of 
 experience in relation to the subject as such, is maintained by 
 Kant in spite of his own recognition that, apart from the 
 consciousness of objects, the consciousness of self would be 
 impossible. 
 
 The contingency of experience in relation to thought as 
 stated by Kant, has two aspects ; it is a general and a 
 particular contingency. In the former aspect, it is asserted to 
 be a contingency that any materials for experience should be 
 given at all ; in the latter aspect, it is asserted to be a 
 contingency that just such materials should be given as are 
 actually supplied in sense. In other words, there is nothing 
 in the nature of the universal determinations of possible 
 experience that makes it necessary that any object should be 
 given, any experience be realised by us ; and there is nothing 
 in these determinations that makes it necessary that the 
 experience realised should be just such as we actually have ; 
 for the nature of our intelligence merely fixes certain conditions 
 as to the kinds of experience which are possible, but withiu 
 these conditions, leaves everything free. Of the first of these 
 kinds of contingency we have already spoken, and it has been 
 pointed out that Kant's owh assertion, that the consciousness of 
 
CHAP. IX. POSTULATES OP EMPIRICAL THOUGHT. 605 
 
 objects is presupposed in the consciousness of self, contains its 
 refutation. For that cannot be said to be contingent for the 
 self without which there would be no conscious self, and strictly 
 speaking, no self at all. We can therefore say, that for the self, 
 the object is a necessary presupposition ; and from this we may 
 immediately infer that it must be given in conformity with the 
 a priori conditions of perception and conception ; or, in other 
 words, the sensational matter taken up into experience must be 
 capable of a synthesis by the imagination which is in conformity 
 with the principles of the pure understanding. But beyond 
 this, the matter may be regarded as free from any determina- 
 tion by thought, and therefore the particular laws of nature, as 
 well as the particular objects to which they apply, might, so 
 far as we can see, be very different from what we find them 
 empirically to be. 
 
 Now, without for the present questioning the general point The idea of the 
 
 general con- 
 
 of view from which this distinction is drawn, we mav notice tmgenoy of 
 
 " objects IS 
 
 that in the second edition of the Critique, Kant introduces a thetMondedi- 
 new determination into the transcendental view of the objects conception oT 
 
 the relation of 
 
 of our knowledge, to which all particular objects are required i^^^r and 
 to conform, and, in fact, extends in one important aspect the «''p«"™<'«- 
 a priori determination of experience. In the first edition he 
 had maintained that all particular experiences must conform to 
 principles which were developed by schematising the categories 
 in reference to time. And, as time is the general form of all 
 perceptions, outward and inward, it seemed to follow that these 
 principles apply indifferently to external and internal experience 
 alike ; though there may be certain other conditions derived 
 from the nature of space, which must be taken into account in 
 applying them to outer experience. But in the second edition 
 of the Critique, we find Kant maintaining that it is only to 
 outer experience under the form of space that the principles of 
 pure understanding must be applied, at least in the first 
 instance, and that any application of them to inner experience 
 is secolidary, if indeed it be possible at all. He thus, as I have 
 
606 THE CRITIQUE OF PURE REASOK. book i. 
 
 said, attempts in one way to extend his conception of the n 
 •priori conditions of experience, in so far as he makes it include 
 the determination by the categories of the matter of sense as 
 given under the form of space, as well as under the form of 
 time ; yet, in another way, he limits that conception, in so far as 
 he asserts that the principles so developed apply only to outer 
 experience, and not in the same sense to inner experience. 
 Outer and inner experience cease to be two parallel kinds of 
 knowledge, and the latter becomes posterior to, and dependent 
 on, the former. The result of this i,s that Kant takes up an 
 altogether different attitude towards the so-called " Idealist," i.e., 
 towards those who deny the reality of external objects as such, 
 or at least the possibility of proving it. To show this, it will 
 be necessary to trace the development of his thought in some 
 detail. 
 Lambert's The method of dealing with the Idealists which Kant 
 
 difficulty jis to 
 
 time. adopts in the first edition of the Critique, may be described 
 
 by saying that he partly lowers internal experience to the 
 level of external experience, and partly raises external ex- 
 perience to the level of internal experience. Lambert's 
 criticism of the Dissertation had already brought Kant face 
 to face with a difficulty that arose out of his view of space 
 and time as forms of perception. Lambert had found it 
 comparatively easy to regard space as merely an ideal form, 
 seeing that it is a form under which wc know things other 
 than ourselves ; but he had found it hard to take the same 
 view of time as the form of inner experience, and to admit 
 that we know even our own states merely as phenomena 
 and not in their real nature. 
 b'l^uaiS'^ In answer to this objection, Kant argues that, as we can 
 inner and determine the relation of things in time a priori, time can- 
 not be a determination of things which belongs to them 
 apart from our perception. Self-perception, as the perception 
 of our states in time, is thus affected by a form which re- 
 duces the object-self to the phenomenal level, to which 
 
 experience. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 607 
 
 things without us are admitted to be reduced by the fact 
 that we know them only through our own ideas. The 
 mediation of a form, under which things are brought in 
 being perceived, is imfhed in both cases; and if it reduces 
 the one into a phenomenon, it must reduce the other 
 also. The error which has led to the idea that things in 
 space are inaccessible to that immediate perception, in which 
 our inner being is present to us, is explained more fully in 
 Kant's comment upon the fourth Paralogism of Eational 
 Psychology. It had been argued by Descartes and others 
 that our consciousness of external things is problematic, 
 because they are outside of us, and therefore we can know 
 them only by their effect upon us. For, " the inference 
 from effect to cause is always doubtful," and it might be 
 that, after ' all, what we referred to something external; was 
 to be regarded simply as the result of something internal. 
 In this way Descartes was led to maintain that all we can 
 be sure of is the play of our own states of consciousness. 
 To this Kant answers that, while in a sense it is true that 
 we cannot go beyond ourselves and what is immediately 
 present in our consciousness, yet the inference drawn in 
 regard to things in space is erroneous ; because it involves 
 a confusion between the externality of things to each other, 
 an externality which belongs to them only as they are 
 present in our consciousness, and an externality to conscious- 
 ness, which would be the negation of such presence. But, 
 if it is once seen that things can be determined as in space 
 only as they are objects of perception to us, it becomes 
 impossible to make such a confusion, or to think that " in 
 : space " means " out of consciousness." The " outness " is 
 recognised as existing between objects v^rhich are for con- 
 sciousness, and not between consciousness and objects. Now, 
 this is what Descartes and his followers overlooked, and 
 therefore their " Transcendental Kealism led directly to Em- 
 pirical* Idealism." Considering space to be a determination 
 
608 THE CRITIQUE OF PURE REASON. book i. 
 
 ,of things in themselves, not only in their relations to each 
 other but also in their relation to the mind as a thing in 
 itself, they were obliged to deny that we can have real 
 knowledge of that, which ex hypothesi is outside of our minds 
 just as one external object is outside of another. On the 
 other hand, Kant claims for his own Transcendental Idealism 
 that, just because it teaches that reciprocal externality can 
 belong to objects only in relation to us who perceive them 
 as in space, it shows that we perceive external objects in 
 space just as directly and immediately as we perceive the 
 states of our own consciousness in time. It may, indeed, 
 be objected that we are liable to illusions, and seem some- 
 times to perceive objects in space which do not reaUy 
 exist; and, that being the case, it may be asked whether 
 space itself with all its contents may not be illusbry. But, 
 the answer is that the very possibility of such an illusion 
 presupposes the reality of space. An illusion is a particular 
 phenomenon referred to the wrong place in the context of 
 experience, but it presupposes the existence of that context. 
 To suppose that space and all outer experiences are illusions 
 would be virtually to deny that consciousness makes a dis- 
 tinction between the different elements of its content, which 
 it obviously does make, on the ground of the nullity of an- 
 other distinction which it cannot possibly make. We cannot 
 divide the contents of consciousness into a content which is 
 in consciousness, and a content which is out of consciousness ; 
 but we certainly can and do make a distinction between 
 objects in space and the successive perceptions or feelings 
 of the sensitive subject, which, as such, are not apprehended 
 as in space, and therefore not as spatially related to each 
 other. It is the vain search for an impossible object ex- 
 ternal to me in the sense of being out of my consciousness, 
 which throws doubt upon the simple fact that there are 
 objects in consciousness which are represented as external 
 to each other, and as different from the ideas which are 
 
CHAP. IX. POSTITLATES OF EMPIRICAL THOUGHT. 609 
 
 referred to my subjective states. But, if there were objects 
 outside of me in the former sense, they would not be in 
 space; for nothing can be in space which is not represented 
 as being in it. It is true that, as in the case of the illu- 
 sion mentioned above, we may represent as existing in space 
 what is not really there ; but this merely means that we 
 have given a wrong interpretation to certain sensations, or 
 that we have referred a phenomenon to the wrong place in 
 the context of experience. But the correction of such an 
 error can come only from the necessity we are under of 
 finding a place for everything in that one context. If I 
 see something which I believe to be a tangible object on 
 the window pane, and if when I put out my hand to touch 
 it I find nothing, I discover that the supposed object does 
 not conform to relations of sight and touch which I have 
 otherwise learnt to hold good for objects of external per- 
 ception ; hence, I am obliged to find another interpretation 
 for the visual appearance in question. But, both the mistake 
 and its correction would be impossible, if there were no 
 ordered connexion of external experience to which pheno- 
 mena could be referred. And if there be such an ordered 
 connexion, that is all that we mean by existence in space. 
 In like manner, all that we mean by existence in time is, 
 that there is a similar connexion of phenomena as repre- 
 sented in it. On the other hand, if by the existence of 
 objects we mean their existence as things in themselves, 
 apart from all conditions in which they are perceived 
 and known, we are not able to say whether the unknown 
 ground of external phenomena is, or is not, different from 
 the unknown ground of internal phenomena ; for we are 
 not able to say anything definite about either. " If any 
 one should ask whether my Psychology is dualistic, I an- 
 swer, ' Certainly, but only in an empirical sense ' ; by 
 which I mean that in the connexion of experience matter 
 is really presented to outer sense as a phenomenal substance, 
 VOL. I. 2 Q 
 
610 THE CRITIQUE OF PURE REASON. book i. 
 
 just as the thinking ego is really presented to inner sense 
 as a phenomenal substance. In both cases, the phenomena 
 must be connected together according to the rules which 
 this category of substance brings into the connexion of inner 
 and outer perceptions, with a view to the constitution of an 
 experience. But, if it be proposed to me to widen the con- 
 ception of Dualism and to take it, as is usually done, in a' 
 transcendental sense, I must point out that by such a course 
 I should deprive myself of any ground for maintaining 
 Dualism, in preference to either of the other alternative sys- 
 tems of Materialism or Spiritualism. To assert such a 
 Dualism would simply be to confuse a difference in our 
 way of representing objects (which in themselves remain 
 unknown) with a difference of these objects in themselves, 
 I myself as represented by inner sense in time, and objects 
 in space outside of me, are indeed specifically different phe- 
 nomena, but this does not enable me to think of them as 
 different things. The transcendental object which lies at the 
 basis of external phenomena, like that which lies at the 
 basis of internal perception, is neither matter nor a thinking 
 being in itself, but a, to us unknown, ground of the phe- 
 nomena which enable us to attain our empirical perception 
 of each of these kinds of object."^ 
 Kant supposes Thcsc Considerations are then applied by Kant to remove 
 
 this view to L J. 1/ 
 
 dlffiouity'^i^ to the difficulty which, on the ordinary theory, was felt as to 
 of mind and" the connexion of mind and body. The whole' difficulty, he 
 thinks, lay in the hypostasis of external phenomena, which 
 were supposed to be things existing outside of our con- 
 sciousness with the same quality which they have in our 
 consciousness. For, when the relation was thus conceived, the 
 question how the things produced the ideas corresponding 
 to them, was beset with difficulty. " To an external object 
 none but external [effects, i.e., changes of place, could be 
 attributed ; nor could it possess any powers except powers 
 
 1 A, 379. 
 
CHAP. IX. POSTULATES OP EMPIRICAL THOUGHT. 611 
 
 to determine spatial relations. But, in us, all effects pro- 
 duced are thoughts ; and to these no relation of place, no 
 change of place, and no spatial form can be assigned. 
 Hence, we completely lose the guiding thread which, in ex- 
 ternal perception, is supplied by the survival of causes in 
 their effects, when we try to follow external objects in their 
 effects upon inner sense. We should, however, reflect that 
 bodies are not things in themselves which are immediately 
 present to us, but phenomena of an unknown object ; and 
 that motion is not the effect of this unknown cause, but 
 merely the phenomenal appearance of its influence upon our 
 sense. Body and motion, then, being neither of them things 
 outside of us but only ideas within us, we do not need to suppose 
 that the motion of matter produces ideas in us ; but motion, 
 with the matter which reveals itself in motion, is itself a 
 mere idea; and the whole self-made difficulty reduces itself 
 to the question how, and through what cause, the ideas of 
 our sensibility are so connected together, that those which 
 we call external perceptions can be represented according to 
 empirical laws as objects without us."^ When this is un- 
 derstood, it becomes unnecessary to resort to any such hypo- 
 thesis as that of pre-established harmony or that of occasional 
 causes, in order to get rid of the idea of a physical influence 
 of matter upon mind ; for, the difficulty of the latter hypo- 
 thesis lay in the fact, that moving matter was taken for a 
 thing in itself and not for a phenomenon. From the point 
 of view we have now reached, we must give another shape 
 to the problem, and ask how a thinking being can have 
 external perceptions. In this shape, however, as Kant main- 
 tains, the problem is obviously insoluble ; and our referenx;e of 
 each perception to a thing in itself as its cause is merely a 
 way of marking the position of a hiatus in our knowledge. 
 The question, indeed, is one which we are so far from being 
 to answer that we cannot even say whether, if the 
 • 1 A. 387. 
 
