Form and meaning in dialog-based computer-assisted language learning

The goal of this thesis is to explore how foreign language learning can be facilitated through the use of intelligent computer-assisted language learning (ICALL) based on natural language processing (NLP) methods. ICALL was provided in the form of a task-based dialog system that gives corrective feedback.We investigated how different parameters of the interaction affect the learning progress. Based on a comprehensive review of existing comparable ICALL applications and the underlying methods and technology, we selected parameters linked to the sophistication and effort required to implement a particular form of interaction and related them to parameters that are based on twomuch debated issues from the field of second language acquisition (SLA). One is the debate that pits form against meaning and leads to a discussion of the extent to which language instruction should focus on linguistic forms and formal correctness as opposed to emphasizing communicative skills and the ability to use the language to make meaning in the real world. Related to that is the second controversial issue which concerns the dichotomy between implicit and explicit knowledge, learning and instruction: How explicit or implicit should instruction be, how does the degree of explicitness affect the development of explicit and implicit knowledge, and how do these two types of knowledge contribute to language skills? These two general issues are condensed into three different experimental conditions, that differ with regard to how much they constrain the learner input and how explicit the feedback is. More precisely, we compare strictly form-focused activities where the learner input is constrained to supply a grammatical target form with generally unconstrained participation in a meaning-oriented task-based dialog. For the latter, we further compare recast and metalinguistic feedback as implicit and explicit types of feedback respectively. The findings of this study indicate that there are small differences in the language skill development afforded by different types of computer-provided instruction. We find that constrained, explicit form-oriented instruction yields in general greater immediate learning gains, while the free, more implicit and meaning-oriented instruction yields more delayed effects. Similarly, comparing implicit recast feedback with explicit metalinguistic feedback we find that the immediate effects are on par but recast feedback leads to greater delayed effects. These differences interact considerably with other parameters of the experimental setting, in particular with the selected target structures. This suggests that the effectiveness of certain types of instruction is highly dependent on the particular content and goal of the instruction. By using current SLA issues as motivation and guide to develop an ICALL system and an experimental framework this work contributes to the yet small field of existing research and development which integrates ICALL and SLA perspectives.