This site is moving to a new domain: check out

“Some of these stories are closer to my own life than others are, but not one of them is as close as people seem to think.” Alice Murno, from the intro to Moons of Jupiter

"Talent hits a target no one else can hit; genius hits a target no one else can see." Arthur Schopenhauer

“Why does everything you know, and everything you’ve learned, confirm you in what you believed before? Whereas in my case, what I grew up with, and what I thought I believed, is chipped away a little and a little, a fragment then a piece and then a piece more. With every month that passes, the corners are knocked off the certainties of this world: and the next world too. Show me where it says, in the Bible, ‘Purgatory.’ Show me where it says ‘relics, monks, nuns.’ Show me where it says ‘Pope.’” –Thomas Cromwell imagines asking Thomas More—Wolf Hall by Hilary Mantel

My favorite posts to get started: The Self-Righteousness Instinct, Sabbath Says, Encounters, Inc., and What Makes "Wolf Hall" so Great?.

Wednesday, July 21, 2010

Group Selection: Feasible or Fashionable--or Both?

I seem to have happened upon a hot-button issue when I posted a review of Sober and Wilson’s 1998 book Unto Others. Many of the objections to the authors’ arguments for group selection that have been pointed out to me (hat tip caynazzo) can be summarized thus: it’s true that group selection cannot be ruled out either logically or empirically, and it must even take place under certain circumscribed conditions; however, the phenomena the theory might be applied to have already been, or probably will be better illuminated by the rival and more robustly supported theory of inclusive fitness, whereby individuals favor others who are likely to carry the most genes in common with them.

Setting aside the possibility that inclusive fitness theory has been more fruitful simply because group selection was ruled out as a matter of orthodoxy for many years, Sober and Wilson explain in their book that these aren’t really rival theories:

“Social interactions among genetic relatives correspond to the nonrandom formation of groups. The significance of relatedness for the evolution of altruism is that it increases genetic variation among groups, thereby increasing the importance of group selection” (77).

They go on to quote W.D. Hamilton: “it obviously makes no difference if altruists settle with altruists because they are related… or because they recognize fellow altruists as such, or settle together because of some pleiotropic effect of the gene on habitat preference” (77). (For the uninitiated, pleiotropy is when a trait that isn't directly selected for comes as a package deal with a trait that is, usually because both are products of the same genes.)

And they conclude that Hamilton “left no doubt that the difference between inclusive fitness theory and group selection theory is a matter of perspective, not process” (77).

The idea is that groups must vary genetically if they are to serve as a unit of selection. The corollary is that there is some genetic similarity within each of the groups that contrasts with the differences between groups. What accounts for the similarity within groups isn’t important to the process of differential reproduction between groups. The group to which I belong can be made up of family members or people who share a set of important traits with me but who aren’t related. Either way the stage is set for group selection.

Still, how powerful and how important selection at the group level is has yet to be determined. And I’m just dipping into the issue here. (Sober and Wilson’s book is 12 years old, and I spent more time on Saul Bellow’s Collected Stories this summer than on evolutionary biology.) I will argue though that there are few reliable markers of kinship that even a species as cognitively sophisticated as humans can use to apportion altruistic favors. True, we have language to describe and label all sorts of relationships, but those labels are frequently applied to nonkin. At least in humans, group (tribal) membership seems a much more salient signal for deciding who deserves unselfish treatment.

But here is what Dr. Larry Kuznar, an anthropologist at my alma mater specializing in evolutionary models, said in response to my inquiry about how Sober and Wilson's ideas are faring:

"Wilson and Sober are at the vanguard of reintroducing group selection to biology. In anthropology, Richard Boyd and Peter Richerson are advancing group selection under the guise of "cultural group selection" (see their book, Not be Genes Alone). How is this holding up in the field? Good question. Wilson and Sober are still a minority view in biology, and I think for good reason.

Boyd and Richerson have had more success in anthropology, but mostly through clever argumentation and by attacking individual selectionist work. I have not seen good robust tests of their ideas. There is a lot of experimental work, even done in different cultures (see Joseph Henrich), but I think that the experiments are rather contrived. Nonetheless, cultural group selection theories are very fashionable among some anthropologists (certainly not all, probably not the majority), behavioral economists, some political scientists, and some evolutionary psychologists. One hears a lot about altruism these days. It has even emerged as one of the competing explanations for suicide bombing."


caynazzo said...

There seems to be a gap between anthropology and evolutionary biology in regards to group selection.

For example, Nicholas Barton's latest textbook titled "Evolution," (graduate level, copywrite 2007) dedicated two pages to group selection:

And excerpt: "The fundamental difficulty with this simple form of group selection is that relationships between neighboring individuals are likely to be weak [...] selection on local groups is typically 2.5-5 times weaker than selection between individuals. A further difficulty is that the number of selective events is much smaller: The number of births and deaths of local demes must be far smaller than the number of individual births and deaths. Thus, random fluctuations (analogous to random genetic drift which scatters demes) strongly interfere with group selection."

There's an updated form of group selection proposed by Sewall Wright called "shifting balance theory" that has demes shift between random drift, selection within demes, and selection between demes, thereby pushing a single large population toward the nearest adaptive peak.

The text concludes: "However, the SBT suffers from many of the same difficulties as simpler forms of group selection. Selection between demes at different adaptive peaks is inherently slow and subject to random events. Moreover, there is little evidence that local demes are in fact at different peaks [...] Group selection cannot disentangle the many differences between local demes, whereas recombination within a sexual population can do so."

That last assertion seems overstated. But that's a textbook for you.

Turning to PubMed.

caynazzo said...

"And they conclude that Hamilton 'left no doubt that the difference between inclusive fitness theory and group selection theory is a matter of perspective, not process'"

I can see that. Because what's the difference in saying altruism acts at the group level (costly to individual, beneficial to group) or saying traits acting at the individual level can encourage group behavior (e.g., altruism).

Dennis said...

I would imagine that once certain mechanisms are in place group selection can take off and become a more important force in evolution. I'm only familiar with the ideas of altruistic punishment and social monitoring from Flesch's "Comeuppance" and PBS "The Human Spark," which demonstrates the work of Karen Wynn, but at the very least there seems to be an innate liking in our species for cooperators.