Model Theoretic Semantics, Proof Theoretic Semantics, and Semantically Flavored Syntactic Features

语法研究的基本目标之一是解释语言中形式和意义的关系。语义学是对意义的研究。语义考虑至少在三个方面发挥作用。首先，语法从句子组成部分的意义中计算出句子的意义，这些组成部分通常是参照模型化情境中的真值来定义的。例如，“No whale flies”为真，当且仅当在建模的情况下，鲸鱼集合和会飞的事物集合的交集为空。第二，句法结构的可接受性可能取决于具有语义风味的形态句法特征。例如，在任何情况下，鲸鱼飞是不可接受的，而在某些情况下，鲸鱼飞是不可接受的，对应于前置短语的否定与非否定特征。这些特征在句法中具有普遍性和理论上的重要性。第三，说话人根据语义知识进行各种推理。例如，没有鲸鱼飞需要没有蓝鲸飞和没有鲸鱼飞高。通常认为，一旦为所有表达式指定了组合模型理论语义，语法和推理就可以自由地享受其成果。但特别是计算语言学家和逻辑学家认为这是不可行的，并开发了证明理论方法。最近的工作进一步表明，证明理论的考虑可能不仅是计算上的优势，但也可能是从“纯”理论语言学的角度启发。 美国国家科学基金会将支持在马萨诸塞州剑桥的2005年美国语言学会暑期研究所举办一个题为“模型理论语义学、证明理论语义学、语义风味句法特征”的讲习班。它的中心目标是探索如何使用证明理论作为模型理论语义和生成语法之间的中介，可以导致在语法/语义接口理论上有趣的见解。从一个证明理论的教程开始，它将通过汇集语义学家，句法学家和理论倾向的计算语言学家，也可能是心理语言学家，来研究语义的这三种用途之间的关系来实现这一目标。该研讨会的广泛影响来自：（i）促进“纯”理论语言学和计算语言学之间的互动，（ii）促进国际合作，（iii）向参加研究所的大量研究生和本科生免费开放，以及（iv）作为纽约大学安娜·萨博尔奇博士和加州大学洛杉矶分校爱德华·斯塔布勒博士研讨会的基础。期刊出版物和专门用于研讨会的网站将有助于在学术界内外传播成果。

One of the fundamental goals of grammatical research is to explain how form and meaning relate in language. Semantics is the study of meaning. Semantic considerations come into play in at least three ways. First, grammar calculates the meaning of a sentence from the meanings of its component parts, defined typically with reference to truth in a modeled situation. For example, No whale flies is true if and only if the intersection of the sets of whales and things that fly is empty in the modeled situation. Second, the acceptability of a syntactic construction may depend on morpho-syntactic features with a semantic flavor. For example, Under no circumstances would a whale fly is acceptable, whereas Under some circumstances would a whale fly is not, corresponding to the negative vs. non-negative feature of the preposed phrase. Such features play a pervasive and theoretically prominent role in syntax. Third, speakers make various inferences based on semantic knowledge. For example, No whale flies entails No blue whale flies and No whale flies high. It is usually assumed that once a compositional model theoretic semantics is specified for all expressions, its fruits can be freely enjoyed by syntax and inferencing. But especially computational linguists and logicians have argued that this is not feasible, and have developed proof theoretical methods. Recent work has furthermore indicated that proof theoretical considerations may not only be computationally advantageous but may also be enlightening from the perspective of "pure" theoretical linguistics. The National Science Foundation will support a workshop, entitled "Model theoretic semantics, proof theoretic semantics, semantically-flavored syntactic features," to be held at the 2005 Linguistic Society of America Summer Institute in Cambridge, Massachusetts. Its central goal is to explore how the use of proof theory as a mediator between model theoretic semantics and generative syntax can lead to theoretically interesting insights at the syntax/semantics interface. Starting with a proof theoretic tutorial, it will achieve this goal by bringing together semanticists, syntacticians, and theoretically inclined computational linguists, possibly also psycholinguists, to investigate the relation between these three uses of semantics. The broader impacts of the workshop derive from (i) fostering the interaction between "pure" theoretical linguistics and computational linguistics, (ii) promoting international collaboration, (iii) being freely accessible to the large numbers of graduate and undergraduate students attending the Institute, and (iv) serving as a basis for seminars by Dr. Anna Szabolcsi at New York University and Dr. Edward Stabler at UCLA. Journal publications and a website dedicated to the workshop will help dissemination of the results inside and outside academia.