(This is a major rewrite of an earlier review, which I decided on reflection was not as clear as it might have been.)
This is a book with a complex context, and it is best to understand something of that context in order to get a clear view of the book. Briefly, Bowles and Gintis have set themselves to resolve one of the most vexing issues in evolutionary theory, that of whether the widespread human trait of altruism toward those who are not close kin can have arisen through natural selection, and if so just how. To do so they must wage war on some views that approach dogma, and they gird and armor themselves with mathematics and factual detail. All this does not make for easy reading, but it is very worth the effort. And it is not necessary to trace all of the details to get a great deal out of it.
In the popular view, the theory of natural selection implies that nice guys always finish last, that it is the strong and ruthless who are fittest, not the cooperative and altruistic. The hyperaggressive Wall St. sociopath is seen as evolution's ideal type. It would seem to follow that altruism cannot be the product of evolution, and thus that natural selection cannot entirely account for the nature of humankind.
Darwin understood all this quite clearly and it troubled him not a little. In a famous passage in The Descent of Man he acknowledged, "It is extremely doubtful whether the offspring of the more sympathetic and benevolent parents, or of those who were the most faithful to their comrades, would be reared in greater numbers than the children of selfish and treacherous parents belonging to the same tribe. He who was ready to sacrifice his life, as many a savage has been, rather than betray his comrades, would often leave no offspring to inherit his noble nature. The bravest men, who were always willing to come to the front in war, and who freely risked their lives for others, would on an average perish in larger numbers than other men."
Darwin argued, however, that the contribution made by the "sympathetic and benevolent" to the survival and success of the group would outweigh the individual advantages of the "selfish and treacherous" : "Let it be borne in mind how all-important in the never-ceasing wars of savages, fidelity and courage must be.... A tribe rich in the above qualities would spread and be victorious over other tribes: but in the course of time it would, judging from all past history, be in its turn overcome by some other tribe still more highly endowed. Thus the social and moral qualities would tend slowly to advance and be diffused throughout the world." Thus was born the concept of group selection.
But nearly four decades ago, group selection died a messy and protracted death, a victim of mathematical analysis of natural selection's mechanisms, the then-new understanding of the molecular basis for transmission of the traits on which natural selection acts, and deeper understanding of the heredity of social insects. I've heard more than one biologist or mathematical biologist say flatly that "group selection is all rubbish." (For a summary and scorecard see Mark E. Borrello, "The Rise, Fall and Resurrection of Group Selection," Endeavour 29, No. 1 (Mar 2005): 43-47.)
In reality, however, it never was that absolute. As the great mathematical biologist John Maynard Smith put it, "The terms group selection should be confined to cases in which the group ... is the unit of selection. This requires that groups be able to 'reproduce,' ... and that groups should go extinct. ... Group selection can maintain 'altruistic' alleles--i.e., alleles which reduce individual fitness but increase the fitness of groups carrying them. The conditions under which this can happen are stringent, so that the main debate concerns whether the process has had evolutionarily important consequences." ["Group Selection," Quarterly Review of Biology 51, No. 2 (Jun 1976): 277-83.]
Bowles and Gintis now return to this debate armed both with new models and new knowledge of the biology and behaviors of our ancestors. The increased puissance of the models derives both from several decades more thought by mathematical biologists armed with the insights provided by extensive computationally-intensive simulation of a kind not feasible in the 1970s. The knowledge of human descent has been augmented by extensive archeological discoveries, elucidated by powerful technologies for exploiting them, together with the entirely new study of human and animal genomes. The book provides a very extensive tour of all of this.
For all our gains in knowledge, there remain huge gaps in our picture of our ancestors and their lives. We still must rely a great deal on inferences that seem plausible in terms of the available evidence but could very well be wrong. It is not possible to say with certainty whether Maynard Smith's stringent conditions were in fact met in the course of human prehistory. Nevertheless, Bowles and Gintis make out a very colorable case that they were met, and that group selection thus endowed our species with its remarkable altruistic and cooperative tendencies. (They prefer to call it multi-level selection; while this seems more precise and descriptive I am not optimistic that it will become standard.)
As an aside, I should remark that this is a field whose terms, such as "altruism" and "strong reciprocity," have an unfortunate tendency to launch some people into hyperbolic rhetorical orbits, as we see in some of the reviews here. But this is really a book about behaviors and mechanisms, leaving us free to take our own views on values.
Bowles and Gintis, together and separately, have published many papers on the subjects treated in the book but so far as I can see the book very largely subsumes all their published work.
While I rather imagine that Bowles and Gintis have more than once felt quite lonely in their efforts, the question of group selection and its influence on the development of altruism has become quite a hot topic, with those taking the positive view having the wind at their back on the whole, at least for now. The evidence for this includes several other books that bear mention. Edward O. Wilson, who was one of those who argued most effectively against group selection from the biological perspective four decades ago, now has published The Social Conquest of Earth, which many of his sometime admirers see as shocking apostasy. Wilson goes briefly over the same ground as Bowles and Gintis but concentrates much of his attention on the case of social insects, his area of deepest expertise.
Wilson's book was preceded by a widely noted and very controversial paper which he co-authored with a prominent younger mathematical biologist, Martin Nowak, who now (with a co-author) has published SuperCooperators: Altruism, Evolution, and Why We Need Each Other to Succeed. This is a non-mathematical exploration of the insights from the mathematical modeling, with references to correlated biology.
Finally, I should mention Christopher Boehm's Moral Origins: The Evolution of Virtue, Altruism, and Shame. Boehm is a social anthropologist, not a biologist at all, who takes up another argument offered by Darwin, that peer pressure and reputation played a decisive role in the evolution of altruism. Boehm does not offer any of the formal game-theoretic models that Bowles and Gintis use (and that underlie Nowak's book), and Bowles and Gintis do seek to use models to deny reputation a place in altruism's evolution. I do not see them as having entirely undermined Boehm's points and I suggest we will see more on the subject.
No doubt we will see much more on the whole issue of altruism's evolution. Surely we have yet to hear the last of the anti-group selection camp, and there is ample room in any event for further discoveries and resulting arguments. But this book seems bound to have continuing importance. It certainly is true that the book is anything but light reading. It's a deep, dense book, but it well repays the effort involved.