Progress in the study of relationships has depended in part on the recognition that relationships have properties not relevant to interactions or to the behavior of individuals, and may require additional principles of explanation. This has led to an emphasis on relationships as linking individuals. In this article we argue that relationship processes occur in the heads of individuals, with the participants having their own idiosyncratic views of the relationship as well as a shared one. The relationship is both affected by and affects the self-concepts of the participants, so that the influences of the self-concept may be critical for understanding the properties and dynamics of relationships Furthermore, consideration of the self-concept can assist in the integration of different but not necessarily incompatible explanations for the same relationship phenomena.In the course of the twentieth century, questions concerning the nature of the "self" have been in and out of fashion. Toward the end of the nineteenth century, when introspection was respectable, the "self" came very much into vogue (James, 1890). With the rise of behaviorism, however, the study of subjective phenomena fell into disrepute among experimental psychologists, and it became the province of clinicians. In recent decades, partly as a result of the cognitive revolution, studies of the "self" are returning to favor (e.g., Baumeister, 1999;Modell, 1993;Scheibe, 1998; Wylie, 1974/79). In studies of relationships, also, the role of the "self" is beginning to receive more recognition but, we suggest, not all that it deserves.One reason that the "self" has not received more overt recognition in studies of relationships lies in the history of research in this area. Although there were of course precursors (Fincham, 1995), the specialized study of personal relationships is only two orWe are very grateful to the referees for their comments on an earlier draft of this article.Correspondence should be addressed to Robert A.Hinde, St. John's College, Cambridge CB2 lTR, UK. three decades old. Before that, psychology was concerned primarily with individuals or with groups of individuals. Such studies of relationships that did exist mostly involved a clinical approach, often top-heavy with theory, which focused on the particulars of individual cases and thus did not lend themselves to the production of generalizations whose validity could be assessed. It became apparent that relationships, of prime importance in the lives of nearly everyone, required study in their own right. Three principles were basic in the subdiscipline that emerged. First, dyadic relationships involve two individuals, and they concern what goes on between them. Second, they depend on cognitive, affective, and behavioral processes, and they involve a series of interactions over time between two individuals such that each interaction is affected by previous ones and perhaps by expectations of further interactions in the future. Third, relationships have properties in addition to...