Wednesday, April 1, 2009

Further reading on group selection

I'm posting in advance of class today because I likely won't have time afterwards. Looking forward to today's discussion on group selection and altruism, as it raises a fascinating tangle of empirical and conceptual issues at the interface of biology, sociology, and philosophy.

Group selection and altruism is a topic on which my thinking has been very much shaped by reading a few key people. So if you're keen on reading more on this topic, I suggest you have a look at:

The work of young British philosophy hotshot Samir Okasha, especially his terrific 2006 book Evolution and the Levels of Selection. This book is already becoming the standard reference on the topic. Most of his papers can be downloaded from his website, so if you don't want to buy or read the book you can easily access individual papers to get snapshots of Okasha's views on key issues.

The work of philosopher Eliot Sober (Marc Ereshefsky's doctoral supervisor), especially his book Unto Others with evolutionary theoretician David Sloan Wilson. Sober basically invented contemporary philosophy of evolutionary biology (in his classic 1984 book The Nature of Selection), and David Sloan Wilson did as much as any biologist to revive group selection as a respectable idea.

The exchange of views between Eliot Sober and evolutionary biologist John Maynard Smith in The Latest on the Best (MIT Press, 1987) is a good non-technical introduction to certain key issues such as what counts as a "group" (yes, that's actually a non-trivial issue).

Richard Dawkins' The Selfish Gene. At the time it was published (1976), this was the final nail in the coffin of group selectionist thinking. Wonderfully clear and forcefully argued, even if some of the conclusions aren't widely held today (which isn't to say that today's group selectionist thinking bears more than a passing resemblance to the naive views that Dawkins was attacking).

Finally, I'll share a little anecdote. As a grad student (this was in the late '90s), I was fortunate enough to have the chance to discuss group selection with Dick Lewontin. Lewontin and Sober wrote a famous 1982 paper about group selection, which addressed the issue that one can often build mathematically-equivalent group-selectionist and individual-selectionist models of the same biological situation. That is, the models make the same predictions about future evolution, but do so by making different assumptions about the underlying causal processes driving evolution. The existence of such predictively-equivalent models could be taken to suggest that the distinction between group and individual selection is merely conventional, a matter of choosing one heuristic perspective over another. In contrast, Sober and Lewontin took a realist view and argued that, in the case of a specific example (heterozygote advantage), the group selectionist model "gets the causal facts right" while the individual selectionist model is simply a mathematical fiction that gives the right predictions for the wrong causal reasons. I'd read the paper and thought a lot about it, and so I was excited to ask Dick Lewontin himself if he still took the same view. He didn't. He'd changed his mind and decided that, at least in the case of the example of heterozygote advantage, that there were no "causal facts" that one could appeal to in order to motivate a choice of one model over the other. I don't know that Lewontin has ever published his revised view, so I don't know if it's widely known that he changed his mind.

Postscript: In Evolution and the Levels of Selection, Samir Okasha argues that Lewontin and Sober's example of heterozygote advantage has actually been misinterpreted by both sides in this realist/conventionalist debate. Okasha argues that Lewontin and Sober arrive at the right (realist) conclusion, but for the wrong reasons. This is a nice example of what I think is Okasha's greatest strength--he's very precise and analytical, and he's good at drawing distinctions that need to be drawn. Okasha's book convinced me that much of the debate and confusion in the group selection literature has arisen because of people failing to draw key distinctions.

No comments: