Cantonese tone production in pre-school Urdu-Cantonese bilingual minority children

Aim: In this study, we examine the production of Cantonese tones by preschool Urdu-Cantonese children living in Hong Kong. Methodology: 21 first language Urdu second language Cantonese children (ages 4-6) and 20 age-matched first language Cantonese children participated in a picture-naming experiment with 86 words (109 syllables in total). Data and Analysis: Acoustic analysis was carried out for perceptually correct and incorrect tone productions of each tone. Comparisons were also made across speaker groups regarding accuracy rates and error patterns. Findings: Overall, first-language Urdu participants had lower accuracy and greater tone confusion than first language Cantonese participants. The pattern is attributable to influence from Urdu prosody, ongoing Cantonese tone mergers, and general sensitivity to phonetic information. Originality: This is the first empirical study on the acquisition of Cantonese tones by children who are heritage speakers of a non-tone language.