The central claim of my in-progress dissertation is that considerations in the contemporary philosophy of mind and epistemology give us good reason to posit the existence of (at least) two different faculties of the mind, distinct in structure and function: perception, on the one hand, and cognition on the other. Put another way: perception and cognition are natural psychological kinds. The argument hinges on the idea that perception, unlike cognition, is belief-independent, and that its being belief-independent is crucial to its functioning.

To that end, the dissertation is largely concerned with drawing out the theoretical implications of belief-independence for perception. Put roughly, a psychological process is belief-independent when beliefs and desires are not indispensable to the representational explanation of that process. A belief or desire is indispensable to the representational explanation of a psychological process when one cannot explain the content of the process’ output without appealing to the content of the belief or desire.

It turns out that belief-independence, as I have defined it, is secured if a psychological system is informationally encapsulated with respect to cognition. Skepticism regarding such a thesis has recently become popular, with various claims to the effect that perception is cognitively penetrated and therefore unencapsulated. As such, much of the dissertation will be concerned with rejecting the idea that there is enough cognitive penetration of perception—that there are what I and others call semantically coherent relations between the content of cognitive states and the content of perceptual outputs—such that would render belief-independence is false. In making this case, I also make a case for the relevance of informational encapsulation to debates over cognitive penetration, which has also recently been questioned in the literature.

From here, I then outline two constitutive functional roles perception is meant to play, roles that, I claim, require the truth of belief-independence in order to for perception to fill them.

The first role is perception’s fundamental functional role of recovering a unique distal scene from underdetermined inputs. It is well-known that perception is subject to a variety of underdetermination problems. Many of these underdetermination problems focus on the computational intractability of recovering a unique distal scene from the impoverished inputs. Because we obviously do recover a unique distal scene from impoverished inputs in a short amount of time, we need an explanation of how this happens. In short, we need an explanation of perception’s tractability, given the fact that perception is a computational process. What I argue is that belief-independence is necessary for the computational tractability of a perceptual process.

The second role is perception’s anchoring character. Perception is anchoring insofar as it makes certain cognitive capacities—the capacity for a posteriori knowledge, the capacity for conceptual singular thought, and the capacity for abstract thought—possible. Perception therefore plays both a semantic and epistemic anchoring role vis a vis cognition. I claim that in order to play these semantic and epistemic roles, perception must be belief-independent.

After arguing for the perception’s anchoring character, I bring everything together to advance an architectural account of the perception-cognition border. Fundamentally, the border between perception and cognition is constituted by restricted patterns of information flow and access. Specifically, the encapsulation of perceptual processes from cognition precludes the content of cognitive states from standing in semantically coherent relations with the contents of perceptual outputs. In virtue of this, it turns out that beliefs and desires—which are, of course, cognitive states—are dispensable to the representational explanation of perceptual processing. As such, perception is belief-independent in virtue of its architectural relation to cognition, that is, in virtue of its being informationally encapsulated from cognition.

The dissertation closes by comparing the account to other extant accounts of the perception-cognition divide and closes by exploring, in miniature, various ways in which my account may helpfully inform various debates in the contemporary philosophy of mind and epistemology, such as: the theory-neutrality of observation, the status of perceptual states as unjustified justifiers, the contents of perception, etc. Downstream, I also hope that my work on perception helps to indirectly characterize some unique and important features of cognition as well, though expounding upon those is another project, one that is beyond the ken of the dissertation.

Outside of my dissertation, I am also deeply interested in thinking about: self-knowledge, intentional action, animal cognition, the nature of representation, computation and artificial intelligence, induction and inference to the best explanation, synchronic and diachronic norms of rational coherence, and the metaphysics of science.