Section 26 Concludes by Consideration of the Question Why Appearances

must conform to the a priori categories. It is no more surprising, Kant claims, than that they should agree with the a priori forms of intuition. The categories and the intuitional forms are relative to the same subject to which the appearances are relative; and the appearances “as mere representations are subject to no law of connection save that which the combining faculty prescribes.”

The summary of the deduction given in section 27 discusses the three possible theories regarding the origin of pure concepts, viz. those of generatio aequivoca (out of experience), epigenesis, and preformation. The first is disproved by the deduction. The second is the doctrine of the deduction and fulfils all the requirements of demonstration. The proof that the categories are at once independent of experience and yet also universally valid for all experience is of the strongest possible kind, namely, that they make experience itself possible. The third theory, that the categories, while subjective and self-discovered, originate in faculties which are implanted in us by our Creator and which are so formed as to yield concepts in harmony with the laws of nature, lies open to two main objections. In the first place, this is an hypothesis capable of accounting equally well for any kind of a priori whatsoever; the predetermined powers of judgment can be multiplied without limit. But a second objection is decisive, namely, that on such a theory the categories would lack the particular kind of necessity which is required. They would express only the necessities imposed upon our thinking by the constitution of our minds, and would not justify any assertion of necessary connection in the object. Kant might also have added, that this hypothesis is metaphysical, and therefore offers in explanation of the empirical validity of a priori concepts a theory which rests upon and involves their unconditioned employment. That is a criticism which is reinforced by the teaching of the Dialectic.

To return now to the omitted sections 22 to 25. Section 22 makes no fresh contribution to the argument of the first edition. Its teaching in regard to pure intuition and mathematical knowledge has already been commented upon. In section 23 Kant dwells upon an interesting consequence of the argument of the deduction. The categories have a wider scope than the pure forms of sense. Since the argument of the deduction has shown that judgment is the indispensable instrument both for reducing a manifold to the unity of apperception and also for conferring upon representations a relation to an object, it follows that the categories which are simply the possible functions of unity in judgment are valid for any and every consciousness that is sensuously conditioned and whose knowledge is therefore acquired through synthesis of a given manifold. Though such consciousness may not intuit in terms of space and time, it must none the less apprehend objects in terms of the categories. The categories thus extend to objects of sensuous intuition in general. They are not, however, valid of objects as such, that is, of things in themselves. As empty relational forms they have meaning only in reference to a given matter; and as instruments for the reduction of variety to the unity of apperception their validity has been proved only for conscious and sensuous experience. Even if the possibility of a non-sensuous intuitive understanding, capable of apprehending things in themselves, be granted, we have no sufficient ground for asserting that the forms which such understanding will employ must coincide with the categories. These are points which will come up for discussion in connection with Kant’s more detailed argument in the chapter on the distinction between phenomena and noumena.

The heading to section 24 is decidedly misleading. The phrase “objects of the senses in general” might be synonymous with “objects of intuition in general” of the preceding section. To interpret it, however, by the contents of the section, it means “objects of our senses.” This section ought, therefore, to form part of section 26, which in its opening sentences supplies its proper introduction. (It may also be noted that the opening sentences of section 24 are a needless repetition of section 23. This would seem to show that it was not written in immediate continuation of it.) The first three paragraphs of section 24 expound the same doctrine of schematism as that outlined in section 26, save that time alone is referred to. The remaining paragraphs of section 24 deal with the connected doctrine of inner sense. Section 25 deals with certain consequences which follow from that doctrine of inner sense.

THE DOCTRINE OF INNER SENSE

We have still to consider a doctrine of great importance in Kant’s thinking, that of inner sense. The significance of this doctrine is almost inversely proportionate to the scantiness and obscurity of the passages in which it is expounded and developed. Much of the indefiniteness and illusiveness of the current interpretations of Kant would seem to be directly traceable to the commentator’s failure to appreciate the position which it occupies in Kant’s system. Several of Kant’s chief results are given as deductions from it, while it itself, in turn, is largely inspired by the need for a secure basis upon which these positions may be made to rest. The relation of the doctrine to its consequences is thus twofold. Kant formulates it in order to safeguard or rather to justify certain conclusions; and yet these conclusions have themselves in part been arrived at owing to his readiness to accept such a doctrine, and to what would seem to have been his almost instinctive feeling of its kinship (notwithstanding the very crude form in which alone he was able to formulate it) with Critical teaching. It was probably one of the earliest of the many new tenets which Kant adopted in the years immediately subsequent to the publication of the inaugural Dissertation, but it first received adequate statement in the second edition of the Critique. Kant took advantage of the second edition to reply to certain criticisms to which his view of time had given rise, and in so doing was compelled to formulate the doctrine of inner sense in a much more explicit manner. Hitherto he had assumed its truth, but had not, as it would seem, sufficiently reflected upon the various connected conclusions to which he was thereby committed. This is one of the many instances which show how what is most fundamental in Kant’s thinking is frequently that of which he was himself least definitely aware. Like other thinkers, he was most apt to discuss what he himself was inclined to question and feel doubt over. The sources of his insight as well as the causes of his failure often lay beyond the purview of his explicitly developed tenets; and only under the stimulus of criticism was he constrained and enabled to bring them within the circle of reasoned conviction. We may venture the prophecy that if Kant had been able to devote several years more to the maturing of the problems which in the face of so many difficulties he had brought thus far, the doctrine of inner sense, or rather the doctrines to which it gives expression, would have been placed in the forefront of his teaching, and their systematic interconnection, both in the way of ground and of consequence, with all his chief tenets would have been traced and securely established.

This would have involved, however, two very important changes. In the first place, Kant would have had to recognise the unsatisfactory character of the supposed analogy between inner and outer sense. As already remarked, no great thinker, except Locke, has attempted to interpret inner consciousness on the analogy of the senses; and the obscurities of Kant’s argument are not, therefore, to be excused on the ground that “the difficulty, how a subject can have an internal intuition of itself, is common to every theory.” Secondly, Kant would have had to define the relation in which he conceived this part of his teaching to stand to his theory of consciousness. But both these changes could have been made without requiring that he should give up the doctrines which are mainly responsible for his theory of inner sense, namely, that there can be no awareness of awareness, but only of existences which are objective, and that there is consequently no consciousness of the generative, synthetic processes which constitute consciousness on its subjective side. It is largely in virtue of these conclusions that Kant’s phenomenalism differs from the subjective idealism of his predecessors. If we ignore or reject them, merely because of the obviously unsatisfactory manner in which alone Kant has been able to formulate them, we rule ourselves out from understanding the intention and purpose of much that is most characteristic of Critical teaching.

The doctrine of inner sense, as expounded by Locke, suffers from an ambiguity which seems almost inseparable from it, namely, the confusion between inner sense, on the one hand as a sense in some degree analogous in nature to what may be called outer sense, and on the other as consisting in self-conscious reflection. This same confusion is traceable throughout the Critique, and is, as we shall find, in large part responsible for Kant’s failure to recognise, independently of outside criticism, the central and indispensable part which this doctrine is called upon to play in his system.

The doctrine is stated by Kant as follows. Just as outer sense is affected by noumenal agencies, and so yields a manifold arranged in terms of a form peculiar to it, namely, space, so inner sense is affected by the mind itself and its inner state. The manifold thereby caused is arranged in terms of a form peculiar to inner sense, namely, time. The content thus arranged falls into two main divisions. On the one hand we have feelings, desires, volitions, that is, states of the mind in the strict sense, subjective non-spatial existences. On the other hand we have sensations, perceptions, images, concepts, in a word, representations (Vorstellungen) of every possible type. These latter all refer to the external world in space, and yet, according to Kant, speaking from the limited point of view of a critique of knowledge, form the proper content of inner sense. “...the representations of the outer senses constitute the actual material with which we occupy our minds,” “the whole material of knowledge even for our inner sense.” (These statements, it may be observed, are first made in the second edition.) As Kant explains himself in B 67-8, he would seem to mean that the mind in the process of “setting” representations of outer sense in space affects itself, and is therefore constrained to arrange the given representations likewise in time. No new content, additional to that of outer sense, is thereby generated, but what previously as object of outer sense existed merely in space is now also subjected to conditions of time. The representations of outer sense are all by their very nature likewise representations of inner sense. To outer sense is due both their content and their spatial form; to inner sense they owe only the additional form of time; their content remains unaffected in the process of being taken over by a second sense. This yields such explanation as is possible of Kant’s assertion in A 33 that “time can never be a determination of outer appearances.” He may be taken as meaning that time is never a determination of outer sense as such, but only of its contents as always likewise subject to the form of inner sense.

This is how Kant formulates his position from the extreme subjectivist point of view which omits to draw any distinction between representation and its object, between inner states of the self and appearances in space. All representations, he says, all appearances without exception, are states of inner sense, modifications of the mind. Some exist only in time, some exist both in space and in time; but all alike are modes of the identical self, mere representations (blosse Vorstellungen). Though appearances may exist outside one another in space, space itself exists only as representation, merely “in us.”

Now without seeking to deny that this is a view which we find in the second edition of the Critique as well as in the first, and that even in passages which are obviously quite late in date of writing Kant frequently speaks in terms which conform to it, we must be no less insistent in maintaining that an alternative view more and more comes to the front in proportion as Kant gains mastery over the conflicting tendencies that go to constitute his new Critical teaching. From the very first he uses language which implies that some kind of distinction must be drawn between representations and objects represented, between subjective cognitive states in the proper sense of the term and existences in space.

