an object, but even constitute the very concept of it? Are
appearances legitimately interpretable in any such manner ? It
was, we may believe, in the process of answering this question
that Kant came to realise that the objects of our repre-
sentations must no longer be regarded as things in them-
selves. For, as he finds, a solution is possible only on the
further assumption that the mind is legislating merely for the
world of sense-experience, and is making no assertion in
regard to the absolutely and independently real. Kant's
method of proof is the transcendental, i.e. he seeks to demon-
strate that this interpretation of the given is indispensably
necessary as being a sine qua non of its possible apprehension.
This is achieved by means of the conclusion already established
through the preliminary steps of the subjective deduction,
namely, that all consciousness involves self -consciousness.
Kant's proof of the objective validity of the categories con-
sists in showing that only by means of the interpretation of
appearances as empirically objective is self -consciousness
possible at all.
The self-consciousness of the subjective deduction, in the
preliminary form above stated, is, however, itself empirical.
Kant, developing on more strictly Critical lines the argument
which had accompanied his earlier doctrine of the transcend-
ental object, now proceeds to maintain in what is at once
the most fruitful and the most misleading of his tenets,
that the ultimate ground of the possibility of consciousness
and therefore also of empirical self-consciousness is the tran-
scendental unity of apperception. Such apperception, to use
Kant's ambiguous phraseology, precedes experience as its
THE OBJECTIVE DEDUCTION 251
a priori condition. The interpretation of given appearances
through a priori categories is a necessity of consciousness
because it is a condition of self-consciousness ; and it is a
condition of self-consciousness because it alone will account
for the transcendental apperception upon which all empirical
self-consciousness ultimately depends.
One chief reason why Kant's deduction is found so baffling
and illusive is that it rests upon an interpretation of the unity
of apperception which is very definitely drawn, but to which
Kant himself gives only the briefest and most condensed
expression. I shall therefore take the liberty of restating it
in more explicit terms. The true or transcendental self has
no content of its own through which it can gain knowledge
of itself. It is mere identity, I am I. In other words, self-
consciousness is a mere form through which contents that
never themselves constitute the self are yet apprehended as
being objects to the self. Thus though the self in being
conscious of time or duration must be conscious of itself as
identical throughout the succession of its experiences, that
identity can never be discovered in those experiences ; it can
only be thought as a condition of them. The continuity
of memory, for instance, is not a possible substitute for
transcendental apperception. As the subjective deduction
demonstrates, self - consciousness conditions memory, and
cannot therefore be reduced to or be generated by it. 1
When, however, such considerations are allowed their due
weight, the necessity of postulating a transcendental unity
becomes only the more evident. Though it can never itself
be found among appearances, it is an interpretation which
we are none the less compelled to give to appearances.
To summarise before proceeding. We have obtained two
important conclusions : first, that all consciousness involves
self-consciousness ; and secondly, that self-consciousness is a
mere form, in terms of which contents that do not constitute
the self are apprehended as existing for the self. The first
leads up to the second, and the second is equivalent to the
assertion that there can be no such thing as a pure self-
consciousness, i.e. a consciousness in which the self is aware
of itself and of nothing but itself. Self-consciousness, to be
1 Memory is only one particular mode in which recognition presents itself in
our experience ; Kant's purpose is to show that it is not more fundamental, nor
more truly constitutive of apperception, than is recognition in any of its other
manifestations. Indeed the central contention of the objective deduction is that
it is through consciousness of objects, i.e. through consciousness of objective mean-
ings, that self-consciousness comes to be actualised at all. Only in contrast with,
and through relation to, an objective system is consciousness of inner experience,
past or present, and therefore self-consciousness in its contingent empirical forms,
possible to the mind. Cf. above, pp. li-ii ; below, pp. 260-3.
