Abstract

One way to theorize about we-identities—the identities that individual subjects have as ‘one of us’—is in terms of the uniformity, interchangeability, and prototypicality of group members. The social-psychological theory of self-categorization epitomizes this approach, which has strongly influenced contemporary phenomenological research on the we. This paper argues that this approach has one important and largely overlooked limitation: the we-identities tied to close personal relationships—exemplified by long-term friendships and romantic partnerships—are based on patterns of interpersonal interaction and integration through which individuals tend to grasp their non-substitutability and complementarity. This limitation suggests that another approach is needed to tackle the we-identities characteristic of close personal relationships. I outline such an approach, by combining resources from classical phenomenology and ongoing research on the socially extended mind.