“Time can never be a determination of outer appearances. It belongs neither to form nor position, etc. On the other hand it determines the relation of representations in our inner state.”

Similarly in those very sentences in which he asserts all appearances to be blosse Vorstellungen, a distinction is none the less implied.

“Time is the formal a priori condition of all appearances in general. Space, as the pure form of all outer intuition, is as a priori condition limited exclusively (bloss) to outer appearances. On the other hand as all representations, whether they have outer things as their object or not, still in themselves belong, as determinations of the mind, to the inner state, and this inner state is subject to the formal condition of inner intuition, that is of time, time is an a priori condition of all appearance whatever. It is, indeed, the immediate condition of the inner appearance (of our souls), and thereby mediately likewise of outer appearances.”

As the words which I have italicised show, Kant, even in the very sentence in which he asserts outer representations to be inner states, none the less recognises that appearances in space are not representations in the same meaning of that term as are subjective states. They are the objects of representation, not representation itself. The latter alone is correctly describable as a state of the mind. The former may be conditioned by representation, and may therefore be describable as appearances, but are not for that reason to be equated with representation. But before the grounds and nature of this distinction can be formulated in the proper Critical terms, we must consider the reasons which induced Kant to commit himself to this obscure and difficult doctrine of inner sense. As I shall try to show, it is no mere excrescence upon his system; on the contrary, it is inseparably bound up with all his main tenets.

One of the chief influences which constrained Kant to develop this doctrine is the conclusion, so essential to his position, that knowledge must always involve an intuitional manifold in addition to a priori forms and concepts. That being so, he was bound to deny to the mind all power of gaining knowledge by mere reflection. If our mental activities and states lay open to direct inspection, we should have to recognise in the mind a non-sensuous intuitional power. Through self-consciousness or reflection we should acquire knowledge independently of sense. Such apprehension, though limited to the mind’s own operations and states, would none the less be knowledge, and yet would not conform to the conditions which, as the transcendental deduction has shown, are involved in all knowledge. In Kant’s view the belief that we possess self-consciousness of this type, a power of reflection thus conceived, is wholly illusory. To assume any such faculty would be to endow the mind with occult or mystical powers, and would throw us back upon the Leibnizian rationalism, which traces to such reflection our consciousness of the categories, and which rears upon this foundation the entire body of metaphysical science.

The complementary negative conclusion of the transcendental deduction is a no less fundamental and constraining influence in compelling Kant to develop a doctrine of inner sense. If all knowledge is knowledge of appearances, or if, as he states his position in the Analytic of Principles, our knowledge can extend no further than sense experience and inference from such experience, either knowledge of our inner states must be mediated, like our knowledge of outer objects, by sensation, or we can have no knowledge of them whatsoever. On Critical principles, consistently applied, there can be no middle course between acceptance of an indirect empirical knowledge of the mind and assertion of its unknowableness. Mental activities may perhaps be thought in terms of the pure forms of understanding, but in that case their conception will remain as purely problematic and as indeterminate as the conception of the thing in itself. It is impossible for Kant to admit immediate consciousness of the mind’s real activities and states, and at the same time to deny that we can have knowledge of things in themselves. The Aesthetic, in proving that everything in space and time is appearance, implicitly assumes the impossibility of direct self-conscious reflection; and the transcendental deduction in showing that all knowledge involves as correlative factors both sense and thought, has reinforced this conclusion, and calls for its more explicit recognition, in reference to the more inward aspect of experience.

As we have already noted, Kant’s doctrine of inner sense was probably adopted in the early ’seventies, and though it is not itself definitely formulated in the first edition, the chief consequence that follows from it is clearly recognised. Thus in the Aesthetic Kant draws the conclusion that, as time is the form of inner sense, everything apprehended in time, and consequently all inner states and activities, can be known only as appearances. The mind (meaning thereby the ultimate conditioning grounds of consciousness) is as indirectly known as is any other mode of noumenal existence. In the Analytic, whenever he is called upon to express himself upon this and kindred points, he continues to hold to this position; and in the section on the Paralogisms all the main consequences that follow from its acceptance are drawn in the most explicit and unambiguous manner. It is argued that as the inner world, the feelings, volitions and representations of which we are conscious, is a world constructed out of a given manifold yielded by inner sense, and is therefore known only as the appearance of a deeper reality which we have no power of apprehending, it possesses no superiority either of certainty or of immediacy over the outer world of objects in space. We have immediate consciousness of both alike, but in both cases this immediate consciousness rests upon the transcendental synthetic processes whereby such consciousness is conditioned and generated. The transcendental activities fall outside the field of empirical consciousness and therefore of knowledge.

Thus Kant would seem to be maintaining that the radical error committed by the subjective idealists, and with which all the main defects of their teaching are inseparably bound up, lies in their ascription to the mind of a power of direct self-conscious reflection, and consequently in their confusion of the transcendental activities which condition consciousness with the inner states and processes which such consciousness reveals. This has led them to ascribe priority and independence to our inner states, and to regard outer objects as known only by an inference from them. The Critical teaching insists on the distinction between appearance and reality, applies it to the inner life, and so restores to our consciousness of the outer world the certainty and immediacy of which subjective idealism would profess to deprive it. Such are the important conclusions at which Kant arrives in his various “refutations of idealism”; and it will be advisable to consider these refutations in full detail before attempting to complete our statement of his doctrine of inner sense.

KANT’S REFUTATIONS OF IDEALISM

Kant has in a number of different passages attempted to define his Critical standpoint in its distinction from the positions of Descartes and Berkeley. Consideration of these will enable us to follow Kant in his gradual recognition of the manifold consequences to which he is committed by his substitution of inner sense for direct self-conscious intuition or reflection, or rather of the various congenial tenets which it gives him the right consistently to defend and maintain. In Kant’s Critical writings we find no less than seven different statements of his refutation of idealism: (I.) in the fourth Paralogism of the first edition of the Critique; (II.) in section 13 (Anm. ii. and iii.) of the Prolegomena; (III.) in section 49 of the Prolegomena; (IV.) in the second appendix to the Prolegomena; (V.) in sections added in the second edition at the conclusion of the Aesthetic (B 69 ff.); (VI.) in the “refutation of idealism” (B 274-8), in the supplementary section at the end of the section on the Postulates (B 291-4), and in the note to the new preface (B xxxix-xl); (VII.) in the “refutation of problematic idealism” given in the Seven Small Papers which originated in Kant’s conversations with Kiesewetter. Consideration of these in the above order will reveal Kant’s gradual and somewhat vacillating recognition of the new and revolutionary position which alone genuinely harmonises with Critical principles. But first we must briefly consider the various meanings which Kant at different periods assigned to the term idealism. Even in the Critique itself it is employed in a great variety of diverse connotations.

In the pre-Critical writings the term idealism is usually employed in what was its currently accepted meaning, namely, as signifying any philosophy which denied the existence of an independent world corresponding to our subjective representations. But even as thus used the term is ambiguous. It may signify either denial of a corporeal world independent of our representations or denial of an immaterial world “corresponding to” the represented material world, i.e. the denial of Dinge an sich. For there are traceable in Leibniz’s writings two very different views as to the reality of the material world. Sometimes the monads are viewed as purely intelligible substances without materiality of any kind. The kingdom of the extended is set into the representing subjects; only the immaterial world of unextended purely spiritual monads remains as independently real. At other times the monads, though in themselves immaterial, are viewed as constituting through their coexistence an independent material world and a materially occupied space. Every monad has a spatial sphere of activity. The material world is an objective existence due to external relations between the monads, not a merely subjective existence internal to each of them. This alternation of standpoints enabled Leibniz’s successors to deny that they were idealists; and as the more daring and speculative aspects of Leibniz’s teaching were slurred over in the process of its popularisation, it was the second, less consistent view, which gained the upper hand. Wolff, especially in his later writings, denounces idealism; and in the current manuals, sections in refutation of idealism became part of the recognised philosophical teaching. Idealism still, however, continued to be used ambiguously, as signifying indifferently either denial of material bodies or denial of things in themselves. This is the dual meaning which the term presents in Kant’s pre-Critical writings. In his Dilucidatio (1755) he refutes idealism by means of the principle that a substance cannot undergo changes unless it is a substance independent of other substances. Obviously this argument can at most prove the existence of an independent world, not that it is spatial or material. And as Vaihinger adds, it does not even rule out the possibility that changes find their source in a Divine Being. In the Dreams of a Visionseer (1766) Swedenborg is described as an idealist, but without further specification of the exact sense in which the term is employed. In the inaugural Dissertation (1770) idealism is again rejected, on the ground that sense-affection points to the presence of an intelligible object or noumenon.

In Kant’s class lectures on metaphysics, which fall, in part at least, between 1770 and 1781, the term idealism is employed in a very different sense, which anticipates its use in the Appendix to the Prolegomena. The teaching of the Dissertation, that things in themselves are knowable, is now described as dogmatic, Platonic, mystical (schwärmerischer) idealism. He still rejects the idealism of Berkeley, and still entitles it simply idealism, without limiting or descriptive predicates. But now also he employs the phrase “problematic idealism” as descriptive of his own new position. This is, of course, contrary to his invariable usage elsewhere, but is interesting as showing that about this time his repugnance to the term idealism begins to give way, and that he is willing to recognise that the relation of the Critical teaching to idealism is not one of simple opposition. He now begins to regard idealism as a factor, though a radically transformed factor, in his own philosophy.

