. 2
( 10)


understanding altogether. And John McCumber argues that the critical
project can survive Kant™s mature insights into re¬‚ective judgment and its
relationship to the sensus communis only through a radical transformation
of its governing ideals.
Rudolf Makkreel shows how the function of re¬‚ective judgment is
essential to the overall cognitive project of reason, but he argues “ contra
Rebecca Kukla

Longuenesse (1998) and some others “ that re¬‚ective judgment is not
always already at work in concept formation and determinative judgment.
Whereas re¬‚ection merely brings disparate particular representations
together to ¬nd what is common for a general concept, re¬‚ective judg-
ment aims “to ¬nd ˜suf¬cient kinship™ among empirical laws to allow them
to be part of a common system.” Re¬‚ective judgment also “seeks univer-
sality whenever we are still left with a remainder of particularity” and
“discerns lawfulness” in what seems contingent. Driven by the principle
of purposiveness, we exercise re¬‚ective judgment in seeking systematicity,
coherence, and unity in nature, all of which are “intrinsically contingent
from the standpoint of the understanding” “ and hence successful re¬‚ec-
tive judgment is essentially pleasurable. Re¬‚ective judgment is thus, in
Makkreel™s view, not subservient to the goals of the discursive understand-
ing, and it does not directly contribute to empirical judgment; instead,
it furthers the ideals of reason by framing our experience through inter-
pretation and orientation. On his reading, the subjective universality of
re¬‚ective judgment is not grounded in actual commonality that we can
presuppose, but is rather a possibility to be cultivated through human
Kirk Pillow seeks to challenge the clean distinction between determi-
native and re¬‚ective judgment in Kant, and to insert creative aesthetic
re¬‚ection and interpretive system-building into the Kantian understand-
ing itself; hence his reading is in direct con¬‚ict with Makkreel™s. Pillow
argues that in order to accommodate the dimensions of human inquiry
that Kant brings to light in the third Critique, including in particular his
account of aesthetic ideas, his original picture of understanding as a fac-
ulty of discursive rules for subsumption has to be revised. Drawing on
the work of Nelson Goodman and Catharine Elgin, Pillow proposes a
new account of the understanding “ one that takes into account how
systems of understanding are “themselves the hard-won products of col-
lective labor [embodying] a history of shared human interests and goals.”
Understanding an object, for Pillow, might encompass not only determin-
ing of which general concepts it instantiates, but also grasping how to use
it, its history, its relationship to other objects and human practices, its
symbolic meanings, its location within a system of property rights, and so
forth. Such understanding is governed not just by the ideal of truth, but
also by a host of cognitive values driven by our collective interests, such
as salience, symbolic resonance, and coherence, many of which have an
ineliminably aesthetic dimension. Pillow concludes that “Kant helps us
see the way to a uni¬cation of the cognitive and the aesthetic, so long
Introduction 31

as we understand cognition more richly than as conceptual subsump-
tion, and so long as we are not tempted by the chimera of pure aesthetic
John McCumber™s deconstructive and constructive intentions are far-
reaching; he promises to reveal Kant™s “helplessness” in the face of the
fact that despite his attempts at mastering his carefully bounded terrain,
“something foreign and unsought, yet intelligent, is surging into [his] phi-
losophy” when he writes the Critique of the Power of Judgment. McCumber
argues that Kant gives us the resources in the third Critique for the begin-
ning of an account of rational revision of our concepts and cognitive goals,
which is an epistemic project distinct from determination, re¬‚ection, or
system-building in accordance with ¬xed principles. While agreeing with
Kant that we need to presuppose a conceptual framework as a condition
for the possibility of experience, he claims that Kant himself gives us rea-
son (despite himself) to see such presupposed frameworks as historically
developing and always open for critical revision rather than as timeless
and a priori. The upshot is that the faculty of reason and its govern-
ing principles end up being “historical through and through,” and that
Kant™s quest for discursive mastery and transcendental epistemic security
are undone by the aesthetic activity that insinuates itself into his theory of
judgment. Hence, according to McCumber, the third Critique transforms
the critical project into something new and suspiciously Hegelian.
Collectively, these essays draw upon the texts, methods, and debates
of both the continental and analytic philosophical traditions, and they
situate Kant in the history of philosophy, not just as a respondent to
the high rationalism and empiricism of the early modern era, but as a
touchstone antecedent to Hegel, Schiller, Schopenhauer, Wittgenstein,
Sellars, Gadamer, Foucault, Davidson, Goodman, and McDowell. They
reveal Kant™s account of judgment, and the dialectics of aesthetics and
cognition within that account, as profoundly relevant to contemporary
debates in epistemology and philosophy of mind, and more narrowly to
debates surrounding the nature of empirical experience and the meta-
physics of normativity. The essays yield important lessons concerning the
ineliminable and yet often problematic place of imagination, sensibility,
and aesthetic experience in perception and cognition.
part one


Thinking the Particular as Contained under
the Universal

Hannah Ginsborg

In a well-known passage from the Introduction to Kant™s Critique of
the Power of Judgment, Kant de¬nes the power or faculty of judgment
(Urteilskraft) as “the capacity to think the particular as contained under
the universal” (Introduction IV, 5:179).1 He goes on to distinguish two
ways in which this faculty can be exercised, namely, as determining or as
re¬‚ecting. These two ways are de¬ned as follows: “If the universal (the
rule, the principle, the law) is given, then judgment, which subsumes the
particular under it . . . is determining. But if merely the particular is given,
for which the universal is to be found, then judgment is merely re¬‚ecting”
(ibid.) As Kant goes on to make clear, the Critique of the Power of Judgment is
particularly concerned with judgment in its capacity as re¬‚ecting rather
than determining. It is concerned, that is, with how we are to ¬nd univer-
sals (which he glosses as rules, principles, or laws) for given particulars.
Despite the fact that the term ˜concept™ does not appear in this set of
de¬nitions, Kant™s discussions of judgment elsewhere make it clear that
this faculty can be identi¬ed at least in part with our capacity to think
particular objects under concepts, in particular empirical concepts.2 The
sense of ˜universal™ (allgemein), then, would appear to be the same sense
that is implied in Kant™s characterization of a concept as a “universal” (or
as it is sometimes translated, a “general”) representation (Logic §1, 9:91).

1 All references to Kant™s works, except for the Critique of Pure Reason, give the volume and
page number of the Akademie edition of Kant™s collected writings (Berlin: De Gruyter,
1902“), with other details as appropriate. References to the Critique of Pure Reason give
the usual A and B pagination. All translations are my own.
2 See especially Section V of the First Introduction to the Critique of the Power of Judgment,

Hannah Ginsborg

To say that a concept is universal or general is to say that it is “common
to several objects” (ibid.), and hence contrasts with an intuition, which
is a singular representation. The question of how we are to think the
particular as contained under the universal would thus appear to be the
question of how we can grasp an individual thing under a concept, that
is, how we can think it as having a feature that can at least in principle
be shared with other objects.3 And re¬‚ecting judgment more speci¬cally
would be concerned with the question not of how we can apply concepts
that we already have, but how we can arrive at concepts in the ¬rst place.
There is, however, another, apparently distinct sense of ˜universal™ that
is also invoked by Kant in describing the exercise of judgment, more
speci¬cally judgment in its capacity as re¬‚ecting. In particular, Kant uses
this sense of ˜universal™ when he describes the claim to agreement made
by a judgment of beauty, although he makes clear that this same claim is
made by cognitive judgments also (see for example Critique of the Power of
Judgment, Introduction VII, 5:191). “Universality” in this sense means, as
he puts it, “validity for everyone” (§8, 5:215). The pleasure in an object
expressed in a judgment of beauty is “universal” (§6, 5:211) because, in
experiencing it, I take it that everyone “ all human beings “ ought to feel
the same pleasure when confronted with the same object. This second
sense of “universal” is unlike the ¬rst in that it alludes not to a plurality of
objects, but rather to a plurality of subjects. Saying that my judgment of
beauty is universal in this sense “ or as Kant also puts it, universally valid “
is a matter of saying that it should be shared by everyone who judges the
My aim in this essay is to sketch a connection between these two senses
of ˜universality.™ I want to suggest that when Kant speaks of judgment as
“thinking the particular as contained under the universal,” he has the
second as well as the ¬rst sense of universality in mind. “Thinking the
particular under the universal” means not only thinking of an object as
having a feature shared in common with a multiplicity of other objects,
but also thinking of one™s own particular response to an object as universal
or universally valid, as one does in a judgment of taste. More speci¬cally,

3 There is also a related question of how we can think a particular concept or law under a
higher-level concept or law; I leave this question aside in the present essay.
4 In their translation of the Critique of the Power of Judgment, Paul Guyer and Eric Matthews
record the distinction between these senses by using ˜general™ for the ¬rst, and ˜universal™
for the second, although with some exceptions (see the translators™ notes at 8 and 66).
In discussing Kant, I will mostly use ˜universal™ for both senses but I will sometimes use
˜general™ for the ¬rst, for example in discussing Hume.
Thinking the Particular as Contained under the Universal 37

I want to suggest that the second, intersubjective sense of universality is
more fundamental in that universality in this second sense makes possi-
ble universality in the ¬rst sense. It is only because we can think of our
responses to objects as universal in the sense of being intersubjectively
valid that we are capable of thinking particular objects under universals
in the sense of subsuming them under concepts that capture what they
have in common with other objects.5

