Wednesday, June 29, 2005

A comment from an anonymous commentator

On your last post about "the argument" from intentionality.Any critique of naturalistic accounts of reference (intentionality, content) needs to grapple with Fred Dretske's 'Knowledge and the Flow of Information' (KFI). He addresses all the concerns in the post about how a physical system could have intentionality: in his words, how to 'bake a mental cake using only physical yeast and flour'.As background, in neuroscience we use the idiom of 'representation' all the time, to refer to 'internal maps by means of which we steer' (Dretske's line). More dryly (and still leaving out some details) in neuroscience, the following are in practice taken to be sufficient criteria for Y to be an internal representations of X (where X is some feature of the world): Y must carry information about X (roughly, X should be predictable from Y), and Y is used by the rest of the system to guide behavior with respect to X. That is, Y is a map of X, and the organism uses Y to steer behavior wrt X. At any rate (and despite some of Vallicella's confident yet unfounded claims to the contrary), we indeed have discovered representational systems in brains that meet the above criteria. This notion of representation is Dretske's starting point, which he spells out in the first two parts of KFI. Then, in the remaining third of his book, Dretske adds to this initial proto-representational core to get things like concepts and full-fledged intensional content.Given that Dretske's story is generally recognized by naturalists (myself included: his books are in the very short list of philosophy books I have read that have actually helped me clarify my ideas as a neuroscientist studying sensory systems) as one of the best stories around about content, I was wondering if in your work on intentionality you have looked at or responded to Dretske's work, in particular KFI? When I read your most recent post, in the back of my mind was Dretske's theories providing obvious responses that make your arguments seem sort of out of touch with the positive naturalist accounts that are out there.As an aside, Vallicella mentions Dretske on his site, here: he does not understand Dretske and ends up attacking a straw man. On that site, he attributes to Dretske the view that Y carrying information about X is sufficient for Y to have X as intentional content, so (e.g.) a thermometer could be said to represent temperature. Dretske takes great pains to point out that this 'thermometer' view of content is false, but doesn't go on (like Vallicella did) to throw out the baby with the bathwater. Rather, Dretske points out that just because natural meaning (basically, information) is not sufficient for intentionality, that doesn't mean that a naturalistic account of intentionality needn't draw at *all* on the resources of information theory. Indeed, Dretske spends the last third of his book discussing how natural meaning can be incorporated into an account of full-fledged intentional content (with all its nice properties like absent referents).