Abstract

The ontic causal-mechanical view of explanation holds that models are explanatory insofar as they convey information about those mechanisms responsible for a phenomenon of interest. A common criticism of this view is that it fails to account for the explanatory practices of science. Scientists do not aim to develop complete and accurate models of causal mechanisms, the criticism goes, but rely on abstract and idealized models to explain instead. I argue that this criticism misses precisely the point that proponents of the ontic view have taken pains to convey – that explanation and understanding are distinct. While an explanation is the target of an explanatory model, certain features of that model (such as its abstract and idealized nature) may lead to greater understanding. The goals of explanation and understanding impose distinct constraints on modeling practices. The ontic view is able to account for this dual set of constraints, while the epistemic view (the position commonly adopted by the critics) fails to do so unless it is supplemented by a set of ontic commitments about the sorts of things that are the proper referent of an explanatory model.