Tuesday, December 14, 2010

Group Selection, Kin Selection, and Moral Tragedy

From Aristotle to Darwin to E. O. Wilson, how one understands human morality and politics seems to depend on how one understands the social insects. The importance of this line of thought in explaining social evolution has recently been highlighted by Wilson's attack on kin selection theory and his defense of Darwin's group selection theory in explaining social evolution from insects to humans.

To explain the biological nature of human beings as political animals, Aristotle compared human beings to other political animals--and particularly, ants, bees, and wasps. The great size and complexity of social insect colonies are comparable to that of human communities.

Darwin saw the same resemblance between human communities and social insect colonies--particularly, in their intricate division of labor. But for Darwin, the existence of sterile female castes of workers among the social insects was "by far the most serious special difficulty which my theory has encountered," as he wrote in the chapter on "Instinct" in The Origin of Species. The obvious problem is that this seems to contradict evolution by natural selection, because the sterile females are sacrificing their reproductive fitness for the reproductive advantage of the queen. Darwin's solution was to propose that "selection may be applied to the family, as well as to the individual, and may thus gain the desired end." He thus implied that natural selection could favor the altruistic sacrifice of individuals for the reproductive good of their group. Moreover, in The Descent of Man, he clearly claimed that human moral and political evolution depended on group selection in war, so that some of the highest moral virtues arose as dispositions that conferred advantages on human groups in violent conflict with other groups. But he left it unclear as to how exactly natural selection could do this.

In the 1960s, William D. Hamilton set out to solve Darwin's problem by developing the ideas of "inclusive fitness" and "kin selection." The basic idea is that individuals can evolve to show altruistic behavior--behavior that is costly to the animal but beneficial to others--if the cost to the actor brings sufficient benefit to sufficiently closely related recipients. Evolutionary selection should favor my acting not only for the sake of benefiting my direct offspring, who share some of my genes, but also for benefiting the offspring of my close relatives, who share some of my genes and therefore might carry my genes for altruistic behavior. In the case of sterile female worker castes, Hamilton argued, this extraordinary altruism could be explained by the fact that the workers are sisters who are more closely related genetically to one another that they are to their offspring, and consequently, it better enhances their inclusive fitness to serve the reproductive activity of the queen in producing more sisters than to have offspring of their own. This happens because among the Hymenoptera, the order of ants, bees, and wasps, the sex-determining mechanism is haplodiploidy, in which fertilized eggs become females, and unfertilized eggs males.

There are serious problems with this theory. One is that some of the eusocial species--such as termites and naked mole rats--don't use haplodiploid sex determination, and therefore Hamilton's theory can't explain the sterility of worker castes among them as it does for Hymenoptera.

But despite such problems, Hamilton's theory of kin selection has been widely accepted among many biological theorists of social evolution over the past 50 years, and especially since the publication of Wilson's Sociobiology in 1975, which adopted Hamilton's theory.

In recent years, however, E. O. Wilson and David Sloan Wilson have been criticizing kin selection theory and defending group selection theory as superior. Most recently, E. O. Wilson co-authored an article with Martin Nowak and Corina Tarnita in Nature (vol. 466, August 26, 2010, pp. 1057-1062) that summarizes some of the reasoning against Hamilton's theory. This article has provoked an intense controversy in the whole community of biologists and psychologists studying the biological evolution of sociality.

Much of the publicity surrounding this controversy turns on the perception that E. O. Wilson has undergone a radical change of mind--from promoting to rejecting Hamilton's theory. But anyone who reads Wilson's Sociobiology carefully (for example, pp. 30, 129) will see that he was never convinced that Hamilton's theory was superior to group selection theory. Over the years, Wilson has given encouragement to proponents of group selection such as David Sloan Wilson; and in recent years, the two Wilsons have co-authored some articles.

It's hard for me to see that one has to choose between kin selection and group selection in explaining social evolution. I suspect that this is a false dichotomy that obscures their complementarity. In fact, one can even argue that kin selection is a form of group selection where the group is constituted by kinship. Here I agree with Samir Okasha, who makes the argument for complementarity in a recent article in Nature (vol. 467, October 7, 2010, pp. 653-55.)

Regardless of whether one favors kin selection, group selection, or some combination of both, there is a deeper moral and political issue here. All of these kinds of Darwinian explanations of social evolution point to the inevitability of tragic conflicts in moral and political life that cannot be resolved by ideal principles of universal love and cooperation.

There is a tendency among modern moral and political philosophers to assume that moral and political life can and should be governed ultimately by some ideal conception of disinterested humanitarianism. One can see this, for example, in John Rawls' appeal to the ideal situation of people agreeing to universal, rational principles of justice in the "original position," where human beings would act as if they were disembodied spirits. One can also see this among religious believers who assume that the teaching of universal love in Jesus' Sermon on Mount is the moral ideal for all human beings, or among secular philosophers like Peter Singer who assume that morality and politics should be guided by an impartial concern for the interests of all sentient creatures.

If our moral and political dispositions have not been created by some cosmic order of the good and the just to conform to some eternal values of love and cooperation, if these dispositions have been created, on the contrary, by natural evolutionary selection, then we can expect that our moral and political lives will be torn by tragic conflicts of interest that cannot be resolved by universal principles of ethics.

