Abstract

In an attempt to show how rational explanation of human and animal behaviour has a place in the scientific explanation of our physical world, Fodor advances the language of thought hypothesis. The purpose of this dissertation is to argue that, contrary to the language of thought hypothesis, we need not possess a linguistic internal representational system distinct from any natural language to serve as the medium of thinking. I accept that we have an internal representational system, but by analyzing Fodor's theory of content, I show Fodor's argument that the internal system must be as expressive as any natural language, which he uses in arguing that the internal system is the linguistic medium of thought, is unsound. Distinguishing an informational theory of content from a causal theory of content, which Fodor conflates, I argue that internal representations, whose content is determined by information they carry, cannot be related in a way that corresponds to semantic associations between terms in natural languages, reflecting actual associations of items in the world. Furthermore, provided certain animal cognition, which is homogeneous with human cognition, can be explained without requiring that the internal system possess anything corresponding to the logical connectives, the internal system need not possess anything corresponding to the logical connectives. I give such an explanation of animal cognition by developing an approach to content in the Rylean/Dennettian tradition, based on the notion of embodied cognition, in which animals embody the hypotheses they entertain in virtue of their total dispositional state, rather than explicitly representing them. It follows that there are two features of natural languages, semantic associations of terms and possessing logical connectives, that the internal system need not have. Hence a rational interpretation of linguistic behaviour need not be derived from an intentional interpretation of the transformations on internal representations, from which it follows that the internal system need not be the medium of thought, which can be a natural language. Also, if the internal system need not possess anything corresponding to the logical connectives, it need not be as natural languages, and hence it need not be linguistic.