Evolution of Preferences in One-shot Prisoner's Dilemma

Can cooperation evolve in one-shot prisoner's dilemma? Utilizing an indirect evolutionary framework that separates objective payoffs and utilities, I show that preferences that are consistent with conditional cooperation in one-shot prisoner's dilemma can evolve along with strictly self-interested preferences. The ability to detect others' preference types is critical for the evolution of reciprocal preferences. In the evolutionary equilibria, both the strictly self-interested preferences and reciprocal preferences coexist and both cooperation and defection are observed. These results complement the folk theorem and the standard evolutionary models by providing an alternative evolutionary logic of cooperation.