Reviewed by
Noah Mark
University of North Carolina at Charlotte
War is the answer. Or, at least, it is the answer to several questions social scientists have been asking for hundreds of years. (Why are societies large? Why are there property rights? Why are there highly productive economies? Why are there taxes?) Bowles and Gintis are not the first to claim that competition, conflict, and war between human groups is the foundation of cooperation and of society. However, their integration of this insight into evolutionary game theory stands to increase the accessibility of this powerful idea to a large number of scholars working in a dominant theoretical perspective that spans the social and biological sciences. This is one reason why I recommend their new book A Cooperative Species: Human Reciprocity and Its Evolution.Bowles and Gintis develop two arguments about human cooperation and how it evolved. First, humans are innately cooperative. We are guided by moral sentiments that evolved long ago. These sentiments compel us to act cooperatively and even altruistically in many situations. We also have a strong sense of fairness and engage in altruistic punishment when we observe individuals violate norms of reciprocity. Cooperation in human societies today and for tens of thousands of years has rested in large part of these moral sentiments. Second, competition between human societies, and in particular conflict and war between human societies, is key to the process by which these moral sentiments evolved and to the process by which attributes of groups, typically called "institutions," that promote cooperation evolved.
Both of these arguments are important, and the wealth of empirical studies Bowles and Gintis review to substantiate their argument makes the book an invaluable resource. Nevertheless, the second of these is the more novel, and how they develop this argument will be controversial, so this is the one I will consider here.
From an evolutionary perspective, the easiest explanation for war is high population-resource pressure. When people have enough land to farm or to hunt and gather on, war is not the answer. But when your group doesn't have enough land to live on, your group can benefit by going to war against some other people and taking their land. Bowles and Gintis incorporate this idea into their model with severe zero-sum competition between groups (societies). When a group goes to war against another group (which each group does about once in every four generations), the group that has more altruistic members (i.e., members who contribute to the war effort) wins. The losing group and all of the members of the losing group fall out of the population. The winning group doubles in size, holding its proportion altruists constant, and then divides into two groups, thereby replacing the defeated group.
This is how war promotes altruism. Groups are paired. The group with the smaller number of altruists (and the smaller proportion of altruists - all groups are the same size) disappears, and the group with more altruists doubles in size and divides into two groups. This group-level process works against the expected individual-level dynamic operating within each group which reduces the proportion of altruists within that group across successive generations. Bowles and Gintis identify conditions under which the group-level selection for altruism can offset the selection against altruism at the individual level.
I anticipate two different reactions to this explanation. Some will say the argument is fatally flawed; the altruism explained is no altruism at all. This reaction follows from recognizing that if contributing to a war effort increases the probability that one's group wins the war, thereby increasing the probability that one survives and reproduces, then contributing to a war effort increases one's own reproductive fitness and is not an instance of altruism.
However, a different reaction can follow from this same recognition that in the Bowles and Gintis model, behaving "altruistically" increases one's own reproductive fitness. The conditions a society faces in this model are severe. On average, a society survives eight generations before it and all its members are destroyed by war. What the analysis brilliantly shows is how these severe conditions can give rise to large-scale cooperation. Under some (less severe) conditions, it is possible for an individual to free-ride, and the temptation to free-ride limits cooperation. However, as populations grew (and as climates changed), some human societies faced conditions so severe that everyone needed to cooperate to survive. When a society went to war, everyone needed to contribute to the war effort. Groups containing shirkers perished. Certainly, contributing to the war effort in this situation is not altruistic, but this might be the key to large-scale human cooperation. Increasing population-resource pressure changed the game. Where members of a society used to face a prisoner's dilemma, they now faced an assurance game. When the opportunity to free-ride disappeared, cooperation exploded.
Return to Contents
of this issue