Here’s one last (I hope) post on the brouhaha about the evolution of social behavior that I’ve covered over the last year or so.
I think E. O. Wilson must be feeling a bit beleaguered about the criticism he’s endured for his relentless advocacy of group selection. Not only was he an author of the Nowak et al. paper in Nature arguing that group selection rather than kin selection was the prime mover of social evolution in insects and humans—a paper that was excoriated by biologists who work on the evolution of behavior—but his new book, The Social Conquest of Earth, which makes the same group-selection argument, has also been strongly criticized.
Perhaps this explains why Wilson took to the pages of the Sunday New York Times to defend his views in an Opinionator piece called “Evolution and our inner conflict.” His thesis is this:
Within biology itself, the key to the mystery is the force that lifted pre-human social behavior to the human level. The leading candidate in my judgment is multilevel selection by which hereditary social behavior improves the competitive ability not of just individuals within groups but among groups as a whole.
“Multilevel selection” is another word for “group selection.”
How did social behavior evolve? Wilson gives a nod to kin selection, the idea that genes can evolve that produce behaviors favoring kin over oneself, so long as the genetic benefits of that behavior among all individuals outweigh the genetic costs to the altruist. (This is weighted by the degree of kinship between actors: behaviors favoring your brothers and sisters over nonrelatives will evolve more readily than those favoring cousins over nonrelatives.) But in the end Wilson dismisses kin selection in favor of selection among human groups:
This seems plausible, but in 2010 two mathematical biologists, Martin Nowak and Corina Tarnita, and I demonstrated that the mathematical foundations of the kin selection theory are unsound, and that examples from nature thought to support kin selection theory are better explained as products of multilevel selection.
Sadly, Nowak et al. demonstrated no such things. In fact, while they did offer what seems to be an alternative theory based on selection among groups, there is not a scintilla of evidence that it explains the evolution of social behavior in either humans or insects better than a kin-selection approach. Their model is in fact unable to rule out kin selection, since they begin with groups that are highly related and do not vary the level of kinship in their models to determine if, as they predict, kinship is not a “driver” of social evolution. So they simply cannot rule out kin selection as a powerful cause of social evolution.
Wilson goes on:
A strong reaction from supporters of kin selection not surprisingly ensued, and soon afterward more than 130 of them famously signed on to protest our replacement of kin selection by multilevel selection, and most emphatically the key role given to group selection. But at no time have our mathematical and empirical arguments been refuted or even seriously challenged. Since that protest, the number of supporters of the multilevel selection approach has grown, to the extent that a similarly long list of signatories could be obtained. But such exercises are futile: science is not advanced by polling. If it were, we would still be releasing phlogiston to burn logs and navigating the sky with geocentric maps.
Well, their mathematical arguments may hold, but others have shown that a). their model is basically one that does involve kin selection, and adds nothing new to the kin-selection perspective, and b). their empirical arguments about the intellectual vacuity of kin selection are serioiusly flawed. Several authors have produced lists of advances in our understanding of nature that have derived from a kin-selection perspective. In contrast, the idea of group selection has added virtually nothing to that understanding.
The number of signatories on the letters critiquing the paper of Nowak et al. is simply an indication of how much opposition their ideas faced in the scientific community. I urge readers with some biology expertise not to count the numbers, but actually read the critiques. No, science isn’t advanced by polling, but it is advanced by the open airing of arguments. All the numbers show is that a lot of scientists—virtually every prominent worker on the evolution of social behavior—have pointed out flaws of the group-selection approach.
For a clearly written critique of Nowak, Tarnita, and Wilson’s group-selection argument for social behavior, and a lucid defense of the value of kin selection, I highly recommend a recent paper by Andrew Bourke (reference below).
But on to the evidence that we humans became altruistic and cooperative by group selection. Wilson makes a strong claim here:
In fact, it seems clear that so deeply ingrained are the evolutionary products of group selected behaviors, so completely a part of the human condition, that we are prone to regard them as fixtures of nature, like air and water.
What are these supposedly group-selected behaviors? Wilson gives three (I quote):
- “Among them is the intense, obsessive interest of people in other people, which begins in the first days of life as infants learn particular scents and sounds of the adults around them. Research psychologists have found that all normal humans are geniuses at reading the intentions of others, whereby they evaluate, gossip, proselytize, bond, cooperate and control. Each person, working his way back and forth through his social network, almost continuously reviews past experiences while imagining the consequences of future scenarios.”
