Language, Thought, and Other Biological Categories: New Foundations for Realism

And probably for related topics. Millikan's account of functions and "Normal" functioning is extremely well-developed, and can provide much needed content to "functionalist" accounts of various kinds, helping greatly in evaluating their merits and their defects. Her notion of a function also enables a unified account to be given of functions for which things are designed and for things which are not designed (obviously without the theological dodge of saying everything is designed).

The book explains why the reproduction of tools, actions, adaptations and habits can matter separately from single instances of use, without referring to any specialized internal mechanisms, as Fodor or Chomsky might require.It thus throws an enormous weight of exemplary philosophical junk (Dennett might claim most of the literature on free will goes here) in the dumpster, by showing that a battery of single counterexamples can be irrelevant to a reproductive motive.The book also defines "function" by referring to reproductive motives, not use motives. A mass of literature referring to function becomes clearer thereby. Dysfunction becomes far less relevant than one might expect when one sees "dysfunction" opposed to "function", as if a law of contradiction applied.I like the formalism in the book, which Millikan seems to have felt compelled to softpedal in her subsequent writings. In a way, Millikan does for "function" here what Abraham Robinson did for infinitesimals. She rehabilitates an aid to intuition, so that people who might be inclined to deny it because it lacks a formal well-definition might have to admit it.The context is biological, i.e., survival and posterity matter more than origins in the mist, a process is step by step, and ideas can persist despite cases of failure.