Study of the Critique reinforces this conclusion. In the Aesthetic Kant teaches the “transcendental ideality” of space and time; and in the Dialectic (in the fourth Paralogism) describes his position as idealism, though with the qualifying predicate transcendental. But though this involves an extension of the previous connotation of the term idealism, and might therefore have been expected to increase the existing confusion, it has the fortunate effect of constraining Kant to recognise and discriminate the various meanings in which it may be employed. This is done somewhat clumsily, as if it were a kind of afterthought. In the introductory syllogism of the fourth Paralogism Descartes’ position and his own are referred to simply as idealism and dualism respectively. The various possible sub-species of idealism as presented in the two editions of the Critique and in the Prolegomena may be tabulated as follows:

{Material {Sceptical {Problematic (the position of Descartes). { { {Sceptical in the stricter and more usual Idealism { { { sense (the position of Hume). { {Dogmatic (the position of Berkeley). {Formal or Critical or Transcendental (Kant’s own position).

The distinction between problematic idealism and idealism of the more strictly sceptical type is not clearly drawn by Kant. Very strangely Kant in this connection never mentions Hume: the reference in B xxxix n. is probably not to Hume but to Jacobi. Transcendental idealism is taken as involving an empirical realism and dualism, and is set in opposition to transcendental realism which is represented as involving empirical idealism. In B xxxix n. Kant speaks of “psychological idealism,” meaning, as it would seem, material or non-Critical idealism.

In the second appendix to the Prolegomena Kant draws a further distinction, in line with that already noted in his lectures on metaphysics. Tabulated it is as follows:

{Mystical, in the sense of belief in and reliance on a supposed { human power of intellectual intuition. It is described as Idealism{idealism in the strict (eigentlich) sense--the position of the {Eleatics, of Plato and Berkeley. {Formal or Critical--Kant’s own position.

This latter classification can cause nothing but confusion. The objections that have to be made against it from Kant’s own critical standpoint are stated below.

Let us now consider, in the order of their presentation, the various refutations of idealism which Kant has given in his Critical writings.

=I. Refutation of Idealism as given in First Edition of “Critique”= (A 366-80).--This refutation is mainly directed against Descartes, who is mentioned by name in A 367. Kant, as Vaihinger suggests, was very probably led to recognise Descartes’ position as a species of idealism in the course of a re-study of Descartes before writing the section on the Paralogisms. As already pointed out, this involves the use of the term idealism in a much wider sense than that which was usually given to it in Kant’s own day. In the development of his argument Kant also wavers between two very different definitions of this idealism, as being denial of immediate certainty and as denial of all certainty. The second interpretation, which would make it apply to Hume rather than to Descartes, is strengthened in the minds of his readers by his further distinction between dogmatic and sceptical idealism, and the identification of the idealism under consideration with the latter. The title problematic which Kant in the second edition applies to Descartes’ position suffers from this same ambiguity. As a matter of fact, Kant’s refutation applies equally well to either position. The teaching of Berkeley, which coincides with dogmatic idealism as here defined by Kant, namely, as consisting in the contention that the conception of matter is inherently contradictory, is not dwelt upon, and the appended promise of refutation is not fulfilled.

Descartes’ position is stated as follows: only our own existence and inner states are immediately apprehended by us; all perceptions are modifications of inner sense; and the existence of external objects can therefore be asserted only by an inference from the inner perceptions viewed as effects. In criticism, Kant points out that since an effect may result from more than one cause, this inference to a quite determinate cause, viz. objects as bodies in space, is doubtfully legitimate. The cause of our inner states may lie within and not without us, and even if external, need not consist in spatial objects. Further, leaving aside the question of a possible alternative to the assumption of independent material bodies, the assertion of the existence of such objects would, on Descartes’ view, be merely conjectural. It could never have certainty in any degree equivalent to that possessed by the experiences of inner sense.

“By an idealist, therefore, we must not understand one who denies the existence of outer objects of the senses, but only one who does not admit that their existence is known through immediate perception, and who therefore concludes that we can never, by way of any possible experience, be completely certain of their reality.”

No sooner is the term idealist thus clearly defined than Kant, in keeping with the confused character of the entire section, proceeds to the assertion (a) that there are idealists of another type, namely, transcendental idealists, and (b) that the non-transcendental idealists sometimes also adopt a dogmatic position, not merely questioning the immediacy of our knowledge of matter, but asserting it to be inherently contradictory. All this points to the composite origin of the contents of this section.

Transcendental idealism is opposed to empirical idealism. It maintains that phenomena are representations merely, not things in themselves. Space and time are the sensuous forms of our intuitions. Empirical idealism, on the other hand, goes together with transcendental realism. It maintains that space and time are given as real in themselves, in independence of our sensibility. (Transcendental here, as in the phrase “transcendental ideality,” is exactly equivalent to transcendent.) But such a contention is inconsistent with the other main tenet of empirical idealism. For if our inner representations have to be taken as entirely distinct from their objects, they cannot yield assurance even of the existence of these objects. To the transcendental idealist no such difficulty is presented. His position naturally combines with empirical realism, or, as it may also be entitled, empirical dualism. Material bodies in space, being merely subjective representations, are immediately apprehended. The existence of matter can be established “without our requiring to issue out beyond our bare self-consciousness or to assume anything more than the certainty of the representations in us, i.e. of the cogito ergo sum.” Though the objects thus apprehended are outside one another in space, space itself exists only in us.

“Outer objects (bodies) are mere appearances, and are therefore nothing but a species of my representations, the objects of which are something only through these representations. Apart from them they are nothing. Thus outer things exist as well as I myself, and both, indeed, upon the immediate witness of my self-consciousness....”

The only difference is that the representation of the self belongs only to inner, while extended bodies also belong to outer sense. There is thus a dualism, but one that falls entirely within the field of consciousness, and which is therefore empirical, not transcendental. There is indeed a transcendental object which “in the transcendental sense may be outside us,” but it is unknown and is not in question. It ought not to be confused with our representations of matter and corporeal things.

From this point the argument becomes disjointed and repeats itself, and there is much to be said in support of the contention of Adickes that the remainder of the section is made up of a number of separate interpolations. First, Kant applies the conclusion established in the Postulates of Empirical Thought, viz. that reality is revealed only in sensation. As sensation is an element in all outer perception, perception affords immediate certainty of real existence, Kant next enters upon a eulogy of sceptical idealism as “a benefactor of human reason.” It brings home to us the utter impossibility of proving the existence of matter on the assumption that spatial objects are things in themselves, and so constrains us to justify the assertions which we are at every moment making. And such justification is, Kant here claims, only possible if we recognise that outer objects as mere representations are immediately known. In the next paragraph we find a sentence which, together with the above eulogistic estimate of the merits of idealism, shows how very far Kant, at the time of writing, was from feeling the need of differentiating his position from that of subjectivism. The sentence is this:

“We cannot be sentient of what is outside ourselves, but only of what is in ourselves, and the whole of our self-consciousness therefore yields nothing save merely our own determinations.”

It is probable, indeed, that the paragraph in which this occurs is of very early origin, prior to the development of the main body of the Analytic; for in the same paragraph we also find the assertion, utterly at variance with the teaching of the Analytic and with that of the first and third Paralogisms, that “the thinking ego” is known phenomenally as substance. We seem justified in concluding that the various manuscripts which have gone to form this section on the fourth Paralogism were written at an early date within the Critical period.

We may note, in passing, two sentences in which, as in that quoted above, a distinction between representations and their objects is recognised in wording if not in fact.

“All outer perception furnishes immediate proof of something actual in space, or rather is the actual itself. To this extent empirical realism is beyond question, i.e. there corresponds to our outer perceptions something actual in space.”

Again in A 377 the assertion occurs that “our outer senses, as regards the data from which experience can arise, have their actual corresponding objects in space.” Certainly these statements, when taken together with the other passages in this section, form a sufficiently strange combination of assertion and denial. Either there is a distinction between representation and its object or there is not; if the former, then objects in space are not merely representations; if the latter, then the “correspondence” is merely that of a thing with itself.

* * * * *

This refutation of idealism will not itself stand criticism. For two separate reasons it entirely fails to attain its professed end. In the first place, it refutes the position of Descartes only by virtually accepting the still more extreme position of Berkeley. Outer objects, Kant argues, are immediately known because they are ideas merely. There is no need for inference, because there is no transcendence of the domain of our inner consciousness. In other words, Kant refutes the problematic idealism of Descartes by means of the more subjective idealism of Berkeley. The “dogmatic” idealism of Berkeley in the form in which Kant here defines it, namely, as consisting in the assertion that the notion of an independent spatial object involves inherent contradictions, is part of his own position. For that reason he was bound to fail in his promise to refute such dogmatic idealism. Fortunately he never even attempts to do so. In the second place, Kant ignores the fact that he has himself adopted an “idealist” view of inner experience. Inner experience is not for him, as it was for Descartes, the immediate apprehension of genuine reality. As it is only appearance, the incorporation of outer experience within it, so far from establishing the reality of the objects of outer sense, must rather prove the direct contrary. No more is really established than Descartes himself invariably assumes, namely, the actual existence of mental representations of a corporeal world in space. Descartes’ further assertion that the world of things in themselves can be inferred to be material and spatial, Kant, of course, refuses to accept. On this latter point Kant is in essential agreement with Berkeley.

