RP or GA? French students' accents are drifting in the Atlantic - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

RP or GA? French students' accents are drifting in the Atlantic

Résumé

The teaching of English pronunciation in French universities has traditionally relied on two standards: General American and Received Pronunciation (Jobert & Mandon-Hunter, 2009: 15); the latter being usually favoured (Glain, 2013: 21; Przewozny, 2018). But a systematic examination of which standards our students actually adopt is still lacking. We therefore investigated accent preferences among French students of English through their oral productions. Twelve phonetic-rich sentences with pronunciation features specifically targeting the RP vs GA dichotomy were designed. They contained ten types of diagnostic features: the TRAP, BATH, LOT and THOUGHT vowels; the <-ile>, <-ization>, and <-ary/-ory> endings, and items potentially displaying T Voicing, rhoticity and Yod-Dropping. The set of twelve sentences included a total of 125 pronunciation features to examine. 304 undergraduate – mainly first and second year – students in the English department were audio recorded reading the sentences in a sound-attenuated booth. They familiarised themselves with the sentences before they produced them. Data from twenty students were thoroughly scrutinised via an auditory analysis performed by four teachers. Their responses were collected thanks to a purpose-built Praat interface allowing them to play back whole sentences or specific words as many times as necessary. For each phonetic feature, three response buttons were available: a UK flag when the RP variant was produced, a US flag for the GA variant, and a question mark for all other cases (e.g. when the item was mispronounced or the listener failed to determine which variant was used). The current analysis consists of 125 tokens × 20 speakers × 4 listeners = 10,000 responses. The aims of this study are threefold: 1) to assess whether auditory ratings were concordant between judges and across features; 2) to determine which accent was favoured by students; 3) to examine whether each speaker was consistent with respect to the RP vs. GA dichotomy. Pairwise agreement between judges ranged from 81.00 to 86.16%; the range rose to 87.24 and 92.40 once we removed question mark responses. By-judge rates of question mark responses over all features varied between 3.88% and 6.60%. The higher such rates were recorded for LOT and THOUGHT while e.g. <-ization> and BATH received fewer question marks. In order to further investigate the link between auditory ratings and oral productions, three vowels were chosen from the BATH, LOT and THOUGHT lexical sets, and manual formant analysis was performed (see figures in Appendix 2). Segmentation at the phone level had been automatically obtained by forced-alignment with WebMAUS. The general pattern shows that the ratings for BATH and THOUGHT were generally consistent with the acoustics, but LOT seemed more problematic. Turning to students’ preferences, the figure in Appendix 3 shows the proportion of RP and GA features for each student. It suggests that most speakers borrowed from both varieties, and that overall, RP and GA were almost equally represented, which follows the trends observed in Yibokou et al. (2019). The whole set of quantitative results in our final presentation will lead us to question our teaching habits.

Domaines

Linguistique
Fichier non déposé

Dates et versions

hal-04138998 , version 1 (23-06-2023)

Identifiants

  • HAL Id : hal-04138998 , version 1

Citer

Emmanuel Ferragne, Anne Guyot Talbot, Hannah King, Sylvain Navarro, Nada Azizi. RP or GA? French students' accents are drifting in the Atlantic. PAC 2023 – Spoken English varieties: interfaces and multidimensional approaches, Apr 2023, Paris, France. ⟨hal-04138998⟩
21 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More