Abstract

We address the issue of the normativity of representation and how Grush might address it for emulations as constituting representations. We then proceed to several more detailed issues concerning the learning of emulations, a possible empirical counterexample to Grush's model, and the choice of Kalman filters as the form of model-based control.