When we explain something we use hypotheses, that is representations, which we formulate on the basis of some background theory or other. What makes such representations explanations are facts in the world. If, for instance, I can appeal to the hypothesis that the second referee has been unfair as a way of accounting for why Claudio is upset. and such a hypothesis is a (good) explanation if and only if the second referee has been unfair. It follows that even if scientific theories can always be

more »

... read in an instrumentalist sense, in using them to formulate explanatory hypotheses we often appeal, implicitly or explicitly, to readings of them that go beyond their empirical content.