Altruism has always been a problem for evolutionists. How does one explain a creature giving up something for another, sometimes its very life? Why, for example, will a monkey give a warning cry that alerts other members of the troop, but that gives away its own position? How could genes governing such behavior persist in the relentless competition for a place in the genome?
The kinds of reasoning used to explain behavior that is good for the group but perhaps not so good for the individual performing it is as old as Darwin. Until George Williams demolished whole classes of argument in his lovely 1966 book, "Adaptation and Natural Selection", it was common to invoke "group selection" as an analog to individual selection, and explain, in a vague, hand-waving sort of way, how altruistic behavior could arise by enhancing the survival of the herd, or school, or flock. And after Dawkins, both the individual and the group were banished from consideration, and the selfish gene reigned supreme.
Only one category of altruism has been taken as consonant with the unit of replication being the gene, namely "kin selection". This is the favoring of relatives: since relatives share genes, helping a gene-mate helps one's own genes, whether or not it benefits one's self. Yet much altruism in nature goes unexplained by kin selection. Think of the soldier who falls on the hand grenade so his (unrelated) buddies can live. There are many more examples from the lives of many creatures, most of whom never saw a war movie. How does one explain the clear patterns of altruistic behavior in animals at all levels of consciousness and cuddliness? Wilson, a biologist, and Sober, a philosopher, dare to think the unthinkable, or at least the unfashionable: is it possible that individuals or groups really do play a replicator role in evolution? They believe that group selection deserves another chance, but this time more rigorously specified.
I was very impressed with the first half of the book, in which they justify a group-selection model for adaptive evolution that can explain a persistent strain of altruism. What they show is that selection can take place at the level of a group of individuals in many more sorts of situations than were thought possible. (A nice bonus of this approach is that kin selection can be explained more simply using this more general context of the group.) Groups, however ephemeral, do have a role to play in selection.
The second half of the book is less convincing, as it involves psychological and philosophical arguments for "psychological altruism" in humans (that is, you not only behave unselfishly, but "want" to behave unselfishly), which, by its very nature, is hard (or very hard) to tease out in experiments, or to introspect to. However, the authors are reasonably convincing that nature would most likely not employ some Rube Goldberg-type of mental devices that depended on hedonism (pleasure-and-pain-driven behavior) to accomplish important tasks, such as child-rearing, but rather build in directly the mechanism to make a parent care to care for its child. In that way, the care of its child would be a primary motivation, rather than an intrumental one (sorry about the jargon!) on the way to getting pleasure or avoiding pain. Parents will find this convincing, as the desire to take care of one's children seems not to depend on how much we "enjoy" doing it.
This book is detailed, conscientious and well-written, but it covers a lot of ground and many of its arguments, especially in the second part, are subtle. So I recommend reading it more than once: this is contentious material. While the authors do not make anything of the political and social implications of their work, these are always waiting in the wings. Altruism, after all, is in direct opposition to selfishness. Many people see in this a political point, and a social point. Those issues are not properly a part of such a work, but do give great interest to its arguments and conclusions. And whether or not its conclusions finally survive intact, this book's arguments and approach seem exemplary and fruitful.