Monday, April 19, 2021

Modal Collapse and Possible Worlds

 Omar Fakhri has an interesting paper on what is known as 'the modal collapse problem' for divine simplicity. Fakhri gives one common version of it: 

(P1) God is intrinsically identical in all possible worlds, i.e. God does not intrinsically change (implied by ADS). 

(P2) The same identical cause brings about the same effect (premise). 

(P3) God is the cause of this universe (premise). 

(C) Therefore, God causes this universe in all possible worlds.

Modal collapse is not actually any kind of problem for divine simplicity ('ADS' in the above stands for 'Absolute Divine Simplicity', which is the name analytic philosophers give to one possible translation of the doctrine of divine simplicity into the preferred vocabulary of analytic philosophy; it is one that I have noted before is not very accurate); setting aside the fact that what counts as identity across possible worlds is a non-trivial question, the argument is based on a straightforward confusion. (P1) appeals to the 'possible worlds semantics' common among analytic philosophers; this makes it natural to think that (P3) is about a possible world. But (P3) is most naturally read on its own being about the actual universe. The actual world is not a possible world in the possible worlds semantics sense, because 'possible worlds' in that sense have to be ways the actual world could be. So all (C) could mean is that every way the actual universe could be involves God causing it. This is not any kind of modal collapse at all; the modal collapse arises from thinking that the actual world is a possible world in the sense used in possible worlds semantics. It is not. But if you assume it is, you get guaranteed modal collapse in any argument in which you assume it.

But Fakhri looks in particular at (P2), arguing that rejecting (P2) leads to unacceptable explanatory gaps. Fakhri rightly notes that it would matter whether you accepted deterministic causes or indeterministic causes; although, to some extent that it is pretty straightforward, since assuming determinism under most metaphysical accounts of determinism directly implies modal collapse. An indeterministic cause can have different possible effects, though. Even if there were only one indeterministic cause, this would necessarily imply that there have to be multiple 'possible worlds' to capture 'ways the actual world can be'. Indeterministic causes are what explain there being distinct possible worlds at all. I think, however, Fakhri gets tangled up in the actual/possible confusion:

Merely pointing to the nondeterministic relation between the cause and effect does not explain the following: what is it about the second possible world that explains why the improbable occurred rather than the probable? ... In the indeterministic case, we are asking ADS for a cross world non-contrastive explanation. The explanatory gap objection is not requesting an explanation of why God brought about this universe rather than another universe. This contrastive question is asking for a reason that would show why God prefers to create this universe over another universe (or no universe at all).

This question seems pretty clearly to confuse the actual world with a possible world. What explains the second possible world (i.e., a possible world in which the improbable happens) is, by definition, the fact that it is possible for some cause to cause what is improbable. There is a possible world for every different possible way the actual world can be, by definition; how probable or improbable it is, is irrelevant to there being such a possible world. But then the question needs to be disambiguated, with the possibilities, "Why does God prefer to create the actual universe over a non-actual one?" (which is not a question anyone needs to worry about, since 'creating a non-actual universe' is not a thing), "Why does God create this possible universe over another universe?" (which is not a question anyone needs to worry about, since creation is of the actual universe, not of possible universes), and "Why does God create anything at all?" The third is a question you could ask, and is equivalent to asking why anything other than God exists; it is not the kind of question, however, that Fakhri has in mind in talking about the explanatory gaps, because it's not a contrastive question.

(P2) in fact directly causes a modal collapse; it implies that no cause is capable of causing more than a unique effect -- that is, each cause has one and only one possible effect, so every effect was the only possible effect of its cause. In reality, if you have allowed an indeterministic cause, you have already denied (P2): you have granted that one and the same cause can have more than one possible effect. That's just what an indeterministic cause is. And every non-collapse account of 'ways the actual world could be' will posit at least one cause such that more than one result is possible from it.