EAPSI: Learning word representations containing foreign sounds may be easier when learning from multiple talkers

学习第二语言对于在当今的全球化世界中进行互动至关重要。掌握一门第二语言的困难很大程度上在于学习它的语音系统。不学习发音系统可能会导致不正确的单词表示，单词表示由较小的发音单位组成：例如，d，ah，g组成单词Dog。了解什么可以使含有外语发音的单词更容易学习，对于开发更好的第二语言教学方法非常重要。许多研究表明，听到多个说话者产生的单词有助于婴幼儿发展更健壮的单词表征。多个说话者也可以帮助同样对语言声音系统缺乏经验的第二语言学习者。然而，第二语言学习者并不是在所有情况下都能从多个说话者中受益的。这个项目将与台湾中正大学的蔡美珍博士合作，探索多个说话者对第二语言学习者有利的情况。中正大学心理语言学实验室的一项行为实验结果将有助于解释为什么多个说话者有助于形成稳健的词汇表征。它还将回答一个长期存在的问题：感知言语在多大程度上与产生言语有关？根据PI以前的研究，假设有经验的学习者在感知新单词时不会因为多个说话者而受益，而没有经验的学习者会受益。相反，假设即使是有经验的学习者在产生新单词时也会继续受益于多个说话者，尽管他们在感知新单词时将不再受益。这样的结果表明，说话者使用不同的单词表征来感知和产生单词。这项NSF EAPSI奖是与台湾国家科学委员会合作资助的。

Learning a second language is crucial to interacting in today's global world. Much of the difficulty in acquiring a second language is in learning its sound system. Failure to learn the sound system can lead to incorrect word representations, word representations being made up of smaller sound units: for example, the sounds d, ah, g make the word dog. Understanding what can make words containing foreign sounds easier to learn is important for developing better methods of teaching a second language. Many studies suggest that hearing words produced by multiple talkers helps infants and young children to develop more robust word representations. Multiple talkers may also help second language learners, who are likewise inexperienced with the language's sound system. However, second language learners may not benefit from multiple talkers in all circumstances. In collaboration with Dr. Jane Tsay at National Chung Chen University in Taiwan, this project will explore the circumstances in which multiple talkers are beneficial for second language learners. Results from a behavioral experiment at the Psycholinguistic Lab at Chung Cheng National University will shed light on the question of why multiple talkers are helpful for forming a robust word representation. It will also speak to the long-standing question: to what extent is perceiving speech related to producing speech? Based on the PI's previous studies, it is hypothesized that when perceiving new words, experienced learners will not benefit from multiple talkers, while inexperienced learners will. In contrast, it is hypothesized that even experienced learners will continue to benefit from multiple talkers when producing new words, although they will cease to benefit when perceiving new words. Such a result would suggest that speakers use different word representations for perceiving vs. producing words. This NSF EAPSI award is funded in collaboration with the National Science Council of Taiwan.