I want to begin laying out this connection by describing a familiar prob-
lem that arises for Kant in connection with the ¬rst kind of universality,
a problem that I shall refer to as the problem of empirical universality
or empirical generality. The problem is that of how to account for the
possession of empirical concepts, that is, concepts that are acquired on
the basis of experience as opposed to originating a priori in our cogni-
tive faculties. Experience for Kant consists in the ¬rst instance of those
representations that come to us because of the way in which our senses
are affected, that is, sensible intuitions. And as Kant emphasizes, sensible
intuitions are, in themselves, singular. To the extent that we regard expe-
rience as consisting in sensible intuitions alone, experience can acquaint
us only with individual things, not with features or properties that they
possess in common with other things. Experience can be the source only
of singular representations, not of representations that are general or
So far, this statement of the problem is too simple. For as Kant makes
clear in the Analytic of the Critique of Pure Reason, experience involves
not just the reception of representations in sensibility, but an activity
of imagination, called synthesis, through which the manifold of sensi-
ble impressions is given order and unity. Experience, understood as the
product of this activity, still consists in intuitions “ that is, of singular
representations “ but these intuitions are structured or synthesized by
the imagination in a way that allows for the representation of generality.
Speci¬cally, Kant holds, the synthesis of imagination proceeds according
to rules or schemata, some of which are a priori and some of which are
empirical. It is in virtue of the a priori rules that our intuitions come to
represent an objective world of causally interacting substances standing

5 I will be defending this claim only for empirical concepts, although I believe that it holds
also for the pure concepts of the understanding.
Hannah Ginsborg

in spatiotemporal relations to one another. These rules are the schemata
in virtue of which the pure concepts of understanding are applicable
to experience. But there are also rules or schemata corresponding to
our empirical concepts, and it is in virtue of their accordance with these
rules that our intuitions come to represent objects as having determinate
empirical features, for example as having qualities like red or belonging
to kinds like dog or house. If we consider experience as consisting not just
in raw, unsynthesized data, but rather as the product of our imaginative
activity, then it would seem that experience does make possible the repre-
sentation of empirical features. For in that case, it would seem that we can
arrive at empirical concepts by re¬‚ecting on, and thus making explicit,
the rules governing our imaginative activity.
But while this quali¬cation is necessary if we are to understand Kant™s
position, it does not resolve the problem. For now we are faced with the
question of the source of these rules. The rules themselves, it would seem,
cannot derive from experience regarded as the product of imaginative
activity, since they are required for the possibility of this activity and of
the experience to which it gives rise. But since they are no less general
or universal than the concepts that they are supposed to make possible,
it is no less problematic to regard them as derived from the raw material
of sensibility.
We can get clearer about the problem by looking at the passage where
Kant appears to offer his most explicit account of the formation of empir-
ical concepts. This is §6 of the Logic, where Kant describes what he calls
the “logical acts” of comparison, re¬‚ection, and abstraction. He illus-
trates these acts, which he ascribes to the understanding, in the following
often-quoted example:

I see e.g. a spruce, a willow and a linden. In ¬rst comparing these objects among
themselves, I notice that they are different from one another with respect to the
trunk, the branches, the leaves and so forth; but now I go on to re¬‚ect only on
what they have in common, the trunk, the branches, the leaves themselves; and I
abstract from their size, shape and so forth; thus I receive [bekommen] a concept
of tree.(§6, note 1; 9:94“5)

The idea behind this example seems to be that we acquire the concept of
a tree by being presented with a ¬nite number of trees and noting both
the features that differentiate them (for example, the shapes and sizes
of their respective leaves and branches) and the features that they have
in common (for example, the fact that they have leaves and branches in
the ¬rst place). By abstracting from the features that differentiate them
and attending to the common features we arrive at the concept of a tree,
Thinking the Particular as Contained under the Universal 39

which presumably can be characterized as the concept of a thing with
leaves, branches, and a trunk.6 But this example does not yield a satis-
fying account of how we arrive at empirical concepts. In the ¬rst place,
the example assumes that we are capable at the outset of recognizing
what is presented to us as having leaves, branches, and a trunk, and this
would seem to presuppose that we possess the concepts leaf, branch and
trunk. So we need to explain the acquisition of these concepts on the
basis of further concepts, and a regress threatens. Now it might be sup-
posed that Kant is in fact committed to the view that sensibility gives us
basic features such as color and shape, and that the operations of com-
parison, re¬‚ection, and abstraction are responsible for the formation of
more sophisticated concepts from these basic ones. On this supposition,
Kant holds something like the compositional view of concepts suggested
by Locke™s distinction between simple and complex ideas, a distinction
taken over by Berkeley and Hume. Sensibility is capable of giving us cer-
tain basic features, or respects in which objects resemble one another;
imaginative or intellectual activity is required only in the formation of
the more complex concepts or ideas through which “ for example “
objects are sorted into higher-level kinds characterized by a multiplicity
of features. But this seems to be precluded by Kant™s familiar view that
intuitions without concepts are “blind,” which suggests that intuitions on
their own could not give us features of objects, even simple features like
color and shape. And an example from student notes on Kant™s logic lec-
tures tells against the compositional picture by suggesting that an activity
of comparison is required, not only for the acquisition of higher-level
concepts characterized by a multiplicity of features or marks, but also for
arriving at the apparently simple concept red.7
A second dif¬culty that arises in connection with the tree example
is that even if we assume that we possess the concepts of leaf, branch,
and trunk, the example gives no indication of why our experience of
the three trees should give rise to a concept involving just these features,
as opposed to the many other features that those three trees have in
common. For example, a spruce, a willow, and a linden typically have in
common that they lack edible fruit, that they afford a degree of shelter
from the rain, that they are composed of woody material, and that insects
live in them. So why do we not attend to these features so as to arrive

6 See B´ atrice Longuenesse™s helpful discussion (1998, 115“16); I agree with her view that
comparison, re¬‚ection, and abstraction should be seen as aspects of a single activity. For
other discussions of this passage see Pippin (1982, 112ff.), Ginsborg (1997a, 53), and
Allison (2001, 21ff.).
7 Wiener Logik, 24:904“5.
Hannah Ginsborg

at a concept that would include the particular trees presented to us, but
also exclude fruit trees and include wooden houses? It is hard to suppose
any explanation for our privileging the tree-characterizing features other
than that we are already in some sense representing the sample objects
as trees, so that possession of the concept tree is already assumed from
the start. Now it might be objected that this problem derives from the
arti¬ciality of the example. In real life we derive the concept tree from
exposure to a much larger sample of trees, and any child who began asso-
ciating the word ˜tree™ with houses, or refusing to apply it to apple trees,
would very quickly be corrected. But what is common to the example and
to real life is that the number of trees we have to go on is ¬nite. And it
is always possible “ using the sorts of maneuvers typi¬ed by Goodman™s
grue and Kripke™s quus “ to come up with any number of features held
in common by a ¬nite group of objects, so that any ¬nite sample can be
regarded as exemplifying any number of nonstandard kinds.
The upshot of this seems to be that we cannot regard the appeal to
comparison, re¬‚ection, and abstraction as constituting Kant™s answer to
the question of how empirical concepts are possible, but only as explain-
ing how concepts we already possess can be clari¬ed or made explicit.8
That is, Kant™s account is not meant to explain how we come to possess
the capacity to represent the objects in question as trees, but rather how
we move from our implicit grasp of them as trees to an explicit under-
standing of the concept tree: that is, a grasp of the concept that allows us
to specify criteria for a thing™s being a tree. Another way of putting the
point is to say that the operation of comparison, re¬‚ection, and abstrac-
tion presupposes that our experience of the trees is already the prod-
uct of synthesis according to empirical schemata. To put the point in
terms of B´ atrice Longuenesse™s useful distinction between two senses of
˜concept,™ it explains how we move from the possession of an empirical
concept understood as a schema or rule for synthesis to possession of an
empirical concept understood as a discursive rule for inference. But this
means that we need to ¬nd another answer to what now emerges as the
more fundamental question about concept acquisition: How are we to
account for our possession of the rule or schema that enables us to see
the presented object as a tree in the ¬rst place?
A suggestive proposal made by Longuenesse and taken up by Henry
Allison is that we can understand the schemata as generated by the very
same act of comparison by which we move from schemata to discursive

8 This is Pippin™s view (1982, 113).
9 1998, 46“7.
Thinking the Particular as Contained under the Universal 41

concepts. Longuenesse understands the act of comparison as a compar-
ison of schemata, but she says that “to compare schemata . . . is ¬rst of all
to generate these schemata” so that “the schemata result from the very acts
of universalizing comparison of which they are the object” (116“17).10
To paraphrase, it is only through our comparison of schemata that the
schemata come into being in the ¬rst place. This formulation is, on the
face of it, paradoxical: How can we compare rules that do not exist prior
to the comparison? But it hints at a bold strategy for resolving the dif¬-
culty: namely, to understand the rules of synthesis as existing not prior to,
but in virtue of, our awareness of our synthesis as rule-governed. In other
words, the activity of re¬‚ection on our synthesis, through which we arrive
at the awareness of it as governed by rules, is precisely what is responsible
for the rule-governed character of our synthesis in the ¬rst place.
As will become clear later, I am very sympathetic toward the general
strategy that I take to be suggested by Longuenesse™s proposal. But I ¬nd
it hard to see how the speci¬c proposal itself can be successful. Even if
we accept the general point that there can be no rules without awareness
of our activity as rule-governed, it is not clear how that awareness can in
turn depend on a comparison of the very rules that it supposedly makes
possible. In other words, it is hard to see how the activity of comparison
that Kant describes in the Logic “ that is, a comparison of perceptually
represented objects to see what they have in common “ could take place
without antecedent schemata, and hence how it could be responsible for
them. For, as we noted, this kind of comparison seems to presuppose
awareness of what is presented to us as having the feature corresponding
to the concept to be made explicit, and that in turn seems to presuppose
a prior synthesis of the manifold according to that concept.
Moreover, Longuenesse herself seems to draw back from this strong
proposal by suggesting that the rule is in some sense present prior to
the act of comparison, although in an attenuated sense. Thus she says
that the rule is present in intuition prior to the act of comparison, albeit
“unre¬‚ected” and “obscure” (118). Although it lacks the “form of uni-
versality,” which it can have only insofar as we have a clear consciousness
of it, it is still, as she puts it, “universal in itself ” (ibid.). In another con-
text, she describes our apprehension in intuition as “guided” by the rule
(49): something that would seem to preclude the possibility of the rule
itself being yielded by a comparison of intuitions, since it would appear
to demand that we grasp the rule prior to our activity of apprehension.
This implies that she is committed, after all, to the presence of the rule

