Representational content in cognitive psychology

Against Stich\u27s recommendation that we purge cognitive psychology of content I argue that ascriptions of representational content are both scientifically legitimate and essential to the continuing success of the cognitive sciences. Yet it is not the ordinary folk notion of content that informs many of these sciences, e.g. experimental cognitive psychology, cognitive ethology, and theory of perception. I develop an approach to representation that builds upon a Dretske-style analysis of representation. However, I reject Dretske\u27s requirement that representational states must covary perfectly with those conditions in the environment they represent. Instead, along with Hatfield and Matthen, I propose that some states represent because, first, they are states of a system the biological function of which is to represent, and, second, in the context of this system their function is to stand in a regular (though not necessarily perfect) correlation with some condition in the environment