Recently John Gibbons published an article titled “Mental Causation without Downwards Causation” (Philosophical Review, Vol. 115. #1), in which he argues that mental properties are only the cause of other mental properties, and never of physical properties. This amounts to a kind of epiphenomenalism, in which the mind runs its own course, and is never really the cause of physical states of affairs (for example my thoughts couldn’t be considered to be the cause of this paper, since it is composed of completely physical properties). I disagree, but I can’t simply dismiss his claims as based on a flawed conception of the mental (say a dualistic conception), since he and I both agree that mental properties supervene on the physical.
Before I begin though allow me to illuminate exactly how I interpret “supervene”. There are several ways to view supervenience, but the definition I will use here is that a property P is a supervenient property when P can be “realized” by different collections of physical properties, and that the only conditions of P being “realized” are physical properties. For example the color green is supervenient on the physical properties of surfaces. Only surfaces with the correct physical properties count as green, and the only reason a surface may or may not count as green are the physical properties. The claim that the mental is supervenient on the physical is of course a bit more complicated than the case of color, mental properties supervene on functional ones, which supervene on the physical. Even though the mental does supervene on the physical the relationship is much more complicated than the case of color, even though they are of the same kind.
With that out of the way we can now address the paper’s claims head on. First I will address the core deduction of the paper which proceeds from two premises. The first premise is that the causes of physical properties are solely physical properties, which I will accept given that I am a materialist, and thus think that the physical universe is causally closed. The second premise is that supervenient kinds are not physical, since they cannot be identified with a physical kind. This I will also accept. From these premises Gibbons concludes that supervenient kinds are not casually relevant to the physical properties of physical kinds. I argue that these premises do not entail this conclusion, because at a particular moment the supervenient property can be said to be identical to the physical properties that realize it, even if in general it is not identical to those properties, and thus be said to be the cause of physical properties. In other words when analyzing a specific system at a specific moment if a supervenient property, say the color green, was being realized by a specific collection of physical properties, say the chemical composition of its paint, then it would be natural to identify that instance of the supervenient property with the properties that are realizing it in the moment, although not in general, that is a specific instance of the color green can be identified with the chemical composition of the paint for a specific object at a specific moment. Perhaps the best way to see this is with an example. Let us consider a ball painted with an unusual green paint. It is dropped into a liquid that reacts only with this specific paint and combusts as a result of that reaction. What was the cause of the ball combusting? Well clearly it was the paint, because if the ball hadn’t been painted in that way it wouldn’t have burnt. However the cause can also be said to be that it was colored green, because if it hadn’t been green it wouldn’t have burnt (because it would have had to be some other color, which wouldn’t have reacted).
Now let me turn to the specific examples that seem to support Gibbons’ conclusion. First there is the example of the bridge. A bolt on a bridge snaps with a loud sound, causing the bridge to collapse. It would seem that the description of the bolt snapping suddenly supervenes on the bolt snapping suddenly and loudly. And, according to Gibbons, the bridge’s collapse is caused only by the bolt snapping suddenly, not suddenly and loudly, thus supporting his conclusions. It is true that the bolt snapping suddenly causes the bridge to collapse. If the bolt hadn’t snapped suddenly the bridge wouldn’t have collapsed. However, it is also the case that if the bolt hadn’t snapped suddenly and loudly the bridge wouldn’t have collapsed ether, and thus the bolt snapping suddenly and loudly can also be considered a cause of the collapse. The reason this is so is because when we consider what would have happened if the bridge hadn’t collapsed suddenly and loudly we must take away both the suddenness and the loudness (not simply one or the other, we are not negating a logical connective here). The pigeon example can be resolved similarly. We are asked to consider a pigeon trained to peck at scarlet. Scarlet supervenes on the color red, and so when the pigeon pecks at a scarlet triangle is the cause the triangle’s being red or being scarlet? Again I would say both, because if the triangle hadn’t been scarlet the pigeon wouldn’t have pecked, and if it hadn’t been red the pigeon wouldn’t have pecked either (because then it certainly wouldn’t have been scarlet), and thus I conclude that both the triangle’s redness and its scarlet-ness are causes of the pigeon’s pecking.
So then, what do I think? Well, I am led to conclude that besides physical to physical causation there is also mental to physical causation, physical to mental causation, and mental to mental causation, and that these last three causal connections all supervene on the physical to physical causal connections. I will however leave a defense of that position for another time. I also encourage you to read Gibbons’ paper for yourself, since it is an interesting piece even if you disagree with its conclusions.