ABSTRACT
Children's use of linguistic expressions to introduce, maintain, or shift reference in narratives has been investigated in several studies (Bamberg, 1987; Hickmann, 1980, 1991; Karmiloff-Smith, 1981, 1985). But the problem of what children have actually learned when they use linguistic forms with referential functions similar to those used by the adults is still open to study. Do children learn to map linguistic forms onto more specific and differentiated functions, in much the same way as in other domains of language acquisition (Bamberg, Budwig, & Kaplan, 1991)? Or do children learn new strategies of discourse processing that enable them to select linguistic forms according to the demands of both the “macro” and “micro” construction of discourse (Karmiloff-Smith, 1985, 1986)?
