A game-theoretic strategy brought to the attention of philosophers by Robert Axelrod, in his The Evolution of Cooperation (1984). The question is how to perform to maximize gain in an interaction that consists of an extended sequence of prisoners’ dilemmas. The program that won the first competition amongst strategies for solving this problem was one that (a) always co-operated unless (b) on the immediately preceding round, its opponent had defected, in which case it defected. The strategy was both nice, in that it never initiated a cycle of defecting, and also forgiving, in that it punished defection only once. The success of the strategy seemed to give some justification for partially retributive social rules, and might explain the common moral tendency to favour them. However, further work has shown that tit for tat is only the optimum strategy in certain specific circumstances.