Doctoral Dissertation Research: The Representation and Processing of Vowel Contrasts

本博士论文项目通过一系列语音识别实验来研究元音的特征是如何在大脑中存储的，以及它们在语音理解中的作用。为了让听者理解口语，大脑必须成功地将听者听到的声音信息与他们所知道的特定单词相匹配。例如，一个以英语为母语的人必须对单词BED和BAD之间的元音区别很敏感，才能听到并在他们的脑海中检索到正确的单词。通过比较美国英语和其他语言的实验结果，该项目分析了这些元音特征在多大程度上可能以特定语言的方式被存储和处理。这项研究不仅对心理语言学领域做出了重要贡献，而且对第二语言学习者，特别是非英语母语学习者和其他语言母语学习者在课堂上进行具有挑战性的元音区分教学具有启示意义。在美式英语中，区分所谓的“tense”和“lax”元音的重要性是众所周知的，比如“beat”和“bit”。然而，目前还不清楚英语元音的紧张程度是如何在大脑中表现出来的，以及这对口语单词识别的影响。至于其他语言，对于某些元音是像英语那样通过紧张度来区分，还是通过其他类型的元音特征来区分，人们存在分歧。该项目结合了“跨模态启动”任务和“辨别”任务，前者要求听者在听到不同的单词后立即识别出一个单词，后者要求听者识别出两个连续的声音/单词是相同的还是不同的。这些实验为大脑对元音特征的分类方式提供了证据，以及这些特征如何导致语音理解过程中不同类型的心理表征的激活：音位（声音结构）、形态（单词结构）和语义（意义）。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This doctoral dissertation project uses a series of speech recognition experiments to investigate how the characteristics of vowel sounds are stored in the mind and the role they play in speech comprehension. In order for listeners to comprehend spoken language, the mind must successfully match acoustic information that the listener hears to specific words they know. For example, a native English speaker must be sensitive to the difference in vowel sounds between the words BED and BAD to hear and retrieve the right word in their mind. By comparing the results of experiments in American English with another language, the project analyzes to what extent these vowel characteristics may be stored and processed in a language-specific fashion. As well as making important contributions to the field of psycholinguistics, this research has implications for the teaching of challenging vowel distinctions in the classroom for second language learners, particularly non-native learners of English and native English learners of other languages.In American English, the importance of the distinction between so-called 'tense' versus 'lax' vowel sounds, such as in the words 'beat' versus 'bit', is well known. However, it is unclear how English vowel tenseness is represented in the mind and the consequence this has on spoken word recognition. As for other languages, there is disagreement about whether certain vowel sounds are distinguished by tenseness, as in English, or by other types of vowel characteristics. The project uses a combination of 'cross-modal priming' tasks, in which listeners must recognize one word right after hearing a different one, and 'discrimination' tasks, in which listeners must identify whether two consecutive sounds/words were the same or different. These experiments provide evidence for the way the mind categorizes the characteristics of vowel sounds, as well as how these characteristics lead to the activation of different types of mental representation during speech comprehension: phonological (sound structure), morphological (word structure), and semantic (meaning).This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.