Linguists and philosophers have traditionally argued that definite constructions presuppose familiarity on the part of the addressee. This article examines empirically the question of what kind of familiarity, in the context of newspaper editorials, this might be. A significant issue, articulated by literacy theorist Walter Ong, is the nature of the reader and whether a writer can know what a reader is familiar with. Taking a case study approach, the author examines definite constructions in 15 editorial articles from the Christian Science Monitor. These constructions are classified, following Brown and Yule (1983), as either re-evoking, new, or inferrable. It is argued that for purposes of studying the writer-reader relationship, the inferrables are most interesting since they indicate what the writer believes the reader is capable of inferring. Ultimately both the new and the inferrable show that writers use definite constructions in accord with genre conventions. The author concludes that such conventions make communication efficient.

Readers and evaluators of children's writing still fall back on deficit explanations; papers are read for signs of what they lack rather than signs of growth. Presented here is a model that predicts how such growth may occur as a logical outcome of language acquisition. Drawing on research done in the past, the article provides a list of the kinds of language learning underway in the elementary school years and suggests that teachers may use this list to anticipate where and how such learning will influence the writing processes of children. Included in the list are sentence syntax, spelling conventions, and discourse grammars, all of which seem to be learned by “creative construction” (hypothesis building and refinement) and, to some extent, memorization. The article argues that children's writing performance is likely to suffer on one or more writing dimensions as the writer selectively attends to other dimensions of the task. For evaluators and teachers there are implications for feedback, for individual agendas, for revision, and for the kinds of conclusions one may draw from the examination of written products.