Proponents of the Standard View who countenance beliefs – non-eliminativists – have the task of showing how beliefs may be scientifically respectable internal states suitable for causal explanation. Since beliefs are identified by content, the task is to show how content may be assigned to internal physical states in such a way that beliefs can be causally explanatory. Although assignment of content to brain states is a purely technical problem – and as a result, this chapter is a fairly technical discussion – it is a problem whose solution is required if there are beliefs as construed by the Standard View. Eliminative materialism, discussed in Chapter 3, is not faced with these problems since it does not recognize beliefs anyway; but the problem of content and causation is an urgent one for noneliminative proponents of the Standard View.

Without trying to survey all the recent work on content, I consider three different approaches to the problem of assigning content to internal states. Two appeal to a language-of-thought hypothesis. The first, proposed by William G. Lycan, tries to show that brain states are syntactically structured entities; the second, proposed by Jerry A. Fodor, looks to a new kind of semantic property – narrow content – to be causally explanatory. The third approach, Fred Dretske's, offers an account of belief as indication, without appeal to a language of thought. I argue that none of these theories is satisfactory: They all have technical (but interesting) difficulties that seem insoluble.