612 THE CRITIQUE OF PUBB REASON. book [. 
 
 power of external perception were taken from us, we should, 
 or should not, be able still to think these or other objects. 
 Hence the . j^gt US restate this argument of Kant a little more freely. 
 
 need for a deus o •' 
 
 coMMt'mind On the ordinary view of the relation of mind and matter,' 
 jrmoved^'^ we have to think of the motion of a body as producing sensa- 
 tion, if not perception, in the mind, just in the same way in 
 which it produces motion in another body. But if we adopt 
 such an idea, we necessarily make a leap from a cause to an 
 effect which is not homogeneous with it; we completely "lose the 
 guiding thread which is supplied by the survival of causes in 
 their effects." There is no identity continuing itself in the 
 change, but rather a complete fierajSaa-ts eiy aX\o yevos. This 
 difficulty, however, which the Cartesians and Leibniz sought to 
 evade by the theories of occasional causes and of pre-established 
 harmony, disappears, as Kant here contends, whenever we 
 realise that both effect and cause are phenomenal ; though the 
 former are phenomena presented under the forms of space 
 and time, and the latter are phenomena presented only 
 under the form of time. Putting it otherwise, when we 
 consider that motion is essentially motion for a perceiving 
 self, i.e., is " an idea," and that the supposed effect, the 
 sensation produced, is simply another event, which also is, 
 only as it is for the perceiving self, i.e., also is an idea, the 
 difficulty of connecting these two kinds of "ideas" seems 
 to disappear or to become less important. They both form 
 links in one context of experience, though the one is represented 
 as an event happening to a phenomenal substance determined 
 in space, while the other is represented as happening to 
 a phenomenal substance which has no determination as in 
 space. Both have to be taken as mere phenomena or ideas, 
 though it so happens that the objects in space "have the 
 illusive characteristic that they appear, as it were, to detach 
 themselves from the soul and hover outside of it." While 
 therefore, we cannot understand "how external perception 
 should be possible for us," we do not find any absolute chasm 
 
CHAP. IX. POSTULATES OP EMPIRIOAL THOUaHT. 613 
 
 opening between it and internal perception such that only 
 a deus ex machina can bridge it over. 
 
 This solution seems at first satisfactory; and no doubt it^"*.!?,'' 
 
 ■^ ' possible to 
 
 does remove the ordinary dualistic difficulty, which gave rise Sne/and"'^ 
 to the theories of "Occasional Causes" and " Pre-established explrienoe? 
 Harmony " ; for it shows that the external object, as well as 
 the self as an object, is relative to the self as a subject, 
 and that in that sense both are " ideas." But we soon find the 
 difficulty returning upon us in another shape. For, according 
 to this view, we are obliged to regard g,ll phenomena as 
 constituting one context of experience, in which all inner as 
 well as all outer phenomena are combined according to the 
 principles of the pure understanding. Now, in the first place, 
 such a view, as Kant indicates in the passage just quoted, 
 would compel us to conceive the individual thinking sub- 
 ject as a substance among other substances, acted on by 
 them and reacting upon them. But, even in the first edition 
 of the Critique, reciprocity is treated by Kant himself, as 
 a category which must be schematised not only in relation 
 to time but also to space, and as therefore a category which 
 can be applied only to things which are determined as existing 
 in space. And in this very chapter in which he speaks of 
 the " thinking ego as a phenomenal substance," we find Kant 
 observing that Psychology and Physics differ greatly ; in so 
 far as in the latter science much truth can be developed 
 a priori out of the mere conception of an extended impene- 
 trable thing, while in the former out of the mere conception of 
 a thinking being nothing can be so developed. " ISTow," Kant 
 goes on, "the reason of this is that, though both are pheno- 
 mena, the phenomenon of outer sense has something standing 
 or permanent in it which furnishes a substratum underlying 
 its changing determinations, and thereby puts into our hand 
 a conception fruitful in synthesis, viz., that of space and a 
 phenomenon in it ; while time, the sole form of inner percep- 
 tion, !ias nothing permanent : and so under the form of time 
 
614 THE CRITIQUE OF PURE REASON. book i 
 
 we can have presented to us only the change of determinations, 
 but no object to characterise by their means. For, in that 
 which we call the soul all is in continual flux, and nothing 
 abides, except (if that could be brought into the question), the 
 pure ego, which is quite simple, — in the sense that the idea of 
 it has no content, no manifold, in it, — and which for that 
 reason seems to represent a simple object. It would, however, 
 be truer to say that it designates or marks off the ego 
 from other objects, than that it represents it."^ "For this 
 ego is neither a. perception nor a conception of any object: 
 it is merely a form of consciousness, which is implied in 
 all perception and conception, and enables us to turn them 
 into knowledge — supposing always that something other than 
 the mere ego is given in perception, as the matter for the 
 conception of an object." ^ But, if this be true, then the 
 category of substance cannot properly be applied to the object 
 of inner experience as such, nor can inner experience form part 
 of a connected whole or context of experience, including both 
 inner and outer. 
 Inner expert- The Same conclusion may be reached in another way. 
 
 ence cannot be , ^^ ^ .^ , , 
 
 co-ordinated Inner experience, as Kant describes it, differs m an im- 
 
 with outward 
 
 ™£°tte°same'^ portaut way from outer experience. The elements of inner 
 a'Jurther"^ ™ experience proclaim themselves at once as " mere ideas," 
 
 reflexion. i .i i i 
 
 while the elements oi outer experience appear at least to 
 be something ■ more. They have, as Kant says, the "illusive 
 characteristic of seeming to separate themselves from the 
 soul, and hover without it." But, Kant argues, the external 
 object, like the idea of it, exists only for consciousness ; and in 
 this point of view they are both " merely ideas." Yet, there is 
 a difference in the two cases. I may be conscious of an object, 
 and then I may direct my attention to the . determination 
 of inner sense in virtue of which that object exists for me ; and 
 thus I may distinguish between the idea as a mere status 
 representativus of the subject, and the object which I know 
 'A. 381. 2 A. 382. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 615 
 
 through it. As Kant explains elsewhere, when I apprehend an 
 object of experience and bring it under a conception, or even 
 when I arbitrarily think an object, I determine my inner sense 
 in a particular way in accordance with my thought. But, I can 
 then proceed to observe this determination, and can say, " I 
 have experienced what is necessary, say, in order to apprehend 
 a figure of four sides in such a way as to enable me to 
 demonstrate its properties. I thus get the empiric conscious- 
 ness of the determination of my state in time by thought."^ 
 But this consciousness of the inner determination involved 
 in the determination of the external object, cannot be taken as 
 the consciousness of an object which is co-ordinate with the 
 external object, or related to it as one phenomenon is to 
 another according to the principles of the connexion of sub- 
 stances in one experience. It cannot be so taken, because the 
 former consciousness is a consciousness of what was involved in 
 the latter, a consciousness of the determination of the inner 
 life of the subject, by which the consciousness of an object 
 is realised. Properly speaking, it is only in this last reflexion 
 that the distinction of inner and outer experience emerges 
 while, at the same time, they are seen to be correlates 
 of each other. But it is impossible that the inner process, 
 involved in the determination of the external object as such, 
 can be treated as a phenomenon which is related according to 
 the Analogies of Experience to the object which through 
 it exists for us. The motion of my body by another body 
 external to it, may, because my body is a sensitive organism, 
 be a condition without which that other body would not 
 for me become an object of perception. But the perception 
 itself cannot be connected with the motion, as that motion 
 is with the motions of other bodies. We have here a relation 
 which we cannot explain without a /xeTa/3ao-t? eh aXXo yevos. 
 Without retracting the abstraction by which we think of moving 
 things apart from their relation to sensitive subjects, we cannot 
 * 'R. X. L261; H. IV. 499. 
 
616 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 explain how the motion of an object and perception as a state of 
 the subject should be connected. And with the retraction of this 
 abstraction we have got into a world which cannot be explained 
 by the Analogies of Experience, without the aid of categories of 
 a higher order. In any case, if we think of a motion as 
 producing a sensation, still more a thought, we have brought in 
 a new factor, which cannot be co-ordinated with the phenomena 
 with which we were previously dealing, as they are co-ordinated 
 with each other. The inner experience to which attention has 
 now been directed, is not another experience to be set beside 
 outer experience, but one which includes it and goes beyond it. 
 Inner experience is outer experience, and something more. 
 
 JSTow, it is obvious that between the periods of the first and 
 second edition of the Critique this difficulty, or at least the 
 conviction that inner experience cannot be simply co-ordinated 
 with outer experience on the ground that both are " ideas," was 
 brought home to Kant.^ 
 dkMve^o"''^ Already in the first edition he had, as has been shown 
 recognised the impossibility 6i directly bringing inner experience 
 under the category of substance. Yet, in the ProlegomeTm, in a 
 passage already quoted, he still speaks of the connexion of 
 external phenomena according to empirical laws as " proving 
 their objective truth, just as the connexion of the phenomena of 
 inner sense proves the reality of my soul" ; ^ and he maintains 
 that we have to "investigate the connexion of phenomena in 
 inner sense according to universal laws of experience." But 
 various causes were already forcing him to realise more 
 
 •■ I express myself thus hesitatingly because Kant never completely,realised 
 the result of his changed point of view. He saw, as we shall find, that it is 
 impossible to bring the soul, as the object of inner experience, under the same 
 categories as the object of outer experience ; he saw also that the former is not 
 parallel with the latter, but presupposes it, and is presented to us by a 
 new reflective movement which goes beyond the consciousness of it ; but he , 
 still speaks of the phenomena of the inner life as part of the general mechanical 
 nexus of nature, and indeed it is this that gives its paradoxical aspect to his 
 view of man's life as at once phenomenally subjected to natural necessity and 
 noumenally freed from it. 
 
 2 Prolegomena, § 49 ; R. III. 106 ; H. IV. 85. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 617 
 
 definitely what he meant by the distinction of inner and outer 
 experience, as well as by the distinction of the objects of both 
 from those things in themselves, or that thing in itself, to 
 which both were to be ultimately referred. 
 
