Euh… but why? Facts and formal representations of French Schwa
Euh... Mais pourquoi? Faits et représentations formelles du schwa en français.
Résumé
Students required to transcribe French words often ask how they can know whether to use <ə>, <ø> or <œ> – the former being a “schwa” and the latter two being strong, lexical vowels (e.g., peur [pœʁ], ‘fear’; peureux [pørø], ‘coward’). Teachers usually answer that “if the word can be pronounced without the vowel, transcribe it as <ə>.” In other words, what determines whether a vowel is a schwa is not its acoustic cue, but its phonotactic behavior. Much phonological work has been devoted to the vowel known as “schwa”. Most of that work acknowledges that the quality of that vowel, at least in the contemporary French of France, is not a central [ə] (Dauses 1973, Malécot & Cholet 1977, Fourgeron et al. 2007). But its exact quality is not clear, beyond the fact that it is front, mid and rounded. The closest to an answer we have found is the much-cited paper by Bürki et al. (2008), which suggests that the realization of “schwa” is distinct from those of both /ø/ and /œ/. However, an obvious problem with this work is that it compares schwa, which is always unstressed, to both unstressed and stressed [ø, œ]. The present talk brings two new pieces to the puzzle, one empirical and one formal. First, experimental evidence is provided to the fact that acoustically, schwa is indistinguishable from /ø, œ/ – contra Bürki et al. (2008). Second, a rationale is proposed for why French uses such a marked vowel as its weakest vowel. Using the tiers of Kaye et al.’s (1985) Element Theory, it is claimed that the realization of schwa is feature-filling on all three available tiers.
Domaines
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|---|
| Licence |