Volume 25, Issue 3-4

This article re-examines the controversial shì … de construction, which is shown to involve different types of structures, with different syntactic properties. The core of the analysis proposed in this article is twofold: (a) shì is a copula, which selects a small clause (with a subject and a predicate), and (b) de marks the presence of two different non-overt operators (a generalized lamba-operator, and an assertion operator). It is argued that the focus reading connected with shì is simply related to its copular/verbal property. Furthermore, Mandarin allows an in-situ focus strategy using phonological prominence; this strategy interacts with shì and its postverbal constituent.

This article deals with the syntax and semantics of intensifiers in Mandarin Chinese. Intensifiers are expressions that are functionally and distributionally comparable to non-argumental x- self in English (e.g., the king himself , or do the work oneself ). Mandarin examples are zìjĭ, bĕnshēn, qīnzì , or qīinkŏu . The descriptive goal is to identify the paradigm of adnominal and agentive-adverbial intensifiers in Mandarin and to determine the individual conditions of their use. On the theoretical side, I apply the analysis of intensifiers as lexicalizations of the identity function first proposed for German to the Chinese case. The identity function takes referential DPs as arguments, or, in the agentive-adverbial use, the agentive Voice head. The identity function in itself is semantically inert, but if the intensifier is focused, alternatives to the identity function come into play. These alternatives are used to explain the contextualization patterns attested for intensifier sentences. Two competing approaches to Mandarin intensifiers are reviewed and found not to be sufficient to account for the data.

This article argues that the distribution of the wh/q -polarity adverb daodi in Chinese cannot be given an analysis in which it moves at LF. It shows that daodi is subject to a structural and locality constraint similar to that for the focus marker shi . They both have the scope-restricting function on the A-not-A predicate or the wh -phrase it is associated with. The long-distance and island effects which daodi apparently exhibits are not due to its being moved, but can be explained as cases of parentheticals or violations of the locality condition on focus or of selectional restriction.

In preverbal position in Mandarin Chinese the morpheme gen sometimes presents strict distributional characteristics, sometimes not. In the first case, it can be either a preposition, equivalent to the comitative marker ‘with’ in English or a conjunction, equivalent to ‘and’. In the second case, it evidences a mixed distribution, partaking of both preposition and conjunctive features. In this paper, gen is treated as an underspecified marker whose values can be predicted from its syntactic behavior and from the types of predicates it is in construction with. It is shown that when it presents mixed distributional characteristics, gen can only be in construction with collective predicates. The semantic features associated with such predicates, which are [+/–eventive], are [+symmetry], hence [–distributivity].

The term “construction” is not a label to be assigned randomly, but presupposes a structural analysis with an associated set of syntactic and semantic properties. Based on this premise, the term “serial verb construction” (SVC) as currently used in Chinese linguistics will be shown to simply refer to any multi-verb surface string i.e,. to subsume different constructions. The synchronic consequence of this situation is that SVCs in Chinese linguistics are not commensurate with SVCs in, e.g., Niger-Congo languages, whence the futility at this stage to search for a “serialization parameter” deriving the differences between so-called “serializing” and “non-serializing” languages. On the diachronic side, SVCs are invoked as a privileged site for verb-to-preposition reanalysis, but it is left open what structure is referred to under this label. A precise structural analysis of both the input and the output structure is, however, indispensable in order to make meaningful statements about language change.

The shi … de construction in Mandarin Chinese is a cover term for at least four distinct constructions. We focus on the remarkable internal properties of the shi … de pattern proper, where both shi and de are present. We concur with a long tradition of scholarship identifying shi … de as a cleft construction, but we show it to have very different properties from it – or pseudo-clefts in languages like English. The crucial distinguishing property of shi … de clefts is that they involve no A′ movement. This is linked to the fact that the presupposition in the shi … de focus cleft involves a projection smaller than CP. It thus lacks a landing site for A′ movement and fails to host material such as negation, normally compatible with full CP clefts in languages such as English. The Mandarin pattern shows that the basic semantic property of clefts, bipartioning into focus and presupposition, is logically independent from A′ movement.

One notable difference between the Chinese nominal phrase and the Zhuang nominal phrase is that in Chinese the demonstrative appears phrase-initially while in Zhuang it appears phrase-finally. In this article, we attempt to derive the different orders of nominal elements in Chinese and Zhuang using the same base order: with the demonstrative generated low. We adopt a base structure in which two D-related projections are involved: DetP, which immediately dominates NP and SP (Specificity Phrase), which dominates NumeP. DetP is where the demonstrative is generated. Differences between Chinese and Zhuang arise from parametric choices such as the head or phrasal status of the demonstrative, what moves (the demonstrative or the NP) and the motivation for movement. We show that a base structure with a low generation of the demonstrative is not only supported by cross-linguistic evidence, it also derives the different orders of nominal elements in a straightforward manner.

This article deals with an indicative-modal asymmetry in Chinese, where a variety of specificity effects are found in fronted object indefinites. To provide a coherent solution under the cartographic approach (Rizzi 1997, Cinque 1999), we propose that there are two types of landing sites for Chinese object fronting: The outer focus position is in the left periphery, and occupied by a specific nominal. By contrast, the inner focus position is located in the peripheral area around v P (a clause-internal focus in Belletti's 2004 terms), where a bare NP is interpreted as specific in realis sentences, but non-specific in irrealis sentences. It is argued that this object specificity follows from a dynamic mechanism of syntax-semantics mapping encoded in the Extended Mapping Hypothesis, and should be treated on a par with Chinese subject specificity. Along this line, we are able to lay out the “topography” of Chinese foci based on the inner-outer dichotomy, and provide a solution to the indicative-modal asymmetry through the interaction between syntactic predication and aspectual quantification.