Susan Schneider

The dissertation advances a theory of psychological explanation in Cognitive Science that is based on the conjunction of three leading, but controversial, views in philosophy of mind and language: a semantics of direct reference, the classical computational theory of the mind, and a theory of mental content that is "broad", . I argue that some psychological kinds are intentional, while others are computational, being expressions in the language of thought.In Part One, I defend the view that intentional kinds are sensitive to broadly referential properties of mental states, being indifferent to the mode of presentation of the state. I offer three arguments for the view that at least some psychological laws should be broad, subsuming states by their referential properties only. In addition, I respond to the leading objection to a referentialist picture of intentional explanation: the problem of Frege cases. In Part Two, I employ expressions in the language of thought to play the theoretical role of guises or modes of presentation , the ways that individuals conceive of referents. I provide a positive account of guises or MOPs, advancing an individuation condition for symbols in the language of thought that is based on the total computational role that the symbol has. Noting that such symbols are not shared, I illustrate how psychological explanation can nonetheless be "public"; that is, I illustrate two ways in which psychological laws can cover different systems. Advantages of this picture of psychological include: a manner of accommodating both of the main intuitions about the Twin Earth thought experiment. Twins have states that play the same roles in their internal cognitive economies, while nonetheless, the states differ in their semantic contents. The basis of a theory of concepts that individuates concepts by both broad content and functional, and more specifically, computational role. Concepts thus have two-dimensions: the role that the concept plays in one's cognitive economy, and a semantic dimension. Openness to various theories of the nature of broad content and to both the Russellian and the hidden indexical theories of attitude ascription

This chapter focuses on one historically important approach to computationalism about thought. According to "the classical computational theory of mind" (CTM), thinking involves the algorithmic manipulation of mental symbols. The chapter reviews CTM and the related language of thought (LOT) position, urging that the orthodox position, associated with the groundbreaking work of Jerry Fodor, has failed to specify a key component: the notion of a mental symbol. It clarifies the notion of a LOT symbol and explores an approach different from the orthodox, Fodorian LOT/CTM. The LOT position is a leading approach to the computational nature of thought. A key feature of LOT is that its mental representations have a combinatorial syntax. A representational system has a combinatorial syntax when it employs a finite store of atomic representations, which can be combined to form compound representations, which can then be combined to form further compound representations.