Anglophonia ()

Awareness of glottal settings for the production of /h/-initial and vowel-initial words in French learners of L2 English

  • Christelle Exare

DOI
https://doi.org/10.4000/anglophonia.4809
Journal volume & issue
Vol. 33

Abstract

Read online

This phonetic study assesses the short-term efficiency of ecological training in glottal awareness for the reading of /h/-initial and vowel-initial words by French learners of L2 English. ‘Glottal awareness’ corresponds to speakers’ ability to produce and recognise three glottal settings: aspiration (e.g. the glottis is wide open for the realisation of the first phoneme of ‘hat’), continuous voicing (e.g. the vocal folds vibrate with no interruption at word boundary for glide linking in ‘two wapples’) and glottalisation (e.g. the vocal folds close abruptly and / or vibrate irregularly for ‘my ˀaunt’). First, 16 French learners of English were recorded while reading 18 target noun phrases. Second, they were given theoretical and practical instruction about glottal settings. The course took place during their English classes in their high school. The goal was to make them aware of compulsory initial aspiration (e.g. the hat), word-boundary continuous voicing (e.g. the japple, two wapples), and word-initial glottalisation (e.g. the ʔapple). Third, the same learners were asked to produce the same 18 target phrases, using particular glottal settings, 17 of them being imposed. They were finally asked to utter the last phrase (‘poor animal’) with the glottal settings of their choice — either continuous voicing or glottalisation — and then say whether they had realised one or the other. The 16 students’ acoustic data were coded and compared. The results show that after training the participants significantly improved their performances in the production of licit onsets: they produced fewer intrusive aspirations, fewer illicit elisions of /h/ and more licit productions at word boundaries. Although the students’ ability to produce correct onsets seems to correlate with their level, no evidence was found as regards i) their being aware of what they produced and ii) a possible correlation between level and awareness. More exploration is necessary to better examine whether multimodal training and reasoning can contribute to the acquisition of automatised phonetic processes.

Keywords