French Learners Audio Corpus of German Speech (FLACGS)

French Learners Audio Corpus of German Speech (FLACGS) Jane Wottawa, Martine Adda-Decker LPP, UMR 7018 CNRS - U. Paris 3 / Sorbonne Nouvelle 19, rue des Bernardins, 75005 Paris jane.wottawa@univ-paris3.fr, martine.adda-decker@univ-paris3.fr Abstract The French Learners Audio Corpus of German Speech (FLACGS) was created to compare German speech production of German native speakers (GG) and French learners of German (FG) across three speech production tasks of increasing production complexity: repetition, reading and picture description. 40 speakers, 20 GG and 20 FG performed each of the three tasks, which in total leads to approximately 7h of speech. The corpus was manually transcribed and automatically aligned. Analysis that can be performed on this type of corpus are for instance segmental differences in the speech production of L2 learners compared to native speakers. We chose the realization of the velar nasal consonant /N/. In spoken French, /N/ does not appear in a VCV context which leads to production difficulties in FG. With increasing speech production complexity (reading and picture description), /N/ is realized as [Ng] by FG in over 50% of the cases. The results of a two way ANOVA with unequal sample sizes on the durations of the different realizations of engma indicate that duration is a reliable factor to distinguish between [N] and [Ng] in FG productions compared to the [N] productions in GG in a VCV context. The FLACGS corpus allows to study L2 production and perception. Keywords: L2 German, L1 German, French learners, realization of [N] 1. Introduction Mastering a second language (L2) means to be able to ex- press oneself and to be understood by native speakers of that language. On one hand, native speakers are able to adapt their perception to accented speech but the learners’ intelligibility has to be high in order to permit native listen- ers to achieve utterance recognition close to 100% (Brad- low and Bent, 2008). On the other hand, the pronunciation of a foreign language is conditioned by the phonological system of the L1. Mastering the phonological system of the L2 improves the communication with native speakers. Flege et al. (1997) state that L2 speech production is linked to the phonemic inventories of both the L1 and L2. The study highlights that erroneous speech production is a re- sult of how close L1 and L2 sounds are in their acoustic properties and that the production skills of L2 learners do not only depend on perception skills in the L2. We recorded the French Learners Audio Corpus of German Speech (FLACGS) to investigate where pronunciation of German differs between German native (GG) speakers and French learners of German (FG). This resource was created to identify pronunciation difficulties of French learners in German. Results of this corpus are used to develop a pro- nunciation training for French learners of German. The long term aim of our research is to develop a training method that improves both pronunciation quality and per- ception accuracy in FG. At the time, we created the FLACGS corpus (2014/2015), we found …

Content truncated to 500 words