252 THE ANALYTIC OF CONCEPTS
possible at all, must at the same time be a consciousness of
something that is not-self. Only one further step is now
required for the completion of the deduction, namely, proo
that this not-self, consciousness of which is necessary to th
possibility of self -consciousness, must consist in empirica
objects apprehended in terms of the categories. For proo
Kant again appeals to the indispensableness of apperception.
As no intuitions can enter consciousness which are not capable
of being related to the self, they must be so related to one
another that, notwithstanding their variety and diversity, the
self can still be conscious of itself as identical throughout
them all. In other words, no intuition can be related to the
self that is incapable of being combined together with all the
other intuitions to form a unitary consciousness. I may here
quote from the text of the second edition : l
"... only in so far as. I can grasp the manifold of the repre-
sentations in one consciousness, do I call them one and all mine.
For otherwise I should have as many-coloured and diverse a self as
I have representations of which I am conscious to myself."
Or as it is stated in the first edition : 2
"We are a priori aware of the complete identity of the self in
respect of all representations which belong to our knowledge ... as
a necessary condition of the possibility of all representations."
These are the considerations which lead Kant to entitle
the unity of apperception transcendental. He so names it for
the reason that, though it is not itself a priori in the manner
of the categories, we are yet enabled by its means to
demonstrate that the unity which is necessary for possible
experience can be securely counted upon in the manifold of
all possible representations, and because (as he believed) it
also enables us to prove that the forms of such unity are the
categories of the understanding.
To the argument supporting this last conclusion Kant
does not give the attention which its importance would seem
to deserve. He points out that as the given is an un-
connected manifold, its unity can be obtained only by syn-
thesis, and that such synthesis must conform to the conditions
prescribed by the unity of apperception. That these conditions
coincide with the categories he does not, however, attempt to
prove. He apparently believes that this has been already
established in the metaphysical deduction. 3 The forms of
unity demanded by apperception, he feels justified in assuming,
1 B 134. 2 A 116. 3 Cf. above, p. 242; below, pp. 258, 332-3.
THE DOCTRINE OF OBJECTIVE AFFINITY 253
are the categories. They may be regarded as expressing
the minimum of unity necessary to the possibility of self-
consciousness. If sensations cannot be interpreted as the
diverse attributes of unitary substances, if events cannot be
viewed as arising out of one another, if the entire world in
space cannot be conceived as a system of existences re-
ciprocally interdependent, all unity must vanish from experi-
ence, and apperception will be utterly impossible. 1
The successive steps of the total argument of the deduction,
as given in the first edition, are therefore as follows : Conscious-
ness of time involves empirical self-consciousness ; empirical
self-consciousness is conditioned by a transcendental self-con-
sciousness ; and such transcendental self-consciousness is itself,
in turn, conditioned by consciousness of objects. The argument
thus completed becomes the proof of mutual interdependence.
Self - consciousness and consciousness of objects, as polar
opposites, mutually condition one another. Only through
consciousness of both simultaneously can consciousness of
either be attained. Only in and through reference to an
object can an idea be related to a self, and so be accompanied
by that self-consciousness which conditions recognition, and
through recognition all the varying forms in which our
consciousness can occur. From the point of view, however,
of a Critical enquiry apperception is the more important of
the two forms of consciousness. For though each is the
causa existendi of the other, self-consciousness has the unique
distinction of being the causa cognoscendi of the objective
and a priori validity of the forms of understanding.
" The synthetic proposition, that all the variety of empirical
consciousness must be combined in a single self-consciousness, is the
absolutely first and synthetic principle of our thought in general." 2
We may at this point consider Kant's doctrine of
" objective affinity." It excellently enforces the main thesis
which he is professing to establish, namely, that the conditions
of unitary consciousness are the conditions of all conscious-
ness. The language, however, in which the doctrine is
expounded is extremely obscure and difficult ; and before
commenting upon Kant's own methods of statement, it
seems advisable to paraphrase the argument in a somewhat
free manner, and also to defer consideration of the transcend-
ental psychology which Kant has employed in its exposition. 3
Association can subsist only between ideas, both of which have
1 Cf. A in. a A 117 n.
3 This transcendental psychology is considered below (p. 263 ff.), in its
connection with the later stages of the subjective deduction. Cf. above, p. 238.
