Language Experience Influences Sociolinguistic Development: The Role of Speaker Race and Language Attitudes on Bilingual and Monolingual Adults' Accent Processing

Abstract

Background/objectives: Speaker race and the listener's language experience (i.e., monolinguals vs. bilinguals) have both been shown to influence accent intelligibility independently. Speaker race specifically is thought to be informed by learned experiences (exemplar model) or individual biases and attitudes (bias-based model). The current study investigates speaker race and the listener's language experience simultaneously as well as listeners' attitudes toward non-native speakers and their ability to identify the accent.

Methods: Overall, 140 White English monolinguals and 140 English/Norwegian bilinguals transcribed 60 Mandarin-accented English sentences presented in noise in the context of a White or East Asian face. Following sentence transcription, participants were asked to rate the strength of the accent heard and completed a short questionnaire that assessed their accent identification ability and their language usage, proficiency, familiarity, and attitudes.

Results: Results show that a listeners' ability to identify an accent and their attitudes toward non-native speakers had a significant impact on accent intelligibility and accentedness ratings. Speaker race by itself did not play a role in accent intelligibility and accentedness ratings; however, we found evidence that speaker race interacted with participants' accent identification scores and attitudes toward non-native speakers, and these interactions differed as a function of language experience.

Conclusions: Our results suggest that bilinguals' sociolinguistic processing may be more in line with a bias-based model than monolinguals.

Keywords: accent intelligibility; bilingualism; psycholinguistics; sociolinguistics; speech perception.

Figure 1

White face (a) and East Asian face (b) taken from the Chicago Face Database.

Figure 2

Transcription accuracy (y-axis) by condition (x-axis). Mean transcription accuracy indicated with a red dot.

Figure 3

Transcription accuracy (y-axis) by accent ID (x-axis). Mean transcription accuracy indicated with a red dot.

Figure 4

Transcription accuracy (y-axis) by attitude (x-axis). Average language attitude score had a range of 0–15 with higher scores reflecting more positive attitudes held toward non-native English speakers.

Figure 5

Accentedness ratings (x-axis) by language group (monolingual vs. bilingual). Accentendness rating was on a scale of 1–7 (1 = no foreign accent, 7 = very heavy foreign accent).

Figure 6

Accentedness ratings (y-axis) by attitude (x-axis), language group, and condition. East Asian face (EAF) condition pictured on the left; White face (WF) condition pictured on the right. Average language attitude score had a range of 0–15 with higher scores reflecting more positive attitudes held toward non-native English speakers.

Figure 7

Transcription accuracy (y-axis) by condition and accent ID (x-axis). Asterisk (*) indicates significant interaction.

Figure 8

Transcription accuracy (y-axis) by attitude (x-axis) and condition. The East Asian face (EAF) condition is pictured in red, and White face (WF) condition is pictured in blue. Average language attitude score had a range of 0–15 with higher scores reflecting more positive attitudes held toward non-native English speakers.

Figure 9

Accentedness rating (x-axis) by condition. Accentendness rating was on a scale of 1–7 (1 = no foreign accent, 7 = very heavy foreign accent).

Figure 10

Transcription accuracy (y-axis) by accent ID (x-axis). Mean transcription accuracy indicated with a red dot.

Figure 11

Accentedness rating by condition. Accentendness rating was on a scale of 1–7 (1 = no foreign accent, 7 = very heavy foreign accent).