A Theoretical and Empirical Study of the Notion of Phase in Generative Grammar

在生成语法的框架内，它解释了人类心智/大脑中语言能力的结构和功能，已经有各种各样的尝试从计算效率的操作中推导出之前框架中提出的计算操作的影响。本研究聚焦于相位的概念，这是最小化程序（最新的生成语法框架）中降低计算复杂性的核心概念之一，并寻求概念上和经验上合理的相位概念定义。具体来说，本研究调查了wh岛和that-t效应。在生成型文献中，人们普遍认为h岛效应包含在最小连接条件下。基于迄今为止未被注意到的事实，即允许从英语间接疑问句中提取特定的wh短语，本研究认为，在最小链接条件下不能包含wh岛效应，但我们仍然需要基于相位概念的传统从属类型解释。它还认为，t效应可以用相位的概念来解释。研究还表明，“that-t”效应的相理论分析也可以解释它们的跨语言变化。

Within the framework of generative grammar, which explicates the structure and function of the language faculty in the mind/brain of human beings, there have been various attempts to derive the effects of the computational operations proposed in the previous framework from computationally efficient operations. This study focuses on the notion of phase, which is one of the core concepts to reduce the computational complexity within the minimalist program, the most recent framework of generative grammar, and pursues a conceptually and empirically plausible definition of the notion of phase. Specifically, this study investigates the wh-island and that-t effects. It is widely accepted in generative literature that the wh-island effects are subsumed under the Minimal Link Condition. Based on the hitherto unnoticed fact that specific wh-phrases are allowed to be extracted out of indirect questions in English, this study argues that the wh-island effects cannot be subsumed under the Minimal Link Condition, but we still need their traditional subjacency-type account based on the notion of phase. It also argues that the that-t effects can be accounted for by the notion of phase. It is also shown that the phase-theoretic analysis of the that-t effects also account for their crosslinguistic variations.