‘WEQ- » L‘

fifi‘iflhi‘fih;\‘a: _.

 

: f
i i’
‘- k.
g ‘ ,.
" a 9' 42-3." 9 A». -.:.~ .. . f";- 4.’ r. f E
ngfigy g§253f3£ 23 gigglei
U ._ " a!” I
mveisuy i

 

This is to certify that the

dissertation entitled
DUB-T17 hm TV va , 071.7er IV IWY , AND INmWPb‘CBJPCT IV I'T‘Y
IN KANT ' S
GRIT IQUT? CF PURE REASON

 

presented by

Jorg Baumgartner

has been accepted towards fulfillment
of the requirements for

-..’ , Til—l"
PhD mgmem .h. osophy

 

 

W #543};

Major professor

Date P“ (~50 7/ 'I 735/

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

MSU

 

 

LIBRARIES
m

 

‘J'nlOcI—l.

RETURNING MATERIALS:
Place in book drop to
remove this checkout from
your record. FINES will

 

be charged if book is
returned after the date
stamped below.

 

1 MA? fin «mg

no?
SEP ‘0 530m
6"

,9
r;

 

 

 

SUBJECTIVITY, OBJECTIVITY, AND INTERSUBJECTIVITY
IN KANT'S

QBIIIQUL‘L DE_ EUBE-B.EASDN

by

Jorg Baumgartner

A Dissertation

submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Philosophy
1

 

 

©1988

JORG BAUMGARTNER

All Rights Reserved

 

 

Abstract

SUBJECTIVITY, OBJECTIVITY, AND INTERSUBJECTIVITY
IN KANT'S
QBITIQLHL .OE- Ell BELBEASQN

by

Jorg Baumgartner

Chapter I contains an examination of the criticisms which some
philosophers (Husserl, Sartre, and Merleau-Ponty) have advanced
against Kant concerning the problem of our knowledge of other
thinking beings. In the course of this examination the nature
and scope of Kant's inquiry is brought into focus: it is a
transcendental inquiry which deals with the a_pri_ri conditions
of the possibility of experience. This means two things: (1) The
question whether there are other thinking beings besides myself
is for Kant not a philosophical (transcendental). but an
empirical question which can be answered only a_pgsterigri, that
is, by actual experience. (2) The a_pri_ri conditions must
nevertheless be such that they do not preclude but leave room for
the experience by which we find out whether there are other
thinking beings besides ourselves. Chapter II develops the
central thought of the Transcendental Deduction in order to deal
with two complementary issues: with Kant's confusions with
respect to the 'subjective', and with the notion of an 'object of

It is argued that objectivity in the

representationsh

 

Critical—transcendental sense ranges equally over the 'outer' and
'inner'. This means that inner experience is objective, i.eH
can be expressed in judgments which possess objective validity
and necessary universality for everybody. Chapter III deals with
the Second Analogy. The main purpose here is to show what types
of objects are constituted by means of the categories: the
determination of a manifold of sensible intuition by means of the
category of cause and effect is its determination as a happening
or as an event. In chapter IV, finally. it is shown that and how
a particular type of empirical objects, namely thinking beings

can be seen as falling under the objects which are constituted by

the categories.

 

DEDICATION

to

Inge and Nicholas

"Hem sonst. als Euch?"

 

ACKNOWLEDGEMENTS

Grateful acknowledgement is made to those who have
given special guidance and encouragement in the writing of
this dissertation: to Rhoda H. Kotzin, whose courses in
philosophy and whose substantive suggestions inspired this
work in every way; to teachers and friends in the Department
of Philosophy at Michigan State University for their help;
to Judy Johnson of the Albion College Library for her aid in
obtaining books; and to Diana Hopkins and John Williams of
the Albion College Academic Computing Center for their

generous and ubiquitous assistance in the preparation of the

computer script.

 

 

TABLE OF CONTENTS

Chapter Page

I. INTRODUCTION........................... 1
Notes to Chapter 1.... ....... . ........ 27

II. SUBJECTIVITY [o]....................... 33
Notes to Chapter II................... 71

III. OBJECTS [ct]............... ..... ....... 75
Notes to Chapter III...... ............ 128

IV. ONE WORLD..............................130

Notes to Chapter IV ..... . ............. 177

V. BIBLIOGRAPHY............ ..... ..........180

 

CHAPTER I
INTRODUCTION
"The objects of transcendental philosophy
are not objects of perception."
(WW XXI, P. 92)
In Kant's Anthgppglggy there is a section "On
Egoism" in which he describes the behavior characteristic
of, for example, the logical and the ethical egoist.
Towards the end he remarks that egoism has as its opposite
pluralism, "that is, the attitude of considering oneself not
as if 00mprising the whole world in one's self, but only as
a citizen of the world." This, Kant says, is how the
science of anthropology states the difference between egoism
and pluralism, and he concludes the section with the
observation that they also form a topic in metaphysics-~the
topic of metaphysical egoism, or solipsism, which discusses
the question "whether I, as a thinking being, have reason to
assume, besides my own existence, also the existence of a
totality of other beings that stand in community with me, i.
e., the existence of what is called 'a world'."1
The reader who with this question in mind turns to
the Critique of Pure Reason will soon be convinced that
nothing is further from Kant's mind than maintaining
metaphysical solipsism, and that the latter is rather one of

those philosophical "scandals" he mentions in the

Introduction (B XL a; NKS 314).2 For although the Cgitigue

 

is, in Bennett's words, "an intensely first—person singular

3

work," a considerable body of text is marked by what
Strawson calls its "collective style."u Indeed, the
locutions in the first-person plural are numerous: Kant
speaks, often emphatically, of "us," “our representations,"
"our sensibility," "our understanding," or of "our
knowledge." And there is no question about the referent of
these pronouns: it is the class of human beings.

Thus Kant says, for example, that "we cannot judge
in regard to the intuition of other thinking beings, whether
they are bound by the same conditions as those which limit
our intuition and which for us are universally valid" (A 27
= B 43; NKS 72). These conditions are the conditions of our
sensibility, and our sensibility is "peculiar to us, and not
necessarily shared in by every being, though, certainly, by
every human being" (A b2 : B 60; NKS 82).

According to the question in the Anthrgpglggy,
however, the plurality of human beings cannot be simply
presupposed as something obvious but demands a
justification. Yet this does not mean egmipsg that the
Qgitiggg has to give it. It has to do so only if the
question "may reasonably be asked" and is not "absurd in
itself and [calls] for an answer where none is required" (A
58 = B 82; NKS 97). But we can expect the question to be a
reasonable one if the Qgitigue shares some basic features

with that framework in which the question arises and from

which it derives its sense. This framework is plainly

 

Cartesian: if not in wording, in essence the question is
the one raised in the Third Meditation and in part answered.
And there it is raised because it is held that a person is a
thinking or immaterial substance which, as such, can know
immediately or directly, although apodeictically so, only
what is 'ig' it in the sense of being a modification of it.
It can have no such indubitable knowledge of what, as matter
of principle, cannot be a modification of it and in this
sense is something '9g3§§'--namely, all other substances, be
they thinking or material.5

There is no question that Kant revises both the
ontology and the epistemology of the Cartesian framework,
and later I will examine both revisions in some detail. At
the moment it is sufficient to say that if anything survives
the ontological revision, it is the contrast of the internal
and external which now recurs in the domain, no longer of
substances, but of appearances. These divide into those of
inner and outer sense. Yet the latter are 'inner' too:
they are representations, and all representations belong,
"whether they have as their objects outer things or
not,...in themselves, as determinations of the mind, to our
inner state" (A 3A : B 50; NKS 77). And this view is open,
it seems, to the traditional objection that Kant formulates
as follows: "I am immediately conscious only of that which

,rntatigns of outer things;

is in me, that is, of my gepr§_

and consequently...it must still remain uncertain whether

outside me there is anything corresponding to it, or not" (B

 

XLa; NKS 3“).

For an answer to this objection,

at Kant's epistemolological revision.

in

as a necessary condition something other than itself.6

Thus in the Analytic of Concepts and

Principles there are arguments for the claim that for a

manifold of representations given in
that each of them belongs to one and
it is necessary that they be brought

be representations of an object held

the tenet that the Cartesian cogito implies,

let us look briefly
The latter consists
has

i. e.,

in the Analytic of

intuition to be such
the same consciousness,
under concepts and so

to be, in some sense of

7

the term, "distinct" from these representations themselves.

And in the Refutation of Idealism Kant argues that the

determination of my existence in time, that is, inner

experience, is possible only on the assumption of outer

experience: it requires, in other words, that some of my

representations be representations of "actual

thingS...outside me" (B 276; NKS 2N5).

It appears that both these claims, if they can be

substantiated, provide Kant with an answer to the question

raised in the Anthggpgiggy. Thus he can say: I have reason

to assume besides my own existence the existence of

 

something else, because "the mere, but empirically
determined, consciousness of my own existence proves the
existence of objects in space outside me" (E 275; NKS 2A5),

or because the fact that I can be conscious of myself as the

same subjeCt of a diversity of representations shOws that

 

3"
. I '

they are representations of objects. But Kant does not say,
and hence answers the original question perhaps only
partially, "whether I, as a thinking being, have reason to
assume" that among the objects of my representations there
are objects of a special type-—thinking beings other than
myself.

Kant's silence has not gone completely unnoticed in
the literature, and in the following sections I will examine
what some philosophers have to say about it. This will give
me also the opportunity to bring to the fore a number of
problems and themes that will occupy the center of my

investigation.

(2)

In Husserl's interpretation, Kant's Copernican turn,
his turn to 'transcendental subjectivity', consists in the
insight that it is necessary to break with the traditional
epistemological procedure that reconstructs experience by
presupposing the following as given: that there exists, as a
matter of course, a world, and that the cognitive subject,
for example myself, occurs IE it as a psychophysical thing
among other things.8 This break is required, Husserl
maintains, because what is given is neither the world nor
myself simpliciter, but only the consciousness of 'the
world' and of 'myself'.9 In other words: that there exists
a world and that I find myself in it is an experience itself

and hence cannot be presupposed in order to explain that

 

 

 

experience.

This experience belongs, according to Husserl, to a
subject that on principle cannot occur in experience, but in
terms of whose 'subjectivity', i. e., cognitive functions,
experience, and there being a world in experience, is to be
rendered intelligible. Hence when Kant expresses the
subjective nature of space and of time by saying that they
are "in us" and would vanish "if the subject...be removed"
(A #2 = B 59; NKS 82), he points to what in Husserl's view
is "an irrefutable state of affairs."10 But this involves
for Kant the problem of exploring how it is possible for a
"cognitive...subjectivity" to arrive not only at a belief
in, but "at a justified conviction about a real world which

it can know."11

And Kant, as Husserl points out, sees that
in order to be khgmn, a world has not only to be real, but
also objective, and that it is objective if and only if it
can be known by any subject. In other words: it is not
sufficient that everybody be in a position to say: "I know a
world in my experience as my_wprlg,...and others know a

world as t eir_wgrlg." What Kant, rather, sets himself

as a task in the gritigyg is to discover the transcendental

 

conditions of the possibility of a "single, individually and

intersubjectively determinable world."13
According to Husserl, Kant's reconstruction of a

"true objectivity within the transcendental subjectivity,"1“

however admirable, has serious shortcomings nevertheless.

Husserl bases his criticism on the claim that "in a

 

transcendental philosophy everything without exception must

'15 Kant's failure to adhere to this

be transcendental.’
norm is why his reconstruction is both incomplete and flawed
by methodological mistakes. He commits such mistakes, for
example, when he designs epistemological theories on the
basis of empirical, psychological doctrines of association
or of reproduction, whose legitimacy cannot be taken for
granted in a transcendental inquiry but must first be
established. And one respect in which the reconstruction is
said to be incomplete concerns Kant's repeated and
unjustified reference to "us human beings." Husserl argues
that Kant fails to consider that in an inquiry like his own
a plurality of subjects is given only in the form of human
beings, and that therefore the "translation of their bodies,
as the translation of all outer nature, into transcendental
appearances"16 yields at first his own ego only, and that
the transcendental possibility of positing other egos has
still to be explored.

Husserl's criticism seems to come to this: The
plurality of subjects to which Kant refers when mentioning

'us' or 'us human beings' is merely presupposed; it is not

 

constituted in that "single, individually and

intersubjectively determinable world" of which the Qgiiiggg
tries to establish the condition of the possibility. I will
come back to this claim later, and at this point remark only
that it is misleading. For we are left with the impression

that Kant is completely unaware of the problem concerning

 

the 'translation' of human beings into appearances. That
this is not so becomes clear once we realize that in a
transcendental philosophy the plurality in question is not a
plurality of other human beings only, but a plurality that
also and equally includes myself. That the constitution of
myself as appearance is at least attempted by Kant is shown
by his all-important distinction between the transcendental
unity of apperception and the empirical self, i. e., the

self that is experienced.17

And it is my empirical self,
according to Kant, that occurs, or is supposed to occur, in

that intersubjective world we are talking about.

(3)

Whereas in Husserl's view Kant does not see and
hence does not investigate the problem of the transcendental
possibility of a plurality of subjects, we may, Sartre
suggests, at first judge the situation differently and take
Kant to be of the opinion that there is no such problem to
be dealt with: the problem of Others does not fall in the
domain of his inquiry, which is a_pri9ri and deals with
"the universal laws of subjectivity," i. e., with laws
"which are the same for all,"18 and with the conditions of
the possibility "only for an object in general."19

Such a view, however, would be misleading. For in
his transcendental inquiry Kant is engaged, according to
Sartre, to determine not only the conditions of an object in

general, but also of various types of objects-—of the

 

 

 

 

physical object for example, or of the beautiful. Hence "if

it is true," Sartre argues, "that the Other represents a
particular type of object which is discovered to our
experience, then it is necessary even within the perspective
of a rigorous Kantianism to ask how the knowledge of others
is possible."20 And the other is a particular type of
object: he is, to be sure, an object in my experience, but
an object which, unlike all other objects, is also a subject
of experience.

Sartre wants to argue for the conclusion that within
the Kantian framework such an investigation comes to
nothing: the concept of the Other, he maintains, is
"neither...a constitutive concept nor...a regulative concept

of my knowledge."21

For the issues to be raised, it is
sufficient to examine Sartre's argument for the first claim
only, which he also puts this way: The concept of the che;
can not constitute our experience."22
First let me point out here that in View of his
actual argument Sartre‘s choice of the term 'constitutive'
is misleading. For we cannot take him to use it in Kant's
sense according to which the forms of intuition or the
categories can be said to be constitutive in so far as they
are a_pgi9£i conditions of objects of experience. Nor can
we take Sartre to employ the term in the sense in which he
himself considers not the gghgept of the Other but the
"23 to be constitutive, i. e., to be "the

"Other ' 5 look

necessary condition of my objectivity."‘u In experiencing

 

myself as looked at, I experience myself as an object, which
in turn is a requirement if I am to be able, upon
reflection, "to conceive of...my properties in the objective

mode"25

--to conceive of them, that is, as they present
themselves to the point of view of the Other. What Sartre's
argument amounts to, is simply the claim that if we
presuppose Kant's account of experience, and if within this
account we mean by 'the Other' not only an object but also a
subject of experience, then the concept of the Other cannot
be exemplified in our experience.

Now in a number of passages in fleihg_ahgfiflgthinghgss
Sartre critically discusses Kant's doctrine of freedom, in
particular the notion of a "non-temporal spontaneity"26
and the notion of a choice on the part of the "intelligible

character."27 It is in the light of these passages that we

 

have to see the preliminary step of Sartre's argument,
which, however unhappily formulated, is a correct and
important one: the problem of the Other does not concern

. . . . 28
his "noumenal" or "intelligible eXistence," because such

 

an existence can, not only in the Other's, but in my own
case as well, "only be thought, not conceived."29 In
correct Critical language Sartre should have said 'not
known'. According to Kant we can think things in
themselves, but we cannot 339w them. We have "no knowledge
of any object as thing in itself, but only in so far as it

is an object of sensible intuition, that is, an appearance"

(B XXVI; NKS 27). The problem, then, of our knowledge of

 

 

11

the Other can concern the Other only as an appearange.

In my View Sartre's argument depends for its crucial
steps on his interpretation of Kant's distinction between
outer and inner experience. And it is in particular Kant's
conception of inner experience that, Sartre says, is the
reason why the knowledge of the Other becomes a problem that
cannot be solved.

First, according to Sartre, inner experience
consists in the knowledge of the appearances of my inner
sense: "of my emotions or of my empirical thoughts."3O To
raise the question of the Other, then, is to ask whether
within the Kantian framework I am justified in asribing to
the other what I can know in my own case by means of inner
experience. In Sartre‘s words: "What I aim at in the Other
is nothing more than what I find in myself."31

Second, in outer experience I obtain knowledge of
the Other "by the presence of organized forms such as
gestures and expression, acts and conducts."32 But
although I ordinarily take these to refer, for example, to
the anger the Other feels towards me, in the Kantian
framework I am prevented from doing so. For in it we are
led to a gap between the Other's anger-behavior and his
anger itself: while the former is public, the latter is
private. For to talk in the Critical sense about the Other's
anger is to talk about an appearance or series of
appearances of "his inner sense...[that] is by nature

refused to my apperception:"33 they are "on principle

 

 

 

 

12

outside my experience and belong to a system which is
311

inaccessible to me"
"35
We might attempt now, Sartre suggests, to bridge the

-—to a system that, in short, "is not

mine.

gap between the Other's anger and his anger-behavior by
connecting them as cause and effect: to consider "the

36 But

redness of Paul's face as the effect of his anger."
the anger the Other feels and the anger—behavior I observe

belong to two giffereht experiences, and Kantian causality

 

"only...links the phenomena of

"37

same
experience. Causality for Kant, Sartre maintains, "is a
unification of the moments of my time in the form if
irreversibility,"38 and there is no way for this causality
to "unify my time with that of the Other."39

If by the Other, therefore, we mean not only an
object of experience, but an object of a particular kind-~i.
e., an object that is also a subject with experiences of its
Own——then the Other cannot be known in a Kantian framework.
His experiences are out of my reach: they are "located

outside of all experience which is possible to me."uO

 

Hence, "the subjective quality of the Other—as-object" turns
out to be in Kant's philosophy a "closed box."1H

In assessing this argument I am struck by the
discrepancy between what Sartre promises to do and what he
actually does. The promise is "to establish the conditions

. . u
of the possibility of the experience of Others." 2 But to

claim this, and to claim it "within the perspectives of a

 

 

13

M3 is to claim that one is engaged in

rigorous Kantianism,"
a transcendental inquiry. And such an inquiry is, in Kant's
words, "occupied not so much with objects as with the mode
of our knowledge of objects in so far as this mode of
knowledge is to be possible a_prigri (A 11 = B 25; NKS 59)-
The scope of a transcendental inquiry is thus defined not by
any conditions, but by the §_p§ig~i conditions of experience
and objects of experience. It is within this scope,
therefore, that the problem of our knowledge of the Other
has to be dealt with.

We would, then, expect Sartre to tell us something
about _a_pgiggi conditions; about space, time, and the
categories, and their function with respect to the
constitution, for example, of "what I find in myself,"414 of
"the Other's anger in so far as it appears to his inner

MS . . . . .
" i. e., of particular instances of inner

sense,
experience. The latter according to Kant, cannot be simply

taken for granted, but "has to be reckoned with the

 

investigation of the possibility of any and every
experience, which is certainly a transcendental enquiry" (A
3A3 = B MO1; NKS 330).

But Sartre, instead of placing his argument on the

.ig§i level, places it plainly on the

transcendental-g_p
empirical—a_pgstrri9ri level: for to talk about the Other's
emotions and about his behavior means, for Kant, to talk
empirically about empirical objects. And once we realize

this, we realize also that the transcendental terminology

 

m.

1"

throughout his argument-~"apperception," "synthesis,"
"synthetic unity“—-is only ornamental. Without it, there is
nothing specifically Kantian about the position under
attack, and no reason why the argument could not be mounted
against, for example, Locke or Hume. Against Kant, however
it idles. And it does so, in my view, because Sartre
completely ignores or fails to understand the "Critical
problem."

In what follows, I will outline this problem and its
solution so far as is necessary for my present purposes. In
the Transcendental Deduction Kant introduces the problem
with the question: what must we mean "by the expression 'an
object of representations'" (A 10A; NKS 13h)? It is the
problem, in other words, of how it is possible for our
representations, which qga representations are "inner
determinations of our mind,“ to acquire, in addition to
their "subjective reality" and "subjective meaning," what
Kant calls "relation to an object," "objective reality," and
"objective meaning" (A 197 = B 2&2; NKS 22M). It is now of
crucial importance to see that the expression 'relation to
an object' does not mean 'relation to an external object':
it concerns, rather. the objectiyity of our representations
whether they are representations of inner or of outer
intuition. And the solution of the problem of their
objectivity is in both cases the same in principle: the

representations must be united in the pure concepts of the

understanding or categories. But the 'objects' thus

 

 

 

 

15

constituted are not particular empirical objects--such as a
chair, for example, or Paul's anger-~but objectivities, i.
e., precisely those 'objects in general' that Sartre
dismisses right at the outset of his discussion of the
Critical philosophy. The problem, then, to which Sartre
should have addressed himself and which he completely
ignores, is a transcendental one and concerns the necessity
of the categories and of the objects constituted by the
categories, e. g. "a happening in general" (A 788 : B 816;
NKS 62M) or something permanent in general, for our
knowledge of particular empirical objects which include 'my

anger‘ and 'the Other's anger' as well.

(4)

In Husserl's opinion, as we have seen, Kant does not
realize that even if he should have succeeded in
establishing the transcendental conditions of a "single,
individually and intersubjectively determinable world,"“6
he ist still faced with the problem "of the transcendental

"“7 Sartre, on the

possibility of positing other egos.
other hand, goes further: he believes to have shown that
Kant cannot solve this problem. We turn now to
Merleau—Ponty, who holds that the problem does not arise for
Kant, but this only because Kant has an inadequate
conception of philosophical reflection.

According to Merleau-Ponty, the very starting-point

of the Critical philosophy is already "the spectacle of the

 

 

 

16

world, which is that of a nature open to a plurality of

thinking subjects."u8 The task, then, which the Critical

philosophy sets for itself is merely to discover "the

conditions which make possible this unique world presented

49

to a number of empirical selves." These conditions are

found in a "transcendental ego"50 or "universal

constituting consciouness."51 And since the Kantian
analysis "is located from the start outside me," that is,
“has nothing to do but to determine the general conditions
which make possible a world for an ego--myself or others
equally-—," the transcendental ego is "just as much other
people's as mine."52 This, Merleau-Ponty maintains, is the
reason why "the problem of the knowledge of other people is

"53

never posed in Kantian philosophy. But if, on the other

hand, philosophical reflection were "not carried outside
itself,"5u if, in other words, Kant had realized that "the
thinking ego can never abolish its inherence in an

'55 he would have been able to achieve a

individual subject,’
more radical self-discovery and other people would have
become a problem for him.

In my discussion of herleau-Ponty I will take issue
with his view of the method of argument which the Critical
philosophy employs. It is an argument which consists in a
£§g£§§§ on a g__fagt9 intersubjective world to the
transcendental ego as the condition which renders such a

world possible. The method, in other words, is the gflgiygig

one, which in "the preparatory exercises" of the Proleggmgh

1m

 

 

 

17

Kant says he used, as opposed to the synthetic method of the
Critigge.56

What I want to point out now is the asymmetry wich
obtains between the regressive or analytic method which
Merleau—Ponty ascribes to the "Critical philosophy," and the
synthetic, i. e., deductive, and hence rigorous and
scientific method of the Qgitigue. For the argument of the
Qgitigue should have as its conclusion a world which is gg
fggtg intersubjective, i. e., the existence of a plurality
of thinking beings or empirical selves. This conclusion,
however, the argument does not have. The basic claim of the
argument is only that its premise, namely the unity of
consciousness, requires that it be consciousness of "objects
in general," not however, of particular empirical objects,
such as tables or human beings.

This becomes clear once we reflect upon the nature
of Kant's inquiry. Its purpose is, at the most general

level, to establish what Kant calls "tggnsgengggtal

 

knowledge" (A11 = B25; NKS 59): by it "we know that—~and
how--certain representations (intuitions or concepts),"
although they are "not of empirical origin...can yet relate
a priggi to objects of experience" (A 56 = B 80f. NKS 96),
And we can know this, Kant claims, because the arguments of
the Transcendental Analytic shOw that only through these
representations "is it possible to gggy anything g§_gh
ggjggt" (A 92 = B 125; NKS 126). In other words: a

transcendental inquiry in Kant's sense concerns the a pgiggi

 

18

conditions which are necessary if there is to be empirical
knowledge. But these a_p§i9§i conditions are all fggmal
conditions, and it is precisely their formal character which
sets a limit to what in a transcendental inquiry can be
known.

Kant claims that the forms of sensibility, space and
time, and the forms of the understanding, its pure concepts
or the categories, make experience and objects of experience
as regards their form possible (cf. A 128; NKS 1M9). But
when he uses the term "form," Kant does not mean, as he
points out in De Mundi Sensibilis, "the outline or some kind

"57

of shape of the object, but, as he says in the Logig,

"the manner in which we kngu the object."58 In the
terminology of the QELLAQEE we can say that these forms are
the respective ways in which a sensible manifold of
representations is ihtuitgg, thgught, i. e., united in one
consciousness, and thus kngyn. This is the reason why in
the Summary of the Transcendental Deduction Kant can say
that "the complete unity" of such a manifold "in one and the
same apperception...constitutes the fggg of all knowledge of
objects" (A 129; NKS 1M9; emph. added).

The sensible manifold itself, however, is outside
the scope of Kant's transcendental inquiry: it denotes not

the formal, but the mageri I element of experience and is g

postegiggi. "In the field of appearance," Kant says, "in

terms of which all objects are giyen us, there are two

elements, the form of intuition (space and time), which can

 

 

19

be known and determined completely a_pri9§i. and the matter
or content" (A 723 = B 751; NKS 583; emph. added). And this
material element, which accounts for the content of our
actual experience, can be given only empirically: "The
matter of appearances, by which things are given us in space
and time, can only be represented in perception, and
therefore a_pgsterigri" (A 720 = B 7N8; NKS 581). What a
transcendental inquiry, then, can be only concerned with,
are the a_pr19:i conditions which a sensible manifold must
fulfill if it is to be united in one consciousness and so
constitute an object of knowledge: "In respect to this
material element we have nothing a_prigri except
indeterminate concepts of the synthesis of possible
sensation, in so far as they belong, in a possible
experience, to the unity of apperception" ( A 723 = B 751;
NKS 583).

We can express Kant's position differently as
follows. We can know a_prigri that if a manifold of
representations is to belong to one and the same
consciousness, then it must be united by means of, for
example, the pure concept of substance and accident. But we

cannot know a_pri9ri what particular kinds of empirical

objects will instantiate this concept: this we find out a

 

pggtgrigri, by means of experience. And the question as to
what we empirically do find out is a question which is
outside the scope of Kant's transcendental inquiry. "The

objects of trancendental philosophy are not objects of

 

 

 

20

59

perception." This, now, is the reason why, in

Merleau-Ponty's words, "the problem of the knowledge of

"60

other people is never posed in Kantian philosophy. The

reason is not, as Merleau-Ponty claims, that the analysis
"is from the start located outside me."61 This claim, too,
has become untenable: the analysis concerns the unity of
consciousness which is the premise of the Transcendental
Deduction (cf. B 131f. NKS 152f.). And the consciousness

whose unity we investigate is, of course, in each case

"mine. I!

(5)

Kant claims that the unity of consciousness implies
experience of objects, i. e., the unification or synthesis
of a given manifold of sensible intuition by means of the
categories. But Kant does not say that the unity of
consciousness requires experience of objects ggg experience
of objects of a special type, namely subjects of experience
other than myself.

His position, thus, is at variance with the views,
for example, of husserl and Strawson. In his Q飣§§123
flgglggtigns Husserl wants to defeat the objection that his
version of transcendental philosophy results in a
transcendental solipsism by arguing that the experience of
an objective world cannot be constituted by me alone,
because objectivity, as actuality for everybody, already

contains the sense 'other subjects'. "The sense of the

 

 

21

being of the world, and in particular of objective nature,
obviously includes being-there-for—everybody, and this is
always co—intended wherever we speak of objective

actuality."62

And among the results of his account of the
experience of someone else, Husserl mentions the following:
"My ego, which is apodeictically given to me...can a_p£igzi
be a world-experiencing ego only by being in communion with
others like himself: by being a member of...a community of
monads."63

And Strawson, in connection with the issue of a

_al§ that

non-solipsistic consciousness, claims in Ihgiyig
the idea of oneself as subject of experience requires not
only experience of something other than oneself, i. e., of
an objective domain, but also that one actually shares it
with other such subjects. For according to Strawson "it is
a necessary condition of one's ascribing states of
consciousness, experiences to oneself, in the way one does,
that one should also ascribe them, or be prepared to ascribe
them, to others who are not oneself."6u

It appears to me that Kant can answer Husserl and
Strawson as follows. The pgggygg of the unity of

consciousness concerns the possibility as to how a manifold

 

of representations "ggn stand together" (B 132; NKS 153) in
one consciousness, or conversely how "it is possible for the
'I think' to accompany all my representations" (B 131; NKS
132). And the sglytigg of the problem consists in what is

necessary for this possibility to obtain: in what the unity

 

22

of consciousness, in other words, implies. Clearly, the
inclusion, among my representations, of representations of
other subjects of experience, cannot solve the problem but
leaves it where it is. The solution, Kant maintains, is the
synthesis of the representations by means of the categories,
and it is these which are a_p;i9§i necessary for the unity
of consciousness. And if it is claimed that among the
categories would have to be, let us say, the pure concept of
a person,65 then every object of experience would have to
fall under this conept.

