Duez, D. (1995). On spontaneous French speech: Aspects of the reduction and contextual assimilation of voiced stops. Journal of Phonetics, 23, 407–427.
This study investigates the reduction of intervocalic voiced stops, aiming to better understand the realization of consonants in spontaneous speech. According to the hypo/hyperarticulation model proposed by Lindblom (1990), speakers would tune their performance to meet the function or formality of their speech, and then they would balance between the need for articulatory clarity and the minimal articulatory efforts. Therefore, the current study predicted that we should see reduction in the consonants of spontaneous speech, since it is a more casual genre and requires less articulatory efforts. The first experiments were conducted to see whether there were consonant reductions in spontaneous speech, and if there were, what kind of reductions they were. Experimental stimuli were segments of VCV structure extracted from the spontaneous speech of two male speakers, and then subjects were asked to listen to the stimuli and write down all they heard from each stimulus. Results showed that subjects indeed misheard some of the consonants, suggesting that these consonants were reduced, but interestingly, it was found that those misidentified consonants usually had the same place of articulation with their original consonants, but realized in a weaker manner (e.g., stops being identified as fricatives). In the second experiment, acoustic analyses were conducted to see if they could find acoustic clues to explain the consonant misidentification, and the results showed that those misidentified consonants did show the corresponding cues (e.g., the /d/ which was misidentified as /t/ would show a longer VOT). The results of the above two experiments confirmed the prediction that the consonants in casual speech would be reduced, and such reductions were usually related to weakening of articulatory gestures.
No comments:
Post a Comment