Relative clauses in Archaic Chinese
上古汉语关系从句研究

Abstract 摘要

As is well known, all relative clauses in Modern Chinese are marked with de. Compared with Modern Chinese, Archaic Chinese seems to have a richer encoding for relative clauses. It has three possible relative markers: zhe, zhi and suo. In this article, we will approach to Archaic Chinese relatives from the perspective of formal syntax. We will argue that zhe is a subject relative pronoun and suo a VP internal relative pronoun, and that zhe– and suo– relatives are derived from the A’-movement of the relevant relative pronoun to the peripheral position, creating an operator-variable relation at LF. Simultaneously, we will use Cecchetto and Donati’s labeling theory to account for a peculiar phenomenon that the same relative pronoun can form both headless and headed relatives. We will further argue that the relative marker zhi, as English that,is an invariant relativizer occupying the C position, and that zhi-relatives involve null operator movement. That is to say, like English type of languages, Archaic Chinese actually involves two separate relativizing strategies: through a relative pronoun or a null operator. The article also contributes to the debate on the analysis of relatives: data from Archaic Chinese favors an Adjunction Analysis to the Raising Analysis (or Complementation Analysis) à la Kayne.

众所周知,现代汉语中所有的关系从句均由“的”引导。相比之下,上古汉语中关系从句的表现形式则更为丰富,共有“者”“之”“所”三种关系化标记。本文将从形式句法的视角探讨上古汉语的上述三种关系从句。我们认为“者”是主语关系代词,而“所”是VP内关系代词,带“者”或“所”的关系从句是通过关系代词A’移动至边缘位置而生成的,并借此在LF层建立了算子–变量关系。同时,我们将采用Cecchetto和Donati的加标理论来解释上古汉语中同一个关系代词既可构成无核关系从句,又可构成有核关系从句这一奇特的语法现象。至于关系化标记“之”,我们认为它与英语中的that相似,是标句词C,而“之”引导的关系从句则是通过空算子移位生成的。也就是说,上古汉语和英语类语言一样,都具有两种独立的关系化策略,即要么通过关系代词实现关系化,要么通过空算子实现关系化。另外,此文还将对目前学界关于关系从句分析方法的争论起到一定的参考作用:上古汉语的语料证明了嫁接分析法要优于Kayne的提升(或称补语)分析法。

Share on facebook
Facebook
Share on google
Google+
Share on twitter
Twitter
zh_CNZH-CN