<< . .

. 4
( : 33)

. . >>

For further clarifications concerning the role of comparison in forming empirical judg-
ments, see KCJ, pp. 113“14.

of substance). Or again, the a priori rule: ˜˜look for what can be repre-
sented by way of a successive synthesis of homogeneous units™™ (this is the
schema for the category of quantity). But in comparing empirical mani-
folds synthesized according to these a priori rules, we become aware of
common patterns of apprehension and we form empirical rules such as:
˜˜look for what can be represented by way of a synthesis that varies around:
four supporting elements (the paws), an oblong-shaped body and pointed
front part (the head), a wagging end-part, a loud sound, and so on™™ (this is
the schema for the empirical concept of dog).
I think that some of the puzzlement Sedgwick expresses comes from
the fact that she confuses what I say about empirical concepts and their
schemata, and what I say about the categories and their schemata. For
instance, she asks: ˜˜How can schemata both guide comparison and result
from comparison?™™18 Here the answer is quite simple: we are not talking
about the same schemata in both cases. The schemata that guide the
comparison are the schemata of the categories. We must have synthe-
sized according to the categories “ looking for homogeneous manifolds,
looking for permanent and changing properties, looking for sequences
in which any change of states ˜˜presupposes something upon which it
follows according to a rule™™ “ in order to come up with representations of
individuals that we proceed to compare in search of empirical rules for
recognition “ empirical schemata.
As an example of the kind of judgment that results from the process of
comparison I have been describing, I would not choose: ˜˜the tree is a
spruce™™ (Sally Sedgwick™s example) but rather: ˜˜all things that have a
trunk, branches and leaves, are trees,™™ or perhaps: ˜˜all trees that have
leaves of such and such a shape are spruces.™™ For what I have been
describing is the process of selecting common features to form empirical
rules for recognition of kinds of things in nature. ˜˜This tree is a spruce™™
would be relevant as an example of application or instantiation of a rule
for recognition thus formed. Moreover, I do not think that an act of
comparison is needed to determine ˜˜which concept is to assume the
place of logical predicate, and which of logical subject™™ (as Sedgwick
suggests). For this is arbitrary: as Kant writes, we might say ˜˜all bodies
are divisible™™ or ˜˜some divisible things are bodies.™™ The place of a con-
cept as subject or predicate in an empirical judgment becomes con-
strained only when we think the object thought under it as ˜˜in itself

Sedgwick, ˜˜Priority,™™ p. 86.

determined with respect to a logical form of judgment,™™ namely sub-
sumable under a category (on this point, see B129). On the other hand,
the act of comparison is needed to determine whether the judgment to
be formed should be an affirmative or a negative judgment (expressing
the agreement or the opposition of concepts under which we represent
objects), a universal or a particular judgment (expressing the identity or
diversity of objects with respect to concepts), a categorical or a hypothe-
tical judgment (expressing a predication under an inner or an outer
Now, what about the categories and their ˜˜acquisition™™? Sedgwick
suggests three senses in which I might be maintaining that the categories
are ˜˜generated™™ in our acts of judging: we become aware of them in our
acts of judging, they are realized in our acts of judging, their form
(universality) is generated out of acts of judging. She adds that in the
third sense there is nothing special about the categories: all concepts are
generated as to their status as universal and reflected representations by
acts of judging. I agree. I spend quite a bit of time explaining just this
point. I am not sure I would endorse either of the first two suggestions,
however: I do not think there is much sense in distinguishing between
˜˜rule™™ and (clear or obscure) ˜˜awareness of the rule™™ in the case of either
schemata or concepts, and I do not think I actually use the expression
˜˜realize the categories.™™ So, what do I mean when I talk of ˜˜generating™™
the categories, and how would I answer Sedgwick™s concern, that the
specificity of categories as a priori concepts seems to be lost if we accept
this point?
Kant himself, actually, is quite explicit about what he calls the original
acquisition of the categories. In his well-known response to Eberhard, he
explains how both space and time, as formal intuitions, and the
categories, are ˜˜originally acquired.™™ The text is worth quoting at some
Impressions are always required in order first to enable the cognitive
powers to represent an object . . . Thus the formal intuition which
is called space emerges as an originally acquired representation
(the form of outer objects in general) . . . the acquisition of which long
precedes determinate concepts of things that are in accordance with this
form. The acquisition of these concepts is an acquisitio derivativa, as
it already presupposes universal transcendental concepts of the under-
standing. These likewise are acquired and not innate, but their acquisition,