10 Page references to Longuenesse are to her Kant and the Capacity to Judge (1998).
Hannah Ginsborg

prior to the act of comparison, so that something other than comparison
is required for its generation. One account she gives of the origin of this
rule appeals to what she calls an “embryonic” form of comparison, which
exists in sensibility itself (114n.25). Similarly, we engage in what she calls,
following Moritz Steckelmacher, a “silent judging,” which is governed
by, and teleologically oriented toward, conscious acts of judging (122).
This suggests that we acquire the schemata not in virtue of the very same
logical comparison that yields empirical concepts, but through a sort of
proto-comparison, which precedes that full-¬‚edged comparison. More
generally, it suggests that we can understand the acquisition of empiri-
cal schemata as a subconscious process, one conceived on the model of
the conscious processes by which we clarify concepts and combine them
in judgments. But if there is a subconscious process responsible for the
initial acquisition of empirical concepts, it is hard to see how we could
understand it on the model of the conscious comparison and re¬‚ection
through which concepts are clari¬ed. For that conscious comparison,
in contrast to the subconscious comparison supposedly responsible for
schemata, depends on our possession of representations that are already
intrinsically conceptual. And it is not clear what it would be for the cor-
responding operations to be carried out on a manifold that is not yet
synthesized according to rules and so presents no general features to
serve as materials for our comparison.11

As many commentators have pointed out, the problem of empirical uni-
versality is not unique to Kant. Kant™s view that experience presents us

11 It might be thought that the question of how empirical schemata are acquired can be
answered by appeal to the activity of transcendental imagination in accordance with the
categories. Longuenesse herself suggests that this is at least part of the answer: A complete
account of how we acquire empirical schemata requires us to consider the “prior activity
of associative imagination, under the guidance of productive imagination” (116n29),
and it is only once we have recognized the role of the categories as “rules for forming
rules” that “we get an answer to the question, How do empirical concepts themselves
emerge?” (51n.25). However, as I have argued in “Lawfulness without a Law” (56“7), we
cannot make sense of synthesis according to the categories unless we can make sense of
it also as governed by empirical schemata, so we cannot appeal to it independently as an
answer to the question of how empirical schemata are applied; moreover, even if we could
make sense of synthesis according to the categories alone, it would not be suf¬cient to
account for the acquisition of empirical schemata. Longuenesse also takes the “concepts
of re¬‚ection” discussed in the Amphiboly to play a role in empirical concept formation
(122ff.), but for reasons similar to those just mentioned, I do not think that they help to
address the problem with which we are concerned.
Thinking the Particular as Contained under the Universal 43

only with particulars is derived from the empiricist tradition represented
by Locke, Berkeley, and Hume, and these philosophers too are faced
with the problem of how to explain our representation of general fea-
tures common to a multiplicity of things. Locke seems to offer an answer
to the problem through his account of “abstraction,” whereby “the mind
makes the particular ideas received from particular beings to become
general” (Essay Concerning Human Understanding, II xi 9). Thus, he says,
“the same colour being observed to day in Chalk or Snow, which the
mind yesterday received from Milk, it considers that appearance alone,
makes it a representative of all that kind; and having given it the name
Whiteness it by that sound signi¬es the same quality wheresoever to be
imagin™d or met with; and thus Universals . . . are made” (ibid.). Or, to
take the more complex example given in Book III of the Essay, children
arrive at the general idea of man by observing “that there are a great many
other things in the World, that in some common agreements of Shape,
and several other Qualities, resemble their Father and Mother, and those
Persons they have been used to . . . wherein they make nothing new, but
only leave out of the complex idea they had of Peter and James, Mary and
Jane, that which is peculiar to each, and retain only what is common to
them all” (III iii 7). But these examples suggest two problems analogous
to those raised by Kant™s tree example. First, they both seem to presup-
pose an antecedent recognition of general features: We have to observe
the “same colour” in milk and snow, and we have to recognize “common
agreements of shape and other qualities” in respect of which individual
human beings resemble one another. Second, even granted that such
basic features of color and shape are given to us, it is not clear how we
can arrive at a complex general idea of man unless we already in some
sense perceive the individuals presented to us as human beings. For oth-
erwise, how could we know which of the many “common agreements” we
have observed in them belong to the concept of man and which do not?
So it seems that, after all, Locke must regard our sensory ideas as pre-
senting us with general qualities and features in spite of their supposedly
“particular” character.
The situation is no different with Berkeley, who, in spite of his vigorous
polemic against the doctrine of abstract ideas, holds essentially the same
view.12 In terms reminiscent of Locke, Berkeley says that “an idea which

12 In claiming that Berkeley™s view is close to Locke™s, I am following Michael Ayers; see Ayers
(1991), I 250“1. Longuenesse opposes Locke™s view on generality to that of Berkeley and
Hume, and sees Kant as to some degree returning to a Lockean view (see 119); as will
become clear, the view presented here disagrees with hers on both of these points.
Hannah Ginsborg

considered in itself is particular becomes general by being made to rep-
resent or stand for all other particular ideas of the same sort” (Principles
of Human Knowledge, Introduction, §12). This is possible insofar as we are
capable of disregarding certain features of the object presented by that
particular idea. For example, if we are carrying out a geometrical demon-
stration about triangles in general, we draw on an idea of some particular
triangle but without invoking in our demonstration such features as the
triangle™s being right-angled or isosceles. This is possible because a man
“may consider a ¬gure merely as triangular, without attending to the par-
ticular qualities of the angles or relations of the sides” (ibid., §16). But if
this account is understood as addressing the problem of empirical uni-
versality, it raises the same dif¬culties we saw with Kant and again with
Locke. First, it is not clear how any general features at all can be given
to us compatibly with the particularity of sensory ideas. Second, even
granted that certain basic sensory features can be given to us, it is still
not clear what allows us to privilege some features rather than others as
contributing to a higher-level property.
What about Hume? In his discussion of abstract ideas in the Treatise,13
Hume claims to endorse Berkeley™s view, which he characterizes as the
view that “all general ideas are nothing but particular ones, annexed
to a certain term, which gives them a more extensive signi¬cation, and
makes them recall upon occasion other individuals, which are similar to
them” (17). The reference to “recalling,” however, suggests that Hume is
going beyond Berkeley; and indeed, his development of the view shows
that this is in fact the case. For the way in which ideas acquire their
“more extensive signi¬cation” depends on a characteristically Humean
mechanism of customary or habitual association. According to Hume™s
account, a particular idea becomes general insofar as it is attached to
a word that in turn is customarily applied to that idea and to others
that resemble it. When we hear the word, it not only calls to mind that
particular idea but also, as Hume puts it, “revives the custom” by which
the word is used to apply to the various resembling ideas. In other words,
the hearing of the word puts the mind in a state of readiness by which any
one of the class of resembling ideas can be called to mind.14 Hume draws
out the implications of this view in his discussion of the use of ideas in
reasoning. When we reason, for example, about the nature of triangles,

13 A Treatise of Human Nature, Book I, Part I, chapter vii. Page references are to the Selby“
Bigge edition (1978).
14 Don Garrett (1997, 63) helpfully gives this class a name: the “revival set.”
Thinking the Particular as Contained under the Universal 45

we have in our mind a particular idea of a triangle, for example the idea of
an equilateral triangle of a certain size; and we initially draw conclusions
about triangles in general based on that particular idea. If, however, we
erroneously draw a conclusion that relies on some feature that is not
universal to triangles, then an idea contradicting that conclusion will
come to mind, leading us to reject it. Thus, if we claim on the basis
of our idea that the angles of a triangle are equal to one another, “the
other individuals of a scalenum and isosceles, which we overlook™d at
¬rst, immediately crowd in upon us, and make us perceive the falsehood
of this proposition” (21).15
Despite the references to custom, which distinguish Hume™s view from
that of Locke and Berkeley, the view is often thought to suffer from the
same problem. Hume begins his account of the formation of general
ideas by saying that “when we have found a resemblance among sev-
eral objects that often occur to us, we apply the same name to all of
them” (20). This seems to imply that the customary use of the name,
and the associated disposition for recalling ideas to mind, depend on
the antecedent recognition of a resemblance among the relevant ideas.
And this in turn seems to assume that we already have a general idea,
namely, one of the respect in which the particular ideas resemble one
another. Put in terms of Hume™s example, the problem is that we cannot
acquire a custom of calling all triangles by the same name, and relatedly a
disposition whereby a particular idea of one triangle calls other triangles
to mind, unless we already possess the general concept of a triangle.16
But although Hume™s reference to “¬nding a resemblance” does appear
to lay him open to this objection, there is another way of understanding
Hume™s view on which the problem does not arise. On this interpretation
of Hume, the acquisition of the relevant custom does not depend on an
antecedent recognition of resemblances among our ideas. Rather, it is
a basic psychological fact about us that our associations of ideas follow
certain regular patterns, so that, for example, the idea of a particular
triangle will naturally call to mind ideas of other triangles in preference,
say, to ideas of quadrilaterals or circles or indeed things that are not plane
¬gures at all. It is because of these natural patterns of association that,
once the word ˜triangle™ has been applied to a representative sample of