E. O. Wilson refers to this as the problem of "moral ambivalency," and it runs through his writing in his book Sociobiology, beginning with the epigram from the Bhagavad Gita. The passage Wilson quotes shows Arjuna doubting the justice of leading his family in a war against another family competing for political dominance, but Lord Krishna (the Lord of the Universe) teaches him that this is his sacred duty. Wilson suggests that an evolutionary theory of social evolution would explain this tragic conflict as arising from a natural history of counteracting pressures on the units of natural selection. Wilson writes:

What is good for the individual can be destructive to the family; what preserves the family can be harsh on both the individual and the tribe to which the family belongs; what promotes the tribe can weaken the family and destroy the individual; and so on upward through the permutations of levels of organizations. Counteracting selection on these different units will result in certain genes being multiplied and fixed, others lost, and combinations of still others held in static proportions. According to the present theory, some of the genes will produce emotional states that reflect the balance of counteracting selection forces at the different levels. (p. 4; cf. pp. 129, 563)

Here, then, is the scientific basis for the tragic realism of evolutionary ethics. We see the evolved moral and political nature of human beings as shaped by countervailing levels of selection, which have created tragic conflicts in our moral emotions that cannot be resolved by rational appeals to universal moral principles. Such moral realism is repugnant to moral utopians who assume that moral and political order must be ultimately guided by universal rules of love and cooperation that can in principle resolve all conflicts.

Oddly enough, while E. O. Wilson accepts this moral realism as a conclusion from his evolutionary account of morality as shaped by group selection, David Sloan Wilson rejects this conclusion while trying to hold on to some utopian vision of universal cooperation.

Some related posts can be found here, here, here, here, and here.


Troy Camplin said...

It is the paradoxical conflicts which drive complexity, so moral tragedy is inevitable. This is why we have tragic art.

I have always disliked the comparison (and implied goal) to social insects. Humans are not eusocial -- it is not part of our evolutionary heritage, and we did not evolve into a eusocial species. Thus, such comparisons can and will only lead us into error.

At the same time, group selection does have much to recommend it. For example, it explains why humans might fight for unifying symbols.

Larry Arnhart said...

E. O. Wilson agrees with you that we are not eusocial. He stresses the point that by comparison with the eusocial insects, human beings shows a distinctly mammalian conflict between the individual and the group.

This conflict is expressed in human history as a tension between individual liberty and social authority that runs through human morality and politics.

Troy Camplin said...

Yes, E.O. Wilson is (ironically, considering he's an entomologist) rather good about making this separation. Many others, however, have not. Especially those with more socialist leanings. I have noticed, however, that mole rats are becoming more popular examples on the left:


Indeed, the tragic conflict is that between our individualism and our sociality -- particularly when we move from individualist to communitarian psychologies (Greek tragedy), and vice versa (Renaissance tragedy).

Anonymous said...

Could group selection be just another form of kin selection, rather than the other way around. Groups are interbreeding organisms, aren't they? The relationships are are bit more distant than first cousins, but don't they (groups) share common genetic bonds?

I don't see how any liberal can take comfort in the eusocial intragroup conduct of mole rat unless of course you can seamlessly exchange members of differing groups without any conflict, but I don't think that would be the case.

In either situation, whether kin or group selection, there seems to be limits on any organisms ability to cooperate and extend altruistic behavior. Am I wrong?

Troy Camplin said...

You are right. The socialists, who do in fact think you can do what you suggest, are wrong.

However, what if "group selection" includes those of other species -- meaning, coevolution? That would hardly be an example of kin selection. And in humans, at least, there was always at the tribal level a consideration of anyone not in one's tribe as being non-human. So it's not necessarily species-specific, as you suggest (extending "kin" out to its outter boundaries). Chimpanzees also don't engage in group selection in regards to other groups. Yet, female chimpanzees will migrate from one group to another, and will then engage in group selection with that group of unrelated individuals. No kin selection there, either.

Anonymous said...


Thank you for your response, as I am trying hard to understand these issues. I appreciate any input you provide.

Regarding your reference to coevolution, could you give me a specific example? Are we talking about truely altruistic behavior extended by one species to an entirely different species apart from symbiotic behavior e.g., animal husbandry, ant and aphid?

And are not chimpanzees highly group selective as evidence by their frequent raiding by males and violence directed at other groups?
While I know that nubile chimpanzee females frequently stray into other groups and become part of the breeding population (which makes sense as far as preventing too much inbreeding within the older group)isn't that the exception? This straying is as though the tribalism emotion (group selection gene(s)) is suddenly turned off in these females, but isn't that the exception. Isn't this behavior an override of the natural strong kin selection (i.e. and expression of the selfish gene)?

What I am seeking is evidence for true altruism and whether altruism can be extended to other groups and, more importantly, survive. From my readings on the subject (I am definitely a student) , I believe that currently the answer is no, but I am always seeking new perspectives.

Again Troy, thanks a million for your help.

Troy Camplin said...

No, female chimpanzees "in general, emigrate at adolescence, between nine and 14 years old"


While males stay with the group. We actually still see this in humans (the son who won't move out of the house; the woman who moves in with her boyfriend who won't move out of his parent's house :-) )

Perhaps coevolution is too strong a term. I am thinking more of those species that have such close interactions that they end up benefiting each other altruistically as well. Oxpeckers which fly off when they detect a lion would warn the animal they are on, for example.

But there is more direct evidence of interspecies altruism:




These provide some evidence for this kind of altruism. Certainly no kin selection in these examples.