See KCJ, ch. 6.

like that of space, is originaria and presupposes nothing innate except the
subjective conditions of the spontaneity of thought (in accordance with the unity
of apperception).20 [emphases in the last sentence are mine]

The idea, then, is this: categories are acquired in that we would form
these concepts neither as rules for synthesis of manifolds in intuition, nor
as ˜˜universal and reflected representations,™™ unless impressions had
triggered our cognitive powers to launch the effort to represent objects.
But they are originally acquired in that both what the rules of synthesis,
and what the universal concepts reflecting these rules are going to be, are
a priori determined by ˜˜the subjective conditions of the spontaneity of
thought™™ (the logical functions of judgment) together with the ˜˜first
formal grounds of sensibility™™ (space and time). In the Critique, Kant
describes this a priori acquisition as an ˜˜epigenesis of pure reason,™™ and he
contrasts his ˜˜epigenetic™™ view of reason both with innatism and with the
idea of an empirical generation of the categories (B167“8). There is thus
no ambiguity at all about the notion. What makes the generation of the
categories unique is that although they are generated (both as rules for
synthesis and as discursive concepts) only under empirical conditions,
their content is determined independently of these empirical conditions
and, indeed, is an a priori condition for the generation of any represen-
tation of empirical objects at all.
What I have said so far should now help me address Henry Allison™s
questions concerning my treatment of the categories in Deduction B.

Deduction B: ˜˜Where have all the categories gone?™™
Deduction B, part one
Allison and I agree that Deduction B is one argument in two main parts.
In the first part, Kant is concerned with proving that the categories are
the intellectual conditions for the representation of an object of sensible
intuition in general. In the second part, he is concerned with show-
ing how the categories relate to the sensible conditions under which

Uber eine Entdeckung, nach der alle neue Kritik der reinen Vernunft durch eine ¨ltere entbehrlich
gemacht werden soll, AAviii, p. 223; trans. Henry Allison, On a Discovery whereby any New
Critique of Pure Reason is to be made superfluous by an older one, in Theoretical Philosophy after
1781 (Cambridge: Cambridge University Press, 2002). For a careful and detailed study of
the ˜˜epigenesis™™ and ˜˜original acquisition™™ of our representations of space and time
according to Kant, see Wayne Waxman, Kant™s Model of the Mind (Oxford: Oxford
University Press, 1991), ch. 7; also chs. 1 and 3.

empirical objects are given. However, Allison and I disagree about the
precise content of the argument in each part.
Concerning the first part, Allison urges that in stressing as I do the role
of the logical functions of judgment, I lose track of the categories alto-
gether. Moreover, he thinks that I read the transition from part one to
part two as a regressive argument that moves from the consideration of
discursive judgment (analyzed in x19) to its conditions in the transcen-
dental synthesis of imagination (xx24 and 26). Against this ˜˜regressive™™
reading, he urges that Deduction B is a progressive argument, from the
elucidation of purely intellectual conditions for the representation of
objects (categories as forms of intellectual synthesis) to the elucidation of
the application of the categories to our sensible intuition and, more
particularly, their role as conditions of the unity of time. Let me consider
each of these two points in turn. The second will provide me with the
transition to Allison™s criticism of my treatment of the categories in the
second part of the B Deduction.
First, do I lose the categories altogether in the first part of the argu-
ment? I do not think so, for the reasons stressed above: I do insist that
logical forms of judgment are forms of the combination of concepts,
whereas categories are universal representations of the synthesis of
intuitions. This difference is strongly present in my reading of the first
part of the B Deduction. I devote a separate chapter (ch. 3) to xx15“18 of
the B Deduction, where Kant argues (1) that any representation of an
object rests on the unity of the synthesis of a manifold in intuition (x15),
and (2) that this unity is to be referred back to the original synthetic unity
of apperception (xx16“18). Only after going through these initial steps
do I submit to close scrutiny x19, where Kant states that the logical form
of judgments is the objective unity of the apperception of the concepts
(note: of the concepts) combined therein, namely the unity by means of
which concepts are related to objects. After devoting four chapters to
analyzing what Kant might mean by this, I conclude:
Kant™s purpose in section 19 is to argue that the logical form of judgment
is the discursive form of the objective unity of apperception whose intuitive
form he described in section 18 as preceding and determining all
empirical-subjective unity of consciousness . . . This is what allows him
to conclude, in section 20, that the unity of empirical intuition, insofar as
it necessarily stands under the original synthetic unity of apperception,
also stands under the logical form of judgment, and thereby under the
categories, since the latter are nothing other than ˜˜concepts of an object,
insofar as the intuition of that object is considered as determined with

