Peter van Inwagen's proposed answer to the special composition question (p. 82, Material Beings) is the following: there is a y such that the xs compose y if and only if the activity of the xs constitutes a life. Now van Inwagen says that he doesn't have "too much to say about what it is for the activities of objects to constitute a life"; he "must leave the notion at an intuitive level and proceed by giving examples" (ibid.). Van Inwagen also has "little to say about the ontology of events" (ibid.), other than that events are datable, locatable, concrete particulars. But I think that we are certainly owed more than what he gives us, and that his proposed answer to the special composition question hides a violation two of van Inwagen's own methodological presuppositions. The first presupposition that his proposed answer violates is van Inwagen's demand that the right hand side of an answer to the special composition question contain no mereological terms. The second presupposition that his proposed answer violates is his demand that an answer to the special composition question be "internal", i.e. that it makes no appeal to anything other than the intrinsic properties of the xs and the relationships among the xs.
Van Inwagen claims in the introduction to Material Beings that he has no notion of parthood that applies to entities other than material objects and quasi-material objects such as the elementary particles of physics. But it seems quite clear to me that the notion of parthood that applies to material and quasi-material objects also applies to events. Many of the things which we think of as events are composite entities, i.e. have other non-overlapping events as parts. A baseball game (an event) is composed of innings (events), which in turn are composed of pitches, hits, strikes, catches, and runs. If a composite entity w, composed of x, y, and z, moves without changing shape (i.e. if x, y, and z continue to bear the same spatial relations to one another), then the movement of w is surely composed of the movements of x, y, and z. And if an entity continuously changes temperature from 10 degrees Celsius to 30 degrees, then its rise in temperature from 10 degrees to 30 degree is surely composed of its rise from 10 degrees to 20 and its rise from 20 degrees to 30.
Now consider van Inwagen's examples of when the activities (events) of x's "constitute" some other event. "The activities of the Household Cavalry and the Life Guards constituted the parade" (ibid.). This seems to me to be a clear case of some events, the activities of the Household Cavalry and the activities of the Life Guard, composing another event, the parade. If I am correct, and van Inwagen's notion of "constitution" is a kind of composition, then van Inwagen has at best given an incomplete answer to the special composition question, for the right hand side of his proposed answer contains a mereological notion. Van Inwagen owes us an answer to the special composition question for events: when do events compose an event? If this notion of "constitution" is not a kind of composition, then I must say that I do not understand this notion.
Before going on to look at whether van Inwagen violates his own "internalism," I wish to correct a few other errors in his formulation of the proposed answer and his account of constitution. The first error I see is that he gives the formulation as a definite description: "the activit[ies] of the xs constitutes a life", which presumably implies that all of the activities of the xs are "part of" (I use scare quotes just in case I am wrong about my thesis that "constitution" is a kind of composition) some life. Now we know empirically that at least some of the particles of which I am presumably composed will continue to exist even after my life is over, and that they continue to participate in events. But it would be wrong to say that the post-mortem activities of the particles that now compose me are "part of" a life: they might be, if those particles come to be part of some other living being, such as the bacteria which will consume this mortal flesh after my death, but this is not necessary. So the proposed answer should be reformulated as follows: there is a y such that the xs compose y if and only if some of the activities of the xs constitute a life. Or, more precisely, to rule out the possibility that there is something composed of "my" particles and some particles out in the Andromeda Galaxy (for some of the activities of these particles constitute a life): there is a y such that the xs compose y if and only if there are some zs such that each of the zs is an activity of one of the xs and the zs constitute a life.
A second error I see is in a parenthetical addendum to one of the examples van Inwagen uses to explicate his notion of event constitution: "the activities of the members of the Household Cavalry and the members of the Life Guards constituted the parade (thus the activities of the xs may constitute a certain event and the activities of the ys constitute that same event, even though nothing is both one of the xs and one of the ys)" (ibid.). I completely fail to see how this could be so: if constitution is a kind of composition, then this cannot be, for then one thing could be composed (at the same time, or over the same interval) of "different stuffs". If van Inwagen were to insist on sticking by this thesis (which is by no means crucial to his position), this seems to me to render constitution even more mysterious than the composition it is used to explain.
The third error I see is in a concession I think he need not make. Alluding to Davidson's example of a sphere which rotates and cools over a certain interval, van Inwagen states the following: "I do not presuppose that if the activities of the xs constitute a certain event at t, then that event is the only event they constitute at t" (p. 83). If the sphere is composed of molecules, and the rotation of the sphere is a distinct event from the cooling of the sphere, then van Inwagen thinks we must say that the activities of the molecules constitute two distinct events. It is again van Inwagen's use of "the" that leads to this concession. (Again, if I am right that constitution is a kind of composition, then to preserve uniqueness of composition we must not make this concession.) Rather than saying that the activities of the molecules constitute two different events, we should say that some of the activities of the molecules (their lowerings of temperature) constitute one event, the cooling of the sphere, and that some other activities of the molecules (their changings of spatial position) constitute another event, the rotation of the sphere.
If I am right that event-constitution is a kind of composition, then this means at worst that van Inwagen has given an incomplete account of composition. But I think there is a more serious difficulty for van Inwagen, in that his proposed answer violates his own internalist maxim: to determine whether the xs compose something, we need only "look at" the xs; we need not know anything about any y not among the xs to know if the xs compose something. Consider now all of the particles that van Inwagen thinks compose me except for one, call it P. Let these particles be the xs under consideration: do these xs compose something? Van Inwagen would certainly maintain that they do not, for he denies that I have any undetached parts other than my cells and "my" elementary particles: if these xs composed something, surely it would be a part of me. So van Inwagen must deny that the activities of these xs constitute a life. But why don't they? If P were to cease to exist, I would continue to live; and (counterfactually) if P did not exist, I still would. The only reason I can see for maintaining that (some of) the activities of these xs do not constitute a life is that there is a y such that (some of) the activities of the xs and y do constitute a life. (Or, to be more precise, there is a y and there are zs such that each of the zs is an activity of one of the xs or of y and the zs compose a life.) We might say that the activities of the xs are not "life-maximal." (Actually, this will not do as a reason for denying that the xs compose something, since the activities of the particles that compose one of my cells are not "life-maximal" either; but (some of) the activities of these particles do constitute a life, according to van Inwagen. I'll assume there's some way to take care of this problem.) But if this were the reason given, van Inwagen would violate his internalism about composition. For to determine whether these xs (all my particles except for P) compose something, we must not only look at (the activities of) the xs, we must look at (the activities) of P.
Back to Brock's Philosophy Page