The Evolutionary Rationality of Social Learning

Abstract

Abstract The field of evolutionary ecology has long been interested in the design and diversity of social learning heuristics, simple strategies that animals use to extract useful information from their social environment. This chapter reviews a slice of this literature, as well as explicitly analyze the evolution of social learning heuristics. The chapter outlines a family of social learning heuristics and analyze their evolutionary performance under two broadly different kinds of environmental variation. As each social learning heuristic also shapes a social environment as individuals use it, the chapter considers the population feedbacks of each heuristic as well. The analyses in this chapter are both ecological and game theoretic. This chapter's analyses are also explicitly evolutionary—heuristics succeed or fail depending upon long-term survival and reproduction in a population, not atomistic one-shot payoffs. As a result, some of the conclusions reflect an evolutionary rationality. For example, heuristics that randomize their behavior can succeed where those that are consistent fail. Overall, however, the approach the chapter reviews here supports the general conclusion that social learning heuristics are likely to be multiple and subtly adapted to different physical, statistical, and social environments.