Possibilities haunt the human sciences. In empirical enquiry, they are suggested by the explanations we offer, and might even be said to support those explanations. (The Armada Invencible, the Spanish thought, was beaten by the weather. ‘I sent it to fight the English’, complained Philip II, ‘not God.’ If, therefore, the weather had been good …) In practical reflection, possibility is itself the point. If we think clearly, can summon the will, and have the right conditions, we can do what it is that we want to do; if not, not. In each kind of case, however, the possibilities remain uncertain. In the first, we cannot usually experiment. In the second, where we sometimes can, the experiment can prove a costly failure. This, it will be said, is why we have theories. They support our explanations and guide our practice. But theories have themselves to be supported. Counterfactuals and other kinds of subjunctive conditional will not go away.

In this book, I consider them. In chapter 1, I briefly consider the ways in which others have done so, and introduce my own argument. This starts from the claim that in History, sociology, and the study of politics, understanding possibility is at the heart of understanding itself. Although this is in a sense a theoretical issue, no theory, I suggest, can resolve it.