A perceptual study of monosyllabic Mandarin tones produced by three-year-old children growing up in Taiwan

Abstract

Wong et al. (2005) reported that monolingual Mandarin-speaking three-year-old children in the U.S. have not produced adult-like Mandarin lexical tones in monosyllabic words. This study adopted the same methods to examine monosyllabic Mandarin tones produced by three-year-old children in Taiwan. Four hundred and thirty eight monosyllabic tones were collected from ten children and seven mothers in Taiwan. These productions and the 92 productions by the four U.S. mothers reported in Wong et al. (2005) were low-pass filtered to retain pitch information, but eliminate lexical information, and were presented to five Mandarin-speaking adults in Taiwan for tone categorization. U.S. mothers' tones were categorized with ceiling accuracies, similar to those judged by U.S. Mandarin-speaking adults reported in Wong et al. (2005). Taiwan mothers' tones were judged with lower accuracies, with the accuracy of tone 2 significantly lower than U.S. mothers'. Taiwan children's four tones were judged with 63%, 24%, 27%, and 74% accuracy, all significantly lower than US and Taiwan mothers except for tone 1. Children made more errors in tones 2 and 3 than tones 1 and 4. These findings were consistent with those in Wong et al. (2005), suggesting similar developmental patterns for children growing up in the U.S. and Taiwan.