The multi-feature hypothesis: Connectionist guidelines for L2 task design

This study focuses on the effects of task type on the retention and ease of activation of second language (L2) vocabulary, based on the multi-feature hypothesis (Moonen, De Graaff, & Westhoff, 2006). Two tasks were compared: a writing task and a list-learning task. It was hypothesized that performing the writing task would yield higher retention and ease of activation of the target words than performing the list task. Translation tests and picture description tests were used to measure acquisition of the target vocabulary. Think-aloud protocols were used to study task performance. Results show that the writing group significantly outperformed the list group on both test types and on both direct and delayed tests. These findings are discussed in the light of task design characteristics and the issue of time on task.