 In the first place, as we have seen, he had already observed J^^Jte^i^pIr-' 
 that inner experience cannot be properly brought under the fio^oharaoter" 
 category of substance, nor consequently under any of the 
 Analogies of Experience which presuppose substance. And in 
 writing his Metaphysical Rudiments of Physics, which wa,s 
 published before the second edition of the Critique, he was led 
 by this and some other considerations, which will presently 
 be mentioned, to renounce his first intention of applying his 
 general theory of experience to Psychology as well as to 
 Physics. For, he argues, in order to an a priori knowledge of 
 definite objects in nature, it is necessary that " we should be 
 able to bring the conceptions of the pure understanding into 
 relation with a corresponding a priori perception, i.e., to con- 
 struct the object of the conceptions or set it before us in an 
 image of perception. Now, all rational knowledge by means of 
 the construction of conceptions is mathematical. While, there- 
 fore, a pure philosophy of nature, i.e., a philosophy which 
 investigates the constitutive elements of the general conception 
 of nature, may be possible without mathematics, it is only by 
 means of mathematics that we can reach a pure doctrine as to 
 the action of definite things of sense, whether material or 
 spiritual." ^ Now " Mathematics is not applicable to the 
 phenomena of inner sense and their laws. At least the only 
 law which we can speak of in this connexion is the law of 
 continuity, which is manifested in the process of inner change. 
 By the application of Mathematics to this law, however, we 
 should get an extension of our knowledge of the soul, which 
 would be related to the extended knowledge of body that 
 Mathematics contributes to Physics, very much as the doctrine 
 of the properties of a straight line is related to the whole of 
 * 1 Metaph. Anfangsgrunde Vorrede; R. V. 309 ; H. IV. 360. 
 
618 THE CRITIQUE OF PURE REASON. book I. 
 
 Geometry. For the pure inner perception, in which the 
 phenomena of the soul would have to be constructed, is time, 
 which has only one dimension. Furthermore, even if we 
 regard Psychology merely as a systematic art of analysis or an 
 experimental doctrine, it cannot rise to the level of Chemistry; 
 for the manifold of inner observation cannot be resolved into 
 its elements, except by a merely ideal division. "We cannot 
 keep the parts separate and again unite them at pleasure. 
 Still less can we subject another thinking subject to our 
 experiments in a way that could produce any satisfactory 
 result. Finally, our observation itself alters the state of the 
 observed object and leads us to misinterpret it. Hence, Psy- 
 chology cannot become more than a historic, and so far as may 
 be, a systematic, doctrine of the nature of inner sense, i.e., it 
 can only be a description of the soul as a natural object, and 
 not strictly a natural science of the soul or even an experi- 
 mental doctrine in regard to it."^ To this we have to add, 
 what Kant says in another (already quoted) passage of the 
 same work, that the phenomena of the inner life have no parts 
 outside of each other, and therefore no parts which are sub- 
 stances like things in space. Hence, it is possible to conceive 
 that by a gradual diminution of intensity our consciousness 
 should be reduced to extinction without any annihilation of 
 substance. 
 
 In this way Kant strips the inner phenomena, which he 
 still regards as objects of experience, of the determinations by 
 which outer objects of experience are characterised ; and makes 
 it impossible to treat the object of inner sense as taking its 
 place on equal terms among the objects of experience. 
 Kanf s two- But there is another point of even more importance in rela- 
 
 fold answer to -^ *■ 
 
 treated wm as *i°^ *° ^^^ cvolution of Kant's thought. In the interval 
 er eeian, j^g^^^ggj^ ^j^g ^^q editions of the Critique, Kant was greatly 
 influenced by the way in which the first edition was inter- 
 preted. Kant's reviewer in the Gelehrfe Anzeiger of Gottingen 
 
 ^ V. 310. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 619 
 
 made it his great charge against the Critique that by it outer 
 sense was brought down to the level of inner sense, i.e., that 
 both were represented as having to do with mere " ideas of our 
 minds"; while he neglected to notice how the empirical reality 
 of the objects of experience is connected with the universality 
 of the principles of the pure understanding. He observed, so 
 to speak, Kant's levelling down, but not his levelling up of 
 
 ;• ^ experience. On the other hand, he treated Kant's Transcen- 
 dental Idealism as fatal to the distinction between phenomena 
 and things in themselves. But Kant " never thought of ques- 
 tioning " that distinction ; nay rather — we may even say — it 
 was one of his main objects to establish it, in order that he 
 might be able to limit the application of the categories of 
 
 I ; necessary relation to objects of sense or phenomena. 
 
 In his notice of this criticism at the end of the Proleqomena, (i) Ttat tiie « 
 
 o > pnon view of 
 
 Kant deals with the former of these points. " The principle of and form" rf^ 
 
 SCT1S6 vindi* 
 
 all true Idealists, from the Eleatic school to Bishop Berkeley," oates the 
 
 objective 
 
 may, he declares, be gathered up in the formula that " all ideas If^^^^Til 
 of things that come to us through sense and experience are f" ™p'"<=*i 
 
 •'; ;; illusory, and that truth is to be found only in the ideas of pure 
 understanding or the pure reason. But, the principle which 
 throughout rules and determines my Idealism is, that all con- 
 ceptions of things derived from pure understanding or pure 
 reason are entirely illusory, and that truth lies only in experi- 
 ence." ^ And then he goes on to declare that, while he agrees 
 
 i ;; with the Idealists in holding that " time and space and what- 
 ever exists in either must be regarded not as things in 
 themselves or their properties, but only as belonging to the 
 phenomena of these things," he differs from them, and especially 
 from Berkeley, in holding that, inasmuch as time and space are 
 pure forms of our sensibility, "they with all their deter- 
 minations can be known by us a priori." Now, there can be 
 no criterion of truth unless particular experiences rest upon 
 universal and necessary laws. The effect, therefore, of Berke- 
 1 K. III. 134 ; H. IV. 122. 
 
620 THE CRITIQUE OF PURE REASON. book I. 
 
 ley's view of the empirical nature of our percepti6ns of space 
 (and time) is, that phenomena are reduced to merely illusory 
 appearances. On the contrary, Kant contends that on his 
 own doctrine time and space, in combination with the pure 
 conceptions of the understanding, prescribe laws a prion, to 
 all possible experience, and thus enable us to distinguish 
 reality from illusion. 
 How far is On this wc mav remark that it is only in view of his last 
 
 Kant just to •' "' 
 
 Berkeley? work, the SiHs, that Berkeley can be reckoned, along with 
 Plato and the other Idealists, as a supporter of the doctrine 
 that real things are apprehended by pure reason, while phe- 
 nomena only are known through sense and experience. In 
 Berkeley's earlier works phenomena, or, as he calls them, ideas, 
 appear as the objects of consciousness, the only real objects 
 which there are or can be. It is true that Berkeley goes on 
 to distinguish " notions " from " ideas," and to refer the latter ■ 
 in so far as they are involuntary to God as their author or 
 cause. It is true also that he treats these involuntary ideas as 
 a language by which God speaks to us. In such conceptions 
 we may find the link that connects the Sensationalism of his 
 earlier, with the Idealism (in the proper sense of the word) of 
 his later, works. On the other hand, Kant, from his point of 
 view, has fair ground for arguing that, because Berkeley has 
 not admitted the a priori character of time and space, and of 
 the principles of pure understanding as determining phenomena 
 according to these forms ; because, in short, he has not seen 
 that particular facts as known presuppose universal principles, 
 he has left himself no criteria to distinguish reality from 
 illusion. It is obvious that, if particular perceptions are not 
 referred to anything beyond themselves, they cannot be re- 
 garded as revealing to us any objective reality. So far, then, 
 Kant's point as against his assailant is simply to show that 
 that assailant had not entered into his view of experience, or 
 appreciated his distinction between perceptions or sensations 
 as mere states of the sensitive subject, and these same percep- 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 621 
 
 tions or sensations as referred to objects by means of the 
 general principles of connexion under which they are brought 
 in experience. 
 
 Kant speaks to the same effect in the third Eemark appended Transcendent 
 to the first part of the Prolegomena, where he argues that, ?^ necT4?y to 
 though time and space are taken over to the subjective side, carreaiTty'"' 
 this does not involve the reduction of phenomena to delusive 
 appearances. A doubt on this point might naturally suggest 
 itself to one who came from Leibniz to Kant. For, while to 
 the former perception was only confused conception, Kant 
 maintained that the distinction of conception and perception 
 was not logical, but 'lay in the genetic origin' of each, and that 
 the special character of the former was due to the manner in 
 which objects affected our sensibility. But does not this turn 
 perception into subjective illusion ? Kant answers, no ; for 
 illusion lies not in phenomena, as such, but in the way in which 
 the understanding interprets them in relation to objects. And 
 "illusion or truth will arise according as we are careful or care- 
 less in connecting the perceptions of sense in time and space 
 according to the rules of the connexion of all knowledge." 
 Thus, " though I hold all ideas of sense together with their 
 forms to be merely phenomena, and time and space to be 
 merely forms of sensibility, which cannot be found outside of 
 sensibility in any objects ; and though, therefore, I cannot make 
 use of these ideas except in reference to possible experience ; 
 yet in this there is nothing which should make me hold them 
 to be merely illusory appearances ; for it is not incon- 
 sistent with their being duly connected together in experi- 
 ence according to the rules of truth." ^ The doctrine of 
 the ideality of time and space, indeed, " is so far from reducing 
 the whole world of sense to an illusion, that it alone explains 
 how the whole science of mathematics should be a priori," and 
 yet should contain not mere brain-spun fancies but objective 
 truth. At the same time, it is the sole means of avoiding the 
 • ■ IK. IIL 48 ; H. IV. 39. 
 
622 THE CRITIQUE OF PURE REASON. book I. 
 
 transcendental illusion, which is at the bottom of the Antinomy 
 of reason and which arises directly from the reference of the 
 characteristics of the phenomena of experience to things in 
 themselves. 
 (2;ThatKanfs In regard to the second point, the supposed denial of the 
 
 transcendental 
 
 iu)Teff™t'^the existence of anything but the thinking being and his states, the 
 totween™ phenomena or ideas present to him, Kant speaks also in the 
 and things in third, but morc fully in the second, Eemark to the first part of 
 
 themselves. 
 
 the Prolegomena, What he has to say in these passages is 
 simply that his kind of Idealism does not affect the distinction 
 between things in themselves and phenomena. It had been 
 acknowledged, he points out, even before Locke's time, but still 
 more after it, that the so-called secondary qualities did not 
 belong to things in themselves. The effect of his own Aesthetic 
 had been only to show that the primary qualities also — exten- 
 sion and solidity and all the characteristics that belong to objects 
 as in space — are phenomenal. " If, therefore, it is unreasonable 
 to call him an Idealist who treats colours not as properties of 
 the object in itself, but as modifications to be explained by the 
 nature of the sense of sight, neither can that name be justly 
 applied to me because I find that other properties and, indeed, 
 all the properties that constitute the perception of a body, belong 
 merely to its phenomenon. For the existence of the thing 
 that appears is not thereby denied, when it is shown that we 
 cannot know through sense how it is constituted in itself"^ 
 On the other hand, if it were required that in order to avoid 
 Idealism Kant should admit that " the idea of space not only 
 corresponds completely with the relation of our sense to the 
 object," — which is his own doctrine, — but " that it is exactly 
 like the object " ; Kant answers that this latter assertion is one 
 to which he can '' attach no meaning whatever, any more than 
 he can attach a meaning to the assertion that the sensation of 
 redness is like the quality of cinnabar which excites it in us." 
 The result of Kant's Aesthetic was to take over to the 
 ' Prolegomena, § 13, Remark Second ; R, III. 46 ; H. IV. 37, 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 623 
 
 side of the subject all the primary as well as the secondary- 
 qualities, and to leave nothing for the object in itself 
 except the blank reference to a " something/' of which 
 we can say only that it is and not what it is. Of 
 course, this left unsettled the question how this " something," 
 even as a point of reference, was to be justified — a ques- 
 tion which became not less difficult after it had been 
 , ;; shown in the Analytic that it is by the understanding that per- 
 ceptions are referred to objects. On that question Kant here 
 says nothing, but merely states that, in carrying over time and 
 space to the subjective side, he did not mean to deny the 
 existence of a thing in itself, independent of our sensations or 
 perceptions. He is, he asserts, as far from such a denial, as 
 ■ those who referred the secondary qualities not to the object, 
 but to the subject as affected by it. 
 With the exception of the slight change of phraseology, which Tte Pniego- 
 
 ■*■ ^ o J. o./ J niena does not 
 
 is involved in calling the Cartesian Idealism " empirical," as dSctiy 
 well as " problematical " or " sceptical," and the Berkeleian ArBt edition of 
 
 _ , . . , tha CritiQv.e. 
 
