The Explanatory Role of Intentional Content in Cognitive Science

Title of Dissertation: THE EXPLANATORY ROLE OF INTENTIONAL CONTENT IN COGNITIVE SCIENCE Andrew Charles Knoll, Doctor of Philosophy, 2015 Dissertation directed by: Professor Georges Rey, Department of Philosophy This work argues that intentional content plays at least two explanatory roles in cognitive science. First, it allows cognitive states to be type-individuated independently of their relations to one another and to mind external phenomena. Secondly, it allows for counterfactual preserving generalizations over states so individuated. Thus, intentional content does not play this explanatory role in highly encapsulated cognitive processes. By contrast, it is necessary to type individuate states that partake in isotropic cognitive processes. This work thus cuts a middle path between those who would eliminate intentional content from cognition altogether, and those who take it to be the ‘mark of the mental.’ Chapter 1 argues that there is no good reason to eliminate intentional content from cognitive science. But, it also argues that there is a coherent notion of computation without representation on offer as well. So, many cognitive processes could be explained as computations over states without intentional content. Chapter 2 argues that many extant accounts of the explanatory role of intentional content end up being otiose. Too often, such accounts are concerned with capturing our intuitions about the proper way to talk about cognitive processes. But, in many cases, this talk can be eliminated from our explanations without loss of explanatory power. Chapter 3 lays out the main argument. Many encapsulated cognitive processes—including early perceptual processes—can be explained in terms of computation without intentional content. In contrast, processes that are open to isotropic revision require their states to be individuated in terms of intentional content. Chapter 4 surveys some objections to this view. One worry is that if cognition is massively modular, then all cognition must be non-intentional. On the contrary, modular processes can also be open to isotropic revision, and thus be amenable to intentional explanation. Chapter 5 provides an example of such a modular process: the phonological system. It argues that states of the phonological system must be individuated in terms of intentional content. Phonological processing thus provides a case study for intentional explanation more generally. THE EXPLANATORY ROLE OF INTENTIONAL CONTENT IN COGNITIVE SCIENCE