ABSTRACT
The main focus of my comments are processes that appear to be contradictory to Brewer's model, but are actually compatible with it. In that way, I hope they suggest expansions of the model, to account for apparent exceptions. The exceptions that I am thinking of are cases in which we adopt or switch to the person-based mode of processing when our goals are impersonal, and others in which we adopt a category-based mode despite self-involving goals. I think such mode switching is affect-driven: Negative affect tends to drive us into category-based processing, and positive affect tends to drive us into person-based processing. I propose that, because the source of the affect that motivates the mode switching is the other's behavior, a context-sensitive and interactional component be added to the model. It can easily accommodate such an addition.