 Idealism " enthusiastic as well as " dogmatic," there is 
 nothing decidedly new in Kant's treatment of the subject in 
 the Prolegomena. But the necessity of defending himself 
 against different misconceptions has forced him to define his 
 position more accurately ; and to bring out clearly the twofold 
 aspect of it, — as, on the one hand, distinguishing between 
 appearance and reality vAthin experience, and, on the other 
 hand, between the phenomenal reality, which is known by us in 
 experience, and the absolute reality of things in themselves. 
 The latter distinction he had not dreamt of denying, but he had 
 hitherto rather taken it for granted than directed any special 
 attention to it, or positively asserted it. The problem of the 
 
 ; i relations of inner and outer experience is not yet touched, 
 though it may be that the discussions just described helped to 
 direct his attention to the subject. And in the only passage 
 
 I f bearing directly on the point, inner and outer experience are 
 treated as organised in the same manner. " Space," we there 
 
624 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 read, " with all the phenomena it contains, belongs to the ideas, 
 the connexion of which according to empirical laws proves their 
 objective reality, just as the connexion of the phenomena of 
 inner sense proves the reality of my soul." "We easily solve the 
 Cartesian doubt (of the reality of outer experience) in common 
 life, by examining into the connexion of phenomena in outer and 
 inner experience according to universal laws." ^ It was after- 
 wards, and probably in connexion with the preparation of his 
 treatise on the " Metaphysical Rudiments of Physics"' that the 
 idealistic question assumed in Kant's mind a new form which 
 made it require a new answer. 
 anotberTiew ^^ ^^^ Kant's Original intention in that treatise to deal not 
 eduion^"^' Only with Physics but also with Psychology ; but, as we have 
 seen, he gave up that idea mainly because in his view inner 
 experience is subjected only to the form of time, and therefore, 
 nothing can be said of the soul a priori except what can be 
 deduced from the fact that its life is a continuous process. 
 Further, either at this time or at least before the second edition 
 of the Critique was published, Kant's attention seems to have 
 been specially directed to the fact that inner and outer experi- 
 ence are not two independent spheres of knowledge. In the 
 first edition, he had already remarked, that " all ideas, whether 
 they have external things for their objects or not, belong as 
 determinations of mind to our inner state, which inner state 
 necessarily falls under the formal condition of inner perception, 
 viz,, time "; and that, therefore, " time is the condition of all 
 phenomena whatever, the immediate condition of the inner 
 phenomena of our soul, and thereby the mediate condition also 
 of external phenomena." ^ But this did not prevent him from 
 speaking of the objects of inner and outer sense as if they were 
 independent objects of experience standing on the same leveL 
 It is true that Kant does not directly apply the categories to 
 determine the relation of inner and outer experience, but he at 
 
 1 Prolegomena, § 49 ; R. III. 106 ; H. IV. 85. 
 
 2 Kritik, A. 34 ; B. 42. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 625 
 
 least speaks as if each might be determined as an empirical 
 connexion by itself; and he nowhere speaks more distinctly of 
 their relation to each other, than in the passage where he tells 
 us that " external phenomena seem, as it were, to detach them- 
 selves from the soul and to hover without," but that this 
 does not hinder them from being phenomena or ideas, like other 
 ideas, which do not so detach themselves. 
 
 In the second edition, on the other hand, Kant seems to see view of the 
 
 second edition 
 
 that inner and outer experience are not two spheres of know- e^erfeSoepre- 
 ledge, separated from each other only as relating to two distinct IxpeST*^' 
 objects which belong to the same context of experience. Inner 
 and outer experience, as he now points out, are identical in 
 matter ; and the determination of what we proleptically call the 
 outer object is presupposed in the consciousness of the process 
 a parte nostra by which this determination is effected. We 
 have, he declares, " from things without us the whole material 
 of our knowledge even for our inner sense." ^ " It is," he says 
 again, "the ideas of outer sense that constitute the proper 
 matter with which we occupy our minds." ^ Now, space was 
 said to be " the form of our receptivity in so far as we are 
 affected by objects and through such affection come to repre- 
 sent them in perception " ; while time in like manner was 
 regarded as " the form, of our receptivity in so far as we are 
 affected by ourselves and thereby come to represent ourselves 
 in perception." "We are not, however, to suppose that these are 
 two separate processes. The true state of the case is exhibited 
 in a passage of the Deduction of the second edition, where it is 
 said that, on the one hand, the affection of our own passive 
 sensibility by the- understanding, which is what Kant calls the 
 " transcendental synthesis of imagination," is necessary ere we 
 can determine any external object as such ; and that, on the 
 other hand, the consciousness of the successive process by 
 which this determination is effected, is what makes us an object 
 for ourselves in inner perception. Kant, indeed, speaks of 
 
 1 B. Pre/ace, XXXIX. ' B. 67-8. 
 
 VOL. I. 2 R 
 
626 THE CRITIQUE OF PURE REASON. book i. 
 
 inner sense being affected in this process, but here he must be 
 regarded as using the term " inner " proleptically ; for if by 
 inner sense we mean the consciousiiess of the self as a subject 
 which passes through these successive states, it is obvious that 
 such a consciousness is the result of reflexion upon the process 
 by which the object is determined. 
 
 " We cannot think a line," says Kant, " without in thought 
 drawing it ; we cannot think a circle without in thought 
 describing it ; we cannot represent the three dimensions of 
 space without setting before our mind's eye three lines which 
 meet at right angles with each other in the same point; we 
 cannot even think of time itself without drawing a straight line 
 as the outward image of time, and directing our attention to the 
 act of synthesis of the manifold whereby we successively deter- 
 mine inner sense, and so to the succession in this determina- 
 tion." We must, therefore, have " motion as an act of the 
 subject," i.e., as " a pure act of successive synthesis by produc- 
 tive imagination of the manifold of external perception," motion 
 " as a synthesis of the manifold in space, ere we can reach the 
 conception of succession, which we do by abstracting from the 
 space-determination thus effected, and attending only to the 
 act whereby we determine inner sense according to its form."^ 
 analysis of' Kant, then, represented the matter as follows : — We have 
 first an affection of sense (which is independent of all activity 
 on our part, and which, for that reason, we refer to the thing 
 in itself). To that affection, as the condition under which it 
 can alone become a perception, or be referred to an object, 
 there attaches itself the form of space. But this reference can- 
 not actually be made unless there be a determination of inner 
 sense, by which the matter so given is successively taken up 
 and combined in the conception of an object. In this process 
 there is a synthesis at once of the matter of sense and of space 
 quanta, which gives rise to a spatial image. In thus deter- 
 mining the object further, we combine one element after another 
 
 IB. 155. 
 
 experience. 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 627 
 
 under conditions of time ; but the consciousness of this succes- 
 sive synthesis, and consequently of time itself, and of the inner 
 life as conditioned by time, is posterior to the consciousness of 
 the object in space and presupposes that consciousness. We 
 become conscious of the affection of ourselves by which we 
 produce the consciousness of the object, only when we direct 
 attention to the process by which that consciousness was pro- 
 duced, and abstract for the moment from the object in space 
 which rises before us in perception by means of the process. 
 Our experience of. ourselves is thus the subjective counterpart 
 of our experience of external objects, and not another experience 
 separate from it. On the other hand, we have to remember 
 that it is only as this subjective counterpart comes into view 
 that objects become external, in the sense of being distinguished 
 from the conscious life of the subject to whom they are suc- 
 • cessively presented ; for " outer " can be thought of only in 
 distinction from " inner." 
 
 The full meanina: of this view of Kant, however, cannot be How tus is to 
 
 ^ be brought in 
 
 seen, unless we combine it with his ideas as to the necessary g^^'t'.™ ?°„ „f 
 i determination of objects by the categories. It is Kant's of pS?e°umier- 
 
 L -IT n • c standing, 
 
 ! I doctrine that no object can be known as such except in so tar 
 as the perceptions, which change upon us every moment, are 
 
 J^ referred to a permanent reality, of which they are regarded as 
 
 s representing the accidents or states. For, if objects were 
 
 regarded as changing with the perceptions, this would be 
 
 equivalent to a denial that objects exist at all. To say that 
 
 an object exists and that it is permanent, amounts from the 
 
 || transcendental point of view to the same thing ; for, to know it 
 as existing I must take it out of its mere momentary presence 
 
 i i in feeling to me as a sensitive subject, and refer it to " con- 
 
 m sciousness in general," i.e., to a consciousness for which the 
 
 particular exists only through the universal or as a determina- 
 
 r tion of the universal. The way in which Kant expressed this, 
 
 as was shown in detail in the last chapter, was to say that the 
 
 conseiousness of the changing is the consciousness of a relation 
 
628 THE CEITIQTjE OF PURE REASOK. BOOK I. 
 
 in time ; that this again can be reached only by a determination 
 of the one time of which all special times are particular limita- 
 tions ; that, further, as time is merely a form of sense which 
 cannot be perceived by itself, there must be something in that 
 which is apprehended as in time to correspond with and repre- 
 sent the unity of time, as distinct from that which represents 
 and corresponds to the succession of times. Or, to put it 
 otherwise, time in itself neither flows nor is permanent ; but we 
 are conscious of its unity as we are conscious of the permanence 
 of objects in it, while we are conscious of its flux as we are 
 conscious of the change of their states or determinations. 
 Kant then proceeds to show that the reference to a permanent 
 object of any determination as a new state or property coming 
 into existence in it, involves its reference to a previous state 
 or determination of it or of other substances as its imiversal 
 condition ; and, finally, that the recognition of the coexistence 
 of objects can only be a recognition of the states of these 
 objects as reciprocally determining each other or finding their 
 universal condition in each other. 
 and^Mrex""^ Now, SO far WO secm to be dealing indifferently with the 
 first rapposed objccts of inner and outer experience. At least, it is only in 
 
 to be equally 
 
 related. dealing with coexistence that Kant found it necessary in the 
 
 first edition to take notice of space as a form of the relations of 
 the objects, or of some of the objects, of which he is speaking — 
 a form which, like time, cannot of itself determine their rela- 
 tions, because it is not perceived by itself ; but which never- 
 theless is a necessary condition under which their relations 
 are determined by the categories. 
 tte°s6 prhioi- ^^^t the " general remark on the system of principles of the 
 thatthey'be^ pure Understanding," which is added in the second edition, 
 
 viewed first as 
 
 the°'^os8ibim ^^™®^ ^^ ^ ^*^P further. In that remark he begins by insist- 
 
 oxperienoe. ™g. ^^ terms similar to those used in the first edition, on the 
 
 need of perceptions to exhibit the reality of the conceptions of 
 
 the understanding, which in themselves are mere forms of pure 
 
 thought or judgment. The proofs of the principles were, as he 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 629 
 
 points out, only deductions of them as conditions of the possi- 
 bility of experience and not derived from conceptions. Thus, 
 we did not get our evidence for the principle of causality out of 
 the mere conception of an event in abstracto as an object of 
 thought, but we showed that it is necessary to the consciousness 
 of it as an object of possible experience. In other words, we 
 asked how an empiric perception could be the means of giving 
 us such knowledge, and we found that it was possible only 
 through its determination by the schematised category. 
 
