In this article, the influence of the prevocalic consonant on vowel duration in isolated French CV[p] sequences is studied. Nine French-speaking male subjects pronounced a corpus constructed using vowels [a,i,u,y] and consonants [p,t,k,b,d,g,fs,f,v,z,3]. Results showed that vowels were longer 1) after voiced than after unvoiced consonants, 2) after plosives than after fricatives and 3) that vowel duration varied according to the place of articulation of the prevocalic consonant only when it was a fricative. Indeed, vowels were longer after labial than after dental fricatives and longer after dental than after palatoalveolar fricatives. These results are in contradiction with the widespread hypothesis that the prevocalic consonant has no influence on vowel duration [3,13,16,8]. They agree with Di Cristo  and Wajskop  on voicing and with O'Shaughnessy  on articulatory mode. It is also hypothesised that if segmentation criteria were comparable, disagreements between published results would disappear in some cases.
Keywords: vowel duration in CV[p] utterances, prevocalic consonants, segmentation criteria.
Bibliographic reference. Thilly, CÚcile (1993): "Influence of prevocalic consonant on vowel duration in French CV[p] utterances", In EUROSPEECH'93, 465-468.