Whether or not there are other subjects besides
myself is for Kant then a question of contingent fact which
can be answered only a_29§t§§igri, that is, by experience.
And this in my view also means that for Kant the question
concerning the relation of those of our empirical judgments
in which we ascribe states of consciousness to ourselves,
and the judgments in which we ascribe states of
consciousness to others, can be only an empirical, not a

philosophical question.

(6)

In my view, then, our 'knowledge of other minds‘ is
for Kant not a philosophical problem, i. e., a problem to
which a transcendental—a_pyi9ri inquiry would have to
provide the solution. In my opinion, Kant more so than any
philosopher before him, is interested in the question as to

what we can mean by the expression "a mind." We can read

 

 

23

the chapter on the Paralogisms as telling us what we must
not mean by this expression, and the arguments of the
Transcendental Analytic as an investigation into the
necessary condition of "a mind" or, in Kant's words, of "a
consciousness in general."66 "Other minds," however, are
for Kant just particular items in the class of empirical
objects. And as there is, for instance, no special
philosophical problem called "our knowledge of tables," so
there is, correspondingly, for Kant no special philosophical
problem about other minds. To be sure, Kant himself points

spgh, can never be found by us

 

out that "thinking beings, gs
among outer appearances, and that their thoughts,
consciousness, desires, etc., cannot be outwardly intuited"
(A 337; NKS 33f.). But we cannot, in the Ordinary sense or
the term, see electrons, for example, either: and this does
not make the question whether or not there are electrons a
philosophical question.

But here we must enter a caveat. We have to take
Kant's claim that the unity of consciousness implies
experience of objects as a formal-transcenental, not as a
material—empirical claim. More explicitly, his thesis is
that the unity of a manifold of sensible representations in
"one and the same apperception" (A 129; NKS 149) is a
synthetic unity which has as its necessary condition that
these representations be united in "concepts of an object in
general," that is, in the categories (B 128; NKS 128). And

so united they obtain, in addition to their being mere

 

 

2H

"modifications of our sensibility" (A 491 = B 519; NKS "39),
relation to an obiect. Thus Kant can say that "it is the
unity of consciousness that alone constitutes the relation
of representations to an object, and therefore their
objective validity and the fact that they are modes of
knowledge" (B 137; NKS 156). It is for this reason that
Kant calls the synthetic unity of consciousness "the form of
all knowledge of objects" (A 129; NKS 149) which must
"precede all experience" and which is "required for the very
possibility in its formal aspect" (A 130; NKS 150). Kant
wants to say, in other words, that there are certain
invariant or ngméii§i§m§2t§ which are necessarily involved
in every instance of empirical knowledge, and that these
elements can be known g_p§igri, i. e., constitute the topic
of a transcendental inquiry. Yet such an inquiry can say
nothing about the actual empirical content of our knowledge.

This content is due to the mgtegialyglement in experience,

that is, to sehsatign. Of the latter Kant says that it is
"the matter of perception," that it "can never be known g
leggl", and that it "therefore constitutes the distinctive

difference between empirical and g_pri9§i knowledge" (A 167
= B 209; NKS 209).

But although it is a contingent matter of fact, i.
e., to be found out "only...a_pgs_e:i_ri, by means of
exp<2rience" (A 721 = B 7H9; NKS 581) what particular kinds

of empirical objects there are, their experience is itself

"possible only in conformity" (A £21 + B 7kg; NKS 581) or in

 

 

25

"accordance with" (A 723 = B 751; NKS 583) the formal
conditions of an experience in general. And this means for
our problem that these conditions: the forms of intuition,
space and time, the categories and the unity of
apperception, must be such that, although they do not
require, they nevertheless do not preclude, but leave room
for that experience by which we find out whether there are
other thinking beings besides ourselves. For if these
transcendental conditions would not allow for such an
experience, Kant would be in an awkward position: the a
prigri nature of his inquiry allows him, on the one hand. to
dismiss the problem at hand as an empirical problem, whereas
on the other hand, according to the very same inquiry, there
can be no empirical solution of it.

And clearly, the transcendental conditions of an
experience in general d9 rule out certain things as

empirically impossible. 0n a-prigri grounds there can be no

 

experience of, for example, miracles, if by a miracle we
mean "an event the cause of which cannot be found in
nature."67 Nor can there be experience of the "power of
§pggtgnegu§ly beginning a series of successive things or
states" (A ""8 = B #76; NKS "12), that is, of freedom. "The
concept of freedom," Kant says, "is a pure concept of
reason...for which there is no corresponding example in any
possible experience, which therefore forms no object of any

"68

theoretical knowledge possible to us. And let us

mention finally that on a_prigri grounds there can be no

 

 

26

experience of our noumenal selves or of ourselves as

persons, 1. e., of "moral personality" which "is nothing
but the freedom of a rational being under moral laws."69
Hence it is only the experience of the "psychological self

. O
as empirical consciousness,"7

that is, of the empirical
self, for which Kant's transcendental inquiry must be shown

to leave room.

In Chapter II, I will examine some of the arguments
of the Transcendental Deduction in order to show that what
Kant calls ’inner experience' is empirical knowledge in the
strict sense of the term, i. e., consists in judgments which
possess what for Kant are equivalent terms, namely objective
validity and necessary universality (for everybody). In
Chapter III, I will discuss the Second Analogy, that is, the
Transcendental Deduction of the category of cause and
effect; my aim is to show what kinds of objects are
constituted by the categories. And in Chapter IV, I will

show that thinking beings other than ourselves can be known

as objects of experience.

 

27

NOTES TO CHAPTER I

1Immanuel Kant, Werke in sechs Baendeh, VI, ed.
Wilhelm Weischedel, Darmstadt: Wissenschaftliche
Buchgesellschaft, 1966, p. 411. Hereafter quoted as Werke.

2

 

ImmanueliKantls_giitigue_9f_PureiBea§9h, trsl.
Norman Kemp Smith, London MacMillan & Co. Ltd., 1956.

3Jonathan Bennett, Kgyt;s_§ialegtic, Cambridge:
Cambridge University Press, 1979, p 69

”P. F. Strawson, The Bounds of Sense, London:

Methuen a Co. Ltd., 1966, p. 197. Hereafter quoted as Thg
Bcundswgf_§en§e.

5Cf. The Philosophical Works of Descagtes, I and
II, trs. Elizabeth S. Haldane and G. R. T. Ross, Cambridge:
Cambridge University Press, 1969, pp. 157-61; p. 2A0.
Hereafter quoted as Haldane.

60f. Jonathan Bennett, Kg

Cambridge University Press, 1966,
Strawson, l22i_2iii. pp. 97-112.

 

tie Analytic, Cambridge:
p. 1; P. F.

_a I
W
O

1
LA)

g
P

7Cf. A 1ou, NKS 13M; A 129, NKS 149; A 191 : B 236,
NKS 220.

8Cf. Edmund Husserl, "Erste Philosophie,"
figssegliang, VII, Haag: Martinus Nijhoff, 1956, p. 225.
Hereafter quoted as Ha VII. Hy translation.

90f. Ha VII, pp. 277, 378.

10Ha VII, p. 226.
‘1Ha VII, p. 225.
12Ha VII, p. 222.
13Ha VII, p. 222.
1uPa v11, p 227
15Ha VII. p 373.
16

Ha VII, PP. 369 f.

Cf. B 157, NKS 168-9; E 406-7, NKS 368-9.

 

 

 

28

18Jean-Paul Sartre, Beingyahg_flothingness, trs.

Hazel E. Barnes, New York: PhilsophicalhzibraryT—1956, p.
225. Hereafter cited as Sartre.

19Sartre, p. 225.

0Sartre, p. 225.

 

21Sartre, pp. 228-9.
22Sartre, p. 227.
23Sartre, p. 269.
2“Sartre, p. 269.
25Sartre, p. 270.
26Sartre, p. 148.
27Sartre, p. A80.
28$artre, p. 225.

29

Sartre, p. 225.

3O
“ Sartre, p. 225.

Q
“1Sartre, p. 226.

32Sartre, p. 226.

 

33Sartre, p. 226.

3“Sartre, p. 226.

35Sartre, p. 227.

36Sartre, p. 226.

37Sartre, p. 226.

38Sartre, p. 226.

39Sartre, p. 226.

0Sartre, p. 227—8.

1HSartre, p. 314.

quartre, p. 225.

lKJ’Sartre, p. 225.

 

 

 

29

1MSartre, p. 226.

l”Sartre, p. 226.

”6Ha VII, p. 222.

u7Ha VII, p. 369.

8
u a. Merleau-Ponty, Phenomenology of Perception,
trsl. Colin Smith, New York: The Humanities Press, 1970, p.
62. Hereafter quoted as Merleau—Ponty.

l‘9Merleau-Ponty, p. 62.

50Merleau-Ponty, p. 62.

51Merleau-Ponty, p. 62.

52Merleau—Ponty, p. 62.

S3Merleau—Ponty, p. 62.

5“Merleau-Ponty, p. 62.

55Merleau-Ponty, p. 61.

56Immanuel Kant, Prolegomena to Any Future
Metaphysigs, ed. Lewis White Beck, Indianapolis-New York:
The Hobbs-Merrill Co., Inc., 1950, pp. 21-2. Hereafter
quoted as ngl. Nevertheless some interpreters have been of
the opinion that the 'transcendental method' is
analytic—regressive. This view springs from the belief that
through Kant's interpolation of certain passages from the
Egglggggehg into the Introduction to the second edition of
the Cpitigge (B 14-24), the latter shifted towards the
Proleggmgna in the sense that its argument, unlike that of
the first edition, has now to be understood as an
analytic-regressive argument (cf. Vaihinger, ggmmentag, I,
u1u-15).

In Kant's own description the analytic-regressive
method starts from a "fact" (Eggl. 23) and then moves by an
analytic exposition backwards to its conditions or grounds,
which are held to gyplgih_hgy the given fact is possible.

We begin with "something already known as trustworthy... and
ascend to sources as yet unknown, the discovery of which
will... explain to us what we knew" (ngl. 22). Thus the
regressive argument of the nglegomehg proceeds from the
fact that there actually are valid synthetic judgments a
pgiggi in pure mathematics, for example, and ascends to
their conditions, namely that space and time are a_p§i9ri
forms of intuition (p. 31 #11-12). And these conditions are
said to explain how this "uncontested synthetical knowledge

 

 

30

g pgiogi" is possible (3:31. 22).

NOw in #117 of his ngik Kant says that the
analytic—regressive method is more suited for popular than
for scientific and systematic purposes. And this clearly
because a regressive argument is ihgghclgsiye in a twofold
sense: neither does the fact, on which the argument rests,
guarantee the truth of the conditions to which it moves, nor
do reversely these conditions, by themselves, constitute a
proof of that fact (of. Wolff, Kahtjsmlheggy, p. 46). The
"Eagtum," therefore, "on which the ££2l§£2m§gé are based, is
not completely indubitable, but vigorously doubted by the
sceptics" (Vaihinger, I, #13). But it is precisely
scepticism that the Critical philosophy is supposed to
defeat: "Criticism alone," Kant says, "can sever the root
of... scepticiSm" (B XXXIV; NKS 32) and, "in the end,
necessarily leads to scientific knowledge, in contrast to
"unfounded assertions" (B 23; NKS 57).

So let us now turn to Kant's other, the
synthetic-progressive method, which according to the ngik
is scientific. This method is the method of deduction that
"goes from the principles to the consequences" (ngyg,
#117): it is rigorous insofar as the truth of the premisses
guarantees the truth of the conclusion. This method, in
contrast to the analytic-regressive, "seeks, without resting
on any fact, to unfold knowledge from its original germs
(£32I., p. 22). In the synthetic procedure, Kant says, we
do not start from facts: on the contrary, these "must
strictly be derived ih_ab§t§§gtg from concepts" (PQQI. p.
26). Kant gives us a hint how both methods are related:
"The £39l§g9mepa...are intended to point out what we have to
do in order to make a science actual, if it is possible"
(Eggi., p. 22). That is to say: whereas the Pygleggmegg
move from the fact of the validity of synthetic judgments a
pglggi in pure mathematics and pure science by an analytic
regress to the conditions that egplaih how that validity is
possible, it is the synthetic—progressive method of the
Qyitigge that at first proves these conditions and so pggyes
the validity of these judgments to be ggtgal. Vaihinger
puts it this way: "In the Eygleggmeha Kant takes as a basis
the validity of mathematics and pure science as
incontestable, indubitable fggtgm: Kant analyzes this fact
in order to discover... its conditions. These conditions
are the grounds of explanation (principia assehgi) of that
in itself incomprehensible fégtgm... Differently, and even
in the opposite way, the gggtigge: it establishes first that
there are such conditions... which it finds ggggg
lgggpgggegtly, i. e., without taking that fggtgm in its
argumentation into consideration. Then it shows that the
iactum can not only be completely and uniquely explained in
terms of these conditions, but that it is moreover
necessarily contained in them and follows from them. In
other words: those conditions are here both, gpguhgsycf
gxplagation (Drincipia essendi : explicatig) and grounds

 

 

10
1%

 

31

2.19.212 (scintimaJrcbangi) for that factum"... Since it also
"furnishes the 222g: of the fagtgm," the "synthetic
procedure is more complete and scientifically more
satisfactory" (Vaihinger, I, 412—13).

But now it appears that this view becomes untenable
as soon as we look at how in the Introduction to the second
edition, Kant formulates the task of the gritigug: for after
asserting that the gritigug has to answer the questions how
pure mathematics and pure science of nature are possible, he
continues in the manner of the Eggleggmena: "Since these
sciences actually exist, it is quite proper to ask how they
are possible; for that they must be possible is proved by
the fact that they exist" (B 20; NKS 56; cf. p. 22). But
clearly, this way of stating the problem can lead us to the
belief that the method of the gritigue is, after all, the
analytic-regressive method of the Ergleggmena. And this
means that at the very outset the gritigpe presupposes, or
takes for granted, what it is supposed to prove-—that these
sciences are actual. The argument of the Qritigug,
therefore, can have only the strength of an exposition or
ggpligatig, but not the apodeictic character of a proof (of.
Vaihinger, I, H12).

According to Vaihinger, however, the view that 8
underwent change to the analytic—regressive method of the
Eroleggmgna is based on both "a factual and methodological
error" (Vaihinger, I, 415). On a factual error because
already in A the status of synthetic judgments a_p§igri in
pure mathematics and in pure science is by no means a
problematic one in that at first Kant grants these judgments
their claim to validity only in order subsequently to decide
upon the legitimacy of this claim. On the contrary, if we
believe Vaihinger, Kant never doubted the validity of such
judgments in pure mathematics, and at least since 1770
entertained the possibility of their validity in pure
science (Vaihinger, I, 388-89). And in the first edition of
the gritigue the validity of both types of judgments is
presupposed as a fact. So Kant mentions, for example in A
A, the "established reliability" of mathematics and its
"shining example of how far, independently of experience, we
can progress in a_p§igri knowledge." And in A 209 Kant says
of the principle of causality that "direct inspection may
show the principle to be actual and true" (my trans.) so
that even "the question how it should be possible, may
therefore be considered superfluous" (NKS 232). It is,
Vaihinger maintains, the very presupposition of their
validity that at first gives rise to the genuine Critical
problem: why is it that I can form valid synthetic judgments
§_2£;9£; about objects? (Vaihinger, I, 395).

For this reason the fact that in the Introduction to
B the validity of these judgments is much more pronounced
than in A should not lead us to commit the methggglggical
error of holding that in B Kant took a turn to the
analytic-regressive methoo of the firoleggmena (Vaihinger, I,

 

 

 

32

#15). In other words: that Kant formulates in B the
Critical problem in Ergleggmgha-fashion, does not mean that
he intends to solve it in the analytic-regressive way. On
the contrary: if in the Brgleggmena the validity of
synthetic judgments a_prigri is basis and m§§fl§ of the
argumentation, in the gritigpg it is neither. But it is,
according to Vaihinger (I, 415), the gbjectiye of the
argumentation, of an argumentation that Kant expressively
calls synthetic, i. e., deductive. This should be taken
seriously. That is to say: we have to look, for example in
the Analytic, for a "principle" or a true premise of which
valid synthetic a_pgi9§i judgments "can strictly be derived"
(232;. 26).

57Werke, III, p. 31.

egke, III, p. “57.

59Kant's Gesammelte Schriften, Band XXI,
herausgegeben von der Preussischen Akademie der
Wissenschaften, Berlin und Leipzig, 1936, p. 92. Hereafter
quoted as WW.

 

 

 

6OMerleau—Ponty, p. 62.
61.
herleau-Ponty, p. 62.

62Edmund Husserl, "Cartesianische Meditationen und
Pariser Vortraege," gussegliana, I, Haag: Nartinus Nijhoff,
1950, p. 124, my trs. Hereafter quoted as Ha I.

63Ha I, p. 166.

611P. F. Strawson, Indiyiguals, London: Hethuen a Co
Ltd., 1965, p. 166. Hereafter quoted as Lngiyiguals.

65Cf. Strawson, Ingiyidgals, p. 101-02.

66For this term see B 1113 (".e_ine_App.e.r_z_§2.t.lgn
ueberhaupt," "ein Bewusstseinrpeperhanpt") and Epgl. M8
("consciousness in general").

67

"Ueber Wunder," NW 18, p. 320.

68"Introduction to the Metaphysics of Morals," in
Kant‘s Critique of Practical Reason fiDQ-Q§D§E-HQIE§_QD-L£§
lh§921_2§_§3big§, trs. T. K. Abbott, London-New
York-Toronto: Longmans, Green and Co., p. 277. Hereafter
quoted as Abbott.

69Abbott, p. 279.

70

O\
O
h)

"Fortschritte," in flgrgg, III, p.

 

CEAPTER ll

SUBJECTIVITY [o]

 

"...a blind play of representations,
less even than a dream."
(A 112; NKS 129)
(1)

I will henceforth refer to the objectivity of an
object in the Critical-transcendental sense as 'objectivity
[ct]; it consists, as we have seen, in the necessary
synthetic unity of a manifold of representations. This
unity is not a function of intuition, by whicn the manifold
is given, but of thought or of concepts. An object of outer
intuition is thus objective [ct] not because it is
represented in space, and there is 10 reason to suppose that
something can not be objectiVe Lot] on the ground that it is
an object of inner intuition, an object, that is, which is
represented only in time (of. A 37“; NKS 342). Synthetic
unity, Kant maintains, is due to the categories: they

"consist solely in the representation of this necessary

 

synthetic unity" (A 79 : B 104; NKS 112,.

He might now be tempted to form a ccrresponcing
notion of what is subjective or, as I shall call it,
'subjective [o],' namely a manifold which does not possess
synthetic unity and the representation of which would lacn

what synthetic unity accounts ior, i. e., ”objectiVC

. . , 1 ,
validity and necessary universality (10F everytocy,." inc

 

—-—-—va

34

representation would, instead, be valid only for the subject
in which it is found, even if, of course, it would be a
representation of an object in space.

In an early stage of the Transcendental Deduction in
the first edition of the Critique Kant, indeed entertains
the thought of a sensible manifold of intuition which does
not possess necessary synthetic unity. In # 13 we fine
three versions of the same idea. (1) "Objects...may appear
to us without their being under the necessity of being
related to the functions of understanding" (A 69; NKS
123f.). (2) "Appearances can certainly be given in
intuition independently of functions of the understanding"
(A 90; NKS 124). (3) "Since intuition stands in no need

whatsoever of the functions of thought, appearances would

 

none the less present objects to our intuition" (A 90; NKS
152).
It is obvious that if the Transcendental Deduction

is to succeed, these claims must be false. For its very

 

strategy consists in showing that consciousness could not
have the formal unity it has, namely that "it must be
possible for the 'I think' to accompany all my
representations" (B 131; NKS 152), unless these
representations were united by the functions or the
understanding, 1. e., by the categories. But such a stategy
would be pointless if it were also maintained that we can be
conscious of a manifold of intuition which does not possess

such a unity. And indeed, at a later stage of the

 

—ii-——--...

35

Transcendental Deduction, Kant rejects the claim. A manifold
"in a state of separation," he says, "and without belonging
to consciousness of myself...is impossible" (A 122; NKS
145). he maintains that the categories "must be reCOgnized
as a_pri9§i conditions of the possibility of experience,
whether of the i_t_iti9h which is to be met with or of the
thought" (A 94; NKS 126; emph. added). And finally he
claims that "all sensible intuitions are subject to the
categories, as conditions under which alone their manifold

can come together in one consciousness" (B 143; NKS 150).

(2)

My proposal has been to consider, as subjective [o], a
manifold of representations which is given in intuition, but
which does not possess necessary synthetic unity. In the
passage last quoted, however, Kant claims, in effect, that
necessary synthetic unity is the only condition under which
a manifold of representations "can come together in one
consciousness." What he rules out, in other words, is that
there can be (one) consciousness of a manifold of
representations which does not possess synthetic unity.
This, he says in the Transcendental Deduction, "is
impossible" (A 122, NKS 145).

It is in my view one of the basic tenets of the
Transcendental Deduction that the formal property of all

consciousness, namely its pgity,is by no means something

obvious, but something of which the transcendental

 

36

possibility has to be explained. And this for the following
reasons.

It is held to be analytic, i. e., to be contained in
its very concept, that a representation can exist or occur
only if there is a subject which is conscious of it. Thus
Descartes says in his Replies that "it is certain that no
thought can exist apart from a thing that thinks, no
activity, no accident can be without a substance in which to
exist."2 Kant, too, subscribes to this view, without
holding, however, that a 'subject' is a Cartesian thinking
substance: "All representations have a necessary relation to
a possible consciousness. For if they did not have this,
and if it were altOgether impossible to become conscious of
them, this would practically amount to the admission of
their non—existence" (A 118a; NKS 142). And what is true of
the genus 'representation,‘ is true of every of its species.
thus intuitions "are nothing to us, and do not in the least
concern us, if they cannot be taken up into consciousness"
(A116; NKS 1111;..3

But while it is analytic that every representation
has a necessary relation to some conscious subject or other,

it is “gt analytic that each of a manifold of
representations has such a relation to one and the same

conscious subject. This, it appears to me, is the reason
why the unity of consciousness must be accounted for. In

the ParaIOgisms Kant himself points out that we can conceive

of a plurality of representations, "for instance, the single

 

 

 

37
words of a verse," as being "distributed among different
beings." But these representations would, Kant continues,

"never make up a whole thought" (A 352; NKS 335). Our point
here is that if we entertain the IOgical possibility of a
plurality of consciousnesses of one representation each, we
are talking only spuriously, and not in any cognitive
relevant sense about both, consciousness and
representations. "If each representation were completely
foreign to every other, standing apart in isolation, no such
thing as knowledge would ever arise. For knowledge is
[essentially] a whole in which representations stand
compared and connected" (A 97; NKS 130). Hence Kant wants
to say not only, as I have repeatedly pointed out, that the
unity of consciousness implies experience of objects, but
also the reverse, namely that empirical knowledge has as its
necessary condition the unity of consciousness.

In the Transcendental Deduction, then, Kant must
address himself to three distinct but connected issues.

1

(1) The possibility of a mgnifglg OI sensible
representations to "stand together" (B 132; NKS 153)in one
Eggsgiggsness. (2) The possibility of ope consciousness of

such a manifold, or, as Kant puts it, of the "thoroughgoing

 

identity of the apperception of a manifold which is given in
intuition" (B 133; NKS 153). (3) The possibility of the
representations to form a "connected whole." In the

following we will try to reconstruct, in its basic features,

 

Kant's solution of these three problems.

 

 

 

(1) A condition of this possibility would be,
instead of each representations of the manifold having a
necessary relation to some conciousness or other, that all
representations of the manifold have such a relation to one
and the same consciousness. It would, then, indeed be
"possible for the 'I think' to accompany" (B 131; NKS 152)
all the representations of the manifold; and the "necessary
relation" which "all the manifold of intuition" is said to
have "to the 'I think' in the same subject in which this
manifold is found" (E 132; NKS 153), could obtain.

(2) It can obtain, Kant claims, only by means of a
synthesis, and his further claim is that this synthesis also
secures the unity of consciousness. The relation of a
manifold of representations to one and the same
consciousness is not a matter of "the empirical
consciousness which accompanies different representations"
and which "is in itself diverse and without relation to the
identity of the subject." This relation, rather, comes
about "not simply through my accompanying each
representation with consciousness, but only in so far as I

njgin one representation with another, and am conscious of

g
the synthesis of them. Only in so far, therefore, as I can
unite a manifold of given representations in pgg
Qgfl§glp§§g§§§, is it possible for me to represent to myself
the identity of the consci0§§p§§§_in [1,e3_phroughoutj these
Ewsentatigns" (B 133; NKS 153).

(3) Now the unification of a manifold of sensible

 

 

39

representations, 1. e., the constitution of its unity, can
be according to Kant not a matter of intuition. in which the
manifold is given, but only of concepts. Concepts are for
Kant "functions of unity among our representations" (A 93 =
B 93; NKS 105f). The concepts, however, in which the
manifold is to be united, cannot be empirical concepts, for
these are derived from experience which presupposes the
unity of the manifold. They are the concepts, already
referred to, which "consist solely in the representation
of...necessary synthetic unity" (A 79 = B 104; NKS 112),
i.,e., the pure concepts of the understanding or categories.
And a manifold of sensible representations which in these
and only in these concepts is united, is an object [ct]. A
manifold which is united, for example, in the empirical

concept of a tree, is in the strict sense of the term not an

 

object [ct].
For our present purposes, then, we must note that a
manifold of representations of sensible intuition can be

accompanied by the 'I think' only if it possesses synthetic

 

unity. Hence what we have called 'subjective [o],' namely a
manifold of intuition which stands together in one
consciousness, but which does not possess synthetic unity,
is transcenentally impossible. And this is the reason why
Kant must reject the thought entertained at the beginning of
the Transcendental Deduction, i. e., that "appearances might
very well be so constituted that the understanding should

not find them in accordance with the conditions of its

 

40

unity" (A 90 = B 123; NKS 124). Now he says: "There might
exist a multitude of perceptions, and indeed an entire
sensibility, in which much empirical consciousness would
arise, but in a state of separation, and without belonging
to a consciousness of myself. This, however, is imoossiolo.
For it is only because I ascribe all perceptions to one
consciousness (original apperception) that I can say of all
perceptions that I am ooosoioos of them" (A 122 = B 145;

emph. added).

(3)
Although the notion of the subjective [o] is a
notion of something which, as we have seen, cannot be

constituted, I will not discard but retain it for heuristic

 

purposes. For by examining what this notion implies, I
intend not only to bring the epistemic function of the
"pure," "original," and "transcendental" unity of

apperception (B 132; NKS 153) into focus, but also to show

 

the incoherence of any account of experience which is based
on this notion.

At times Kant himself adopts such a strategy, i. 6.,
assumes for the sake of argument of what is subjective [0].
Thus in a letter to Herz he assumes with respect to a
manifold of representations that "I am...conscious of each
individual representation, but not of their relation to the
unity of representation of their object, by means of the

synthetic unity of their apperception."u On this

 

 

N1

assumption, however, the representations would, Kant
maintains, never "represent objects. They would not even
reach that unity of consciousness which is necessary for
knowledge of myself (as object of inner sense). I would not
even be able to know that I have [them]; consequently for
me, as a knowing being, they would be absolutely nothing.
They could still carry on their play in an orderly fashion,
as representations connected according to empirical laws of
association, and thus even have an influence upon my feeling
and desire, without my being aware of them...This might be
so without my knowing the slightest thing thereby, not even
what my own condition is." 5

Now in this passage Kant suspends, as it were, what
in the opening sections of the Transcendental Deduction in B
he establishes as "the principle of the original §l§£fl§£l£
unity of apperception" (B 137; NKS 156). This principle says
that "all oy representations in any given must be subject to
that condition under which alone I can ascribe them to the
identical self as my representations, and so can comprehend
them as synthetically combined in one apperception through
the general expression ' I think'"(A B 138; NKS 157).

The condition, to which Kant here refers, is, as we
have seen, the condition of their synthesis in a pure
concept. It is in such a concept, Kant wants to say, that
representations first of all can form "a whole" or "stand
compared and connected," that is, be modes of knowledge or

have relation to an object. "Knowledge," Kant says,

 

 

 

mi“ " ’"T

M2

"consists in the determinate relation of given
representations to an object; and an oojoot is that in the
concept of which the manifold of a given intuition is
EEILQQ" (B 137; NKS 156). To suspend the principle, then,

means that our representations would fail to represent

objects: they would, as Kant says in the same letter, "be

nothing to us, that is, not objects of goowlmogo at all."6
But they would also have no "relation to the
identity of the subject." For the latter requires according

to the principle precisely their synthesis. “The thought
that the representations given in intuition belong to me,
is...equivalent to the thought that I unite them in one
self—consciousness, or can at least so unite them; and
although this thought is not itself the sloohosis of
representations, it presupposes the possibility of that
synthesis. In other words, only in so far as I can grasp
the manifold of the representations in one consciousness, do

I call them one and all slog" Otherwise, Kant continues, "I

should have as many-coloured and diverse a self as I have

 

 

representations of which I am conscious to myself" (E 13h;
NKS 154). I could, as Kant puts it in his letter to Herz,
not even know "what my own condition is."

What we can mean, then, by the notion of the
subjective [o], is a manifold of ciconnected
representations, standing apart in isolation, each of which
is accompanied by a different empirical consciousness. But

with this, Kant wants to say, we can in the oogoitive sense

 

 

43

mean nothing at all. For such a manifold has no cognitive
function: on the basis of the subjective [c] "we should
have," as Kant writes to Herz, "knowledge neither of

ourselves nor of other things." 7

(H)

We have seen that in an early stage of the
Transcendental Deduction in A Kant maintains that the notion
of what I call 'subjective [01' has a legitimate
constitutive employment. This claim, however, he later
dismisses: what the notion of the subjective [o] purports to
refer to, is, as he says, "impossible." And in the meantime
we have seen, too, that these two claims cannot be made
compatible. But it is nevertheless instructive to
understand why Kant gets entangled in this contradiction, a
contradiction of which he apparantly is never completely
aware: for although we find only a few traces of it in the
revised edition of the Critiqoe (cf. B 1h1—U2; NKS 158-59),
the contradiction reappears, as we shall see, at the basis
of an untenable doctrine in the Prolegomona.

