Abstract

This is a critical examination of Jerry Fodor's language of thought hypothesis . Fodor postulated a system of mental representations in which mental processes are supposed to be coded. The language of thought is alleged to be a discrete representational system which has a finite vocabulary, a compositional syntactic structure and a set of semantic rules that determine the meanings of the expressions. This mental language is supposed to be innate and instantiated in the brain, where neural objects or processes serve as the physical signs of the language. In the first chapter I trace the origins of the LOT hypothesis, showing the influence of generative grammar and Turing's algorithms on Fodor's work. I also show how functionalist philosophy of mind arose as an alternative to philosophical behaviorism and reductionistic materialism. In the second chapter I argue against Fodor's computational theory of propositional attitudes by questioning the very idea of physical symbols that have intrinsic semantic content and showing that his theory does not capture our normal notions of belief and desire. In chapter three I examine in some detail the hypothetical structure of LOT and I argue that there cannot be a language where the syntax perfectly mirrors the semantics, a language where there is no place for vagueness or ambiguity. I examine in chapter four the alleged innate character of LOT, and I argue that Chomsky's poverty of the stimulus argument does not support the thesis of innate concepts. Moreover, I argue that Fodor's innatism is inconsistent with his causal theory of meaning for LOT terms. In chapter five, I examine Fodor's arguments against connectionism, the main rival of the LOT hypothesis in contemporary cognitive science, and I argue that Fodor is wrong in believing that thought has the characteristics of constituency, productivity and systematicity in virtue of some natural law, but that it is rather a conceptual, or grammatical, truth that thought has those characteristics. It is not a factual issue whether thought has a symbolist or a connectionist structure