This research examined the effects of shadowing and elicited imitation (EI) on L2 speaking. In EI, learners wait for three seconds (longer than the phonological loop lasts) after they hear the stimulus before repeating. The results of our experiment suggest that these similar activities may involve different cognitive processing.
Communicative activities are in the mainstream of L2 pedagogy today, leaving non-communicative oral practices to seem somewhat outdated in the field of SLA. However, for lower-level learners in the EFL context, non-communicative practices, including shadowing and elicited imitation (EI), can be effective activities. Shadowing is assumed to train the subvocal rehearsal mechanism in the phonological loop of working memory, which conceivably retains perceived phonological information for one to two seconds. On the other hand, under EI, learners are instructed to wait three seconds (longer than the phonological loop lasts) before repeating, thus requiring them to understand the stimulus both structurally and semantically in order to reproduce what they have heard. In this presentation, we report how the effects of these two kinds of oral practices differentially emerged in spontaneous speaking among lower intermediate level learners. The experiment was conducted in a Japanese computer science university in 2014 and 2015. One group with 30 participants had shadowing practice for 15 to 20 minutes once a week over a 14-week long semester. The other group with 32 participants was provided with EI practice under the same condition. We compared the complexity, accuracy, and fluency of their spontaneous speaking elicited via a semi-direct speaking test before and after the experimental period. The results show that the effect of shadowing is stronger in fluency measures (shorter silence, fewer self-corrections, improved speech rate and phonation time ratio), whereas EI facilitated learners to produce longer sentences (more words per AS-unit) with limited effects on fluency (improvement in dysfluency). We believe the discrepancy occurred due to the different nature of shadowing and EI: shadowing focuses on the phonological loop and working memory, while EI induces learners to reproduce sentences as a pseudo-speaking experience and leads to faster access to their explicit knowledge.
