Jerry Fodor's ambition in Variations is polemical rather than scholarly.

Hume Variations
Jerry A Fodor
Clarendon Press

He thinks that sorting out what's worth saving in Hume's 'theory of ideas' is a useful way of crystallising some of the central issues in contemporary philosophy of mind. Hume's Treatise of Human Nature, it emerges, is the 'foundational document of cognitive science,' and Hume an early proponent of the 'Representational Theory of Mind', which Fodor happens to endorse.

Consider this joke, due to Groucho Marx: "I shot an elephant in my pyjamas." Pause. "How an elephant got in my pyjamas I just can't imagine." Laughter. According to Fodor, there's a convincing and commonsensical story to be told about the mechanics of Groucho's witticism. We can parse the first sentence in two ways, but we naturally assume that Groucho meant to say that (he in his pyjamas) (shot (an elephant)). The punchline works, of course, by showing us that it wasn't what he meant. Fodor points out that this explanation depends on our seeing that what Groucho had in mind was (shot (an elephant in my pyjamas)). And it can't be the case that what Groucho thought is ambiguous in the way that what he said was, for that would put Groucho and the listener in the same boat. It is thoughts, in other words, that "disambiguate statements". The philosophical moral of the story (and the joke) is that equivocation is not intrinsic to thought or "mental representation", but afflicts instead the inference from what someone says to what he has in mind, which depends on the context of the utterance.

Now, an important feature of Fodor's conception of the mind and mental phenomena is his view that our ordinary beliefs about such things, which are sometimes dismissed by philosophers as so much 'folk psychology', are actually in pretty good shape, and that the job of philosophy is not to correct them but rather to show how they hang together. Take the story about Groucho. It is intuitively plausible, yet many philosophers, especially 'pragmatists' and 'Wittgensteinians', reject it. Fodor maintains that their reasons for doing so rest on a fundamental confusion; and worse still, that the explanations they offer run counter to our intuitions about the relationship between what we say and what we think.

On Fodor's account then, pragmatists and Wittgensteinians take the content of mental representations to be context–dependent. (Relatedly, they hold that to possess a concept is to know how to use it — that's why they're called 'pragmatists'; whereas for representationalists like Hume and Fodor, to have the concept DOG just is to be able to think about, to represent in one's mind, dogs.) But getting Groucho's joke seemed to depend on it being the case that what he meant was independent of any contextual fact. Why, then, are pragmatists prepared to deny what strikes us as obvious? Namely, that statements, which can be equivocal, get their content from thoughts, which can't. Fodor suggests two reasons. First, pragmatists assume that we're faced with a stark choice here: either the content of thought is intrinsic or it is "relativised to a context of interpretation". Fodor argues that this doesn't exhaust the options available, and that to say, as some representationalists do, that thoughts depend on something or other for their content is not at all the same as saying that the content of a thought is what an interpreter–in–context takes it to be.

The second reason why pragmatists have dismissed representationalism has to do with concerns in the theory of knowledge. Early representationalists combined a psychological claim about the mental particulars that mediate our perceptions with an epistemological claim to the effect that only mental things can be perceived. A claim from which, as Hume himself showed, scepticism about our ability to perceive non–mental things surely follows. Fodor doesn't, as far as one can tell, doubt the cogency of these concerns, just their relevance to the science of the mind. "Epistemology," he says, "hasn't really got much to do with psychology." Today representational accounts of the mind run unencumbered by the need also to do duty as theories of knowledge. Pragmatist critics of such accounts habitually confuse epistemological issues with psychological ones, and it is not the least of Fodor's achievements to have made this clear to the general reader in a book whose brevity does not come at the expense of rigour, clarity or ambition.