Perceptual identification of talker ethnicity in Vancouver English

Abstract

Studies of ethnolinguistic variation typically begin by describing the speech production variables used to index social groups. In this study, we begin with indexical recognition – the perceptual identification of speakers’ self‐identified ethnic groups – to determine whether speakers produce ethnolinguistic variation and whether listeners are sensitive to it. Speech samples were recorded from thirty individuals from Metro Vancouver who self‐identified as Chinese, East Indian, or White Canadian. These utterances were used in a perception task where listeners categorized speakers’ ethnicities. Listeners’ social networks were labeled according to the ethnic group with which they reported spending the most time. Analyses indicate that while speakers vary in their productive expression of ethnolinguistic variation, listeners are consistent in their labeling choices. Listener accuracy was higher for voices from the listeners’ reported social group and White voices. These results suggest that familiarity with ethnic groups through social networks and mainstream culture influences indexical recognition.

The Effect of Ethnicity on Identification of Korean American Speech

“…If we assume that ethnically Korean listeners have, overall, the most experience listening to Korean voices, that non-Asian listeners have the least experience, and that non-Korean Asian listeners fall somewhere in between, then we could conclude that greater experience hearing Korean American voices leads to greater accuracy in identifying Korean ethnicity in unfamiliar voices. These results accord with similar findings in Hanna (1997), Newman and Wu (2011), and Wong and Babel (2017). Although it is of course possible for a non-Korean listener to have extensive experience with Korean voices, or for a Korean listener not to have heard many Korean voices in their life (e.g., transnational Korean adoptees raised in a predominantly non-Korean neighborhood), we believe in the safety of this assumption: out of 105 listeners, 86 described themselves as something other than "Asian" for their own ethnic self-identification.…”

“…Hanna (1997) hypothesized that a high-rising pitch contour throughout an utterance is characteristic of Asian American voices, while Newman and Wu (2011) analyzed several potential acoustic cues, including breathy phonation, Voice Onset Time (VOT), and formant quality of prevocalic /ô/, which all showed up in various combinations in the Asian American speech samples. Wong and Babel (2017) sampled Asian Canadian speakers, rather than Asian American speakers, but found some similar acoustic properties that were potentially attributable to a transfer effect from Cantonese into (North) American English, including a less-constricted postvocalic /ô/, /D/-stopping (e.g., this pronounced as dis), and less reduction in unstressed syllables.…”

“…Finally, Wong and Babel (2017) conducted a perceptual study of Chinese, East Indian, and White Canadians raised in the Metro Vancouver region of Canada. Listeners heard two-second clips of spontaneous speech from these speakers and were given a forcedchoice identification task (White, Chinese, or East Indian).…”

Nativeness perception and speaker voice as predictors of (non-)native English speaker evaluations in four ELF contexts

“…The listeners who had not been made aware of the accent origins of the speakers did not display these speaker evaluation patterns, which also suggests that, in general, identifying specific accents is challenging to listeners. This is supported by research that shows that, beyond general distinctions between (non-)native accents, or a standard or non-standard accent, listeners often struggle to correctly identify the exact accent they are listening too, unless they are familiar with a particular accent (Gnevsheva 2018;Nolan 2003;Brunner 2009;McGorrery & McMahon 2017;Nejjari et al 2012;2020;Wong & Babel 2017).…”

Language contact within the speaker: Phonetic variation and crosslinguistic influence

“…The Cantonese-speaking population in Canada is a diaspora that has extended roots (Yu, 2013), but the evidence as to whether the English of Cantonese-English bilinguals in Canada carries any unique signatures comparable to Hong Kong English is somewhat mixed. Controlled laboratory-elicited read sentences suggest pronunciation differences, if extant, are not perceptible to listeners (Babel & Russell, 2015), but more natural semi-spontaneous speech from early Cantonese-English bilinguals is categorized as associated with Chinese identity at levels well above chance (P. Wong & Babel, 2017). Hoffman and Walker (2010) examined the speech patterns of ethnically Chinese individuals in Toronto who were either first generation Cantonese-dominant speakers (n = 10) or second or third generation English-dominant speakers from historically Cantonese-speaking families (n = 23).…”