15 For an illuminating discussion see Broughton (2000).
16 See Kemp Smith (1940, 260). Henry Allison raises this objection and also a related one:
How can the idea of an isosceles or scalene triangle, called to mind, be recognized as a
counterexample unless we already recognize it as a triangle? (2001, 23).
Hannah Ginsborg

triangles, we will become disposed to apply it to triangles generally; and,
relatedly, that when we entertain hypotheses involving the word ˜trian-
gle™, it is precisely ideas of triangles that we are disposed to call to mind
as potential counterexamples. “Finding a resemblance” among triangles,
on this reading, does not precede the acquisition of the correspond-
ing disposition; rather, acquiring the disposition is just what ¬nding the
resemblance consists in.17
If we understand Hume in this way, then his account of empirical
generality is very different from that of Locke and Berkeley, and different
in a way that bypasses the problem. Hume™s view does not presuppose
the representation of empirical generality, but rather accounts for it by
exploiting the generality of a custom or disposition. Reverting to the tree
example from the previous section, we represent the general concept tree
insofar as we entertain an idea of one particular tree accompanied by a
state of readiness to call to mind ideas of other particular trees: a state of
readiness that is in turn possible because we have acquired a disposition
to associate different ideas of trees with the same general term and hence
with one another. Such a disposition is general because it is inde¬nite in
scope. The ideas we are disposed to call to mind in connection with our
initial particular idea and its associated general term need not be limited
to ideas of trees we have actually experienced, still less to ideas of trees
that have been expressly associated with the word ˜tree.™ But this does
not prevent our acquiring the disposition on the basis of exposure to a
limited sample of trees. And if acquisition of the general idea or concept
can be identi¬ed with acquisition of the disposition, then the problems
noted in connection with Kant™s use of the example can be avoided. On
coming to associate the word ˜tree™ with spruces, willows, and lindens,
most human beings will in fact form a disposition such that the same

17 The dispositionalist position ascribed to Hume on this interpretation has some af¬n-
ity to the “psychological nominalism” that, according to Wilfrid Sellars, we arrive at
through “modifying” Hume™s view (§29 of “Empiricism and the Philosophy of Mind,”
1963, 160“1). Broughton allows it as a possible reading but does not herself endorse it.
One commentator who does endorse a dispositionalist reading of I vii is Ayers (1991, I,
257), although his account differs from the present view, and from that of most other
commentators, in taking Hume™s concern in the passage to be not the problem of empir-
ical generality, but rather the problem of how a priori knowledge is possible. It is not
essential to the argument of this essay that the interpretation described here does in fact
correspond to Hume™s view. However, in spite of the fact that the reading does not ¬t
perfectly with the text of I, vii itself, I think there is a case to be made for its adoption
on the grounds that it coheres well with Hume™s naturalistic outlook overall and, in par-
ticular, with his denial that there is any difference in principle between human reason
and the reason of animals (see III, xvi).
Thinking the Particular as Contained under the Universal 47

word will call to mind fruit trees but not wooden houses. And that is just
to say “ on the interpretation of Hume I am suggesting “ that they will
acquire the general concept tree.
So understood, however, Hume™s account suffers from another kind of
problem. We can describe the problem by saying that, even though there
is generality in the account, it is in the wrong place: It does not enter into
the content of our ideas, but rather it is external to them. The account
is supposed to explain how a particular idea can become “general in its
representation”: how it is that in having the idea of a particular tree or
triangle, I come to represent the general property of being a tree or a
triangle. But why should the representational character of a particular
idea be transformed in this way simply by being accompanied by a state
of readiness to call to mind other particular ideas? We might try to answer
this question by saying that it is the awareness of my own state of readiness,
rather than the particular idea itself, that constitutes my possession of the
general idea. In entertaining the particular idea of a tree or triangle, we
might say, I feel myself impelled or driven to call to mind other ideas
of trees or triangles, and it is in that feeling that the representation of
an object as a tree, or as a triangle, consists. However, this answer is
unsatisfactory, since the representation of a general feature of objects
seems to require more than the awareness of a subjective tendency to
associate ideas. Perhaps the idea of a willow tree always brings to mind
childhood picnics or the thought of linden trees inevitably reminds of
me of Berlin; and perhaps I am well aware of these patterns of association
among my ideas. This does not mean that I recognize a feature common
to willows and picnics or to lindens and Berlin. Even if I can explain my
tendencies of association in terms of objective relations among the things
represented by my ideas (for example, that my family used to have picnics
near a willow tree or that Berlin™s most famous avenue is planted with
lindens), my awareness of those tendencies does not constitute a grasp of
any feature or relation belonging to the objects represented. All that I am
aware of is something about my own psychological makeup, and it is not
clear how such an awareness could ever amount to the representation of
general features belonging to things independently of me.18

18 Perhaps the account could be modi¬ed to accommodate these cases of idiosyncratic
association by supposing that awareness of an associative tendency amounts to the rep-
resentation of a general feature only if I can rule out the tendency™s being due to some
particular quirk of my psychology. Thus modi¬ed, the account says that I represent some-
thing as a tree if I not only call to mind other trees in association with it, but also take
myself, in so doing, to manifest a tendency that is part of human nature, in the sense that
Hannah Ginsborg

We have now considered two pre-Kantian, and more speci¬cally empiri-
cist, positions on the question of empirical generality or universality.
The ¬rst is that of Locke and Berkeley, and it can also be ascribed to
Hume if Hume™s view is understood according to what we might call an
˜intentionalist™ reading. The second is the position occupied by Hume
if we understand him on the alternative, ˜dispositionalist™ reading. On
the ¬rst position, despite the supposed particularity of our sensory ideas
(or in Hume™s case, impressions), they present us not only with individual
things, but also with general features common to a plurality of things. On
the second position, the particularity of ideas remains unimpaired, and
the possibility of general ideas is accounted for in terms of the possession
of general dispositions to associate particular ideas in determinate ways.
But neither of these positions is satisfactory: On the ¬rst, the representa-
tion of empirical generality is invoked too soon, whereas on the second,
it fails to make any appearance at all.
Is there any alternative? I think that there is, and I want to characterize
it by taking as a starting point the dispositionalist reading of Hume.19 Let
us go back to the suggestion that we can account for my entertaining the
general idea tree by supposing that I have an idea of some particular tree,
coupled with a state of readiness to call to mind ideas of other trees. And
let us suppose that on some occasion I do have a particular idea of a tree,
say a linden, and that due to my having the relevant disposition, an idea of
some other tree, say a sycamore, comes to mind. Now, as we saw, the prob-
lem with this suggestion is that, even if we add some kind of awareness
of being impelled to think of the sycamore, the most that this account
can give us is a recognition of a certain psychological tendency in myself.
It does not give us what we want, namely the recognition of something
common to the linden and the sycamore. But what if we supplement the
suggestion by adding that, when the idea of the sycamore comes to mind,
I take its appearance in my mind to be appropriate? More speci¬cally, what
if we say that I take the idea of the sycamore to be the upshot not merely
of a certain tendency in myself, but of a tendency that is universally valid

it is common to all or most human beings. But the question remains: Why should that
amount to representing a feature or property of things, as opposed to a psychological
tendency in myself (albeit one shared by human beings in general)? Even if we rule out
idiosyncratic associations, it is hard to see how simply being aware of a tendency to call
certain ideas to mind could amount to the awareness of a general feature that the objects
of the ideas have in common.
19 From now on I shall refer to Hume, taken on this reading, as “Hume” tout court.
Thinking the Particular as Contained under the Universal 49

in Kant™s sense: a tendency that everyone ought to feel when entertain-
ing the idea of a linden? If we amend the suggestion in this way, then
we can address the problem by saying that my awareness goes beyond a
recognition of actual psychological processes and tendencies in myself.
In contrast to the problematic examples of willows and picnics, or lindens
and Berlin, I take it not only that I myself have a tendency to associate
the idea of the linden and the idea of the sycamore, but also that this
association between ideas is appropriate, or conforms to an intersubjec-
tively valid standard governing how these ideas ought to be associated.
I take it that these ideas are not merely associated in my own mind, but
that they belong together in the sense that everyone ought to feel the
same tendency to associate them as I do. And this makes it much less
implausible to suppose that my awareness could amount to a grasp of an
objective feature shared by the sycamore and the linden themselves.
Now I want to propose that this amended suggestion represents, at
least in part, Kant™s solution to the problem of empirical generality or
universality. More precisely, I want to see Kant as adopting a Humean view,
but with two signi¬cant modi¬cations. First, Kant expands the role that
Hume had ascribed to the association of ideas, holding that dispositions
to associate ideas are required not just for general thought and belief,
but also for perception itself. So for Kant, it is not just in thinking about
trees that we are in a state of readiness to call to mind particular ideas
of trees; rather, the very perception of a tree involves the activation of a
disposition to call to mind previous representations of trees. Second, and
more importantly for the purposes of this essay, Kant gives the Humean
view a normative twist. My perception of a tree not only involves my being
in a state of readiness to call to mind “ or in Kant™s terms to “reproduce” “
representations of other trees; it also involves my taking it that, insofar
as I do call ideas of other trees to mind, I am doing what I and everyone
else ought to be doing under the circumstances. The generality of my
disposition is thus, so to speak, incorporated into my perception rather
than remaining external to it as on the Humean view. I see the tree as a
tree in virtue not merely of my state of readiness to call to mind previously
perceived trees in connection with it, but also of my awareness that this
state of readiness is appropriate given my present perceptual situation.20

20 I do not mean to claim here that the two conditions mentioned in this sentence are
suf¬cient for the representation of generality. There are many cases in which someone
might associate ideas in a certain way, and take her associations to be appropriate rather
than idiosyncratic, without her representing the objects of her ideas as having a common
Hannah Ginsborg

