Some unresolved problems in Cantonese grammar

Abstract

Despite growing sophistication in studies of the grammar of Cantonese, some rather basic issues remain poorly understood. Two such problems are addressed in this talk. One basic problem involves the description and analysis of 'bare classifier' Noun Phrases taking the form [Classifier Noun]. In Cantonese such phrases allow definite interpretations as in (1): (1) 間 屋 太 貴 gaan2 uk1 taai3 gwai3 'The house is too expensive.' At least three approaches are prominent: (i) Cantonese allows omission of a demonstrative, allowing the 'bare classifier' structure to be assimilated to a general or 'universal' Chinese grammar (Chao 1968); (ii) classifiers function as definite articles (Chao 1949) or as D (Cheng & Sybesma 1999, 2012); (iii) classifiers themselves may have referential properties as in Miao (Bisang 1993, 1999), a property which may have entered Cantonese through substrate influence (Matthews 2006). These proposals will be reviewed critically. A second problem involves the status of complementizers. Two recurrent proposals are that (i) items such as Cantonese 話 waa6 are complementizers (Chui 1994, Yeung 2006); (ii) Sentence-final particles such as 喎 wo5 occupy the COMP position (Law 1990, Li & Sybesma 2007). These proposals are in conflict inasmuch as they imply left-headed and right-headed CPs coexisting in the same language, and even in the same clause in cases such as (2): (2) 佢 同 我 講 話 冇 錢 喎 keoi5 tung4 ngo5 gong2 waa6 mou5 cin2 wo5 'He told me had had no money' From the perspective of grammaticalization, 話 waa6 has been treated as a semi-complementizer (Chappell 2008). How is such a semi-complementizer to be treated synchronically? Evidence mentioned by Chappell argues against the possibility that such elements form a CP with the following clause, calling for an alternative analysis.