254 THE ANALYTIC OF CONCEPTS
occurred within the same conscious field. Now the funda-
mental characteristic of consciousness, the very condition of
its existing at all, is its unity ; and until this has been
recognised, there can be no understanding of the associative
connection which arises under the conditions which con-
sciousness supplies. To attempt to explain the unity of
consciousness through the mechanism of association is to
explain an agency in terms of certain of its own effects. It is
to explain the fundamental in terms of the derivative, the
conditions in terms of what they have themselves made
possible. Kant's argument is therefore as follows. Ideas do
not become associated merely by co-existing. They must
occur together in a unitary consciousness ; and among the
conditions necessary to the possibility of association are
therefore the conditions of the possibility of experience.
Association is transcendentally grounded. So far from
accounting for the unity of consciousness, it presupposes the
latter as determining the conditions under which alone it can
come into play.
"... how, I ask, is association itself possible? , . . On my
principles the thorough - going affinity of appearances is easily
explicable. All possible appearances belong as representations
to the totality of a possible self-consciousness. But as this self-
consciousness is a transcendental representation, numerical identity
is inseparable from it and is a priori certain. For nothing can
come to our knowledge save in terms of this original apperception.
Now, since this identity must necessarily enter into the synthesis of
all the manifold of appearances, so far as the synthesis is to yield
empirical knowledge, the appearances are subject to a priori con-
ditions, with which the synthesis of their apprehension must be in
complete accordance. . . . Thus all appearances stand in a thorough-
going connection according to necessary laws, and therefore in a
transcendental affinity of which the empirical is a mere conse-
In other words, representations must exist in conscious-
ness before they can become associated ; and they can exist
in consciousness only if they are consciously apprehended.
But in order to be consciously apprehended, they must
conform to the transcendental conditions upon which all
consciousness rests ; and in being thus apprehended they are
set in thoroughgoing unity to one another and to the self.
They are apprehended as belonging to an objective order or
unity which is the correlate of the unity of self-consciousness.
This is what Kant entitles their objective affinity ; it is what
1 A 113-14.
THE DOCTRINE OF OBJECTIVE AFFINITY 255
conditions and makes possible their associative or empirical
This main point is very definitely stated in A 101.
"If we can show that even our purest a priori intuitions yield
no knowledge, save in so far as they contain such a connection of the
manifold as will make possible a thoroughgoing synthesis of
reproduction, this synthesis of the imagination " [which acts through
the machinery of association] "must be grounded, prior to all
experience, on a priori principles, and since experience necessarily
presupposes that appearances can be reproduced, we shall have to
assume a pure transcendental synthesis of the imagination" [i.e.
such synthesis as is involved in the unity of consciousness] "as
conditioning even the possibility of all experience." l
In A 1 2 1-2 Kant expresses his position in a more
ambiguous manner. He may seem to the reader meroJy to
be arguing that a certain minimum of regularity is necessary
in order that representations may be associated, and experience
may be possible. 2 But the general tenor of the passage as a
whole, and especially its concluding sentences, enforce the
stronger, more consistent, thesis.
" [The] subjective and empirical ground of reproduction accord-
ing to rules is named the association of representations. If
this unity of association did not also have an objective ground,
which makes it impossible that appearances should be apprehended
by the imagination except under the condition of a possible synthetic
unity of this apprehension, it would be entirely accidental that
appearances should fit into a connected whole of human knowledge.
For even though we had the power of associating perceptions, it
would remain entirely undetermined and accidental whether they
would themselves be associable ; and should they not be associable,
there might exist a multitude of perceptions, and indeed an entire
sensibility, in which much empirical consciousness would arise in my
mind, but in a state of separation, and without belonging to one
consciousness of myself. That, however, is impossible. For only
in so far as I ascribe all perceptions to one consciousness (original
apperception), can I say in all perceptions that I am conscious of
them. There must therefore be an objective ground (that is, one
that can be recognised a priori^ antecedently to all empirical laws
of the imagination) upon which may rest the possibility, nay the
necessity, of a law that extends to all appearances. ..."
