| Non-reductive physicalism (NRP) is the theory that, although all properties supervene on physical properties, our theories about mental states do not reduce to theories in any base science. Though historically a popular thesis in the philosophy of mind, it has come under attack for its failure to account for mental causation. Specifically, authors such as Jaegwon Kim argue that non-reductive physicalism's ontological claims are inconsistent with its claims of realism, together with some plausible assumptions about causation. Here I provide a provisional defense of non-reductive physicalism.;One of the major problems with non-reductive accounts of mental causation is that they fail to provide any reason for counting mental kinds as real, projectible kinds---that is, there are arguments that statements about mental kinds are not supported by the positive instances, not do they allow inductive generalizations. I contend that this problem arises only if we assume that projectibility is a matter of a statement's central terms referring to kinds defined by instantiation of internal physical properties. I discuss reasons for rejecting this assumption and provide an alternative account whereby projectible statements can refer to "constrained kinds". Such kinds can feature in projectible statements even though members of the kind differ in their physical properties. Thus, there can be statements of psychology that do not reduce to statements in a base science (as the non-reductive physicalist claims) but that are still projectible (i.e., supported by their positive instances and allowing further inductive inferences).;The alternative account is then applied to a specific example from motor neuroscience: the representation of motor intentions in area M1 of the brain. I also provide some ideas as to how to apply the account to other examples in neuroscience, as well as other sciences (biology, economics, etc.). Even if non-reductive physicalism turns out to be implausible for other reasons, the theory of constrained kinds given here has potential applications to a number of philosophical problems and could be used to understand the ontology of a number of "special sciences". |