To see what might lead us to understand Kant in this way, let us go back
to the discussion of Kant that we left at the end of Section II. We saw in
Section I that the procedures of comparison, re¬‚ection, and abstraction
described in the Logic do not by themselves explain how the acquisition
of empirical concepts is possible. Rather, they presuppose that we already
possess empirical concepts in the form of schemata, that is, rules for the
imaginative synthesis of the manifold. They explain how we arrive at an
explicit understanding of these concepts, one that enables us, for exam-
ple, to grasp that something counts as a tree if it is the kind of thing
that has leaves, branches, and a trunk. But they do not explain how we
come to be able to see something as a tree in the ¬rst place, since that
is accounted for in terms of the imagination™s activity in accordance with
rules. We ¬nd Kant™s most detailed account of this activity in the section
of the ¬rst edition Transcendental Deduction entitled the “Threefold
Synthesis.” The account of imaginative synthesis that Kant gives in this
section is extremely complex, but at its center is an activity that he calls
the “reproduction of the sensible manifold”. This is an activity of recalling
previous perceptions, where the recall involved is of two different kinds.
In the ¬rst, we call to mind the perceptions that immediately preceded a
current perception in order to form a coherent image. For example, in
order to perceive a line, we must “reproduce” the previously perceived
segments alongside the currently perceived segment. In the second,
we call to mind representations of previously perceived objects of the
same kind as the one we are now perceiving. This allows us to repre-
sent the object of our current perception as having features that do not
impinge on our senses at the time of perception, but that we nonetheless
perceive as belonging to the object. For example, I can perceive a body
as impenetrable even though I do not touch it because, in perceiving it
visually, I also call to mind perceptions of other bodies in which their
impenetrability did impinge on my senses. Similarly, I might see a distant

feature: for example when the perception of a tree calls to mind birds or lumber. What
is required further, on the view I am presenting, is that the subject™s awareness of the
appropriateness of her associations be ˜primitive™, that is, not based on the prior appre-
ciation of some fact about the world that legitimizes the association. In the example
given, the subject presumably takes her associations to be appropriate on the grounds
that birds live in trees and trees can be made into lumber. But in the kinds of cases that I
take Kant to have in mind as accounting for the representation of generality, the subject
cannot cite any reason for the appropriateness of her associations. I discuss this ˜prim-
itive™ appreciation of appropriateness in Section V, in connection with the question of
whether one can take one™s associations to be appropriate without antecedently grasping
a rule in virtue of which they are appropriate.
Thinking the Particular as Contained under the Universal 51

tree as having leaves even though I am suf¬ciently far away that a homo-
geneous mass of green would make the same sensory impression. What
allows me to see it as having leaves, as opposed to being draped with
green fabric, is that in seeing it, I reproduce previous representations of
trees in which the distinctness of the leaves directly affected my sense
Both these kinds of reproduction clearly have some af¬nity with the
association of ideas as Hume conceives it. And the second, in particu-
lar, is reminiscent of the kind of association invoked in Hume™s account
of how particular ideas become general. But the differences may seem
too pronounced for it to be possible to assimilate Kant™s view to a ver-
sion of Hume™s, even taking into account the two modi¬cations I men-
tioned earlier. The most important difference has to do with the rule-
governed character of reproduction. For Kant, at least as he is standardly
understood, our imaginative activity of reproducing representations is
not the effect of habit or custom, as the corresponding associations are
for Hume, but is carried out in accordance with a previously grasped
rule. In other words, imagination is guided in its reproductive syntheses
by understanding: and this guidance takes place in virtue of our grasp of
concepts, both pure and empirical. This presents a sharp contrast with
Hume, for whom associations of ideas are not guided by any intellec-
tual faculty but are simply a result of blind dispositions, like those of
But we can also read Kant in a way that brings him closer to Hume
while still doing justice to the rule-governed character of our reproduc-
tive associations. For the claim that our activity of imagination is governed
by rules does not necessarily imply that our activity must be guided by
those rules. Nor does it imply that the activity cannot be, as on Hume™s
view, the expression of natural dispositions of the kind that are shared by
animals. On the reading that I am proposing, the activity of reproductive
synthesis, like the association of ideas for Hume, is simply something that
we are naturally disposed to do. It is a natural psychological fact about
human beings that, if shown a certain number of trees, they will develop
a disposition such that the perception of one tree will tend to call to mind
other previously perceived trees. What makes the corresponding associa-
tions rule-governed is not that they are guided by a speci¬c, antecedently
grasped rule, but rather the fact that we take them to have normative sig-
ni¬cance. The associations are rule-governed because in carrying them
out I take myself to be doing not only what I am disposed to do, but
also what I (and everyone else) ought to do. That is, I take my actual
Hannah Ginsborg

associations, blindly habitual though they are, to manifest conformity to
a normative standard applicable to everyone. The rule-governedness of
my associations is thus a function of my taking them to be rule-governed,
which is in turn a function of my taking my natural dispositions as exem-
plifying a universally valid norm.
Part of the appeal of this reading is that it offers an answer to the
question that remained unsolved at the end of Section I. That was the
question of how to account for our possession of the rules governing
the synthesis of the manifold, in particular those rules identi¬able with “
or at least corresponding to “ empirical concepts. Seeing something as
a tree requires that we synthesize the manifold according to a certain
rule corresponding to the concept tree. But how could we come to grasp
such a rule antecedently to an experience in which we see something as a
tree? The dif¬culty here dissolves if we reject the assumption that the rule
must be grasped antecedently to the experience, and more speci¬cally to
the synthesis that makes the experience possible. Once this assumption
is rejected, we do not need to explain how the rule can be acquired
antecedently to the synthesis. Instead, we can say that the rule is acquired
insofar as the subject acquires the disposition that makes the relevant
kind of synthesis possible. I acquire the rule tree, and hence become
capable of seeing things as trees, by acquiring the disposition to associate
different representations of trees with one another and, more speci¬cally,
to reproduce past perceptions of trees when a particular tree is presented
to me.21 But this is possible only because I take a certain attitude toward
the disposition, namely, that the associations I am disposed to carry out
in accordance with the disposition conform to a normative standard that
is universally valid. It is only because I regard my actual associations as
expressing how I (and everyone else) ought to associate representations
that my coming to be disposed to associate representations in that way
amounts to the acquisition of a rule according to which they ought to be

21 This is somewhat oversimpli¬ed. For one thing, it applies only insofar as concepts are
observational. To the extent that a concept is theoretical, possession of that concept,
even in a minimal sense, will require more of a capacity to articulate criteria. Second,
depending on the context, we might invoke more or less stringent requirements for
concept possession: For example, we might say that a child has the concepts of solid, liquid,
and gas if she can reliably sort things into the appropriate categories while imposing more
demanding requirements on a student of advanced chemistry.
Thinking the Particular as Contained under the Universal 53

In the previous section, I suggested that we view Kant™s account of empir-
ical generality as a modi¬cation of the dispositionalist view I ascribed to
Hume. Grasping an empirical concept involves, as on Hume™s account,
the possession of a disposition to associate one™s representations in cer-
tain determinate ways; but it also involves taking one™s associations to be
as they ought to be, that is, to manifest conformity to normative stan-
dards. Reading Kant in this way helps us to see how his identi¬cation
of empirical concepts with rules for synthesis can serve as an answer to
the problem of empirical generality. For it suggests that this identi¬ca-
tion need not require the possession of concepts prior to synthesis, but
merely that the subject be capable of regarding her activity of synthesis
in normative terms. However, this reading rests on a philosophical pre-
supposition that is likely to strike readers as problematic. My reading is
based on the suggestion that we can account for a subject™s grasp of a rule
in terms of her adopting a normative attitude toward her mental activity.
This suggestion presupposes that we can make sense of her as adopting
this normative attitude without in turn assuming that she grasps a speci¬c
rule to which her mental activity is subject. But it might be protested that
this is impossible. How can I take an association of ideas to be appropri-
ate if I don™t antecedently have in mind some speci¬c rule with which it
accords? For example, how can I take my association of the idea of linden
with the idea of a sycamore to be appropriate if I do not already think of
the association as governed by the concept tree?
This protest might re¬‚ect two different kinds of worry. The ¬rst stems
from the fact that a single idea, for example the idea of a linden, might
be associated on various occasions with ideas of many different kinds of
things: ideas of lindens, of other deciduous trees, of other trees more gen-
erally, of living things, and so on. Given this, it might seem that recalling
the idea of a sycamore constitutes an appropriate association only on the
assumption that the operative rule is, say, tree or deciduous tree. In another
context, say one in which the linden is presented as an example of its
particular species, the association with a sycamore would be inappropri-
ate. Generalizing this ¬rst worry, it might seem that, depending on how
the context is characterized, any arbitrarily speci¬ed association might be
made out to be either appropriate or inappropriate. For example, if the
linden is presented as an example of something that is wooden, harbors
insects, and has no leaves in winter, then the association with the idea of
Hannah Ginsborg

a house would seem to be appropriate and the association with the idea
of spruce would seem to be inappropriate. So it might seem that the view
I am suggesting does not avoid the problem of concept acquisition noted
in Section I.
However, this worry, at least in its generalized version, overlooks the
crucial role I am ascribing to natural dispositions. The account I am sug-
gesting depends on the idea “ implicit in Hume “ that our capacity to
form associative dispositions is limited. As Hume points out in the chapter
of the Treatise discussed earlier (Book One, I vii), “the idea of an equilat-
eral triangle of an inch perpendicular may serve us in talking of a ¬gure,
of a rectilinear ¬gure, of a triangle, and of an equilateral triangle” (20).
This is possible because each of the corresponding terms corresponds to
a speci¬c disposition: “all these terms,” he says, “excite their particular
habits” (ibid.). But it seems clear that the list could not be expanded
inde¬nitely, since there is a limited number of habits that we are natu-
rally inclined to form in connection with the idea Hume describes. Even
though the equilateral triangle Hume describes is an instance of the con-
cept equilateral or ¬ve-sided that disjunctive concept does not correspond
to a natural disposition: We do not, without special training, form the
habit of associating equilateral triangles with, say, irregular pentagons
and regular hexagons to the exclusion of oblongs and isosceles triangles.
Given this kind of limitation, there is no reason why we cannot say of each
of this ¬nite set of habits or dispositions that the corresponding associ-
ations are appropriate. It is true that if someone misidenti¬es a linden
as a sycamore because her disposition to associate ideas of various kinds
of trees leads her to call to mind the idea of a sycamore when presented
with a linden, she is doing something inappropriate, namely making a
false claim. But that does not show that the association itself cannot be
regarded as appropriate. For the association is just a particular manifes-
tation of the disposition in virtue of which she sees the linden as a tree:
and the actualization of that disposition is appropriate no matter what
the context.22
There is, however, a second and more abstract worry that might be
raised about the presupposition under discussion, namely that it is inco-
herent. It might be claimed that it simply does not make sense to suppose

