修飾語認可及漢語定冠詞詞組結構：以屬性檢驗為架構的分析

Modifier Licensing and Chinese DP: A Feature Analysis

在北京話這類沒有顯形定冠詞和不定冠詞的漢語中，數詞前的修飾語會造成一個名詞組得被詮釋為殊指，數詞後的修飾語則不會有相同的效應。這些及其他有關修飾語分布和指涉方面的特性得以妥善說明，如果能將表示無指和定指／殊指這兩種不同的屬性分置於D 和F 這兩種不同的中心語。D 和F 的屬性設定將認可不同類型的修飾語，而且相關設定將在LF 結構經由相關成分的隱形移位來檢驗。這樣的分析方式除了可以說明漢語修飾語及指示詞、數詞、量詞、名詞之間的共存限制，也可以說明北京話和廣東話名詞組之間的指涉異同。這樣的分析還同時對漢語有下列三個意涵：(A) 只有如名詞類的代表體物的實詞及如D 和F 類的純粹表達名詞指涉的虛詞中心語能認可特定類型的修飾語，(B) 無指名詞組的功能範疇投射比定指／殊指名詞組的功能範疇投射還要更高一個層面，以及 (C) 空號D 中心語的認可條件比空號F 中心語的還要更嚴格。

In Chinese languages like Mandarin, without overt determiners and articles, the presence of a pre-numeral, not post-numeral, modifier may obligatorily result in the specific reading of the noun phrase. These and other distributional and referential properties of Mandarin modifiers may be accounted for under an assumption that the features associated with non-specific indefiniteness and definiteness/specificity may be split into two distinct functional heads D and F, each of which may license different kinds of modifiers and may be checked at LF by an operation of covert movement of nominal expressions marked with the relevant feature. An account along this line of thought may not only capture the co-occurrence restrictions between various types of Chinese modifiers and demonstratives, numerals, classifiers, nouns; it may also explain the referential distinctions between Mandarin and Cantonese noun phrases. Three implications may be found with such an analysis: in Chinese, (a) only the lexical category N, which denotes the entity, and functional heads like D and F, which may be associated strictly with the reference interpretation of the entity denoted by N, may license modifiers of certain kinds; (b) the indefinite non-specific non-bare noun phrase may project to a higher functional projection than the definite/specific non-bare noun phrase; and (c) the licensing of an empty D may be more restrictive than the licensing of an empty F.