respect to the logical functions of judgment™™ (section 14) or ˜˜universal
representations of synthesis™™ (section 10).21

So I do make the distinction quite explicitly, and I do stress how Kant™s
opening argument concerning the necessity, for any representation of
object, of a unity of synthesis of intuition under the unity of apperception
is what allows him to move from asserting that the manifold of intuition is
brought to the unity of apperception by way of the logical form of judg-
ment, to asserting that this manifold is subject to the categories. I admit,
however, that some of my formulations tend to blur the distinction
between logical functions of judgment and categories. This is because
I put great emphasis on the fact that absent any sensible manifold to be
synthesized, all that remains of the categories are logical functions of
judgment. And even with a manifold to be synthesized, to understand
each and every one of the categories we need to relate it to the specific
form of judgment toward which it guides the synthesis of manifolds in
intuition. This is what it means to say that a category is a ˜˜concept of an
object, by means of which the intuition of this object is considered as
determined with respect to a logical function of judgment.™™
In fact, it is precisely because I give so much importance to the relation
of categories to the synthesis of intuitions that in KCJ I indicate my
disagreement with Henry Allison™s view according to which in the first
part of the B Deduction, the object Kant is concerned with is only an
object in sensu logico, in contrast to the second part where Kant is sup-
posed (according to Allison) to be concerned with the sensible object, the
object given in the forms of our sensibility, space and time. Against this
view, I maintain that already in the first part of the Deduction, the notion
of an object is to be analyzed as involving (1) the ˜˜undetermined object of
an empirical intuition™™ (the appearance of the Transcendental
Aesthetic), (2) the object of the synthesis of appearances (cf. x17 of the
Deduction, at B137: ˜˜the object is that in the concept of which the manifold
of a given intuition is united™™ “ all emphases are Kant™s), and (implicitly)
(3) the transcendental object, namely the object we presuppose as exist-
ing, and by reference to which we seek agreement among our synthe-
sized representations.22

KCJ, p. 185.
For the same reason (in addition to textual reasons), in KCJ I express doubts about Henry
Allison™s suggestion that the distinction between Objekt and Gegenstand on the one hand,
objective validity and objective reality of the categories on the other hand, is relevant to the
transition from part one to part two of the B Deduction. I think in both parts the categories

Nevertheless, I agree with Allison in maintaining that in the first part
of the Deduction, the categories are considered as pure intellectual
concepts of the unity of synthesis of any intuition, as long as the latter
is sensible (receptive, not spontaneous). The forms of our sensible intui-
tion, space and time, do not play any specific role in the argument. By
contrast, in the second part they do come into the foreground. Allison
and I also agree that for this to be a significantly new move in the
argument, part two has to be more than the specification to the case of
our sensibility, of an argument first made in the general case of all
sensibly conditioned intellect. So what is new about part two of the
In my view, the answer is this: in part one, Kant argues that the
categories, albeit originating in the understanding alone, are concepts
under the guidance of which the synthesis of any sensible intuition
achieves the kind of unity that allows it to be related to an object repre-
sented as distinct from our representation of it. In part two, he argues that
space and time themselves, the forms of our sensibility, stand under the
very same unity of apperception whose discursive forms are the logical
forms of judgment, and in which the categories thus originate as ˜˜concepts
of an object, by means of which the intuition of the object is considered as
determined with respect to a logical function of judgment.™™
Contrary to Allison, I do not think that the decisive step in part two is
x24, namely Kant™s explanation of what he calls the ˜˜figurative synthesis™™
(synthesis speciosa) or transcendental synthesis of imagination. I take x24
to be a transition section, one that is certainly extremely important in
that it introduces the notions that will be essential to the second part of
the argument: figurative synthesis, affection of inner sense by the under-
standing. But part two of the argument, properly speaking, does not
occur until x26. Kant himself states this quite explicitly, not once but
twice, each time stressing that in part two he is going to consider ˜˜the
manner in which things are given™™ (B144“5, B159).
Is such a move from part one to part two of the Deduction a regressive
argument? One may want to describe it in this way, since after all things
do need to be given before they are thought (synthesized and reflected
under concepts according to the logical forms of judgment). So my
account might be read as a regression from conditions of thought to

