Computational and psycholinguistic investigations of covert dependencies

语言是了解人类内心活动的一扇窗户。以前的大量工作研究了说英语的人是如何理解他们的母语的，以及说英语的孩子是如何发展他们的阅读、写作和会话技能以成为足够的英语使用者的。对于不同语言的说话者在语言交际中是否采取了质的或量的不同策略，我们知之甚少。对潜在跨语言差异的良好理解直接关系到与美国社会相关的各种重要问题，包括开发更有效的外语学习和教学策略，以及从外国文本、网站和书面或口头语料库中提取信息的复杂技术工具。这个项目研究的语言是普通话，作为中国的官方语言，由于中国的经济和政治权力的日益增长的影响，普通话是一种对美国具有重大国家利益的语言。中国移民也是居住在美国的最大的外国出生群体之一。当前项目的特定经验域是关于中文问题的结构。该项目由两个主要组成部分组成。首先，一系列的行为实验将调查汉语使用者如何根据汉语的特殊语法特性，决定如何最好地记忆、回忆和预测汉语疑问句中编码的语言信息。其次，将开发详细的计算和统计模型，以提供一个框架来解释这些行为。还将开发与中文语料库合作的技术工具。行为方法和计算方法的结合使得回答关于英语和汉语之间存在哪些差异，以及语言使用者如何适应这些差异以最大限度地提高两种语言的交际效率的问题成为可能。该项目的很大一部分内容还包括对学生进行尖端研究问题和方法的培训，并与中国的领先语言研究人员和机构建立密切的国际合作。

Language is a window into the inner working of the human mind. A great deal of previous work has examined how English speakers understand their mother tongue, and how English speaking children develop their reading, writing and conversational skills to become sufficient English users. Comparatively little is known about whether speakers of different languages adopt qualitatively or quantitatively different strategies in language communication. A good understanding of potential cross-linguistic differences bears directly upon a variety of important issues relevant for U.S. society, including the development of more efficient learning and teaching strategies for foreign languages, and of sophisticated technological tools to extract information from foreign text, websites, and written or spoken corpora. The language under investigation in this project is Mandarin Chinese, which, as the official language of China, is a language with significant national interest to the U.S., due to the growing influence of China's economic and political power. Chinese immigrants also form one of the largest foreign-born groups living in the United States. The particular empirical domain of the current project is on the structure of questions in Chinese. The project consists of two major components. First, a series of behavioral experiments will investigate how Chinese speakers, conditioned by the particular grammatical properties of the Chinese language, make decisions about how to best memorize, recall and anticipate linguistic information encoded in Chinese questions. Second, detailed computational and statistical models will be developed to provide a framework to explain these behaviors. Technological tools will also be developed to work with Chinese corpora. The combination of the behavioral and computational methods makes it possible to answer questions about what differences exist between English and Chinese, and how language users adapt to these differences so as to maximize communicative efficiency in each language. A substantial part of the project also includes training students in cutting-edge research questions, methods, and establishing close international collaboration with leading language researchers and institutions in China.