«ABSTRACT I argue that Kant’s distinction between the cognitive roles of sensibility and understanding raises a question concerning the conditions ...»
limited version of premise (2):
(2´): Both empirical intuition and judgment have a common source of unity—namely, discursive activity.
Restricting the argument of Intellectualism to merely empirical intuition has the advantage of avoiding the considerations militating against its application to the forms of intuition. Moreover, with regard to mereological relationships, Kant seems committed to understanding empirical intuition differently than pure intuition.
He says, “I call an extensive magnitude that in which the representation of the parts makes possible the representation of the whole” (A162/B203). Since Kant claims all intuitions are extensive magnitudes (A162/B201), he would seem to be clearly committed to at least the restricted version of premise (2) and thus the restricted version of Intellectualism.
However, there are problems with the restricted version of the argument. First, Kant’s argument in the Axioms of Intuition, from which the above quote is taken, applies only to determinate appearances, in contrast to his introduction of the notion of an appearance as the indeterminate object of an empirical intuition (A20/B34).39 This would suggest that we do indeed have intuition prior to any determination (where the latter presumably occurs via a synthesis by the understanding) and thus, contra (2´), that empirical intuition exists independently of any uniﬁcation offered by the understanding.
Sutherland (“Kant’s Axioms”) offers a similar diagnosis of the relation between the Aesthetic and the Principles. To the extent that we differ it is with respect to an understanding of this notion of determination. I discuss this issue of determination further with respect to Kant’s idea of “combination” (Verbindung) in §3 below.
94 journal of the history of philosophy 53:1 January 2015 Second, premise (2´) appears somewhat ad hoc. Kant presents the conditions for the unity of the forms of intuition in a way that requires continuity of structure between the forms of intuition and particular empirical intuitions. After all, Kant arrives at the pure forms of intuition by abstracting from all the empirical characteristics of a given empirical intuition (A20–21/B34–35). This implies that in any given empirical intuition, the purely formal characteristics of intuition are also present. So once the unity of the forms of intuition has been conceded, it is no longer clear why all intuitions do not possess a form of unity independent of the understanding. If Intellectualism is going to object that empirical intuition stands in special need of uniﬁcation in the relevant sense, then the burden is on Intellectualism to tell us why unity is required at the point of empirical intuition rather than pure intuition.
A third worry concerns the dialectical motivation for endorsing Intellectualism.
Part of the attraction of Intellectualism lies in its promise of making sense of the argument of the B-Deduction. According to Intellectualism, fulﬁlling this promise requires the demonstration of the dependence of the forms of intuition (i.e. space and time themselves) on synthesis. This is presumed to be the argument of §26.40 According to Intellectualism, if the representations of space and time themselves depend on the categories, then since anything that appears in space and time will be subject to the same fundamental conditions for representation as space and time themselves, such appearances will necessarily be subject to the categories.41 But once the claim that the unity of the pure forms of intuition is fundamentally discursive has been given up, the Intellectualist’s proposed interpretation of the role that the forms of space and time play in the Deduction no longer holds. For if I am right, then the original aesthetic unity of space and time cannot depend on synthesis, categorical or otherwise. So Intellectualism no longer has the resources to provide an interpretation of the argument of §26 that could account for the supposed necessity with which the categories apply to objects of experience, and thus lacks a crucial motivating element.42 We are thus left in a rather difﬁcult position. Intellectualism is by far one of the most widespread and well-developed interpretations of Kant’s critical philosophy generally, and of the argument Transcendental Deduction in particular. If the two steps of the argument I laid out above are correct, then either Kant incoherently endorses both Intellectualist and anti-Intellectualist claims in his critical philosophy, or Intellectualism is, despite its other attractions, a deeply problematic interpretive position, which may need to be abandoned.
