ABSTRACT
In philosophy, linguistics, and psychology it often is assumed that good theoretical work should keep a clear distance from applied teaching and intervention settings. The assumption of this article and of its associated program of research is that just the opposite strategy pays off—good theoretical work can be grounded in data collected from applied settings (cf. Anderson, 1987). A second central orientation for this chapter is that both applied and theoretical work need finer differentiation than we have so far seen. In order to achieve the kind of differentiated theories and differentiated interventions argued for, it is essential to recognize that many key questions are open rather than settled. This position is close to that of Elizabeth Bates and her colleagues with respect to dissociable mechanisms in language acquisition (Bates, Bretherton, & Snyder, 1988; see also chapter 2 of this volume). Premature closure, the treatment of important questions as if a little bit of rhetoric and data settle them when neither argumentation nor data have been conclusive, can seriously hamper progress in intervention and theory. Accordingly, a third theme for this chapter is that many key issues in language acquisition remain open. A final theme concerns the process of language acquisition itself: We argue that advances in development can occur very rapidly if the child has access to even a little bit of the right evidence. To account for this phenomenon a rare-event learning mechanism (RELM) is hypothesized and described.
