Acceso usuarios
This critical review examines the relationships between writing and working memory. The first part of the review proposes a definition of writing through the presentation of Hayes and Flower's classic model (1980) and the different writing processes. Then, in order to show the progressive integration of the notion of working memory in writing research, recent models, elaborated by Hayes (1996) and Kellogg (1996) are described. A series of studies, testing the relevance of Kellogg's model is discussed. The critical analysis of these experimental data leads to questions about the role of Baddeley's working memory model (1986) to account for (a) the dynamics of processing during writing and (b) the evolution of writing expertise. The second part of this review attempts to exceed these two limits in discussing the interest of theoretical ideas about working memory that differ from Baddeley's model. On the one hand, some activation models (Anderson, 1983 ; Cowan, 1993), that account for the dynamics of processing are presented, illustrated by different experimental work and analysed in the framework of text production. On the other hand, the adaptation made by McCutchen (1996) of Just and Carpenter's (1992) Capacity Theory and the adoption by Kellogg (1999) and McCutchen (2000) of Ericsson and Kintsch's (1995) theory about Long-Term Working Memory are discussed as potential theoretical paths to explain the development (in children) or the strengthening (in adults) of writing expertise. Finally, the conclusion, whose main objective is to sum up the different ways to conceive relationships between writing and working memory, focuses on the increasing importance of long term memory, considered as a resource register, storing the different pieces of knowledge implied in writing processing.
