Investigating the interaction between voice quality and plosive production in Australian Englishes: Acoustic features of vowel-/t/ sequences

Abstract

A number of researchers have posited a link between voice quality and consonant production (e.g., Keating and Esposito 2007). This work investigates the connection between the acoustics of voice quality (laryngeal behaviour) and consonant realization (supralaryngeal behaviour) in Australian Englishes. Focusing on speech produced by 52 Australian English speakers (both “mainstream” and Aboriginal English speakers), we show that in vowel-/t/ sequences “breathy” t-categories (e.g., affricates, fricatives) co-occur with breathier vowels (including those with pre-aspiration), while laryngealized t-categories (e.g., ejectives, glottal stops) co-occur with creakier vowels. Vowels preceeding breathy /t/ categories have stastistically significantly higher F0 and H1–H2 values compared to vowels preceding canonical /t/. Meanwhile, vowels preceding glottal t-categories generally have lower mean spectral tilt measures (H1–H2, H1–A1), signifying increased creakiness in the vowel prior to glottal variants. We also report on dynamic analyses of H1–A1, paying attention to acoustic features of the last 25% of the vowel where peaks/troughs occur depending on consonant quality. A relationship between the voice quality of the vowel and the (broad) consonant type is demonstrated, however, it is not one-to-one. Rather voice quality appears to assist in creating the right conditions for particular consonantal variants. Keating, P. and C. Esposito. (2007), “Linguistic voice quality”, UCLA Working Papers in Phonetics, 105: 85-91.