Issues in second language acquisition and language processing

Abstract

The focus of this special issue is on the interface between second language acquisition (SLA) and language processing. The four articles in this issue were originally presented at the workshop on 'Issues in SLA and language processing' held in June 2000 at the University of Paderborn, Germany. The articles tackle language processing from three rather different points of view. This diversity of perspectives is not atypical of research in second language (L2) processing which has in the past included such issues as procedural skills, attention, input and the L2 processor. The assumption that L2 acquisition is constrained by processing represents the basis for several approaches to SLA. This view constitutes a basic assumption in work on L2 input processing (e.g., van Patten, 1996), in research on L2 skill acquisition (e.g., McLaughlin et al. , 1983; Levelt, 1978; McLaughlin, 1987; Hulstijn, 1990; Schmidt, 1992), in work on operating principles (e.g., Andersen, 1984), in the 'competition model' (e.g., Bates and MacWhinney, 1982), in Clahsen's (1984) L2 processing strategies as well as in my own work on processability (Pienemann, 1998). The one thing that the articles in this issue have in common is that they all attempt to integrate aspects of processing into a theory of SLA. Hulstijn focuses on the integration of the notion of automatization into acquisition and representation. Carroll addresses the interface between L2 input and linguistic knowledge, and the two papers by Hâkansson et al. and by Di Biase and Kawaguchi explore the constraints assumed to be imposed on L2 development by the architecture of the emerging L2 processor (Pienemann, 1998) in a number cross-linguistic settings. In this article Hulstijn argues that current theories of the representation, acquisition and automatization of linguistic knowledge are at present largely incompatible due to fundamental differences in the underlying, assumptions made by the symbolist and the connectionist schools of thought. He makes the point that a coherent theory of SLA needs to reconcile some of these differences in order to benefit from the complementary strengths of both approaches. Hulstijn offers a speculative proposal towards