Collocation analysis can be used to extract meaningful linguistic information from large-scale corpus data. This paper reviews the methodological issues one may encounter when performing collocation analysis for discourse studies on Chinese. We propose four crucial aspects to consider in such analyses: (i) the definition of collocates according to various parameters; (ii) the choice of analysis and association measures; (iii) the definition of the search span; and (iv) the selection of corpora for analysis. To illustrate how these aspects can be addressed when applying a Chinese collocation analysis, we conducted a case study of two Chinese causal connectives: yushi ‘that is why’ and yin’er ‘as a result’. The distinctive collocation analysis shows how these two connectives differ in volitionality, an important dimension of discourse relations. The study also demonstrates that collocation analysis, as an explorative approach based on large-scale data, can provide valuable converging evidence for corpus-based studies that have been conducted with laborious manual analysis on limited datasets.
The exact nature and derivation of patient-subject constructions (PSC) in Chinese are still at dispute in literature. Based on the restriction of manner adverbial modification and the nonexistence of the manner reading of zenme ‘how’ observed in Chinese PSC, a morphosyntactic analysis has been provided. We argue that the seeming action verb V in PSC is not a real main verb, but a verbal root to be introduced into the derivation after syntax via external morphological merger. The real main verb of PSC in syntax is a covert light verb ∅BEC, which selects a nominal phrase (NP) as its specifier (Spec) and a resultative phrase (RP) as its complement. BECP is further selected by an aspect (Asp) head le. To satisfy the extended projection principle (EPP), the NP at [Spec, BECP] moves to the [Spec, TP] in syntax. After syntax, the resultative (R) head-moves to ∅BEC at the phonological form (PF) to satisfy the phonological requirement of ∅BEC, forming R-∅BEC; then, a bare verbal root merges with R-∅BEC at PF to denote the manner of the change of state. Due to the phonological requirement of le, V-R-∅BEC head-moves to le, producing the right order of PSC. The two elided forms of PSC can be derived similarly. This research suggests that covert light verbs and morphology may play an interactive role in the derivation of some “typical” constructions in Chinese.
This article proposes a pro analysis for split questions (SQs) in Chinese, dissimilar to the biclausal account employing focus movement and deletion in Arregi 2010 and the one employing the silent head in Kayne 2015 and Tang 2015. SQ consists of a wh-clause and a tag clause. We argue that the entire SQ is an information/confirmation-seeking question, represented by a Speech Act Phrase (SAP)-shell structure (Speas and Tenny 2003; Oguro 2017, etc.) with wh-clause in its specifier and the tag in its complement. The tag of Chinese SQ is a base-generated clause, [pro (copula) tag ma/ne], composed of an empty subject pro, an optional copula, a tag, and a final particle, instead of being derived from a fully-fledged structure parallel to the wh-part akin to those of English and Spanish SQs. Such a pro analysis overcomes difficulties encountered in the other accounts regarding the distribution of the final particles and their clause-typing, the optionality of the copula, the ubiquitous uses of tag, the connectivity effects, and the island-insensitivity. Analytically, two seeming variants of SQ imply that the derivation of an SQ depends on whether its tag moves and whether a copula exists.
This paper analyzes verb reduplication in Mandarin Chinese under a lexicalist framework. By adopting the Lexicalist Hypothesis proposed by Chomsky (1970), a distinction has been made between syntactic and morphological verb reduplications by means of five tests: productivity, le insertion, categorial stability, transitivity, and input/output constraints. It is found that the AA and ABAB patterns of verb reduplication have relatively high productivity and regular syntactic behaviors, whereas the AABB pattern of verb reduplication shows extremely low productivity and syntactic idiosyncrasy. Given these observations, this paper proposes that the AA and ABAB patterns should be syntactic verb reduplications derived at the syntactic level, whereas the AABB pattern should be morphological verb reduplication formed in the lexicon. The two types of verb reduplications have different generative mechanisms.
