Notions of ‘representation’ and the Diverging Interests of Philosophy and Empirical Research

Many of you may have had a neighbour or a grandfather who was a fanatic about tools. 1 Grandpa owned everything Sears ever made, and took it as a sign of moral turpitude if you used, say, a flathead screwdriver to take out a Phillips-head screw. Now I confess that I still do not own a Phillips screwdriver, and also use my flathead screwdriver to open paint cans. But Grandpa did rub off on me a little bit: I eventually became a philosopher, and take it that an important part of what philosophers do consists in the examination of conceptual tools. What I am going to do here is look at one of the main conceptual tools used in cognitive science: namely, the notion of representation. Due to limitations of time, the discussion will be be very programmatic in nature, and much of the detail that might be supplied in a longer discussion is left out. I hope to persuade you of three main points: First, there is real unclarity and even ambiguity in the usage of the word 'representation'. Second, different notions of representation are suited to different tasks—in particular, that some philosophical projects require a stronger sense of the word than do most empirical theories. Third, it is possible to adopt a weak construal of the word 'representation' that lets us have a good scientific psychology and lets us have intentional realism too, but without providing a representational explanation of intentionality. Now my interest in clarifying the very notion of representation may come as a bit of a surprise. There seems to be a widespread assumption in cognitive science circles that we know perfectly well what we mean when we say something " is a representation, " and hence the only interesting questions are about what sorts of representations are used, how they are realized in the brain, and how they get this peculiar property of representing something. And the assumption that we know what it is to be a representation is quite understandable, since the notion of representation looks like an old friend. After all, we have known about representation as long as we have known about symbols and pictures and maps and leitmotifs. Anybody who knows English knows what we mean when we say that a picture or a symbol is a representation. While this attitude is perfectly natural, it also involves some implicit (and arguably …