Providing explanations of the phenomena we experience is a major accomplishment of science. A theory should be good for making sense of the world, for explaining why things happen as they do, if it is good for anything at all. Scientific curiosity begins with why questions. Why does this happen? Why are things this way? Scientific explanations are answers to why questions. Explanations are an accomplishment, and they are worth working for because they enhance our understanding of the world.

Explanations, in other words, are desirable in themselves as valuable ends. You want your theories to explain things because that's what you want, just as you might buy shiny apples because shininess is what you want. Explanation is a desirable end, but it is also a means to something else. It is useful as an indicator of truth, as shininess might be an indicator of freshness and good taste. If an apple is shiny it is likely to be fresh, so even if shininess for shininess' sake didn't matter, shiny apples would be desirable for taste's sake. Similarly, it is a good thing for a theory to be explanatory, not just for explanation's sake but for truth's sake as well. A theory that explains phenomena is more likely to be true than one that doesn't. That, at least, is the claim if explanation is an external, truth-conducive virtue.