Study on the interface between syntax/semantics and phonology in terms of prominence

本研究取得了以下关于日语和英语话语中表示突出信息的韵律的结果（即，首先，我们从文本数据中挑选出相关的表达，并编辑了一个可以在计算机上使用的语言语料库。其次，我们研究了一个语言模型，整合词汇，语音，句法，语义和语用信息确定韵律的基础上，中心语驱动短语结构语法（HPSG）和underspecified的方法，由Pierrehumbert和Beckman倡导。在此基础上，我们提出了话语焦点生成的具体过程和语法规定。更具体地说，我们已经表明，我们可以指定一般和简洁的信息，在每个组件，即词汇，语法，语音和语用，并可以定义它们之间的最小接口。我们还明确指出，HPSG和未充分指定的方法是合适的框架，构建这样一个architecture.We认为，这项研究是重要的，因为它的跨学科和综合的方法，在该领域中，只有有限的问题集迄今已得到解决。

The research has achieved the following results concerning prosody in Japanese and English utterances representing information on prominence (i.e., new information or focus).First, we picked out relevant expressions from text data and edited a linguistic corpus which is available on computers. Second, we investigated a linguistic model integrating lexical, phonological, syntactic, semantic, and pragmatic information determining prosody on the basis of Head-Driven Phrase Structure Grammar (HPSG) and the underspecified approach advocated by Pierrehumbert and Beckman. Furthermore, based on the above model, we proposed specific processes and grammatical stipulations to generate focus in utterances. More specifically, we have shown that we can specify general and concise information in each component, i.e. lexicon, syntax, phonology and pragmatics, and can define minimal interfaces between them. We have also made clear that HPSG and the underspecified approach are suitable frameworks for constructing such an architecture.We believe that this study is of importance because of its interdisciplinary and integrated approach taken in the field where only limited sets of issues have hitherto been addressed.