What Incongruent Counterparts Show

In a recent paper, Robert Hanna argues that Kant’s incongruent counterparts example can be mobilized to show that some mental representations, which represent complex states of affairs as complex, do so entirely non-conceptually. I will argue that Hanna is right to see that Kant uses incongruent counterparts to show that there must be a non-conceptual component to cognition, but goes too far in concluding that there must be entirely nonconceptual representations that represent objects as existing in space and time. Kant is deeply committed to the thesis that no representation of a complex state of affairs as complex can be entirely non-conceptual. For Kant, all representations of complex states of affairs as complex (including those of incongruent counterparts) are conceptually structured. I present an interpretation of the Transcendental Aesthetic according to which Kant not only aims at Leibnizian and Newtonian accounts of space and time, but also Hume’s. Hume’s account fails to make representations of complex states of affairs sufficiently determinate. Kant offers an account later in the Critique that is meant to correct this failing by requiring that all representations of complex states of affairs as complex be conceptually (inferentially) structured. In his recent paper, ‘Kantian Non-Conceptualism’, Robert Hanna argues that Kant’s incongruent counterparts example can be mobilized in an argument for the conclusion that ‘essentially non-conceptual content exists’. That is, he argues that this example proves that some mental representations, which represent complex states of affairs as complex, do so entirely non-conceptually. His argument is (roughly) as follows. Incongruent counterparts are qualitatively identical. Therefore, any concept that applies to one of a pair of incongruent counterparts applies to the other. Therefore, a representation of a pair of incongruent counterparts as distinct cannot be an entirely conceptual representation. Therefore, this representation must be entirely non-conceptual. Clearly, the move from the penultimate step of this argument to its conclusion is invalid, but I will argue that it is exactly the move that Hanna makes. So, Hanna is right to see that Kant uses incongruent counterparts to show that there must be a non-conceptual component to cognition, but moves beyond what his premises can support in concluding that there must be entirely non-conceptual representations of items as existing in space and time. Instead, I will argue, Kant is deeply committed to the thesis that no representation of a complex state of European Journal of Philosophy ]]]:]] ISSN 0966-8373 pp. 1–18 r 2011 Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. affairs as complex (including any spatio-temporal state of affairs) can be entirely non-conceptual. That is, for Kant, all representations of complex states of affairs as complex are conceptually structured. Therefore, any representation of incongruent counterparts as (spatially or temporally) distinct from one another must be conceptually structured. This thesis is at odds with a certain reading of the Transcendental Aesthetic according to which Kant argues that our intuitions of space and time are entirely non-conceptual. I will thus offer an interpretation according to which, even in the Aesthetic, representations of space and time must be conceptually structured. The case for this interpretation will center on the notion that while Kant scholars have long focused on Kant’s attempts to refute Leibnizian and Newtonian accounts of space and time in the Aesthetic, they have largely overlooked the fact that Kant also there seeks to refute Hume’s account. I will argue that by examining the reasons that Kant gives in the Aesthetic for rejecting Hume’s account of how we represent space and time, we can see that he must also be committed to the thesis that the representations of space and time are all conceptually structured. The argument that I will trace through the Aesthetic goes (roughly) like this. Hume thought that we represent items as being spatially and/or temporally related to one another by placing representations of these items into spatial and/ or temporal relationships with one another. This account fails to make representations of complex states of affairs sufficiently determinate (cf. Landy 2009a). The account offered by Kant later in the Critique, in the Metaphysical Deduction, is broadly speaking a kind of inferentialism and is meant exactly to correct this failing of Hume’s (cf. Landy 2009b). Kant’s inferentialism requires that all representations of complex states of affairs as complex be inferentially (conceptually) structured. Thus, we must understand Kant’s project in the Aesthetic not as positing non-conceptual representations, but rather as bracketing the conceptual aspects of representation to emphasize and scrutinize their nonconceptual aspects. My procedure here will be as follows. First I will present and critique Hanna’s argument, in the process suggesting an alternative to the conclusion he draws from Kant’s premises. Next I will present evidence that this alternative is, in fact, Kant’s position by showing that his position in the Transcendental Aesthetic can be read as a direct response to Hume’s account of how we represent complex states of affairs as such. Finally, I will sketch a broad-strokes picture of how the rest of the Critique fills in the details of Kant’s response to Hume by constructing a theory of representation according to which non-conceptual representations (which do not represent anything as complex) are combined to form conceptually-structured representations (which do). To begin, then, I will turn to Hanna.