It is by no means surprising that Kant’s first critics, puzzled and bewildered by the obscurer and more difficult portions of the Critique, should have based their interpretation of Kant’s general position largely upon the above passages; and that in combining the extreme subjective idealism which Kant there advocates with his doctrine that the inner life of ever-changing experiences is itself merely ideal, should have come to the conclusion that Kant’s position is an extension of that of Berkeley. Pistorius objected that in making outer appearances relative to an inner consciousness which is itself appearance, Kant is reducing everything to mere illusion. Hamann came to the somewhat similar conclusion, that Kant, notwithstanding his very different methods of argument, is “a Prussian Hume,” in substantial agreement with his Scotch predecessor.

II. =“Prolegomena,” Section 13, Notes II and III.=--In the Prolegomena Kant replies to the criticism which the first edition of the Critique had called forth, that his position is an extension of the idealism of Descartes, and even more thoroughgoing than that of Berkeley. Idealism he redefines in a much narrower sense, which makes it applicable only to Berkeley

“...as consisting in the assertion that there are none but thinking beings, and that all other things which we suppose ourselves to perceive in intuition are nothing but representations in the thinking beings, to which no object external to them corresponds in fact.”

In reply Kant affirms his unwavering belief in the reality of Dinge an sich

“...which though quite unknown to us as to what they are in themselves, we yet know by the representations which their influence on our sensibility procures us.... Can this be termed idealism? It is the very contrary.”

Kant adds that his position is akin to that of Locke, differing only in his assertion of the subjectivity of the primary as well as of the secondary qualities.

“I should be glad to know what my assertions ought to have been in order to avoid all idealism. I suppose I ought to have said, not only that the representation of space is perfectly conformable to the relation which our sensibility has to objects (for that I have said), but also that it is completely similar to them--an assertion in which I can find as little meaning as if I said that the sensation of red has a similarity to the property of cinnabar which excites this sensation in me.”

Kant is here very evidently using the term idealism in the narrowest possible meaning, as representing only the position of Berkeley, and as excluding that of Descartes and Leibniz. Such employment of the term is at variance with his own previous usage. Though idealism here corresponds to the “dogmatic idealism” of A 377, it is now made to concern the assertion or denial of things in themselves, not as previously the problem of the reality of material objects and of space. Kant is also ignoring the fact, which he more than once points out in the Critique, that his philosophy cannot prove that the cause of our sensations is without and not within us. His use of “body” as a name for the thing in itself is likewise without justification. This passage is mainly polemical; it is hardly more helpful than the criticism to which it was designed to reply.

In Section 13, Note iii., Kant meets the still more extreme criticism (made by Pistorius), that his system turns all the things of the world into mere illusion (Schein). He distinguishes transcendental idealism from “the mystical and visionary idealism of Berkeley” on the one hand, and on the other from the Cartesian idealism which would convert mere representations into things in themselves. To obviate the ambiguities of the term transcendental, he declares that his own idealism may perhaps more fitly be entitled Critical. This distinction between mystical and Critical idealism connects with the contents of the second part of the Appendix, treated below.

III. =“Prolegomena,” Section 49.=--This is simply a repetition of the argument of the fourth Paralogism. The Cartesian idealism, now (as in B 274) named material idealism, is alone referred to. The Cartesian idealism does nothing, Kant says, but distinguish external experience from dreaming. There is here again the same confusing use of the term “corresponds.”

“That something actual without us not only corresponds but must correspond to our external perceptions can likewise be proved....”

IV. =“Prolegomena,” Second Part of the Appendix.=--Kant here returns to the distinction, drawn in Section 13, Note iii., between what he now calls “idealism proper (eigentlicher),” i.e. visionary or mystical idealism, and his own.

“The position of all genuine idealists from the Eleatics to Bishop Berkeley is contained in this formula: ‘All cognition through the senses and experience is nothing but mere illusion, and only in the ideas of pure understanding and Reason is there truth.’ The fundamental principle ruling all my idealism, on the contrary, is this: ‘All cognition of things solely from pure understanding or pure Reason is nothing but mere illusion and only in experience is there truth.’”

This mode of defining idealism can, in this connection, cause nothing but confusion. Its inapplicability to Berkeley would seem to prove that Kant had no first-hand knowledge of Berkeley’s writings. As Kant’s Note to the Appendix to the Prolegomena shows, he also had Plato in mind. But the definition given of “the fundamental principle” of his own idealism is almost equally misleading. It omits the all-essential point, that for Kant experience itself yields truth only by conforming to a priori concepts. As it is, he proceeds to criticise Berkeley for failure to supply a sufficient criterion of distinction between truth and illusion. Such criterion, he insists, is necessarily a priori. The Critical idealism differs from that of Berkeley in maintaining that space and time, though sensuous, are a priori, and that in combination with the pure concepts of understanding they

“...prescribe a priori its law to all possible experience: the law which at the same time yields the sure criterion for distinguishing within experience truth from illusion. My so-called idealism--which properly speaking is Critical idealism--is thus quite peculiar in that it overthrows ordinary idealism, and that through it all a priori cognition, even that of geometry, now attains objective reality, a thing which even the keenest realist could not assert till I had proved the ideality of space and time.”

V. =Sections added in Second Edition at the Conclusion of the Aesthetic.= (B 69 ff.)--Kant here again replies to the criticism of Pistorius that all existence has been reduced to the level of illusion (Schein). His defence is twofold: first, that in naming objects appearances he means to indicate that they are independently grounded, or, as he states it, are “something actually given.” If we misinterpret them, the result is indeed illusion, but the fault then lies with ourselves and not with the appearances as presented. Secondly, he argues that the doctrine of the ideality of space and time is the only secure safeguard against scepticism. For otherwise the contradictions which result from regarding space and time as independently real will likewise hold of their contents, and everything, including even our own existence, will be rendered illusory. “The good Berkeley [observing these contradictions] cannot, indeed, be blamed for reducing bodies to mere illusion.” This last sentence may perhaps be taken as supporting the view that notwithstanding the increased popularity of Berkeley in Germany and the appearance of new translations in these very years, Kant has not been sufficiently interested to acquire first-hand knowledge of Berkeley’s writings. The epithet employed is characteristic of the rather depreciatory attitude which Kant invariably adopts in speaking of Berkeley.

VI. =“Refutation of Idealism” in Second Edition of the “Critique.”= (B 274-9, supplemented by note to B xxxix.).--The refutation opens by equating idealism with material idealism (so named in contradistinction to his own “formal or rather Critical” teaching). Within material idealism Kant distinguishes between the problematic idealism of Descartes, and the dogmatic idealism of Berkeley. The latter has, he says, been overthrown in the Aesthetic. The former alone is dealt with in this refutation. This is the first occurrence in the Critique of the expression “problematic idealism”: it is nowhere employed in the first edition. Problematic idealism consists in the assertion that we are incapable of having experience of any existence save our own; only our inner states are immediately apprehended; all other existences are determined by inference from them. The refutation consists in the proof that we have experience, and not mere imagination of outer objects. This is proved by showing that inner experience, unquestioned by Descartes, is possible only on the assumption of outer experience, and that this latter is as immediate and direct as is the former.

=Thesis.=--The empirically determined consciousness of my own existence proves the existence of objects in space outside me.

=Proof.=--I am conscious of my own existence as determined in time. Time determination presupposes the perception of something permanent. But nothing permanent is intuitable in the empirical self. On the cognitive side (i.e. omitting feelings, etc., which in this connection are irrelevant), it consists solely of representations; and these demand a permanent, distinct from ourselves, in relation to which their changes, and so my own existence in the time wherein they change, may be determined. Thus perception of this permanent is only possible through a thing outside, and not through the mere representation of a thing outside. And the same must hold true of the determination of my existence in time, since this also depends upon the apprehension of the permanent. That is to say, the consciousness of my existence is at the same time an immediate awareness of the existence of other things outside me.

In the note to the Preface to the second edition occurs the following emphatic statement.

“Representation of something permanent in existence is not the same as permanent representation. For though the representation [of the permanent] may be very changing and variable like all our other representations, not excepting those of matter, it yet refers to something permanent. This latter must therefore be an external thing distinct from all my representations, and its existence must be included in the determination of my own existence, constituting with it but a single experience such as would not take place even internally if it were not also at the same time, in part, external. How this should be possible we are as little capable of explaining further as we are of accounting for our being able to think the abiding in time, the coexistence of which with the variable generates the conception of change.”

The argument of this note varies from that of B 274 ff. only in its use of an ambiguous expression which is perhaps capable of being taken as referring to things in themselves, but which does not seem to have that meaning. “I am just as certainly conscious that there are things outside me which relate to my sense....”

In B 277-8 Kant refers to the empirical fact that determination of time can be made only by relation to outer happenings in space, such as the motion of the sun. This is a point which is further developed in another passage which Kant added in the second edition.

