One of Kant’s central contributions to the philosophy of mind consists in his articulation (if sometimes only implicitly) of a necessary condition for rationality that I shall call “cognitive control.”
By “rationality” I mean (at least) the ability to engage in the kinds of activities required for conceptual representation, judgment, inference, and the fixation of rational attitudes (e.g. belief and desires).
Cognitive control is a necessary (and perhaps sufficient) condition for a subject to count as rational. An agent has control over their various cognitive processes just in case the agent plays a particular kind of decisive efficient causal role in the outcomes of cognitive processes, whether they be, e.g., new beliefs that are the result of inference, or bodily actions that are the result of some choice. I propose that this decisive causal role is manifested most clearly in Kant’s conception of the free attentive control an agent has over her cognitive resources, in which the agent’s determination of how to distribute her attention settles which sorts of cognitive processes she engages in and how she does so. I argue that this view complements, rather than conflicts with, contemporary scientific understanding of humans and other animals. In particular, it articulates a richer conception of the difference between rational and non-rational forms of abnormality than the somewhat deflationary tendencies of cognitive psychology might otherwise allow, without either over-intellectualizing rationality, or positing an unbridgeable gulf between the rational and non-rational. The Kantian view allows that that the capacity for rationality may extend beyond the human species, but it also helps clarify what is so significant in rational human life.
There is a philosophical tradition that regards rationality as essentially self-conscious. What this means is open to interpretation. But a variety of versions of this thesis construe rationality as dependent on possession of a set of meta-representational (or sometimes “meta-cognitive” abilities).1 It is important to distinguish Kant’s thesis that rationality requires cognitive control, and thus a kind of metacognition, from the thesis that rationality requires meta/representation/.
“metacognition” can designate eitehr a cognitive process that monitors or controls other cognitive processes or a cognitive process that is or involves self-directed metarepresentational states. I take the philosophical tradition I am targeting to be largely concerned with the latter. My notion of “cognitive control” is compatible with the former conception of metacognition. For discussion see [@carruthers2014a]. ↩︎