TASK-BASED INTERACTION AND ACQUISITION

Investigating the empirical link between task-based interaction and acquisition: A meta-analysis

Investigating the empirical link between task-based interaction and acquisition: A meta-analysis

Introduction

Second Language Acquisition (SLA), as a sub-discipline of applied linguistics, is still a very young field of study. While it may not be possible to identify its precise starting point, many researchers would agree that the late sixties marked the onset of an intense period of empirical and theoretical interest in how second languages are acquired. Much of this research has been directed at understanding and contributing to more effective instructed language learning.

In addition to the numerous studies that have investigated the effects of instruction on learning (Norris and Ortega's meta-analysis published in 2000 identified 79 studies), much of the theorizing about L2 instruction has been specifically undertaken with language pedagogy in mind, for example Krashen's Monitor Model (Krashen, 1981), Long's Interaction Hypothesis (Long, 1996), DeKeyser's skill-learning theory (DeKeyser, 1998), VanPatten's input processing theory (VanPatten, 1996; 2002) and my own theory of instructed language learning (Ellis, 1994) all address the role of instruction in L2 acquisition.

However, the research and theory do not afford a uniform account of how instruction can best facilitate language learning. There is considerable controversy (see Ellis, forthcoming). In particular, there is no agreement as to whether instruction should be based on a traditional focus-on-forms approach, involving the systematic teaching of grammatical features in accordance with a structural syllabus, or a focus-on-form approach, involving attention to linguistic features in the context of communicative activities derived from a task-based syllabus or some kind of combination of the two. Nor is there agreement about the efficacy of teaching explicit knowledge or about what type of corrective feedback to provide or even when explicit grammar teaching should commence.

These controversies reflect both the complexity of the object of enquiry (instructed language acquisition) and also the fact that SLA is still in its infancy. Given these controversies, it might be thought unwise to attempt to formulate a set of general principles of instructed language acquisition. Hatch's (1978a) warning - 'apply with caution' - is as pertinent today as it was some thirty years ago. Nevertheless, I think there is a need to try to draw together a set of generalisations that might serve as the basis for language teacher education, and I am not alone in this, for Lightbown (1985; 2000) has felt and responded to a similar need. If SLA is to offer teachers guidance, there is a need to bite the bullet and proffer advice, so long as this advice does not masquerade as prescriptions or proscriptions (and there is always a danger that advice will be so construed) and so long as it is tentative, in the form of what Stenhouse (1975) called 'provisional specifications'. I have chosen to present my own provisional specifications in the form of 'principles'. I do not expect that all SLA researchers or all language teachers will agree with them. I hope, though, that they will provide a basis for argument ...