It appears that the reason for Kant's quandry is to
be found in the peculiar method which he employs in the
Critique and elsewhere. The Eritigoe, De Vleeschauwer
writes, "is based on the method of isolating the factors
that occur in knowledge."8 Similarly Vaihinger talks about
Kant's "decomposing method which consists in considering

"9

each of the cognitive faculties at first by itself. Thus

m—J

 

 

 

44

Kant says, for example, that "in the transcendental
aesthetic we shall l§2l§£§ sensibility, by taking away from
it everything which the understanding thinks through its
concepts"(A 22 = B 36; NKS 67), and that in "a
transcendental logic we isolate the understanding...[by]
separating out from our knowledge that part of thought which
has its origin solely in the understanding" (A 62 = B 87;
NKS 100).

Kant owes the use of this procedure not merely to
considerations of exposition, but rather to a philosophical
reason which in the Transcendental Doctrine of Method he
sets forth as follows: "It is of utmost importance to
isolsoo the various mooes of knowledge according as they

differ in kind and in ogigio, and to secure that they not be

 

confounded owing to the fact that usually, in our smoloymopo

of them, they are oomoi eo. What the chemist does in the

analysis of substances...is in still greater degree

incumbent upon the philosopher" (A 8H2 : B 870; NKS 660;

 

emph. added). Kant's method of isolation, in other words,
is based on the idea that although our knowledge is, or
"usually" is, knowledge only ss "ooojopo53oo" (A 271 : B
327; NKS 283) of elements, it is a conjunction of elements
none of which is reducible to the other. Thus he says at
the beginning of the Transcendental Logic that "intuition
and concepts constitute...the elements of all our knowledge,
so that neither concepts without intuition in some way

corresponding to them, nor intuition without concepts. can

 

 

I—E—nw-n—w

45

yield knowledge" (A 50 B 7“; NKS 92). And he adds that
although it is only through their "union" that knowledge can
arise, these elements "cannot exchange their functions" and
that it is therefore necessary to "carefully separate and
distinguish the one from the other" (A 51-2 = B 75-6; NKS
93). Kant's admonition is well founded. For he asserts here
one of the basic tenets of the Critical philosophy, namely,
in Strawson's words, the "original duality of intuition and
concept."10 And he asserts it against his predecessors,
against Descartes, Leibniz, and Spinoza, against Locke,
Berkeley and Hume, all of whom in his view in one way or
another "confounded" intuition and concept or assimilated

the one to the other. "Leibniz ; tollooooslisoo appearances,

 

just as Locke soososlisso all concepts of the understanding"
(A 271 = B 327; NKS 283).

Kant's method, however, is subject to a
misunderstanding which is due to the ambiguity of terms such
as "isolation" or "separation." Given his fondness to

compare his own procedure with that of the chemist,11 we

 

might be led to believe that an element of knOwleoge can
occur in separation. Kant, it appears, is aware of this
danger. Here is a remark from his Reply to Eberhard which,
although it concerns the topic of abstraction, in my view
nevertheless pertains to the problem at hand: "One does not
abstract a oooosot as a common predicate, but one abstracts

in the oss of a concept £32m the diversity which falls under

it. The chemists alone are in a position to abstract

A6

something, when they extract a liquid from other materials
in order to have it separately; the philosopher abstracts
from what in a certain use of a concept he does not want to
consider."12

Hence to "isolate" an element of knowledge, for
instance intuition, should not be taken to mean that we come
in possession of it, as does the chemist of an element,
namely as something which can exist independently of the
other element, that is, of concept. In other words: Kant's
method of isolation is supposed to result not in real, but
in L£§£§Q§flgfifl§§i distinctions. These mos; be drawn, as we
have seen, in order not to confound the soooifio epistemic
contributions of each of the elements or sources of
knowledge. It is in this respect, indeed, that "intuition
stands in no need whatsoever of the functions of thought" (A
90—1 = B 123; NKS 12A). But it is ooly in this respect. We
are not permitted, in other words, to do what Kant
apparantly does against his own warning, namely to confuse
the transcendental independence of intuition with its real
independence. And it is this mistake that, in my view, is
responsible for the uncritical claim according to which, as
we have seen, "objects...may appear to us without their
being under the necessity of being related to the functions
of understanding," for the claim, that is, that the notion
of what is subjective [0] does have a legitimate

constitutive employment.

 

 

 

 

“in i -
A7

(5)

It is this same mistake to which, I think, we must
ultimately trace Kant's notorious distinction between
judgments of perception and judgments of experience that he
develops in the #18-20 of the Broiegomeoa, to which he
refers obliquely in #19 of the Transcendental Deduction, and
that he states summarily in #AO of the Logio. Although this
distinction cannot be maintained as it stands, it can
nevertheless be reconstructed along Critical lines, as can
the equally untenable notion of subjective validity which
according to Kant belongs to a judgment of peception.

Both types of judgments are empirical in the sense
that they have their ground in "what is given to sensuous

intuition," that is, "in immediate sense—perception."13 A

 

judgment of experience, however, in contradistinction to a
judgment of perception, is not only empirical in this sense,
but expresses experience in the Critical sense of the term,

that is, "empirical knowledge" (B 165; NKS 173). And this

 

is because in such a judgment the perceptions are subsumed
under a pure concept of the understanding and thus possess
necessary synthetic unity. In a judgment of experience
Kant wants to say, the perceptions are not represented as
they contingently happen to belong to each other in a
subject's perceptual state, "in my state or in several
states of mine, or in the perceptual state of others,"1u

but as they are necessarilLrelei§9-§mgbz_ib§b§elyes. It is

this "necessary connection"15 among the given perceptions

 

us

themselves that, according to Kant, constitutes their
"reference to an object."16
This gives us the opportunity to render the notion
of an object [ct] more precise. Such an object is the
representation, in a pure concept or category, of the
synthetic unity of the given perceptions, that is, of their
necessary relation among themselves. It is, therefore, an
opjgotiyo representation, a "mode of knowledge," or a
oognition. And this representation is not only numerically,
but also in kind, "distinotj from any one of the given
perceptions, or a collection or association thereof. The
given perceptions are sonsiplo representations, their
object, however, consists in their representation in a pure
gonoopt. It is this state of affairs to which, in my view,
Kant refers when in the Transcendental Deduction he says of
perceptions that are represented as they necessarily belong
to one another, that "they are combined io_tno_ooj§ot, no
matter what the state of the subject may be" (5 1A2; NKS
159). And now we can see, too, why for Kant an objective
representation is also a universal representation, 1. e.,
why what he considers in the Prolegomena as the defining
mark of a judgment of experience, namely "objective
validity" and "necessary universality (for everybody)," are
"equivalent terms."17
One of Kant's examples of a judgment of experience
is "The sun warms the stone." Its empirical ground are the

perceptions of the sun and of the warmth of the stone as

 

 

 

 

‘49

they succeed one another "in my subject."18 But it is

only by the synthesis of these perceptions in a pure
concept, 1. e., in the pure concept of cause and effect,
that they are determined as they are necessarily connected

among themselves, and thus, in Kant's words, "changed"19 or

'20 It is for this reason that

"converted into experience.‘
the judgment is not only empirical, but expresses empirical
knowledge, and is hence, not only valid for "my subject,"21
but for all subjects: it "hold[s] good for us and in the

same way for everybody else."22

In my view the most useful and least ambiguous

definition of a judgment of perception is to be found in
Kant's Logio, where he says that such a judgment is about

"23 A judgment of

our "perceptions oos perceptions.
experience, we have said, consists in the representation of
their object, i. e., in the representation of their
necessary synthetic unity. A judgment of perception, Kant
wants to claim, consists in something else, and this
precisely because of what it is a judgment about. "The
consciousness of perceptions," Kant says in the fiofloxiooso,
"relates all representations only to ourselves as
modifications of our state; they are thus, unconnected among
themselves,... constitute no knowledge, and have no relation
to an object."2u And this view, it appears, is reflected

here in the Erolegomoos where Kant says that a judgment of

 

perception requires "no pure concept of the understanding,"

but only "the logical connection of perceptions in a

 

50

thinking subject." 25

Thus in the judgment 'The room is
warm' I "merely compare perceptions and connect them in a
consciousness of my state."26 Such a judgment is,

.27

therefore, only "subjectively valid,‘ 1. e., its validity

is "limited to the subject" who makes it, and "to its state
at a particular time."28 A judgment of experience, by
contrast, is valid, as we have seen, "no matter what the
state of the subject may be."

It has often been remarked and hardly needs to be
repeated that Kant's doctrine of judgments of perception in
the Eoolsgomoos is incompatible with his account of judgment
which he gives in the goioioos. There he says that a
judgment "is nothing but the manner in which given modes of
knowledge are brought to the objective unity of
apperception" (B 141; NKS 159). Clearly, if this is
correct, then a judgment of perception is not a joogmoot.

More important for our purposes, however, is to see that the

 

 

doctrine is in violation with an already familiar claim of
Kant, namely the claim which in the revised version of the
Transcendental Deduction he puts as follOws: "All sensible
intuitions are subject to the categories, as conditions
under which alone their manifold can come together in one
consciousness" (B 1A3; NKS 160). And this means that a
judgment of perception is but an instance of what we have
called 'subjective [o]', something, that is, which cannot be
constituted: a manifold of sensible intuition which stands

together in one consciousness, but which does not possess

51

synthetic unity.

(6)

It is the notion of the subjective [0], then, which
provides us with the clue for the Critical reconstruction of
a judgment of perception: the reconstruction must be based
on the realization that the unity of consciousness, and
whatever is necessary for this unity, cannot be
circumvented.

The central thesis, as we have seen, of the
Transcendental Deduction is that the unity of consciousness
implies a synthesis of representations. But this means that
the order of constitution which we find in the goologomoos
is invalid. For there Kant claims not only that there are
these two kinds of judgments, but moreover that "all our
judgments are at first merely judgments of perception," and
that "we do not til afterwards give [some of] them a new

9

reference (to an object),"2 and so convert them into
judgments of experience. According to the thesis of the
Transcendental Deduction, however, we must say: unless a
manifold of perceptions possesses synthetic unity, there can
be no "unity of consciousness...in the manifold of
perceptions" (A 112; NKS 139). They could not even "enter
the mind or be apprehended" (A 122; NKS 1M5) and hence would
be "nothing at all" (A 120; NKS 1AA). Without synthetic

unity, therefore, they would not only fail to refer to an

object, but would not even oos perceptions be available as

 

 

 

— ‘ '4

52

the topics of a judgment of perception.

But this means that as a matter of transcendental
necessity also a judgment which is about perceptions qua
perceptions only, must be a judgment in the strict Critical
sense of the term (B 141; NKS 159), i. e., a judgment in
which the perceptions are subsumed under a pure concept. It
means, to put it differenty, that also the "consciousness of
perceptions" which, as we have seen, "relates all
representations only to ourselves as modifications of our
state," must be understood in the strict Critical sense: as
a consciousness which requires the pure concepts as the
conditions of what for Kant is the form of all
consciousness, namely its unity.

In order to see what this involves we shall look at
a letter of Kant's to Beck. The problem under consideration

is "the relating of a representation. qua determination of

 

the subject, to an object distinguished from it. by which
means it becomes a cognition and is not merely a feeling."30
Under discussion, in other words, is the claim that if a
representation "is to be a cognition. a rslatigu to
something else (something other than the subject in which
the representation inheres) befits the representatiom
wherey it becomes communicable to other people; for
otherwise it would belong merely to feeling... which in
itself is not communicable."31 With the basic feature of

Kant's answer we are already familiar: the relating of

representations to an object consists. he says, in "the

 

 

grasping (apprahausig) of the given manifold,... its
reception in the unity of consciousness" by means of'a
Synthesis, and thus in the "representation of a
composite."32

I now want to point out the following. If an
object's being "distinguished" from the subject would
consist in its being an 99j§gt_iu_§paua, or if, in other
words, the relation of a representation to "something other
than the subject" would have to be construed as a relation
to such an object, then there could be no cognition of
representations taken qua determinations of the subject
only. and they would therefore not be "communicable."

We have seen, however, that for Kant the relation of
a manifold of representations to an object consists solely
in the representation of their necessary synthetic unity in
a pure conept. And it is the representation of this unity.
i. e., of the object [ct], which is. as I have said,
"distinct“ from the representations themselves. It is this
representation, therefore, and not an object in space, which
constitutes what here in his letter Kant calls the
"ggmggbiug_gtb§r than the subject." But this means that
even representations taken QB; determinations of the subject
only, can have a relation "to an object distinguished from
[them]:" they can be united in a pure concept and thus be
represented as "contained in-a_§iasl§-r§preseatatign" (A 99;
NKS 132). And this representation is not only numerically,

but also in kind distinct from the representations qua

 

 

5M

determinations: it is the representation of an object [ct],
that is, a cognition and hence communicable.

But the notion of an object [ct] does not only align
for representations qua determinations to be "connected in
the object:" they must be connected. Kant stresses this
point in the letter to Herz which we have already mentioned.
Representations, he says, without the functions of the
understanding, would be "nothing to us, that is, not objects
of kugwlggga at all; we should have knowledge neither of

"33

ourselves nor of other things. He stresses, in
particular, that the representations "would not even reach
that unity of consciousness that is necessary for knowledge

"3” And knowledge of

of mysself (as object of inner sense).
my representations qua determinations is, of course, but a
special type of knowledge of myself.

Kant distinguishes here two kinds of empirical
objects: "myself (as object of inner sense)" and "other
things." To the former Kant refers also, e. g., as the
"object of inner perception," or the "object of inner

"35 We are dealing here, in other

empirical intutition.
words, with what in the Palalogisms Kant calls an
empirically "iuugr object," 1. e., an object which "is
represented only in-i&§_tim§:relatipn§" (A 373; NKS 3H8).
Correspondingly, we can take "other things" to fall under
what Kant calls an empirically "aztgrual object," that is,

an object which also is "represented in apgggfl (A 373; NKS

3H8).

 

 

 

   

 

55

It appears that we can now say the following: the
notion of an object [ct] is nautral with respect to, or
ranges over 99th, empirically external and empirically inner
objects. The objectivity [ct] of an object, thus, consists
not in its being represented in space, and the fact that an
object is represented in time only and hence inner, does not
amount to its being 'subjective'.

To the neutrality of the notion of an object [ct]
there corresponds the neutrality of the principle of the
unity of apperception itself: it ranges over auy_giy§n
intuitign. "All my representations in any given intuition
must be subject to that condition under which alone I can
ascribe them to the identical self as my representations" (B
135; NKS 155). The same point is made in #2” of the
Transcendental Deduction where Kant says that "apperception
and its synthetic unity is... far from being identical with
inner sense. The former, as source of all 99m91335193,
applies to the manifold of iutuijigu_ig_gaggral, and in the
guise of the gagaggrias, prior to all sesnsible intuitiom
to objects iu-geu§ral" (B 15h; NKS 166). What the principle
stipulates, in other words, is that a manifold of 99th,
outer and inner intuition, must satisfy the condition under
which alone there can be consciousness of it, namely the
condition of synthetic unity in a pure concept. Otherwise,
the manifold of githgy type would be "merely a blind play of
representations, less even than a dream," that is, "without

an object" (A 112; NKS 139) and hence incommunicable in

 

 

56

principle.

Conversely, however, we can say: if consciousness in
the Critical-transcendental sense, i. e., a consciousness
whose fundamental property, or form, is its unity "in" or
"throughout" a manifold of representations, requires their
unity in an object [ct], then whatever I can be said to be
conscious of, is communicable in principle. And this in
particular if what I am conscious of. is a manifold of inner
intuition only, e. g., representations taken qua

determination of the subject only.

(7)

Throughout these sections of the Prglgggmena the
scope of the Briugiplg, i. e., that it ranges over "the
manifold of intuitions in general," is obscured by Kant's
equivocation on the term 'object'. There is, in my view, no
doubt that when he claims that the given perceptions
acquire, by their subsumption under a pure concept

"reference to an object,"36

he uses the term 'object' not
only in the sense of an object [ct], that is, as denoting a
manifold of any description which has the formal property of
possessing synthetic unity, but also in the sense of an
empirical object which is represented in space.37

This ambiguity reveals itself not only in the way in
which Kant contrasts the expressions 'object' or 'objective'
on the one hand with such phrases as, for example, "my

«38

present state of perception, "a connectiofl of

 

57
perceptions in my mental state,"39 and "a consciousness of
my particular state”0 on the other hand: these states are

all 'inner' states. The ambiguity is also contained in what
Kant says about a certain class of judgments of perception,
i. e., a class whose members cannot be converted into
judgments of experience: e. g. "The room is warm, sugar
sweet, and wormwood bitter."u1 And they cannot be
converted even though a concept of the understanding were
superadded, "because they refer merely to feeling which
everybody knows to be merely subjective and which of course
can never be attributed to the pyjagt and consequently never
become objagtiya."u2 Clearly, Kant is here talking about
an gxterual object. And what he is saying derives from
on§u3 of his views about sensible qualities that in the

Aesthetic we find in The Transcendental Exposition of the

 

Concept of Space, and according to which these qualities
cannot be regarded as "objective determinations" or

"properties of things, but only as changes in the subject"

 

(A 29 = B MS; NKS 73). On this view sensible qualities are
to be treated not as determinations of 'objects,‘ but of the
'subject,' and the sentence, 'The room is warm,‘ can be
rendered epistemologicaly more perscpicuous with the
sentence, 'I have a sensation of warmth in this room.‘

Now, given the expository character of the

glgggggua and their conception as "a sketch and textbook,n

kit-u

it suggests itself to illustratg the objective validity

and necessary universality (for everyone) of a judgment of

my“

58

experience by a judgment which is, as for example, "The sun
warms the stone," about empirically external objects, 1.
e., about objects which are "represented in spagg" (A 373;
NKS 3H8). In contrast to what Kant calls an empirically
"iuugr object," the sun and the stone are, as it were, there
for everyone to see.

But such an illustration obscures not only, as I
have pointed out, the scope of the Briuuiplg, i. e., its
range over "the manifold of intuition in general," but also,
and more importantly so, the precise character of the uuiuy
of the manifold which the Briugiplg requires and hence why
the Principle can have the range it does: It is a unity in
a pure category or concept, that is to say, a unity which
consists entirely in its being tbgugnt. This is one of the
fundamental claims of the Critiqug, and we have to be clear
about what it entails.

To say of a manifold of representations that it

 

possesses synthetic unity or that it is "combined in the
object" (B 130; NKS 151) is to say that its being so
combined is 99} a matter of aauaiuility, but of the
ungarataugiug: "The combination (ggujuuutig) of a manifold
in general can never come to us through the §§u§9§" (B 129;
NKS 151). On the contrary: "All combination...be it a
combination of the manifold of intuition, empirical or
non-empirical, or of various concepts, is an act of the
understanding. To this act the general title 'synthesis'

may be assigned" (8 130; NKS 151). Inigkiug, for Kant then,

 

59

i§_99mpiga&ign. And combination itself is defined as "the
representation of the synthetic unity of the manifold" (B
131; NKS 152), that is, as a judgmeut. In the Bruiagumgua
Kant says: "The business of the senses is to intuit, that of
the understanding to think. But thinking is uniting
representations in one consciousness... The unity of
representations in one consciousness is judgment. Thinkng,
therefore, is the same as judging or referring
representations to judgments in general."u5

But if the unity of a manifold in general is
synthetic, i. e., is something which is, qua unity, due nut
to sensibility but to an act of the understanding and thus
is something which is thuugbt or jugggg, then it belongs to
the very essence of this unity that there can be no
intuition whatsoever of it. "By means of gut§r_§§u§§, a
property of our mind," Kant says in IE§D§Q§99§DLQl
A§§Ib§519. "we represent to ourselves objects as outside um
and all without exception in spaga." By means of "iuuar
sense...the mind intuits itself or its inner state," and
ting is the "form in which the intuition of inner states is
possible" (A 22 f.= B 37 f.; NKS 67; emph. added). But as
something which is essentially thgugbt, synthetic unity
precisely does not admit of this: to be represented, by
means of outer sense, in space, and by means of inner sense,
in time. What only cau be so represented is the manifold in
so far as it is giyeu_iu_i_tuitiuuz not, however. its unity.

Kant thus can talk about "outer intuition" and "inner

 

 

 

 

 

60

intuition" (A 33 f.= B 50 f.; NKS 77)--but it would be
senseless, that is, strictly absurd, to talk about an outer
or inner synthetic unity.“6 In other words: the
distinction between the "inner" and the "outer," and thus
between inner and outer objects of knowledge depends
entirely on the typg of the manifold of intuition that is
either temporal or spatial, or both: but it does nut depend
on the unity of the manifold.

Hence, although the judgment "The sun warms the
stone" is about empirically external objects, 1. e., about a
manifold of representations which are, qua "determinations
of the mind" in time and which "have for their objects"
things in space (A 3n = B 50; NKS 77), the synthetic unity
or objectivity [ct] of these representations, 1. e., the
cognition expressed in the judgment, is nothing in space and

time at all.“7

It is nothing 'outer' and it is, in
particular, nothing 'inner.’ And the same holds, mutatia
mutauuis, for what Kant calls 'judgments of perception:' for
those judgments which according to him are about
representations qua determinations of the mind, i. e., about
representations, which are in time only. Their synthetic
unity is nothing in time either: the cognition, therefore,
of our inner states is not itself an inner state.

Thus it is for the objectivity [ct] of a manifold of
sensible intuition and therefore for the objective validity
and universality (for everybody) of auy empirical judgment

§§r;§;;!_;gmatgrial whether the representations united are

 

 

61

in time only or whether they have, in addition, for their
objects things in space. We can make an even stronger
claim: if we grant Kant that the two "fundamental sources of
the mind" (A 50 = B 7h; NKS 92), namely, sensibility and
understanding, "cannot exchange their functions," if, in
oher ords, "the senses can think nothing" (A 51 = B 75; NKS
93). and if the synthetic unity of a manifold is a unity
that is thgught, then it is iugguggiyaplg that a manifold's
being a manifold of either outer or inner sensible intuition
should in any way respectively either strengthen or weaken
the objectivity [ct] of the judgment.

The fact that the synthetic unity of a manifold, the

 

bearer of objectivity [ct], is a unity in a pure concepm
that is, a function put of sensibility, but of the
understanding, explains not only why it is that the
Epiugipla can range over the manifold of outer and inner

intuition. It is also the reason why, furthermore,

 

objectivity in the Critical sense cuts across those
contrasts that are associated with the 'inner' and 'outer':
the contrasts, for example, between the mental and the
physical, between the private and public. In each of these
cases the objective [ct] is ugithar. Thus it is misleading
to assimilate or to identify the objective [ct] with the
publig. And this not only because the objective [ct] is
nothing spatial or physical, but because we would be at a
loss if within Kant's transcendental inquiry we were to form

the contrasting notion of what is private. What objectivity

62

[ct] alone can be contrasted with is subjectivity [0]. But
this is a purely heuristic notion which signifies something
whose constitution is transcendentally impossible: namely, a
unitary consciousness of a manifold of representations which
does up; possess synthetic unity. Such a manifold would
only be sguagg and not thuught and could, therefore. not be
known. "I would not even be able to know that I have
[it]."u8 That is to say, we are talking about
representations which are not even private in the sense that

in principle only I and nobody else could know them.

(8)

Kant says that in a judgment of perception "I
connect (perceptions) in one consciousness of my state.”9
We know that he is talking about "perceptions qua

perceptions"50 or about perceptions qua "modifications of

our state,"51 that is to say, about a sensible manifold of

 

i599: intuition only. On our reconstruction we know

 

furthermore: it is required by the 'I think' or the
friggipla that the connection of this manifold be ayuubggjg,
which is to say, be thought in a pure concept or "brought to
the objective unity of apperception" and thus judged (B 1A1;
NKS 159). And we know finally. as a consequence of this
that the manifold in question as thought or aa judged, i.
e., its gggujtigu. is objective [ct], that is, constitutes
an objective [ct] representation which is not identical

with, but distinct frOm, the perceptions themselves: it is

63

not something 'inner.'52 And this, as we have seem

because the objectivity [ct] of the representation, 1. e.,
its content as that which is thgught in it, is a function of
the uuugrataugiug and gut of sensibility, and thus incapable
of being represented in time at all. Hence, objectivity
[ct] and distinctness [ct] or, in Kant's repeated phrase
the notion of "an object distinct from our representations"
is neither precluded by their being representations of iuuep
nor does it require that they be representations of putgr
sensible intuition.53
Ultimately, then, it is the satisfaction of the
transcendental condition of the unity of consciousness as
spelled out in the Eriugiple, i. e., the synthetic unity of
a manifold of any description, which accounts for the fact
that the empirical "consciousness of my state"5u is a
cognition or knowledge in the strict Critical sense of the
term. It is, to be sure, knowledge of an inner manifold,
but is, qua_kngwl§uge, precisely nut inner, but objective
[ct] and thus in principle communicable to others. And this
is the reason why any empirical judgment in which such a
cognition is expressed, must have "objective validity" and
its equivalent, "necessary universality (for everybody)."55
"The synthetic unity of consciousness," Kant says, "is...an
objective condition of all knowledge" (B 138; NKS 156) and
hence, also of the knowledge of our inner states (ngugth).

At times Kant refers to an unsynthesized manifold of

inner sensible intuition as being merely "a feeling which

 

 

64

everybody knows to be merely subjective...and which of
course can...never become objective,"56 1. e., as something
which has no cognitive worth: which is, as he says in the
Critique_gf_duggmgut "available for no cognition whatever,
not even for that by which the subject gggnizgs itself."57
This, however, is false. Any manifold, we have shown, which
does not possess synthetic unity, is subjective [o], i. e.,
something of which there can be no consciousness of. But
what can Kant mean by a feeling of which I cannot be
conscious and of which, therefore, I cannot "even know that
I have it?" It is precisely this question which is glossed
over in the letter to Beck where in a curious turn of the
phrase Kant says that an unsynthesized manifold of inner
sensible intuition is a "feeling (of pleasure or
displeasure), which iu_ij§§lt [au_§igh] cannot be
communicated." But within Kant's transcendental inquiry
there is no warrant to talk about 'a feeling in itself.‘ He
can, to be sure, talk about what is a_pcigri and
constitutiya, and about what is empirically constitutgu. He
can, furthermore, talk about what as a matter of principle
cannot be constituted: the subjective [c] which "would be
nothing to me" and whose decisive importance for the
argument of the Transcendental Deduction lies in the fact
that it is the only alternative to the unity of
consciousness. But the legitimate discourse about what

939395 be constituted does not allow us the inference that

there is something QBQQD§§1$B$9§. What I have already

65

pointed out and what I call the 'gritisal_di§iun9tion' says
the following: a manifold of any description either
possesses synthetic unity and is constituted, or it is
"nothing at all" (A 120; NKS 1""). There is thus, no
"feeling in itself." It is either "nothing at all" or I can
be conscious of it. And if I am conscious of it, then the
transcendental condition of the unity of consciousness must
be fulfilled. But then a feeling i§ "available... for
cognition," that is, for the empirical knowledge of myself.
And this knowledge cannot fail to be objective [ct]: it "is

valid (gummuuigablg) for everyone."58

(9)

We are now finally in a position to say in which
sense judgments of perception are 'subjective.' Kant
claims, as we have seen, that they are "merely subjectively
valid," that their "validity is... limited to the subject"
and "to its state at a particular time:" they "hold good
only for us (that is, our subject)." What Kant, according
to our reconstruction, should have said is the following: a
judgment of perception is called 'subjective' for the sole
reason that it is about the sensory state of th9_§ubj§9t who
makes it, in contradistinction to a judgment in which a
subject judges something pupa: than its state. This sense
of 'subjective' we shall henceforth call 'subjective [s].'

But otherwise such a judgment is a judgment [ct] and hence

 

may“

66

has objective validity and necessary universality for
everybody.

The subjective [s], thus, is objective [ct]. In
other words: within the domain of all judgments [ct] the
91£29r9n§13_§p§§if19a of a reconstructed judgment of
perception does not consist in its cognitive status or
value, in its being, as Kant says, "merely subjectively
valid," but exclusively in its mattgr. It is a member of
that class of judgments in which a manifold of iuugr
intuition only is brought to the objective unity of
apperception.This is the kind of manifold we are dealing
with whenever we talk about the subjective [3]. But as
being brought to the unity of apperception, i. e., judged,
such a manifold is, as I have shown, no longer inner but
objective [ct]. And I want to maintain that the subjective
[s] is 93: of two senses in which within the scope of Kant's
Critical inquiry anything can be legitimately called
'subjective.‘ It denotes the subjective insofar as it can
be constituted. The only other legitimate sense is the one
in which space, time, and the categories are respectively
said to be the "subjective conditions of all our intuitions"
(A "9 = B 66; NKS 86) and the "subjective conditions of
thought" (A 89 = B 122; NKS 12"). What is subjective in
this sense is so because it has its ground or "its seat in
the subject only. as the formal character of the subject" (B
u1; NKS 70): it is a_prigri. But it is at the same time

§c§g§gggg§uual, that is, the necessary and universal

67

condition of objectivity [ct].