Given that we’re social animals (see below), it is to an individual’s advantage to suss out one’s fellows in the group, and to your advantage to label potentially damaging or helpful individuals as such, for they can help or hurt not only you, but your relatives. There is no individual disadvantage that I can see to this behavior. Au contraire: any individual who can better read other people could have a reproductive advantage. So this can evolve via individual selection; there’s no need for group selection.
- “A second diagnostic hereditary peculiarity of human behavior is the overpowering instinctual urge to belong to groups in the first place. To be kept in solitude is to be kept in pain, and put on the road to madness. A person’s membership in his group — his tribe — is a large part of his identity.”
Yes, but why does this suggest group selection? It’s a common mistake, one pointed out by Steve Pinker in his Edge essay, to mistake adaptations for living in groups with the notion of group selection. Cooperation can evolve by several forms of individual selection, including associating with others for protection or the ability to bring down larger prey, or through reciprocal altruism. The advantages of grouping are not only multifarious, but widespread in animals. And if you evolve grouping behavior, then you will likely evolve the urge to want to stay in a group. None of that needs group selection to evolve. Given the relative difficulty of group selection compared to individual selection, it would be extraordinary if the huge variety of animals who group all evolved that trait, and the penchant to aggregate, via group selection.
And finally, some xenophobia,which is probably the flip side of the second behavior given above:
- “All things being equal (fortunately things are seldom equal, not exactly), people prefer to be with others who look like them, speak the same dialect, and hold the same beliefs An amplification of this evidently inborn predisposition leads with frightening ease to racism and religious bigotry.”
No problem for individual selection here: if you’ve evolved to be cooperative with others in groups, but simply haven’t encountered people who are different, it is to your advantage to be suspicious of them, for they haven’t passed the test of familiarity. Or, if groups compete with each other for food, individuals might also selected to be wary of those belonging to other groups. That involves group competition, but not group selection. A group-selection explanation would invoke xenophobia being maladaptive within groups, but a trait that has spread to all humans because xenophobic groups simply killed off the groups of “nice people.” That doesn’t sound so plausible to me.
In a book review by Michael Price of Brunel University, which I’ve discussed here, Price criticizes the view that human cooperation and other social traits must have evolved by group selection, pointing out the many features of human cooperation that suggest its evolution by selection among individuals. I won’t reprise the evidence here, but go to Price’s review (or my summary) and read the paragraph that begins as follows:
The view that group selection is needed to explain most human cooperation seems inconsistent with the fact that over the past several decades, most successful research on this cooperation has theorized that it is produced by individual-level adaptations (Price, 2011; Price and Johnson, 2011).
Price then enumerates the features of human “cooperation” that suggest it confers advantages to individuals. And if it does that, then there is no need to invoke selection among groups, which posits that “cooperative” traits are really disadvantageous for individuals within groups.
Nevertheless, Wilson blithely paints a group-selection scenario whereby the traits above, presumably disadvantageous in individuals, spread throughout the species:
Probably at this point, during the habiline period, a conflict ensued between individual-level selection, with individuals competing with other individuals in the same group, versus group-level selection, with competition among groups. The latter force promoted altruism and cooperation among all the group members. It led to group-wide morality and a sense of conscience and honor. The competitor between the two forces can be succinctly expressed as follows: within groups selfish individuals beat altruistic individuals, but groups of altruists beat groups of selfish individuals. Or, risking oversimplification, individual selection promoted sin, while group selection promoted virtue.
“Probably”? We have no evidence that altruism and cooperation evolved in this way, and suggestive evidence it evolved by individual selection—and not only in humans. Sadly, Wilson, though at least expressing some uncertainty with the word “probably,” is severely misleading people about the evolution of cooperation. So are others, like Jon Haidt and D. S. Wilson. The force with which the adherents of group selection push their theory contrasts sharply with the paucity of evidence that their favorite process actually operates in nature. It’s puzzling. Well, maybe not so puzzling given that scientific egos and Templeton funding are on the line.
Bourke, A. F. G. 2011. The validity and value of inclusive fitness theory. Proc. Roy. Soc. Biol. 278:3313-3320.