Norman Kemp Smith.

A commentary to Kant's 'Critique of pure reason,' online

. (page 11 of 72)
Online LibraryNorman Kemp SmithA commentary to Kant's 'Critique of pure reason,' → online text (page 11 of 72)
Font size
QR-code for this ebook

the Introduction of the second edition make no essential change,
except merely in the emphasis with which Kant's belief in the
existence of valid a priori synthetic judgments is insisted
upon. As has already been stated, only by virtue of this initial
assumption is Kant in position to maintain that there is an
alternative to the strict synthetic method. 1\ie problem from
which he starts is common to both methods, and for that reason
the formulation used in the Prolegomena can also be employed
in the Introduction to the Critique. Only in their manner of
solving the problem need they differ. 2 Kant's Critical problem
first begins with this presupposition of validity, and does not
exist save through it. 3 He does not first seek to discover

1 Cf. Vaihinger, i. p. 394. Cf. above, p. 28.

2 Cf. Vaihinger, i. pp. 415-17.

3 Paulsen objects that if synthetic a priori judgments are valid without
explanation, they do not need it. For two reasons the objection does not hold.
(a) Without this explanation it would be impossible to repel the pretensions of


whether such judgments are valid, and then to explain them.
He accepts them as valid, but develops a method of argument
which suffices for proof as well as for explanation. The
argument being directed to both points simultaneously, and
establishing both with equal cogency, it may legitimately be
interpreted in either way, merely as explanation, or also as
proof. Kant does not profess or attempt to keep exclusively
to any one line of statement. Against the dogmatists he
insists upon the necessity of explaining the validity of a priori
synthetic judgments, against the sceptics upon the possibility
of proving their validity. And constantly he uses ambiguous
terms, such as * justification ' (Recktfertigung), ' possibility/
that may indifferently be read in either sense. But though
the fundamental demand which characterises the synthetic
method in its distinction from the analytic thus falls into the
background, and is only occasionally insisted upon, it is none
the less fulfilled. So far as regards the main argument of the
Critique in either edition, the validity of synthetic a priori
judgments is not required as a premiss. It is itself inde-
pendently proved.

The manner in which Kant thus departs from the strict
application of the synthetic method may be illustrated by
an analysis of his argument in the Aesthetic}- Only in the
arguments of the first edition in regard to space and time is
the synthetic method employed in its ideal and rigorous form.
For the most part, even in the first edition, instead of
showing how the a priori character of pure and applied
mathematics follows from conclusions independently estab-
lished, he assumes both pure and applied mathematics to be
given as valid, and seeks only to show how the independently
established results of the Aesthetic enable him to explain and
render comprehensible their recognised characteristics. This
is not, indeed, any very essential modification of the synthetic
method ; for his independently established results suffice for
deducing all that they are used to explain. The validity of
mathematics is not employed as a premiss. Kant's argu-
ment is, however, made less clear by the above procedure.

Further difficulty is caused by Kant's occasional employ-
ment, even in the first edition, of the analytic method. He
several times cites as an argument in support of his view

transcendent metaphysics (cf. A 209 = 6 254-5 ; A 283 = 6 285). (b] This solution
of the theoretical problem has also, as above stated, its own intrinsic interest and
value. Without such explanation the validity of these judgments might be
granted, but could not be understood. (Cf. Prolegomena, 4-5 and 12 at the
end. Cf. Vaihinger, i. p. 394.)

1 Cf. Vaihinger, ii. p. 336. The argument of the Analytic, which is still more
complicated, will be considered later.


of space the fact that it alone will account for the existing
science of geometry. That is to say, he employs geometry,
viewed as valid, to prove the correctness of his view of
space. 1 Starting from that science as given, he enquires what
are the conditions which can alone render it possible. These
conditions are found to coincide with those independently
established. Now this is a valid argument when employed
in due subordination to the main synthetic method. It
offers welcome confirmation of the results of that method. It
amounts in fact to this, that having proved (by application
of the transcendental method) the mathematical sciences to
be valid, everything which their validity necessarily implies
must be granted. Kant's reasoning here becomes circular,
but it is none the less valid on that account. This further
complication of the argument is, however, dangerously apt
to mislead the reader. It is in great part the cause of the
above division among Kant's commentators. The method
employed in the Prolegomena is simply this form of argument
systematised and cut free from all dependence upon the
transcendental method of proof. 2