(10)

I conclude this chapter with a remark on the
unreconstructed subjectivity which, as Kant says in the
frglgggmgua, belongs to a judgment of perception. Under
discussion, in other words, is the correct philosophic
'location,' if any, of the indisputable state of affairs
that the judgment 'I have a feeling of warmth in this room,‘
for example, does not mean that "I or any other person shall
always find it as I do now."59

The decisive point to remember is that a subjective
[s] judgment is, as is any judgment [ct], objective [ct]: it
cannot be made unless the manifold of inner intutition "is
brought to the objective unity of apperception" (B 1A1; NKS
159). But then it is "valid" or "communicable" as is a
judgment whose matter consists in a sensible manifold of
inner aug outer intuition. Only if, therefore, the a_prigri
conditions of the possibility of experience are fulfilled,
are we in a position to discover and to assess py
experience, i. e., a-pgstgrigri, the ampirigal subjectivity
or objectivity of our empirical judgments. That the room
feels warm to me is an empirical fact, and that not
everybody "shall always find it as I do now," is likewise an
empirical fact. Facts like these can be found out "only...
§_99§§§rigri, by means of experience" (A 721 = B 7A9; NKS

581). This, however, is possible only "in accordance with"

68

the a_pgi9§i conditions of the possibility of experience (A
723 = B 751; NKS 583). But it is with these a_pri9ri
conditions that philosophy proper, according to Kant, is
concerned: it is trauagaufigutal (A 56 = B 80 f.; NKS 96).
We will adopt a term from the section on the Amphibqu and
say that the subjective and objective validity of which Kant
talks in the Prolegomena. have no "transcendental" but only
an empirical "lggajign" (A 268 = B 32"; NKS 281) and are
therfore, strictly speaking, of no philosophical interest at
all.

In analogy to a distinction which Kant makes in the
Introduction to the second edition of the Critiqua, we can
talk of the "99mparatiyg" (B 3; NKS A") or relative
empirical objectivity or subjectivity of an empirical
judgment. From this we distinguish its
Critical-transcendental objectivity and intersubjectivity
which are "strigt" (iuigu): the Critical disjunction says
that a manifold of intuition either possesses synthetic
unity or is nothing at all. By contrast all empirical
judgments can be graded or ranked according to their
relative empirical subjectivity and objectivity. The
paradigmatic case of the latter would be an empirically
universal judgment about objects in space, that of the
former a judgment of perception, i. e., a judgment which is
about myself and my state at a particular time.

But even withiu the domain of the empirical

subjective is room for differentiations. "The taste of

 

69

wine" (A 29 = B "5; NKS 73) may be "different for different
men" (A 29 = B AS; NKS 73) and may "appear differently to
every observer" (A 30; NKS 73) and thus be subjective. But
it may also not vary and hence be objective if, let us say,
the community of observers is one of vintners. On the other
hand, the description, in terms of colors, of a landscape by
an acyanobleptic would, according to Kant, be subjective
twice over: because it is a function, as is any such
description, of "the particular constitution of the
sense-organs" (A 29; NKS 74), and because in this case we
are dealing with a defect in this constitution. But
nevertheless: the description could not be made and could
not be found out to be subjective [3] unless it were
objective [ct].

Considerations such as these, however, have only an
empirical, but no "transcendental location" and should,
therefore, not occur on the arguments of the Irauaugugautai
Aesthetic and of the Iransoendental-tosig. Their prooer
location is Kant's own empirical Authrupglugy where, it
appears, he carefully distinguishes
empirical-anthropological questions from transcendental
ones. Thus what would be illegitimate in transcendental
inquiry, is perfectly legitimate in an anthropological one:
to rank, for example, the five bodily senses according to
their relative subjectivity and objectivity. Three of these
senses, i. e., the senses of touch, sight, and hearing "are

more objective than subjective, that is, as empirical

70

intuitign they contribute more to the knowledge of the
external object than they animate the consciousness of the

affected organ."60

The other two, the senses of taste and
smell, "are more subjective than objective, i. e., what we
represent by means of them is suited more for the enjoyment
of the external object than for its cognition. This is the
reason why with regard to the former it is easy to
communicate with others. With respect to the latter,
however, although the external empirical intuition is the
same... the manner in which the subject is affected by it,

can be quite different."61

"*3
.4

NOTES TO CHAPTER II

1Prol. p. M6.

2Haldane, II, p. 64.

30f. A 10M, IMQS 13“; A 491 ::EB 519, NKS 1L39; A 492
: B 520, NKS U40.

”Arnulf Zweig, ed. and trs., Kant_Philosoph1§gl

Corresponden0§r11§9;93, Chicago: The University of“ Chica ago
Press, 1967, p. 15A. Hereafter quoted as Zweig.

Zweig, 153—5“.

 

6Zweig, p. 153.
7r .,
Aweig, p. 153-
0B. J. De Vleescnauwer, Langegygtign
Transcendentai_jg§;§ ;_'Qeuyre de Jfifli, ll, Antwerpenmparis.
'S Gravenhage, 1937, p. 178 my tr fereafter quoted as De
Vleesohauwer.
9H. Vaihinter. COMEEDEfiE-éE-b93§31_5I}L}1_S§£
figiggg_y§§nun§t, ll, Stuttgart Berlin Leipzig: Union
Deutsche Verlagsgesellschaft, 1692 p 120 Eereafter
quoted as Vaihinger.
1O , \r
49329;-9i1§§y§§, P 9!
11Cf. BXXa, NKS 2M; £391., 115; Abbott, 262.
12

1u£_2l., t “9
15££9l., p “6
16££91.. p Ab
172.9i., p “6
18Pr2. . t “9
19

72

Zweig, p. 216.

Zweig, p. 216.
Zweig, p. 153.
u
Zweie, p. 153.
erhe, VI, p. U30; my trs.

_I_“..Q_AI1 p0 “6.

37in other words: Kant uses the term 'object' (a) in

the constitutive sense and (b) as referring to something

constituted, and so blurs the distinction between
transcendental and empirical discourse.

gel” 1)- L17.
39§_2l., p. RS.
£§2l., p. A8.
EAL, p. 117.

£rol., p. 47 note, emph. added.

“3 The competing view is to be found in the second
edition of the Critique where Kant says that "the predicates
of appearance can be ascribed to the object itself, in
relation to our sense, for instance, the red color of the
scent to the rose" (B 70a; NKS 89). Or in the Logic where

 

73

it is asserted that "'The stone [is] warm' is a judgment of
experience" (Logic, #u0). And a mixed view is set forth in
the Critique of Judgment where Kant distinguishes between
objective and subjective sensation. "The green color of the
meadows belongs to objeceixe sensation, as a perception of
an object of sense." A eeejeeeige sensation, by contrast,
is a feeling, that is, a representation which is "referred"
not to the object, but "simply to the subject:" it "must
always remain merely subjective and can constitute
absolutely no representation of an object," and "is
available for no cognition whatever, not even for that by
which the subject £2£Qi§§§ itself" (C.J. ed. Bernard, 49).
I will sflgw that this latter claim is blatantly false.
Ehgl., p. 130.

l“5_Prol., P. 52.

u6It is therefore perhaps misleading to say that
"the content of inner experience" must, in order to be
"knowledge of the self (expressible in judgments)...be
eategeeigeg without being QQJQQLifiigg in the normal sense
(viz., ascribed by outer sense to a spatial object" (Beck,
Essays on Kant and Hume, 48).

u7This is the reason why for Kant the objectivity
[ct] of the judgment 'The body is heavy' can eegee result
from the fact that "the two representations have eleeye been
conjoined in my perception, houevee_e§£ee that perception be
repeated" (B 142; NKS 159 emph. added). The very same point
is made by Heidegger: "The perceptual judgment
(flegeeeeeeegeegeeil) 'every time when...then', does not
gradually change over after a sufficient number of
observations, into the experiental judgment
(fieieegeegeeggeil) 'if...then‘"... "The experiential
judgment demands in itself a new step, another way of
representing the given, that is, in the concept," which is
an "essentially different representation." (flhat ;e_e
Ihih8?. 139).

uszweig, p. 153.

 

“9proi., p. us.

Sofierke, 111, p. 3uu.

51ww, XVIII, p. 386.

52Hence Kant's attempt to a Critical Revision, in
the gegelegieee, of the traditional, broadly Cartesian
notion of 'geee§_pe§eeeeieg' which is notoriously ambiguous.
For our purposes here, we can single out two complementary
claims of Kant's. The first is that "inner experience in
general... or (inner) perception in general... is not to be

 

7n

regarded as empirical knowledge, but as knowledge of the
empirical in general" (A 343 = B 401; NKS 330), that is, of
a given manifold of sensible intuition. Hence 'inner
experience' gee knowledge consists in the synthetic enit ,
by means of a pure concept, of a manifold of ieeeg intuition
in one consciousness. But this knowledge is precisely gee
inner, but objective [ct] and therefore communicable. For
Descartes, on the other hand, it ie inner and hence
incommunicable: because what inner perception in the
broadest sense is claimed to be the perception of, namely
all the "modes of thought" which "reside (and are met with)
in me" (Med. III, HR I, 157), is a manifold which does not
possess synthetic unity. Kant's ether claim is that inner
perception "in the narrowest sense of that term" (A 367; NKS
345), the "perception of the self" which is expressed in the
proposition '1 think' (A 342 = B 400; NKS 329) is nothing
inner either. For the 'I think' is the "mere apperception"
which is "transcendental" and which "serves only to
introduce all our thought, as belonging to consciousness" (A
3A1 = B 399 f.; NKS 329).

53This is the reason why the fieieeeeiee_e§_lgeeli§m
cannoe occur in the Transcendental Deduction and why the
latter is in no need of a supplementation by it. Authors
who think that it needs this supplementation, e.g. Strawson
and Beck, have, in my view, a wrong conception of
objectivity [c] and in particular of distinctness [ct]: in
one way or another they conceive of them in eeeeie; terms.

511E 01., p. 48.

£rol., p. 46.
Brol., p. 47 note.

57J. H. Bernard, ed., Kant's Critique of Judgment,
London: Macmillan and Co., Ltd., 1914, p. 49.

58Zweig, p. 216.
592204., p. 47.
“genie, VI, p. nus.
“Beige. VI, p. 446.

Chapter III
OBJECTS [ct]
"That something happens..."
(A 191 = B 236; NKS 220)
(1)

My discussion in the last chapter of the Beieeiple
resulted in a number of claims the most important of which
can be summed up as follows. I have shown that unity of
consciousness requires objectivity [ct], namely, a given
manifold of intuition of any description which possesses
synthetic unity. The latter, I lave said, consists in the
fact that the manifold is united or thought in a pure
concept of the understanding. This accounts not only for
its objectivity [ct], but also for what is equivalent,
namely its intersubjectivity in priciple. Thus I could
claim, consciousnss is always consciousness of objects [ct]
which are communicable or intersubjective, and this even if
the manifold should be a manifold of iegee intuition only.
The principle of the unity of apperception, I have pointed
out, ranges equally over the manifold of outer e_e inner
intuition: its scope is the manifold of intuition in
general. And I have claimed that the strength of the
principle derives from the fact that the eeiy alternative to
the unity of consciousness is subjectivity [o], i. e., a

manifold which does gee possess synthetic unity. The

 

 

76

constitution of such a manifold I have shown to be
transcendentally impossible. My judgments, in other words,
about what Kant calls empirically "inner objects," are as
objective [ct] and intersubjective as are our judgments
about objects in space. An unsynthesized manifold would, as
Kant says, be "nothing at all" (A 120; NKS 144) or "nothing
to me" (B 132; NKS 153). This led me to formulate the
Critical disjunction according to which a manifold of
intuition either possesses synthetic unity or is nothing at

all.

(2)

But how much of all of this has Kant really
established? Let us look at #17 of the Transcendental
DeduCtion in B wnere Kant makes the following two claims.
He maintains, first, that "insofar as the manifold
representations of intuition are Llifil to us, they are
subject to the... supreme principle of the possibility of
all intuition in its relation to sensibility." This
principle says that "all the manifold of intuition should be
subject to the formal conditions of space and time" (E 136,
NKS 155). Secondly, Kant maintains that insofar as the

leanings in one

representations "must allow of being
consciousness, they are subject to... the supreme principle
of the same possibility, in its relation to understanding"

according to which "all the manifold of intuition should be

subject to conditions of the original svnthetic unity of

 

_"‘"V

77

apperception" (B 136; NKS 155). For this claim, I have
argued at length in the preceding chapter. But it says no
more than that the unity of consciousness in a manifold of
representations requires their synthetic unity and that the
latter is the result of an act of combination which "is an
affair of the understanding alone, which itself is nothing
but the faculty of combining e_eeieei, and therefore of
bringing the manifold of given representations under the
unity of apperception" (B 134 f.; NKS 154).

To these two claims we must now add a third one
which we find in the short section 20 and to which Kant
himself refers as "the transcendental dedvegieee" (B 159;
NKS 170): namely, the claim that the categories whose table
he has established in the Metaphysical Deduction (A 70-83 =
B 95 : 109) in fact ege the conditions of synthetic unity,

i. e., "the conditions under which alone" the manifold of

 

all sensible intuitions "can come together in one

 

consciousness" (B 143; NKS 166). They eee, it is claimed,
"concepts of combination" (B 131; NKS 152), and precisely
for that combination which the ££lfl§lpl§ requires, and
"without which nothing can be thought or known, since the
given representations would not have in common the act of
the apperception 'I think,‘ and so could not be apprehended
together in one self—consciousness" (E 137; NKS 156). But
this claim, it appears, is established by Kant neither in
the Transcendental Deduction of the first nor in the

Transcendental Deduction of the second edition of the

78

In A the claim is simply a consequence of what Kemp

Smith in his Commentary calls the "identification of

apperception with understanding."1 "Ihe_gei§y_ef
eppereepeeee," Kant says, "in relation to the synthesis of

 

iLasinaLien__i§_lhefingerstang’ing. . . I n t h e u n d e r s t a n d i n a
there are then pure e_pe;e£; modes of knowledge which
contain the necessary unity of the pure synthesis of
imagination in respect to all possible appearances. These
are the eeeegegiee, that is, the pure concepts of
understanding" (A 119; NKS 143).

Now Adickes says that in #20 in B we find the
simplest and clearest expression of the deduction which Kant
could achieve.2 And De Vleeschauwer maintains that here
Kant succeeds in demonstrating for the first time "the

"3 The deduction in #20

objectivity of the categories.
"distinguishes itself from the deduction of the first
edition insofar as there the categories are deduced neither
as synthetic forms of apperception (objective deduction) nor
as e_pe;egi forms of objects (subjective deduction). It is
governed by the notion of judgment and by the idea of the
identity of the categorical and the judgmental act. Thus
the metaphysical and the transcendental deduction are now
tightly joined to each other. In 1781 their connection was

quite weak.")4

With this asseSSment, Wolff partially agrees. In

his discussion of #19 and of #20 he says that "Kant

 

 

79

introduces the categories by way of a doctrine of judgment.
As judgment is the assertion of a unity of representations
by means of the copula 'is', the functions of unity
expounded in the Table of Judgments will yield the modes of
combination in general (#19). When applied to a manifold of
sensible intuition, they give the modes of synthesis or
Table of Categories (#20). This way of leading into the
categories efter a discussion of the unity of apperception
is much superior to the original discussion in the
Metaphysical Deduction."5 But Wolff disagrees, and in my
view, rightly so, with both Aidckes and De Vleeschauwer: "It
is still the case, however, that the Tables of Judgments and
Categories are eg_hge at this point in the argument.

Rather, Kant should modify his exposition slightly so that
he claims here merely the exietepee of such tables, and not
yet their possession. In practice, he does just this, for
the second edition Deduction makes no more detailed
reference to particular categories than did the first
edition."6 That is to say: "the objectivity of the
particular categories" is precisely net demonstrated. And
Kemp-Smith says that "the Analytic of Concepts supplies no
proof of the validity of particular categories, but only a
quite general demonstration that forms of unity, such as are
involved in all judgment, are demanded for the possibility
of apperception. The proofs of the indispensableness of
epggifig categories are first given in the Analytic of

"7

Principles. I will now turn to a discussion of the

8O

"Principle of Succession in Time, in accordance with the Law

of Causality" (B 231; NKS 218).

(3)

It is worth our while to begin the discussion of
Kant's Second Analogy with a remark concerning its place in
the main argument of the Analytic. In The Discipline of
Pure Reason, Kant characterizes the way we prove
transcendental synthetic propositions, that is, the
principles of the pure understanding, to the effect that
"our guide is the possibility of experience” (A 783 z B 811;
NKS 621). In particular, Kant says with respect to the
Second Analogy, according to which "all alterations take
place in conformity with the law of the connection of cause
and effect" (B 232; NKS 218), that its "proof proceeds by
showing that experience itself, and therefore the object of
experience, would be impossible without a connection of this
kind" (A 783 : B 811; NKS 621). Whatever ramifications
Kant's proof of the Second Aial0gy has for the "General
Problem of Pure Reason" (B 19; NKS 55), its significance for
his strict epistemological—transcendental inquiry consists
in the assertion that the law of the connection of cause and
effect is a necessary condition for the possibility of
experience.

Now this claim is but a specification of "the
general principle of the analogies" which in its revised

version says that "experience is possible only through the

81

representation of a necessary connection of perceptions" (B
218; NKS 208). This tenet, however, is but a reformulation
of the conclusion of the Transcendental Deduction.
"Experience," Kant maintains, "is knowledge by means of
connected perceptions" (B 161; NKS 171), or "the
thoroughgoing synthetic unity of perceptions is... the form
of experience" (A 110; NKS 138). And this conclusion, it
will be remembered, is held to follow from the 'I think,‘
that is, from the necessary unity of consciousness. The
same point is made by Kant himself at the beginning of the
Analogies: he ties what he is about to say to the main
thread of his argument: "The general principle of the
Analogies," he says, "rests on the necessary unity of
apperception" (B 220; NKS 209). That it so "rests" means
that it, and the three Analogies themselves, are implied by,
and hence necessary for, the unity of apperception.

It is the first-edition version of the Principle of
the Analogy which tells us in some detail in which sense the
Analogies go beyond and deepen, the result of the
Transcendental Deduction: "All appearances are, as regards
their existence, subject a pripri to rules determining their

relation to one another in 939 time" (A 177; NKS 208). That
is to say: the representation of a necessary connection of
perceptions, of which the revised version of the Principle
talks, is a representation, by means of a pripri rules, of a

necessary temporal connection among perceptions, of a

manifold of perceptions "as it exists objectively in time"

82

(B 219; NKS 209). And it is precisely this representation,
we ought to note, in terms of which Kant finally gives the
most explicit answer to the question which he first asks in
his letter to Herz8 and which keeps recurring in the
Analytic, namely, what "we mean by the expression 'an object
of representations'" (A 104, NKS 134; A 129, NKS 149; A 155
= B 194, NKS 192; A 197 : B 242, NKS 224). And the answer

is that to confer "re

IF
W
'rr
4—:
b
b
k-r
b
m
b
b
1U
L;
W
b
k-r

upon our
representations" consists in "subjecting the representations
to a rule, and so in necessitating us to connect them in
some specific manner... as regards their tim§:relatign§" (A
197 = B 242 f.; NKS 224; emph. mine).

And again it is of importance for Kant that we see
how the temporal character of synthetic unity, or of
necessary connection, which here in the AnalOgies comes to
the fore, is related to the Transcendental Deduction and its
central thought, that is, the necessary unity of
apperception. All the manifold of intuition, he says here,
"must, as regards its time-relations, be united in the
original apperception." This, he continues, "is demanded by
the a_pri9ri transcendental unity of apperception, to which
everything that is to belong to my knowledge (that is, to my
unified knowledge), and so can be an object for me, has to
conform" (A 177 : B 220; NKS 209 f.). Now with this claim
we are already acquainted: in the opening sections of the
Transcendental Deduction in B, Kant argues that "the

synthetic unity of consciousness... is a condition under

which every intuition must stand in order to b§g94§_a

1:3

object for me" (B 138; NKS 156). And this tenet Kant makes
immediately explicit as the Principle of Apperception, which
is "the highest principle in the whole sphere of human
knowledge" (B 135; NKS 154), and which says that "all my
representations in any given intuition must be subject to
that condition under which I can ascribe them to the
identical self as my representations, and so can comprehend
them as synthetically combined in one apperception through
the general expression, ';_thihk'" (B 138; NKS 157).

The "condition" which the Principle of Apperception
in general terms refers to, is "synthesis," that is, the
synthesis of representations in "a concept" (B 137; NKS
156), in a category (B 143; NKS 160), or in a concept of an
object in general (A 93 2 B 126; NKS 126). And to establish
in this way synthetic unity of a manifold of representations
is to unite them in one consciousness. Here in the
Analogies of Experience, synthetic unity becomes specified
as the "synthetipmpnity in the time—relations of all
perceptions," and the individual Analogies are said to be
rules of synthesis by which this unity is brought about, i.

7 f. : E

R1

e., "rules of universal time—determination" (A 1

220; NKS 210). Hence the category for example, that is

,_;

under discussion in the Second Analogy is but a particular
instance of that general condition of synthesis of which the
Principle of Apperception talks. Therefore, to prove the

Second AnalOgy, is not gnly to prove that the category of

_.—_._

84

cause and effect is required for the possibility of
experience: that it is, in other words, one of the
indispensable and "fundamental concepts by which we think
objects in general for appearances" (A 111; NKS 138), but
also to prove that this category is, ultimately, a necessary

condition of the unity of consciousness itself.

(4)

The initial and perhpas central difficulty of the
Analogies consists in Kant's notion of the time—order of
representations in apprehension which is "always successive"
(A 198 : B 243; NKS 224), "accidental" (B 219; kKS 209),
"arbitrary" (A 193 : B 238; NKS 222), "undetermined" (A 193

: B 238; NKS 221) and "g Digitiyg" (A 193 = B 238; NKS 221),

but which is ggjegtifieg or rendered "gbjectiye... by
reference to a rule" (A 195 : B 240; NKS 223; emph. added)
or category such that the "time-order" of the

representations is now a "necessary order" (A 200 : B 245;

NKS 226; A 194

I
U"
m
w
KO
H
7‘

(n
m
m
[L

emph. added).

The Lenminué

—— —.

angug of cbjectivation, however, the
notion of a subjective time-order of representations, is
dubious: it is inconsistent with the result of the
Transcendental Deduction. Thus Kant asserts in A that
"according to [the principle of the unity of apperception]
all appearances, without exception, must so enter the mind
or be apprehended, that they conform to the unity of

apperception. Without synthetic unity in their connection,

85

this would be impossible" (A 122; NKS 145). And the
corresponding claim in B is that "all sensible intuitions
are subject to the categories, as conditions under which
alone their manifold can come together in one consciousness"
(B 143; NKS 160). But given this tenet, Kant is faced with
the following dilemma:

(1) If the representations gg conform to the unity of
apperception, i. e., possess synthetic unity and are subject
to the categories, then their order in consiousness can be
no other than an gpjgctiyg [ct], that is, necessary [ct]
order.

(2) If they do 22; conform to the unity of apperception
and do not possess synthetic unity, then they cannot be
"apprehended," "enter," or "come together in" one
consciousness, and ng_tim§;g£g§£ whatsoever can be assigned
to them.

I have shown that a manifold of representations
which stands together in one consciousness but that does not
possess synthetic unity is subjegtiyg [0], that is,
something whose constitution is transcendentally impossible.
And Kant's notion of a subjective time—order of
representations is, it appears, subjective [o]: it is an
order in consciousness, but not a necessary order. But to
give up necessity, that is, synthetic unity, is to give up
everything. The result is £9; one consciousness of a
multiplicity of representations whose time-order would be

subjective instead of objective [ct]. The result, rather,

 

86

is a multiplicity of consciousnesses of one representation
each (of. B 134; NKS 154; of. A 352; NKS 335): and in this

u

case the very notion of a temporal green of representations
no longer applies.

Accordingly, {ant's doctrine of a subjective
time-order of representations in apprehension has to be
abandoned. At first sight this seems to be fatal: it has
been suggested, for example by Wolff, that "the central
insight of the Second Analogy is the distinction between

"9'

subjective and objective timeuorder, and that ”the proof
of the Analogy... demands two contrasting orders, both of
. 1 "1O . . _ _ - .

them cons010ts. Similarly Strawson argues that it is
essential for {ant to distinguish "two sets of relations:
(1) the time-relations between the objects which the
perceptions are to be taken as perceptions of; (2) the
time-relations between the members of the (subjective)

. j -. - , "11 ._ _ .
series of perceptions themselves. Without this
distinction, Strawson claims, Kant could not distinguish
between "objects and perceptions of objects" and "all
attendant notions would collapse too: i. e., the notion of a
subjective or experiential route through an objective world,
the possibility of empirical self-consciousness, the
necessary self—reflexiveness of experience, hence the very

. h . . , 12
notion ox experience itself."

One way out of the dilemma, and one which Kant

 

 

87

himself takes, consists in the introduction of an additional
synthesis which accounts for the subjective time-order of
representations in consciousness. Thus Kant talks about,
for example, the "apprehension" of representations as "their
reception in the synthesis of imagination" (A 190 = B 235;
NKS 219). He mentions a "synthesis of apprehension" (A 192
= B 237; NKS 221) and "a subjective synthesis of
apprehension" (A 195 = B 240; NKS 223). And he refers to
the "synthetic faculty of imagination" (B 233; NKS 218) or
to the "synthesis of the manifold by the imagination" (A 201
= 246; NKS 226). But there are two considerations which
count strongly against Kant's assumption of a second
synthesis.

First: whatever these syntheses elsewhere in the
giggigug are supposed to accomplish, they are definitely no;
claimed to generate a sgbjg_tiy§ time-order of
representations. 0n the contrary, in the Subjective
Deduction in A they form essential moments of the Synthesis
of Recognition in a Concept. In other words: they are
required for the synthetic unity of a manifold of intuition
in a pure concept. Thus Kant says, for instance, that a
manifold of intuition, if "unity of intuition" is to arise
out of it, "must first be run through, and held together."
This "act" he calls the "synthesis of apprehension," and
this synthesis is necessary for a manifold to "be
represented as a manifold, and as contained in_a_§ing;§

representatign" (A 99; NKS 131 f.). And the latter notion

 

Co
(0

is, ironically, precisely the one which in the Second
Analogy Kant needs in order to distinguish between a
succession of representations in apprehension and the
representation of an objective succession.

Second: it is the very point of the Transcendental
Deduction that for a manifold of representations "to come
together in one consciousness" not any synthesis is
required, but their synthesis by means of the categories
"alone" (B 143; NKS 160). In no other way is the unity of
consciousness possible. 0f any other synthesis, therefore,
the following can be said: if it is thought to take place
pgigr to the synhesis by the understanding, we are asked to
entertain the meaningless notion of a time—order of
representations which is not in consciousness. And if it is
thought to take place afteg the synthesis by the
understanding, then the time-order of the representations
will already be a necessary order. In othe words: any other
synthesis, in addition to the synthesis in a pure concept,
is either no solution of the problem at all, or else there

is no problem for it to solve.

The other way out of the quandry derives Ircm a
distinction which Kant draws at the beginning of the Seconc
Analogy. The distinction concerns the notion o; a
representation and is crucial: Kant makes it in order to

answer, finally, the question as tc wn~

91
(1
P:
('

I:
(D

89

Critical-transcendental sense, has to be understood by "an
oject of representations" (A 104; NKS 134).

The distinction under consideration is the one first
formulted by Descartes, who in his Third Meditation

distinguishes between ideas taken "only in themselves."13

as "certain modes of my thoughts,"1u and as unrelated "to

anything beyond,"15 and ideas 'properly so called' which

are, "so to speak, images of the things"16 and which are,
therefore, related to something "else beyond themselves."17
The very same distinction underlies, for example, Kant's
claim in the Transcendental Aesthetic according to which
"all representations. whether they have for their objects
outer things or not, belong, in themselves, as
determinations of the mind, to our inner state" (A 3n = B
50; NKS 77). It is to be found in the body of the Second
Analogy where Kant talks about representations qua
"modifications" (A 197 = B 2A2; NKS 224) or "inner
determinations of the mind" (A 197 = B 242; NKS 224) in
contradistinction to "an object for these representations (A
197 = B 2N2; NKS 22A). And finally we have the distinction
at the beginning of the Second Analogy where Kant refers to
representations GEE "objects of consciousness" and
representations qya_repr§§§ntatig_§. that is, "viewed... in
so far as they stand for an object" (A 190 = B 235; NKS
219).

Of decisive importance now is the interpretation of

Kant's further claim, namely that representations qua

 

90

objects of consciousness "are not in any way distinct from
their apprehension" (A 190 = B 235; NKS 219). It appears
that Paton, for example, and in particular Wolff, are
correct when they take Kant to mean here that
representations qua objects of consciousness do not differ
from their apprehension with respect to their timezgrger.
That is to say: the time-order of representations in
apprehension is their time-order qua objects of
consciousness. Wolff calls this their "subjectiye
t1m§=92ger."18 And from it we can now distinguish their
99J§9$ig§_tim§:9rd§r which, according to Wolff, is "their
order qua representations, which is to say, the order in
objective time of the states or events of which they are

19

representations." It is the time order in which they are
"connected in the object" (A 190 = B 235; NKS 220).

Thus, when Kant says in the proof of the Principle
of the Analogies, for instance, that in apprehension
representations "come together only in accidental order,"
that is, "without necessity determining their connection,"
he is talking about representations qua objects of
consciousness and their subjective time-order. But when he
continues to say that in experience [ct] "the manifold (of
representations) has to be represented... as it exists
objectively in time" (B 219; NKS 209), he refers to
representations gga_repre§entgtign§ and their objective
time-order. The whole point, of course, for Kant's quandry

is the fact that both orders can, in Wolff's words, "exist

 

 

 

91

in consciousness."20 Representations, he says, have a
"double nature." As objects of consciousness "they are
apprehended in a subjective time-order. But as

representations, they form the content of empirical
judgments which assign them to an objective time-order."21

In my opinion, however, this solution does not
survive scrutiny. To be sure, we can talk about the "double
nature" of representations and there is, as we shall see, a
legitimate use for this notion. What is questionable,
however, is the inference to the further claim that they
have this double nature also "in consciousness," a claim
from which, if correct, the doctrine of the two time-orders
"in consciousness" would immediately follow. But the
inference, in my View, is fallacious: it rests on an
unresolved ambiguity of the expression 'in consciousness,‘
an ambiguity of which, it appears, not only Wolff is the
victim, but in the first place, Kant himself.

