In reading the recent issue of Faith and Philosophy I came across an interesting argument, tucked away in a footnote, in an article by Graham Oppy. In this article, Oppy is continuing his debate with Koons on Koons's version of the cosmological argument. The passage to which the following is a footnote is on reasons why nontheists would reject the view that the universe, as the sum of all wholly contingent events, has a cause.
Here's a sketch for another argument, this time for the conclusion that no theist ought to accept the claim that every wholy contingent event has cause. This argument relies on the assumption that theists must buy into the free willd efence against arguments from evil. (That assumption, in turn, can be underwritten by Mackie's famous argument: if free will is compatible with determinism, then God could--and hence should--have made a world in which everyone always freely chooses the good.) The free will defence relies on the assumption that people have libertarian freedom, i.e. it relies on the assumption that, when people make free choices, there is nothing in the world which determines or causes those choices. So, consider an occasion on which a person X freely chooses A rather than B. Plainly, the event of X's freely choosing A rather than B is a wholly contingent event--but, as a result of doctrinal commitments elsewhere, the theist is required to deny that there is a cause of X's freely choosing A rather than B. So, by the theist's own lights, it simply isn't true that every wholly contingent event has a cause. Given that theists have good reason to reject the claim that every wholly contingent event has a cause, they are hardly well placed to insist that non-theists ought to accept it.
(Graham Oppy, "Faulty Reasoning about Default Principles in Cosmological Arguments," Faith and Philosophy (April 2004) 249n10.)
This is, as I said, an interesting argument, but I'm inclined to think it does not work. For the libertarian is actually not committed to the claim that there are no causes to their choices; they are only committed to the claim that some choices are not sufficiently determined by causes in the world. By 'being sufficiently determined by a cause' I mean: there is no sufficient cause, i.e., a cause which is a sufficient condition, for this choice in particular. There may, however, be other kinds of causes, e.g., necessary causes; and one form of necessary cause is that there be adequate cause for a choice. And this is one of the big differences between the libertarian and the determinist, I think; the determinist is committed to the claim that all adequate causes are sufficient causes, the libertarian is not. So the libertarian can perfectly well allow that there is a cause for X's choosing A rather than B; namely, that cause or set of causes adequate for choosing A. What the libertarian will deny is that this cause or set of causes adequate for choosing A necessitate the choosing of A; they suffice (in the ordinary colloquial sense) for the choosing of A, but do not suffice (in the technical philosophical sense) for the choosing of A. So I don't think a commitment to incompatibilist free will can be converted into an argument for the claim that some wholly contingent things are uncaused.