Learning by individuals from the behavior of others and imitation pervade the social life. Issues related to such learning have been debated since the beginning of the social sciences, more than a century ago. However, in the last ten years or so they have stimulated a revival and very active research in economics, with extensions to other “human sciences” (sociology, psychology, political science). The purpose of this book is to give an account of these studies. Perhaps it will induce others to enter the field and provide them with some training.

The setting is one of rational agents with limited information who share that information with others through their actions. The properties of the learning process are analyzed froma theoretical point of view, but some empirical studies are discussed in relation to the theoretical results.

Special attention is devoted to the pathologies of social learning by rational agents. Herds appear to be obvious examples of failures of social learning. Indeed, herds, fads, bubbles, crashes, and booms are cited as proofs of the irrationality of individuals. However, most of these colorful events will appear in the models of rational agents studied in this book.

The assumption of rationality may seem a bit narrow. Indeed, at this stage of the evolution of research, the concept of rationality itself is beginning to be seriously investigated. In this book, the usefulness of the assumption goes beyond the standard “benchmark” justification: a recurrent issue will be that despite the rationality of individual behavior, and often because of that rationality, the process of social learning may be inefficient or fail completely.