Knotted and Entangled: New Light on the Identification, Classification and Value of Language Related Episodes in Collaborative Output Tasks

Abstract

Our study is of the collaborative output of students engaged in grammar dictation tasks (Dictogloss) in a British EFL context. Our key aim was to investigate the effect of pre-task focus on form work on task output processes. The main focus of this paper, however, is on how we refined the framework of Language Related Episodes (LREs) used by other researchers to analyse L2 output processes, and produced our own taxonomy of episodes. We also discuss the role of collaborative output in L2 development, particularly the use of metalanguage. We argue that analysis based on LRE counts, although valuable, fails to capture completely the complexity of the interaction, and go on to discuss the 'value' and 'nature' of episodes. The former term is concerned with differences in the length of episodes and the extent of learners' engagement with the language item concerned. The latter reflects the messy nature of L2 interactional data and the difficulty of 'untangling' it to identifying episodes. It deals with features like discontinuity, embedding and overlapping.