语法性与语用性——汉语名词短语作有定解读的实质

Grammaticality and Pragmaticality: On the Nature of Definitely Interpreted NPs in Sinitic Languages

本文秉承近年来兴起的新描写主义研究范式之精神(胡建华2006, 2009, 2010, 2016, 2017;李汝亚2008;隋娜、胡建华2016, 2017等),通过对比英语中[the+N]有定短语与汉语中可作有定解读的种种名词短语(如光杆名词、光杆量词、[量+名]短语、[量+量+名]短语、[数+量+名]短语和[指+量+名]短语等),重点探讨语言学界有关量词的两极认识即"实词说"与"冠词说"的根源所在。笔者指出:(1)由于定冠词the是语法化了的有定标记,即"有定性"是其内在的语法属性,[the+N]短语无论是脱离语境还是身处句中不同位置,皆能作有定理解,本质上是"语法有定表达式";(2)汉语的光杆名词、光杆量词和[量+名]结构等脱离语境时与"有定性"没有任何关系,既不能作有定解读也不能作无定解读,而在一定的语用条件下既可作有定解读又可作无定解读,因此它们作有定解读时本质上是"语用有定表达式";(3)论元位置上的各种名词短语可统一分析为DP,无需移位能够占据D位置的是定冠词、指示词、量化词(如"所有""每")等,而其他成分如光杆名词、光杆量词或重叠量词等,则需要进行中心语移位后才能占据D位置。笔者进一步指出,语言学研究要严格区分语法性与语用性:语法性主要是指语言成分的固有属性,是与语境等外在因素无关的恒常表现;相反,语用性则是语言成分在具体语境里的随机表现。国内外语言学界将汉语量词等同于英语定冠词的分析,是将语用现象误作语法现象的结果。

Under the research paradigm of New Descriptivism （ see Hu 2006, 2009, 2010, 2016, 2017; Li 2008; Sui and Hu 2016, 2017 inter alia）, the present paper addresses the nature of a variety of definitely interpreted NPs in Sinitie languages, with aparticular focus on the issue of why there exists a gulf between traditional grammarians and contemporary linguists with regard to the grammatical status of numeral classifiers. Having carefully examined the relevant data, I contend that the analysis of equating Chinese numeral classifiers with English definite article is the consequence of mistaking an essentially pragmatic phenomenon for a grammatical one. Specifically, it is argued that （ 1 ） ＂definiteness＂ is the defining property of English definite article but not Chinese numeral classifiers, so [ the＋N] expressions can denote definiteness out of context and can receive a definite reading irrespective of their syntactic position in a sentence, whereas bare nouns, bare classifiers and [ CI＋N] phrases have nothing to do with ＂（in） definiteness＂ out of context, and if placed in a proper context, can be interpreted as definite or indefinite; （ 2 ） the analysis of equating Chinese numeral classifiers with English definite article suggests that Chinese should possess hundreds （ or dozens if only sortal classifiers are counted） of definite articles which differ from one another phonologically, semantically and morphologically. In the forseeable future, this would never happen to a typical classifier language like Chinese; （ 3 ） in terms of syntactic analysis, definitely interpreted NPs of various kinds can be analyzed as DP where the position of D should be occupied by definite articles, demonstratives, pronominals, strong quantifiers and so on, whereas bare nominal, bare classifier, and numeral should move to D for a definite interpretation. Finally, it is proposed that definitely interpreted NPs should be plausibly classified into two types, namely, ＂grammatically determined definite expressions＂ （which are beaded by elements with the inherent property of ＇ definiteness＇ such as definite articles and demonstratives） and ＂pragmatically determined definite expressions＂ （which, albeit lacking an element with the inherent property of ＇ definiteness＇, can receive a definite readinR via a pragmatic context）.