Prosody in a contrastive learner corpus

This article presents the detailed description, constitution and rhythm analyses of an oral English L1/L2 learner corpus spoken by both native speakers of British English (L1) and French learners of English (L2). This contrastive learner database called ANGLISH was designed at the University of Provence at the Parole et Langage laboratory for a PhD project. So far, it has been used to study the rhythm of learners’ productions. The method used is a comparative analysis of French and native speakers’ productions using several recently-developed rhythm metrics with focus on the reading part of the ANGLISH corpus. The final aim is to set up an objective evaluation which could be implemented in computer-assisted systems for teaching English prosody.