I had a very long conversation yesterday with one of our graduate students regarding Weak Supplementation: he finds Weak Supplementation plausible while I find it implausible. Anyway, I felt this was one of those really good philosophical conversations where you get to something at the root of the issue. So, here's where I felt we got to: It is crucial to how one thinks about parthood whether one sees a close connection between parthood and composition. In particular, it is crucial whether one accepts this thesis:
- Necessarily, if an object has proper parts, it is composed of them.
- x is a proper part of y if and only if (and if so, because) there are zs that compose y such that x is one of the zs.
- The zs compose y if and only if (and if so, because) every one of the zs is a part of y and every part of y overlaps at least one of the zs.
On the other hand, I am quite sceptical of (1). I think of parts as derivative from the whole. Instead of wanting to know when things compose a thing, I want to know when a thing has proper parts. One way to put the question, in the case of material substances, is this: Given a thing and a region of space (or spacetime), under what circumstances is there a part that exactly fills that region? And a tempting answer, somewhat reminiscent of van Inwagen's "life" answer, is that this happens when the thing has a function fulfilled precisely in that region.
I then see no reason why the whole would have to be composed of its proper parts. I think I can imagine a stories like the following being true. There is something that looks like a normal human hand. It has five fingers as proper parts. And that's all. There is no such part as the fingerless palm, since the fingerless palm just doesn't have the right kind of functional unity, and I shall suppose that in this scenario there are no cells or particles. The hand then has six parts: the hand itself (an improper part) and the five fingers. Since the hand isn't composed of the fingers, the hand isn't composed of its proper parts. More interestingly, perhaps, I might be persuaded to think of the tropes of an immaterial substance as its proper parts, but I wouldn't want to say that the substance is composed of its tropes--that would be a bundle theory--nor would I want to say that the substance is composed of its tropes and a bare particular. So, then, an immaterial substance might have proper parts, but it wouldn't be composed of them.