The Number + Classifier in Mandarin [Number + Classifier + de + NP] can have a property reading—describing the property of the NP, or a quantity reading— expressing the quantity of the entity. The two mainly differ in: (a) de required in the property construction but optional in the quantity construction, and (b) de licensing NP-deletion in property, not quantity constructions. These differences are due to their structures and the derivation of de’s. The [Number + Classifier] expressions in the property construction are just like other prenominal relative clauses modifying the NP [[Number + Classifier] relative clause + de + NP]. De is basegenerated as a head that can license its complement (the modified NP) to be null. In the quantity reading, the structure should be [Number + [Classifier + NP]], where Classifier is a head taking the NP as its complement. De is phonologically inserted —a way to reflect focus-encoding on the quantity via the strategy of phonological phrasing. As focus can also be encoded by stress or pause, de-insertion does not always apply (=(a)). When the NP is missing, proper phonological phrasing is achieved without de. That is, the apparent failure of de licensing NP-deletion in the quantity construction is actually due to the non-application of de-insertion (=(b)). As de is not inserted phonologically when the NP is null, NP-deletion should apply before de-insertion. This ordering follows straightforwardly from the approaches that derive ellipsis structures via deletion-in-syntax or base-generation of empty categories. Support for these claims comes from the corresponding data in Taiwanese, whose tonal behavior reveals a clearer picture.

P-Insertion and Ellipsis

Despite the more than half-century study by numerous linguists on the properties of the attributive marker de and the structure of the nominal phrase [XP de YP] in Mandarin Chinese, a consensus has yet been reached. This work provides a new perspective to the issues through the investigation of the properties of the corresponding marker e in Southern Min-Taiwanese. The empirical support is built on the structural factors determining the occurrence of e and its tone values in different contexts. The study helps provide answers to (i) how the seemingly conflicting facts in constituency structures [XP e YP] can be accommodated: some facts indicate that e forms a unit with the preceding XP but some others suggest that it forms a constituent with the following YP; and (ii) what principles determine the tone changes of e. The core of the answers is the observation that Southern Min-Taiwanese allows the sequence e 0 -e 5 , indicating that the structure of [XP e YP] actually can be [[XP e 0 ][e 5 YP]]. Correspondingly, the structure of the Mandarin nominal phrase [XP de YP] can also be the result of reduction from two de’s, thereby resolving the debate on whether Mandarin [XP de YP] should be [[XP de] YP] or [XP [de YP]].

de in Mandarin ↔ e in Taiwanese

This paper examines the interpretational properties and distributional patterns of adjectival modification with de in Mandarin, aiming to revive the adjunction approach to adjectival modification, against the alternative approach which takes adjectives as uniformly occupying the Specifier positions of different functional projections. Specifically, unlike previous studies which primarily focus on the distributional patterns of a single [Adj+de] sequence within a nominal expression, this paper mainly investigates collocated [Adj+de] sequences, and reveals that they not only show the same distributional patterns as a single [Adj+de] sequence, but may also give rise to two different interpretations, an observation which has gone unnoticed in the literature. Based on these observations, it is proposed that collocated [Adj+de] sequences are best analyzed on a par with a single [Adj+de] sequence as AP adjuncts, which can target three different positions, including a Determiner Phrase (DP), a Numeral Phrase (NumP) and a Noun Phrase (NP). As for the two readings available to collocated [Adj+de] sequences, they are proposed to be derived from different structures.

Adjectival Modification in Mandarin: Evidence for the Adjunction Analysis

This paper addresses some relevant explanations concerning de and de constructions in Chinese. Based on a detailed description of the main features of de, a novel analysis of the syntactic status of de and the internal structure of de constructions is proposed. It is argued that de, as a functional category, cannot be separated from the genitive constituent, and hence it cannot function as a head. In effect, the nature of de constructions is determined by the definite article, numerals or quantifiers and has nothing to do with de per se. De is a linking marker, which is used to connect the constituents preceding and following it, including the constituents with features of [+N] and [-V]. De can adjoin to the Spec or complement. As a consequence, in terms of structural segmentation, de should be segmented either with the Spec or with the complement, but it is only a clitic of the Spec or complement. Since de is not the head of de constructions, it cannot project as DeP.

/pdf/the-syntactic-status-of-de-and-the-internal-structure-of-de-1b838xecod.pdf

The Syntactic Status of de and the Internal Structure of de Constructions in Chinese

The focus of this paper is on the syntactic status of de in Chinese nominal phrases under the generative frameworkIt is argued that being a functional category,de(的)is not necessarily treated as a head unless there is independent supporting evidence from semanticsIt is also argued that the modifiers to which de is attached should be analyzed as adjuncts rather than specifiers,according to the criteria listed in Duffield(1999)The discussion in this paper hopefully can resolve some of the debates on the status of de in the literature

Tang Sze-win

Papers

On de as a head in Chinese