Huang (2015) characterizes “Modern Chinese as a language of high analyticity at multiple levels” and demonstrates “a ranking of relative analyticity among the three dialects: Cantonese > Mandarin > TSM”. This paper argues that Teochew (cháoshànhuà, 潮汕話), another variety of Min, different from TSM, shows more synthetic performances than Mandarin. Chomsky’s “productivity” criterion (1970) helps distinguish lexical operations from syntactic ones. In this spirit, this paper will illustrate its arguments from two perspectives -- lexical and syntactic operations. When it comes to lexical operations, analyses on both the semantic changes within the same categories and the categorial shifts will be made. Besides, syntactic discussions on emphatic inflection, bare classifier phrases, verb-object order and other variants of V-movements in Teochew will also be demonstrated. All analyses will be put under the theoretical framework of generative grammar with the help of a cartography approach. For analyses at the lexical layer, this paper adopts Si’s 司富珍 (2012, 2017a, 2017b, 2018) XW structure, trying to capture the synthesis performances of the Teochew lexicon. As for syntactic operations, the split-CP hypothesis of Rizzi (1997, 2001, 2004) and Rizzi and Bocci (2015), the CL-to-D hypothesis of Simpson (2005), the light verb approach of Chomsky (1995) and the split-light verb hypothesis of Si 司富珍 (2018) will be used as references. Through comparative studies with Mandarin, Cantonese and other languages like English, this paper will conclude that Teochew is a dialect with higher synthesis compared with Mandarin.
In Mandarin Chinese, the string of three overt elements in a row, a locative, a verb, and a nominal, asserts the existence of the entity denoted by the nominal in the location. This paper argues that the verb is contained in an adjunct, whereas the locative in its base position and the nominal establish a matrix predication relation. Thus, instead of the overt verb, the head of the matrix predicate of the construction is null. Moreover, a new analysis is provided to explain the obligatory argument sharing between the verb and the matrix predication of the construction. Furthermore, the paper argues that the agent of a transitive verb in certain types of embedded clauses needs to be Case-licensed by either the v of the selecting verb, as in an ECM construction, or a local c-commanding functional element, such as a complementizer, as in the English infinitive for construction. This Case-licensing explains why the transitive verb in the string has no agent. The research shows that the syntactic strategies to license abstract Cases in Chinese are similar to the ones found in other languages. Finally, the paper argues that the post-verbal -zhe is an adessive marker when it occurs in a non-progressive context.
In the literature about processing of relative clauses (RCs), subject relatives (SRs) are reported to be easier than object relatives (ORs) in a number of languages, but the status of prenominal ORs in languages where the object follows the verb (SVO) is still partly controversial. This study explores the production of RCs in Cantonese in two elicited production experiments and two corpus studies. In the first elicited experiment, an overwhelming preference for SRs was observed. In two corpus studies where the context and the feature of arguments were uncontrolled, the reverse pattern was observed. In order to reconcile the two datasets, we speculate that what counts in object dependencies is the featural endowment of the subject, as in the intervention hypothesis implemented in Friedmann et al. 2009. A second elicited experiment was run to test this hypothesis. The results suggest that production of RCs in Cantonese displays a subject preference in general and that object dispreference is modulated by featural mismatch.
In this paper, I first introduce what inalienable possession structure (IPS) is cross-linguistically as well as how to form an IPS in Mandarin Chinese, i.e., pronoun + body part or kinship term, etc. With the help of postverbal IPS, I relate the lack of plural pronominal possessor in IPS, which is never discussed in the literature, to the prohibition of distributivity over distributivity, i.e., the semantic anomaly of distributive plural possessor over the stubborn distributivity inherent to Chinese IPS nouns. I also argue that the requirement of a plural pronominal possessor seen in the IPS of public places, spatial directions, and professional titles is a result of stubborn collectivity shared by these nouns. In the end, I discuss the association between the distinction of inalienable and alienable nouns and that of active and stative verbs.
One controversy in the study of the Chinese shenme ‘what’-based rhetorical question (shenme-RQ for short) is how it takes on a negative interpretation. This paper attempts to apply enthymeme or rhetorical syllogism to the deduction of negative meaning of the shenme-RQ. Triggered by the shenme-RQ, or one of its words or phrases, the hearer extracts the explicit premise, fills in the premise that is implicit either in the context or in her or his encyclopedic knowledge, and deduces the conclusion, the negative meaning of the shenme-RQ. According to what premises are left out, the paper also explores the deduction patterns of the negative meaning of shenme-RQs and proposes a procedure for obtaining the negative interpretation. That said, the negative meaning of the shenme-RQ will be entrenched in the mind of its users and conventionalized in the Mandarin Chinese community via repeated use.