现代汉语的对比性特征——基于词性和句法关系清单的分析

Distinguishing Features of Mandarin Chinese as Against English in POS and Syntactic Inventories

本研究基于英汉语平衡语料库,采用数据驱动的描写方法观察词性和句法关系的宏观分布特征,并为汉英语言研究提出量化分析思路。研究发现,名词和动词是英汉语句子扩展采用的主要词性手段,相应地,定中关系和状中关系是英汉语句子扩展采用的主要句法手段。对比分析显示,英汉语句子扩展手段存在差异:英语更依赖介词系统,故更多呈现静态关系;汉语更多调用动词来编码相应的语义关系,成就了汉语的“动态”特征。此外,汉语句法成分间依存距离短于英语,主要是因为汉语句段铺排更多依赖时序,而英语多使用后置修饰成分。研究表明,数据驱动的研究不仅能检验我们的语言直觉,还能为微观语言对比和分析提供参照。有描写性数据的支持,汉语研究和中外语言对比研究有望拓展新的疆域。

The study adopts a data-driven approach to the description of the macroscopic linguistic features characteristic of Chinese and English, and the data subject to the description comes from the comparable corpora (LCMC and FLOB/FROWN). The study assumes that the statistical analysis could contribute to a better understanding of the major features of Chinese and it may open up quantitative approaches to contrastive linguistic study. The study finds that, in English and Chinese, nouns and verbs play the most active roles in sentence expansion, and correspondingly, modifier-noun relations and adverbial-verb relations play equally active roles. The difference between Chinese and English is also obvious: English tends to use more prepositions, which are used to express static relations;compared with English, Chinese tends to resort to verbs encode what is expressed by prepositional phrases in English, which contributes to the dynamic feature typical of Chinese. The study also finds that the dependency distance of Chinese is shorter than that of English, for the latter is prone to using post-positioning elements. Owing to such difference, in E-C translation, the translated Chinese, if it is in line with the norm of Mandarin usage, will resort more to verbs than to modifier-head construction in sentence expansion, as can be seen in the use of adverbial-head, subject-predicate and predicate-object constructions. In addition, the fact that different translations encode the same information indicates there is flexibility in language use, which is reflected in its ability of using various linguistic devices to transfer the same or similar meaning. Considering that, the trade-off among the distributions of elements in language inventories is a common practice in language use. The study concludes that data-driven approach to contrastive analysis could not only testify to our intuition about a language in its general sense, but also contribute empirical evidence for microscopic analyses. And under the support of data from linguistic inventories comparison, the boundary for language studies is expected to be expanded.