Recent perspectives on American sign language. H. Lane & F. Grosjean (Eds.). Hillsdale, N.J.: Lawrence Erlbaum Associates, 1980. Pp. v + 170.

Bereiter's framework for modelling different kinds of writing. The major experimental methods it describes are the protocol analysis of Hayes and Flower, and Gould's timing methods for examining letter writing processes. Yet much remains to be done. Critically needed at this time are adequate experimental tests of the proposed theoretical frameworks. For example, are the proposed writing processes (e.g., generating, organizing, and translating) empirically identifiable in the sense that different experimental factors (e.g., nature of topic) could be shown to affect them differentially? In addition, although several chapters in this book relate writing to the basic psychological processes of problem solving, it must also be characterized in relation to other basic processes like memory (see Kintsch, 1980; Black, Wilkes-Gibbs, and Gibbs, in press). Nonetheless, this book makes a convincing case that writing is an important phenomenon to study, and suggests ways in which we might continue to do so.