A reader asked why I didn't mention supervenience in my recent posts on property dualism. He opines that "the notion was invented to make sense of the position you are arguing against." Let's see.
My Problem With Property Dualism Roughly Stated
I take a property dualist to be one who maintains all of the following propositions:
1. There are irreducibly mental properties.
2. There are irreducibly physical properties.
3. All particulars are physical particulars.
4. Some but not all particulars instantiate both irreducibly mental and irreducibly physical properties.
My problem, roughly, is that I don't understand how a physical particular (a brain, a region of a brain, a brain event, or state, or process) can instantiate one or more irreducibly mental properties. Why should there be a problem? Well, if a physical particular is exhaustively understandable in terms of physics (and the sciences based on it) then there is just nothing irreducibly mental about it, in which case it cannot instantiate an irreducibly mental property. I am making the following assumption:
A. If a nonrelational predicate P is true of a particular x, then there must be something in or about x that grounds P's applicability to x.
So if 'feels pain' is true of a physical particular, and 'feels pain' picks out an irreducibly mental property, then there must be something irreducibly mental about that physical particular. Otherwise there would be nothing in or about the particular that could render the predicate true of the particular. But if there is something irreducibly mental about a physical particular, then that particular is not physical in the sense of being exhaustively understandable in terms of physics.
I find (A) to be self-evident. For suppose you were to deny it. Then you would be countenancing the following: there is some particular x that instantiates a property P-ness even though the nature of x excludes P-ness. You would be countenancing, for example, an electron (which is course a negatively charged particle) which yet instantiates the property of being positively charged. If a particular has a an intrinsic (non-relational) property, then that property expresses what the particular is, its nature (in a broad sense of this term).
Now we have to see whether the notion of supervenience can help me with my problem.
The problem for the nonreductive physicalist is that he must avoid both eliminativism and reductionism but without falling into epiphenomenalism, emergentism, or (of course) substance dualism. Epiphenomenalism cannot accommodate the fact that mental phenomena sometimes enter into the etiology of physical events, while emergentism and substance dualism leave physicalism behind. The problem is to somehow secure the reality, the causal efficacy, and the irreducibility of the mental while maintaining the dependence of the mental on the physical. Nice work if you can get it! What the physicalist needs, it seems, is a dualism of properties together with the idea that the mental properties somehow nonreductively depend on the physical ones. But how articulate this dependency relation?
Enter supervenience. The basic idea is that mental properties are not identical with, but merely supervene upon, physical properties in the way in which ethical properties have been thought (by G. E. Moore, R. M. Hare, and others) to supervene upon natural properties. Suppose A and B are both ethically good. It does not follow that there is any one natural, non-disjunctive, property with which goodness can be identified. Perhaps A is good in virtue of being brave and trustworthy, whereas B is good in virtue of being temperate and just. Goodness is in this sense "multiply realizable." A and B are both good despite the fact that their goodness is realized by different natural properties.
Nevertheless, (i) a person cannot be good unless there is some natural property in virtue of whose possession he is good, and (ii) if a person is good in virtue of possessing certain natural properties, then anyone possessing the same natural properties must also be good. Given that A-properties supervene upon B-properties, the "supervenience T-shirt" might read: "No A-property without a B-property" on the front; "same B-properties, same A-properties" on the back. As Jaegwon Kim puts it, "The core idea of supervenience as a relation between two families of properties is that the supervenient properties are in some sense determined by, or dependent upon, the properties on which they supervene." (Jaegwon Kim, "Epiphenomenal and Supervenient Causation," in Supervenience and Mind: Selected Philosophical Essays, Cambridge University Press, 1995, p. 98.)
Kim's preferred way of cashing this out is in terms of strong supervenience. Let A and B be families of properties closed under such Boolean operations as complementation, conjunction and disjunction. A strongly supervenes on B just in case:
(SS) Necessarily, for any property F in A, if any object x has F, then there exists a property G in B such that x has G, and necessarily anything having G has F.
Applying (SS) to physicalism, we may define the determination thesis of strong supervenience physicalism as the view that, necessarily, (i) for any mental property M, if x has M, then there is physical property P such that x has P, and (ii) necessarily, anything having P has M.
But how does this help me with my problem? If x has M and M is an irreducibly mental property, then, by assumption (A) above, x is at least in part mental, and not wholly physical where 'wholly physical' means 'exhaustively understandable in terms of physics and the sciences based on it.' This problem is not solved by telling me that x cannot have a mental property without having a physical property, and that anything having that physical property must have the mental property. For my problem is precisely how x, which is wholly physical, can have an irreducibly mental property in the first place.
One might respond along the following lines. "Look, the whole idea here is that mental properties are functional properties. So when we say that x, a brain event say, has a mental property, all we mean is that it stands in certain causal relations to sensory inputs, behavioral outputs, and intervening brain events. So what makes the brain event mental is simply the relations in which it stands to inputs, outputs and other brain events. Once you grasp this, then you grasp that the brain event can be wholly physical in nature despite its having a mental property. Mental properties are not intrinsic but relational."
Unfortunately, this won't do. Felt sadness has an intrinsically mental nature that cannot be functionally characterized. A subsequent post will spell this out in detail.
Besides, if the property dualist holds that mental properties are really relational, then, strictly speaking, he is not a property dualist. He is not maintaining that there are two sorts of properties, but that mentality consists of relations and that there are no monadic mental properties. Furthermore, his talk of irreducibility must mean only that that type-type identities fail, that for every mental property there is not one unique physical property with which it is identical. Irreducibility boils down to multiple realizability. Mental 'properties' are irreducible in that they are multiply realizable.