 But Kant then goes on to add a thought which was not con- 
 tained in the first edition of the Critique. " Yet more note- 
 worthy is it that in order to understand the possibility of 
 things in conformity with the categories and so to exhibit the 
 objective reality of the latter, we need not merely perceptions 
 but even, in all cases, external perceptions. When, e.g., we take 
 the pure conceptions of relation, we find that in order to supply 
 something permanent in perception which corresponds to the 
 conception of substance, and so to exhibit the objective reality 
 of that conception, we require a phenomenon in space; because 
 space alone is determined as permanent, while time, and so all 
 that is in inner sense, is in constant flux. Again, in order to 
 exhibit change as the perception corresponding to the conception 
 of causality, we must take for our example motion, as change in 
 space ; and without such an example, it would be as impos- 
 sible for us to set change before our mind's eye in a perception, 
 as it is to comprehend it in pure thought. For change is a 
 combination of contradictory determinations in the existence of 
 one and the same thing. Now, it is quite impossible for reason 
 without an example, to make it comprehensible that from a 
 given state of a thing an opposite state should follow. Nay, we 
 may add that it is something to which, without a perception, we 
 could attach no intelligible meaning. Now, the perception 
 required is the perception of the movement of a point in space. 
 It is the existence of such a point in different spaces (as a 
 sequfence of opposite determinations) which first enables us to 
 
630 THE CRITIQUE OF PTJEE REASON. book I. 
 
 intuitively realise a change ; for, in order afterwards to make 
 inner changes of state thinkable, we must first image time (the 
 form of inner sense) as a line and the inner change as the 
 drawing of the line — thus calling in the aid of external percep- 
 tion to make comprehensible the successive existence of our- 
 selves in different states. And the reason of this is, that all 
 change presupposes something permanent in our perception 
 without which it could not be perceived as change, and that in 
 inner sense no such permanent perception is to be found. 
 Finally, the possibility of objects that conform to the category 
 of recvprocity is equally incomprehensible to pure reason. 
 Hence, we are unable to understand its objective reality without 
 perception and indeed external perception. For, how are we to 
 think it as possible that, if several substances exist, there 
 should be something in the existence of each which follows 
 necessarily from the existence of the others, so that from a 
 determination of one, we can argue to necessary presence of 
 connected determinations in the others ? And this is what is 
 required for reciprocity, however difficult it may be to under- 
 stand such a relation between substances which, as such, have 
 isolated independence. Hence, Leibniz, when he attributed a 
 certain commercium to the substances of the world, had to bring 
 in the Deity to mediate between them, seeing that from their 
 existence taken by itself such a relation rightly seemed to him 
 to be incomprehensible. We may, however, easily make such 
 community of substances intelligible to ourselves, if we repre- 
 sent them as in space, i.e., as objects of external perception. ' 
 For space contains already in the a priori idea of it certain 
 formal relations of externality, in which lies the possibility of 
 real relations of external action and reaction between sub- 
 stances, and so of reciprocity. In like manner, it might easily 
 be shown that the possibility of things as quanta, and so the 
 objective reality of the category of quantity, can be exhibited 
 only in external perception, and can only in the second instance 
 be applied to inner sense also." ^ 
 
 1 B. 291 seq. 
 
CHAP. IX. POSTULATES OP EMPIRICAL THOUGHT. 631 
 
 If we bring this statement into relation with what has The problem 
 
 of knowledge 
 
 been said before, we see how it is to be understood. The J""**-- '^l'*" 
 Principles of Understanding had to be proved indirectly, betlkenasthe 
 by reference to possible experience. Kant here adds that the pSmty 
 
 of outward 
 
 they can be proved, in the first instance at least, only in '^^vonence. 
 relation to possible external experience; i.e., that we have 
 first to ask, not simply how objects in general with their 
 changes and coexistences are possible experiences, but how 
 outward objects with their external changes or motions and 
 their coexistence in one space are possible experiences. This 
 is necessary, because time itself can be represented only by 
 a line in space ; and this again is a consequence of the fact 
 that time is in constant flux, while space has its parts per- 
 manently determined in relation to each other. Hence, in 
 space alone can we seek for that which corresponds to the 
 category of substance ; and determination by that category 
 is the basis of all other determination of objects as in time. 
 If, therefore, it be necessary that there should be something 
 in objects, as Kant argues, to represent time itself as that 
 in which all special times succeed each other, that something 
 cannot be represented as existing merely in time ; it must 
 be also in space ; for that which is determined as merely 
 in time would be merely changing, or rather it would be 
 mere vicissitude without permanence and so would not even 
 be known as changing. Hence, an experience of things as 
 merely in time would be impossible, or, impossible at least 
 as a primary experience. 
 
 The effect, then, of Kant's " remark " is to show a depend- 
 ence of determination in time upon determination in space, 
 which necessitates a kind of double schematising of the cate- 
 gories in order to their application to experience. Inner 
 experience or experience of our states as successive in us, 
 in so far as it is an experience under the form of time alone, 
 is dependent for its possibility on the possibility of outer 
 expfcrience ; and therefore, immediately and in the first instance, 
 
632 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 our question must refer to the possibility of external experi- 
 ence only. 
 How tar does Does this iu anv way affect Kant's argument ? The answer 
 
 this affect the J •/ a 
 
 fheprindpies? ^^^^ ^® *^^*'' ^hile it does not alter the general argument as 
 to the necessity of the principles of understanding to deter- 
 mine for us objects of experience, it does involve a correction 
 of his first statement of it, and especially of the idea of two 
 separate kinds of experience, each of which has a different 
 object. For, supposing that idea were correct, it would 
 immediately follow that each of these objects must be 
 conceived as connected with the other according to the Ana- 
 logies of Experience. It now appears, however, that the 
 consciousness of the internal object is the consciousness of 
 the process in us as sensitive beings, by which the conscious- 
 ness of the external object is realised. The former is, therefore 
 not a new independent consciousness added to the conscious- 
 ness of the object, but simply a correction of the latter in so far 
 as it leaves out of account the process a parte nostra by which 
 it is realised. 
 
 Purport of the And this brings us to consider Kant's use of the principles 
 
 Refutation of 
 
 Idealism. thus established in his " Eefutation of Idealism." The object 
 of that Eefutation, as he tells us, is once for all to give the 
 coup de grdce to the theory which treated external experience 
 as more doubtful than internal experience, and to prove 
 conclusively that "we have experience and not imagination 
 of external things." This proof, Kant thinks, cannot be 
 given except by showing that " our inner experience, which 
 Descartes did not question, is itself possible only under pre- 
 supposition of outer experience." The dogmatic idealism of 
 Berkeley was already, as Kant held, disproved by the Aesthetic; 
 for Berkeley's disproof of the reality of external objects was 
 really directed against an erroneous view of external experience, 
 which asserted it to be an experience of things in themselves. 
 So conceived, space with all the objects it conditions could 
 easily be shown to be unintelligible and contradictor}'; and 
 
OHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 633 
 
 therefore, " the good Berkeley " might be pardoned for denying 
 their existence or treating them as illusions of imagination;^ 
 for, as Kant argues in another place, the moment we yield 
 to the transcendental illusion, and treat time and space as 
 absolutely real, we find ourselves face to face with the anti- 
 nomies of reason which tempt us to treat them as altogether 
 
 ■ ' illusory. The Aesthetic, abandoning the idea that external 
 
 ; things are things in themselves, is able to establish their 
 reality as objects of experience, and to refute this crude form 
 of Idealism. But this does not yet get rid of the problematic 
 Idealism of Descartes, which asserts nothing but only argues 
 that "it is impossible to prove by immediate experience any 
 existence outside of our own." In the first edition, Kant 
 thought it enough to say that all empirical reality is trans- 
 cendentally ideal, and that, though in this sense ideal, the 
 external is as real, — i.e., it has the same kind of reality — as 
 the internal, since its phenomena form a connected context of 
 experience under the Analogies of Experience. But this still 
 left room for a doubt, in so far as it admitted two kinds of 
 experience, of which the inner included the outer: for why 
 should it not be said that outer experience has no existence 
 
 ^' except as part of inner experience ? Now, Kant attempts to 
 
 ['] meet this difficulty by showmg that we cannot have inner 
 
 experience except on the presupposition of outer experience; 
 
 I and that, therefore, it is vain to attemjjt to reduce the latter to 
 a fiction, which is real only as an element of the former. 
 
 The proof which Kant gives has excited much discussion, and ^^J^jf^g ™ 
 it is undoubtedly expressed with considerable ambiguity. This objeete. *" 
 ambiguity is partly due to the necessity Kant is in of speakmg 
 of external objects as real in opposition to mere ideas, a mode of 
 , I speaking which verbally at least brings him into collision with 
 his own statement that all experience is of phenomena and so, 
 in that sense, of ideas. He has, therefore, been supposed to be 
 here asserting for phenomena a kind of reality which before 
 * 1 B. 71 • Third general Remark to the Aesthetic. 
 
634 THE CRITIQUE OF PUEE REASON. book i. 
 
 he had allowed only to things in themselves ; even although he 
 states explicitly, at the beginning of the Refutation itself, that he 
 is seeking to prove merely that we experience external things 
 and do not imagine them. Now, as we shall see afterwards, 
 there is a certain sense in which we may say that, according to 
 Kant, the reality of external phenomena involves a reference of 
 them to things in themselves. But, immediately and in the 
 first instance, what Kant is attempting to show is that inner 
 experience is so dependent upon outer experience that the denial 
 of the reality of the latter must carry with it the (to Descartes 
 impossible) denial of the reality of the former. What Kant 
 says, in fact, is that there is no determination of any object in 
 time, which does not involve primarily the determination of the 
 object itself as permanent. Now, inner experience has no 
 permanent, because it is an experience of a flux of perceptions. 
 The conscious self, indeed, underlies the flux of perceptions, as 
 the unity in relation to which they are determined, or rather, 
 as the unity which determines them as elejnents in an experi- 
 ence or objective consciousness ; but with the consciousness of 
 the self perceptions can be united only when they are deter- 
 mined as such elements. And, as this unity cannot furnish the 
 manifold through the determination of which in accordance 
 with the categories a consciousness of objects is generated, 
 so it cannot itself be its own primary object ; but that 
 primary object must be something in distinction from which, 
 and in relation to which, it becomes conscious of itself. It 
 follows, therefore, that it is only in so far as we bind to- 
 gether a manifold as in space and determine it by the category 
 of substance, that we can have knowledge of a permanent 
 objective reality, in relation to which we can become conscious 
 of the successive process of our inner life. 
 •*/S"'Hf';* There is a certain difficulty about this argument, which 
 
 of the Refuta- ** o J 
 
 andthidSec™' arises from Kant's isolating the category of substance (as he 
 
 S'atemciTtofit. docs uot isolatc it iu the passage previously quoted). The 
 
 determination of the external object as a permanent substance, 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 635 
 
 however, is for Kant the fundamental basis of all other deter- 
 minations of it : it is that in which it is first distinctly taken 
 out of the succession of our states and set over against them. 
 Hence, according to Dr. Arnoldt, what Kant really seeks to 
 show is that, "while the transcendental unity of my original 
 apperception progressively forms and rounds off one half of my 
 complete consciousness, i.e., all my perceptions of material 
 things and processes in space, into an empirically real sphere 
 of external experience, it also, along with this and by the same 
 continuous process, shapes the other half of my complete con- 
 sciousness, i.e., all my perceptions of psychical objects and 
 processes in time, into another empirically real sphere of inter- 
 nal experience."^ This view I take to be in substance correct, 
 in so far as it asserts that the consciousness of the external 
 world and of our inner life unite to form one experience, 
 "which could not be inward, if it were not at the same time, 
 outward." But such words might easily be misunderstood, if 
 it were supposed that .the two forms of experience exactly 
 
 ' porresponded to each other. That, of course, would not be 
 consistent with Kant's own assertion that the category of sub- 
 stance cannot be applied to the phenomena of inner sense. 
 In the first edition of this book, objection was taken to the 
 
 I Refutation of Idealism, on the ground that Kant's proof that 
 the permanent substance is the necessary substratum of all 
 time-determination, implied that only the changes of that very 
 substance which is perlmanent can be objects of knowledge to 
 us in experience. In the Refutation, however, Kant seems to 
 argue that, while we cannot bring the thinking subject itself 
 under the category of substance, the permanence of an external 
 substance is sufficient to enable us to determine the sequence 
 of states in ourselves as thinking subjects. And this objection 
 will undoubtedly hold good, if we suppose that the self with the 
 
 I states and procession of its inner life is simply to be treated 
 as an object like other objects. The result, however, to 
 • 1 E. Arnoldt, Kant nacJi Krnio Fischer's neuer Parsiellung, p. 38. 
 
