Understanding the effects of readers on writing development requires prior conceptualization of the relationship between writers and readers. Recently two major schools of thought have emerged concerning this relationship: social constructionism and social interactionism. Influenced by Saussure's structuralist concept of la langue as a set of language norms and Durkheim's concept of social fact, social constructionists emphasize normal, standard discourse: Their key principle is empirical consensus, their unit of analysis is the canon, and their level of social analysis is the community. By contrast, social interactionists, following Bakhtin, focus directly on situated, heteroglossic discourse and seek to characterize la parole, or language use, in useful terms: Their key principle is reciprocity between conversants, and their unit of analysis, as well as their level of social analysis, is the communicational dyad (writer-reader pairs; speaker-listener pairs). Because they focus on whole writer-reader communities, social constructionists deal with the effects of readers on writers in general terms, and they typically reify readers into “the Reader” when dealing with individual cases. By contrast, social interactionists, who concern themselves with describing actual, individual writers and readers, often in ethnographic studies, must articulate a principled analysis of writers and readers without analytic access recourse to group norms. This article first contrasts social constructionist and social interactionist approaches to the problem of discourse and then examines recent social interactionist studies concerning the effects of readers on writers' development, including investigations of word-segmentation skills, peer conferencing, and instructional discourse.