In keeping with a considerable number of seventeenth
and eighteenth—century theories, I interpret the expression
first in the 99$919£i§§ll§§§§9b¥§i§§1 sense. A
representation, viewed not in its representational function,
but in respect of its existence or the kind of entity it is,
is 'in consciousness' or 'in the mind' because there can be
no representation unless there is a mind which is conscious
of it. "No thought," Descartes says, "can exist apart from
a thing that thinks."22 "Ideas," according to Locke, "are

nothing but perceptions of the mind. "23 And Berkeley

 

states that the "existence of an idea consists in its being
perceived," which is to say that ideas "cannot exist
otherwise than in a mind perceiving them."2

Without a doubt, we must understand Kant's claim in
the same sense, that "in itself" a representation has "no
objective reality," but "exists only in being known" and is
"otherwise nothing at all" (A 120; NKS 144); or his
assertion that "if it were altogether impossible to become
conscious of [representations], this would practically
amount to the admission of their non-existence" (A 108a; NKS
142); and finally, all those passages in which he states
the doctrine of Transcendental Idealism and maintains that
representations "have...no independent existence outside our
thought" (A 491 : B 519; NKS 433) or that they "cannot exist
outside our mind" (A 492: B 520; NKS 440), but only "in us"
(A 372, NKS 347; A 42 : 859; NKS 440).

It seems that Kant here is firmly committed to the
traditional ontOIOgical/metaphysical conception of a
representation being iin consciousness' or 'in the mine.‘

In what follows, iowever, I shall show that this sense is at
variance, if not incompatible, with the transcendental
nature of his argument. He may, B§£A§P§, be allowed to make
use of this sense simply for reasons of exposition, that is,
for an intelligible formulation of the problem of
objectivity [ct]. But where he sets forth the solution in
the Transcendental Deduction and in the Analogies, it can no

longer be taken for granted. In particular the

 

 

 

93

ontological/metaphysical sense of 'in consciousness' cannot
be used to procure representations and their subjective
time-order a 'subrepetitious entry' as it were, into
consciousness, that is, an entry without their conforming to
the condition of its unity.

We find now in the Transcendental Deduction and in
the Analogies a number of passages in which the discussed
sense of 'in consciousness' has not been set completely
aside and still holds its sway. Kant says, for example,
that representations "enter the mind" (A 122, NKS 145),
"stand" (B 132, NKS 153) or "come together in one
consciousness" (B 143; NKS 160), that they are "generated in
the mind successively" (A 190 : B 235; NKS 220) or simply
"in us"(A 197 : B 242; NKS 224). These claims, one might
say, are couched in a more or less metaphorical language and
doubtlessly can be phrased in a less offending way. But one
can also say that here the language betrays a defect in the
substance, that is, in the argumentation, and therefore has
to be looked at without charity.

What is at issue in these passages are not only the
conditions, in Critical-transcendental terms, of experience
or of empirical knowledge, but also of a mind or of
consciousness itself. The transcendental possibility of
both is inextricably linked by the very structure of the
argument: the categories, it is claimed, precisely in so far
as they unite a manifold of representations in an object

[ct], render also the unity of consciousness in these

 

 

 

.d._-—x-. a~a- __x g-a..« V

914

representations first of all possible. And this unity is,
according to Kant, the form of consciousness in general, i.
e., of any consciousness whatever.

But if the transcendental constitution of
consciousness itself is here indeed the issue, then Kant
cannot legitimately at the same time presuppose
'consciousness' or 'mind' in the ontological/metaphysical
sense, and make use of it as something which, for example,
representations "enter" or "in" which they already "are."

On the contrary: we must say that nothing can be said to be
"in consciousness" prior to its constitution, that is,
unless the conditions of its unity are fulfilled. And it is
in the light of these strictures that we should look at the
passages at hand.

Thus Kant says that representations cannot "enter
the mind" unless "they conform to the unity of apperception"
(A 122; NKS 145). The latter clause, clearly, belongs to
the transcendental discourse proper of the Deduction. The
former clause, however, which derives from the
ontological/metaphysical sense of 'in consciousness,‘ is
illegitimate in the present context, which purports to be
transcendental. In fact, the damage in this particular
passage is negligible precisely because of the
Egghéggggental condition which Kant places on the
onological/metaphysical notion of representations entering
the mind. As a consequence this notion dissolves or cancels

itself: what Kant is claiming here is simply that there can

 

 

 

1.. .v 7"w--"V'

95

be no consciousness unless the condition of its unity is
satisfied. The damage is severe, however, when Kant says in
the Second Analogy with respect to the subjective time-order
of representations that they "are generated in the mind
successively" (A 190 z B 235; NKS 220). For Kant has no
right to refer to anything being generated "in the mind" if
the argument concerns the transcendental possibility of a
mind, that is, of its unity, in the first place.

The same objection applies, in my view, to the
introductory passage (A 197 : B 242; NKS 224) to the fourth
proof of the Second Analogy where Kant now explicitely
accounts for the subjective time-order in terms of the
ontological/metaphysical sense of representations being 'in
consciousness.‘ "We have representations in us," he says,
"and we can become conscious of them. But how far this
consciousness may extend...they still remain mere
representations, that is inner

this or_that relatigh_9§ time" (emph. added). Clearly, the

 

subjective time-order of representations in consciousness is
here a corollary of the ontolggigalzgetaphysipal sense of
'in consciousness'. Kant then shifts ground and turns to
the seisiemglgaiaalzirapsnsnjspial question. namely. as to
how it "comes about that we posit an oject for these
representations, and so, in addition to their subjective
reality, as modifications, ascribe to them some mysterious

kind of objective reality." The question is, in other

words, how it is possible for a representation to acquire

96

"objective meaning in addition to its subjective meaning
which belongs to it as determination of the mental state."
And Kant's answer, which no longer will do, is that the
representations, which are "in us" and which as
modifications of the mind stand "in this or that relation of
time," must be subjected to a rule such that they are
"necessitated in a certain order as regards their
time—relations," which is to say, possess an objective [ct]

time—order.

(7)

Although the passage under consideration (A 197 = B
242; NKS 224) is a failure in the sense that Kant does not
succeed to account legitimately for the subjective
time-order of representations 'in consciousness,‘ the
passage is nevertheless of interest precisely because of the
admixture of ontolOgical/metaphysical and
epistemological/transcendental considerations that it
contains. Let us, therefore, examine it further.

I begin by exploring a different and uncritical
answer to the question concerning the "object for these
representations," that is, the answer given by
Transcendental Realism as Kant refers to the metaphysical
doctrine which he claims to be inherent in Descartes's
Empirical Idealism. For the transcendental realist, the
objects corresponding to our representations, Kant says, "in

the manner in which they are represented, as extendec beings

 

97

or as alterations," gg_hay§ an "independent existence
outside our thoughts," that is, they are not "mere
modifications of our sensibility," but "self-subsisting
things" or "things-in-themselves" (A 491 = B 519; NKS 439;
cf. A 369 f.; NKS 346 f.). And furthermore, according to
Kant's account, it is precisely because they "exist
independently of us" and have "an existence by themselves,"
that the objects of our representations are said to be
"outside us" or "external" (A 369; NKS 346). The
representations, on the other hand, have no independent
existence; their "reality... depends on immediate
consciousness" and are therefore "never to be met with save
in us" (A 372; NKS 3A8) and said to "exist only in us" (A A2
: B 59; NKS 82).

According to this view we can now distinguish two
different time-orders. The first is the time-order of the
independently existing objegts or our representations: it is
a time—order "apart from" or "outside us" (A 373; NKS 346).
The other is the time-order "in us" or "in the mind." And
if we now say that for the transcendental realist the domain
of the 'objective' is the domain of the independently real
and external, and by contraSt the domain of the merely
'subjective' the domain of the only dependently real or
inner, then the first time-order will be an objective, and
the second will be a subjeCtive time-order. This, then, is
the historical setting, in outline at least, of Kant's claim

in the passage under discussion, namely that our

 

98

representations stand Qua "inner determinations of our mind
in this or that relation of time:" the setting is an
ontological/metaphysical one, and so is the claim itself.
It is not necessary to rehearse here the various
arguments, both metaphysical and epistemological, that have
been advanced to the effect that within this framework the
expression 'an objeCt corresponding to our representations'
is ultimately incoherent or internally inconsistent; why on
this view the "objective reality" of our representations,
namely that they represent something other than themselves,
is, as Kant himself says, "mysterious," and consequently,

the notion of a representatipn itself. Of importance here

 

is only that for Kant an 'object of representations is not

an independently real object, and that it is not the latter

‘

"which prevents our modes of knowledge from being hapha'ard
and arbitrary" (A 104; NKS 13A) and so makes them gngwlegge,
which is to say, gpjfictiye [ct]. Objectivity [ct], for
Kant, is not grounded in the independently real, but in
necessity. An object [ct], he claims here in the Second
Analogy, consists in the representation, by means of a rule
or category, of the necessary temporal relation among
representations themselves. And the necessity of this
relation is, as is "all necessity, without exception,...
grounded in a transcendental condition" (A 106; NKS 135),
namely in the transcendental unity of apperception.

We see now quite clearly that in the passage under

consideration, Kant is confused about what can be

 

99

legitimately contrasted with what. The contrast of
objectivity [ct]: of the "objective reality," the "objective
meaning," and the necessary and therefore objective
time-order of our representations is not their merely
"subjective reality" or "subjective meaning," or their
subjective time-order, i. e., their standing Qua "inner
determinations of the mind in this or that relation of
time." Objectivity [ct] is a Critical-transcendental notion
and can only be contrasted, as I have shown, with
subjectivity [o]. The subjectivity here, however, which
Kant ascribes to the "reality" and the "meaning" of our
representations is an ontglggigal1ggtaphysigal notion: it
derives solely from the fact that a representation depends,
for its existence or its reality, on a mind or a subjept
that is conscious of it.

What is at issue here, however, is not only Kant's
confusion of contrasts. It appears that I can go one step
further and say that in his Critical~transcendental inquiry
there is no room for this type of subjectivity, because
there is no room for the notion of a representation qua
modification or determination. And if this notion has to be
given up, then we will have to give up the doctrine of the
double nature of representations itself.

To say, in the ontological/metaphysical scheme under
consideration, that representations are, in the existential
sense, mocifications or determinations, is to say that they

necessarily belong to, or inhere in, a sibsta;ge. By the

100

latter we understand, if we abstract from the concurrence of

God, a thing which, in Descartes's words, for example, "so

. . . . . "25
eXists that it needs no other thing to eXist, but on
which for their existence the representations depend: "it is

certain that no thought can exist apart from a thing that
thinks;... no accident without a substance in which to
exist."26

The point here is that on this view, and if we take
the notion of a representation quu modification seriously,
the BELLE of consciousness turns out to be analytic: it is
the consequence of the concept of a thinking substance and
of a representation quq mocification of such a substance.
Representations are 'in the mind' because of their nature
quq mooifications, and to say that a plurality of
representations is in one mind is to say that there exists,
as Kant himself characterizes the view, an "I" in which, as
their "common subject" they "inhere" (A 350; NKS 334).

But then the objection will be: if representations
quu modifications, quq "inner determinations of our mind"
are indeed indispensable for Kant, then his central, and
anti—Cartesian claim according to which the unity of
consciousness is synthetic, is unintelligible, and the
principle of the synthetic unity of consciousness
superfluous: "All uu representations in any given intuition
must be subject to that condition under which alone I can
ascribe them to the identical self as qy repres=ntations,

and so can comprehend them as synthetically contined in one

101

apperception through the general expression, '1 think'" (B
138; NKS 157). But if, on the other hand, we take the
principle seriously, as we certainly must, then we ought to
eliminate from Kant's transcendental argument, as an
extraneous element, the notion of a representation qua
modification and everything that this notion entails: that
our representations, as we have seen, are 'inner' and
'subjective,’ and that they have an 'inner' and 'subjective'
time—order that subsequently becomes objectified (A 197 : B
2A2; NKS 22“).

Now it goes almost without saying that in the
Paralogism Kant denies what the notion of a representation
quu modification implies, namely the view that in his own
words, "everyone must...necessarily regard himself as
substance, and thought as [consisting] only [in] accidents
of his being, determinations of his state" (A 349; NKS 333).
This is not the place to examine in detail Kant's discussion
of the First Paralogism. The suggestion is only that his
attack of the claim that "I, as thinking being (soul), am
substance" (A 3A8; NKS 333) could have included, and quite
obviously so, an attack of the notion of a representation
quu modification. For clearly, this notion makes sense only
in conjunction with this, as Kant calls it, quuutig View.
That is to say: we can apply the notion of a modification
when we talk, dogmatically, for example, about "the real
subject in which thought inheres" (A 350; KKS 334), "the

subject in itself" (A 350; NR8 33A; or about the "l" as

 

"self-subsisting being" or "suusta_g§" (B A07; NKS 369).

But we can no longer apply it when we talk in Critically
correct language, that is, talk for instance about "the
constant logical subject of thought" (A 350; NKS 33A),
"consciousness... as the transcendental subject" (A 350; KKS

33M), the "I of apperception" (B 407; NKS 369), or about the

"I in every act of thought" (B A07; NKS 369).

(8)

I have said above that the initial difficulty of the
Analogies consists in Kant's notion of a suujgutiue
time-order of representations, which is claimed to be in
consciousness and to serve as the basis of objectivation
[ct]. This notion, I have pointed out, is incompatible with
the principle of the synthetic unity of apperception, which
allows only of one, that is, qujeptiye [Ct] time—order of
representations in consciousness. And the notion, as we
have shown, can be saved neither by the introduction of an
additional synthesis, nor by invoking the traditional
doctrine of the double nature of representations.

In my view, the difficulties in the Transcendental
Deduction and here in the Analogies that surround some of
Kant's uses of the terms 'subjective,‘ ‘objective,’ and
their COgnates, result from the fact that the notion of
objectivity [ct] is a uonutitutiyg notion and as such has as
its proper contrast the notion of subjectivity [c] which

signifies something which as a matter of principle qguuqtque

 

 

 

103

constituted. Thus, being constitutive of experience,
objectivity [ct] can be exemplified in experience, that is,
is exemplified in any empirical judgment. For given the
identity of the categories and the logical functions of
judgment in so far as the latter "are employed in the
determination of the manifold of a given intuition" (B 143;
NKS 160), an act of judgment is an act of objectivation
[ct]: it is "nothing but the manner" in which such a
manifold is "brought to the objective unity of apperception"
(B 141; NKS 159) and thus determined. But there can be no

judgments which would exemplify the 'undetegmineg‘ or
subjectivity [o]: in which, let us say, a manifold of
sensible intuition would be brought to the 'subjective unity
of apperception.‘ In the transcendental discourse at hand,
this term is meaningless: it involves, to be sure, not only
a logical, but an epistemic contradiction. Yet Kant
nevertheless uses the term (cf. B 139; NKS 157), and uses it
in contradistinction to the gbjegtiye unity of apperception,
as he correspondingly distinguishes "the objective unity of
representations from the subjective" (E 1M2; NKS 159).

This, it seems, is not a matter of careless terminology on
Kant's part, but is due, rather, to the difficulties which
the exposition of the notion of objectivity [ct] is faced
with, and this precisely because the notion requires that
its only legitimate contraSt, namely subjectivity [0],
cannot be constituted: it is not in the domain of discourse

possible to us.

 

 

104

We can, to be sure, characterize the subjective [o]
from the 'ggtsig§,' as it were. Kant does so himself: he
describes it, in opposition to the "unitary experience" (A
123; NKS 146) in "which representations stand compared and
connected" (A 97; NKS 130), as "a turmoil" of
representations (A 111; NKS 138, my trans.) which fail to
"constitute... knowledge... and consequently would be for us
as good as nothing" (A 111; NKS 138). It is this state of
absolute cognitive blindness to which Kant refers when he
talks about "a blind play of representations, less even than
a dream" (A 112; NKS 139). And we can go further and say
that lack of synthetic unity among

. . . . 2 ., , . . .
implies the 'iragmentation' 7 ox tne "numerical unity oi...

the representations

apperception" ‘A 107; NKS 136) and therefore the
impossibility of selfuconsciousness. Not only, thus, would
my representations fail to "represent objects" other than
myself or to yield "Knowledge of myself (as object of inner
sense)," but I would "not even be able to Know that I have"

38
. . c. . _
representations. But we cannot tell a story from the

'insige' or give a description of what subjectivity [o], i.
e., a ‘turmoil' or 'a blind play of representations' would
be like. For any such description presupposes at least a
unitary consciousness: but if we grant this, then we must
also grant the rest that is necessary for such a
consciousness, that is, the synthetic unity of he
representations.

In other words: if according to #19 of the

105

Transcendental Deduction objectivity [ct] is of necessity
the defining mark of our discourse, that is, of our
thegretigal language, then subjectivity [0] cannot be
expressed in this language. It is this impossibility which
is the reason for the systematic inadequacy and for the
final failure of all our attempts to say anything about, or

to illustrate, the subjective [0]: we cannot, as it were,

step outside the domain of objectivity [ct].

We have come across this state of affairs already in
the previous chapter where I discussed Kant's doctrine of
judgments of perception. Such a judgment is supposed to
serve as a contrast to objectivity [ct] in the sense that it
expresses an unsynthesized manifold of sensible intuition in

one consciousness. But a manifold of this description is

 

not subjective, but subjective [o], and, therefore, strictly
ineffablg. This is of course the reason why from the very
beginning Kant has to admit that the doctrine is
self—defeating: he calls, what by hypothesis is not brought
to the unity of apperception and hence is not a judgment, e.
g., the expression 'The room feels warm to me,‘ by its
correct name, namely a judgment. Our reconstruction of a
judgment of perception, if we recall, consisted in showing
that it is necessarily ijectiye in the

Critical—transcendental, but subjectiye in the empirical

sense or subjective [s]: that the room feels warm to me does

 

106

not mean that, in Kant's words, eyegybgdy else in the room
should "find it as I do now."29

The suggestion now is that we have here an example
of a possible procedure to deal with an item which Kant
considers to be subjective, but wich upon examination turns
out to be subjective [0]: namely to try to relocate it in
the domain of the constituted, that is, of experience [ct].
As such it will have satisfied the transcendental conditions
of the possibility of experience, and therefore be, first of
all, objective [ct]. It remains to be seen, then, in which
empirigal sense of the term the item in question is alsp
subjective: but the point is that in whichever sense it will
turn out to be so, it cannot of course be contrasted with
objectivity [ct], because it is itself objective [ct].

Let us now turn to the subjectiye time-order of
representations "in apprehension" (A 192 = B 237; NKS 237)
which according to Kant is "always successive" (A 189 : B
235; NKS 219; cf. B 225, NKS 213; A 190 = B 235; NKS 220; A
198 : B 243, NKS 224): be it "the apprehension of the
manifold in the appearance of a house" (A 190 : B 235; NKS
220) or that of a ship moving down stream (A 192 : B 237;
NKS 221). The subjective timeuorder or "subjegtiye

successipn" is, therefore, "entirely undetermined, and does
not distinguish one appearance from another," 1. e., "it

does not prove anything as to the manner in which the

manifold is connected ;

b
b
'C
p
W
b
In
»
—8
w
U.)
ll
m
m
to
a»
'7':
(I)
m
m
_l

f.; cf. 8 233, NKS 219; A 190 = B 235, NKS 220; emph.

 

 

107

added). In other words: "the subjective succession by
liéfiii" (A 193 = B 238, NKS 221; emph. added) does not tell
me whether the manifld "in the object" is successive or
coexistent: whether I am presented by an empirical object,
let us say, of the type 'event' or of the type 'thing,‘ In
order to know this, the manifold of apprehension must be
"represented as an object:" it must be brought "under a rule
which distinguishes it from every other apprehension and
necessitates some one particular mode of connection in the
manifold" (A 191 : B 236; NKS 220).

Now, for the sake of the argument, we can grant Kant
the claim that the time-order of a manifold of
representations in apprehension is always successive, yet
must ask why and in which sense this order should also be
bjegtiye. If by the latter we mean 'not objective [ct],'
then the manifold turns out to be subjective [o]: of such a
manifold, however, no consciousness is possible as a matter
of principle. But then I cannot even know that the
time-order of the manifold is one of succession.

Let us, therefore, discuss the grounds for the other
claim, namely that the manifold is spgpessiye. In order to
know this, I must be able to become conscious of it as
successive. This point, it appears, is admitted by Kant,
inadvertently though, when he says that "I am conscious only
that my imagination sets one state before and the other
after, not that one state precedes the other in the object"

(B 233; NKS 219). Such a consciousness, however, requires,

108

as does on Critical-transcendental principles all
consciousness in the strict sense of the term, for its very
possibility the synthesis of the manifold by means of the
categories.

The claim that "it must be possible for the 'I
think" to accompany all my representations" (B 131; NKS 152)
implies that all my representations must allow of being
objectified [ct]. The claim does pg; mean, however, that
all of them must allow of being objectified [ct] in space,
i. e., be representations of an emprically external object.
But it does mean, in conjunction with the further claim,
namely that "all representations, whether they have as their
objects outer things or not, belong, in themselves... to our

inner state" (A 34

I!

B 50; NKS 77), that all aw
representations must allow to be synthesized or connected,

as Kant puts it in the Egglegomena, in '‘a consciousness of

“__—-—__

n30

my particular state. And it is such a consciousness

which is required if I am to find out or to kngw
empirically, as in the case at hand, anything about my
representations "in_th§m§§;ye§" as opposed to the ijegts of
which they are sometimes the representations.

At this point it is of crucial importance to make
EAEilglé the correct transcendental/epistemological Urammar
involved in the expression 'a consciousness of my state.‘
What is brought to the objective unity of apperception and,

therefore, determined and gnpwg, is my empirical self in the

sense that it is predicated and claimed to be true 9

*1)
H'
V‘

 

109

that it apprehends representations successively. This, of
course, is an a_p9§t§pigri claim: what I know a_prigri is
that representations "are in time, and necessarily stand in
time-relations" (A 34 = B 51; NKS 77). That the latter in
apprehension are always relations of succession, is known by
experience, that is, by inner experience. The claim,
therefore, is a contingent claim. Its proper domain is, as
Kant argues in the Bacalpgisms, "empirical psychology" (A
343 = B 401; NKS 330), and it is, as some commentators
maintain, as a matter of fact false.

But in my view the point is that the expression
'spbjggjiye succession' has to be reconstructed as 'the
succession of representations in_§upj§9L§.' The phrase
"subjective succession in apprehension" (A 193 = B 238) is,
in other words, an empirical claim to the effect that
representations in_m§ or in any_§ubj§9t are apprehended
successively. And this order can be contrasted only with
the temporal order of what is not in me, which here can mean
only the temporal order of the empirically-§¥terpal-Qbiests
of which these representations are the representations of.
But to maintain that the time-order of representations is
successive in subjects does not mean that it is therefore
'subjective' in an epistemological sense, i. e., subjective
as opposed to objective [ct]. To draw this conclusion would
mean that, correspondingly, we would have to ground the
'objective' in the object. that is, ground it pntglggigally.

In this case, however. we would have abandoned the Critical

 

 

110

stance.

I am now in a position to state less ambiguously one
of the theses which Kant advances in the Analogies. The
thesis can not be that there is a subjective time—order
which is "rendered objective by reference to a rule" (A 195
: B 240; NKS 223). Given Kant's central claim, that all
consciousness is necessarily consciousness of the objective
[ct], such a time—order is systematically elusive. Rather
we can take Kant to assert the following. We know from
outer experience that the time~order of the objects of our
representations is sometimes one of coexistence, and at
other times one of succession. From inner experience,
however, we know that the time-order of the representations
"in themselves," 1. e., the order of their apprehension ig_a
subject, is always successive. This orcer, hence, is
"entirely undetermined," that is, "§y_it§ei§... does not
prove anything as to the order in which the manifold is
conneCteo in the object" (A 193 : B 238; NKS 221 f.; emph.
added). If experience, therefore, were to be construed
§g4;r§ly in terms of a succession of representations, there
would be no way in which we could make intelligible to
ourselves two types of representations which as a matter ox
fact we go have: namely the representation of a succession
and of a co—existence of representations.

Thus, Kant's disagreement with htse concerns not the

‘

truth of the tenet SCOOPQINL to whicn we zinc in us nothing

"but different perceptions, which succeed each other... ano

A

which are in a perpetual flux and movement."31 Of this,
Kant himself says, "I am conscious" (B 233, NKS 219). The
disagreement, rather, is that for experience "successive
perceptions"32 are not sufficient. "Experience," Kant
says in the Principle of the Analogies, "is possible only
through the representation of a BeC§§§§£l

.— . . ——

perceptions" (A 176 : B 218; NKS 208; emph. added).

(10)

There remains, however, still one sense of
'subjective,‘ which according to some of Kant's commentators
applies precisely to my reconstructed notion of a succession
of representations in a subject. According to this view,
which derives from the second argument for the Second

AnaloBy (A 189—94 : B 2 M-39; NKS 21

\O

-22), a succession of

U)

representations is subjective in the sense that the subject
can determine the order in which the representations succeec
one another. Thus the manifold of the appearance of a
house, for instance, can be represented "either from right
to left or from left to right" (A 192 z B 238‘ NKS 221),
either from top to bottom or from bottom to top. "In the

series of these perceptions," Tant 5 ya, there is "no

{u

determinate order specifying at what point I nust begin to
connect the manifold" (A 193 z B 238, KKS “21). The order,
in other words, in this series is "entirely undetermined" {A

, "altogether arbitrary" Litip.}, and

\u

193 = B 238; iCKS 22

reversible (cf. A 19 f. : B 239; hhS 222}.

L A.)

112

Let us now consider "an appearance which contains a
happening" (A 192 = B 237; NKS 221), e. g., a ship moving
down stream. In this case, Kant claims, "my perception of
its lower position follows upon the perception of its
position higher up in the stream, and it is impossible that
in the apprehension of this appearance the ship should first
be perceived lower down in the stream and afterwards higher
up (A 192 : B 237; NKS 221). If, in other words, we are
dealing with an appearance which contains an objective

succession such that A precedes B, then A and B must be

 

represented in this order. The representation of 5 must
follow the representation of A, and the representation of A
cannot follow upon the representation or B, but can only
precede it. "The order in which the perceptions succeed one
another in apprehension is in this instance determined, anc
to this order apprehension is bound down" (ibid.): it is an
irreversible and a "necessary order" (A 193 : E 238; NKS
221).

In my View, it is unjustified to take Kant to assert
in this passage that the respective properties of the two
orders make the first a subjeptiye and the second an
gbjegtiye [ct] order. This View is expresssed by Gram most
emphatically. "The distinction,” he says, which "Kant makes

between a subjective and objective time order comes to this.

H

The time order in which a manifold s presented is
subjective if I can vary it at will. This is illustrated by

the example of our apprehension of a house... There are

 

113

other sequences, like that of a ship's movement downstream,

which we can allegedly not vary. And the order in which

such a manifold is presented is for Kant objective."33
There is, first of all, no textual evidence for such

a suggestion. Kant says, to be sure, that when we perceive

an event, "we must derive the subjegt'ye_igpges§ipn of

1H

apprehension from the gbjegtiyehsppgessigl of appearances"
(A 193 : B 238; NKS 221). But this does not mean that when
we perceive what is not an event, e. g., a house, where no

such 'derivation' is possible, the order in the succession

is subjective. For neither is Kant's claim here out of the

 

1

ordinary, nor is it a particularly Qgipipal claim. All he
is saying, according to my previous reconstruction, is that
when we experience an event, the order in which the
perceptions follow one another in_£he_§ppjegt is determinec
by the order in which the appearances follow one another in
the object. And this state of affairs does not render the
order in the subject gbjeptiye [ct]. Had Kant thought it
were so, the existence of the Second Analogy would be
unintelligible; he would be committed to the view that for
our knowledge of an objective succession, as opposed to our
knowledge of an objective coexistence, appgehensipn is
sufficient: for in this case the latter "is bound down" (A

192 : B 237; NKS 221), and its order is an irreversible
order.
That an irreversible order is not an objective [ct]

order becomes clear once we examine parts of the present

 

11H

argument itself, and in particular if we look at it as
Kant's answer to Hume. There is no doubt that both must be
taken to agree that we know that a certain event AB occurs:

"I see a ship move down stream." Both differ, or course, in

their age up; of this knowledge. But even here we find at

least partial agreement: they concur in the view that
"nothing is really given us save perception(s)" (A 493 = B
521; NKS 441) "...in this or that relation of time" (A 197 :
B 242; NKS 22“). And it seems that Hume would furthermore
not quarrel with Kant's other claim, namely that in the
experience of an event the order in the succession of our
perceptions is an irreversible order. A crucial step in
Kant's argument, however, consists in the tenet that "the

‘\

objectivenpelation or

appearances...is not determined

through mere perception" (B 23A; NKS 219). And this means

J

that a succession of perceptions, eye; the order in which

2

1H

——.——

they occur is an irreversible order, is not suffipien; for
our knowledge of the event AB. And to say this is to say
that an irreversible order is n_t an objective [ct] order.
It is for my purposes instructive to look at this
matter from a different angle. It appears that in the case
of an event the property Kant assigns to the succession of
perceptions, brings with it, as it were, an illusiph of
objectivity [ct]: its order is determined, necessary, and
irreversible. But we must not overlook the fact that this
order belongs to the succession pgly within the frarework of

empirical realism in whose language a considerable part of

 

115
the argument is stated.“

34 The succession of perceptions in
the subject, Kant says, that is,

in me, is derived from
determined by, the succession
i.

appearances

e., in the course of the ship down stream.
course,

or
of

in the object,

This, of
is an empirical and hence a posteriori claim,
it is based on both,

and
outer and inner experience.

Thus it
presupposes the a_prig§i conditions of the possibility of
experience as already fulfilled.

