Abstract

A causal relationship is proportional just in case the cause is neither overly nor insufficiently detailed with respect to its effect. Two influential principles that appeal to this notion are: (i) all genuine causation is proportional, and (ii) causal explanations that appeal to proportional relationships are better than those that don’t. Both principles have featured fruitfully in the philosophy of science. The former has been used to argue that the sort of high-level causes identified by the special sciences are not ‘outcompeted’ by causes at the level of fundamental physics. The latter has been used to account for the scientific preference for explanations of high-level phenomena in terms of causes that are situated at a similarly high level.

In spite of their successes, I identify several apparently disparate problems with both principles, some well-recognised in the literature, others novel to this paper, and tell a story that renders these problems legible as natural consequences of the structure of proportionality cases. I thus unify the seemingly independent threats these problems pose to principles (i) and (ii). On my view, the root of the problem with proportionality constraints is that they fail to take into account how both overly and insufficiently detailed events can cause their effects by way of the proportional causes of those effects. I close by drawing some lessons about the connection between explanatory grain and explanatory distance, and say something about how ‘close’ explainers should be to that which they explain.