Vowel-to-vowel coarticulation
Bob McMurray, Jennifer Cole, Cheyenne Munson, and Gary Linebaugh
Talkers vary their vowel sounds when speaking, but rather than being random these variations happen
systematically depending on many contextual factors. The other vowels that are pronounced in proximity to the
target vowel are one such factor. For example, we expect an /ε/ near an /i/ to be
more fronted and raised compared to an /ε/ near an /æ/. The vowels
/ε/ and /ʌ/ in particular may undergo coarticulation from many directions
because of their central position in vowel space, illustrated at right.
In this research project we have examined the systematic production of vowel-to-vowel coarticulation accross word boundaries. For instance, the vowel in wet is produced slightly differently when pronounced in the phrase "wet eagle" instead of the phrase "wet octopus." The figure below is from our production study in which talkers pronounced words with /ε/ and /ʌ/ vowels followed by, or in the context of, words containing /i/, /æ/, and /ɑ/ vowels. The figure shows the effect of these context vowels on the two target vowels.
Production data has shown that the effect of vowel-to-vowel coarticulation accross word boundaries is present, but the context vowel is only one of many factors that affects the target vowel. In fact, the effects of the context vowel can only be seen after accounting for variance due to other factors, such talker identity and the consonant between the vowels. A parsing process, in which variance is assigned to different factors, can allow for the extraction of as much information as possible from the input. Further experiments are needed to see whether listeners employ this type of parsing.
Publications:
Cole, J. S., Linebaugh, G., Munson, C., & McMurray, B. (2010). Unmasking the acoustic effects of vowel-to-vowel coarticulation: A statistical modeling approach. Journal of Phonetics, 38, 167-184.
McMuuray, B., Cole, J. S., & Munson, C. (in press). Features as an emergent product of perceptual parsing: Evidence from vowel-to-vowel coarticulation. R. Ridouane and N. Clement (Eds.) Where do Features Come From?
Presentations:
Cole, J. S., Linebaugh, G., Munson, C., & McMurray, B. (2007, October).
Vowel-to-vowel coarticulation across words in English: Acoustic evidence.
Paper presented at the 13th Mid-Continental Workshop on Phonology. Columbus, OH.
abstract
Cole, J., McMurray, B., Linebaugh, G., & Munson, C. (2007, October). Parsing acoustic variability as a mechanism for feature abstraction. Paper presented at Where do Features Come From? Phonological Features in the Brain, the Mouth, and the Ear. Paris, France.