Nothing is greater than God in any sense. But a whole is greater at least in some sense than a proper part. Therefore:
That (3) follows is pretty easy, too. Consider the true proposition that God created elephants (or that God exists and elephants exist). If this has a truthmaker, that truthmaker contains God as a part. But that truthmaker cannot just be God, since if x is a truthmaker for p, then that x exists entails that p is true, while that God exists does not entail that God created elephants. So, the truthmaker would have to contain God as a proper part, which would violate (1). The argument leaves open the possibility that all true propositions are made true by one or more entities, so that the proposition that God created elephants might be made true by God and elephants (not considered as a composite object, but simply as a plurality). But it's still the case that the proposition lacks a truthmaker.
Finally, (4) follows from the observation that there are de re propositions about God, such as that God has created us.
- God is not a proper part of any whole.
- Unrestricted compositionality is false.
- Some true propositions have no truthmaker.
- At least some de re propositions do not contain the object that they are about as a part.
That (3) follows is pretty easy, too. Consider the true proposition that God created elephants (or that God exists and elephants exist). If this has a truthmaker, that truthmaker contains God as a part. But that truthmaker cannot just be God, since if x is a truthmaker for p, then that x exists entails that p is true, while that God exists does not entail that God created elephants. So, the truthmaker would have to contain God as a proper part, which would violate (1). The argument leaves open the possibility that all true propositions are made true by one or more entities, so that the proposition that God created elephants might be made true by God and elephants (not considered as a composite object, but simply as a plurality). But it's still the case that the proposition lacks a truthmaker.
Finally, (4) follows from the observation that there are de re propositions about God, such as that God has created us.
I think this view deep-sixes all truthmaker theories, for the following reason. "God created elephants" and "elephants created God" are different propositions. One is true, the other is false. The truthmaker for the true one cannot just be the plurality of God and elephants since you have to distinguish between the propositions. So if there is a truthmaker, it is a state of affairs or a set or ordered pair or some other structured entity. But any such structured entity will have God as a part, which ex hypothesi cannot be. So there can be no truthmaker for "God created elephants." And it's a pretty short step to no truthmakers at all.
ReplyDeleteI think that's too quick. Truthmaker theorists will have a structured entity like God creating elephants, which differs from elephants creating God. This structured entity won't just be the mereological sum of God and elephants.
ReplyDeleteSurely, by the way, singular existential claims have truthmakers.
ReplyDeleteGranted, the structured entity might not be a mereological sum. What puzzles me is how it could avoid having God as a part.
ReplyDeleteAnd on singular existentials: obviously this is the most tempting case of truthmaker theory. But if the theory can't generalize successfully, it seems to me one ought to ditch the whole thing.
I have no strong views on truthmaker theory, by the way.
Probably couldn't, so it's subject to the problem in my post. But I thought you were pointing out another problem, the problem with non-symmetric relations (which some people are indeed worried about).
ReplyDeleteBy the way, it's my feeling that truthmakers are getting to be passé in the metaphysics community. Some people still talk of them, but there is more and more of a switch to talk of grounding. Grounding is hot stuff these days.
ReplyDeleteI am not sure what I think of grounding. Dan Johnson did an excellent dissertation with me that defended a concept of "ontological explanation" that does some of what grounding is supposed to do, but that also has some interesting advantages. (The most obvious formal difference is that if p grounds q, then p entails q but not conversely; but Johnson's ontological explanation has two-way entailment.)