“...in order to understand the possibility of things in conformity with the categories, and so to demonstrate the objective reality of the latter, we need not merely intuitions, but intuitions that are in all cases outer intuitions. When, for instance, we take the pure concepts of relation, we find firstly that in order to obtain something permanent in intuition corresponding to the concept of substance, and so to demonstrate the objective reality of this concept, we require an intuition in space (of matter). For space alone is determined as permanent, while time, and therefore everything that is in inner sense, is in constant flux. Secondly, in order to exhibit change as the intuition corresponding to the concept of causality, we must take as our example motion, i.e. change in space. Only in this way can we obtain the intuition of changes, the possibility of which can never be comprehended through any pure understanding. For change is combination of contradictorily opposed determinations in the existence of one and the same thing. Now how it is possible that from a given state of a thing an opposite state should follow, not only cannot be conceived by any reason without an example, but is actually incomprehensible to reason without intuition. The intuition required is the intuition of the movement of a point in space. The presence of the point in different spaces (as a sequence of opposite determinations) is what first yields to us an intuition of change. For in order that we may afterwards make inner changes likewise thinkable, we must represent time (the form of inner sense) figuratively as a line, and the inner change through the drawing of this line (motion), and so in this manner by means of outer intuition make comprehensible the successive existence of ourselves in different states. The reason of this is that all change, if it is indeed to be perceived as change, presupposes something permanent in intuition, and that in inner sense no permanent intuition is to be met with. Lastly, the possibility of the category of community cannot be comprehended through mere reason alone. Its objective reality is not to be understood without intuition and indeed outer intuition in space.”

In this passage Kant is modifying the teaching of the first edition in two very essential respects. In the first place, he is now asserting that consciousness of both space and motion is necessary to consciousness of time; and in the second place, he is maintaining that the categories can acquire meaning only by reference to outer appearances. Had Kant made all the necessary alterations which these new positions involve, he would, as we shall find, have had entirely to recast the chapters on Schematism and on the Principles of Understanding. Kant was not, however, prepared to make such extensive alterations, and these chapters are therefore left practically unmodified. This is one of the many important points in which the reader is compelled to reinterpret passages of earlier date in the light of Kant’s later utterances. There is also a further difficulty. Does Kant, in maintaining that the categories can acquire significance only in reference to outer perception, also mean to assert that their subsequent employment is limited to the mechanical world of the material sciences? This is a point in regard to which Kant makes no quite direct statement; but indirectly he would seem to indicate that that was not his intention. He frequently speaks of the states of inner sense as mechanically conditioned. Sensations, feelings, and desires, are, he would seem to assert, integral parts of the unitary system of phenomenal existence. Such a view is not, indeed, easily reconcilable with his equating of the principle of substance with the principle of the conservation of matter. There are here two conflicting positions which Kant has failed to reconcile: the traditional dualistic attitude of Cartesian physics and the quite opposite implications of his Critical phenomenalism. When the former is being held to, Kant has to maintain that psychology can never become a science; but his Critical teaching consistently developed seems rather to support the view that psychology, despite special difficulties peculiar to its subject matter, can be developed on lines strictly analogous to those of the material sciences.

We may now return to Kant’s main argument. This new refutation of idealism in the second edition differs from that given in the fourth Paralogism of the first edition, not only in method of argument but also in the nature of the conclusion which it seeks to establish. Indeed it proves the direct opposite of what is asserted in the first edition. The earlier proof sought to show that, as regards immediacy of apprehension and subjectivity of existence, outer appearances stand on the same level as do our inner experiences. The proof of the second edition, on the other hand, argues that though outer appearances are immediately apprehended they must be existences distinct from the subjective states through which the mind represents them. The two arguments agree, indeed, in establishing immediacy, but as that which is taken as immediately known is in the one case a subjective state and in the other is an independent existence, the immediacy calls in the two cases for entirely different methods of proof. The first method consisted in viewing outer experiences as a subdivision within our inner experiences. The new method views their relation as not that of including and included, but of conditioning and conditioned; and it is now to outer experience that the primary position is assigned. So far is outer experience from being possible only as part of inner experience, that on the contrary inner experience, consciousness of the flux of inner states, is only possible in and through experience of independent material bodies in space. A sentence from each proof will show how completely their conclusions are opposed.

“Outer objects (bodies) are mere appearances, and are therefore nothing but a species of my representations, the objects of which are something only through these representations. Apart from them they are nothing.” “Perception of this permanent is possible only through a thing outside me, and not through the mere representation of a thing outside me.”

The one sentence asserts that outer objects are representations; the other argues that they must be existences distinct from their representations. The one inculcates a subjectivism of a very extreme type; the other results in a realism, which though ultimately phenomenalist, is none the less genuinely objective in character. This difference is paralleled by the nature of the idealisms to which the two proofs are opposed and which they profess to refute. The argument of the Paralogism of the first edition is itself Berkeleian, and refutes only the problematic idealism of Descartes. The argument of the second edition, though formally directed only against Descartes, constitutes a no less complete refutation of the position of Berkeley. In its realism it has kinship with the positions of Arnauld and of Reid, while, in attempting to combine this realism with due recognition of the force and validity of Hume’s sceptical philosophy, it breaks through all previous classifications, formulates a profoundly original substitute for the previously existing theories, and inaugurates a new era in the theory of knowledge.

As already pointed out, Kant restates the distinction between the subjective and the objective in a manner which places the problem of knowledge in an entirely new light. The subjective is not to be regarded as opposite in nature to the objective, but as a subspecies within it. It does not proceed parallel with the sequence of natural existences, but is itself part of the natural system which consciousness reveals. Sensations, in the form in which they are consciously apprehended by us, do not constitute our consciousness of nature, but are themselves events which are possible only under the conditions which the natural world itself supplies. The Cartesian dualism of the subjective and the objective is thus subordinated to the Critical distinction between appearance and reality. Kant’s phenomenalism is a genuine alternative to the Berkeleian teaching, and not, as Schopenhauer and so many others have sought to maintain, merely a variant upon it.

The striking contradiction between Kant’s various refutations of idealism has led some of Kant’s most competent critics to give a different interpretation of the argument of the second edition from that given above. These critics take the independent and permanent objects which are distinguished from our subjective representations to be things in themselves. That is to say, they interpret this refutation as based upon Kant’s semi-Critical doctrine of the transcendental object (in the form in which it is employed for the solution of the Antinomies), and so as agreeing with the refutation given in the Prolegomena. Kant is taken as rejecting idealism because of his belief in things in themselves. This is the view adopted by Benno Erdmann, Sidgwick, A. J. Balfour.

As Vaihinger, Caird, and Adamson have shown, such an interpretation is at complete variance with the actual text. This is, indeed, so obvious upon unbiassed examination that the only point which repays discussion is the question, why Benno Erdmann and those who follow him should have felt constrained to place so unnatural an interpretation upon Kant’s words. The explanation seems to lie in Erdmann’s convinced belief, plainly shown in all his writings upon Kant, that the Critique expounds a single consistent and uniform standpoint. If such belief be justified, there is no alternative save to interpret Kant’s refutation of idealism in the manner which Erdmann adopts. For as the subjectivism of much of Kant’s teaching is beyond question, consistency can be obtained only by sacrifice of all that conflicts with it. Thus, and thus alone, can Erdmann’s rendering of the refutation of the second edition be sustained; the actual wording, taken in and by itself, does not support it. Kant here departs from his own repeated assertion, in the second hardly less than in the first edition of the Critique, of the subjectivity of outer appearances. But, as Vaihinger justly contends, Kant was never greater than in this violation of self-consistency, “never more consistent than in this inconsistency.” Tendencies, previously active but hitherto inarticulate, are at last liberated. If the chrysalis stage of the intense brooding of the twelve years of Critical thinking was completed in the writing of the first edition of the Critique, the philosophy which then emerged only attains to mature stature in those extensions of the Critique, scattered through it from Preface to Paralogisms, which embody this realistic theory of the independent existence of material nature. For this theory is no mere external accretion, and no mere reversal of subordinate tenets, but a ripening of germinal ideas to which, even in their more embryonic form, the earlier Critical teaching owed much of its inspiration, and which, when consciously adopted and maturely formulated, constitute such a deepening of its teaching as almost amounts to transformation. The individual self is no longer viewed as being the bearer of nature, but as its offspring and expression, and as being, like nature, interpretable in its twofold aspect, as appearance and as noumenally grounded. The bearer of appearance is not the individual subject, but those transcendental creative agencies upon which man and nature alike depend. Both man and nature transcend the forms in which they are apprehended; and nothing in experience justifies the giving of such priority to the individual mind as must be involved in any acceptance of subjectivist theory. Though man is cognisant of space and time, comprehending them within the limits of his consciousness, and though in all experience unities are involved which cannot originate within or be explained by experience, it is no less true that man is himself subject to the conditions of space and time, and that the synthetic unities which point beyond experience do not carry us to a merely individual subject. If man is not a part or product of nature, neither is nature the product of man. Kant’s transcendentalism, in its maturest form, is genuinely phenomenalist in character. That is the view which has already been developed above, in the discussion of Kant’s transcendental deduction. I shall strive to confirm it by comparison of the teaching of the two editions of the Critique in regard to the reality of outer appearances.

