Intentionality and supervenience

III. Intentionality and the Supervenience Strategy

Another very popular view, which has even been accepted by some Christians, is a nonreductive materialist position. On this view, intentional states are not eliminated, they are not reducible to physical states, they are, however, supervenient upon physical states. Mental states are not identical to physical states, but given the state of the physical, there is only one way the mental can be.

Of course, earlier I indicated that supervenience of all non-physical states on physical states is part of what it takes for a world-view to be naturalistic. However, if mental states can be reductively analyzed in terms of physical states, then the supervenience is simply obvious. A difference in B requires a difference in A because, in the final analysis, Bs just are As. Again, if the B-states are eliminated from the ontology, then we don’t have to worry about a difference in B that is not guaranteed by a difference in A. However, for many, perhaps most philosophers who believe in a broadly materialist world-view, the reductionist and eliminativist positions are both implausible. For these philosophers, the supervenience relation has a job to do, it explains how it is possible for everything to be in the final analysis physical while at the same time maintaining the irreducibility and the autonomy of the mental realm.

Philosophers often distinguish between weak supervenience and strong supervenience. According to weak supervenience, B-properties weakly superven on A-properties if and only if things that are alike in their A-properties are always alike in their B-properties. What this establishes is a constant conjunction between A-properties and B-properties. It does not really show that there is anything about the A-properties that guarantees that the B-properties will always be the same. Nevertheless, we must remember what caused problems for reductionist accounts of mental states. The physical, I maintained, is incurably indeterminate with respect to propositonal states. Whatever story we tell at the physical level is compatible with a multiplicity of stories at the mental level. This kind of constant conjunction claim, however, explains little. There is, for example, a constant conjunction between increases in the homicide rate in New York City and increases in the rate of ice cream consumption. We could say that the homicide rate supervenes on the rate of ice cream consumption, but we will have explained nothing. We will not have shown that ice cream consumption is responsible for homicides, or vice versa, or whether these are just two unrelated effects of a common cause (an increase in the city’s temperatures).

I should add that a good deal of confusion in the discussion of neuroscientific discoveries and their relation to the philosophy of mind often occurs at this point. What neuroscience if often able to do is provide correlations between certain mental states and activity in certain parts of the brain. These are often taken as proof of materialism, but there is no good reason why dualists should not expect these correlations to exist. Further, it must be emphasized that correlation between mental states and physical states is not the same as identification of mental states with physical states.

Strong supervenience is the claim that B-properties strongly sueprvene on A-properties just in case things that are alike in A-properties must be alike in B-properties. On this view the supervenience isn’t just a brute conjunction, it is necessarily so. However, as an attempt to explain anything, this seems inadequate as well. Religious expalnations are often taken to task as being god-of-the-gaps explanations, this just seem to me to be a necessity-of-the-gaps explanation. “Why, if Jones’s beliefs could be 5 or 6 different ways given the physical, or perhaps, given the physical, Jones could be a zombie with no beliefs at all, does Jones have the beliefs he has?” If the answer is “Well, there’s this strong supervenience relationship that exists between the physical and the mental, so it’s necessary, it looks as if we are taken no closer to an explanation as to why Jones has the beliefs he has.

Why does the supervenience relation exist, if it does? It is pure dumb luck? Is it a Leibnizian pre-established harmony set up before the foundation of the world by God? (This might not be naturalistically acceptable). Presumably, it is not a physical relation, so why does it exist? Unless there is something about the physical that guarantees that the mental be only one way, the supervenience relation needs to be explained.

There is what James Stump calls a “classic reflexivity problem” for the suprevenience theorist. For supervenience theory, everything is either physical, or supervenes on the physical. So, the supervenience relation is going to have to be either physical or supervene on the physical, if supervenient physicalism is true. But does it. Stump summarizes an argument originally presented by Lynch and Glasgow to contend that the supervenience relation itself cannot be admitted into the supervenient materialism’s ontology, which I have altered slightly for the sake of congruence with previous discussion:

1. For physicalist, all fact must be materialistically acceptable. That is, th eyare facts about physical things, or about things which are ontologically distinct from the physical, but strongly supervene on the physical.

2. There must be some fact—the explanation—in virtue of which B-properties supervene on A-properties; call the S-facts. What kind of facts are S-facts? There are two options for materialistically respectable facts:

a) They themselves could suprevene on A-properties. But then there is an infinite regress problem, for now we have to explain this new supervenience relations, which in turn needs to be explained, and so on ad infinitum. So this is no good.

b) Or, the S-facts could not just be further A-properties, that is, facts about the physical entity. But then these facts do not bridge the explanatory gap betweent he B-facts and the A-facts.

Perhaps the supervenience theorist can simply accept the suerpvenience relation as an unexplained brute fact. If so, as Stump suggests, the apparent explanatory advantage of materialism over dualism, based on parsimony, is dissipated. In addition, there are more problems for this position when we come to the problem of mental causation.

Intentionality is more than just a puzzle for naturalism, it is a deep and profound problem distinct from, and as serious as, the “hard problem” of consciousness. Reduction of understood intentional states and propositional intentional states seems to be inherently impossible. Elimination of those states eliminates states essential to the operation of the natural sciences on which the credibility of naturalism is founded. Non-propositional successors to propositional attitudes cannot do the job assigned to them. Supervenient materialism commits the materialist to a materialistically unacceptable relation between the physical and the mental, and, as we shall see, presents serious problems in accounting for mental causation.

Theories of the universe that make the mental basic fact of reality, such as theism, pantheism, or idealism, do not have the problem of unacceptably terminating explanatory chains were mental states. Thus the problem of intentionality provides one good reason for preferring a broadly mentalistic world-view to a broadly materialist world-view.