In other words:
in question is an

the order
experienced or constituted order,

and
therefore cannot be said to be constitutive of
itself.

experience
And indeed, as soon

as we enter the domain
Critical-transcendental inquiry proper

of“ (ant'zs
prigri conditions of the possibility

and investigate the

of
"objective succession of

our knowledge of *
221),

an

appearances" (A 193
nothing

7‘
_

_ s 238; NKS
any longer can be said to be "derived." The
transcendental turn in the present
view, with the

argument occurs,
"otherwise,"

in my
which introduces
"by itself," or in

in the subject is

the claim that
the transgehgeptal sense, the succession
"entirely undetermined" and "altogether
arbitrary" and not prove

anything as

"does
which the

to the manner in
manifold is connected in the object" (A 193 = B
238; NKS 221). In the transcendental sense the succession
Q- . fi' it is a given manifold of
intuition which does not possess synthetic unity,
denies it all reference
still calls it a

correctly

and Kant
to an gbiect. But he
"subjective succession," that is, a

116

succession in a subjggt. All reference to a subject,
however, ought equally to be dropped, since otherwise the
succession as a manifold which does not possess synthetic
unity, according to the argument of the Transcendental

Deduction, turns out to be subjective [o].

(11)

We concluded our discussion of the last of various
senses of 'subjective' with the remark that what gives rise
to the problem of the Second Analogy is the radical
indeterminateness of what Kant calls in contradistinction to

\

"the ablesiiie_supps§aion of

appearances," "the spbjgctiye

 

succession (of representations) in apprehension" (A 193 = B
238; NKS 221), and which on our reconstruction became a
succession of representations in a ‘a subject.‘ This
manifold is held by Kant to be "arbitrary" or "undetermined"
in the sense that "by itself" it "does not distinguish one
appearance from another," or "does not prove anything as to
the manner in which the manifold (of appearnce) is connected
in the object" (A 193 : B 238; KKS 221).

For the intelligibility of the problem, however, it
is not required that the manifold of representation is in
sgggessign. The impression to the contrary derives from
Kant's repeated insistence that the apprehension of a

manifold is "always successive" (cf. A 198 z B 2A ; NKS

LI)

22A). It is sufficient to recognize that the temporal order

which obtains in a manifold of representations, be it one of

117

succession or of simultaneity, is not necessarily the same
as the temporal order which "belongs to the manifold in the
appearnces themselves" (A 190 : B 235; NKS 220). Thus we
might be inclined to think for instance that the
successiveness of a manifold of representations "proves"
that the represented manifold of appearance is in succession
and so constitutes a happening or an event. But mere
successiveness is inconclusive: for if I scan, for example,
"the manifold in the appearance of a house which stands
before me," my representations will succeed one another.

Yet this does not mean "what no one will grant," namely,
that "the manifold (in the appearance) of the house," e. 5.,
its roof and its basement, is "also in itself successive" (A
190 = B 235; NKS 220), that is, an event.

We can render the indeterminateness under discussion
more precisely in the following manner. A manifold of
representations is "entirely undetermined" (A 193 : B 238;
NKS 221) in the sense that "this or that relation of time"
(A 197 = B 242; NKS 224) which in it obtains, "by itself" or
"alone" (B 225; NKS 213) affords no evidence whatsoever as
to whether the manifold of appearance, which the
representations are the representations of, is in itself or
"as object of experience... bassistent or in sunsessipl" C”
225; NKS 213). It is the neutrality with respect to this
alternative, an alternative, that is, between two different
basic types of objeCts, to which Kant refers when he says of

the successiveness of our representations that it does "not

118

distinguish one appearance from another," or "does not prove
anything as to the manner in which the manifold (of
appearance) is connected in the object" (A 199 = B 238; NKS
238).

The epistemological prgplem, then, to which the
Second Analogy addresses itself, is how, in general terms,
an "entirely gndetermined" manifold of representations

becomes degermiheg. Specifically, it is the problem on
which grounds it is to be decided that it represents a
manifold of appearance which is successive, i. e., an
"appearance which contains a happening" (A 192 = B 237; NKS
221) or an event, namely "that something, or some state
which did not previously exist, comes to be" (A 191 : B 236;
NKS 220 f.). Let us, with respect to its §g_ptign, first
look at the short summary Kant gives of it towards the end
of the gripigpe and where in my view he expresses the basic
idea of his argument much more concisely than anywhere in
the AnaIOgy itself. He says: "In the Transcendental
Analytic... we derived the principle that everything which
happens has a cause, from the condition under which alone a
concept of a happening in general is objectively
possible—~namely, by showing that the determination of an
event in time, and therefore the event as belonging to

experience, would be impossible

(I)

ave as standing under such

a dynamical rule" (A 788 2 B 816; NKS 62A). Kant claims

(1)

here, in other words, the fact that our concept of

happening in general is "objectively possible," that is, in

119

the terminology of the Postulates, has "objective reality"
and hence refers to a possible object of experience and not
to "mere fancies" (cf. A 222 : B 269; NKS 241), is subject
to a transcendental condition: namely, as he says in the
same paragraph, to "the synthetic condition of the
possibility of the object in accordance with this concept"
(A 781 = B 815; NKS 624). This condition is the "dynamical
rule" to which Kant alludes, i. e., the principle of
causality. And it renders the object of the concept
possible in so far as it makes the "determination of an
event in time," which is to say, "the experience... of
anything as happening" (A 195 = B 2H0; NKS 223) first of all

possible.

(12)

Our reconstruction of Kant's argument for this claim

is based mainly on the first three proofs of the AnaIOgy.35
I begin by recalling what {ant considers to have
demonstrated "by reference to the appearance" of a ship
which moves down stream (A 192 : B 237; NKS 221), and of a

"house which stands before me" (A 192 : B 235; NKS 220).
With respect to the former Kant says that "the order in
which the perceptions succeed one another in apprehension is
in this instance determined" (A 192 = B 237; NKS 221),
namely by the order of the events of which they are the
perceptions. And with respect to the example of the house,

Kant maintains that in this case there is, by contrast, "in

120

the series of perceptions... no determinate order specifying
at what point I must begin in order to connect the manifold"
(A 193 = B 238; NKS 221).

Commentators have pointed out, however, that these
two claims are only contingent claims. That is to say: "the
physical and physiological setting of a perceptual
situation"36 can be imagined to be such that the,orCer in
the succession of a manifold of representations is not
determined by, or "bound down" (A 192 z B 237; NKS 221) to,
the order of the "appearances in their succession" (A 195 :
B 240; NKS 223), i. e., of the events. Or the perceptual
situation can be conceived to be such that the order in the
manifold of representations is determined, althOUgh what is
represented are appearances which coexist. That a manifold
of representations be order-determined or order-indifferent
is, therefore, not a necessary condition for its
representing, respectively, an objective succession or an
objective coexistence. Nor is it, as I have already shown,
a sufficient condition: "it does not prove anything as to

the manner in which the manifold (of appearance) is

P—o

38;

SRS 222).

R)

connected in the object" (A 193 : B
Let us see now how this state of affairs affects
Kant's argument. Its purpose, I have said, is to show under

which condition an "entirely undetermined" (A 193 = B 2 8;

U)

NKS 221) manifold of representations can be said to
represent an cbjective succession, i. e., constitute the

experience of an event.

(1) Although an objective succession does not imply an
order-determined manifold of representations, the latter
nevertheless, if it is to represent an objective
succession, according to Kant, must at least be assumed or

taken to be order-determined: "For mere succession in my
apprehension, if there be no rule determining the succession
in relation to something that precedes," that is, if the
succession is believed to be order-indifferent, "does not
justify me in assuming any successigh in the object" (A 195
: B 2A0; NKS 223; emph. added).

(2) Let us suppose that the manifold is
order-indifferent and let the representations Ba and hb
represent, respectively, the appearances A and B. Then A
can be represented either before or after B. To be sure,
this is "by itself" (A 193 : B 238; ;NKS 221) not conclusive
proof, i. e., sufficient for the coexistence of A and B, in

which case we would contradiCt our assumption of "a

succession in the object" (A 19

l
U

- D 2H0; NKS 223). BUt

\

1

this assumption is nevertheless unjustified if the manifold
is supposed to be order-indifferent: for there is no basis

for assuming of either appearance, A or E, that it must

('Y‘

occur after, and cannct be coexistent with, the other, tha
is, "that it follows" (A 195 : B 2&0; KKS 223; emph. added).
However, if we take, as Kant says we must, the manifold to

be order-determined, then what is represented by Ra, namely

the appearance A, must precede what is represented by Eb,

the appearance B, which is to say that 3 must follow upon A.

122

(3) Now, to say that B must follow upon and cannot
precede A implies the "recognition" (A 196 = B 2A1; NKS 22A)
of a condition under which alone this temporal relation
between A and B is possible, namely that they stand to each
other in "the relation of cause and effect" (A 202 = B 2A7;
NKS 227). In order that "the objective relation" (B 23A;
NKS 219), Kant says, between the appearances A and B "be
known as determined, the relation between the two... must be
so thought that it is thereby determined as necessary which
of them must be placed before, and which of them after, and
that they cannot be placed in the reverse relation... The
concept which carries with it a necessity of synthetic
unity... (is) in this case the concept of the relatign_9£
gau§§_ang_9££§gt, the former of which determines the latter
in time as its consequence" (B 23“; NKS 219).

(A) I have said earlier that the epistemological problem
to which the Second Analogy addresses itself, results from
the indeterminateness of a manifold of representations: its
being "entirely undetermined" (A 193 = B 238; NKS 221) with
respect to an object of either two basic types, 1. e., an
object of the type 'coexistence' or of the type
'succession.' I can now say: to claim of a manifold of
representations, Rm and Rn, that Rn represents, not an
appearance which coexists with what is represented by Rm.
but an "appearance which contains a happening" (A 192 = B
237; NKS 221), that is, an eyegt, implies the assumption

that what Rn represents, namely the event En, follows upon

and cannot precede the event Em which is represented by Rm.
But to say that the temporal relation between Em and En is
determined and cannot "be placed in the reverse" (B 23A; KKS
219), i. e., that En always or "invariably" (A 193 = B 239;
NKS 222 / A 198 = B 2A3; NKS 225) follows upon Em, is to say
that En "follows according to a rule" (A 195 : B 2A0; NKS
222; A 198 = B 253; NKS 225). This rule is "that the
condition under which" En "invariably and necessarily
follows is to be found in what precedes" it, that is, in Em
(A 220 = B 201; NKS 226). Invariable and necessary
succession in time, however, is the schema of causality (cf.
A 14“ z B 183; NKS 185). Therefore: a representation can be
said to be a representation "of an event (1. e., of anything
as happehigg)...only on (the) assumption” that "appearances
in their succession, that is, as they happen, are determined
by the preceding state" (A 195 : B 240; NKS 222;.

(5) From here let us now look at Kant's answer to Hume.
"The accepted view," Kant says with respect to Hume, "Is
that only through perception and comparison of events
repeatedly following in a uniform manner upon preceding
appearances are we able to discover a rule according to
which certain events always follow upon certain appearances,
and that this is the way in which we are first led to
construct for ourselves the concept of cause" (A 195 : B
240; NKS 223). There is no doubt that Kant agrees with the
first part of this statement which conterns, we can say, the

empirical procedures by which we discover particular causes

12A

for particular events. ”To obtain any knowledge
whatsoever... of special laws concerning those appearances
which are empirically determined,... we must resort to
experience" (B 165; NKS 173). Kant's disagreement is with
the second clause wich is a claim about the formation of the
concept of cause. He argues, as we have seen, for the
following: that a representation Rn of a manifold represents
an event or has "relation to an object" (A 197 : B 242; NKS
22A) of the type succession, is gagigaple only if it stands
under the Rule of the Second Analogy according to which "I
must refer" Rn "necessarily to something else which precedes
it," i. e., to Hm, "and upon which it follows in conformity

9; NKS

U)

with a rule, that is, of necessity" (A 19A : B 2
222). It is precisely this procedure, namely the
application of the category of cause and effect to fin and
Rh, by which an object of the type 'succession,‘ that is, an
event, is first of all corstituted.

This, however, means two things. It means, first,
that the rule of cause and effect, or the concept of cause,

cannot in the proposed manner be derived from experience.

(.o
.J
C)
:3
J

oi the rule, a

(n

11"

s—I

On the contrary: the implicit "reco

f.
u

condition of the synthetic unity of appearances in time, has

been the ground of experience its=lf, and ras therefore
preceded it a_p§igri" (A 196 : L 2A2, 1L3 22A). Arc it

means, second, that a hunean manifold of impressions, since
it does let stand uncer this rule, is ghgepitaple: "bx

itself," 1. 9., without the rule, "it does not prove

125

anything as to the manner in which the manifold is connected
in the object" (A 193 = B 238; NKS 222), namely whether any
one impression is an impression of an eyent in the first
place. Thus the fact that one impression follows
"repeatedly... in a uniform manner upon" (A 196 = B 2A0; NKS
223) some other impression does not preclude the possibility
that the sequence of events which they purport to be the
impressions of, is "a sequence that occur(s) solely in the
imagination" (B 23A; NKS 219), and that in fact they are the
impressions of an object of the type 'coexistence.' "Mere
succession in my apprehension," Kant says, "if there be no
rule determining the succession in relation to something
that precedes, does not justify me in assuming any

succession in the object" (A 195 = B 2A0; NKS 223).

(13)

I have already indicated that Kant places the
Analogies of Experience in the context of the main argument
of the Analytic by claiming that their "general principle...
rests on the necessary unity of apperception" (B 220; NKS
209). We can now go further and say of one of the three
special principles, namely of the principle of causality or
of "the law of the connection of cause and effect" (B 232;
NKS 218) that it likewise so "rests," i. e., is implied by.
and hence a necessary condition of, the unity of
consciousness.

This claim, however, brings us back to the

126

Transcendental Deduction itself. The central idea of the
latter, as expressed in #16, is that the ‘I think' or the
unity of consciousness requires "a synthesis of
representations" (B 133; hKS 153). In #20, Kant arrives at
the conclusion that the categories are the conditions of
this synthesis: "All sensible intuitions are subject to the
categories, as conditions under which alone their manifold
can come together in one consciousness" (B 143; MKS 160).
Now synthetic unity of a manifold of representations of any
description is Kant's most general definition of an object
[ct], i. e., of an object of knowledge. The categories,
therefore, are conditions of knowledge and of objects of
knowledge. Specifically, they are conditions of experience
and of objeCts of experience, if we are dealing with a
manifold of perceptions. "Experience," Kant says in #26
whose title is ‘Transcendental Deduction,‘ "is knowledge by

means of connected perceptions" (B 161; RKS 171}. This

m
(A

tenet, however, stands at the beginning of the Analogies
their principle: "The principle of the analogies is:
Experience is possible only through the representation of a
necessary connection of perceptions" (B 216; HHS 206). The
version in A is more detailed in the sense that it speciries
this necessary connection as a connection in time: f
experience to be possible it is required that "all
appearances are, as regards their existence, subject a

prior; to rules determining their relation to one another in

 

7‘

one time" (A 117; NKS 206). The principle of causality,

Kant claims, is such a rule, and what the Second Analogy
proves is that this rule is an a_p§i9ri condition of the
possibility of experience. He may recall Kant's answer to
Hume: "Recognition of the rule, as a condition of the
synthetic unity of appearances in time, has been the ground
of experience itself, and has therefore preceded it a
Bfilgnl" (A 197 = B 242; NKS 224). It is with the proof of
this claim, then, and not sooner, that the Transcendental
Deduction of a particular category, 1. e., of the category

of cause and effect, finally comes to its conclusion.

 

128

NOTES TO CHAPTER III
1Norman Kemp Smith, A Commentary_t9_£ant1§
'Critique of Purgufleason', New York: The Humanities Press,

1962, p. 258. Hereafter quoted as NKSC.

2 . .. u . -
Erich Adickes, lmmanuel_gan

 

_i§-gritis_d§r_r§iaan

Vernunft, Berlin: Mayer a Mueller, 1889, p. 149

9

JDe Vleeschauwer, III, p. 150, my trs.

u , ii.

De Vleeschauwer, iil, p. 151.

5R0bert Paul Wolff, fiantlsilhegri_nf-n§nial
Actiyity, Cambridge: Harvard University Press, 1963, p 186.

Hereafter quoted as Wolff.
6. n .

HolIf, p. 1oo.

7KKSC, p. 3M3.

8m .
aweig, p. 712.

9Holff, p. 275.
1Owoix‘f, p. 279.
110_ A ,
otrawsou, danis§i2:_§:p§s, p. 12%
12 p , c , n H
Strawson. Basa9§-li-§§o§s. t 124
13. r
haldane, I, p. 159.
u -, - -
1 Haldane, l, p. 160.
15

Haldane, l, p. 159.

f

10. _
haldane, l, p. 159.

17Haldane, I, pp. 159-60.

A,

bWolff, p. 2M2.

1

\C)

Holff, pp. 2&3-HA.

20': NW . F
{\Olli, P. 2000

21Holff, p. 280.

22- . _V ,_
Haldane, ll, p. ct.

129

D
JJohn Locke, An Essay Concerning_fipmah
Understanding, ed. A. S. Pringle-Pattison, Oxford: The
Clarendon Press, 1969, p. 16.
24 , . . . . . . .
Berkeley's Philosophicalmflritings, ed. DaVid L.

Armstrong, London: Collier-hacmillan, Ltd., p. 62.

C?
ZJHaldane, I, p. 239.

26Haldane, II, p. 64.

 

27 . , . - .
I owe this term to Rhona H. KotZin.

Z‘qeig’ p. 135’

29222l., p. 46.

0, ., a M .
3 Prol., p. A8. Cf. "Did the S e or noenigsberg
Have No Dreams?" in Lewis White Beck, 2 aysmpp_ganthang
Hume, New Haven and London: Yale University Press, 1978, pp.
47 f.

m w
Rom

31David Hume,
Selby—Bigpe, Oxford: T

-ti§§ gliflpmsn-

-§ __. _ L
a 1

_ h a ‘14 ed. L. A.
'rendon Press,

tp_s,
960, p. 233.

2. . _
3 _L§§§l§§, P. 253-

33. -_, F“ . _ ,, n

AOLLke S. Gram, iant.-§129lgnil_sn_ _

Bvanston: Northwestern University Press, 1968, p.

gqu. Lewis White Beck, "Six Short Pieces on the
Second Analogy," §§§QY§MPDi£By§_§B§_S§BE." New Haven and
London: Yale University Press, 1978, p. 1A1. Hereafter
quoted as Beck.

? 2i, ,

“ B 232-34. A 189~95 : 8 234—39. A 19u—95 : B
239-A0. am indebted, in part, to Lewis White Beck, "Once
More unto the Breach: Kant's Answer to Hume," in Deck, pp.
130-35.

H 0'1

Wolff, p. 266; cf. beck, pp. 147-50.

Chapter IV
ONE HORLD
"The Other‘s anger is "outside
my experience and belong[s] to
a system which is inaccessible
to me. Sartre, 226
"There is one sinble experience."
(A 110; NKS 138)

At the beginning of the present inquiry I quoted a
passage from Kant's Anthwgpglpgy in which he considers it to
be a "metaphysical" problem "whether I, as a thinking being,
have grounds to assume, besides my own existence, the
existence of a totality of other beings that stand in
community with me, i. e., the existence of what is called ‘a
world'."1 To call this a metaphysical problem is, for
Kant, to say that it is a problem that falls in the domain

of the theoretical or "speculative...enployment of pure

reason" (A 8A1 : B 869; hhS 659) wnich answers the question

U)
U‘.
\
O

C 5.

"What can I know?" (A 805 L ;3, AKS 6

it is, thus, a

3

problem to which the Cgitigp« is supposed to provide the

answer. Let us how conclude this inquiry by asking how some
OI its results pertain to the question at hand. in order to
do so I shall consider two interpretations, A and B, of the
question: the first is Cartesian, the other

Critical—transcendental.

131

A

I will, first, interpret it in
Cartesian sense according to which, as

refers to a thinking substance in

which

its original,

we have seen, it
all

"activites...fallihg under the description of thought,
perception or the mind" reside as its "accidents”2 or
modifications. Of these, and iny of these, such a
substance or subject is said to be immediately conscious.

According to Descartes some of

H

awareness are, so to speak, images of

calls them ‘ideas‘ and bives as

a

man or of a chimera, of heaven,

God." ideas, in other besides

words,

modifications, have representational

a

purport to refer to sOmethihb "beyond

objeCts", that which ”exist outsid

‘17.”
.Lo,

exist 'incependehtiy‘.

with this v’

The problem

in the Iknct that

for knowing that

than themselves. Lant this

philb

follows: "I

that is, of

consequently...i
1::

outside

Li

these immediate

Objects of

the things." He

thonght of a

.33‘

("
._

angel, or (even) of

being mere

function: they

a

..: , , ‘ R : J
themseiVes', to

A .
” " or which

(1)

(f
U}

, it is éFbUcQ,

iCiGz‘l'C

(f.
C
Pf.
'—

grounds

()

SOL.

(‘i
F).
O
:3
O)
1" a
0

PI.

C‘.

n.»

ectfloer

not.”

‘31"

132

Hence the "existence of things outside us...must be accepted
on mere faith" (B XL, NKS 34). And even the causal argument
that infers the existence of external things from "inner
perception, taking the inner perception as the effect of
which something external is the proximate cause" (A 368;
NKS 3A5) is not satisfactory proof. It does not stand up
against the objection that it is "far from certain that, if
the perception exists, there exists also the object
corresponding to it" (A 371; hkS 3A7). And this is because
"the inference from a given effect to a determinate cause is
always uncertain since the effect may be due to more than
one cause. Accordingly, as regards tie relation of the
perception to its cause, it always remains doubtful whether
the cause be internal or external, whether, that is to say,
all the so—called outer perceptions are not a mere play of
our inner sense, or whether they stand in relation to actual
external objects as their cause" (A 360; LES 3A5).

Under the Cartesian interpretation, then, of the
expression 'a thinking being‘, it is only the enistence of

myself qua thinkinb substance that is apodeictically

certain. The existence of "a totality of other beings that
stand in community with me," i. e., of a world," is,
however, in principle open to doubt. ”I myself with all my
representations," hant says, am on the Cartesian view
somethinb that is "inmediately perceived" and whose

"c

existence "does not allow 01 tein. doubted.“ sat with

regard to objects to whicn these representations ptrport to

_I
_\
U ‘1

refer, we must conclude with the Fourth Paralogism "that we
can never, by way of any possible experience, be completely
certain as to their reality" (A 369, NKS 345; A 367, NKS

344).

(2)

This "scandal to philosophy” (EXL; LKS 34) is,
according to Kant, due to a metaphysical doctrine which he
calls Transcendental Realism, "which rebarcs time and space
as something given in themselves, independently of our

sensibility" (A 369; NKS 346). Hence "outer appearances" or

LA)

"the objects of outer sense” (A 71; MRS 347) are likewise
interpreted as "thihpsmin—thenselves, which exist
independently of us and of our sensibility" and in this
sense are said to be "outide us” (A 363; hhS 346. Cf. A 371;
NhS 347). The expressions 'external' and ‘ottside us‘ refer
on the View under consideration to "that as thin

exists apart from us" (A J? ; hhS 348). And it is precisely

LL)

this, according to hant, that is the source of the
epistemological difficulties into which this metaphysical

doctrine "inevitably falls.” for; since "what 's without is

(1

not in me, I cannot encounter it in my apperce"tion, nor

therefore in any perception.” “he latter is, "properly
rebardeo...merely the determination of apperception". In

the strict sense of the term, "I on not, therefore, in a

v . _ . _ ... .. ‘ _- q ‘\ ‘ - - W _. i. ‘ ‘7 A 1 ._ _. ‘ .' I. I: " -\-f '1‘ - I.
pOSltiun to pegpgggy external thin S" .n "be, “he as

—— . - \_.. \ d

P\
U.)
V}

These difficulties can be avoided, Kant claims, if
we adopt the doctrine of Transcendental Idealism for which
he has argued "from the start" (A 370; HHS 3A6), that is,
already in the Transcendental Aesthetic. This doctrine is
that "appearances are to be retarded as beinp, one and all,

representations only, not things in themselves, and that

1)

time and soace ar therefore only sensible forms of our
intuition, and not determinations given as existing by
themselves, nor conditions of objects viewed as thinbs in
themselves" (A 369; NR8 3A5. Cf. D).

On this view, Kant saintains, the ”existence of
matter," or of objeCts that, in Descartes's formulation,
"exist outside me," can be accepted "on the unaided
testimony of our mere selruconsciousness,” (A 370; NKS 3M6),

i. e., without the help of a causal inference. "There

.0
f)
,_

be no question, the argument :oes, "that I am conscious of

O

my representations, these representatiox {no I myself, who

E
w

have the representations, therefore exist. thernal objects
(booieS), however, are here appearntes, and are therefore
nothinb but a species of my representations, the objects 01
which are somethinb only through these representations.
Apart from them they are nothinb. Thus enternal thincs
exist as well as I syself, anc both, indeed, upon the
immediate witness of my sell—consciousness" (A 37C~71; HRS

.Flw :'~:.- t
t: vnamb .‘V -h—l

n
V L.

p—v
b
_
h
._I
’_ I
L.

[-1

3A6~A7). henCe the difference is no

myself that I can be said to be 'ismeciatelv' conscious Cl.

135

This term is now pointless: "the only difference is that the
representation of myself, as the thinking subject, belongs
to inner sense only, while the representations which mark
extended beings belong also to outer sense" (A 370-71; NKS
346-47). This difference. since time and space are,
respectively, the forms of inner and of outer sense, can
also be expressed as follows: the representation of myself
is an empirically 'inner' object in so far as it is
"represented only in its tim§:rglatigp§," whereas the
representation of an extended being constitutes an
empirically 'external' object, 1. e., an object that is also
"represented in_§pa9§." But both, time and space, are
transcendentally ideal: they are not "to be found save in
9§" (A 373; NKS 3A8).

I will not discuss here the various reasons why
Kant's argument fails to refute the Cartesian position: why,
for example, "it refutes," in Kemp Smith's words, "Descartes
only by virtually accepting the still more extreme position
of Berkeley."5 Kant himself came to realize this: we have
the Refutation of the Second Edition and the note pertaining
to it in the Introduction which make pronounced use of the
results of the Analytic, in particular of the Analogies.
But is appears to me, nevertheless, that the doctrine of
Transcendental Idealism is at least sufficient to expose
some of the mistakes involved in the Fourth Paralogism: I
hope that their removal will help us see our original

question in the proper, that is to say, in the

Critical-transcendental perspective.

(4)

Let us, then, examine one claim about "the topic of
the rational dOCtrine or the soul" (A 3M4 — B U02; NKS 330},
the claim that underlies the Fourth ParalOgism as an
unquestioned presupposition and according to which "'I' as
thinking being" or "the soul" am said to be "in relation to
possible objects in space” (A 3AM ; B 402, was 330—31).

lf we interpret, first, the expression 'objects in
space“ in the transcencental—realistic sense ano nean by
'objects‘ tnings—in-themselves, ant by 'space‘ a
determination that exists by Itselt or that "attaches to the
objects them-selves" (A 26 a L #2; ins 71): ii, in other
woros, we taae the expression to refer to "what asmtgiybwiy
itsggf exists apart ir0m us" (A 373, hLS 3ht), then the
claim unuer discussion ttrns out to be a coLhatit assertion.
"ho one," Kant says, "is in a position to tecice what an
unknown object may or nay not be able to to“ (A 392, ths
3J9).

but if we give, insteac, the expression 'objects in
space' a transcencentai-ideaiistic interpretation, then tat
claim of the rational psycholoList LO longer names sense.
For we cannot say that the cOngtlve subject ”is in relation
to possible ochcts -n sgate,” ;i tact: tqlo ‘nterpretation
these objects are mere rep‘esentations ant mete: ‘in us',
amp the a_priggi representations o; .patt ant ti.e ”swell in

137

; NKS 3148) .

LU

us as forms of our sensible intuition" (A 37
"No one,” Kant argues, "coulo dream of holoing that what he
has come to recognize as mere representation, is an outer

cause" (A 390; NKS 58}. The only relations conceivable are

LL)

those that obtain among the items which are iin us', i.e.,
which appear to a subject. These fall for Kant, as we have
seen, unoer the notion of the "empirical object, which is
called an external object if it is representeo in_spa§§, anc
an inner object if it is reprsenteo only in its

tit -§ela‘ipgs” (A 373; NKS gut). Thus Kant says that "in
the connection of experience matter, as substance in the

(field of) appearance, is really Liven to outer sense, Just

as the thinking ‘1‘, also as substance in the (field of)
appearance, 's given to inner sense. Further, appearances

in both fields nust oe connected with each other accoroint
to the rules which this category introouces into that

as of our inner perceptions

O
C‘
r:
O
C
(.1.
(L;
m
(r
S:
(D
F J
'- l

connection of

whereby they constitute one experience" (A

L l
- 7
\n
H-
r
(A
( A‘-
LN
*-
C
k“.
Pk‘

he are how in a position to expose the basis or the

Fourth Paralogisn: the cau al relation observec to obtain

(a

(U

among what Kant calls "empirically external objects" (A

Clllrll‘xlliL

(I
'0
‘—
(1‘.-

‘H o C. 4-- — --«-..'.- .3ar. : .-
u“ I . ‘ . g. ‘3‘ vs:
PR5 Que) lb Cohen t0 Ducal; 3&1}: t--L,'e'-

y; the latter

c.)

subject and 'objects in space'. That is to s
are taken to be the cause of the representations that the

. ' . 4 ‘ ‘ z- I ‘ ~ A r ‘\ I v. '- .-\ u -,- H " t" -~' ~ I- —'
SUUJGLC .Lb bulO LO [la‘y‘c ;L__ ullEfiln. ibis, Cl kOuZ

In
m
Fl
'1
«(1
F.)
*1
(D
(it

the hypostatization of

things in

particular point

to regard then

by themselves

which

thinking subjept_

BEEéEén9§§

m

emphasis

What

added

11 O C

outside

19 I?

- /'

Kant

the philosophical

objects is

relation we observe to obtain

himself as we

stricture: we

the Transcendental

object upon

affected by it

not necessary

FJ.

here; but it

be the same

”‘0
Cit.)

in the chapter

it

we

Anthrphaigsy i

first of all,

not to

E

"we

they exist in

maintains
account of

be

know,

1.1110,

’

to discuss

5

18

reasonable

c- Q
(Au

D§

Aesthtic

themselves.