The whole matter is, however, still further complicated by
the distinction, which we have already noted, between real and
ideal possibility. Are the given synthetic a priori judgments
valid? That is one question. Can the Critical philosophy
discover, completely enumerate, and prove in a manner never
before done, all the possible synthetic a priori principles ?
That is a very different problem, and when raised brings us to
the further discussion of Kant's transcendental method. The
question at issue is no longer merely whether or not certain
given judgments are valid, and how, if valid, they are to
be accounted for. The question is now that of discovering
and of proving principles which have not been established
by any of the special sciences. This shifting of the problem
is concealed from Kant himself by his omission to distinguish
between the undemonstrated axioms of the mathematical
sciences and their derivative theorems, between the principles
employed by the physicist without enquiry into their validity
and the special laws based upon empirical evidence.

1 Cf. A 46-9 = 6 64-6. The corresponding sections of the Prolegomena^
Vaihinger contends, were developed from this first edition passage, and the
transcendental exposition of space in the second edition from the argument of
the Prolegomena.

2 The synthetic method of argument is, as we shall see later, further extended
in the Analytic by being connected with the problem of the validity of ordinary
experience. But as the mathematical sciences are proved to have the same con-
ditions as neither more nor less than the consciousness of time, this also allows
of a corresponding extension of the analytic method. The mathematical sciences
can be substituted for the de facto premiss by which these conditions are proved.



As regards the mathematical axioms, the problem is fairly
simple. As we shall see later, in the Aesthetic^ they do not
require a deduction in the strict transcendental sense. They
really fall outside the application of the transcendental
method. They require only an " exposition." But in regard
to the fundamental principles of natural science we are pre-
sented with the problem of discovery as well as of proof.
Unlike the axioms of the mathematician, they are frequently
left unformulated. And many postulates, such as that there
is a lex continui in natura, are current in general thought, and
claim equal validity with the causal principle. Kant has thus
to face the question whether in addition to those principles
employed more or less explicitly by the scientist, others, such
as might go to form an immanent metaphysics of nature, may
not also be possible.

B. (a) * Psychological and logical possibility. Both have to
be recognised and accounted for. Let us consider each in

(i) Psychological possibility. What are the subjective
conditions of a priori synthetic judgments ? Through what
mental faculties are they rendered possible ? Kant replies by
developing what may be called a transcendental psychology.
They depend upon space and time as forms of sensibility,
upon the a priori concepts of understanding, and upon the
synthetic activities by which the imagination schematises
these concepts and reduces the given manifold to the unity
of apperception. This transcendental psychology is the
necessary complement of the more purely epistemological
analysis. 2 But on this point Kant's utterances are extremely
misleading. His Critical enquiry has, he declares, nothing in
common with psychology. In the Preface to the first edition
we find the following passage : " This enquiry . . . [into] the
pure understanding itself, its possibility and the cognitive
faculties upon which it rests . . ., although of great importance
for my chief purpose, does not form an essential part of it." 3
The question, he adds, "how is the faculty of thought itself
possible? ... is as it were a search for the cause of a given
effect, and therefore is of the nature of an hypothesis [or
' mere opinion '], though, as I shall show elsewhere, this is not
really so." The concluding words of this passage very fairly
express Kant's hesitating and inconsistent procedure. Though
he has so explicitly eliminated from the central enquiry of

1 Cf. above, p. 43.

2 What follows should be read along with p. 235 ff. below, in which this dis-
tinction between the "subjective" and "objective" deductions is discussed in
greater detail.

