Cognition Is (Mostly) Representational. So What?
Most cognitive phenomena are and should be explained with recourse to representational mechanisms. While this is contested, the Good Regulator Theorem makes it clear: adaptive control of behavior will require representations of some form or another. However, the debate over representation is misleading because the notion of cognitive representation does not go very far in helping cognitive science to attain deep theoretical understanding of cognition. My focus will be on how far you can go on to unify explanations of cognitive phenomena by relying on the idea that they rely on representations. I argue that one does not go very far, but that's at least some start. But with the mechanistic approach to representation, you get a bit more heuristics, and you can understand the link from representation to control to behavior.