Causation: Reductionism Versus Realism

Any adequate approach to causation must provide accounts of causal laws, and of causal relations between states of affairs, or events, and, in each case, one is confronted with the choice between reductionism and realism. With respect to causal laws, the relevant issue concerns the relation between causal laws and the totality of events. According to reductionism, causal laws are supervenient upon the total history of the world. According to realism, they are not. With respect to causal relations, the central issue is whether causal relations between events are reducible to other states of affairs, including the non-causal properties of, and relations between, events. The reductionist holds that they are; the realist that they are not. These choices between reductionist and non-reductionist approaches to causal laws and causal relations are surely among the most fundamental in the philosophy of causation. But in spite of that fact, they have received very little discussion. For, although there have been exceptions, the history of the philosophy of causation since the time of Hume has been largely the history of attempts to offer reductionist accounts of causal laws and of causal relations, and most philosophers have been content simply to assume that a reductionist approach to causation must be correct. In this paper, I shall argue that reductionist accounts of causation are exposed to decisive objections, and that the time has come to explore realist alternatives.