An Optimality Theoretic Approach to Formal Semantics and Pragmatics

这个为期三年的项目开始讨论OT中最大的问题之一--可选性。针对英语、法语、东京日语和关西日语中互补词的删除，得出中和方法上级其他方法的结论。由于主要关注的是V提升和补足语缺失之间的关系，因此我遵循了Otani和惠特曼（1991）提出的日语V到I运动。但我也证明了，即使语言中没有V到I的运动，中性化方法仍然比其他方法更好。在第二年，我注意到了不可表达性，这是OT中的另一个重大理论问题。我讨论的现象是英语和意大利语中的多个wh-疑问句，这是语法-语义界面中的一个问题。本文从双向优选论和单向优选论的句法和语义角度出发，指出了Legendre等（1998）提出的PARSE（wh）分析方法在经验和概念上存在的问题，并提出了一种替代分析方法。Chierchia（1995）认为语义学产生了驴代词的强和弱解读，而偏好是由语用学决定的。这种方法可以在OT中捕获。生成两个阅读作为候选，通过约束排序选择最优输出。作为一种语用约束，我提出了一种语用约束，即“努力最小化”，即“为完成输入输出句所描述的目的而付出必要和充分的努力”。因此，本研究在优选论的框架下考察了一个句法问题、一个句法-语义界面问题和一个语义-语用问题。然而，我无法讨论一个语用学上恰当的现象。这将是未来的研究课题。

This three-year project started to discuss the optionality, one of the biggest issues in OT. Focusing on the complementizer deletion in English, French, Tokyo Japanese, and Kansai Japanese, it was concluded that the neutralization approach was superior to other approaches. Since the main concern was the relation between V raising and complementizer deletion, I followed the V-to-I movement in Japanese proposed by Otani and Whitman 1991. But I also showed that even without the V-to-I movement in the language, still the neutralization approach is better than anything else.In the second year, I paid attention to the ineffability, another big theoretical issue in OT. The phenomenon I discussed was multiple wh-questions in English and Italian, which is a problem in the syntax-semantics interface. From the perspective of bidirectional OT as well as monodirectional OT syntax and semantics, I pointed out the empirical and conceptual problems of the analysis with PARSE(wh) proposed by Legendre et al. 1998, and proposed an alternative analysis.In the final year, I discussed a case of semantics-pragmatics interface; the interpretations of pronouns in donkey sentences. Chierchia 1995 argues that semantics generates both the strong and the weak readings of donkey pronouns and the preference is determined by pragmatics. This approach can be captured in OT. The two reading are generated as candidates, the optimal output is chosen by the constraints ranking. As a pragmatic constraint, I proposed a pragmatic constraint, called Effort Minimization, which says, "Make a necessary and sufficient effort to accomplish the purpose described by the in out sentence".This study, thus, investigated a syntactic problem, a problem of the syntax-semantics interface, and a problem of the semantics-pragmatics in the framework of OT. I was not able to discuss a pragmatics proper phenomenon, however. This will be a future research topic.