Abstract

There is a question how a causal approach to explaining the representational powers of belief can work together with a view according to which beliefs have truth conditions determined by their logical form. In Chapter I, I point out that truth makers of representations should not be language dependent entities if the aim is to show how representation is possible. I introduce a causal account of belief as being the kind of account which can comply with this. ;I examine features of a view according to which language is essential to belief. An argument for the view that beliefs must have a language like structure is the argument from the possibility of rationality. I argue that on one interpretation of the view in question, this is not the kind of thing that avoids Humean associationism and puts reason into the thought. ;I find that a causal account of the referents of terms in a mental sentence is not a workable causal approach to belief. I show how causal atomism contrasts with other theories and what other kind of causal account of belief you could have. ;In Chapter V I examine a particular causal account of representation, that of Stampe's, treating it as an example of a nonatomistic account of belief. I show how such a view responds to the problem of false belief. I argue that the general form of answer the causal theory gives is the same that certain noncausal theories give. ;I turn to the question what it would be for a belief to have semantic complexity on a causal account. On one way in which a belief could have semantic complexity--by representing more than one object--there is an apparent problem. Assumptions about the epistemology of representation and assumptions about the relation the form a belief takes to what it represents appear to lead to the conclusion that only true beliefs can have semantic complexity. I suggest one possible way out of this problem.