The evolutionary turn in game theory

In a conference bar a few years ago, I was talking with an economist who specializes in applications of game theory. I mentioned that I was interested in the conceptual problems associated with the assumption that players have common knowledge of their rationality. H is reaction was surprise that anyone still thought seriously about that issue at all. With something like the embarrassment with which people of my age look at photographs of themselves dressed in the fashions of the early 1970s, he confessed that there had been a time, not so very long ago, when he had been engrossed in what has come to be called the reŽ nement programme, in which questions about the meaning of ‘common knowledge’ and ‘rationality’ are central. But, of course, not now. Now, did not everyone recognize that this programme had been a silly waste of time, and that all the most interesting work in game theory was based on an evolutionary approach? I was not surprised by this response: it is a commonplace. But at a deeper level, the fact that so many game theorists now say things like this is surprising. In the last ten or so years, game theory has executed a sharp change of direction, away from models based on assumptions of ideal rationality and towards models that are inspired by evolutionary biology. The assumption of individual rationality has usually been interpreted as the most fundamental assumption of economics; to give it up is surely a huge step to take. That biology should be taken as the natural science on which to model economics is itself a major break with tradition. From the neoclassical revolution onwards, theoretical economics have taken physics, with its aim of explaining everything in terms of a few simple mathematical principles, as their inspiration. Biology is a much messier discipline; it deals with evolutionary processes that are path-dependent, subject to historical contingencies, and in many respects inherently unpredictable. Taken at face value, then, the evolutionary turn in game theory is a momentous theoretical revolution, a sledge-hammering of hard cores, a Journal of Economic M ethodology 8:1, 113–13