Edited by Zhisheng (Edward) Wen and Mohammad Javad Ahmadian
[Task-Based Language Teaching 13] 2019
► pp. 279–302
Chapter 12. Task-induced second language development
A micro-genetic study
The present study provides a fine-grained analysis of how a 22-year-old Japanese English learner (Miwako) structured and restructured her language resources across a sequence of tasks which progressively increased in referential demand. It provides a picture of how language was tailored to communicative task demands through performance. The method of microgenetic analysis employed (Siegler, 2006) considers fluctuations in variability as well as the path, rate and breadth of change in the form-to-function mappings used in completing the task. It also considers sources of variability and possible mechanisms of change. In so doing, the discussion focuses on the emergence of two language forms that allowed the speaker to complete the task more effectively and efficiently (relative clauses and comparative structures) and relates this developmental variation to general fluctuations in CALF. The results indicate that over the course of the task sequence the learner’s L2 use showed evidence of:(1) an initial state of stabilized language use in completing easy versions of the task, (2) the destabilization of these routine linguistic strategies and a period of high variability in response to increased referential demand, and (3) subsequent organization of novel linguistic strategies in response to repeated exposure to these increased demands. The analysis thus demonstrates how task-induced L2 variation related to local-level change in the language the speaker used. It also shows how the processes of change over approximately an hour of task performance were parallel to those that have been documented over months (Larsen-Freeman, 2006; Verspoor, Lowie & van Dijk, 2008; Spoelman & Verspoor, 2010). Finally, it was found that CALF provided a complementary perspective that was broadly consistent with the microgenetic analysis in this case but masked variables in task-induced L2 change that may confound findings in other cases.