Miller, L. J., & Grosjean F. (1997). Dialectal effects in vowel perception: The role of temporal information in French. Language and Speech, 40(3), 277–288.
Presentation: Sarah
Summary: Shelly
One of the functions temporal information serves is to contrast long vowels from short ones. For example, in English, /i, æ, u/ can be distinguished from /ɪ ɛ ʊ/ by their duration, which co-varies with spectral information. However, in standard French, previous studies indicated that vowel duration only plays a minor role in this regards, and this distinction seems to exist only between /ɔ/-/o/ pairs. Following this rationale, the authors of the present study predicted that, given a spectrally ambiguous vowel between /ɔ/-/o/, English listeners should rely on duration for identification. If the ambiguous vowel is longer, than it will have more chances to be identified as /o/, and vice versa. However, for Standard French listeners, it is predicted that they may count on spectral cues more because durational cue is very minor in their phonological system. To test this prediction, the authors conducted an experiment by generating 10-step /kɔt/-/kot/ series at three levels of length as stimuli, and asked subjects to identify each stimulus as either cotte or côte (English listeners were aided with notes as cut or coat). Results showed that in the series of long duration, more tokens were identified as /o/ by English listeners, but for French listeners, they showed no differences across three levels of length, which means that they only utilized spectral information. The authors wanted to see whether this bias in weighting of perceptual cues can also be found across dialects in the same language, so they recruited Swiss French speakers as subjects for the second experiment. Different from Standard French, in Swiss French temporal cues play a more prominent role in vowel contrast. Therefore, if the above experiment is run on Swiss French speakers, one should expect that speakers of Swiss French should be more like English speakers and depend on durational cues more than speakers of Standard French. The results of the second experiment fulfilled this prediction. Based on the above findings, the authors concluded that the present study not only demonstrated different perceptual weightings across different languages, but such differences can also be found across dialects.
Presentation: Sarah
Summary: Shelly
One of the functions temporal information serves is to contrast long vowels from short ones. For example, in English, /i, æ, u/ can be distinguished from /ɪ ɛ ʊ/ by their duration, which co-varies with spectral information. However, in standard French, previous studies indicated that vowel duration only plays a minor role in this regards, and this distinction seems to exist only between /ɔ/-/o/ pairs. Following this rationale, the authors of the present study predicted that, given a spectrally ambiguous vowel between /ɔ/-/o/, English listeners should rely on duration for identification. If the ambiguous vowel is longer, than it will have more chances to be identified as /o/, and vice versa. However, for Standard French listeners, it is predicted that they may count on spectral cues more because durational cue is very minor in their phonological system. To test this prediction, the authors conducted an experiment by generating 10-step /kɔt/-/kot/ series at three levels of length as stimuli, and asked subjects to identify each stimulus as either cotte or côte (English listeners were aided with notes as cut or coat). Results showed that in the series of long duration, more tokens were identified as /o/ by English listeners, but for French listeners, they showed no differences across three levels of length, which means that they only utilized spectral information. The authors wanted to see whether this bias in weighting of perceptual cues can also be found across dialects in the same language, so they recruited Swiss French speakers as subjects for the second experiment. Different from Standard French, in Swiss French temporal cues play a more prominent role in vowel contrast. Therefore, if the above experiment is run on Swiss French speakers, one should expect that speakers of Swiss French should be more like English speakers and depend on durational cues more than speakers of Standard French. The results of the second experiment fulfilled this prediction. Based on the above findings, the authors concluded that the present study not only demonstrated different perceptual weightings across different languages, but such differences can also be found across dialects.
