The role of gender and ethnicity in the specification of American English vowels

This study explores acoustic variation in vowel phoneme production for adult speakers of General American English. Peterson and Barney (1952) conducted an analysis of the acoustic features of American English vowels in order to map the phonological vowel space; Hillenbrand et al. (1995) replicated and extended their classic study. Both studies presented acoustic measures for men and women. The findings across the two studies include target values for various acoustic cues in the speaker productions, including fundamental frequency (F0), formant frequencies, and duration measures. However, no mention is made of speaker ethnic background, and studies have since suggested that there is evidence of both perceptual and acoustic variations with respect to speaker ethnicity [Thomas & Reaser, (2004)]. The purpose of the current research is to replicate and extend the findings of the previous studies to determine whether differences exist with respect to ethnicity that do not arise from dialect features. Speakers were recorded producing words and sentences containing /hVd/ target stimuli. Speaker productions were screened for dialectal features. Temporal and spectral steady-state vowel measures of acoustic variation (F0, formant values, and vowel duration) were obtained for all speakers. Results are presented for the variations across the speaker gender/ethnic groups. [Hope College Psychology Department, The Carl Frost Center for Social Science Research, The Jacob E. Nyenhuis Student/Faculty Cooperative Grant