Using Google Translate’s Speech Features for Self-Regulated French Pronunciation Practice

Abstract

[EN] This study investigated the pedagogical use of Google Translate’s Text-To-Speech (TTS) and Automatic Speech Recognition (ASR) capabilities for improving L2 French students’ pronunciation of orthographic {-ent} endings in French: while {-ent} is silent in verbal constructions (e.g. “(ils) résident” [re.zid] ‘(they) reside’), the same sequence is pronounced [ã] in other forms such as nouns and adjectives (e.g. “(un) résident” [re.zi.dã] ‘(a) resident’). Twenty beginner learners of French completed self-regulated homework activities using Google Translate to listen to (to develop sound awareness and aural perception) and produce phrases (to develop oral production) containing the target {-ent} forms. A pretest, immediate posttest, and delayed posttest measured awareness, perception, and production of {-ent} pronunciation. Results of repeated-measures ANOVA indicate no significant improvement in awareness or perception, but a significant increase in production scores from pretest to each of the two posttests. The findings suggest that integrating Google Translate’s built-in speech technologies into L2 pedagogy can improve the oral production of French grapheme-to-phoneme rules, such as those observed in {-ent} sequences.