IV—Causes and Coincidences

A cause is something which ensures that its effect is no coincidence. This remark is a platitude but it is one which finds no place in the most popular philosophical analyses of causation. And if we assume that (a) coincidences have no causal explanation and (b) causation just is causal explanation, then we must deny that a coincidence has a cause. But it is not mere common sense and my adherence to (a) and (b) which leads me to deny causes to coincidences. Any adequate theory of causation must enable us to distinguish the causally efficacious properties of objects from their causally inefficacious properties. I shall argue that current theories fail in this regard and they fail because they allow that coincidences have causes. To avoid their difficulties, we must adopt the Aristotelian view that nothing can bring about a coincidence.' In the first section of the paper, I say something in support of (a) and (b). In the second section, I try to reformulate the traditional analysis of causation so as to enable it to distinguish causal from constitutive relations. The third section introduces two counterexamples to the traditional theory which do not involve relations of constitution and criticises a number of attempts to deal with these examples. The final section suggests that we deal with them by denying causes to coincidences.