Are Asian Language Speakers Similar or Different? The Perception of Mandarin Lexical Tones by Naïve Listeners from Tonal Language Backgrounds: A Preliminary Comparison of Thai and Vietnamese Listeners^*

* This research was supported by an 11th Hakuho Foundation Japanese Research Fellowship (2016–17) and a 2018 Endeavour Research Fellowship. I thank the Editor in Chief of Australian Journal of Linguistics and two anonymous reviewers for their thoughtful comments, Mariko Kondo and Kaori Idemaru for giving me access to their phonetic laboratories at Waseda University and University of Oregon, respectively, and all the participants for making this study possible. Portions of this study were presented at the Japanese Society for Language Sciences 19th Annual International Conference (JSLS2017) at Kyoto Women’s University, Japan.

Abstract

Mandarin is one of the most representative tonal languages in the world with four tone categories (Tone 1 (T1): high level (ā); Tone 2 (T2): high rising (á); Tone 3 (T3): dipping (ǎ); Tone 4 (T4): high falling (à)). Learning Mandarin tones is known to be difficult for speakers from diverse linguistic backgrounds. The perception of Mandarin tones by naïve, non-native listeners from two tonal languages with a larger tone inventory than Mandarin—Thai and Vietnamese—was examined. The listeners’ discrimination accuracy of six tone pairs (T1–T2, T1–T3, T1–T4, T2–T3, T2–T4, T3–T4) was assessed and compared to that of native speakers of Mandarin on the one hand and Australian English on the other hand. The Thai and Vietnamese groups were clearly less accurate than the Mandarin group and showed a different pattern of results from each other. The Australian English group was less accurate than the Thai group only for T2–T4 and did not differ from the Vietnamese group for any of the pairs. Taken together, these findings suggest that first language tone knowledge may not necessarily be facilitative and that lack of experience with lexical tones may not disadvantage listeners from non-tonal language backgrounds in processing unfamiliar tones.

Keywords:

• Cross-language Speech Perception
• Mandarin Lexical Tones
• Thai
• Vietnamese
• Australian English

Notes on Contributor

Kimiko Tsukada completed her PhD in Linguistics at Macquarie University and postdoctoral research in Australia and the USA. Her research interests include experimental phonetics, cross-language speech production/perception and second/foreign language speech learning. Her research participants include speakers of Arabic, Burmese, Cantonese, English, Italian, Japanese, Korean, Malayalam, Mandarin, Mongolian, Norwegian, Persian, Telugu, Thai, Turkish and Vietnamese. Email: [email protected]

ORCID

Kimiko Tsukada http://orcid.org/0000-0001-6365-3322

Notes

* This research was supported by an 11th Hakuho Foundation Japanese Research Fellowship (2016–17) and a 2018 Endeavour Research Fellowship. I thank the Editor in Chief of Australian Journal of Linguistics and two anonymous reviewers for their thoughtful comments, Mariko Kondo and Kaori Idemaru for giving me access to their phonetic laboratories at Waseda University and University of Oregon, respectively, and all the participants for making this study possible. Portions of this study were presented at the Japanese Society for Language Sciences 19th Annual International Conference (JSLS2017) at Kyoto Women’s University, Japan.

1 The bilingual experimenter was born in Harbin, China and migrated to Australia before completing primary education. She is proficient in Mandarin as well as English and uses both languages in professional settings.

2 Apart from excluding tokens that had obvious background noise, there was no manipulation on the selected tokens before presenting them as testing stimuli. In other words, tokens were clearly identifiable as the indented tone, but contained natural phonetic variations.

3 In future research, it would be desirable to form more homogeneous groups of participants in terms of their estimated L1 proficiency.

4 An anonymous reviewer raised the possibility that replaying may have affected accuracy. Although the possibility cannot be excluded, it is unlikely to be serious (Tsukada 2011). Allowing replays would be more practical, as it is a natural part of speech communication and also helps to reduce participants’ anxiety.