22 It might be objected that the idea of a natural disposition is itself problematic, or at least
cannot bear the weight that is being placed on it in this account. I will not try to address
this line of objection here, except to say that my appeal to natural dispositions in the
context of this account derives some support from Graeme Forbes™s (2002) defense of
a dispositionalist account of rule-following.
Thinking the Particular as Contained under the Universal 55

that we can think of a thing as conforming to a normative standard with-
out ¬rst having in mind the idea of a speci¬c rule or standard to which it
is antecedently subject. In the present context, the claim would be that
the very idea that my mental activity is as it ought to be presupposes the
antecedent idea of a rule or concept that dictates how it ought to be. But
it is not obvious why this should be so. It seems to me that we do in fact
often take our associations to be appropriate without being able to rec-
ognize speci¬c respects in virtue of which they are appropriate. Indeed,
the possibility of this kind of normative awareness is routinely assumed in
introducing children to new concepts. Six-year-olds learn the concepts
solid, liquid, and gas, say, by being presented with objects that they are
asked to sort into kinds: Does the chalk “belong with” the stone, the bot-
tle of water, or the balloon? This kind of procedure relies not just on the
child™s being mechanically disposed to sort objects in a particular way, but
also on a primitive appreciation that what she is doing is appropriate: She
recognizes that the chalk should go with the stone even if she cannot say
anything about why it should. What is going on here is not that the child
already grasps that the chalk and the stone are solid as opposed to liquid
or gaseous, and therefore should be classed together: Rather, the child
is inclined to sort the chalk and the stone together, and implicitly takes
her inclination to re¬‚ect how they ought to be sorted. Her appreciation
of the appropriateness of her sorting inclinations “ that is to say, of her
associative dispositions “ does not presuppose possession of the concept
solid, but it provides the basis on which that concept can be acquired. To
the extent that her sorting inclinations in fact lead her to discriminate
solids from liquids and gases, her recognition of their appropriateness
amounts to a recognition of her activity as both governed by and con-
forming to a rule: a rule that she can initially specify only by the example
of her own activity, but that she will later be in a position to articulate as
the concept solid.
One might be puzzled here about how a subject can take her activity
to be governed by a rule that is, in the ¬rst instance, picked out through
the example of that very activity. In order for her to take her activity to be
governed by a rule, she must be able to make sense of the possibility that
what she does might fail to accord with the rule; but how can what she
does fail to accord with a rule that is exempli¬ed by her activity itself ?23

23 This objection has been put to me in terms of Wittgenstein™s remark about a “private
language” at Philosophical Investigations §258: “One would like to say: whatever is going
to seem right to me is right. And that just means that here we can™t talk about ˜right.™”
Hannah Ginsborg

An initial answer is that, while she cannot take what she does at any given
time both to exemplify a rule and to fail to accord with that same rule, she
can still make sense of the idea of the rule™s being contravened, namely
by considering the possibility that she might act differently. For she may
take it that, if she were to act differently, she would fail to accord with
the rule that she now recognizes as governing her activity. She would, as
she sees it, be acting wrongly because she would not be acting this way.
By the same token, she may take others to be failing to accord with the
rule now exempli¬ed by her behavior, and she may take herself to have
contravened that rule on previous occasions. That is to say, she may take
others, and herself on other occasions, to be failing to do as they ought
because they are failing to do as she is doing now.
This answer might seem inadequate to address the dif¬culty. For surely,
it might be objected, the subject must recognize that, if she were to act
differently, she would still be according with a rule exempli¬ed by what
she would be doing in that counterfactual situation. Similarly, she must
recognize that others who act differently are according with rules that
are exempli¬ed by what they are doing. So it would seem that she is not
in a position to make sense of anyone™s ever failing to act as they ought:
No matter what others do, she must take them to be doing as they ought
in the sense that they are according with a rule exempli¬ed by their own
activity. And that undermines the idea that her own activity exempli¬es
a rule that is universally valid.
But the assumption underlying this objection is mistaken. If a subject
takes what she herself does in a certain situation to conform to a rule that
it exempli¬es, she will not recognize another subject™s divergent activity
as also conforming to a rule that it exempli¬es; rather, she will deny
that the other subject™s activity exempli¬es a rule at all. If, in the context
of the kind of sorting exercise I described earlier, Alma sorts the chalk with
the stone but sees another child, Bruno, sorting it with the balloon, she
will not take it that his behavior is governed by a rule that it exempli¬es,
because she does not take there to be any rule that prescribes that the
chalk ought to be sorted with the balloon. In taking it that she is sorting
the objects as they ought to be sorted, and thus as anyone ought to sort
them, she excludes the possibility that someone who is presented with the
same objects, but who sorts them differently, is also doing as he ought.
She will thus take Bruno to be failing to do as he ought, either in the sense
that there is no “ought” applicable to his behavior at all (she may think
that he has opted out of the exercise and is engaged in random play) or
in the sense that he has violated the rule that does govern his behavior,
Thinking the Particular as Contained under the Universal 57

namely the rule exempli¬ed by her own sorting activity. Whether or not
she takes his activity to be rule-governed at all, it does not, by her lights,
exemplify a rule.
It might seem problematic here that Alma has no criterion for deter-
mining that it is Bruno rather than herself who is mistaken in taking his
activity to exemplify a rule, and hence Bruno rather than herself who
must be counted as failing to do as he or she ought. For it is equally open
to Bruno to take what he does to exemplify a rule for sorting the objects
in question, and hence to deny that Alma™s sorting activity exempli¬es a
rule. We seem to be faced with the possibility of multiple subjects sorting
objects or associating representations in different ways, each taking her
own activity to exemplify a rule, and none in a position to establish the
legitimacy of her claim against those of the others. So how can any one
subject, recognizing that possibility, take her own sorting or associative
activity to be as it ought to be? Lacking a criterion, she seems to be in
no position to defend her claim in the face of disagreement from others,
and that seems to undermine the intelligibility of her claim to be doing
as she ought in the ¬rst place.
One part of my response here is simply to deny that the absence of a
criterion of correctness undermines the possibility of a subject™s intelligi-
bly taking herself to be doing as she ought. Two subjects can genuinely
disagree about what is appropriate in a given case “ and hence make con-
¬‚icting claims about which one is mistaken in taking his or her activity
to exemplify a rule “ without a criterion™s being available to resolve that
disagreement. But the other part of the response is to draw attention
once again to the role played in my account by the idea of natural dispo-
sitions and, in particular, the idea of such dispositions as shared. For the
most part, human beings naturally converge in the ways they are inclined
to sort objects and, correspondingly, to associate representations: If they
did not, we could never come to attach a common meaning to words like
˜tree™ and ˜solid.™ So disagreements like that between Alma and Bruno
rarely arise, and, if they do, they tend to be quickly resolved. With fur-
ther exposure to examples and other kinds of training, Bruno™s sorting
dispositions will naturally come into line with Alma™s and ours, so that
he will come to agree with Alma that his earlier sorting behavior failed
to be as it ought to be. The point here is not that a subject can use the
idea of “what comes naturally” as a criterion for determining whether or
not she is associating her representations as she ought. Rather, it is that
we all naturally tend to associate our representations in the same ways, so
that the need for such a criterion does not arise. The fact of our shared
Hannah Ginsborg

natural dispositions enables us to agree on which rules are exempli¬ed
by our activity overall, and hence on a shared set of concepts.
I have been defending the possibility of a subject™s adopting a nor-
mative attitude to her mental activity without any antecedent grasp of
a concept or rule determining how that activity ought to be. But could
Kant allow such a possibility? The answer is that he not only could but
does, and this brings us back to the central thesis of the essay. For his
account of judgments of beauty in the Critique of Judgment explicitly relies
on the idea that we can conceive of our mental activity to be as it ought
to be without conceiving it as governed by a speci¬c rule or concept.24
As I noted at the beginning of the essay, a judgment of beauty makes a
claim to universal validity. In taking something to be beautiful, I take it
that everyone ought to judge it in the same way that I do. But judgments
of beauty have a further feature that at ¬rst sight seems to stand in con-
¬‚ict with their universal validity. They are what Kant calls “subjectively
grounded”: Instead of ascribing an objective feature to the thing, as a
cognitive judgment would do, they re¬‚ect the subject™s own response to
the object, a response that consists, more speci¬cally, in a certain activity
of the subject™s imagination. So in making a judgment of beauty, I take it
that everyone ought to respond imaginatively to the object as I do. But
I do so without ascribing to the object a feature in virtue of which that
response is universally called for, and hence without taking the appro-
priateness of my imaginative activity to depend on its conformity to an
antecedently speci¬ed rule. A subject who judges an object to be beautiful
thus takes her mental activity to be appropriate in the primitive way that
I have described: In Kant™s words, she sees her judgment as “the example
of a universal rule which cannot be stated [die man nicht angeben kann]”
(Critique of Judgment §18, 237).
While I do not have space to go into the many complications of Kant™s
account of judgments of beauty, I want at least to note that Kant™s treat-
ment of them indicates his acceptance of the kind of normative attitude
under consideration. For Kant holds that such judgments are both intel-
ligible and in principle legitimate. The mere fact that we make judg-
ments of beauty shows that we do, under certain circumstances, take our-
selves to respond appropriately to objects, but without taking ourselves