Cf. Longuenesse, Capacity to Judge; Pereboom, “Kant’s Deductions.” For an especially clear discussion of this point, see Grifﬁth, “Perception and the Categories,” §10. Interpretations such as Longuenesse’s (which advocate a pre-conceptual synthesis that accounts for representation of space and time) will be more complicated in the story they tell concerning the necessity of the categories, but the general strategy and importance of §26 of the B-Deduction in that interpretive strategy is the same.
This is true even if we admit that Kant seems to require that the categories be in some way necessary for the representation of empirical objects. Certainly that seems to be the thrust of Kant’s argument in the B-Deduction (cf. B143, B161, B165). The point I make above against Intellectualism is simply that the most obvious strategy for achieving this conclusion—according to which the pure forms of intuition depend for their generation on the activity of the intellect—is vitiated once Intellectualism gives up premise (2).
two kinds of unity in the CRITIQUE OF PURE REASON To be sure, conviction concerning this point is hard to come by in the absence of an alternative positive account of the argument of the Deduction. Kant’s emphasis there on the unifying role of categorical synthesis gives the impression that the unity of intuition is ineluctably linked to the discursive unity provided by the categories (cf. B143–44, B152, B160–61). This issue demands further discussion.
But ﬁrst, in the next subsection, I discuss a series of objections that may arise in reference to Kant’s claim concerning our having intuitions of inﬁnite wholes. I then go on, in §3, to provide a sketch of how we might understand the relation between the aesthetic unity of intuition and a distinct notion of unity—namely, discursive unity—that I take to be the topic of the Deduction. Finally, I show that the controversial argument at B160–62 (discussed above), is compatible with this alternative reading.
2.3 An Objection: Totality and Unity One might object that my argument concerning the primacy of the intuitive and holistic representation of space commits Kant to the view that we have an actual intuition of the inﬁnite totality of space (and correspondingly time). This would be at least prima facie problematic for three reasons.43 First, it seems cognitively implausible, at least for ﬁnite minds. Second, it is incompatible with Kant’s view that we cannot represent an actual inﬁnity (B40, A432/B460). Third, it also seems incompatible with the receptive nature of sensibility, since receptivity requires affection of the senses, and it is implausible that an inﬁnite whole affects our senses. Let us take these in turn.
First, it certainly seems true that we do not perceptually intuit space as an inﬁnitely large object, since this is both phenomenologically and cognitively implausible.44 Indeed, it is difﬁcult even to conceive of the phenomenological character of being presented with an inﬁnitely large object. But it is not necessary (or proper) to tie pure intuition to perception, for the third and fourth arguments of the Metaphysical Exposition do not entail this absurd claim. Instead, all they phenomenologically require is that any perceptual experience of a space, or of something shaped and located in space, requires a representation of that space as merely a limitation of a larger whole. So for any experience of a bounded region, that experience is made possible in virtue of a representation of a larger region that encompasses the bounded one, and so on. Moreover, this way of conceiving of the perceptual presentation of particular spaces is sufﬁciently problematic for empiricist views, which must argue that space represented as inﬁnite in magnitude depends on a construction out of particular perceived spaces or spatial points rather than being given as a condition for their representation (e.g. Locke, Essay, II.xiii.4, 168). Hence interpreting Kant’s claim in the way I have suggested still leaves it with bite against his empiricist opponents.
This brings us to the second problem, that the holistic grasp of space as an inﬁnite given magnitude (B39–40) requires that we are able to cognize an actual inﬁnity, which is something Kant denies. The magnitude of a thing must be thought
Thanks to Thomas Land for discussion of these issues.
See Shabel, “Reﬂections on Space,” 52 for expression of a similar worry.
