| Summary: | 碩士 === 國立臺灣師範大學 === 英語學系 === 106 === The thesis aims to investigate the bi-comparatives in Mandarin. The main focus is on the debate between the Direct Analysis (DA) and the Reduction Analysis (RA) in terms of the analysis of the underlying structure of the post-bi constituent.
On the one hand, the Direct Analysis claims that the underlying structure of the post-bi constituent is just what we see on the surface. In short, DA does not assume the operation of comparative deletion in a bi-comparative, which means that under DA, there is only one token of the predicate of comparison in the underlying structure. However, it is found that DA fails to account for certain cases of bi-comparatives which need two tokens of the comparison predicate in the underlying structure, such as bi-comparatives with preposed objects, bei long passives, ba disposal constructions, and verb copying constructions.
On the other hand, the Reduction Analysis argues that the post-bi constituent involves a structure richer than its surface structure. In other words, RA assumes the operation of comparative deletion, which elides the predicate of comparison inside the post-bi constituents. However, as pointed out in Xiang (2003), RA encounters three problems: the lack of subcomparatives, the lack of embedded standards, and the distribution of dou in bi- comparatives. Although Erlewine (2017) and Hsieh (2017) offer some solutions to the three problems in their proposal, this thesis finds that their analyses are not without problem.
Besides DA and RA, Liu (2011) proposes a Hybrid Analysis, which advocates that there are two types of bi-comparatives: phrasal and clausal bi-comparatives. Some criteria are provided in Liu (2011) to distinguish between these two types of bi-comparatives, but this thesis argues that the criteria for the distinction need some modifications.
A revised Hybrid Analysis is proposed in this thesis, and there are two major arguments in the current proposal. First, the distinction between the two types of bi-comparatives depends on whether the post-bi constituent(s) form a complete constituent or not. Second, the lack of embedded standards and the distribution of dou in bi-comparatives are accounted for by means of the No Embedding Constraint for bi-comparatives and syntactic haplology respectively under the revised Hybrid Analysis.
|
|---|
