No mental property is identical with a physical property.
All properties strongly supervene2 on physical properties.3 In other words: necessarily, for all properties P, all objects x, and all times t: if x has P at t, then, for some physical property P, x has P at t and, necessarily, for all y and all times t , if y has P at t, then y has M at t.
These assumptions, and Kim’s arguments, concern two different kinds of entities: properties and events. Like Kim, we assume that events are concrete instances5 of properties—not “instances” in the sense of objects that have the properties, but instances in something like the sense of havings of properties by particular objects at particular times.6 We take the following schema (at least when suitably restricted) to be a platitude: the event that is the having of property P by object xat time t exists if and only if x has P at t.7
According to Kim, Irreducibility, Supervenience, and Closure are shared commitments of all nonreductive physicalists. Thus Kim is posing the dilemma: either reject nonreductive physicalism or reject Causal Efficacy, viz., accept epiphenomenalism.
But things aren’t quite so simple, as there is a further premise in the argument which is not a characteristically nonreductive physicalist assumption:
No single event can have more than one sufficient cause occurring at any given time, unless it is a genuine case of overdetermination.8
Kim says this is a “general metaphysical constraint” (p. 22), so presumably he thinks we should accept it whether we are nonreductive physicalists or not. Whether Exclusion is true or not is not perfectly obvious, and we’ll return to this matter in section 7. For now let us simply note that the appeal of the principle is perhaps due to the fact that it sounds tautologous: it sounds a lot like the claim that every event has at most one sufficient cause occurring at a given time unless it has more than one sufficient cause occurring at that time.
3. Other general metaphysical constraints
Though the assumptions Kim names and calls his “substantive premises” (p. 41) end here, his argument requires several other assumptions that, like Exclusion, are not clearly nonreductive physicalist commitments. We assume that he regards them, too, as “general metaphysical constraints”. One such assumption is what we will call: