定中结构的两分和“的”的语义功能

Two types of attributive nominal phrases and the semantic function of de from the perspective of semantic type theory

围绕"的"的有无,汉语的定语一直有区分"限制性/描写性"(或"区别性/描写性"等术语)的说法,但是各种定语两分分析都存在界定模糊、操作性差的缺点,也不能完全解释"的"字隐现的规律。本文采用语义类型理论,尝试在语义层面对有"的"和无"的"的定中结构的语义差别给出明确清晰的区分。"的"的隐现区分了两种语义类型:不用"的"的定中结构是e类型语义成分,特指名称(专有名称或种类名称);使用"的"的定中结构是<e,t>语义类型,指称具有特定属性的个体所组成的集合。"的"的语义功能是一个<e,t>语义类型转换算子。本文的分析可以广泛解释"的"字隐现的难题,证明了语义类型理论的有效性,也有助于观察相关问题的跨语言类型学差异。

There is a tradition of dividing Chinese attributives into two parts: restrictive vs. descriptive attributives.Such a distinction, however, is vague and causes contradiction to the nominal system. On the basis of semantic type theory, this paper proposes that the presence and absence of de is semantically driven: attributive nominal phrases with de are always of type <e, t>, denoting a set with members sharing the same properties, while those without de are of type e, with a reference to a kind name or a proper name. It is claimed that de is an unselective type shifter to type <e, t>. The analysis in this paper gives a unified explanation to the complicate behavior of de in Chinese nominal phrases and demonstrates the usefulness of semantic type theory. It also helps explain the typological differences between Chinese and other languages.