AllData.xlsx

This study investigated how talker variability, training density, and working memory capacity interact to modulate the acquisition of Japanese vowel length contrasts by advanced Mandarin-speaking learners. Twenty native Mandarin speakers who had achieved JLPT N1 proficiency participated in a perceptual training experiment. In this experiment, two variables, namely talker variability (high vs. low) and training density (high vs. low), were manipulated. Pre-training and post-training tests were conducted to evaluate participants’ discrimination accuracy and reaction times. Additionally, working memory capacity was measured using digit span and reading span tests. While accuracy improved significantly post-training across conditions, the patterns of reaction time (RT) uncovered nuanced interactions between the training parameters and working memory. The RT results indicated that individual differences in working memory played a crucial role in modulating the learning efficacy. Learners with high working memory capacity reaped greater benefits from the cognitively challenging high variability/high density condition. In contrast, learners with low working memory capacity thrived under the less challenging, high variability/low density condition that effectively alleviated cognitive overload. These findings contribute significantly to refining theoretical models of second language phonological acquisition. They demonstrate that the quality of input interacts dynamically with cognitive capacity, ultimately determining the process of perceptual restructuring. Even for learners at an advanced proficiency level, it may be imperative to design targeted training programs tailored to individual cognitive profiles to persistent phonological plateaus. The results of this study hold important implications for optimizing phonetic training approaches, especially when catering to learners with diverse cognitive capacities.

本研究旨在探讨说话人变异性、训练密度与工作记忆容量三者的交互作用，如何调节以普通话为母语的高级日语学习者对日语元音长度对立的习得。本研究招募了20名已取得日本语能力测试（Japanese-Language Proficiency Test，JLPT）N1级水平的普通话母语者，参与一项知觉训练实验。本次实验共操纵两个自变量：说话人变异性（高vs.低）与训练密度（高vs.低）。研究分别在训练前与训练后开展测试，以评估被试的元音辨别准确率与反应时。此外，采用数字广度测验与阅读广度测验测量被试的工作记忆容量。尽管各实验条件下的训练后准确率均得到显著提升，但反应时（RT）结果揭示了训练参数与工作记忆之间的精细化交互模式。反应时分析表明，工作记忆的个体差异对学习效果的调节起到关键作用。工作记忆容量较高的学习者，从认知负荷较高的高变异性/高密度训练条件中获益更多；与之相对，工作记忆容量较低的学习者则在难度更低的高变异性/低密度训练条件下表现更佳，该条件可有效缓解认知过载。本研究结果为完善第二语言语音习得的理论模型提供了重要支撑。研究证实，输入质量与认知容量之间存在动态交互，最终决定语音知觉重构的过程。即便对于处于高级水平的学习者，设计贴合个体认知特征的针对性训练方案，或许仍是应对持续性语音学习高原期的必要举措。本研究结果对优化语音训练方法具有重要实践意义，尤其在面向具备不同认知能力的学习者开展教学时。