Abstract

Farah argues that nonlocal models explain clinical data better. However, the locality assumption does not seem so implausible if different sorts of data are taken into account. In particular, priming experiments using evoked response potentials support modularity. I describe some ERP studies relevant to this issue