International Journal of Chinese Linguistics - Volume 3, Issue 1, 2016

Tone sandhi in Wuxi Chinese involves “pattern substitution,” whereby the base tone on the first syllable is first substituted by another tone, then spread to the sandhi domain. We conducted a wug test to investigate native Wuxi speakers’ tacit knowledge of tone sandhi and found that the substituion aspect of the sandhi is not fully productive, but the extension aspect is, and sandhi productivity is influenced by the phonetic similarity between base and sandhi tones. These results are discussed in the context of how phonological opacity, phonetic naturalness, and lexical frequency influence phonological learning, and a grammatical learning model that can predict Wuxi speakers’ experimental behavior is proposed.

This paper examines the derivation of two types of A′-dependencies — relative clauses and Left-Dislocation structures — in the framework of Minimalist Program based on Mandarin data. Relatives and LD structures demonstrate many distinct syntactic and semantic properties when they contain a gap and a resumptive pronoun respectively. A thorough study of the relevant data reveals that when a gap strategy is adopted, island effects and crossover effects are always observed, irrespective of whether the relevant gap is embedded within a relative clause or within an LD structure; on the contrary, when the resumptive strategy is adopted, a sharp distinction is observed between these two structures. A resumptive relative clause gives rise to island effects and crossover effects systematically; by contrast, a resumptive LD structure never gives rise to these effects. In the Minimalist Program, island effects and crossover effects are not exclusively used as diagnostic tests for movement since the operation Agree is also subject to locality constraints. I will argue that a relative clause containing either a gap or an RP and an LD structure with gap are derived by Agree and they are subject to the locality condition whereas a resumptive LD structure is derived by Match that is an island free operation and it is not subject to the locality constraint. Multiple Transfer and multiple Spell-Out are possible in an Agree chain, but not in a Matching chain. The choice of the derivational mechanism depends on the interpretability of the formal features attached to the Probe and to the Goal in the relevant A′-dependencies.

This paper develops the empirical and theoretical basis for the necessity in admitting the operation of verb-stranding ellipsis (VVPE) in Chinese. I present new arguments showing that, though two analytic possibilities — null argument analysis and VVPE analysis — are in principle available in the grammar of Chinese, they can be differentiated in specific syntactic environments. In particular, I show that the existing null argument approaches would have difficulty in accounting for the following facts: disjunction of multi-constituent elements under negation, the difference of island effects in the presence of a linguistic antecedent, the verb identity requirement and the possibility of having part of the idiomatic expression as the missing gap. Therefore, it is argued that VVPE must remain a viable syntactic operation in Chinese when a null object analysis is unavailable.

The derivational differences of the fragment answers in Mandarin Chinese lie in whether a fragment moves or not. Under the movement and ellipsis analysis (Merchant 2004), fragment answer to wh-question moves to SpecFocP, followed by TP ellipsis. In contrast, fragment answer to yes-no question or for correction is a base-generated structure, [pro copula fragment]. The analysis is supported not only by the existence of the copular verb and the fragment answers to questions involving the passive constructions and preposition stranding but also by cross-linguistic evidence.

The ka-construction in Taiwan Southern Min is well-known for its polysemy. This study argues that with evidence from extra-argumentality, ka-NP position, thematic relationship, passivization, over-generalization, and dialectal difference, ka should at least be categorized into three different types, i.e., light verb, applicative, and preposition, instead of a uniform one. The first type involves the disposal patient/theme and the goal/source constructions where ka is a light verb taking a reduced VP complement. The second one applies to the benefactive/adversative construction where ka is a high applicative head mediating the relation between the ka-NP and a VoiceP. The third one accounts for the non-gapped dative construction where ka is a preposition heading a PP that modifies a VP. By so doing, the polysemy is simply an interplay between different types of ka and their corresponding syntactic structures.

As an empirical study, linguistic investigations rely heavily on the selection of suitable subject speakers as representatives of a language community, which in turn underline the necessity of determining representativeness. Through explicit description of the parameters that define a target population, this paper argues that (i) representativeness can be most straightforwardly captured using the same parametric description; and (ii) the degree of representativeness of the subject increases with the degree of satisfaction of the parameters. On this footing, this paper exemplifies the fundamental quality of the parameters of representativeness across different linguistic research topics, purposes and even methodologies. Evidently, the clarification on what counts as representative through the use of parametric description is a founding stone to ensuring reliability of language data in the establishment of any linguistic theorizing.