L;

U)

construed

is

for

the faculty

some of

OH

I"; O W

‘\

1.

CO

the Crit

l

the

consider the questi‘

C

V
..A

CO

13

these objects: they are treated as

Kant is very clear about this

hypostatise outer appearances and come

representations but as things existing

us, with the same quality as that with

, and

here, in other words, is that

our consciousness of empirical

in terms of the caus

between these objects. Lant

not immune from violating his own

instance, at the very begin

the claim that "the eff

of

U)

"5

representation,

the implications of this tenet

(l
,.
CT

seat of then will

('1
(‘t
,._.
,4

to expec‘

the that are under attack

(1)

position

Paralogisn.

139

limits of a transcendental inquiry.

(6)

Such an inquiry is a pripri. Transcendental

(1’)

philosophy, Kant a"s in ortsghritte, is "the doctrine of

the possibility of all g_priori knowledge in general, which
,. . k . . , , 6 , , .._
doctrine is the Critique of Pure Reason'. And aCCOPClnL

f)

to the Qritigue itselI transcendental knowledge is defined
as being "occupied... with the mode of our knowledge of

object in so far as this mode of knowledge is to be possible

§_p§ip§i" (B 25; HRS 59). Leaving the Transcendental

Aesthetic aside, it is occupied, in other words, "with our a
prior; concepts of objects in general."7
(7)

These concepts are not concepts by which we
distinguish, for example, between chairs and tables. This
we do by the respective empirical concepts that are cerivec
from experience, i. e., are éilpelsfipll- An §_pripgi
concept of an object in general is basic: it is required to
confer upon a manifold of sensible intuition those invariant
determinations we think belong to an object as such. And
there are essentially two. Che is that the naniiclo CLS a
relation to an object in the sense of something
"distinguished from the subject,” or of "s nethir;C other
than the subject in which the representation inheres.” inc
other characteristic is that an object in this sense is al;o

1N0

objective or intersubjective, i. e., "valid (communicable)
for everyone."8 The Critical-transcendental notion of an
object does justice to both these features: and object [ct]
consists, we have shown, in the representation of the
necessary synthetic unity of a manifold of sensible
intuition. And it is the categories that impart this unity.
They are, thus, "functions of unity among our
representations" (A 59 = B 93; NKS 105-6), or concepts
"which consist solely in the representation of this
necessary synthetic unity" (A 79 = B 10“; NKS 112). And in
so far as the different categories are representations of
different modes of unity, they can be said to be
constitutive of the different modes in which a manifold of
sensible intuition can become objective [ct] at all. The
categories of relation, for instance, represent synthetic
unity as necessary connection: here we have, as Kant says. a
"synthesis of the manifold so far as its constituents
pessssarilx-9e19pg_t9_99e_angtn§r. as, for example. the
accident to some substance, or the effect to the cause" (B

201; NKS 198, note).

(8)

A transcendental inquiry, thus, is not concerned
with now we arrive empirically at a certain piece of
empirical knowledge that we express in a synthetic a
pggggpigp; judgment such as, for instance, "The body is

heavy." or "The sun warms the stone." Such judgments are of

141

interest only in so far as they raise the question, not 01
the empirical but of the a_p£iggi conditions of their
possibility. Among these conditions are the a_pripri
concepts of objects in general in which, Kant claims, a
‘anifold of sensible intuition is united—~i.e. thought--and
hence known §_pnigri. If, therefore, a nanifold is thought
in the category of substance and accident, for example, the
cognition consists not in our ascription, let us say, of the
predicate 'heavy' to a particular body. What is represented
a_prigri and thus comes to be known is that the sensible
representations are representations of something £§£M§Q§9L
or abiding in general. And similarly, if we represent a
manifold of sensible intuition in the category of cause and
effect, we do he; assign, for instance, causal efficacy to
the sun in relation to the warmth of the stone. The a
prigri cognition is that the representations are
representations of succession, i. e., of something happening

or of an §X§DLo

(9)

We are now in a position to give at le

Q1
0)
(r
'(‘1
Q!
"S
('f
0
*5

our original question a Critical-transcendental
interpretation. in part the question concerns, "the
existence of a totality of Other beings which stand in
community with he." Clearly, if by 'beings' (gesen; we mean
such things as tables, electrons, or flashes cf lightning,

then the question, lie outside the scope of a

U!

 

142

transcendental inquiry and cannot be answered by it: it is

an a_2ri2£i

inquiry. Whether there are tables or lightning
is a question of fact that can be answered only a
postgrigri, that is, by experience. But it can be answered
only "in accordance with" (A 723 : B 751; NKS 583) the a
prior; conditions of the possibility of experience. And
this means tht we cannot have empirical knowledge of a table
or of a flash of lightning unless we know a_prigri,-—i.e. by
the application of a category to manifold of sensible
representations--that the latter are representations of
something permanent or of something that happens. For in
respeCt to this alternative a manifold of sensible intuition
by itself is, as we have seen in the discussion of the
Second Analoev. completely inanisrninals or insiffereni.
The question, therefore, as to the existence of a totality
of other beings can in a Critical-transcendental inquiry
mean only the question whether the categories qpa_a_pripri
concepts of objects in general indeed do effect a

differentiation in the described sense of the éiven manifold

of sensible intuition: a differentiation, that is, into
mooes in which it can become objective or an object for us.

In the Critique this question is the question of the

O
"5
[.4
(‘f

objective validity of the categories. is one thing

to say that these concepts are "represented in the nine

~—I

n
U".
E
U)
.3
v»
u 3
\ ,4
I

completely a pripri" \A 156 : b It is quite

another thing, however, to show that they are not "a mere

play of imagination or of understandinp“ ;r :39 - E 216, LKc

1H3

259), but "relate to an object" and so have "meaning and
significance in respect to it" (A 155 B 194; NKS 192). And
to show this, i. e., the objective validity of the
categories, means proving "their necessary application to
the objects of experience" (A 156 : 195;NKS 193). And this
proof is all that talk about grounds for assuming the
existence of a totality of other beings-~i.e., of a
world--in a Criticalwtranscendental inquiry comes to. This

inquiry is a p igr; and as such conce‘ns not the existence

31

of world, but the universal and necessary conditions for the
constitution of a world in its "forna' aspect"9 in the
first place. And also in the domain of t‘anscencental
constitution in this sense belongs the question not of what
exists but of what we must mean by the notion of existence,

'\

and what we mean by it, will be a gerral Leature of the

I

world. In other words: that the sun warms the stone, for
instance, is not such a leature. But what Kant calls a

principle, 1. e., a synthetic a_p§igri principle of pure

L)
("T

understanding, namel" that "everythinb that happens, th-
is, bebins to be, presupposes something upon which it
follows accordinb to a rule" (A 169; hnS a1b) does express
such a feature. Similarly, it is not a formal determination
of the world that there are tables. The Principle, however,
which says that "that which is bound up with the naterial
conditions of experience, that is, with sensa ion, is

v? ”N, f o .' -’ "- "I F, ‘ e~ -. -' n 'r- - f" .. r ‘
- I.» (.60, th (:35) QOCS state ; iC*ru;c.J.

r\ )
_\
CI"

.... ,..‘ ,ln
EUtLAgi" (n

 

determination of the world.

11m

(10)

In the last section I have shown that part of Kant's
question in the gnthrgpglpgy becomes in his
Critical-transcendental inquiry the question of the
objective validity of the categories. The other part of the
question, which I have so far neglected, concerns the
expression '1 as thinking being'. In what follows I shall
show that it is the very structure of the proof of the
objective validity of the categories, that allows me to give
the expression 'I as a thinking being' a
Critical~transcendental interpretation.

Earlier in my investigation I have said that this
proof, which comprises the arguments of the Transcendental
Deduction and of the Principles chapter, has as its core the
claim that the necessary unity of any finite consciousness
in a manifold of representations of sensible intuition
implies the synthetic unity of these representations by
means of the categories. The central tenet, in other words,
is that the unity of consciousness requires as a matter of
transcendental necessity that the manifold be constituted a;
an object [ct]; as somethin; that possesses "objective

_;h ., _ . "11
alley (Ior eVerybtcy;.

U I

validity and nice

(I)

sary univer
I have argued above that the proof of the objective
validity of the categories issues in the Principles, i. e.,
in the Synthetic Principles of Pure Understanding that

concern the formal determinations of the world. i can now

1M5

say, correspondingly, that Kant's claim that the necessary
unity of consciousness implies a synthesis of
representations (Cf. B 133; NKS 153) and is, therefore,
synthetic, denotes an equally formal, a_prigri and hence
universal and necessary determination of any consciousness
of such a world. And it is this determination that in my
view forms one, and the fundamental one, of two
interpretations to which the expression '1 as a thinking
being' in Kant's Critical-transcendental inquiry is
susceptible.

Kant's claim of the necessary synthetic unity of
consciousness has likewise, as I have already pointed out,
the status of a principle: it is the "Principle of the
necessary unity of apperception" and is indeed "the highest
principle in the whole sphere of human knowledge" (B 135;
NKS 15H)._ And the closest tie between it and the principles
of pure understanding is that at least some of the latter
are claimed to be derived from the former: "the general
principle of the three analogies," Kant says, "rests on the
necessary pnity of apperception" (B 220; NKS 209). It is,
in other words, the very form of consciousness as such, 1.
e., its necessary synthetic unity, which requires that it be
consciousness of an objective [ct] domain, that is, of a
world tha formal features of which are described by the
principles of pure understanding: of objects which possess
extensive and intensive quantity, which are substances that

have changing accidents and that stand in causal interaction

1M6

with other substances.12

This thesis does not amount to a straightforward,
affirmative answer to the question Kant raises in the
Anthropolggy~—namely, whether "I as thinking being, have
grounds to assume, besides my own existence, the existence
of a totality of other beings that stand in community with
me, i. e., the existence of what is called 'a world'."
Kant‘s thesis, rather, undercuts the question by maintaining
that its first term, 'I as thinking being', cannot be taken
for granted, but already presupposes, as the condition of
its very p ssibility, what the question is a question
about~—that is, a 'world‘. For what Kant's thesis says is
that as matter of transcendental necessity there can oe, in
the terminology of the question, no ’I'I unless there is a
'world‘. In Critical~transcendental terms we find this
dependency expressed at the very ceginning of the
Transcendental Deduction: the ”thoroughgoing identity of the

apperception cf manifold union is given in intuituion

contains," Rant says, ”a synthesis of representations" (s
133, NKS 153;. And this Claim, which we save exanined in
detail earlier, is really Lant‘s thesis in its most

rueimentary form. The fact, however, that the identity or

apperception stands under ccncitisn has to be kept in Lint

[u

when in the course or the Analytic the notion o: a syntnek

(II
F"
(’8

of representations is developed. Thus when Kant deternines
naturwa:nl its formal unity as "the synthetic unity or the

manifold of appearances according to rules" ;A 12?; figs 1

I—
fix
\

1117

we have to note that these rules, i. e., the categories,
precisely in so far as they are constitutive of a nature or
of a 'world', are also constitutive of the identity of

apperception itself.

(11)

Let us now modify the original question of the
Anthrgpglggy and take the 'totality of other beings" to
include beings of a particular kind-—houses, for example, or
thinking beings, or electrons. And let us now investigate
what, it will mean, in light of the answer Kant has already
given, to ask whether we have "grounds to assume the
existence" of any such particular being.

Kant's answer, remember, in brief, was the thesis
that the unity of apperception implies an objective [ct]
domain. If explicated, the thesis will "instruct us... as
to what it is that can be known as an object of experience"
(B 165; NKS 173), namely a manifold of sensible intuition
which possesses synthetic unity by means of the categories.
The thesis does not instruct us, however, in regard to the
particular instances of an object of experience: they cannot
"be geriyeg from the categories" (8 165; NKS 173). With
these we "become acquainted" (my trans.) only a_pgst§rigri,
that is, "we must resort to experience" (B165; NKS 173).

Here Kant is saying that the categories are
necessary but not sufficient conditions for actual

experience. They are not sufficient: the categories are

1H8

only "indeterminate concepts of the synthesis of possible
sensations" (A 723 = B 751; NKS 583). Actual experience
requires, in addition, sensation as the material element of
experience. But they are necessary: actual experience
itself is possible only "in accordance with these concepts
as rules of an empirical synthesis" (A 723 : B 751; NKS
583). And this means two things. (1) We can show, on a
prigzi transcendental grounds, that something can not be
known as an object of experience and hence can not be a
member of the objective [ct] domain, by demonstrating that
its concept is not in conformity with the Principle of Pure
Understanding. Thus the referents of what Kant calls 'ideas
of reason' cannot occur in the objective [ct] domain: God,
for example, is according to the First Postulate not
empirically possible. Similarly, according to the Second
Analogy there is no room for miracles in the objective
domain, if by a miracle we mean "an event whose cause cannot

be found in nature."13

(2) We cannot prove, of course, on a
prior: transcendental grounds, what will be known as an
object of experience: the Principles, Kant says, cannot
"exhibit a_pri9;i any one of their concepts." 1. e., the
categories, "in a specific instance; they can only do this a
pgstgrigri, by means of experience" (A 721 = B 7&9; NKS
581). But if a specific instance is experienced, then it
must be experienced "in accordance with" the categories "as

rules of an empirical synthesis" (A 723 = B 751; NKS 583).

And this we must in principle be able to show of every

1N9

specific instance, and to do so amounts to what in Kant's
Critical-transcendental inquiry we have to mean by 'to have
grounds for' or by 'to give an account of' the instance in
question.

Thus we can say that my seeing a house that stands
before me--i. e., something that is relatively enduring or
permanent-~requires, according to the First Analogy, the
category of substance and accident._ And the reason for this
is what I have called the indeterminateness of a manifold of
sensible representations: I have said that it is, with
respect to any one representation a§_giyen, undecidable
whether it is the representation of something that is
permanent or of something that happens. It is the
differentiation of the indifferent given manifold for
example into a "happening in general" (A 788 = B 16; NKS
62A) or into something permanent in general that, according
to Kant, is effected by categories. These are a_nrinni and
tpgngggnggnjal differentiations that make our empinigal
distinctions first of all possible: without them, Kant wants
to say, we could not, empirically, distinguish between the
ship that moves downstream and the house that stands next to
the stream. Hence it is such a_prigri differentiations that
a transgengentgl account of a specific empiringl instance
must be concerned with. And it need be concerned with
nothing more. Thus, for example, what in the case of the
ship comes to be known [ct] is a happening or an alngraxinn

that, according to the Second Analogy, demands the category

150

of cause and effect. Moving ships, in other words, can be
known as objects of experience and therefore are members of

the objective [ct] domain.

(12)

Before we examine the membership of thinking beings,
let us pause briefly and put the problem in context with the
main tenets of the Transcendental Deduction.

There is only gne objective [ct] domain. "There is
one single experience," Kant says, "in which all perceptions
are represented as in thoroughgoing and orderly
connection...When we speak of different experiences, we can
refer only to the various perceptions, all of which, as
such, belong to one and the same general experience" (A 110;
NKS 138). And there can be only one experience because
according to the Principle of the Unity of Apperception,
"all appearances, Hi§b995_§59§DL199. must so enter the mind
or be apprehended, that they conform to the unity of
apperception. Without synthetic unity in their connection,
this would be impossible," i. e., I could not say "that I am
conscious of them" (A 122; NKS 1&5, emphasis added).

This tenet entails that for every putative
empirical concept we must be able to show that the
experience from which it is derived is possible in
accordance with the categories. We might nnt be able to do
so because the concept in question upon closer examination

turns out to be "fictitious" (A 222 = B 269; NKS 2H1). Or

151

the concept might upon inspection turn out to be not
empirical at all, but a transcendental idea of which the
purported reference to an empirical object rests on a
"dialectical illusion" (A 63 = B 88; NKS 100). The same
holds for thinking beings: they either belong to the
objective domain, or else are 'nothing at all'. What we are
not permitted to do is to widen the scope of the objective
domain by adding another "original and primitive concept" (A
82 = B 108), i. e.,.a category to the list we already have.
Such a move is forbidden primarily not because Kant claims
that the list is complete and exhaustive ( A 79 = B 105, NKS
113), but because the objection will be that the experience
in question is the exception which the principle of the
ynilx-9f-app§rcaptign rules out: that appearances can.
therefore, enter the mind witbnn; being subject to the
categories. All we can do is supplement, adjust or refine
some of the membership-criteria for the objective [ct]
domain that we already have._ And we can see Kant himself
doing so: suppose it is argued that electrons can not be
known to exsist because the notion of an electron is
inconsistent with the Second Enstnlgtg according to which
"that which is bound up with the material conditions of
experience, that is, with sensation, is nngnal" (A 218 = B
266; NKS 239). Here is Kant's refinement: "The postulate
bearing on the knowledge of things as BQLQfil does not,
indeed, demand immediate penneptinn (and, therefore,

sensation of which we are conscious) of the object whose

152

existence is to be known. What we do, however, require is
the connection of the object with some actual perception, in
accordance with the analogies of experience, which define
all real connection in an experience in general" (A 225 = B

272; NKS 2u2-u3).

(13)

I have argued above that Kant's reconstruction of a
Cartesian 'thinking being' results, in part, in the notion
of the transcendental unity of apperception. If I now show
that thinking beings can occur in the objective [ct] domain,
I can of course not try to show this for the transcendental
unity of apperception itself. The latter can, to be sure,
be legitimately investigated in a transcendental inquiry,
but cannot itself be an object of what this inquiry tries to
establish the n_npinni conditions of—-that is, empirical
knowledge. To call this unity 'transcendental' is to say
that it is §_nrigni necessary for the possibility of
experience: it cannot, therefore, itself be experienced.1u

In objective [ct] domain, then, we must take the
expression 'a thinking being' to refer to a particular type
of empirical object among other such objects. And this is,
besides the sense 'transcendental unity of apperception',
within the confines of the Critique of Pure Theoretical
Reason, all we can mean by the expression. Following Kant,
we now understand by a thinking being in this, empirigal

sense, (1) a being that has 'thoughts, consciousness,

153

desires etc.," all of which "belong to inner sense" (A 357;
NKS 338-u0), and (2) a being that "as man" is also "an
object of the outer senses" (B “15; NKS 373).

In what follows I shall establish the membership in
the objective [ct] domain first of myself as such a being,
and subsequently of other such beings: I will show, in other
words, how I myself and Others "can be known as ab object of
experience" (B 165; NKS 173). Such a procedure I have
called a 'transcendental account'. It is to be expected
that the transcendental account' of myself as "an object of
the outer senses" (8 N15; NKS 373) will follow the general
lines which I have indicated above. But so will the account
of myself as an object of inner sense. For inner experience
or empirical knowledge of myself has the same requirements
as experience in general--namely, unity of apperception, the
categories, and a given manifold of sensible intuition: "in
order to know ourselves, there is required in addition to
the act of thought, which brings the manifold of every
possible intuition to the unity of apperception. a
determinate mode of intuition, whereby this manifold is
given" (B 157; NKS 168). The "limiting condition" (B 159;
NKS 169), i. e., the fact that in this case the manifold can
be represented only in time as the form of inner sense, may
complicate the account, but will not affect its nature.

Thus the transcendental account, for example of a change of
our mood, will not differ in principle from the account of

the changing positions of the ship. That is to say: in this

154

case, too, we are not concerned with emninigal distinctions
the account of which, according to Kant, belongs to
empirical psychology and to anthropology. We are concerned
not with what Kant calls the "specific instance" (A 721 = B
7H9; NKS 581), which is a_ng§§§rigpi, but with what is a
pninri necessary for it, i. e., with transcendental
differentiations of a sensible manifold of intuition, which
is in this case a manifold of inner intuition only.

To give a transcendental account of Others qna
"object[s] of the outer senses" (B 415; NKS 373) should
likewise pose no particular problem. But the difficulties
appear to be formidable if we attempt to do so with Others
taken as beings that have "thoughts, consciousness, desires,
etc." (A 357; NKS 338-39). Here is the apparent obstacle:
"In the Transcendental Aesthetic we have proved, beyond all
question, that bodies are mere appearances of our outer
sense and not things in themselves. We are therefore
justified in saying that our thinking subject is not
corporeal; in other words, that, inasmuch as is it
represented by us as object of inner sense, it cannot, in so
far as it thinks, be an object of outer sense, that is, an
appearance in space. This is equivalent to saying that
thinking beings, a§_§n9h, can never be found by us among
outer appearances, and that their thoughts, consciousness,
desires, etc., cannot be outwardly intuited. All these
belong to inner sense" (A 357; NKS 338-39). Part of this

passage is, according to Kant, fallacious and the source of

155

a paralogism: of the inference, namely, that we must regard
"souls as quite different entities from their bodies" (A
358; NKS 339). I have chosen this passage because what it
says "seems to be so natural" (A 357; NKS 339) as to give
rise to an argument equally fallacious and for the same
reasons: I cannot, it is said, have grounds to believe that
there are beings other than myself that have "thoughts,
feeling, desire, or resolution," for these are

"never...objects of outer intuition" (A 358; NKS 339).

(14)

Nothing prevents our giving one and the same
transcendental account of both: of Others as objects of the
outer senses and of myself as such an object (cf. B A15; NKS
373). "If I want to determine my location in the world as a
human being," Kant says, I must "view my body in relation to

other bodies outside me."15

My body, as well as the body
of Others, must be considered as what Kant calls
"empirically external objects," or as things which are to be
found in space" (A 372; NKS 348).

Since the focus of much of the subsequent discussion
will be on the categories of relation, I will deal with the
problem at hand in terms of the category of magnitude (A 1&2
= B 181; NKS 183), which I have so far neglected. In
particular, I have not examined the proof of its objective

validity that results in the Principle of the Axioms of

Intuition. This Principle says, as stated in A, that "all

156

appearances are, in their intuition, extensive magnitudes"
(A 162; NKS 197), and in B. that "all intuitions are
extensive magnitudes" (B 202; NKS 197). In what follows I
will not enter into a detailed evaluation of this proof, but
restrict myself to conveying its main idea, and only so far
as it pertains to the problem under investigation.

Kant begins the proof with a reference to the
Transcendental Aesthetic: "Appearances, in their formal
aspect, contain an intuition in space and time, which
conditions them, one and all, a_prigri" (B 202; NKS 197-98).
I will return subsequently to the problem as to the
relationship between Kant's doctrine in the Aesthetic and
his doctrine here in the Analytic. But it seems that Kant
here refers to the claim that space and time are a_prigri
representations that as such necessarily underlie,
repectively, all outer appearances, and all appearances as
their fnnm (cf. A 23-2N : B 38-39; NKS 68; A 30-31 = 8 N6;
NKS 74).

We not only intuit appearances, however. in space
and time, or in time only, but we represent them as
occupying a particular space or time. And it is this, i.
e., the constitution of our "representations of a
determinate space or time" (B 202; NKS 198) that Kant is
concerned with here in the Analytic. At issue, in other
words, is the possibility of those of our empirical
judgments in which we talk, for example, of a thing's size.

of its distance to something else, or of the length of a

157

happening and the frequency in which it occurs. Also at
issue, hence, is the possibility of such judgments in which,
in some fashion or other, we predicate determinate spaces or
times of our pgnies. Thus we say: "The cut in his finger
was deep, and the wound took a week to heal." Or: "I am 20
inches taller than my son," which for Kant is an empirical
judgment about myself and my son qna "objects for the outer
senses" (B 415; NKS 373)-~that is, about our bodies.

Kant is now saying, on the grounds of the results of
the Transcendental Deduction, that the representation of an
appearance as occupying a determinate space or time requires
a synthesis. The appearance "cannot be apprehended, that
is, taken up into empirical consciousness, save through that
synthesis of the manifold whereby the representations of a
determinate space or time are generated" (B 202; NKS 298).
This synthesis, too, is successive: it is one of "part to
part," and as such it involves the repeated position, i. e.,
the addition of parts of space or of time (A 163 = B 203-0“;
NKS 198-99; cf. Rfl. 5726; AA 18, 337). The synthesis
consists, in other words, in the "combination of the
nnmngenenns manifold and consciousness of its synthetic
unity " (B 202-O3; NKS 198; emph. added).

According to the Transcendental Deduction, however,
every synthetic unity requires a pure concept or category in
which a manifold first of all becomes united and hence
representable as an object. And here this means:

"Consciousness of the synthetic unity of the

158

manifold...homogeneous in intuition in general, in so far as
the representation of an object first becomes possible by
means of it, is, however, the concept of a magnitude" (B
203; NKS 198). Every category is, for Kant, a concept of a
unity, i. e., a concept that serves as a rule for, and hence
governs, a particular combination of the manifold. And so
does, Kant maintains, the category of magnitude: the
particular type of synthesis of which it is the rule and
which it governs is "the successive synthesis of part to
part" (A 163 = B 204; NKS 199), the repeated positing of the
same, that is, of the manifold in so far as it is
homogeneous. It is, in other words, the "concept of a
magnitude" in which "the unity of the combination of the
manifold (and) homogeneous is thought" (B 203; NKS 198).

All appearances, Kant has claimed, are intuited as
occupying a determinate space or time. The representation,
however, of a determinate space or time consists in the
representation of the manifold homogeneous in the category
of magnitude. Hence objects, as appearances, "are all
without exception mggnitnn§§, indeed, egggnsiye magnitudes"
(B 203; NKS 198), i. e., magnitudes that are generated by
the successive addition of their parts.

Let us pause here and compare this principle of the
pure understanding with Kant's doctrine in the
Transcendental Aesthetic. There he argues not only for the
claim that space and time are the a_pn19ni forms of

intuition, but also that they are intuitions themselves.

159

What Kant wants to point out, it appears, is that the
representation of space, for example, is not "a general
representation (repreaenta§i9_per_nota§_9gmmnn§§)" or a
-pnsept. but a "sinanlar representation (representation
si_gnlani§)" or an intnitig_:16 a representation, in other
words, "which can be given only through a single object" (A
32 = B 47; NKS 75). And the reason why the representation
of space "is not a discursive or...general concept..., but a
pure intuition" (A 2A-25 = B 39; NKS 69), is that "space is
essentially one" (A 25 = B 39; NKS 69). For we can, Kant
says, "in the first place...represent to ourselves only one
space; and if we speak of diverse spaces, we mean thereby
only parts of one and the same unique space. Secondly,
these parts cannot precede the one all-embracing space, as
being, as it were, constituents out of which it can be
composed; on the contrary, they can be thought only as in
it" (A 25 = B 39; NKS 69).

Commentators have of course observed that it is
difficult to reconcile such a claim with the doctrine of the
Axioms of Intuition. For their principle says that "all
intuitions are extensive magnitudes" (B 202; NKS 197), and
of the notion of an extensive magnitude Kant gives the
following definition: "I entitle a magnitude extensive when
the representation of the parts makes possible, and
therefore necessarily precedes, the representation of the
whole" (A 162 = B 203; NKS 198). And commentators have

said, too, that this discrepancy can be explained by the

160

fact that for expository reasons in the Transcendental
Aesthetic the emphasis must be on the contributions of
sensibility, whereas those of the understanding are
neglected. This is, to be sure, correct, but we must
nevertheless locate the precise problem.

Kant cannot say of the representation of a
particular empirical object, e.g., of a table, that it is an
intuition: if it were, we could be said to know the object
independently of the categories and the unity of
apperceptiohn. What he is saying, of course, is that such a
representation involves both, intuition _gg concept: it
requires, that is, a sensible manifold of intuition, the
forms of intuition, and the categories. All of these we can
now call (adopting Kant's own terminology) 'elements' in
empirical knowledge or in the knowledge of an empirical
object. And let us note further, as we have seen, that the
general procedure by which Kant arrives at these elements is
one of isolatign (cf. A 11 = B 36; NKS 67), and that the
purpose of this procedure is to isolate the elements as they
differ with respect to their origin, their 99gnitiye_natur§
ang_fgngtign. Thus when here in the Transcendental
Aesthetic Kant calls the representation of space "a
necessary a_pri9ri representation" (A 24 = B 38; NKS 68)
and, as we have seen, a (pure) intuition, part of what he
means is this: that this representation is subjective in the

sense that it is due to the cognitive subject, that it is an

"objective" representation (of. A 28 = B 39; NKS 69) insofar

161

as it is constitutive of an object [ct]--and that in
contradistinction to other, equally constitutive
representations--it is not a concept, but an intnitign.
What I suggest here is that it is both necessary and
legitimate to treat, for example, of the representation of
space 'in isolation': provided, that is, we keep in mind
that the claim according to which this representation is an
intuition can then be only a claim about an element in our
knowledge of an object [ct]. On Critical grounds it cannot
be a claim to the effect that we know space by intuition
alone.

It appears that on the whole Kant is in the
Transcendental Analytic more aware of the crucial difference
between an element that is constitutive of empirical
knowledge, and such knowledge itself, than he is in the
Transcendental Aesthetic. For this difference can be seen
to be expressed in all those numerous passages in which it
is argued that the categories by_tnem§§ly§§ or in
'isolation' "are nothing but fgrm§_9£_tn9ngnt" (A 248 = B
305; NKS 266), or that pure categories, i. e., "the
categories, apart from the condition of sensible intuition,
of which they contain the synthesis, have no relation to any
determinate object" (A 2H6; NKS 264). In the Transcendental
Aesthetic, by contrast, the corresponding emphasis with
respect to the forms of intuition appears to be lacking.

