In Physics 2.4–6, Aristotle offers an account of things that happen “by luck” and “spontaneously”. Many of these things are what we might think of as “lucky breaks”: cases where things go well for us, even though we don’t expect them to. In Physics 2.5, Aristotle illustrates this idea with the case of a man who goes to the market for some reason unrelated to collecting a debt he is owed. While he is there, this man just so happens to (...) run into his debtor and get his money back—which is what he “wanted” all along. This case has a number of features that have proved puzzling. Most notably, Aristotle seems to think that while “it happened accidentally to him to come and to do this for the sake of getting the money”, nevertheless the man “did not come for the sake of this”—though he “would have”. What must such a proceeding be like to be described in these ways? Physics 2.4–6 makes several important moves towards answering this question. One of them—among the earliest and seemingly a foundational one—is the claim that “for the sake of something are as many things as could be done by thought and by nature”. The aim of this paper is to identify the role that this claim plays in Aristotle’s account of spontaneous proceedings. It defends two main claims. First, Aristotle’s suggestion that spontaneous proceedings could be done by nature and by thought is the product of a more general strategy. His approach is to sketch an intuitive pattern of explanation for proceedings that happen for the sake of something in the ordinary—non-spontaneous—way, and then to try to extend that model to the case of things that do so spontaneously, preserving as much of it as possible. Applying this strategy reveals that whereas nature and thought in fact function as efficient causes of ordinary proceedings that are for the sake of something, this cannot be the case in spontaneous ones. Instead, nature and thought could function as their causes. Second, I argue that Aristotle’s implementation of this strategy is rooted in his account of causation and especially Physics 2.3’s list of “modes” of causes. When he claims that spontaneous proceedings “could” be done by nature and by thought, he is saying that nature and thought are their efficient causes in capacity. In the practical case, spontaneous proceedings are caused by agents who have certain capacities, connected with their desires and abilities. In the right circumstances, these agents would actively exercise those capacities in acting for the sake of a given end. (shrink)