Schopenhauer, to whom this new development of the Critical teaching was altogether anathema, the cloven hoof of the Hegelian heresies, denounced it as a temporary and ill-judged distortion of the true Critical position, maintaining that it is incapable of combination with Kant’s central teaching, and that it finds no support in the tenets, pure and unperverted, of the first edition. Kant, he holds, is here untrue to himself, and temporarily, under the stress of polemical discussion, lapses from the heights to which he had successfully made his way, and upon which he had securely established, in agreement with Plato and in extension of Berkeley, the doctrine of all genuine philosophical thinking, the doctrine of the Welt als Vorstellung.

We may agree with Schopenhauer in regarding those sections of the first edition of the Critique which were omitted in the second edition as being a permanently valuable expression of Kantian thought, and as containing much that finds no equally adequate expression in the passages which were substituted for them; and yet may challenge his interpretation of both editions alike. If, as we have already been arguing, we must regard Kant’s thinking as in large degree tentative, that is, as progressing by the experimental following out of divergent tendencies, we may justly maintain that among the most characteristic features of his teaching are the readiness with which he makes changes to meet deeper insight, and the persistency with which he strives to attain a position in which there will be least sacrifice or blurring of any helpful distinction, and fullest acknowledgment of the manifold and diverse considerations that are really essential. Recognising these features, we shall be prepared to question the legitimacy of Schopenhauer’s opposition between the teaching of the two editions. We shall rather expect to find that the two editions agree in the alternating statement and retraction of conflicting positions, and that the later edition, however defective in this or that aspect as compared with the first edition, none the less expresses the maturer insight, and represents a further stage in the development of ideas that have been present from the start. It may perhaps for this very reason be more contradictory in its teaching; it will at least yield clearer and more adequate formulation of the diverse consequences and conflicting implications of the earlier tenets. It will be richer in content, more open-eyed in its adoption of mutually contradictory positions, freer therefore from unconscious assumptions, and better fitted to supply the data necessary for judgment upon its own defects. Only those critics who are blind to the stupendous difficulties of the tasks which Kant here sets himself, and credulous of their speedy and final completion, can complain of the result. Philosophical thinkers of the most diverse schools in Germany, France, and England, have throughout the nineteenth century received from the Critique much of their inspiration. The profound influence which Kant has thus exercised upon succeeding thought must surely be reckoned a greater achievement than any that could have resulted from the constructing of a system so consistent and unified, that the alternative would lie only between its acceptance and its rejection. Ultimately the value of a philosophy consists more in the richness of its content and the comprehensiveness of its dialectic, than in the logical perfection of its formal structure. The latter quality is especially unfitted to a philosophy which inaugurated a new era, and formulated the older problems in an altogether novel manner. Under such conditions fertility of suggestion and readiness to modify or even recast adopted positions, openness to fuller insight acquired through the very solutions that may at first have seemed to satisfy and close the issues, are more to be valued than the power to remove contradictions and attain consistency. This is the point of view which I shall endeavour to justify in reference to the matters now before us. In particular there are two points to be settled: first, whether and how far the argument of the second edition is prefigured in the first edition; and secondly, whether and to what extent it harmonises with, and gives expression to, all that is most central and genuinely Critical in both editions.

In the first place we must observe that the fourth Paralogism occurs in a section which bears all the signs of having been independently written and incorporated later into the main text. It is certainly of earlier origin than those sections which represent the third and fourth layers of the deduction of the first edition, and very possibly was composed in the middle ’seventies. Indeed, apart from single paragraphs which may have been added in the process of adapting it to the main text, it could quite well, so far as its refutation of idealism is concerned, be of even earlier date. The question as to the consistency of the refutation of the second edition with the teaching of the first edition must therefore chiefly concern those parts of the Analytic which connect with the later forms of the transcendental deduction, that is to say, with the transcendental deduction itself, with the Analogies and Postulates, and with particular paragraphs that have been added in other sections. We have already noted how Kant from the very first uses terms which involve the drawing of a distinction between representations and their objects. Passages in which this distinction occurs can be cited from both the Aesthetic and the Analytic, and two such occur in the fourth Paralogism itself. Objects, he says, “correspond” to their representations. A variation in expression is found in such passages as the following:

“...the objects of outer perception also actually exist (auch wirklich sind) in that very form in which they are intuited in space....”

Such language is meaningless, and could never have been chosen, if Kant had not, even in the earlier stages of his thinking, postulated a difference between the existence of an object and the existence of its representation. He must at least have distinguished between the representations and their content. That, however, he could have done without advancing to the further assertion of their independent existence. Probably he was not at all clear in his own mind, and was too preoccupied with the other complexities of his problem, to have thought out his position to a definite decision. When, however, as in the fourth Paralogism, he made any attempt so to do, he would seem to have felt constrained to adopt the extreme subjectivist position. Expressions to that effect are certainly very much more common than those above mentioned. This is what affords Schopenhauer such justification, certainly very strong, as he can cite for regarding subjectivism as the undoubted teaching of the first edition.

When, however, we also take account of the very different teaching which is contained in the important section on the Postulates of Empirical Thought, the balance of evidence is decisively altered. The counter-teaching, which is suggested by certain of the conflicting factors of the transcendental deduction and of the Analogies, here again receives clear and detailed expression. This is the more significant, as it is in this section that Kant sets himself formally to define what is to be understood by empirical reality. It thus contains his, so to speak, official declaration as to the mode of existence possessed by outer appearances. The passage chiefly relevant is as follows:

“If the existence of the thing is bound up with some perceptions according to the principles of their empirical connection (the Analogies), we can determine its existence antecedently to the perception of it, and consequently, to that extent, in an a priori manner. For as the existence of the thing is bound up with our perceptions in a possible experience, we are able in the series of possible perceptions, and under the guidance of the Analogies, to make the transition from our actual perception to the thing in question. Thus we discover the existence of a magnetic matter pervading all bodies from the perception of the attracted iron filings, although the constitution of our organs cuts us off from all immediate perception of that matter. For in accordance with the laws of sensibility and the connection of our perceptions in a single experience, we should, were our senses more refined, actually experience it in an immediate empirical intuition. The grossness of our senses does not in any way decide the form of possible experience in general.”

Now it cannot, of course, be argued that the above passage is altogether unambiguous. We can, if we feel sufficiently constrained thereto, place upon it an interpretation which would harmonise it with Kant’s more usual subjectivist teaching, namely as meaning that in the progressive construction of experience, or in the ideal completion which follows upon assumption of more refined sense-organs, possible empirical realities are made to become, or are assumed to become, real, but that until the possible experiences are thus realised in fact or in ideal hypothesis, they exist outwardly only in the form of their noumenal conditions. And as a matter of fact, this is how Kant himself interprets the teaching of this section in the process of applying it in solution of the antinomies.

“Accordingly, if I represent to myself the aggregate of all objects of the senses existing in all time and all places, I do not set them, antecedently to experience, in space and time. The representation is nothing but the thought of a possible experience in its absolute completeness. Since the objects are mere representations, only in such a possible experience are they given. To say that they exist prior to all my experience, can only be taken as meaning that they will be met with, if, starting from actual perception, I advance to that part of experience to which they belong. The cause of the empirical conditions of this advance (that which determines what members I shall meet with, or how far I can meet with any such in my regress) is transcendental, and is therefore necessarily unknown to me. We are not, however, concerned with this transcendental cause, but only with the rule of progression in that experience in which objects, that is to say, appearances, are given. Moreover, in outcome it is a matter of indifference whether I say that in the empirical progress in space I can meet with stars a hundred times farther removed than the outermost now perceptible to me, or whether I say that they are perhaps to be met with in cosmical space even though no human being has ever perceived or ever will perceive them. For though they might be given as things in themselves, without relation to possible experience, they are still nothing to me, and therefore are not objects, save in so far as they are contained in the series of the empirical regress.”

But though this is a possible interpretation of the teaching of the Postulates, and though further it is Kant’s own interpretation in another portion of the Critique, it is not by any means thereby decided that this is what the section itself actually teaches. Unbiassed study of the section, in independence of the use to which it is elsewhere put, can find within it no such limitation to its assertion of the actual independent existence of non-perceived bodies. We have to remember that the doctrine and solution of the Antinomies was completed prior to the writing of the central portions of the Critique. The section treating of their solution seems, indeed, in certain parts to be later than the other main portions of the chapter on the Antinomies, and must have been at least recast after completion of the Postulates. But the subjectivist solution is so much simpler in statement, so much more fully worked out, and indeed so much more capable of definite formulation, and also so much more at one with the teaching developed in the preceding chapter on the Paralogisms, that even granting the doctrine expounded in the section on the Postulates to be genuinely phenomenalist, it is not surprising that Kant should have been unwilling to recast his older and simpler solution of the Antinomies. In any case we are not concerned to argue that Kant, even after formulating the phenomenalist view, yields to it an unwavering adherence. As I have already insisted, his attitude continues to the very last to be one of alternation between two opposed standpoints.

But the most significant feature of Kant’s treatment of the argument of the Postulates still remains for consideration. It was in immediate succession to the paragraph above quoted that Kant, in the second edition, placed his “Refutation of Idealism” with the emphatic statement that this (not as in the first edition in connection with the Paralogisms) was its “correct location.” It is required, he says, as a reply to an objection which the teaching of the Postulates must at once suggest. The argument of the second edition in proof of the independent reality of material bodies, and in disproof of subjectivism, is thus given by Kant as a necessary extension and natural supplement of the teaching of the first edition.

