Doctoral Dissertation Research: Learning of Novel Phonetic Categories After Training in Perception and Production

人类拥有产生和理解言语的非凡能力。 为了成功地学习一门语言，一个人必须了解特定语言中的声音类别是什么，以及是什么使各种声音彼此不同。 例如，“r”和“l”在英语中属于不同的声音类别，但在其他一些语言中则不然。 学习如何区分这些声音并发出它们是学习语言的重要组成部分。目前，人们对于听众如何学习感知语音类别已经了解很多。然而，人们对说话者如何在生产中学习语音类别以及一种模态（即感知或生产）中的学习如何与另一种模态中的学习相关联知之甚少。 感知和生产之间的关系可以有多种形式。 以一种方式学习总是可以引导以另一种方式学习（即，一个人必须能够感知到一种区别，然后才能产生这种区别）。 两种模式的类别形成可以并行发生（即感知和产生可以同时学习新的声音）。 或者，以一种方式学习实际上可能会阻碍另一种方式的学习（例如，语音生成中的学习可能会破坏语音感知中的学习）。 本研究旨在区分这些假设。本研究中的实验旨在训练参与者新的声音类别，训练重点是感知或产生。 然后参与者在感知和生产方面都受到测试。 弄清楚新的声音类别是如何在两种模式中形成和访问的，不仅可以深入了解新声音类别的学习过程，还可以深入了解这两种模式之间的关系。这项研究的结果在理论和应用领域都具有重要意义。 更好地理解类别是如何形成的将有助于为第一语言和第二语言学习者提供学习理论。 了解语音感知和生成之间的关系将为感知和生成中语音处理模型的开发提供信息。 在应用领域，了解感知和产生的界面，特别是在声音类别的层面上，对于第二语言教师将是有益的。 此外，在学习（或重新学习）新的声音类别时，临床医生（例如言语病理学家）将受益于更好地理解这两种模式如何相互作用。 例如，如果患者言语产生有障碍，他们也许能够利用其完整的感知能力来重新训练受损的模式。

Humans possess a remarkable ability to produce and understand speech. In order to successfully learn a language, an individual must learn what the sound categories in a particular language are and what makes various sounds different from one another. For instance, "r" and "l" belong to different sound categories in English, but not in some other languages. Learning how to differentiate between these sounds and produce them is an important part of learning a language. Currently much is known about how listeners learn to perceive the speech sound categories; however, significantly less is known about how speakers learn about speech categories in production and how learning in one modality (i.e., perception or production) is related to learning in the other modality. The relationship between perception and production could take many forms. Learning in one modality could always lead learning in the other (i.e., an individual would have to be able to perceive a distinction before they could produce it). Category formation could occur in parallel for both modalities (i.e., both perception and production could learn about new sounds simultaneously). Alternately, learning in one modality could actually hinder learning in another modality (e.g., learning in speech production could disrupt learning in speech perception). This study aims to differentiate among these hypotheses. The experiments in this study are designed to train participants on novel sound categories with the training focusing on either perception or production. Then participants are tested in both perception and production. Untangling how new sound categories are formed and accessed in the two modalities will give insight not only into the learning process of new sound categories but also into how the two modalities are related. The findings of this study will be significant in both theoretical and applied domains. Having a better understanding of how categories are formed will help inform theories of learning for both first- and second-language learners. Understanding how speech perception and production are related will be informative in the development of models of speech processing in both perception and production. In the applied domain, understanding how perception and production interface, particularly at the level of sound categories, will be beneficial for second-language teachers. Furthermore, clinicians, such as speech-language pathologists, will benefit from a better understanding of how the two modalities interact when learning (or re-learning) novel sound categories. For instance, if patients have an impairment in speech production, they may be able to use their intact perceptual abilities to re-train the impaired modality.