Kant is not merely asserting that the associableness of
ideas, and the regularity of connection which that implies,
must be postulated as a condition of experience. That would
be a mere begging of the issue ; the correctness of the
1 Cf. above, p. 229. a c f> A JQQ . I
256 THE ANALYTIC OF CONCEPTS
postulate would not be independently proved. Kant is really
maintaining the much more important thesis, that the unity
of experience, i.e. of consciousness, is what makes association
possible at all. And since consciousness must be unitary in
order to exist, there cannot be any empirical consciousness in
which the conditions of association, and therefore of reproduc-
tion, are not to be found.
A further misunderstanding is apt to be caused by Kant's
statement that associative affinity rests upon objective affinity.
This seems to imply, in the same manner as the passage
which we have just considered, that instead of proving that
appearances are subject to law and order, he is merely
postulating that an abiding ground of such regularity must
exist in the noumenal conditions of the sense manifold. But
he himself again supplies the needful correction.
" This [objective ground of all association of appearances] can
nowhere be found, except in the principle of the unity of appercep-
tion in respect of all forms of knowledge which can belong to me.
In accordance with this principle all appearances must so enter the
mind, or be so apprehended, that they fit together to constitute the
unity of apperception. This would be impossible without synthetic
unity in their connection, and that unity is therefore also objectively
necessary. The objective unity of all empirical consciousness in
one consciousness, that of original apperception, is therefore the
necessary condition of all (even of all possible) perception ; and the
affinity of all appearances, near or remote, is a necessary consequence
of a synthesis in imagination which is grounded a priori on rules." *
The fundamental characteristic of consciousness is the
unified form in which alone it can exist ; only when this
unity is recognised as indispensably necessary, and therefore
as invariably present whenever consciousness exists at all, can
the inter-relations of the contents of consciousness be properly
If this main contention of the Critical teaching be accepted,
Hume's associationist standpoint is no longer tenable.
Association cannot be taken to be an ultimate and in-
explicable property of our mental states. Nor is it a property
which can be regarded as belonging to presentations viewed
as so many independent existences. It is conditioned by
the unity of consciousness, and therefore rests upon the
" transcendental " conditions which Critical analysis reveals.
Since the unity oj" consciousness conditions association, it
cannot be explained as the outcome and product of the
mechanism of association.
1 A 122-3.
CATEGORIES HAVE FACTUAL VALIDITY 257
In restating the objective deduction in the second edition,
Kant has omitted all reference to this doctrine of objective
affinity. His reasons for this omission were probably twofold.
In the first place, it has been expounded in terms of a tran-
scendental psychology, which, as we shall find, is conjectural
in character. And secondly, the phrase " objective affinity "
is, as I have already pointed out, decidedly misleading. It
seems to imply that Kant is postulating, without independent
proof, that noumenal conditions must be such as to supply
an orderly manifold of sense data. But though the doctrine
of objective affinity is eliminated, its place is to some extent
taken l by the proof that all apprehension is an act of judg-
ment and therefore involves factors which cannot be reduced
to, or explained in terms of, association.
There are a number of points in the deduction of the first
edition which call for further explanatory and critical com-
ment. The first of these concerns the somewhat misleading
character of the term a priori as applied to the categories. It
carries with it rationalistic associations to which the Critical
standpoint, properly understood, yields no support. The
categories are for Kant of merely de facto nature. They have
no intrinsic validity. They are proved only as being the
indispensable conditions of what is before the mind as brute
fact, namely, conscious experience. By the a priori is meant
merely those relational factors which are required to supple-
ment the given manifold in order to constitute our actual
consciousness. And, as Kant is careful to point out, the
experience, as conditions of which their validity is thus
established, is of a highly specific character, resting upon
synthesis of a manifold given in space and time. That is to
say, their indispensableness is proved only for a consciousness
which in these fundamental respects is constituted like our
own. 2 And secondly, the validity of the a priori categories,
even in our human thinking, is established only in reference
to that empirical world which is constructed out of the given
manifold in terms of the intuitive forms, space and time.