96 journal of the history of philosophy 53:1 January 2015 through a synthesis of its parts, and an inﬁnitely large thing requires an inﬁnite synthesis (A433/B461; cf. A162–63/B203). The representation of an inﬁnite magnitude would therefore require completion of an inﬁnite task—namely, the synthesis or combination of all the members of the inﬁnite totality—something that, according to Kant, is impossible.45 The important thing to note in reply to this objection is that, for Kant, there are two different ways in which a whole can be represented, and only one of those ways is problematic. There is a difference between the conceptual representation of space as an inﬁnite magnitude, which would require precisely the problematic form of part-whole discursive cognition that Kant highlights, and the intuitive representation of the whole, which does not require the problematic representation of it through a representation of its parts.46 Kant explicitly characterizes space as a totum, that is, as a whole that is prior (in the sense of grounding) to its parts.47 One should actually not call space a compositum but rather a totum, because the parts of space are only possible through the whole, rather than the whole being made possible through the parts. (A438/B466; cf. R 3789, 17:293; R 5299, 18:147) The status of space as a whole or totum in this sense means it does not fall prey to the kind of objection Kant raises against our cognition of the natural world as a whole in the mathematical Antinomies. There the issue concerns a “totum syntheticum,” or synthetic whole that is constructed from its parts. Henry Allison
puts the issue this way:
[T]he problem is that the rule or procedure for thinking a totum syntheticum [i.e. a whole which is not prior to its parts] clashes with the one for thinking an inﬁnite quantity. The former demands precisely what the latter precludes, namely, completability (at least in principle).48 So the problem Kant raises for the grasp of inﬁnite magnitudes such as the extent of the natural world is a problem only for a certain kind of whole—one that depends on its parts. Since space and time, as forms of intuition, do not have this feature, they are exempt from the worry. This is not to justify Kant’s claim that space and time are indeed such wholes, but it does show that his exposition of space and time as inﬁnite given magnitudes need not conﬂict with his discussion of problems related to the discursive cognition of inﬁnite wholes in the Antinomies.
Finally, there is the issue of receptivity and the forms of intuition. Kant is very clear that space and time, as forms of intuition, are ideal and so do not disclose See Bennett, Kant’s Dialectic, 120–21.
It is important that one keep in mind here the distinction between a conceptual representation whose extension is inﬁnite and a conceptual representation whose content or intension is inﬁnite.
Kant’s worry is speciﬁcally directed against conceptual representation that involves an inﬁnitely complex intension, or series of conceptual marks. Conceptual representation by ﬁnite minds is compatible with a concept’s applying to an inﬁnite number of things or being a mark in an inﬁnite number of subordinate concepts. See the sources cited in note 20 above.
For contemporary discussion of the notion of grounding in metaphysics, see Schaffer, “On What Grounds What” and “Monism.” For discussion of the relationship between metaphysical notions of grounding and Kant’s epistemology, see Smit, “Apriority.” Idealism, 370. Bell (“Propositional Unity”) also highlights the importance of distinguishing in Kant different notions of the part/whole relation.
two kinds of unity in the CRITIQUE OF PURE REASON the natures of any objects as they are themselves. But the claim that we have a holistic cognitive grasp of the basic characteristics of space and time suggests that something is indeed given to the subject. Such givenness implies a modiﬁcation of the subject’s receptivity, which seemingly contradicts the claim that the forms of intuition are ideal.
I suggest the following resolution. First, characterizing intuition as holistic and using this to explain our grasp of space and time is completely compatible with the idea that space and time are forms of intuition, for it helps explain how these forms are intelligible as ways of being receptive to objects. The ﬁrst (space) is the way of being receptive to objects distinct from the subject; the second (time) is the way of being receptive to the subject’s own modiﬁcations. Since space and time are ways in which a subject is receptive, they need not be considered as independent entities capable of affecting the subject herself. So they may be given as the subject matter of cognition without requiring that they be the result of a modiﬁcation of the subject.
Second, the claim that the intuition of space is of an inﬁnite whole is compatible with that given whole having a form of being that is entirely dependent on the subject.49 So there is nothing about the claim that space may be given to the subject that is incompatible with the ultimate subject-dependence of the forms of intuition.