Voice quality as a marker of ethnicity in New Zealand: From acoustics to perception¹

Abstract

This study is the first acoustic analysis of voice quality in the two main ethnic dialects of New Zealand English. In a production experiment, narratives from 36 speakers were analyzed and H1‐H2 spectral tilt measures were calculated for each vowel. The results provide instrumental evidence for impressionistic claims about the differing voice quality features of the two main ethnic groups, showing that Maori English speakers are creakier than European New Zealanders. A perception experiment was also carried out to determine the perceptual salience of voice quality for the identification of speaker ethnicity. The results of regression analyses confirm that listeners are sensitive to the phonation differences, and are able to rely on phonation cues in an ethnic dialect identification task. The study demonstrates the role of voice quality as a critical sociolinguistic variable, and highlights the importance of listeners’ previous dialect exposure in terms of sensitivity to prosodic cues. Ko tēnei pūrongo te tirohanga tuatahi ki ētahi āhuatanga o te reo e puta ana i te korokoro o ngā kaikōrero o ngā reo ā‐iwi e rua o te Reo Pākehā i Aotearoa. I tētahi whakamātau i āta tirohia ngā oropuare o ētahi kōrero nō ngā kaikōrero 36, ā, kua tatauria ngā ine H1‐H2 e kīia ana ko te spectral tilt. Ko ngā putanga he taunakitanga mō ngā whakaaro o te tāngata e pā ana ki te rerekētanga o ngā reo o te hunga Pākehā me te hunga Māori, e whakaatu hoki ana he kekē atu te reo o ngā mea e kōrero ana i te ‘Māori English’ i te reo e kōrerohia ana e te iwi Pākehā. He whakamātau whakarongo i whakahaerehia kia kite mena he āwhina tēnei āhuatanga o te reo kia mōhio te kaiwhakarongo ko wai te iwi o te kaikōrero. Ko ngā whakaputanga o ngā tatauranga e kī ana kei te tino mārama tēnei āhuatanga o te reo e puta ana i te korokoro o te tāngata, ā, ka taea ēnei rerekētanga te whakamahi kia whakawehea te tangata Māori i te tangata Pākehā. Nā tēnei ka kitea he āhuatanga motuhake te reo o te korokoro, ā, he mea nui hoki mena kua tino wāia te kaiwhakarongo ki ngā reo ā‐iwi. [Māori]

Social categories are shared across bilinguals׳ lexicons

“…Linguistic experience also plays a role in the perception of ethnic varieties. In the New Zealand context, for example, high previous exposure to the Māori English variety has been shown to have a beneficial effect on perception and accuracy in ethnolect identification tasks even when listeners only have access to suprasegmental information in the speech signal, such as rhythm (Szakay, 2008a), intonation (Szakay, 2008b), and voice quality (Szakay, 2012b).…”

“…Māori English GOAT-monophthongization (Maclagan et al, 2008b) appears to be a recent innovation that Warren and Bauer (2004) do not yet mention in their survey of Māori English, but is unlikely to be affected by contemporary phonological patterns in Māori. Listeners are aware of these differences between the two codes, at least insofar as they are able to identify ethnic identify from auditory stimuli (Szakay, 2012b).…”

The quantitative prevalence of creaky voice (vocal fry) in varieties of English: A systematic review of the literature

“…as [20] did for speakers of Standard Southern British English); another is to extract acoustic measures like spectral tilt, from which the degree of perceptual 'creakiness' (and 'breathiness') can be inferred (e.g. as [21] did for speakers of New Zealand English). Such studies are not included in this review because they do not offer an answer to the question what is the prevalence of creaky voice in English?…”

Intonational variation and incrementality in listener judgments of ethnicity

“…' Foulkes, Docherty, Khattab, and Yaeger-Dror (2010) found that listeners' identification of Tyneside children's gender was affected by two continuous measures (amplitude and F0) as well as several categorical measures; however, the authors also report significant correlations between amplitude and F0 in stimuli, suggesting potential issues with collinearity in the modeling procedure. In terms of voice quality, Szakay (2012) found that in New Zealand, ethnic identification was affected by several continuous voice quality measures; speakers with higher mean H1-H2 (a measure of creakiness) were likelier to be identified as Māori.…”