There is therefore reason for concluding that the same preconception which has led to such radical misinterpretation of Kant’s Refutation of Idealism has been at work in inducing a false reading of Kant’s argument in the Postulates, namely the belief that Kant’s teaching proceeds on consistent lines, and that it must at all costs be harmonised with itself. Finding subjectivism to be emphatically and unambiguously inculcated in all the main sections of the Critique, and the phenomenalist views, on the other hand, to be stated in a much less definite and somewhat elusive manner, commentators have impoverished the Critical teaching by suppression of many of its most subtile and progressive doctrines. Kant’s experimental, tentative development of divergent tendencies is surely preferable to this artificial product of high-handed and unsympathetic emendation.

INNER SENSE AND APPERCEPTION

We are now in position to complete our treatment of inner sense. When the inner world of feelings, volitions, and representations is placed on the same empirical level as the outer world of objects in space, when the two are correlated and yet also at the same time sharply distinguished, when, further, it is maintained that objects in space exist independently of their representations, and that in this independence they are necessary for the possibility of the latter, the whole aspect of the Critical teaching undergoes a genial and welcome transformation. Instead of the forbidding doctrine that the world in space is merely my representation, we have the very different teaching that only through consciousness of an independent world in space is consciousness of the inner subjective life possible at all, and that as each is “external” to the other, neither can be reduced to, or be absorbed within, the other. The inner representations do not produce or generate the spatial objects, do not even condition their existence, but are required only for the individual’s empirical consciousness of them. Indeed the relations previously holding between them are now reversed. It is the outer world which renders the subjective representations possible. The former is prior to the latter; the latter exist in order to reveal the former. The outer world in space must, indeed, be regarded as conditioned by, and relative to, the noumenal conditions of its possibility; but these, on Kant’s doctrine of outer and inner sense, are distinct from all experienced contents and from all experienced mental processes. This will at once be recognised as holding of the noumenal conditions of the given manifold. But it is equally true, Kant maintains, in regard to the noumenal conditions of our mental life. We have no immediate knowledge of the transcendental syntheses that condition all consciousness, and in our complete ignorance of their specific nature they cannot legitimately be equated with any individual or personal agent. As the empirical self is only what it is known as, namely, appearance, it cannot be the bearer of appearance. This function falls to that which underlies both inner and outer appearances equally, and which within experience gains twofold expression for itself, in the conception of the thing in itself = x on the one hand, and in the correlative conception of a transcendental subject, likewise = x, on the other.

But with mention of the transcendental subject we are brought to a problem which in the second edition invariably accompanies Kant’s discussion of inner sense. The ‘I think’ of apperception can find expression only in an empirical judgment, and yet, so far from being the outcome of inner sense, preconditions its possibility. What then is its relation to inner sense? Does not its recognition conflict with Kant’s denial of the possibility of self-conscious reflection, of direct intuitive apprehension by the self of itself? The pure apperception, ‘I think,’ is equivalent, Kant declares, to the judgment ‘I am,’ and therefore involves the assertion of the subject’s existence. Does not this conflict on the one hand with the Critical doctrine that knowledge of existence is only possible in terms of sense, and on the other with the Critical limitation of the categories to the realm of appearance? How are such assertions as that the ‘I think’ of pure apperception refers to a non-empirical reality, and that it predicates its existence, to be reconciled with the doctrine of inner sense as above stated?

As we have already observed, Kant’s early doctrine of the transcendental object was developed in a more or less close parallelism with that of the transcendental unity of apperception. They were regarded as correlative opposites, the dual centres of noumenal reference for our merely subjective representations. Kant’s further examination of the nature of apperception, as embodied in alterations in the second edition, was certainly, as we shall find, inspired by the criticisms which the first edition had called forth. His replies, however, are merely more explicit statements of the distinction which he had already developed in the first edition between the transcendental and the empirical self, and that distinction in turn was doubtless itself largely determined by his own independent recognition of the untenability of his early view of the transcendental object. Though it is much more difficult to differentiate between the empirical and the transcendental self than to distinguish between the empirical object and the thing in itself, both distinctions are from a genuinely Critical standpoint equally imperative, and rest upon considerations that are somewhat similar in the two cases.

One of the chief and most telling criticisms directed against the teaching of the first edition was that Kant’s doctrine of a transcendental consciousness of the self’s existence, i.e. of the existence of a noumenal being, “this I or he or it (the thing) which thinks,” is inconsistent with the teaching of the Postulates of Empirical Thought. In that section, as also later in the section on the theological Ideal, Kant had declared most emphatically that existence is never discoverable in the content of any mere concept. It is revealed in perception, and in perception alone, in virtue of the element of sensation contained in the latter.

“...to know the actuality of things demands perception, and therefore sensation.... For that the concept precedes perception, signifies the concept’s mere possibility; the perception which supplies the content [Stoff] to the concept, is the sole criterion [Charakter] of actuality.”

Yet Kant had also maintained that the ‘I think’ is equivalent to ‘I am,’ and that in this form, as an intellectual consciousness of the self’s existence, it precedes all experience. The teaching of the Postulates is, however, the teaching of the Critique as a whole, and such critics as Pistorius seemed therefore to be justified in maintaining that Kant, in reducing the experiences of inner sense to mere appearance, destroys the possibility of establishing reality in any form. Appearance, in order to be appearance, presupposes the reality not only of that which appears, but also of the mental process whereby it is apprehended. But if reality is given only in sensation, and yet all experience that involves sensation is merely appearance, there is no self by which appearance can be conditioned; and only illusion (Schein), not appearance (Erscheinung), is left. To quote Pistorius’ exact words:

”[If our inner representations are not things in themselves but only appearances] there will be nothing but illusion (Schein), for nothing remains to which anything can appear.”

Kant evidently felt the force of this criticism, for in the second edition he replies to it on no less than seven different occasions. In three of these passages the term Schein is employed, and in the note to B xxxix the term Erdichtung appears. This shows very conclusively that it is such criticism as the above that Kant has in mind. The most explicit passage is B 428:

“The proposition, ‘I think,’ or ‘I exist thinking,’ is an empirical proposition. Such a judgment, however, is conditioned by empirical intuition, and the object that is thought therefore underlies it as appearance. It would consequently seem that on our theory the soul is completely transformed, even in thinking [selbst im Denken], into appearance, and that in this way our consciousness itself, as being a mere illusion [Schein], must refer in fact to nothing.”

Kant, in his reply, is unyielding in the contention that the ‘I think,’ even though it involves an empirical judgment, is itself intellectual. “This representation is a thinking, not an intuiting,” or as he adds, “The ‘I think’ expresses the actus whereby I determine my existence.” Existence is therefore already given thereby. Kant also still maintains that the self thus revealed is not “appearance and still less illusion.”

“I am conscious of myself ..., not as I appear to myself, nor as I am in myself, but only that I am.” “I thereby represent myself to myself neither as I am nor as I appear to myself. I think myself only as I do any object in general from whose mode of intuition I abstract.”

Kant’s method of meeting the criticism, while still holding to these positions, is twofold. It consists in the first place in maintaining that the ‘I think,’ though intellectual, can find expression only in empirical judgments--in other words, that it is in and by itself formal only, and presupposes as the occasion of its employment a given manifold of inner sense; and secondly, by the statement that the ‘existence’ which is involved in the ‘I think’ is not the category of existence. Let us take in order each of these two points.

Kant’s first method of reply itself appears in two forms, a stronger and a milder. The milder mode of statement is to the effect that though the representation ‘I am’ already immediately involves the thought of the existence of the subject, it yields no knowledge of it. Knowledge would involve intuition, namely, consciousness of inner determinations in time, which in turn would itself presuppose consciousness of outer objects. As a merely intellectual representation,

“...this ‘I’ has not the least predicate of intuition which, in its character of permanence, could, somewhat after the manner of impenetrability in the empirical intuition of matter, serve as correlate of time determination in inner sense.”

The stronger and more definite mode of statement is that the ‘I think’ is an empirical proposition. Though it involves as one factor the intellectual representation, ‘I think,’ it is none the less empirical.

“Without some empirical representation supplying the material for thought, the actus, ‘I think,’ would not take place....”

The empirical is indeed “only the condition of the application or employment of the pure intellectual faculty,” but as such is indispensable. This is repeated in even clearer terms in B 429.

“The proposition, ‘I think,’ in so far as it amounts to the assertion, ‘I exist thinking,’ is no mere logical function but determines the subject (which is then at the same time object) in respect of existence, and cannot take place without inner sense....”

This admission is the more significant in that it follows immediately upon a passage in which Kant has been arguing that thinking, taken in and by itself, is a merely logical function.

The real crux lies in the question as to the legitimacy of Kant’s application of the predicate existence to the transcendental subject. Its employment in reference to the empirical self in time is part of the problem of the Refutation of Idealism in the second edition; and the answer there given is clear and definite. Consciousness of the empirical self as existing in time involves consciousness of outer objects in space. But as Kant recognises that a transcendental ego, not in time, is presupposed in all consciousness of the empirical self, the question whether the predicate of existence is also applicable to the transcendental self cannot be altogether avoided, and is indeed referred to in B 277. The attitude to be taken to this latter question is not, however, defined in that section.

