Whose Contingency Is It Anyway?

William F. Vallicella

once accused David Bentley Hart of conflating two distinct notions of contingency in his contingency argument for God. Specifically, Hart seems to move from the idea that something could have been otherwise or not at all to the claim that such a thing is therefore dependent upon an external cause. Let’s call the first kind of contingency "modal contingency" and the second "dependent contingency."¹ Bill argues—and I agree—that it’s too quick to move from one to the other; just because something could have been otherwise or not at all doesn’t necessarily imply that it "hangs" upon something else, so to speak. We need some justification for linking these two understandings of contingency; otherwise, Hart’s cosmological argument will not hold up. (Hart isn’t the only one who makes this mistake; In How to Think About God, Mortimer Adler similarly assumes, without significant justification, that because the physical universe could have been otherwise, that it must have an outside cause.)

Another way to see the distinction is this: it seems reasonable to ask whether things that occur in some possible worlds but not others (if one will excuse possible worlds semantics) require an external cause or determining factor. If this question makes sense to you (as it does to me), it's likely because you recognize the two different notions of contingency.

So, why might one think that things which occur in some but not all possible worlds require a cause? In other words, why think whatever is modally-contingent is also dependently-contingent?

There are several ways this linking project might be accomplished.

First, one might appeal to what we could call "metaphysical intuition": doesn’t it seem obvious (or perhaps self-evident, or at least strongly plausible) that things which could have been otherwise or not at all must be determined to be as they are by something external?²

Honestly, I think this response is perfectly satisfactory, but I doubt it would convince a determined skeptic. Not to say it shouldn’t, but we know how skeptics can be… ; )

Second, one could defend a broad causal or explanatory principle that would, at minimum, wrap around all modally contingent things, justifying the need for an external cause. Here, something like the Principle of Sufficient Reason (PSR) could do the trick. This is one of the approaches I take in The Best Argument for God, so, obviously, I think it’s a good one. (You can see my book for more details.)

Third, one could roll up their sleeves and work out a theory of modally contingent beings—examining their nature, for instance—and see if justification for their requiring a cause necessarily follows. Here’s an idea (and as usual, expect some Thomism from me): perhaps, following Aquinas, we come to understand beings as modally contingent because there is a real distinction between their essence and their act of existence. Thus, they are possibly non-existent even if they exist, since nothing about their essence guarantees their existence. In that case, it seems reasonable to look for an external cause of their existence.

But does that really solve the issue? Why not just say it’s a brute fact that these essences have existence, rather than attribute it to a cause? If brute facts were acceptable (were they?) before concerning why some things occur in some but not all possible worlds, why can’t they be acceptable here? If not, then we’re just back to justifying something like the PSR—which means this strategy is functionally no different from the first, however interesting that interim metaphysical theorizing may have been.

Good point! So, let’s try another approach. What if it turns out that such metaphysical composites are actually contradictory structures unless they are caused in their existence? If so, that would provide a PSR-free linking path, justified through metaphysical analysis. What would this contradiction amount to? Barry Miller thought he could draw out the contradiction that uncaused composites are at once conceptually constructible from parts and yet not conceptually constructible from parts. It’s a terribly complicated argument embedded in a terribly difficult book to locate these days, From Existence to God; however, it has the redeeming features of being both brilliant and undoubtedly sound.

But all that is a matter for another time. The only point I wanted to make in this article is that Vallicella is correct in his criticisms of Hart’s approach to God while offering some ways to patch it.

1. Vallicella’s understanding of something being dependently contingent seems equivalent to it being “deeply dependent” or strongly grounded.

2. Of course, it’s difficult to define exactly what metaphysical intuition is; nevertheless, people often appeal to it when they believe a claim seems self-evident or strongly plausible, even if it’s hard to provide a strict logical proof. For example, the idea that "everything that begins to exist must have a cause" might be something we grasp through metaphysical intuition.