636 THE CRITIQUE OF PURE REASON. book i. 
 
 which Kant's argument really points is that in the deter- 
 mination of external objects in space, (the basis of all 
 which determination is the application of the category of 
 substance,) there is implied a process of the inner life; 
 but that the consciousness of that process, which is involved 
 in the distinction of inner and outer life, logically pre- 
 supposes the process itself, and therefore the determination 
 of objects in space as such. The inner life is thus in a sense 
 parallel with the outward, but in another sense, it includes 
 and goes beyond it, implying a further reflexion which is not 
 present in the determination of the outward object as such. 
 Hence, it is only what was to be expected if we find 
 that the categories, which were sufficient to determine the 
 external world of objects regarded simply in itself, fail to be 
 adequate when we come to regard our consciousness of that 
 world as constituting the contents of our inner life. Unfortun- 
 ately, Kant, regarding the movement of reflexion as a movement 
 of abstraction, does not see this. Hence this impossibility — the 
 impossibility of bringing the inner life under the principles by 
 which outward objects are determined — seems to him to reduce 
 Psychology below the level of science, in the strict sense of the 
 word; whereas it really is an indication that these principles 
 are inadequate to any science that does not consider the 
 phenomena of matter abstractly, apart from their relation 
 to life and mind. But, to this point we shall return after 
 we have first discussed another point that requires elucidation 
 in the Refutation of Idealism. 
 the^e%taeio» 1 s^i^i above that there was a sense in which a reference to 
 
 can be said to , i * * j. i i > ±r\ t_ • i* j. i 
 
 refer to oxter- things m themsclves as at the basis oi external experience is 
 
 nal objects as , 
 
 things 'in iiivolved in that Refutation. Thus, Kant says that " if with 
 
 themselves, "^ *' 
 
 the intellectual consciousness of my existence expressed in the 
 'I am ' of pure apperception which accompanies all my judg- 
 ments and acts of understanding, I could at once combine a 
 determination of my own existence through intellectual percep- 
 tion, the consciousness of a relation to something without me 
 
caAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 637 
 
 would not be necessary for such determination." ^ Now, in this 
 it is implied that the division of the consciousness of the self 
 from that of the not-self, and the dependence of the former 
 upon the latter, really result from the fact that affections need 
 first to be given; and that it is only afterwards that the subject, 
 taking them up successively and binding them together into 
 one consciousness of-a world in space, can, in opposition to the 
 world so determined, become conscious of itself as a unity in 
 aU this process. While, therefore, it is not true that Kant 
 supposes us to become conscious of ourselves in opposition to 
 permanent things in themselves, and what he is speaking of is 
 the relation of two empirical objects, or rather of two aspects of 
 our experience \^ yet it is true that ultimately the dualism in 
 experience is in his mind connected with the opposition 
 between the ego in itself and the thing in itself ; for the latter 
 is " the ground " to which the materials of experience are 
 
 IB. Preface., XL. 
 
 ^ The great difficulty in Kant seems to be that he does not distinguish these 
 two views. His argument really points to "two aspects" of experience, 
 regarded as a determination of objects and regarded as involving a subjective 
 process. In this point of view, inner experience is simply outer experience under 
 a reflexion that brings both sides into relation to each other. But Kant con- 
 stantly speaks as if it were a distinction of two objects in one experience, though 
 he qualifies this by pointing out that the object of inner experience cannot be 
 determined in the same way as the objects of outer experience. The inade- 
 quacy, however, as I contend, belongs here not to the object of inner experience, 
 but to the categories under which it is proposed to bring it. As inner experience 
 is outer experience and something more, it requires higher categories to explain 
 it. The suggestion of some of the above remarks is due to Dr. Staudinger's 
 Nawmerm, but since writing them I have read a luminous article by Dr. Vaihinger 
 on Kant's Refutation of Idealism, in the Strasshurger Abhandlungen znr Philo- 
 sophie (1884), in which the most of the difficulties in relation to Kant's language 
 are discussed with great acuteness. Dr. Vaihinger dwells especially on the fact 
 that Kant in the first edition was discussing mainly the relation of all experience 
 to the transcendental ego, and, therefore, generally spoke of all phenomena as 
 ideas; while in the second edition he dwells more upon the opposition of ideas, 
 as states of the empirical ego, to external objects, both of which are in a sense 
 " ideas " as they are existences for the self. The confirmations of this view 
 which Vaihinger derives from the imperfect treatise on which Kant was 
 engaged in the last years of his life are very remarkable. He notices also 
 that in that treatise Kant shows a tendency to approximate to Fichte, which 
 confirms the view suggested above. 
 
638 THE CRITIQUE OF PURE REASON. book l. 
 
 attributed, just in so far as these materials are passive affections, 
 given to the mind in sense and not supplied by its own 
 spontaneity. And the work of that spontaneity is simply to 
 construct the materials given into an experience of external 
 objects, in opposition, and in relation, to which it then becomes 
 conscious of the process of self-determination involved in that 
 construction. 
 How Kant Jq what scusc Can we recognise this view of Kant as a true 
 
 proceeds from ° 
 
 torf^™"'"' conception of the relations of inner and outer experience ? Por 
 object as an the Ordinary consciousness, the object stands in relation to the 
 
 idea or pheno- 
 menon to the subiect, but is quite independent of it: and the idea that it is 
 
 determination i ^ ' ± x 
 
 objecUn"dis- SO independent readily clothes itself in the metaphor of spatial 
 our ideas. extsmality. The object is thus conceived to be outside of the 
 subject, as one object in space is outside of another. On this 
 the first criticism which was made by Kant from his " transcen- 
 dental " point of view was that the object is out o/the conscious 
 subject only as being an object for it, which it distinguishes 
 from itself; but that it cannot possibly be spatially external to 
 it, since objects are spatially external to each other only /or 
 consciousness, and there is no meaning in speaking of their 
 being spatially outside of consciousness or of consciousness as 
 being spatially outside of them. Indeed, such a relation of 
 externality between consciousness and objects, if it were con- 
 ceivable at all, would make it impossible that the latter should 
 be objects /or it. But this suggests the further thought, that the 
 consciousness of the self and of the not-self are correlative, t.«., 
 that it is only in relation to the object determined as the not- 
 self that we are conscious of ourselves. We are conscious 
 of ourselves only as we distinguish the object from our- 
 selves and determine ourselves in distinction from, yet in 
 relation to, it. Now, Kant's way of recognising this truth we 
 have already seen. He conceives of the self as a unity, which 
 yet cannot be conscious of itself except in relation to a mani- 
 fold which it successively takes up into itself and combines in 
 relation to itself, in this manner constituting an object which it 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. I5'6\) 
 
 distinguishes from itself and from the process of its own inner 
 life. In this way, beginning with the dualism of the ego and 
 its affections, to which a parte nostra the form of space is con- 
 ceived to attach itself, he is led to suppose that, on the one hand, 
 the ego acting through its categories successively combines the 
 matter so given in relation to itself and so becomes conscious 
 of an objective world in space, while again, in opposition to the 
 world in space so determined, it by a further process becomes 
 conscious of itself and its process in time. 
 
 The error which underlies this view is one to which reference Btt^™^?* 
 
 xnis vis vv IS 
 
 has frequently been made, and one the explanation of which confu8ion"f* 
 
 1- pT7->iii *^® regressive 
 
 carries us very deep into the mechanism ot Kant s thoughts, method with 
 
 •^ -*- a method of 
 
 It may be shortly restated thus : — Kant conceives that inner abstraction. 
 experience is something more abstract than outer experience, 
 just as the pure consciousness of the self is something more 
 abstract still, which is reached by going back to the ultimate 
 unity involved in all experience. The ego, acting upon inner 
 sense and taking up the given manifold, is conceived as con- 
 structing outer experience, which again gives rise to a conscious- 
 ness of inner experience, when we abstract from the result and 
 attend only to the process whereby outer, experience is formed ; 
 and this, iinaUy, gives rise to a consciousness of the pure self, 
 when we abstract even from this process and attend only to the 
 unity which makes it possible. In truth, however, as was shown 
 in a former chapter, we cannot go back on the process whereby 
 intelligent experience was formed, in the way of tracing how 
 thought, successively taking up the data of sense, determines 
 them by the categories, and how in this way a merely sensitive 
 consciousness is turned into an intelligible experience. If we 
 attempt to do so, we inevitably become the victims of the 
 dilemma that the data of sense so taken up must be conceived 
 as, prior to the application of the category, either having or not 
 having the qualification which the category brings to it ; and 
 to admit this dilemma is to make the process of experience 
 eith«r useless or impossible. We can escape such an alternative 
 
640 THE CRITIQUE OP PURE REASON. BOOK l. 
 
 only if we remember that the development of consciousness and 
 self-consciousness is one in which there is no possibility of 
 setting one element or process as distinctly prior or posterior 
 in time to another. Por, it is only _ as separated from the self 
 that the object can be related to the self, and only as related 
 to the self that the object can be separated from it. What, on 
 the one side, is a reference of a state of the self to an object, 
 may be described on the other side, as a separation of the self 
 from the object. Perception and conception thus come into 
 being as at once separated from, and referred to, each other; 
 or if we say that the former exists prior to the latter, it cannot, 
 as so existing, be conceived as having the characteristics which 
 it has when so distinguished and referred. Development 
 involves not only an addition to, but a transformation of, that 
 from which the development begins. "We may say, if we like, 
 that, if we had a series of sensations by themselves, these must 
 be taken up and combined in relation to the unity of the self, 
 ere they could give rise to the consciousness of an object ; and 
 we may say further that, only as this is done, can we rise to a 
 consciousness of the self as distinguished from the object. We 
 must, however, remember that the latter process is not a new 
 process which begins when the other is done ; but that it is only 
 for the self-conscious ego that the object is fully determined 
 as an object, and that the construction of what is called outer 
 experience is the construction of inner experience as well. Or, 
 to put it otherwise, it is the full development of inner experi- 
 ence and the correction of the partial abstraction by which 
 outer experience is, in its earlier imperfect forms, separated , 
 from it, that first show us what outer experience really is. 
 Ideas cannot Kant thiuks of the object of inner experience as the sncces- 
 
 be treated aa 
 
 thinking "sub- ^^'^^ °^ °^^ °^^ idcas as states of our individual subjectivity. ■ 
 inner"iife*M a Properly speaking, however, ideas cannot be treated as states ; 
 
 series of aucli .,..-,,,..., i 
 
 states. of an mdividual subjectivity, but only sensations ; for ideas 
 
 imply a reference to the " I think " on the one side, and to the 
 object on the other. An idea always stands for something is 
 
CHAP. TX. POSTULATES OF EMPIRICAL THOUaHT. 641 
 
 a symbol of something else than itself, to the ego that has the 
 idea. But for that reason, it must be contemplated as other 
 than a state of the subject for which it is. In becoming an 
 idea, therefore, a state of the subject has ceased to be such a 
 state; and, on the other side, the subject in turning a sensa- 
 ytion, which may in a sense be called a state of the feeling 
 subject,^ into an idea, has ceased to be a merely feeling subject, 
 and become a conscious self. Hence, though we may say that 
 our sensations are the means whereby we know the world, we 
 must recognise that, in so far as they remain sensations, there 
 is no world for the ego to know and no ego to know it. It is 
 a paralogism, therefore, when Kant ranks sensations as a kind 
 of ideas, and treats sensations and ideas indiscriminately as 
 states of the self. For the self is a subject of ideas, and so in 
 the proper sense a self, only as it separates between itself and 
 what are called its states and turns them into ideas through 
 which it is conscious of objects. Now, if this be true, we 
 cannot go back on any constructive process, in which an in- 
 telligible experience is manufactured by the understanding out 
 of a series of sensational states. This, indeed, is involved 
 directly in Kant's admission that the ego is conscious of itself 
 only as it is conscious of the identity of its act in determining 
 objects. It follows that the mere flux of sensations in us as 
 sensitive subjects can in no way become objective for us as 
 such a series. In that sense inner experience, if we can call it 
 so, has already disappeared in the development of a thinking 
 self. Our inner experience is just our outer experience on its 
 inner side, or it is an experience in which that inner side is 
 specially reflected on. And, on the other hand, we must 
 remember that, though such distinct reflexion may be want- 
 ing, there is no outer experience which is not also an inner 
 experience ; or, in other words, that the determination of things 
 
 ^ I.e., it would be such a state if we could suppose that there is such a thing 
 as a merely feeling subject (which is not implicitly a thinking subject), or that 
 there is such a, thing as a mere sensation (which is not implicitly a percep- 
 tion). ' 
 
 VOL. I. 2 S 
 
642 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 as objects in time and space through the .categories cannot 
 be separated from a consciousness, though it may be an 
 undeveloped consciousness, of their relation to the subject, 
 which in distinction from them is conscious of itself We 
 can no more have an outer experience without an inner 
 experience, than we can have a consciousness of the mere 
 particular as such without the universal, though in both 
 cases it is possible that we may have the former without 
 reflecting upon the latter. 
 The mistake of Now, in reflectin" upon the inner life, as such, we are 
 
