Abstract
"Difference-making" accounts of causation have recently been put to use in solving the "exclusion problem": the problem of explaining how mental properties can be causes if they supervene on but are not identical with their physical realisers. One question that received insufficient attention in the recent literature, however, is that of the identity criteria for mental properties. If mental properties are individuated according to their causal role, as traditional functionalism has it, they fail to be sufficiently distinct from their effects to render them apt candidates for causal relevance. But if we try to individuate them in some other way, we run the risk of being able to arbitrarily identify gerrymandered properties that will satisfy the difference-making criteria for causal relevance. Thus, I shall argue, the exclusion problem is still alive and kicking.
