Functions are Explanations: An Ode

George C. Williams’ death makes me reflect a little on adaptationism, which has had such a profound influence on me and the way that I – and many others – think about biology and psychology.

Encapsulating the logic of adaptationism briefly is difficult, but at its heart, the lesson of his book, Adaptation and Natural Selection is that by hypothesizing an evolved function, you commit to a pattern of observable data, rendering your hypothesis falsifiable. If you think that the thing in the center of a human chest is for pumping fluid, well, it better darn well have valves to stop the backflow of fluids. This logic lies at the heart of evolutionary biology, and of course evolutionary psychology. (If you think jealousy is for deterring infidelity, it better have properties that make it good for doing so.)

The idea remains misunderstood, perhaps most famously in charges that hypotheses about function are “just so” stories, a change made by Stephen Jay Gould, and echoed later by others, including other prominent scholars such as Noam Chomsky and, more recently, Jerry Coyne, who claimed that evolutionary psychology consisted of “untested – and probably untestable – speculations.” Other critics echo such lines, and work in evolutionary psychology is often casually dismissed by critics.

This is not, of course, to say that all hypotheses about function are equally good. Hypotheses about function can be suspect because they are inconsistent with known facts, are stated too broadly to make good predictions, do not identify functions in the biological sense, and so on.

Nonetheless, it seems that there remains a gap between the way adaptationist logic as laid out by Williams is used by evolutionary psychologists and the way it is understood by people reading it. One problem is that it often appears that consumers of evolutionary psychology think that practitioners care about how a trait evolved, and that our hypotheses are intended to answer this question. This explains why “behavior doesn’t fossilize” is such a common refrain.

There is a sense in which Williams’ idea has enjoyed rousing success. Certainly a case can be made in the context of the that biology is currently practiced. But it seems to me that Williams’ work is not done, and probably many of us in the field are responsible. We have not been able to articulate Williams’ insights about the relationship between theory and data in a way that allows critics to understand the epistemic foundations of our field.

It makes me think we need to work harder do so. If for no other reason, than we owe it to the memory of George C. Williams.

• Tweet