24 For a fuller discussion of this point, see Ginsborg (1997a), especially Section IV, and
Ginsborg (1997b), especially Section V. In those articles I make a distinction between
primitive and derivative ascriptions of normativity, which is intended to address worries
of the kind discussed earlier.
Thinking the Particular as Contained under the Universal 59

to conform, in so responding, to a speci¬c rule or standard governing
the perception of the object. Moreover, he argues, we are entitled to so.
As long as my pleasure in an object is disinterested, which he takes as
implying that it does not depend on any “private condition” that sets me
apart from other human beings (Critique of the Power of Judgment §6, 5:211),
we are entitled to claim that all other human beings ought to respond
to the object in the same way that we do. I take this to suggest, in the
¬rst place, that Kant himself wants to make room for the possibility of
normative claims that do not presuppose speci¬c rules. In the second
place, albeit more speculatively, I take it to point to precisely the kind of
move embodied in what I have called Kant™s ˜normative twist™ on Hume.
For our entitlement to make judgments of beauty appears to depend on
our being entitled to take a normative attitude toward our mental activity
more generally. As long as my mental activity is not in¬‚uenced by any
factors that set me apart from other human beings, Kant appears to sug-
gest, then I can legitimately take it as representing a standard that all
human beings, myself included, ought to meet. And if that is so, then to
the extent that my dispositions to associate representations are indepen-
dent of my desires and of other contingent features of my psychology, I
can take them as exemplifying normative rules that apply to all human

25 This talk of “entitlement” may suggest a further, and still more general, worry about the
view I am ascribing to Kant. Suppose that I am right to interpret Kant as holding that
we regard our mental activity as exemplifying normative rules and that this accounts for
the possibility of grasping empirical concepts. This does not in itself seem to show that we
are entitled to take this normative attitude toward our mental activity. We are thus left with
the question of how we can legitimately take our mental activity to exemplify normative
rules, regardless of whether or not we actually do so as a matter of psychological fact.
But this question can be answered, I think, by appeal precisely to the dependence of our
grasp of empirical concepts on our adoption of this normative attitude. In other words,
we are entitled to regard our mental activity as exemplifying normative rules precisely in
virtue of the fact that our doing so is a condition of the possibility of empirical concepts,
and hence of cognition more generally. I ¬nd at least a hint of this answer at §21 of
the Critique of Judgment where Kant says that cognitions and judgments “must . . . allow
of being universally communicated . . . for otherwise they would be altogether a merely
subjective play of the powers of representation, just as skepticism demands” (5:238) and
that the “universal communicability of our cognition must be assumed in every logic and
every principle of cognition that is not skeptical (5:239). The point can be made vivid by
asking what it would be for this normative attitude to fail to be legitimate. In the case of
speci¬c concepts whose legitimacy might be called into question, in particular the pure
concepts of the understanding, we can make sense of the idea that we are not entitled
to use them: Experience might fail to present us with objects to which they apply. It is
the task of the Transcendental Deduction in the Critique of Pure Reason to rule out that
possibility. But the general principle that we are entitled to take a normative attitude
Hannah Ginsborg

We are now in a position to see the connection between the two senses
of “universal” invoked at the beginning of this essay. When Kant char-
acterizes judgment as the “capacity to think the particular as contained
under the universal,” he means to refer, at least in part, to the capacity
to think particular objects under empirical concepts. But if the view I
have attributed to him is correct, he takes this capacity to require that
we be able to think the particular under the universal in another sense,
namely, that of being able to regard certain of our psychological responses
to objects as universally valid. This suggests that the most fundamental
characterization of judgment should not be as a capacity to think objects
under concepts, as suggested by the ¬rst sense of “universal,” but as a
capacity to regard one™s mental responses to objects in normative terms,
as suggested by the second sense. For it is only by virtue of taking a nor-
mative attitude to one™s mental activity that one can regard it as governed
by rules, which in turn is required for recognizing the objects we perceive
as falling under empirical concepts.26

toward our mental activity does not purport to be an objective principle, so it does not
make sense to suppose that objects could fail to accord with it. Any attempt to show
that it is not legitimate would itself have to appeal to a normative principle governing
our mental activity and would thus be self-defeating. The point here is related to Kant™s
claim that the deduction of taste (which, as I understand it, rests on the general principle
under discussion) is “easy, because it does not have to justify any objective reality of a
concept” (Critique of the Power of Judgment §38, 5:290).
26 Earlier versions of this essay were given at the 2002 France“Berkeley Conference on
Kant and Normativity and at the University of Chicago. I am grateful to members of the
audiences on those occasions for comments and discussion, and in particular to Janet
Broughton, James Conant, John Haugeland, John MacFarlane, and Charles Travis. The
essay bene¬ted also from discussions with Janet Broughton, Quassim Cassam, Alva No¨ , e
Seana Shiffrin, and Jay Wallace, and from Rebecca Kukla™s valuable substantive and
editorial comments.

The Necessity of Receptivity
Exploring a Uni¬ed Account of Kantian Sensibility
and Understanding

Richard N. Manning

In the Critique of Pure Reason, Kant attempted to explain just how the sensi-
ble matter provided by intuition contributes to the content and ground-
ing of empirical judgment. But many commentators, both Kant™s con-
temporaries and ours, have found his answer ultimately unsatisfactory,
and have laid blame on his apparently fundamental distinction between
sensibility (the receptive faculty of intuition) and understanding (the
spontaneous faculty of concepts). For example, Reinhold (1789)1 found
the distinction and the dualisms it engenders so problematic that he pro-
posed that the idea of representation, common to both sensibility and
understanding, should supplant it as an ultimate grounding principle
for transcendental idealism. And Davidson™s famous rejection of the very
idea of a conceptual scheme (Davidson 1984) is targeted at the distinc-
tion between conceptual and experiential elements in thought, which he
takes Kant™s distinction to entail (Davidson 1999, 51). But perhaps the
commentators have been wrong, not in ¬nding fault with the idea that
these faculties and their contributions to experience and judgment are
fundamentally distinct, but in attributing that idea to Kant. In this essay,
I explore this theme. I shall ¬rst illuminate the dif¬culties for Kant™s
account as typically understood, from the standpoint of the question
of how sensibility could possibly provide the sort of grounding or guid-
ance for the understanding™s operations that could ever yield objective
empirical judgment.2 I shall then turn to John McDowell™s recent effort

1 The legitimacy of Reinhold™s complaint is discussed and debated in Ameriks (2000) and
Larmore (2003).
2 In so framing the matter, I take the lead from the early chapters of Sellars (1992).

Richard N. Manning

to overcome these worries. McDowell™s approach is to attribute to Kant,
contrary to the standard view, an originally uni¬ed view of sensibility and
understanding. As I shall endeavor to show, such a uni¬ed view, whether
or not it captures Kant™s actual view, faces a threat of idealism, which
Kant himself took very seriously and which, despite repeated efforts, he
was never fully satis¬ed that he refuted.

1. the structure of kantian cognition
and the problem of guidance
Kant says, “In whatever way, and through whatever means a cognition may
relate to objects, that through which it relates immediately to them . . . is
intuition. This, however, takes place only insofar as the object is given
to us; but this, in turn, is possible only if it affects us in a certain way”
(A 19, B33).3 Intuition is thus essentially a mode of receptivity. Our spe-
ci¬c and exclusive mode of being affected by objects in intuition is sensi-
bility. Kant de¬nes the understanding negatively as a nonsensible faculty
of cognition, and positively as the faculty of concepts. Since sensibility
alone rests on affections, the understanding is a spontaneous rather than
a passive or receptive faculty. Concepts are rules for the classi¬cation of
representations and are thus essentially general in relation to what they
classify. Intuitions, in contrast, are themselves wholly particular in the
sense that they cannot be multiply instantiated.4 Though distinct, each
of these two faculties is essential to all cognition. “Without sensibility no
object would be given to us, and without understanding none would be
thought. Thoughts without content are empty, intuitions without con-
cepts are blind” (A51/B75). By ˜content™ here, Kant means the sensible
matter provided in intuition.
But how can these two essential elements of cognition ever come
together in a way that could yield objective empirical judgments? Such
judgments “ subsumptions of intuitive representations under concepts “
amount to commitments directed toward objects in a world that is not of
our making and are answerable for their correctness to the way that those
objects are. If our judgments are to be in this way answerable to objects, it
seems they must be in some way guided or constrained by the character of

