In this paper I shall put forward a version of the so-called logical connection argument to try to show that desires cannot be causes of actions. Now for anyone who wishes to use that kind of argument to show this, the ideal way of proceeding would be to first set out a complete analysis of the causal relation, and then to go on to argue that the relation between desires, certain other conditions, and actions fails to match some essential feature (...) revealed by that analysis. But I cannot proceed in this way; for I do not possess any full-blooded analysis of the causal relation. Nonetheless, I do not think that this places me in a position of real weakness. For the argument I shall present requires the truth of only one claim with respect to the nature of the causal relation. This is the claim that the causal relation is a contingent relation. I think that this claim is a relatively uncontroversial one; I am going to assume that it is true; and the argument which follows will exploit that claim. (shrink)