Functional explanation and the problem of functional equivalence.

The legitimacy of functional explanations in biology is threatened by a problem first identified by Hempel: the problem of functional equivalence. In order for the prevalence of a trait to be explained by its function, the function would have to explain why that very trait is prevalent and not some other functionally equivalent trait. But functions alone cannot meet this explanatory demand. I argue that this is a problem not only for Nagelian deductive-nomological models but also for etiological models of functional explanation. I contrast these models with a dual model of adaptive explanation and design explanation. This dual model largely circumvents the problem of functional equivalence, but divests functions of much explanatory power.