3 All parenthetical citations are to Kant™s Critique of Pure Reason unless otherwise indicated.
4 This crucial distinction between intuitions and concepts is indicated by the fact that Kant
proves that space is not a concept “ which is not to say that we have no concept of space “
on the ground that it is not multiply instantiable (all spaces being a part of space rather
than instances of distinct spaces).
The Necessity of Receptivity 63

our relation to those objects. Yet the sole basis for our relation to them is
through their effect on our sensibility in intuition, and sensible intuition
is itself blind. How could a blind sensible intuition provide such guiding
Kant is, of course, well aware of the prima facie problem he faces in
bringing together the sheer receptivity of intuition with the spontaneity
of the understanding in a way that could yield objective empirical judg-
ment. Indeed, this is the central problem of the ¬rst Critique™s Analytic of
Principles. There, Kant describes two mental functions “ the synthesis of
the imagination and schematization of concepts “ in trying to bridge the
apparent gap between sensible intuition and understanding.
Synthesis is “the action of putting different representations together
with each other and comprehending their manifoldness in one cogni-
tion” (A77/B103). While “the manifold of representations can be given
in an intuition that is merely sensible, i.e., nothing but receptivity,” “the
combination (conjunction) of a manifold in general can never come
to us through the senses” (B129). Mere intuition, while it “provides a
manifold,” can never give unity “without the occurrence of synthesis,”
which “runs through and takes together this manifoldness” (A99). This
allows the manifold contained in an intuition to be apprehended as such,
that is, as a manifold, and is a condition on the possibility of cognition
(A78“9/B104). Kant refers to the synthesis as “a blind but indispensable
function of the soul” (A78 B103).
Kant attributes the synthesis of the manifold of speci¬cally sensible
intuition, which he calls the “¬gurative synthesis,” to the imagination,
which, he claims, belongs to sensibility insofar as it is concerned with intu-
ition (B151). But he makes clear nonetheless that imaginative synthesis
belongs more appropriately to spontaneity. Imagination is a determining,
and not merely determinable, faculty and it is therefore allied to under-
standing (B151“2) Synthesis “is an act of the spontaneity of the power
of representation, and, since one must call the latter understanding, in
distinction from sensibility, all combination . . . is an action of the under-
standing” (B130). The synthesis of intuition by imagination “is an effect
of the understanding on sensibility” (B152, emphasis added). So far as it
is spontaneous, Kant calls the synthesis of the imaginative “productive.”
In light of the operation of productive synthesis, “we can represent
nothing as combined in the object without having previously combined
it ourselves . . . among all representations, combination is the only one
that is not given through objects but can be executed only by the subject
itself” (B129“30). Now since our sole mode of affection by objects is
Richard N. Manning

through this sensible manifold, and since particulars are present in the
intuition of the manifold only through the operation of the synthesis
of imagination, particulars, as such, are not present in the manifold of
intuition at all.
The problem of guidance gains purchase at the initial moment of the
cognitive process: the transition from mere sensation to intuited appear-
ance. All appearance already involves not mere receptivity, but the order-
ing and placing of that matter of sensation in relation, according to space
and time, the a priori forms of intuition (A19“21/B33“5). Sensation, as
the effect of the object on us, cannot be present to us spatiotemporally
independent of the activity of our mental faculties. How then can sheer
sensory content, which is not itself spatiotemporally arrayed, determine
the spatiotemporal structure of its appearance? The determinate charac-
ter of the sort that might conceivably provide guidance to or constrain
the way the manifold is taken up as a manifold seems to be the product
of synthesis. Yet guidance from the contents of the sensible manifold itself
seems required if synthesis is to be a part of a process that yields objective
understanding of objects sensed.5
The chapter on the schematism of concepts in the Analytic of Princi-
ples arguably presents the next stage in the progress from sheer receptiv-
ity of the sensible manifold to full, spontaneous concept application in
empirical judgment. There, Kant tells us that in all applications of con-
cepts to objects, the representation of the object must be “homogeneous”
with the concept, in the sense that “the concept must contain that which
is represented in the object” (A137/B176). The pure concepts, he says,
are “entirely unhomogeneous” in comparison with empirical intuitions.
While some of what Kant says suggests that schemata are required only
for the application of pure concepts, it is nonetheless clear that schemata
are required for the application of empirical concepts as well. Noting that
no image could be adequate to the pure concept of a triangle precisely
because images are particular in a way that the concept is not, Kant con-
cludes that what is required is the schema of a triangle, which “signi¬es
a rule of the synthesis of the imagination” (A141/B180). He then insists
that “even less does an object of experience [an appearance] or an image
of it ever reach the empirical concept, rather the latter is always related
immediately to the schema of the imagination as a rule for the deter-
mination of our intuition in accordance with a certain general concept”

5 Indeed, sheer sensibility does not even provide an occasion for synthesis, since occasions
are in time, and the temporal character of intuition is itself a product of synthesis.
The Necessity of Receptivity 65

The process of schematization is, Kant says, “a hidden art in the depths
of the human soul” (A141/B180“1). “The schema of sensible concepts,”
he continues, “is a product, and as it were a monogram of pure a priori
imagination, through which and in accordance with which the images
¬rst become possible, but which must be connected with the concept, to
which they are never themselves fully congruent, always only by means of
the schema they designate” (A141“2/B181). So schemata are represen-
tations of general procedures, and either are or signify rules that make
possible the subsumption of images under concepts. Images themselves
cannot make this possible, since they cannot “attain the generality” of
concepts. Since, in Kant™s view, schemata can make this possible, they evi-
dently do attain to that generality. And Kant indeed identi¬es schematism
as “a procedure of the understanding” (A140/B179, emphasis added), the
faculty of concepts. But it is this very generality of concepts that indicated
the need for an intermediary between concepts and intuited objects in
the ¬rst place.6 Schemata seem to be general in just the same way that
concepts are, hence equally heterogeneous with objects on this score,
and equally in need of additional terms through which they can relate
to them. In this way, schemata seem to be “ like third men in the Pla-
tonic context “ too much like the very elements between which they are
supposed to mediate to pull off the task.
To take stock, the situation appears to be this: Kant recognizes a
moment of sheer receptivity in our cognition, which moment is crucial to
our being able to think of objects at all. However, in order for receptivity to
be able to provide us with objects for thought, the received sensible man-
ifold must be worked over by the imagination through synthesis to yield
discrete, re-identi¬able particulars as contents of sensory representation,
which can then be related to concepts through schemata. But what war-
rants the claim that the matter of sheer receptivity has guided, or so much
as constrained, the formation of such contents, and hence the formation
and character of the schemata bringing them to conceptualization? And
if this claim is unwarranted, then we seem likewise unwarranted to claim
that judgment can amount to an objective response to our receptivity.
In both synthesis and schematization, all the guiding work seems to be
done by the operation of spontaneity. In each case, the product seems
to owe all of its character (if not its existence), insofar as that character

6 See here the introduction to the second book of the Analytic, where Kant argues that
judgment cannot be a matter of mere rule application, on pain of in¬nite regress
(A133/B172). This regress initiates the need for a speci¬cally transcendental logic
designed to link understanding, as a faculty of rules, to judgment as the faculty of sub-
suming something under rules.
Richard N. Manning

is relevant to its subsumption under concepts in judgment, to the fac-
ulty of concepts. Sheer presynthesized sensibility is too unstructured, as
Kant seems to conceive it, to guide the operations of spontaneity at all.
It cannot have a voice in the outcome of synthesis, namely schematiza-
tion, and, consequently, judgment, for it is dumb as well as blind. Since
concepts without intuitions are empty, and since intuition seems to have
no voice in concept application, we seem to be forced to admit that our
concepts, and hence our thoughts, are after all empty, of nothing, hence
not even possibly right or wrong.

2. rejecting sheer sensibility: unifying intuition and
Apparently, we must reject the idea of sheer unstructured sensibility. In his
Science and Metaphysics: Variations on Kantian Themes, Wilfrid Sellars (1992)
argues that Kant equivocated between two senses of ˜intuition™: When he
appeals to intuitions in satisfying empiricist and anti-idealist demands for
input from a world not of our making, he conceives them as sheer recep-
tivity, and when he appeals to them to explain the legitimacy of the appli-
cation of concepts whose ground is in the understanding, he conceives
them as the postsynthetic presentation of (at least proto-conceptually
structured) spatiotemporal particulars. But the former sort of intuitions
could not possibly guide the synthesis of the latter. Sellars contends that
Kant should have conceived sheer receptivity in terms of nonconcep-
tual impressions that possess properties analogous to those of the objects
whose conceptualization they guide. These states and their properties are
in the nature of theoretical, explanatory posits that, because their prop-
erties are conceived on analogy with those of empirical objects, explain
the structure of our perceptions of these.
While Sellars intends his suggestion as a correction rather than an
interpretation of Kant, he does think that Kant was correct in recogniz-
ing the need for guidance from some form of receptivity that is sheer in
being utterly prior to and independent of the operations of the faculty of
concepts. John McDowell (1998 “ hereafter “HWV”), in response to both
the Kantian problematic as Sellars formulates it and his dissatisfaction
with Sellars™s correction, has denied that this need for sheer sensibility
is genuine or that Kant believed it was. Recognizing the hopelessness of
trying to bridge any notion of intuition as sheer nonconceptual receptiv-
ity with understanding, McDowell claims that, despite appearances,7 “the

7 Pun intended.
The Necessity of Receptivity 67

idea that perception involves a ¬‚ow of conceptual representations guided
by manifolds of ˜sheer receptivity™ is not Kantian at all” (HWV 452). On
McDowell™s reading of the ¬rst Critique, Kantian intuitions themselves are
“shapings of sensory consciousness by the understanding” (HWV 462).
Kant™s insight, he claims, is that “the very idea of a conceptual repertoire is
the idea of a system of capacities that allows, as it were at the ground level,
for actualizations in which objects are immediately present to the subject”
(HWV 463). By ˜objects™ here, McDowell does not mean some kind of
sensible particulars that demand further shaping by the understanding
to be actualized as empirical objects, but full-on empirical objects them-
selves. Thus these shapings of sensory consciousness involve the very same
capacities of the understanding that are in play in the full, spontaneous
act of judgment. The contribution sensibility makes to experience and
judgment is not even notionally separable from the spontaneity of the
understanding as the faculty of concepts.
This reading obviously represents a radical shift from the conception
of the interplay between sensibility and understanding as I have sketched
it in this essay so far. And it alters it in a way that makes the problematic
I have exposed appear illusory. For if there are no sheer manifolds given
in intuition, then it makes no sense to demand that such sheer mani-
folds provide guidance in the application of concepts. If what is ˜given™
in intuition is already conceptually structured, then the notion that an
application of concepts to it in full, spontaneous judgment might be cor-
rect or incorrect seems to require no further defense. Assuming that it
can sustain criticism, then, McDowell™s reading of Kant is tempting.


. 2
( 10)