Analysis of Interaction Sequences in Computer-Assisted Language Learning

* A frequently cited research advantage of computer-assisted language learning (CALL) is the built-in data-collecting methods that can document learners' interactions as they work on learning activities (e.g., Bland, Noblitt, Armington, & Gay, 1990; Doughty, 1992; Jamieson & Chapelle, 1987). The suggestion is that such data can provide researchers with detailed information about learners' interactions and performance. However, these technical capacities for collecting interaction data need to be accompanied by well-motivated procedures for describing and interpreting them in terms of their value for language development (Chapelle, 1994). When the goals of CALL activities include learners' engagement in some form of negotiation of meaning (Long, 1985), procedures for data interpretation can borrow from interactionist research on task-based language learning that theorizes observable features of interaction expected to be positive for language development. The methods of interactionist task-based research rely on focused discourse analysis that attempts to identify instances of learners' interactions expected to be beneficial for language acquisition on the basis of specified theoretical tenets developed over the past 15 years (Gass, 1997; Gass & Madden, 1985; Long, 1996; Pica, 1994). Central to the theory is the claim that learners need to attend to, or notice, linguistic characteristics of the target language in order to acquire them (Robinson, 1995;