en fr

Acquisition of prosody in LX (alias "L2"): the accentuation (primary and secondary accents) in LX French by multilingual adult speakers of L1 Turkish

Acquisition de la prosodie en LX (alias « L2 ») : l’accentuation (accents primaire et secondaire) en français LX par des locuteurs adultes plurilingues de turc L1

Abstract : This presentation contributes to the domain "Language acquisition" of the international congress of the AFLS 2019, which is about different varieties of French (« Le français d’ici, de là, de là-bas »). This experimental acoustic study is indeed about the acquisition of the initial and final accents in a variety of LX French as a third or fourth language (Dewaele 2017) spoken by L1 Turkish multilingual young adult speakers living and studying in the town of Rouen in Normandy (about one hour and a half in the West direction from Paris, France). Given the similarities and differences between the French initial and final accents and the Turkish primary and secondary accents, regarding their locations, functions and acoustic features (Lewis1967/2000, Fónagy 1979, Sezer 1983, Van der Hulst & van der Weijer 1991, Göksel & Kerslake 2005, Levi 2002, 2005, Di Cristo 2016), the aim is to examine how the accents are realized in French by these Turkish speakers. The contrastive analysis of the accentual systems of both languages in the theoretical frame of the autosegmental metrical phonology (Pierrehumbert 1980, Beckman & Pierrehumbert 1986, Pierrehumbert & Beckman 1988) suggests that, compared with French as a first language, the Turkish speakers realize more phonetic differences than phonological ones (Mennen 1999, 2007). The presentation will show that the first results of the acoustic analysis with Praat confirm our hypotheses: at the phonological level, we'll see (i) which locations and functions the French initial and final accents have in the accentual phrase, (ii) if they are realized with the relevant acoustic features (F0 and duration?), and (iii) if the French initial accent is reinterpreted or not as a Turkish non final primary accent (Sezer 1983); at the phonetic level, we'll see which lengthening and F0 variations are realized over the accented syllables. These first results apparently confirm the predictions of the “stress deafness” Hypothesis (Dupoux & al. 1997, 2010) and of the Markedness Differential Hypothesis (Eckman 1977, 2008).