THE NON-REDUCTIONIST'S TROUBLES WITH SUPERVENIENCE

Abstract

Most physicalists today believe that mental properties can be neurologically realized in many different ways. If some brand of functionalism is correct, then an internal event realizes a mental property in virtue of playing the right functional role, and not because of its intrinsic physical features. This allows that events that differ in neural type might realize the same mental property, the only constraint being that each of these event can play the functional role definitive of that mental property. The belief in multiple realizability is one of the main reasons for the popularity of non-reductive physicalism. If mental states are multiply realizable at the neural level, then they are also multiply realizable, and to a far greater degree, at more basic levels of physical structure ‐ e.g., the molecular, atomic, and subatomic levels. Of course, the properties of the physical sciences are also multiply realizable; the property of being a carbon atom, for instance, is multiply realizable with respect to subatomic features. However, the problem with mental properties is that they, unlike chemical properties, are thought to be multiply realizable with respect to the entire range of features mentioned by the physical sciences. That is, for any mental property M, and for any event of physical type P, it is possible that M is realized by an event of some physical type other than P. If so, and if this possibility is nomological as well as metaphysical, then mental properties are irreduciblein the sense that (NR1) there are no nomologically true biconditionals connecting mental properties with physical properties (i.e., properties of the physical sciences). Given that property-identity requires biconditionals that are (at least) nomologically true, it would also follow that