The Syntax Of Classifiers Within The Framework Of Phase Theory

This paper tries to conduct a study on classifiers in the framework of phase theory. Based on Abney’s DP-Analysis, classifiers have been well recognized as heads in nominal phrases in previous syntax studies, which is also supported by the studies on the semantics. Svenonius (2003) assumes that there are at least two phases in noun phases: DP and nP, corresponding to CP and vP in clauses. Svenonius (2003) proposes a more complex model of phases in nominal phrases as in Op—[Q—K—Num*—[n—N. In a similar vein, Cai (2009) proposes D as a head instead of Q as in Op*—[D—K—Num*—[n—N. Based on their assumptions, I propose a different model, which is along the line D*[Q—K—F*—[n—N.Though the previous scholars have examined the syntax and semantics of classifiers thoroughly, they never study within the framework of phase theory. Cai (2009) starts the analysis with phase theory, assuming a parameter on nP shell between classifier languages and non-classifier languages. However, his study is conducted via classified discussions, which violates the principle of minimalist and raises many uninterpretable linguistic phenomena. To make the research more convincing in a unified way, I assume that QP and nP are two phases in nominal phrases, that the projections of nominal phrases are divided into [N+Num+Cl] sequence and [Num+Cl+N] sequence, and that n is a nominal attractor, attracting the noun (N1) in [N+Num+Cl] sequence and the classifier (N2) in [Num+Cl+N] sequence. This assumption is more tolerating. It simplifies the parameter setting, and can analyze most languages in their derivations. Based on my proposal, I work out the derivations of phases in nominal phrases in classifier languages with the data of some different languages and dialects such as Mandarin, Cantonese, English, Japanese and Thai. In my analysis, I focus on NP movement which well explains for [Num+Cl+N] form in nominal phrases and hold that the original sequence for NP is [N+Num+Cl]. In this way, we get an overview of the derivation of phases in DP structure in classifier languages. Thus, I get a conclusion that C1is nominal at essence and that it belongs to nP, lying at the bottom of the tree diagram.