And Kant's claim that the representation of space is

an intutition is ambiguous. If it is, as I have indicated,

162

a claim about the epistemic nature of an element in
empirical knowledge, then Kant is undoubtedly entitled to
it. If, however, it is a claim to the effect that our
kngylggge of space consists in an intuition, then we are
entertaining a view as uncritical as the view that the
representation of a table is an intuition. And this second
interpretation suggests itself by the fact that space is
said to be represented as "one" or as "unique" (A 25 = B 39;
NKS 69)._ According to the Transcendental Deduction,
however, the unity of space is not given in intuition, but
thought in the pure concepts of the understanding. And Kant
does correct himself: "Space represented as
99399;...contains more than mere form of intuition; it also
contains ggmbinatign of the manifold, given according to the
form of sensibility, in an intnitiye representation" (B
160n; NKS 170; Kant's emph.) Clearly, here in the
Transcendental Deduction Kant wants to stress two things:
(1) that the representation of space as an object requires,
as does the representation of any object [ct], the necessary
synthetic unity of a manifold by means of a pure concept;
and (2) that the representation of space, taken in isolation
or as element only, is not a conceptual but an intuitive
representation.

We must side, then, with Kant's view in the
Analytic, and in particular adopt his view expressed in the
Axioms. There he has argued, as we have seen, that the

representation of a determinate space is not a matter of

163

intuition alone: he has argued, in other words, that the
sensible manifold as given in intuition is nngetermineg with
respect to extensive magnitude. Getting back to the
original problem in this section, I can now conclude the
following. The judgment 'I am 20 inches taller than my son'
is an empirical judgment about myself and my son gna
"objects of the outer senses" (B A15; NKS 373), that is,
about our bodies. This empirigal judgment requires,
according to the Principle of the Axioms of Intuition, the
category of magnitude: and what in terms of the latter first
of all comes to be determined and hence known [ct] are
extensive magnitudes. My body, therefore, and my son's
body, can be known as objects of experience and hence are

members of the objective [ct] domain.

(15)

In the chapter on Subjectivity I have shown that
inner experience, too, is subject to the principle of the
necessary synthetic unity of apperception; the latter ranges
not only over the manifold of outer intuition, but over
"representations in any given intuition" (B 138; NKS 157;
emph. added). Thus inner experience consists, as does outer
experience, in the representation of an object [ct]: it is
expressed, in other words, in a judgment in the strict sense
of the term (cf. B 1A1 f.; NKS 159). in a judgment,
therefore, that possesses "objective validity and necessary

universality (for everybody)".17 And this means that my

16R

judgments about myself as a being that has "thoughts,
consciousness, desires, etc.," all of which "belong to inner
sense" (A 357; NKS 338 f.), are made in accordance with the
categories. I can know myself, therefore, as an object of
inner experience, and as such an object I am a member of the
objective [ct] domain.

The fact that the principle of apperception, the
"highest principle in the whole sphere of human knowledge"
(B 135; NKS 15"), applies to both outer and inner intuition
means that the particular principles govern equally both
outer and inner experience. I have argued in the last
section that the judgment "The cut in my finger was deep,
and the wound took a week to heal," requires, according to
the Principle of the Axioms of Intuition, the category of
quantity. And the same holds for the judgment "The pain in
my finger lasted one hour," which is about an empirically
inngr object and expresses the consciousness of my
particular state: what by means of the category is here
first of all represented and hence known [ct] is a
"determinate time-magnitude" (A 153 = B 203; NKS 198).
Similarly Kant claims that the category of cause and effect
unites not only a manifold of outer but also of inner
intuition: what in our awareness of a change in our mood,
for example, comes to be known [ct] is an inner happening or
an "inner alteration," i. e., "the successive existence of
ourselves in different states" (B 292; NKS 255). Such a

succession, too, is an objective [ct'] that is, a causal

165

succession. Inner alterations, in other words, must be seen
as falling under the principle according to which "all
alterations take place in conformity with the law of the
connection of cause and effect" (B 232; NKS 218).

My emphasis on the fact that the particular
Principles apply to both outer and inner experience suggests
concluding this section with two observations. The first is
about Kant's view of the status of empirical psychology, the
other will be a comparison of Strawson and Kant.

(1) It appears that the "System of the Principles of
the Pure Understanding" (A 150 = B 189; NKS 189) must be
taken to be coextensive with what in the Introduction to the
QIiLiQBS and in the frglgggmgna Kant calls "the pure part of
natural science" (B 18; NKS 5A) or the "pure" and "universal
science of nature" (B 20; NKS 56).18 This science is said
to be pure in the sense that the Principles are synthetic a
prinri, and strictly uniyersal in the sense that it "must
bring nature in general, whether it regards the object of
the outer or that of the inner sense (the object of physics

19 On this

as well as psychology)" under these principles.
view, then, empirical psychology is on par with physics: it
too is, and for the very same reasons, a genuine natural
science. Here is what Kant says at the beginning of the
Paralogisms: "If our knowledge of thinking beings in
general... were based on more than the eggitg, if we

likewise made use of observations concerning the play of our

thoughts and the natural laws of the thinking self to be

166

derived from these thoughts, there would arise an empirical
psychology, which would be a kind of physiglggy of inner
sense, capable perhps of explaining the appearances of inner
sense" (A 347 = B NOS; NKS 332). This passage has a
cautionary note that I think reflects Kant's other view of
empirical psychology: the view that he expresses in the
Anfangsgrnende and according to which empirical psychology
can never attain the status of a genuine natural science,
because mathematics is not applicable to the appearances of
inner sense.20 This position, however, is at variance with
Kant's doctrine in the Critiqng. For he claims not only, as
he does in the General Note on the System of the Princples,
that the category of quantity "can...be applied...to inner
sense," although, to be sure, "only through the mediation of
outer intuition" (B 239; NKS 256). In the Axioms, moreover,
Kant claims that "all appearances are...intuited as
aggragates." And the fact that they are so intuited, i. e.,
"as complexes of previously given parts" (A 163 = B 204; NKS
199) is said to be the very reason why they allow of
mathematical treatment. 21

(2) In the chapter 'Persons' of his book,
Ingiyidnals, Strawson discusses the problem of how one can
ascribe states of consciousness to others. Referring to the
historical origin of the problem itself, Strawson says: "if
the things one ascribes states of consciousness to, are

thought of as a set of Cartesian egos to which only private

experiences can, in correct logical grammar, be ascribed,

167

then this question is unanswerable and this problem
insoluble."22

It appears that Kant would agree with part of what
Strawson here claims. We can say that to the "private
experiences" there corresponds what in all those passages in
which Kant pgses the Critical problem of an "object of
representations," he calls (in contradistinction to such an
object) representations in the sense of "modifications" or
"inner determinations of the mind" (cf. A 197 = B 232; NKS
22"). These by_tnem§§ly§§ indeed would not be

23

"communicable." But we have already seen earlier that
Kant wants to go further: he is saying that a mind or a
subject could not even know that it has representations in
this sense._ And this is because they do not "reach [the]
unity of consciousness,"2u because, that is, they are not
"synthetically combined in one apperception" (B 138; NKS
157). This, however, would mean that the very notion of a
subject, and the connected notion of the ascription of
representations to it, could no longer be entertained. But
then the problem at hand would not only be "insoluble" but
could not even be stated.

Strawson himself does not want to give us a solution
of the "problem of other minds," but argues instead for a
position according to which the problem does not arise.25
Central to his position are two tenets. Strawson claims,

first, that the concept of a person is "the concept of a

type of entity such that both predicates ascribing states of

168

consciousness n9 predicates ascribing corporeal

characteristics, a physical situtation etc. are equally
applicable to a single individual of that single type."26

He calls predicates of the first kind, e.g. "is in pain,"
P-predicates, and those of the second kind, e.g. "is in the
drawing room," M-predicates,.predicates, that is, "which are

also properly applied to material bodies."27

Secondly,
Strawson maintains that "one ascribes P-predicates to others
on the strength of observation of their behaviour; and that
the behaviour-criteria one goes on are not just signs of the
presence of what is meant by the P-predicate, but are
criteria of a logically adequate kind for the ascription of

the P-predicate."28

Now an obvious difference between Strawson and Kant
is that for Kant there is no single type of entity to which
both M-predicates and P-predicates are applicable. These we
employ, Kant maintains, in our judgments, respectively,
about empirically external objects, including our bodies,
and about empirically inner objects, such as our inner
states and episodes. In contrast to Strawson, then, Kant
holds that a 'person,‘ i.e. the empirical self, is something
'inner': it is an object not of outer, but of inner sense,
and in so far as time is the form of inner sense, it is
time, and not space, which is the "mode of representation of

myself as object" (B A; NKS 79).

 

169

I share this particular view with Bird whose own
comparison between Strawson and Kant is based mainly on the
Refutation of Idealism. But Kant's argument there shows
only, according to Bird, that inner discriminations
presuppose the discrimination of "outer, spatial objects,"29
not, however, that "the concept 'person' has the meaning
that Strawson, but not Kant, ascribes to it." For the
latter, Bird maintains, empirical "persons are essentially

"30

possessors of consciousness. And although Kant rejects,

as does Strawson, Descartes' account of persons, he still

holds "that a person is essentially that to which inner

. . 31

characteristics belong."
And this brings us to the central difference, in my

view, between Kant and Strawson. I have said that Strawson

wants to argue for a position in which the sceptical problem

with respect to other minds does not arise. His remarks,

therefore, Strawson assures us, should not be understood as

a "solution" of this "problem," or as a "general
philosophical 'justification'" of our beliefs about
others.32 Such "justifications" are bound up with a set of

connected distinctions which his own concept of a person, as
we have seen, is supposed to undercut: the distinctions,
that is, between mind and body, the mental and the physical,
the inner and outer, the subjective and objective.

Now we have seen that Kant, too, at times talks in a
Cartesian fashion about representations qna "inner

determinations of our mind" that have only "subjective

170

meaning" (A 197 = B 242; NKS 22"). This way of talking
allows Kant to state the Critical problem in terms of the
"relation to an object" of these inner determinations, and
of their acquiring "objective meaning" and "objective
reality" (A 197 = B 2A2; NKS 22"). One result of my
discussion of the Second Analogy, however, has been that at
the center of the Critical problem is not the 'inner', but
how, on the presupposition that nothing is given to us save
representations (of. A A83 = B 521; NKS A31), there can be,
not knowledge of enter objects, but objects of kngulegge at
all. And this is because of what I have called the
undeterminateness of a manifold of representations as given
in sensible intuition. If the sensible manifold is
undetermined, i. e., net thought in the category of cause
and effect, there is no question of our having knowledge of
particular empirical happenings or events, be they
represented in space or in time only-—that is, be they
empirically inner objects. The central problem, therefore,
is for Kant not the ascription of M-predicates and
P-predicates, but the a_prjgrj possibility of ascribing
predicates te_eny§ning_a§_ellz the possibility, in other
words, of "the unity among our representations," i. e., of
jngggments (A 69 = B 93; NKS 105-06). The actual ascription
of predicates is a_pQ§§erjgri, i.e., discovered and learned
by experience and empirical procedures, and hence for Kant

outside the scope of a philosophical inquiry which for him

is a_pr19ri.

171

(16)
I will finally answer the question how, according to
Kant, we "ascribe," as Strawson puts it, "states of

33

consciousness... to others. It is the question which, as
we have seen, in Husserl's view Kant failed to explore: the
question concerning the transcendental possibility of
positing other egos. And according to Sartre it is the
problem which Kant cannot solve: "the subjective quality of
the Other-as-object," he claims, turns in the Critical
philosophy out to be "a closed box."3u
For this investigation it is the problem of showing
how thinking beings can be known as objects of experience
and hence be members of the objective [ct] domain. The
apparent obstacle, I have pointed out, is that "thinking
beings, a§__nen, can never be found among outer appearances,
and that their thoughts, consciousness, desires, etc.,
cannot be outwardly intuited. All these belong to inner
sense: (A 357; NKS 338). Kant, however, says also that
although "we cannot intuit" the thoughts of thinking beings,
we can "indeed intuit their signs [in the field] of
appearance" (A 359; NKS 340). And the second part of his
Antnrepglggy, as the title page instructs us, deals with
"the ways of knowing the inner of human beings by means of
the outer."35

Such knowledge, however, is according to Sartre,

precisely for what the Critical philosophy leaves no room.

172

If we talk, for instance, about Paul's anger, we are talking
about a series of appearances of "his inner sense" which "on
principle are outside my experience and belong to a system

36

which is inaccessible to me." And we are not allowed to
consider Paul's behavior as the effect of his anger because
the behavior which I observe and the anger which Paul feels
belong to two different experiences: and Kantian causality,
he says, "only... links the phenomena of ene_ang_the_§eme
experience."37

We have seen that knowledge of the existence of
things in general requires the satisfaction of the Second
Postulate according to which "that which is bound up with
the material conditions of experience, that is, with
sensation, is aetnel" (A 218 = B 266; NKS 239). Clearly,
the Postulate as it stands is of no help for the problem at
hand: I cannot have, Sartre will say, sensations of the
appearances of Paul's inner sense. Let us then look again
at Kant's refined version of the Postulate: "The postulate
bearing on the knowledge of things as eetngl does not,
indeed, demand immediate PSEQQPEiQB (and, therefore,
sensation of which we are conscious) of the object whose
existence is to be known. What we do require, however, is
the connection of the object with some actual perception, in
accordance with the analogies of experience, which define
all real connection in an experience in general" (A 225 = B
272; NKS 242-N3).

Indeed it appears that this refined version of the

173

Second Postulate allows us to infer the existence of Paul's
anger from the perception of his behavior. Here is Kant's
example: "Thus from the perception of the attracted iron
filings we know of the existence of a magnetic matter
pervading all bodies, although the constitution of our
organs cuts us off from all immediate perception of this
medium. For in accordance with the laws of sensibility and
the context of our perceptions, we should, were our senses
more refined, come also in an experience upon the immediate
empirical intuition of it. The grossness of our senses does
not in any way decide the form of possible experience in
general" (A 226 = B 273; NKS 243). The objection, of
course, will now be that what makes the problem under
consideration a problem is not "the gneesneee of our
sesnes," but the fact that, as Kant puts it, our thoughts
and our feelings "can never [be] objects of outer intuition"
(A 358; NKS 339). But this fact does not decide the form of
a possible experience in general either. For this form
consists, as we have seen, in the synthetic unity, by means
of the categories, of a manifold of sensible intuition in a
consciousness in general. It consists, in the two cases at
hand, in a manifold's of outer intuition being united, i.

e., thought in the pure concept of cause and effect. It
would be absurd and the existence of the Second Analogy a
puzzle, were we to think that this unification to be a
unification of particular empirical objects of which some

belong to "outer sense" but due to the "grossness of our

174

senses" eenne} be seen, and of which some cannot on
principle be geen, because they belong to "inner sense."
What Kant has argued for is that the representation of the
given manifold of intutition in the category of cause and
effect makes "the experience of an event [i. e., of anything
as happening)" (A 195 = B 240; NKS 223) first of all
possible. This means that only if, as Kant puts it very
carefully, "the manifold in the appearance" (A 190 = B 235;
NKS 220) of the iron filings or of Paul is experienced as a
happening, is it possible to look for and to discover a
pesterigri. i. e., by experience and by whetexer-empirieal
procedures the respective causes, be they emprically inner
or outer.

Because the experience of "anything as happening"
requires e_nrinri application of the category of cause and
effect, the e_pgeteninri judgment 'Paul is angry' is made
"in_eeeergenee_yi§n" the principle of the Second Analogy.

It is for thie reason alone that it expresses empirical
knnnlegge, or possesses what for Kant is equivalent, namely
objective validity and necessary universality (for
everybody). "The relation of cause and effect is the
condition of the objective validity of our empirical
judgments, in respect of the series of perception, and so of
their empirical truth" (A 202 = B 247; NKS 227).

I have shown in the chapter on subjectivity [c] that
objective validity and necessary validity (for everybody)

belong also, and for the same reason, to the judgment 'I am

175

angry'. Hence my judgment 'Paul is angry' and Paul's
Judgment 'I am ansry' express the same_empirisal-epsnisien.
Thus the fact that my judgment is one of outer experience
and that Paul's judgment is one of inner experience does
also not decide "the form of a possible experience in
general." This form is, as we have said, necessary
synthetic unity, and it is the latter which gives us
objectivity and universality: "There would be not reason for
the judgments of other people agreeing with mine if it were
not the unity of the object to which they all refer and with
which they all agree; hence they must all agree with one
another" (Prol. 46). Now the categories, we have seen, are
not only the forms of the unity of the object, 1. e., of a
manifold's being "combined in the object, no matter what the
state of the subject may be" (B 142; NKS 159). but also the
conditions of the form of consciousness in general, 1. e.,
of its necessary unity. And it is this state of affairs
which necessarily ties a plurality of empirical subjects
into a common world, i. e., into a world which is menifeln
as to its metnen, but ene as to its fern. It is this state
of affairs, too, which is the reason why there is only "one
single experience," and why when we speak "of different
experiences," e. g., Paul's and mine, "we can refer only to
the various perceptions, all of which, as such, belong to
one and the same general experience" (A 110; NKS 138), whose
sameness consists in its "form," 1. e., in the "synthetic

unity of appearances according to concepts" (A 110; NKS

176

138). Paul's anger, thus, does net in Sartre's words,
"belong to a system which is inaccessible to me:" we both
inhabit one and the same. And in order to do so, it is not
necessary, as Sartre thinks it is, that I unite my
representations with Paul's. What is necessary, and also
sufficient, is that Paul and I bring, in a judgment, our
representations "to the object unity of apperception" (B

141; NKS 159).

 

 

NOTES TO CHAPTER IV
1(1 h .
_I_V_ Ping, VI, p. 1111.
2. _ .
Haldane, II, p. 64.
3
haldane, I, p. 159.
4. - M A
5. . _
hKSC, p. ‘04—05.
6 , _ . _ . __ .
Lenne, III, p. 604 my trs.
711“ I - ’13 \
“I, v, p. CJ (A. 12).
8“ ._ q ,
hweig, p. 216.
9.
In the fingle b-§fl§ Kant says
aspect of nature... is... the conformity
objects of experience and, so far as is
their negensayy conformity" (44}. To es

is to establish the necessary

to all objects of experience. And in th
tha "na ture, considered merely as natur
dependent upon these categories as

its necessary conformity to law (na tn§e__
§n_'e c_tata)" (E 165; [HHS 173).

O” .

Kawit‘s
Postulate is:
in connection

an object exist

with perc

eption,

application

more explicit formulation of
s or is,

ac
that is,

tha

the original

t the "formal
to law of all
known. a-2ri9ri.
tablish the latter
of the categories

e fininigne he claims
e in general, is

LPOUHO Oi
I orftjal i_‘t§_r‘

the Secczd

tual "if it stands
with sensation as

material supplied by the senses, and through perception is
determined by means of the uncerstanding" (A 234 z B 266;
“its (.5;
'11 A ,
.13 .L..O_.-_L-7 Po 1+6-
12.. , ‘ _ , h._ , . . -v .
LI. h. J. Paton. Lani;s_neta£hysins-ei_iaperisr;e.
ll, London: Georg Allen u Union Ltd; flew York: Humanities
Dress, Inc., 1970, p. 336. Eerafter quoted as Paton.
13.. a.-- I,
(BIL:, A\\';.L.Ll, {Jo 3&0.
14_ u,. .. _ ,.__ h, , .., , ., _
it is this idea, i Cnlflh, that lies at the basis
of Kant's arguments in the Paralogisn Chapter ano links then
to the Trazms endental Deduction. Thus he says. n, cannot
know as an object that which I the t pr suppose in orcer tr
know any object" (A 402; ER; 365). cr L“ naintains that
"the unity of consciousness, which underlies the
categories," that is, "the subjeCt of Le“ CECECOFlES,"

178

cannot be known. ”The subject of the categories cannot by
thinking the categories acquire a concept of itself as an
object of the categories. For in order to think them, its
pure self—consciousness, which is what is to be explained,
must itself be presupposed" (B 421—22; NKS 377). This is
what Kant means when in the Transcendental Deduction he
calls the 'I think', which must be capable of accompanying
all other representations," also "eniginalgapneneepnignz" it
is "that self-consciousness which... cannot itself be
accompanied by any further representation" (B 231; NKS 153).
If it were so accompanied, it could be gnngn and hence occur
in the objective domain. But then it would also no longer
be arisinal.

151irsg, V1, p. 259.
16i_:ia, 111, p. 521.
17-49;, p. 46.
182£2l-, p. 43.
19£r0l., p. ”3.
20E_££§, V, p. 16.
21,

h. H. Welsh. £an£i§_§ritie
Edinburgh: At the University Press,

 

9 .
”4wCig, p. 216.
n, . a
Lweig, p. 153,
25_ .
inniXinnals. p. 112.
26s . . ., q
liaisinnais. p. 101-Oa.
27- . .. .
lL£}”}9Eél§a p. 104.
26. ,. . 4 -_ - .,
infliX4Q4§i§, p. 100.
2 J ’ J "r ' r \. . 'r -
Graham BlFC, gannfl§_lhe9§y”§;_nnnglenge, new iork;
The Humanities Press, 1962, p. 182. hereafter quoted as
' r

Bi c.

(J)

OBirc, p. 186.

U)

179

LLELLOCRAPHE

Beck, Lewis fihite. ”Six Short Pieces on th
of Experience," in Leggys_panant ‘

nd An'IOgy
3'1 3
haven and London; Yale University Press,

161'.

r

o L
ref“ ..-,
9(Ua £51“-

.,_

~~~-mu—» "Lid the Sage of Koenigsberg Have to Dreams?" in
s e , g_1 " . Lew Haven and London:
le UniverSity ‘ress, 1978, pp. 38n60.

Bennett, Jonathan. gant1§

__ finelygip. Cambridge. Cambridge
University Press

Cameridgc. Cambridge

y re . , .
berteley. Georae. Esrselsiis-lhii:snpnieaiizritisas. he.
T'avid 1. Atmstrong. London: CollieruLecnillah,

('2?
(.1

'rd, Crtiunu. next"

Descartes, hen-. it
vols. rs.

i _ p_ , _
northwestern University Press, 1906.

alsitr, ntuolf. santsteaispa- “ilceghei“‘ GCOPL Cit;
Verlagsbuethancl‘nay 19b1°

deicnann, Lucien. L&_epnnnnvnie_npnngnewe§_i:pggye;g_gngn
- .'.--. ‘ ‘ 7‘ 1' r
_A_L £33 .1! a I C‘. I1 .L 5.4 o t) o ' LI 0 , 1 J L o
brat Loltie Q 'a‘t Oxtoi‘“\ ant tr: 1 IFlOP'
‘ 1 ' ‘ ' "-’ ° .t.-.._.:..‘-_.:.- -.-‘-._._ 51...; J... _ .5. .. - - L‘ ’ :2. __.. ._ _ -- .1‘ °
,~ , fl, a. .’ .... u. 7 , —— e- s " ‘ i ‘ V ‘o— _ /;
hvanston. hOILUhechPh uniVersi 3 .re-s, 19 t.

.‘

" » ' ' . - , V p -,a- - ' \ ,\ "‘ . .. _ -. _
ilk? 193-c. be I‘ a 1‘01 b —'-h - 4.4.1:_L_ 5‘11}; 54.0 E‘-.._1JI _O_u_. 53:. 39?. .1_«_U.‘«_<-‘«rlt.f 3:3:- -
-- A . .3. -. —. ' , ' - , ' .' -\- - - ~ .- r, . A. -- . .. " r’
tranhlufu an rain. Vittorio nlOchPnaLh, 18:1.
1: _ ~' ~. I— (’3 r- (; - - ,. .. i i ,
°- ‘ ‘ ‘ d .- N "7 o £A£AE-L_ :LE_ _C&_ -lia_l_LA;__ 6 11‘ L: I U- c L o L) C {fl Lr'v‘as, J! o ’ L . e
-. -~ 1“ \ “ .‘ I" ‘ 1‘ A r- .‘ x - ' ,_ r‘ I r < . ‘ J ‘
\ILJI‘C. i/t‘lxbuCli Evita. c111 anal bl; L steel.” 1. enhuiiu.
9 J J O
" s - 1 ‘ v s” I:- ~ - -.- :‘e v’ 1'"
emitsub. AALLJI; “L. “'cly vb., 1:01.
heurion, Lister. “Lie seweisstruktur von “ants
. ,i . : ., s , - ,. _. - .’ s . .. .. .- . , - (i
Ll“ C La 5 L C 14". L Ad L4 CL .L \—I L C C. U “u b l UIA . ' ... Ii ¢&;_oi ’ :4 U 0 U L i." ;._ Kt
'- ~ - o c I (\f. \‘ . 7“.. - - '. “.- . .' " a. C .11: '-.r 2 -
ELfiQLQQ. Ask/91 A) o ALLCi/ LIELQM'CD £4. 1. 1 Us. L“, 1 ‘3‘. .. ‘ r r).

90-104.

~e. Ed. L. A.
rendon Press, 1960.

heme, David. A Tneatip

'che Leoitationen und Fariser

husscrl, Edmund. "Cart is
inn; I. Haag: Kartinus hijhoff,

as a
Vortraege." Buseen

1950.

- ~~~~--. "Erste Philosophie.” fineéeniifie Vll. Haag:
Lartinus Hijhoff, 1956.

Immanuel. Imnnnnel _,_ H
Ed. Erich Adickes. Berlin:

,_
a
H.
("1"

h.

_---.---, gantis_§ritinpe_9;_lra9ti§al_£§aspi-ans-ch;r
H ass-ey-iheiilepriiei-Bthins- T‘s. T. K. Abbott.
‘ Tork~Torontoz Longmans, Crccn and Co

~~~~~~~-. Zant:s-§ri&icpe_pf-innanent. Trs. J. H. bertare.
London: Lacnillan and Co. Ltd., 1914.

—--~--—-. lrplsapnsna i9lfljl_fy&§£§_L§§9PLE§A§§o Be. Lewis
White neck. 1 cianapolis~NeJ lork. The Scbbs~herill
5

 

——~~---—. grinin_eer_gfgn§j_j;rgpngg. Hamburg; Felix
neiner, 1956.

r: lessen. Trs.

—-------. immar9;iiiarils_§1ipii:;_pi__p _ _
Korma. hemp Smith. LOncon. Lacnillan u Co., Ltd.,
1956.

—~--~- , 42;5;_ " _ _ _ _ 9'. Vilheln Geischeccl.

L.
liche Buchgesellscnaft, 1966.

CPS.

‘ .. Q . ' . I t a ‘ ,_ .-. _ ‘v ,_ fi _, . .-. ,‘_ __ ‘H‘ _‘ _ _ 1- ,

news snlpn. Leruan. A_gesnsni2£iiip_ianijs-i.iiinps-pi-ipli
‘a 'r I“ t.,,,.,-°.-‘ ,_ A ./n
__flg_g_u_§)_£i. C‘n lOY'lx lne Ethniaillclb'fi PPCSS, IC-Oc.

- ‘ ' t». 1 x. 1‘ ’5 ~ ‘ ' 7 ~ ‘~ -. " ' < -. v "6' ‘ "' '
LJCFIICI ’ k; . {‘111t; .LLAJL’C‘QDLUlJ. .L C3 0* LCIA ksbn. vC’ ‘
‘... - s ' t '. .- -..-- '— ~- ' ~~ ' \--' '
UCIKAUULiUut-. " In $25.11.)“ f-_C}4_L:_i_b_§_isj_g_c._y. cu. LCo“..LU .{-lot,
7" .\, 1' .‘x (1‘ It ‘ -. v . I" ., ,2 b' ," _ .’ A I‘ r\ - f. 1 l
He‘d“ 14396-119, J—L- Until bOLl‘b, 1905; k-‘p. ago-4e.
..IN V ‘_ , _. _. '.. __ ' 13" .~__, ‘ _ -- , 4 .‘Fy!
._-....__..-....—. Lilli. .LQULLQL‘.\. iCl-lbuili LJQOI'aQ' -_-.;x..., [3,41,

[(otziih hhooa.1u "han
Empirical l 0
Int ernatignel
112*119.

on the Hole of Sensation in
dge " In Akten_ deemfié

Locke, John. _1h1hssay Co
A. S. Pringle Pat
Press, 1969.

9§rfiifii.EPE§D-UBC§IP
on. Oxford: The C

hartin, Gottfried. Imaanuel_3n.t. Berlin: Walter de
c

Gruyter a Co., 196).

LerleauvPonty, Laurice. Pnengnenglnhy
Colin Smith. New lork: The .u

i. J- iantjs_.l . i -- s
lork: The humazlities Press, 1970.

rj-I

Paton,

houssett, Bernarc. L
Paris: Lihrar

é_- __ l- ____l.
i

Sartre. JcannPa ul inintlant-npthinane§s. Tr“. neuei
Barnes. hew Iork: Philosophical Library, 95 .

Strawson, Peter Frederick. litiyinnele. London: Le thuen t
Co., Ltd., 1965

~~~--—-~. Engrhgpnng_gi_;en§e. Loncon; nethuen a Co.,
Ltd., 1966.

Scllars, hilfriec. Seiente_ann_rptaphysins. London:
Routledge a Regan Paul, 1968.

Veininser hens. Qfiemfillgjiék-i§fliféi51lléfl-£3£_f§}&ffi
Ye _rnnn_t, 2 vols. Stuttgart-Berlin-Leipzlg. Uxion
Deutschc Veriagsztsellanaft, 1781—92.

'1 .._....,‘ .. v‘ -- ’ - ._ ,6 ‘,_,‘ , ,dr.

Visesuhahher t. J. ee. LL_J§£§QEJ93_$r9999§;§§::9g?_y:y£
l OSCVifi 93_Lant, 3 vols. Antwerpen—Paris ‘S
GI‘E‘iV‘Cll 1116.58, 1954*37.

talsh, h. h. gentjsjgrinieisn_p§_;et9fgjgip§. Ldlnturgh
At the Univers ity Press, 1975.

Wolff. hopert Paul. Lééfiléijflfiprji$i_L§P§§}-£9$3Y3l3-
Cahbrnge; Earvard University Press, 1913.

 

"Illllllljilllllllr