No abstract

A previous study [H. Sussman, H. McCaffrey, and S. Matthews, J. Acoust. Soc. Am. 90, 1309-1325 (1991)] of American English CV coarticulation showed a remarkably linear relationship between onset frequencies of F2 transitions, plotted on the y axis, in relation to the F2 midvowel "target" frequencies, plotted on the x axis, for CVC tokens with initial [b d g] and ten medial vowel contexts. Slope and y-intercept values of regression functions fit to these scatterplots ("locus equations") were shown to serve as statistically powerful phonetic descriptors of place of articulation. The present study extends the locus equation metric to three additional languages--Thai, Cairene Arabic, and Urdu--having both two and four place contrasts for syllable-initial voiced stops. A total of 14 speakers (Thai = 6, Arabic = 3, Urdu = 5) produced 1740 CVC tokens that were acoustically analyzed using MacSpeech Lab II. Strong linear regression relationships were found for every stop category across all speakers. Slopes and y intercepts systemically varied as a function of place of articulation. Cross-language comparisons of stop place categories were performed but variability of slope and y intercept means tempered conclusions concerning the existence of CV "phonetic hot spots."

Perceived vowel height has been reported to vary inversely with the distance (in Bark) between the first formant frequency (F1) and the fundamental frequency (F0) [H. Traunmüller, J. Acoust. Soc. Am. 69, 1465-1475 (1981)]. Syrdal and Gopal [J. Acoust. Soc. Am. 79, 1086-1100 (1986)] observed that naturally produced [+high] and [-high] vowels tend to divide at a critical F1-F0 distance of 3-3.5 Bark, corresponding to the bandwidth of the "center of gravity" effect [L. Chistovich and V. Lublinskaja, Hear. Res. 1, 185-195 (1979)]. In the present study, listeners identified three sets of synthetic vowels varying orthogonally in F1 and F0 and ranging from /i/-/I/, /I/-/epsilon/, and /epsilon/-/ae/. For the /I/-/epsilon/ set, which corresponds to the [+high]/[-high] distinction, there was a relatively sharp identification boundary located at an F1-F0 distance of 3-3.5 Bark. However, for the /epsilon/-/ae/ and /i/-/I/ sets, which occupied regions where the F1-F0 distance was always greater than or always less than 3 Bark, vowel labeling varied more gradually as a function of F1-F0 distance. Also, F1-F0 distance was a better predictor of labeling performance than F1 alone only for the /I/-/epsilon/ set. Possible sources of the F1-F0 distance cue for vowel height are discussed.

Perceived vowel height has been reported to vary inversely with the difference (in Bark) between the first formant frequency (F1) and the fundamental frequency (F0) [H. Traunmüller, J. Acoust. Soc. Am. 69, 1465–1475 (1981)]. Syrdal and Gopal [J. Acoust. Soc. Am. 79, 1086–1100 (1986)] observed that naturally produced [+high] and [−high] vowels tend to divide at a critical F1–F0 difference of 3–3.5 Bark, corresponding to the range of the spectral center of gravity effect [L. Chistovich and V. Lublinskaja, Hear. Res. 1, 185–195 (1979)]. In the present study, listeners identified three sets of synthetic vowels varying orthogonally in F1 and F0 and ranging from /i/–/i/, /i/–/eh/, and /eh/–/æ/. For the /i/–/eh/ set, which spanned the [+high]/[−high] distinction, there was a steep identification boundary located at an F1–F0 difference of 3–3.5 Bark. However, for the /eh/–/æ/ and /i/–/i/ sets, which occupied regions where F1–F0 was always greater than or less than 3 Bark, vowel labeling varied more gradually as a function of F1–F0. In fact, for the /i/–/i/ set, F1–F0 was actually a poorer predictor of identification performance than F1 alone. [Work supported by NIDCD.]