 Berkeley, and ' n f > 
 
 tioS*ofit?"°°' necessarily led to retract the abstraction under which the outer 
 life is generally regarded, i.e., to take into account the fact that 
 outward objects exist only for a subject. But, in doing so, we 
 are apt with Berkeley to fall into the opposite abstraction of a 
 merely inner life which has no reference to an outer life. For, 
 how did Berkeley reach his so-called Idealism ? The effort of 
 Locke to remove from external objects all that is merely sub- 
 jective, had led to a view of those objects in which they are 
 divested of all secondary qualities, and in which the objective 
 world is reduced to a universe of mere matter and motion, a 
 Newtonian system of mechanically related substances. Berke- 
 ley's criticism of this view is simply to point out that for a 
 sensitive subject such a world can exist only through its own 
 affections, and therefore cannot be known to exist apart from 
 them. The Kantian answer is that, while for such a subject 
 there would be no external world as such, neither would there 
 be any consciousness of sensations as states of the self The life 
 of a purely sensitive being is not for it an inner life, i.e:, not a 
 consciousness of a series of states of its own being, any more 
 than it is a consciousness of an outer world of objects. On 
 the other hand, the self-conscious being which has an inner life, 
 cannot separate it from the outer life which it presupposes. Its 
 inner life is not the consciousness of a series of sensations as 
 such, but of perceptions or ideas which refers to external 
 objects. We are indeed entitled, with Berkeley, to correct the 
 
uhat. ia. l-USTUliATES OF EMPIRICAL THOITGHT. 643 
 
 abstraction of the ordinary consciousness, by which external 
 objects are taken as things in themselves which have no essen- 
 tial relation to the sensitive subject. And we are entitled 
 with him to reject the distinction of the primary from the 
 secondary qualities of things as objective from subjective ; for 
 feelings of touch are no more independent of sense than any 
 other sensations. But then we are entitled equally to point 
 out that, if we reduce the inner life to mere sensations, it ceases 
 to be a consciousness of an inner any more than of an outer 
 world ; and, therefore, such a reduction cannot be said to prove 
 that there is no outer world, but only an inner world. We 
 cannot say that the inner series of sensations constitutes 
 the fact, and that whoever goes beyond it to assert an outer 
 world is going beyond that of which we have immediate con- 
 sciousness ; for we have not in this sense a consciousness of 
 our inner life except as mediated by a consciousness of the 
 outer world. And, if the process whereby a thinking con- 
 sciousness goes beyond sensations in referring their content to 
 external objects is to be treated as illegitimate, equally must 
 we treat as illegitimate the process by which this sensitive 
 content so transformed is referred to a self. To reduce ex- 
 perience to a purely inner experience, would, therefore, be 
 suicidal, if it meant the reduction of the external object to a 
 mere inward state of the subject : for, with this reduction, the 
 subject itself would disappear, or, what is the same thing, 
 would cease to be object to itself. Hence, the transcendental 
 reflexion which calls attention to the relativity of the external 
 object to the subject, must not be interpreted as if it reduced 
 that object to the feelings of a sensitive subject ; for these do not 
 constitute an inner any more than they constitute an outer ex- 
 perience, and they can be transformed into an inner experience 
 only by a process which presupposes the determination of out- 
 ward objects as such. What it really refutes is the Materialism 
 which not only maintains the reality of the external object, but 
 refuses to take into account the subject for which it is real. 
 
644 THE CRITIQUE OF PTJEE REASON. BOOK I. 
 
 And what it tends to prove is not Sensationalism, but what we 
 may call Spiritualism, or Idealism, in the proper sense — the doc- 
 trine that the reality of the material object lies ultimately in its 
 necessity as an element in the evolution of spirit.^ If, on the 
 other hand, we say that the external object is relative to the 
 * feelings of the sensitive subject, it must be in the sense that the 
 unity of the sensitive life, into which all the manifold stimuli 
 from the external world are taken up and, so to speak, ab- 
 sorbed, contains in it the germ and anticipation of the unity 
 of spiritual life, i.e., of a life which implies the division and 
 reunion of object and subject in consciousness and self-con- 
 sciousness. In determining an animal as such, we recognise 
 two sides of its being; we recognise it as an object in space, 
 and yet as having in it a centre of sensation, in reference to 
 which the spatial externality of the parts of its body is con- 
 tinuously in process of being cancelled. In its inner life, the 
 externality of space is reduced to the unity of a process in 
 time, and the difference of its outward existence disappears in 
 the unity of a sensitive subjectivity which passes through a 
 series of states. In this sense it might be said to live merely 
 an inner life, which, however, is inward only for us who con- 
 trast the two aspects of its life. And in our own life, in which 
 the contrast of inner and outer is actually made, the inner life 
 has ceased to be inward, in the sense of a series of sensitive 
 states unrelated to objects, in the very process by which in 
 another sense it has become inward, as the life of a conscious 
 subject. Hence, we can refute the so-called Idealism, which 
 denies the reality of the external world and reduces it to the 
 sensations of the individual subject, by the very process by 
 which we establish the true Idealism, viz., the doctrine that an 
 external object is not seen in its whole truth until its relation 
 
 1 We may describe the error of Berkeley as essentially this, that he confuses 
 the true Idealism, i.e., the refutation of Materialism by the proof that matter 
 as an object is relative to the conscious self, with the so-called Idealism which 
 is really Sensationalism, and which is as inconsistent with the reality of spirit 
 as of matter. 
 
CHAP. IX. POSTDLATES OF EMPIRICAL THOUGHT. 645 
 
 to the self is considered, and until indeed it is regarded as an 
 element in the process of spiritual life. 
 
 The result then is to show the defect of three imperfect i°"™ ^^v^"- 
 
 ^ ence=outer 
 
 theories as to the nature of inner experience, all of which Sempiated 
 have left traces in the language of Kant. We cannot JSt'elunglnd 
 
 . . • . ., . thought. 
 
 treat inner experience as the consciousness of another object 
 which forms part of the same context of experience with 
 external objects; nor, again, can we treat it as a separate 
 kind of experience, which is capable of being brought under 
 the same categories, though perhaps with a less definite 
 result owing to the nature of the form of time; nor, 
 finally, can we regard it as a consciousness of the process 
 whereby sensations are developed into an intelligible conscious- 
 ness. Kant, though he often speaks as if the first or the 
 second of these alternatives were true, yet in his Refutation of 
 Idealism seems to fall back upon the third. What he un- 
 doubtedly shows is, that inner experience cannot be brought 
 into line with outer experience, or determined in the same way. 
 But the reason lies, not, as Kant seems to imply, in the more 
 abstract nature of inner experience (as given merely under the 
 form of time) but in the fact that, with the recall of the 
 abstraction which we use in determining the objects of outer 
 experience as such, — in other words, with the introduction 
 of the idea that such objects can exist only in relation to 
 a perceiving subject, — we make a necessary fxera^aa-K eh aXXo 
 yevoi, a change to another sphere of thought, in which it 
 is impossible to use the categories of external relation except 
 in subordination to categories of a higher order. Even the 
 action of the environment on a living being cannot be truly 
 conceived according to such categories as those of causality 
 and reciprocity, taken in their ordinary sense. Still less can 
 we treat in that way the relation of the objects known 
 to the self that knows them. In so far as the objects with 
 which we start are essentially objects related to .feeling 
 and* knowledge, we cannot separate the determinations 
 
646 THE CRITIQDE OW PURE REASON. book i. 
 
 they have as objects in space from the further determin- 
 ation which comes to them from such relation. And this 
 further determination, with the application of higher categories 
 which it involves, is not an external addition to our knowledge 
 of objects as in space, but a step toward the discovery of 
 the ultimate meaning or reality of that knowledge. In this 
 point of view, therefore, it is evident that to contemplate 
 our experience as inner experience is simply to enrich our 
 outer experience by bringing in the thought of its relation 
 to feeling in ourselves as sensitive subjects. We cannot, how- 
 ever, conceive ourselves simply as sensitive subjects in relation 
 to an environment, without still making abstraction from 
 the fact that as sensitive subjects we are objects for ourselves, 
 i.e., that we are conscious of our sensations only as we 
 distinguish them from and relate them to the unity manifested 
 in self-consciousness. And this implies that our inner life 
 is not seen in its truth when it is taken as merely sensitive. 
 What, however, we usually mean by an inner experience, is the 
 whole consciousness of ourselves as not only sensitive but 
 thinking subjects ; and this cannot be distinguished from our 
 consciousness of the world ; it can be so distinguished only 
 in the sense that it is the consciousness of the world in a 
 less abstract, and. therefore in a higher and more complete 
 form. In this sense the consciousness of the self is just 
 the consciousness of the world fully developed. 
 vtowSae"™ Now, as has been indicated already, the conception of 
 
 vrtie™we"" development has this difficulty about it that it contains some- 
 regard it as the 
 environment thing vcTv like a contradictiou. It is a conception applied to a 
 
 ofadeveloping o ^ sr rr 
 
 organism. being which exists in space and time, but only in so far as its 
 existence is the negation of that externality which we attribute 
 to all that is in space and time. Thus we think of that 
 which develops as externally related to an environment, in 
 which, however, it finds the means of its self-maintenance. 
 The external relation prepares us to expect the loss of both 
 terms in a third or resultant term ; but the developing being 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 647 
 
 subordinates the external environment to itself, and makes the 
 conditions that seem to limit it a means to the maintenance 
 and aggrandisement of its own being. "We may say, then, 
 of such a being that it implies, to start with, an external or 
 negative relation to that which yet it requires, and in which 
 alone it can find itself realised. In the same way, we might 
 show that a developing being exhibits to us a process of 
 life, which seems at first to be a process of mere change, 
 but which shows itself on consideration to be a change bent 
 back upon itself. Thus we are forced to recognise that it is 
 impossible to deal with a developing being, either as externally 
 conditioning and conditioned by. other things, or again as 
 passing through a series of changes according to the law of 
 causation, without bringing in, in the one case, the negation 
 of the mere externality or spatial relation of the organism 
 and the medium, and, in the other case, the negation of 
 the merely causal sequence of its states. 
 
 A further complexity arises when the being to which we Further 
 
 change when 
 
 apply the idea of development is a self-conscious being. For, in we ™w it as 
 such a case, the nature of the being to be developed seems still 7x41 70"^" 
 more directly to involve the negation of that difference which an/is the 
 
 means through 
 
 yet seems to be presupposed. The consciousness of self is the "hich its sen- 
 
 consciousness 
 
 consciousness of a unity which is presupposed in the determina- '^ d^^ioped- 
 
 tion of objects as a world in space and time. We cannot, 
 
 however, suppose a difference, such as exists between things as 
 
 in space and time, to exist in that for which alone time 
 
 and space are. Yet, on the other hand, in so far as we admit 
 
 that such a consciousness is gradually developed, we are obliged 
 
 to regard the subject of it as passing through states in time 
 
 and standing in relation to objects which externally affect it. 
 
 Our first solution of the difficulty will naturally be to say that 
 
 the developing being presupposes the externality which yet it 
 
 negates, and that it presupposes the successive determination of 
 
 the subject which yet is a conscious subject only as it cancels 
 
 succession in itself. But the difficulty returns that, in taking 
 
648 THE CRITIQUE OF PORE REASON. book I. 
 
 this view, we seem to be making the subject of the consciousness, 
 for which time and space alone are, itself an object in time 
 and space, while yet we regard the process of its existence as 
 one in which both time and space are negated. To put it more 
 directly, a developing consciousness is conceived as passing 
 through a series of stages, yet, just so far as its development is 
 for itself, — and it cannot be its development, strictly speaking, 
 till it is for itself, — it neutralises this change. For to the sub- 
 ject as thus self-conscious, all its life is one. In the same way, 
 its existence in space as externally related to other objects is its 
 existence, only in so far as these objects and its own existence 
 are equally for it; and this involves that, as a self-conscious 
 being, it does not hold an external relation to them. We can 
 get over this difficulty only by recognising that, while space and 
 time and all objects in them exist only for a self-conscious 
 subject, and while, in so far as I am a conscious self, they exist 
 for me, yet that I am a derived self-consciousness, and so far 
 must be regarded as an object, and not as a subject ; though it is 
 only as a subject that I am in the proper sense an ego or self, — 
 a being which can say " I." 
 seSconaoloua- I^ t^is idea of the process of knowledge as a process of 
 development were worked out, it would afford a solution of 
 most of the difficulties which necessarily spring up in connexion 
 with Kant's view of it. For his opposition of an ego in itself 
 and a thing in itself, which are beyond experience, but to 
 which the constituent elements of experience are referred, is 
 made necessary by the supposed dualism in experience between 
 self-consciousness and the consciousness of objects. And this 
 dualism rests on the idea, that the return upon self in self- 
 consciousness is a process of abstraction which presupposes the 
 consciousness of objects, but which is negatively related to that 
 consciousness. Hence, the reference of that in the object which 
 thought in its pure unity excludes, to a thing in itself, corre- 
 sponds, on the other side, with the conception of the unity 
 of self-consciousness as purely analytic. On the view stated 
 
 ness, 
 
CHAP. IX. POSTULATES Or EMPIRICAL THOUGHT. 649 
 
 above, however, the return upon self in self-consciousness is 
 a positive movement by which the consciousness of objects 
 is completed ; and, indeed, to seek to know things as they are 
 in themselves, is just to correct the abstraction of the con- 
 sciousness of objects by raising it to the form of self-conscious- 
 ness. A few more words of explanation on this point may be 
 useful. 
 