Their validity is a merely phenomenal validity. They are
valid of appearances, but not of things in themselves. The
a priori is thus doubly de facto : first as a condition of brute
fact, namely, the actuality of our human consciousness ; and
1 Cf. B 140-3 ; B 151-2 ; B 164-5 5 an< ^ below, p. 286.
2 Here again the second edition text is more explicit than the first : "This
peculiarity of our understanding, that it can produce a priori unity of apperception
solely by means of the categories, and only by such and so many, is as little
capable of further explanation as why we have just these and jio other functions
of judgment, or why space and time are the sole forms of our possible intuition."
B 145-6. Cf. above, pp. xxxiii-vi, xliv, 57, 142, 186 ; below, pp. 291, 411.
258 THE ANALYTIC OF CONCEPTS
secondly, as conditioning a consciousn'ess whose knowledge is
limited to appearances. It is a relative, not an absolute
a priori. Acceptance of it does not, therefore, commit us
to rationalism in the ordinary meaning of that term. Its
credentials are conferred upon it by what is mere fact ; it does
not represent an order superior to the actual and legislative for
it. In other words, it is Critical, not Leibnizian in character.
No transcendent metaphysics can be based upon it. In
formulating this doctrine of the a priori as yielding objective
insight and yet as limited in the sphere of its application, the
Critique of Pure Reason marks an epoch in the history of
scepticism, no less than in the development of Idealist teaching.
There is one important link in the deduction, as above
given, which is hardly calculated to support the conclusions
that depend upon it. Kant, as we have already noted, 1 asserts
that the categories express the minimum of unity necessary
for the possibility of apperception. A contention so essential
to the argument calls for the most careful scrutiny and a
meticulous exactitude of proof. As a matter of fact, such
proof is not to be found in any part of the deductions,
whether of the first or of the second editions. It is attempted
only in the later sections on the Principles of Understanding,
and even there it is developed, in any really satisfactory
fashion, only in regard to the categories of causality and
reciprocity. 2 This proof, however, as there given, is an
argument which in originality, subtlety and force goes far to
atone for all shortcomings. It completes the objective deduc-
tion by developing in mastexly fashion (in spite of the diffuse
and ill-arranged character of the text) the central contention
for which the deduction stands. But in the transcendental
deduction itself, we find only such an argument if it may be
called an argument as follows from the identification of
apperception with understanding.
"The unity of apperception, in relation to the synthesis of
imagination, is the understanding. ... In understanding there are
pure a priori forms of knowledge which contain the necessary unity
of pure synthesis of imagination in respect of all possible appearances.
But these are the categories, i.e. pure concepts of understanding." 3
The point is again merely assumed in A 125-6. So also
in A 126 :
"Although through experience we learn many laws, these are
only special determinations of still higher laws, of which the highest,
11 Cf. above, pp. 252-3.
2 The second Analogy embodies the argument which is implied in, and
necessary to, the establishment of the assertions dogmatically made in A 111-12.
3 A 119.
CATEGORIES VALID FOR APPEARANCES 259
under which all others stand, originate a priori in the understanding
itself. . . ."*
Again in A 129 it is argued that as we prescribe a priori
rules to which all experience must conform, those rules cannot
be derived from experience, but must precede and condition
it, and can do so only as originating from ourselves (aus uns
" [They] precede all knowledge of the object as [their] intellectual
form, and constitute a formal a priori knowledge of all objects in so
far as they are thought (categories)."-
But this is only to repeat that such forms of unity as
are necessary to self-consciousness must be realised in all