3 A x-xi.


the Critique all psychological determination of the mental
powers, statements as to their constitution are none the less
implied, and are involved in his epistemoiogical justification
alike of a priori knowledge and of ordinary experience. If we
bear in mind that Kant is here attempting to outline the
possible causes of given effects, and that his conclusions
are therefore necessarily of a more hypothetical character
than those obtained by logical analysis, we shall be pre-
pared to allow him considerable liberty in their formulation.
But in certain respects his statements are precise and definite
the view, for instance, of sensations as non-spatial, of time
as a form of inner sense, of the productive imagination as
pre-conditioning our consciousness, of spontaneity as radically
distinct from receptivity, of the pure forms of thought as
not acquired through sense, etc. No interpretation which
ignores or under-estimates this psychological or subjective
aspect of his teaching can be admitted as adequate. 1

(2) Logical or epistemoiogical possibility. How can synthetic
a priori judgments be valid? This question itself involves a
twofold problem. How, despite their synthetic character, can
they possess truth, i.e. how can we pass from their subject
terms to their predicates? And secondly, how, in view of
their origin in our human reason, can they be objectively
valid, i.e. legislate for the independently real ? How can we
pass beyond the subject-predicate relation to real things? This
latter is the Critical problem in the form in which it appears
in Kant's letter of 1772 to Herz. 2 The former is the problem
of synthesis which was later discovered.

(b) (i) Possibility of explanation and (2) possibility of exist-
ence. (i) How can synthetic a priori judgments be accounted
fort How, despite their seemingly inconsistent and apparently
paradoxical aspects, can their validity (their validity as well as
their actuality being taken for granted) be rendered compre-
hensible ? (2) The validity of such judgments has been called
in question by the empiricists, and is likewise inexplicable
even from the dogmatic standpoint of the rationalists. How,
then, can these judgments be possible at altt These two
meanings of the term ' possible ' connect with the ambiguity,
above noted, in the term ' how.' The former problem can be
solved by an analytic method ; the latter demands the applica-
tion of the more radical method of synthetic reconstruction.

(<:) Eeal and ideal possibility. 3 We have to distinguish
between the possible validity of those propositions which the
mathematical and physical sciences profess to have established

1 This is a criticism to which Cohen, Caird, and Riehl lay themselves open.
2 Cf. below, pp. 219-20. 3 Cf. above, pp. 49-50.


and the possible validity of those principles such as that of
causality, which are postulated by the sciences, but which the
sciences do not attempt to prove, and which in certain cases
they do not even formulate. The former constitute an
actually existent body of scientific knowledge, demonstrated
in accordance with the demands of scientific method. The
latter are employed by the scientist, but are not investigated
by him. The science into which they can be fitted has still to
be created ; and though some of the principles composing it
may be known, others remain to be discovered. All of them
demand such proof and demonstration as they have never
yet received. 1 This new and ideal science is the scientific
metaphysics which Kant professes to inaugurate by means of
the Critique. In reference to the special sciences, possibility
means the conditions of the actually given. In reference to
the new and ideal metaphysics, possibility signifies the con-
ditions of the realisation of that which is sought. In view of
this distinction, the formula How are synthetic a priori judg-
ments possible? will thus acquire two very different meanings,
(i) How are the existing a priori synthetic judgments to be
accounted for? (2) How may all the really fundamental judg-
ments of that type be exhaustively discovered and proved ?
Even in regard to immanent metaphysics Kant interprets the
formula in both ways. This is due to his frequent confusion of
immanent metaphysics with the principles of natural science.
Its propositions are then regarded as given, and only their
general validity calls for proof. It is, however, in the problem
of ideal possibility that the essential problem of the Critique
lies ; and that is a further reason why it cannot be adequately
dealt with, save by means of the synthetic method.

Experience. Throughout the Introduction the term ex-
perience 2 has (even at times in one and the same sentence)
two quite distinct meanings, (i) as product of sense and under-
standing acting co-operatively, and (2) as the raw material
(the impressions) of sense. Considerable confusion is thereby

Understanding and reason 3 are here, as often elsewhere in
the Critique^ used as equivalent terms. Throughout the entire
two first sections of the Introduction to the second edition the
term reason does not occur even once. As first mentioned, 4
it is taken as the source of metaphysical judgments.

1 Cf. Vaihinger, i. p. 405. The existing sciences can, as Vaihinger says, be
treated en bloc, whereas each of the principles of the new philosophy must be
separately established.

2 A i. 3 A 1-2. 4 B6 = A


General (a priori) truths have an inner necessity and must be
clear and certain by themselves. 1 These statements are not in
accordance with Kant's new Critical teaching. 2 They have
remained uncorrected from a previous way of thinking. This
must be one reason for the recasting of this paragraph in the
second edition.

