Monday, November 8, 2021

Top-down mereology and the special and general composition questions

Van Inwagen distinguishes the General Composition Question:

  • (GCQ) What are the nonmereological necessary and sufficient conditions for the xs to compose y?

from the Special Composition Question:

  • (SCQ) What are the nonmereological necessary and sufficient conditions for the xs to compose something?

He thinks that the GCQ is probably unanswerable, but attempts to give an answer to the SCQ. Note that an answer to the GCQ immediately yields an answer to the SCQ by existential quantification over y.

There are two main families of mereological theories:

  • Bottom-Up: The proper parts explain the whole.

  • Top-Down: The whole explains the proper parts.

Van Inwagen generally eschews talk of explanation, but the spirit of his work is in the bottom-up camp.

It’s interesting to ask how the GCQ and SCQ look to theorists in the top-down camp. On top-down theories, the xs that compose y are explained by or identical to y. It seems unlikely to suppose that in all cases there would be some relation among the xs that does not involve y which marks the xs out as all parts of one whole. That would be like thinking there is a necessary and sufficient condition for Alice, Bob and Carl to be siblings that makes no reference to a parent. Therefore, it is likely that any top-down answer to the SCQ must make reference to the whole that is composed of the xs. But if we can give such an answer, then it is very likely that we can also give an answer to the GCQ.

If my plausible reasoning is right, then on top-down theories either:

  1. An answer can be given to the GCQ, or

  2. No answer can be given to the SCQ.

1 comment:

danielm said...

This jives with the general thrust of my Master's thesis where I attempt a mereological formalization of living things in loosely Aristotelian terms. The identity of substances is taken to be basic and primary (i.e., I reject extensional identity); the identity of two parts x and y is parasitic on the identities of the substances they are parts of. Given some x and some y, x = y only if there is some S of which both are parts. This condition is, of course, insufficient to establish identity between x and y; both still need to occupy the same place in the whole, as it were, but any such determination also defers to the identity of the substance in like manner.