are related to the object of empirical intuition. However, in the first part, the argument
rests on the nature of the categories as forms of thought. In the second part, in contrast, it
rests on the nature of space and time as the a priori forms of our sensible intuition. See
KCJ, pp. 110“11, n. 14.

what is more primitive: conditions of presentation in sensibility.
However, I should caution that I do not think that the argument pro-
ceeds simply by retreating, or regressing, from forms of discursive
understanding (part one) to the syntheses of imagination as their pre-
condition (part two). The argument is more radical than this if, as I have
just suggested, it moves from forms of thought to forms of givenness.
Nevertheless, Allison is correct in pointing out that I talk of a ˜˜retreat™™
from the forms of discursive understanding to the syntheses of imagina-
tion. But when I use this expression, what I describe is the transition, in
my own book (KCJ), from part two (where I consider the logical forms of
judgment) to part three (where I consider the transcendental syntheses
of imagination, and thus not only xx24 and 26 of the Deduction, but also
the System of Principles of the Pure Understanding).23 When describing
Kant™s argument in Deduction B, what I say is that part two of the
argument is a revisiting, in light of the argument of part one, of ˜˜the
manner in which things are given,™™ namely the forms of intuition, space
and time, that were first expounded in the Transcendental Aesthetic.24
Kant™s point is that space and time themselves, which have been
described in the Transcendental Aesthetic as forms of intuition and
pure intuitions, are now revealed to be the product of the ˜˜affection of
sensibility by the understanding,™™ namely by the unity of apperception
as a capacity to judge. And so, by the mere fact of being given in space
and time, all appearances are such that they are a priori in accordance
with the categories, and thus eventually subsumable under them.
However, I am aware that I am not making my case any better in Allison™s
eyes by proposing to read the second part of the B Deduction in this way.
Our most fundamental disagreement bears precisely on this point.
So I now consider Deduction B, part two.

KCJ, p. 197.
See KCJ, pp. 212“16. I do not claim to be especially radical in my reading. It is Kant™s thesis
that I describe as radical, not my reading of it. What I hope on behalf of the latter is that it is
accurate. Nor do I make any claim to being the first to defend such an interpretation.
Predecessors include e.g. Hegel: see Glauben und Wissen, in G. W. F. Hegel, Gesammelte
Werke, Deutsche Forschungsgemeinschaft, ed. Rhein-Westfal. Akad. d.Wiss. (Hamburg:
F. Meiner, 1968“); Faith and Knowledge, trans. Walter Cerf and H. S. Harris (Albany: SUNY
` ´
Press, 1977). Pierre Lachieze-Rey: see L™Idealisme kantien, 3rd edn (Paris: Librairie philo-
sophique Vrin, 1972); Wayne Waxman: see Kant™s Model of the Mind. The originality I claim
for my view is my emphasizing the relation between the unity of apperception and the
logical functions of judgment, and my relating the unity of space and time to the ˜˜unity
that precedes the category of unity™™ (B131, in x15 of the B Deduction). More on this below.