 Attention has already been directed to the way in which J^j'^o'^''^^.^^ 
 Kant's conception of the analytic nature of the judgment of nlTuminon. 
 self-consciousness leads us to the conception of a Noumenon. 
 This noumenon is identified with the thing in itself, (which at 
 first appeared merely as the unknown cause to which we refer 
 sensations), and gives a new meaning to it. The two characters 
 which the thing in itself thus plays are, indeed, closely con- 
 nected : for it is because the pure spontaneity of the ego 
 expresses itself in the analytic judgment that it can become 
 synthetic only in relation to given matter ; and, on the other 
 hand, it is just for the same reason that the utmost synthesis 
 reached in experience cannot correspond to the pure unity of 
 thought, and that thought is obliged to set up its own ideal in 
 opposition to experience. 
 
 To put this more definitely, the thing in itself appears at Neoessity of a 
 first simply as the ground to which we refer our sensational ""^y^^nd 
 data, which are given as the material for the process of ex- thoughtand 
 
 r^i • 1 p ^ 1 1 T • r knowledge, as 
 
 perience. Obiection has often been taken to the application oi absolutely 
 
 ^ •' . . . defined by 
 
 the category of cause to the relation of sensation to the thing ^*"'- 
 in itself, and Kant seems to have been aware of this objection. ^ 
 What, however, fortifies him against it is just this, that the 
 activity of thought is conceived by him as in itself analytic, or 
 as synthetic only through its application to the matter of sense 
 through its forms ; which again carries with it the consequence 
 
 1 This objection was first taken by Jaoobi ( Werke, II. p. 303, seq.). Kant's 
 answer is practically contained in the following passage. " We think of some- 
 thing of which in itself we have no conception, but which we think of as 
 standing in a relation to all phenomena analogous to that in which they stand 
 to»each other." (A. 674 ; B. 702.) 
 
650 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 that in this synthesis difference can be only externally com- 
 bined ; and this in turn involves the contingency of experience 
 in relation to thought, and the contingency of the particular 
 elements of experience in reference to each other. Lastly, in so 
 far as the thinking ego becomes conscious of itself in this pro- 
 cess, it becomes conscious of this double contingency as opposed 
 to its own pure analytic self-assertion of its unity, in which 
 differences disappear ; and it is thus driven to set up an ideal of 
 knowledge which contrasts with the reality of experience, in the 
 same way that the pure consciousness of self contrasts with the 
 empirical consciousness. In a former chapter, I pointed out more 
 definitely the difficulties which arise from the analytic conception 
 of thought in itself — which force Kant to schematise thought 
 in relation to time, in order to reach the conceptions of intensive 
 and extensive quantity, involving as these do the synthesis of 
 unity and difference, of affirmation and negation. I pointed 
 out, further, that the reflective categories, involving the idea of 
 a duality of correlative elements, were illegitimately, on Kant's 
 principles, ascribed to pure thought apart from its deter- 
 mination in relation to time, and could only be so ascribed 
 because Kant starts from the judgment. In fact, Kant him- 
 self sees that apart from the schematism they fall back into the 
 identity. Lastly, I pointed out that for the same reason, i.e., 
 because the diff'erence is extraneously brought into thought, the 
 return of thought into unity with itself becomes impossible. 
 Hence, the categories of modality, which are meant to express 
 the correlation of mind with its object, cannot be carried 
 beyond the conception of relation already expressed in the 
 reflective categories ; i.e., the mind must be conceived in relation 
 to its object as in one aspect involving it, yet in another aspect 
 repelling and excluding it. For, self-consciousness is possible 
 only in relation to the object, as the consciousness of an identity 
 of thought with itself in determining it as an object ; yet, in 
 such determination the object is opposed to the consciousness 
 that so determines it. If we bring this iiito relation with what 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 651 
 
 was said before of the synthetic judgment as a combination of 
 the unity of thought with a matter extraneous to it, or of a 
 matter already united with the unity of thought with a new 
 matter, it becomes clear that for such combination there is 
 wanted some middle term, and that this want of mediation 
 always sends us back in search of a previous matter whose 
 union with thought is more direct. But this previous matter 
 in its turn, as matter given for thought and not by it, requires 
 mediation. On the other hand, it is only, as Kant allows, 
 in relation to a matter, that the identity of self appears as the 
 jvdgment of self-identity or self-consciousness. Hence, as an 
 explicit self-consciousness, the ego involves the presence to it of 
 that matter which it opposes to itself, and which it is ever by 
 this regress seeking to bring to unity with itself. What this, 
 however, would seem to involve, viz., that beyond the analytic 
 and the synthetic judgments, as the condition of their 
 possibility, or beyond the opposition of the subjective 
 unity of thought and the objective unity of knowledge 
 there lies a unity which embraces both, is hidden from 
 Kant. Kant, therefore, is unable to find in the idea of 
 self-consciousness more than the suggestion of an ideal of 
 knowledge, which cannot be realised in experience as a know- 
 ledge of empirical objects. If, however, we consider the 
 judgment of self-consciousness as what it really is, i.e., as itself 
 involving a synthesis which is also an analysis, as in fact 
 transcending the distinction of analysis and synthesis which 
 Kant draws, we shall have no dif&culty in regarding the 
 necessary distinction of the judgment of self-consciousness 
 from the judgment of consciousness '{i.e., the judgment 
 which expresses the knowledge of the object) as not exclud- 
 ing their unity, but rather presupposing it. And if this 
 be the case, the ideal of knowledge suggested by self-conscious- 
 ness will not be irreconcilable with the experience with which 
 it is contrasted, but will rather be a necessary correction and 
 Completion of such experience. 
 
652 THE CRITIQUE OF PURE REASON. BOOK I. 
 
 So^'th?rtat ^^^^ ™^y ^^ brought into relation with what has been 
 tion'^ot th"'™' already said of the principles of modality, and especially of the 
 theticaiiy opposition of real possibility to reality. That opposition, as I 
 
 necessary, i.e., x ./ ./ ^ ^ 
 
 coibinattoT pointed out, could not be taken in the sense that perception is 
 ^etermSed'' Something which must be added to thought in order to deter^ 
 mine as real that which is already known as really possible. 
 On the contrary, the reference of a perception to an object, 
 which alone brings it under the category of reality, involves its 
 connexion with other objects according to the analogies of 
 experience. The defect of this determination, )iowever, is that 
 it can never be complete, and hence all the reality we can 
 reach may be said to be only hypothetically necessary, i.e.,. 
 determined as real only in relation to something else which is 
 not yet so determined. ■ 
 ontoreoognise Now, the meaning of this is that, in terms of a consciousness 
 
 in the self the ,.,. .ii ii ^ n- .ji 
 
 principle in which IS guidcd Only by the reilective categories, there is no 
 
 which the 
 
 differentiation fl^g^i determination of anything. Everything so determined 
 
 begins and to "^ J o J o 
 
 returns.* appcars as a link in a chain or network of conditioned con- 
 ditions extending ad infinitum in time and space, as an endless 
 series which nowhere finds a point of attachment to which it is 
 made fast. But, then, the ego for which this chain exists and 
 for which each perception is referred to another as a link in the 
 chain, cannot itself be taken as such a link, nor can it be con- 
 ceived as external to all the links as they are to each other. If 
 this is so, the hypothetical necessity which we predicate of the 
 links of the chain in relation to each other, cannot be the last 
 word to be said even about these links. The relation to the 
 self, which is implied in the consciousness of the world under 
 such conditions, must be taken in to complete the conception 
 of the world. This, however, implies that the chain exists 
 only in relation to the self, which returns into itself — i.e., is 
 conscious of itself — only as it combines all phenomena pre- 
 sented to it in one world ; or, in other words, that the chain 
 itself is to be taken as a factor in the process whereby self- 
 consciousness is realised. If vCft look at the chain by itself, we 
 
CHAP. IX. POSTULATES OF EMPIRICAL THOUGHT. 653 
 
 find that we are driven continually to seek a further link, while 
 yet no last link can be found. But when we look at it in 
 reference to the subject, we see that this endless progress is 
 itself the result of the imperfection or abstractness of the 
 principle, the application of which leads to it. The endless 
 progress really is a factor in a life which is founded in itself 
 and never removes from itself. In this point of view Kant's 
 error is, that he stops short with the conception of self-con- 
 sciousness as setting up an ideal for experience which it cannot 
 realise, and does not go on to see that just in so far as self- 
 consciousness is attained the ideal is realised ; for that which, 
 taken by itself, is a mere external combination of coexistent and 
 successive phenomena which no-where and no-when limits itself, 
 has become a unity of correlative elements in a whole which is 
 limited by itself. No doubt there are new difficulties involved 
 in the fact that this self-consciousness, as it appears in us, is 
 only in process of development ; but these difficulties do not 
 affect the truth of the relativity of the chain of phenomena as 
 such to the self which brings their diversity together in the con- 
 sciousness of its own unity. 
 
 From the point of view we have now reached, we can review Tie correction 
 
 -*- of our first 
 
 the movement of .Kant's thought in relation to experience, or ^ZfaS""^^ 
 
 ,, 1 . , t , . • 1 • mathematical 
 
 rather we can see what that movement mvoives in a way principles by 
 
 the analogies 
 
 impossible to Kant himself. The simplest attitude of thought of experience. 
 toward objects — that in which there is least of reflexion upon 
 the conditions under which they are known to us — is that in 
 which they are determined as objects in space passiiig through 
 changes in time, and in which they are supposed to be given in 
 perception as such objects, so that for us is left only the task of 
 analysis. On this view the mental process seems to be confined 
 to a formal activity of thought, by which the quantity and 
 quality of objects is exactly ascertained. But, as yet, we do not 
 reflect on them except as objects given each in a single per- 
 ception. A second reflexion, however, teaches us that without 
 the determination of objects as j^.ermanent, and of their changes 
 
'9 
 
 654 
 
 THE CRITIQUE OF PURE REASON. 
 
 The correction 
 of the view of 
 phenomena 
 under the 
 analogies of 
 experience by 
 the higher 
 categories 
 which deter- 
 mine objects 
 as related to 
 the self. 
 
 as determined by law, there would be no objects for us ; or, in 
 other words, there would be no things or events identified as 
 the same on their recurrence in perception, and the world would 
 not be for us one world in one space and time. 
 
 When, however, such principles are seen to be necessary for 
 knowledge, the question cannot but arise, whether we are 
 entitled to apply them to perception : a question which can be 
 satisfactorily answered only when it is shown that our per- 
 ception really involves these principles, and that they spring 
 from the same activity of the self by which objects were 
 quantitatively and qualitatively determined, an activity which 
 is not merely formal in the one case any more than in the 
 other. But this again brings to light a new characteristic of 
 things which gives still further qualification to them, viz., their 
 essential relation to the self. And, if in taking the first step 
 we were led to recognise that all objects in space and time 
 must be necessarily related to each other according to the 
 Analogies of Experience, we are now forced to go further and 
 to recognise that time and space, the world of objects so related, 
 cannot be adequately understood unless we regard it as essenti- 
 ally related to a conscious self, and as a necessary element in 
 its self-consciousness : or, in other words, unless we regard the 
 world in space and time as essentially the manifestation of a 
 spiritual principle. 
 
 It is the change which comes over the empirical conscious- 
 ness in this point of view with which we have to deal in Kant's 
 Diahctic. 
 
 END OF VOL. I. 
 
 PBINTEE BT EOBERI MACLEHOSE, AT THE UNIVEKSITV PRESS, OLASOOW.