An Argument Against the Unification Account of Explanation
Abstract: This paper argues that an increase in the known unifying power of a theory is often not accompanied in an increase the perceived quality of its explanations. The theory may explain many new things, but it does not explain the old things any better just because it now explains the new things. This strongly suggests that unification accounts of explanation are mistaken. I conclude with a discussion of the explanatory role of unification in science. A surprising consequence of the discussion is that unification has explanatory role only in a non-Humean world, a world with real causal powers and laws of nature.
See An Argument Against the Unification Account of Explanation (PDF)