The Development of Voice Quality in Second Language Speech Acquisition: A Case Study of the Parallel Speech Corpus of Chinese Natives and Arabic Natives.

Abstract

Objectives: The primary objective of this study is to investigate whether the voice quality in second language (L2) speech production changes over time as learners progress in their L2 acquisition.

Methods: A total of 83 Arabic native speakers learning Chinese (59 males and 24 females) and 62 Chinese native speakers learning Arabic (23 males and 39 females) participated in the study. The participants had varying durations of L2 learning (DOL) experience. Each participant recorded word lists in both their first language (L1) and L2, and these recordings were analyzed for voice quality. The acoustic parameters examined included jitter, shimmer, F₀ mean, and F₀ range. A method for calculating the differences in voice quality between L2 and L1 was employed to dissipate voice difference. A linear mixed-effects model was fitted in R to analyze the effects of DOL, gender, and native language background on the L2-L1 voice quality differences.

Results: A significant triple interaction effect of DOL, gender, and native language background was observed for jitter, shimmer, and F₀ mean. These findings indicate that the effect of DOL on voice quality in L2 speech production is influenced by both gender and language background. The general trend showed that, with increasing learning duration, participants exhibited greater jitter and shimmer, while F₀ mean and F₀ range progressively approached the target language.

Conclusions: The findings demonstrate that voice quality in L2 speech production changes over time with learning duration. These changes are influenced by gender, and since the language-specific nature of voice quality in L2 production, the impact of learning duration on voice quality is also influenced by the learners' native language background.