Written training tasks are better than oral training tasks at improving L2 learners’ speech production

Clara Solier (1) , Cyril Perret (2) , Lorraine Baqué (3) , Christiane Soum-Favaro (1)

1 Octogone - Octogone Unité de Recherche Interdisciplinaire
2 CeRCA [Poitiers] - Centre de recherches sur la cognition et l'apprentissage [UMR 7295]
3 UAB - Universitat Autònoma de Barcelona = Autonomous University of Barcelona = Universidad Autónoma de Barcelona

This study examined the effect of written input on the production of word-final vowels (oral: /i/, /e/; nasal: /ɑ̃/, /ɔ̃/) by 100 native Moroccan learners of French as a second language, relying on a pretest/posttest paradigm. During pretest, participants performed a word repetition task. Then, participants were assigned to five experimental (training) conditions, varying in the modality of stimulus presentation (oral vs. written) and in the modality of response (oral vs. written), before their oral production of the word-final vowels was evaluated in the posttest. Results clearly showed that posttest pronunciation accuracy was affected by the presence of orthographic information in the training task. The copy training task was the most efficient at improving posttest accuracy. Results indicate that orthography, and more specifically the written production copy task, helps L2 learners’ pronunciation more efficiently than phonetic correction.

Domaines: Psychologie; Linguistique

Clara Solier, Cyril Perret, Lorraine Baqué, Christiane Soum-Favaro. Written training tasks are better than oral training tasks at improving L2 learners’ speech production. Applied Psycholinguistics, 2019, 40 (6), pp.1455-1480. ⟨10.1017/S014271641900033X⟩. ⟨hal-02444016⟩