Even with (unter) our experiences there is mingled knowledge
which must be of a priori origin. 3 Kant is here distinguishing
the immanent a priori, such as that involved in any causal
judgment, from the transcendent a priori dwelt upon in the
next paragraph. The latter is expressed through metaphysical
judgments, such as * God exists,' ' the soul is immortal.'

Original concepts and judgments derived from them. 4 Cf. B

Pure. In the title of the section the term pure 5 (rein) is, as
the subsequent argument shows, taken as exactly equivalent
to a priori. As Vaihinger notes, the adjective apriorisch
had not yet been invented. The opposite of pure is here
empirical (empirisch)?

All our knowledge begins with experience. 7 This is a stronger
statement than any in the corresponding paragraphs of the first
edition. Had Kant proceeded to develop its consequences,
he would have had to recast the entire Introduction, setting
the problem of empirical knowledge alongside that of the a
priori? As it is, he is forced 9 to subdivide the absolutely a
priori into the pure and the mixed. 10

By objects which aflfect (riihren) our senses. The raw material
of sensuous impressions. 11 These incidental statements call for
discussion. Cf. below, pp. 80-8, 120-1, 274 ff.

A knowledge of objects which we call experience. 12 Kant
does not keep to this definition. The term experience is still
used in its other and narrower sense, as in the very next
paragraph, when Kant states that knowledge does not,
perhaps, arise solely from experience ( = sense impressions).

In respect of time. 13 This statement, taken as an account of
Kant's teaching in the Critique, is subject to two reservations.
In the Aesthetic^ Kant sometimes claims a temporal ante-
cedence for the a priori. And secondly, the a priori is not for
Kant merely logical. It also possesses a dynamical priority. 15

Even experience itself is a compound. 16 The " even " seems
to refer to the distinction drawn in A 2 between the immanent
and the transcendent a priori? 1

1 A 2. 2 Cf. above, pp. xxxv, 36 ff. ; below, pp. 565-7. 3 A 2.

4 A 2. 5 B i. 6 Cf. below, p. 55. 7 B I.

8 Cf. below, p. 54. 9 B 2-3. 10 Cf. below, p. 55. u B i.

12 B i. " B i. 14 Cf. below, p. 88 ff. 15 Cf. below, p. 237 ff.

16 B i. 17 Cf. below, pp. 55-6.


It is therefore a question whether there exists such knowledge
independent of experience. 1 This question was not raised in the
first edition. 2 The alternative methods, analytic and synthetic,
are discussed above, p. 44 fif.

Such knowledge is called a priori and is distinguished from
empirical knowledge. 3 Throughout the Introduction, in both
editions equally, Kant fails to state the problems of the
Critique in a sufficiently comprehensive manner. He speaks
as if the Critique dealt only with the absolutely a priori,
in its two forms, as immanent scientific knowledge and as
transcendent speculation. It also deals with the equally
important and still more fundamental problem of accounting for
the possibility of experienced Our empirical knowledge involves
an a priori element, and may not therefore be opposed to a
priori knowledge in the manner of the passage before us.

This term a priori is not yet definite enough. 5 It is frequently
employed in a merely relative sense. Thus we can say of a
person who undermines the foundations of his house that he
might have known a priori that it would collapse, that is,
that he need not wait for the experience of its actual fall. But
still he could not know this entirely a priori ; he had first to
learn from experience that bodies are heavy, and will fall when
their supports are taken away. But as dealt with in the
Critique the term a priori is used in an absolute sense, to
signify that knowledge which is independent, not of this or
that experience only, but of all impressions of the senses.
Thus far Kant's position is comparatively clear; but he proceeds
to distinguish two forms within the absolutely a priori, namely,
mixed and pure. The absolutely a priori is mixed when it con-
tains an empirical element, pure when it does not. (" Pure " is
no longer taken in the meaning which it has in the title of the
section. 6 It signifies not the a priori as such, but only one
subdivision of it.) Thus after defining absolutely a priori
knowledge as independent of all experience, Kant takes it
in one of its forms as involving empirical elements. The
example which he gives of an absolutely a priori judgment,
which yet is not pure, is the principle : every change has its
cause. " Change " is an empirical concept, but the synthetic
relation asserted is absolutely a priori. In the next section 7
this same proposition is cited as a pure judgment a priori
"pure" being again used in its more general meaning as synony-
mous with a priori. This confusion results from Kant's
exclusive preoccupation with the a priori, and consequent