Deduction B, part two
Allison objects to two main points in my interpretation of this second
part: my identifying the ˜˜formal intuitions™™ of x26 of the B Deduction
with the ˜˜forms of intuition and pure intuitions™™ of the Aesthetic, and my
claim that when Kant defines synthesis speciosa as an affection of sensibility
by the understanding, he means affection by the capacity to judge. Let
me consider each point in turn.
First, form of intuition and formal intuition.
I maintain that when Kant describes space and time as ˜˜formal
intuitions,™™ in the footnote to x26 of the Transcendental Deduction,
he is describing the very same space and time he characterized as
˜˜forms of intuition™™ or ˜˜pure intuitions™™ in the Transcendental
Aesthetic. I am not maintaining that the Transcendental Deduction
calls for a ˜˜revision™™ of the Transcendental Aesthetic. The term I use is
˜˜re-reading™™: what I think is that everything that was said in the
Transcendental Aesthetic about the nature of space and time stands,
but it is brought into new light by the argument of the Deduction.
Indeed, when Kant says, in x26, that space and time ˜˜are represented
with the determination of the unity of the manifold,™™ he immediately
adds: see the Transcendental Aesthetic (B160).25 And then he goes on:
this unity presupposes a synthesis by means of which ˜˜(in that the
understanding determines the sensibility), space and time are first
given as intuitions™™ (B161n). Here he refers us back to x24, where he
explained the ˜˜affection of sensibility by the understanding™™ as being a
synthesis speciosa, or the transcendental synthesis of imagination (see
B151“2). Space and time, then, are forms of sensibility, just as Kant
maintained in the Transcendental Aesthetic. But they are forms of a
sensibility affected by the understanding, and thus they are the product
of synthesis speciosa, the transcendental synthesis of imagination. And
I must say that it seems to me quite reasonable to maintain that the
unity, unicity (there is only one space and one time), and infinity of time
and space “ all features attributed to them as pure intuitions, in the
Transcendental Aesthetic “ are features we imagine or anticipate and
thus project as preconditions of the unity of experience. It strikes me as
quite reasonable to maintain that, on the one hand, the qualitative
features of spatiality and temporality depend on our sensibility, which

The same was said at B136n, B140, B137. But only now is the point brought into the
argument with full force.

thus provides ˜˜first formal grounds™™ of the ordering of sensations that
yields appearances; and that, on the other hand, the unity, unicity,
and given infinity of space and time “ and thus space and time themselves,
as intuitions in which all appearances are combined and ordered “
are products of our imagination. This is no revision of the
Transcendental Aesthetic. The latter allowed for this further develop-
ment, indeed mentioned it explicitly in the B edition, where Kant intro-
duced the idea of a ˜˜self-affection™™ of the cognitive subject, in striking
parallel to the idea of synthesis speciosa introduced in the Transcendental
Deduction (cf. B68“9).
In support of my proposal that the ˜˜forms of intuition™™ of the
Transcendental Aesthetic turn out just to be the ˜˜formal intuitions™™
resulting from what Kant calls, in the B Deduction, the ˜˜affection of
sensibility by the understanding,™™ I observe that it would be a mistake to
suppose that ˜˜form of intuition™™ is universally opposed to ˜˜formal intui-
tion™™ as what is indeterminate to what is determinate. The reason this
would be a mistake, I maintain, is that Kant™s notion of form is a rela-
tional one, always paired with matter. And in this pairing, form means
˜˜determination,™™ matter ˜˜undetermined™™ (and determined by the form)
(see A266/B322). To this, Allison objects that the opposition between
form of intuition and formal intuition, in the footnote to B160, is an
opposition between what remains ˜˜indeterminate™™ and what is ˜˜deter-
minate.™™ Moreover, he points out that Kant also mentions ˜˜form of
intuition™™ as what is indeterminate elsewhere (B154). Of course I agree
with that. As Allison acknowledges, I myself insist that in the footnote to
B160 the ˜˜form of intuition™™ is indeterminate by comparison to the
˜˜formal intuition™™ which is determined by the ˜˜affection of inner sense
by the understanding.™™ What I add, however, is that the opposition so
understood cannot hold universally and cannot be an argument for
opposing formal intuition to form of intuition in all cases. My suggestion
is that ˜˜form™™ should always be understood in context, and in connection
with the specific matter for which it is the form. Thus the form of
intuition as mere ˜˜formal ground™™ (in On a Discovery) is a form for a
matter, sensations as mere affections of which we are not even conscious.
The formal intuition as providing ˜˜not only the manifold, but the unity
of the manifold™™ (B 160n) is a form for the matter of appearances. Recall
that, in the Transcendental Aesthetic, Kant says of the appearances that
their matter is ˜˜that which corresponds to sensation™™ and their form is
space and time as forms of our sensible intuition (and themselves pure
intuitions). In a footnote to the Transcendental Dialectic, Kant explicitly

equates ˜˜form of intuition™™ and ˜˜formal intuition™™: ˜˜Space is merely the
form of outer intuition (formal intuition)™™ (B457n).26
Second, affection of sensibility by the capacity to judge.

<< . .

. 4
( : 33)

. . >>