“The two persons iterated Prisoner's Dilemma,” says Robert M. Axelrod, “is the E. coli of the social sciences.” An analogue of the Prisoner's Dilemma is found in a variety of fields, ranging from evolutionary biology and arms control policy to networked computer systems and political philosophy. Consequently, when unrelated fields have analogous structures, the results in one are deemed applicable to the other. It might appear, then, that the problem of group rationality, namely, “How should scientists cooperate with one another to form or organize an effective group?” is so remarkably similar, say, to the problem of social justice, “How should individuals cooperate with one another to form or organize a just group?” or to the problem in evolutionary theory, “How do species reach an equilibrium state with neighboring species?” that one might unsuspectingly conclude: A little tinkering with an adequate theory of social justice or evolutionary theory would produce an adequate theory of group rationality. Nothing is further from the truth. Or so I shall argue.

Here is what I shall attempt in this chapter. The problem of group rationality, I want to show, is a unique problem. It is not a problem that can be solved in evolutionary terms (section I), nor is it a problem solvable by the resources of game theory. Since this latter is a significant task, I address it in two sections (sections II and III), each section dealing with a distinctive treatment of the analogy between game theory and group rationality.