In the first edition Kant has insisted that the categories as pure forms of the understanding, in isolation from space and time, are merely logical functions “without content.” Interpreted literally, this would signify that they are devoid of meaning, and therefore are incapable of yielding the thought of any independent object or existence. As merely logical forms of relation, they presuppose a material, and that is supplied only through outer and inner sense. Such is not, however, the way in which Kant interprets his own statement. It is qualified so as to signify only that they are without specific or determinate content. They are taken as yielding the conception of object in general. Passages in plenty can be cited from the first edition--passages allowed to remain in the second edition--in which Kant teaches that the pure forms of understanding, as distinct from the schematised categories, yield the conception of things in themselves. This view is, indeed, a survival from his earlier doctrine of the transcendental object. In all passages added in the second edition the consequences of his argument are more rigorously drawn, and the doctrine of the transcendental object is entirely eliminated. It is now unambiguously asserted that the pure forms of understanding, the “modes of self-consciousness in thinking,” are not intellectual concepts of objects. They “yield no object whatsoever.” The only object is that given through sense. And since in thinking the transcendental subject we do, by Kant’s own account, think an “object,” he is led to the conclusion, also explicitly avowed, that the notion of existence involved in the ‘I think’ is not the category of the same name. So also of the categories of substance and causality.

“If I represent myself as subject of thoughts or as ground of thinking, these modes of representation do not signify the categories of substance or of cause....”

The notion of the self, like the notion of things in themselves, is a concept distinct from all the categories.

This conclusion is reinforced by means of an argument which is employed in the section of the first edition on Paralogisms. Apperception is the ground of the possibility of the categories, and these latter on their side represent only the synthetic unity which that apperception demands. Self-consciousness is therefore the representation of that which is the condition of all unity, and which yet is itself unconditioned.

“...it does not represent itself through the categories, but knows the categories and through them all objects in the absolute unity of apperception, and so through itself. Now it is, indeed, very evident that I cannot know as an object that which I must presuppose in order to know any object....”

This argument recurs in B 422.

“The subject of the categories cannot by thinking the categories acquire a conception of itself as an object of the categories. For, in order to think them, its pure self-consciousness, which is what was to be accounted for, must itself be presupposed.”

It is extremely difficult to estimate the value and cogency of this argument. Many objections or rather qualifications must be made before it can be either accepted or rejected. If it be taken only as asserting that the unity of self-consciousness is not adequately expressible through any of the categories, it is undoubtedly valid. If, further, the categories be identified with the schemata, it is also true that they are not applicable in any degree or manner. The schemata are applicable only to natural existences in space and time. Self-consciousness can never be reduced to a natural existence of that type. On the other hand, if it is not self-consciousness as such, but the self-conscious subject, which on Kant’s view is always noumenal--“this I or he or it (the thing) which thinks”--that is referred to, and if we distinguish between the categories strictly so called, that is, the pure forms of understanding, and the schemata, it is not at all evident that the self-conscious subject may not be described as being an existence that is always a subject and never a predicate, and as being related to experience as a ground or condition. These indefinite assertions leave open alternative possibilities. They do not even decide whether the self is “I or he or it.” In so far as they advance beyond the mere assertion that the self rests upon noumenal conditions they are, indeed, incapable of proof, but by no Critical principle can they be shown to be inapplicable. When, therefore, Kant may seem to extract a more definite conclusion from the above argument, he advances beyond what it can be made to support.

Kant is here influenced by the results of the ethical enquiries with which in the period subsequent to 1781 he was chiefly preoccupied. He believed himself to have proved that the self, as a self-conscious being, is a genuinely noumenal existence. That being so, he was bound to hold that the categories, even as pure logical forms, are inadequate to express its real determinate nature. But he confounds this position with the assertion that they are not only inadequate, but in and by themselves are likewise inapplicable. That is not a legitimate conclusion, for even if the self is more than mere subject or mere ground, it will at least be so much. When ethical considerations are left out of account, the only proper conclusion is that the applicability of the categories to the self-conscious subject is capable neither of proof nor of disproof, but that when the distinction between appearance and reality (which as we shall find is ultimately based upon the Ideas of Reason) has been drawn, the categories can be employed to define the possible difference between self-conscious experience and its unknown noumenal conditions. Any other conclusion conflicts with the teaching of the section on the Paralogisms.

It is important to observe--a point ignored by such critics as Caird and Watson--that in the sections under consideration Kant most explicitly declares self-consciousness to be merely “the representation of that which is the condition of all unity.” He maintains that this representation, as standing for “the determining self (the thinking), is to be distinguished from the self which we are seeking to determine (the subject which thinks) as knowledge from its object,” or in other words, that, without special proof, unattainable on theoretical grounds, “the unity of thought” may not be taken as equivalent to the unity of the thinking subject. They may be as diverse as unity of representation and unity of object represented are frequently found to be. We may never argue from simplicity in a representation to simplicity in its object.

But to return to the main thesis, it may be observed that these arguments, with the exception of that which we have just been considering from the nature of self-consciousness, lead to the conclusion that the categories are as little applicable to the thing in itself as to the transcendental subject. Even the argument from the necessary and invariable presence of self-consciousness in each and every act of judgment is itself valid only from a point of view which regards self-consciousness in the manner of Kant’s early semi-Critical view of the transcendental subject as an ultimate. But if, as is maintained in the section in which this argument occurs, viz. that on the Paralogisms, self-consciousness may be complexly conditioned, and may indeed have conditions similar in nature to those which underlie outer experience, the categories may be just as applicable, or as inapplicable, to its noumenal nature as to the nature of the thing in itself. It is noticeable that in the second edition, doubtless under the influence of preoccupation with ethical problems, some of Kant’s utterances betray a tendency to relax the rigour of his thinking, and to bring his theoretical teaching into closer agreement with his ethical results than the theoretical analysis in and by itself at all justifies. This tendency was, of course, reinforced by the persisting influence of that view of the transcendental subject which he had held in the middle ’seventies, and from which he never completely emancipated either his language or his thinking. Indeed in several of the passages added in the second edition Kant even goes so far as to adopt language which if taken quite literally would mean that the ‘I think’ is an immediate consciousness of the mind’s purely intellectual activity--a view which, as we have seen, is altogether alien to the Critical position. It would, as he argues so forcibly elsewhere, involve a kind of experience which does not conform to Critical requirements, and which would lie open to the attacks of sceptics such as Hume.

In B 157-8 the difficulties of Kant’s position are again manifest. Speaking of the representation of the self, he declares that “I am conscious of myself ..., not as I appear to myself, nor as I am in myself, but only that I am.” This may seem to imply that existence is predicable of the transcendental self. He adds that though the determination, i.e. specification in empirical form, of my existence (mein eigenes Dasein) is possible only in inner sensuous intuition, it is “not appearance and still less mere illusion.” But in the appended note it is urged that my existence (Dasein) as self-active being is represented in purely indeterminate fashion. Only my existence as sensuous, and therefore as appearance, can be known, i.e. can be made determinate.

The problem is more directly and candidly faced in the note to B 422. That note is interesting for quite a number of reasons. It reveals Kant in the very act of recasting his position, and in the process of searching around for a mode of formulation which will enable him to hold to a transcendental consciousness of the self’s existence and at the same time not to violate the definition of existence given in the Postulates, i.e. both to posit the transcendental self as actual and yet to deny the applicability to it of any of the categories. After stating that the ‘I think’ is an empirical proposition in which my existence is immediately involved, he proceeds further to describe it as expressing “an undetermined empirical intuition, i.e. perception,” and so as showing that sensation underlies its assertion of existence. Kant does not, however, mean by these words that the existence asserted is merely that of the empirical self; for he proceeds:

“...existence is here not a category, which as such does not apply to an indeterminately-given object.... An indeterminate perception here signifies only something real that is given, given indeed to thought in general, and so not as appearance, nor as thing in itself (Noumenon), but as something which actually [in der That] exists, and which in the proposition, I think, is denoted [bezeichnet] as such.”

The phrases here employed are open to criticism on every side. Kant completely departs from his usual terminology when he asserts that through an “indeterminate perception” the self is given, and “given to thought in general” as “something real.” The contention, that the existence asserted is not a category, is also difficult to accept. It is equally surprising to read that its reality is given “neither as appearance nor as thing [Sache] in itself (Noumenon)”; for hitherto no such alternative form of real existence has been recognised.

But to press such criticisms is to ignore the spirit for the sake of the letter. Kant here breaks free from all his habitual modes of expression for the very good and sufficient reason that he is striving to develop a position more catholic and comprehensive than any previously adopted. He is seeking to formulate a position which, without in any way justifying or encouraging the transcendent employment of the categories, will yet retain for thought the capacity of self-limitation, that is, of forming concepts which will reveal the existence of things in themselves and so will enable the mind to apprehend the radical distinction between things in themselves and things experienced. But he has not yet discovered that in so doing he is committing himself to the thesis that the distinction is mediated, not by the understanding, but by Reason, not by categories, but by Ideas. As I have already indicated, this tendency is crossed by another derived from his preoccupation with moral problems, namely, the desire to defend, in a manner which his Critical teaching does not justify, the noumenal existence of the self as a thinking being.

THE TRANSCENDENTAL ANALYTIC