1 B 2. 2 Cf. above, p. 27 . 3 B 2.

4 Cf. above, pp. 39 ff., 53 ; below, pp. 57-8, 222 ff., 241, 286-9.

5 B 2-3. 6 Cf. above, p. 53. 7 A 9-10 = 6 13.


failure to give due recognition to the correlative problem
of the empirical judgment. The omitted factor retaliates
by thus forcing its way into Kant's otherwise clean-cut
divisions. Also, it is not true that the relative a priori falls
outside the sphere of the Critical enquiry. Such judgment
expresses necessity or objectivity, and for that reason demands
a transcendental justification no less urgently than the abso-
lutely a priori. The finding of such justification is, indeed,
the central problem of the Analytic!

The subdivisions of the a priori may be tabulated thus :

Relative, e.g. every unsupported house must

A priori knowledge

( Mixed, e.g. every change has its cause.

Absolute^ Pure, e.g. a straight line is the shortest
[ distance between two points.

The term pure (rein) thus acquires a second meaning dis-
tinct from that defined above. 2 It is no longer employed as
identical with a priori, but as a subdivision of it, meaning
unmixed. Its opposite is no longer the empirical, but the im-
pure or mixed. Owing, however, to the fact that "pure" (in its
first meaning) is identical with the a priori, it shares in all the
different connotations of the latter, and accordingly is also
employed to denote that which is not relative. But "pure" has
yet another meaning peculiar to itself. The phrase " inde-
pendent of experience" has in reference to "pure" an
ambiguity from which it does not suffer in its connection with
"a priori" (since mathematical knowledge, whether pure or
applied, is always regarded by Kant as a priori}. It may
signify either independence as regards content and validity, or
independence as regards scope. The latter meaning is
narrower than the former. By the former meaning it denotes
that which originates, and can possess truth, independently of
experience. By the latter it signifies that which is not only
independent of sense but also applies to the non-sensuous.
In this latter meaning pure knowledge therefore signifies
transcendent knowledge. Its opposite is the immanent. The
various meanings of " pure" (four in number) may be tabulated
as follows :

(a) (i) A priori: independent of experience as regards origin
and validity. (Its opposite = empirical.)

(2) Absolutely independent of experience. (Its

opposite = relative.)

(3) Unmixed with experience. (Its opposite =
impure or mixed.)

1 Cf. above, p. 39 ff., and below, pp. 286-9. 2 ? 53 5 cf. also pp. 1-2.


(b) (4) Independent of experience as regards scope = tran-
scendent. (Its opposite = immanent.)

All these varied meanings contribute to the ambiguity of
the title of the Critique. Kant himself employs the title in all
of the following senses :

1. Critique of absolutely pure a priori knowledge, deter-
mination of its sources, conditions, scope and limits.

2. Critique of all a priori knowledge, relative as well as
absolute, in so far as it depends upon a priori principles,
determination, etc.

3. Critique of all knowledge, whether a priori or empirical,
determination, etc.

4. Critique of transcendent knowledge, its sources and

Further meanings could also be enumerated but can be
formulated by the reader for himself in the light of the ambi-
guities just noted. 1 The special context in each case can alone
decide how the title is to be understood. If a really adequate
definition of the purpose and scope of the Critique is sought
by the reader, he must construct it for himself. The following
may perhaps serve. The Critique is an enquiry into the sources,
conditions, scope and limits of our knowledge, both a priori and
empirical, resulting in the construction of a new system of
immanent metaphysics; in the light of the conclusions thus
reached, it also yields an analysis and explanation of the tran-
scendental illusion to which transcendent metaphysics, both as
a natural disposition and as a professed science, is due.

Kant further complicates matters by offering a second
division of the absolutely a priori? viz. into the original and
the derivative. Also, by implication, he classes relative a priori

Online LibraryNorman Kemp SmithA commentary to Kant's 'Critique of pure reason,' → online text (page 11 of 72)