新加坡华语句末语气词 ：在语料库基础上与普通话的对比统计 = Statistical comparison of Singaporean and Mainland Mandarin Chinese : a corpus-based study on sentence final particles

• Main Author: 白欣彦 Pak, Xin Yan
• Other Authors: School of Humanities and Social Sciences
• Format: Final Year Project
• Language: Chinese
• Published: 2014
• Subjects: DRNTU::Humanities
• Online Access: http://hdl.handle.net/10356/55632
• Institution: Nanyang Technological University

关于普通话语气系统的研究已有不少(王力 1985 等),但新加坡华语作 为汉语变体之一,学术界目前还未对其语气系统进行深入且系统的探讨。此外, 前人关于句末语气词的研究大都着重于其所表达的语气功能,而对其他可能影响 语气词的选择和使用的因素(如语体类型等)分析不多。因此,本文在大型语料 库的基础上,对华语和普通话中的句末语气词的选择和使用情况从多个因素进行 统计分析,并讨论句末语气词在华语和普通话中的异同。另外,本文也探讨华语 中特有的句末语气词(咩,咯,hor),从而提供一个较为全面的华语句末语气词研究。本文从 Gigiword 汉语语料库中的新华网语料库和新加坡联合早报语料库 中,找到普通话和华语中使用频率高的五个句末语气词(呢、吗、吧、啊、 啦),并为每个词随机抽取 100 条例句做进一步统计分析。本文在现有语料库信 息基础上,对每条例句进行标注,观察六个可能影响句末语气词的使用的因素: 语用功能、搭配的语气副词、语体、否定词、句型和句末语气词的省略。本文使 用统计软件 R 中的 polytomous 软件包(Arppe 2012,cf. Han et al. 2013, Appre 2008, Lau 2013) 对这些标注信息进行分析。统计结果表明: 1. 本文提出的六个因素能有效区分五个句末语气词在单一汉语变体的差异 2. 本文提出的五个句末语气词在两个汉语变体中的选择和使用相似, 但也存在差异,其主要差异为语气副词的搭配、句末语气词出现的 语体和句末语气词与陈述句的关系。本文希望此研究除了可以补充现有与语气词相关研究的不足之外,也可以 介绍使用统计软件 R 来探讨普通话和汉语变体(华语)之间的差异的研究方法。Modal particle is one feature of Mandarin Chinese (Wang 1985; among others). Desipte the various studies on modal particles in Mainland Mandarin Chinese (MMC), the studies on Singapore Mandarin Chinese (SMC) are few, let alone comparative studies of MMC and SMC. Furthermore, previous studies focused mainly on the functions of modal particles whereas the factors influencing their uses (such as syntactic category, genre) were seldom investigated. Hence, this study aims to use a corpus- driven statistical approach to analyze and compare the uses of modal particles in SMC and MMC, particularly sentence final particles (SFPs). This study also investigates SFPs that are unique to SMC (meh, lor, and hor) to provide a more comprehensive study on SMC SFPs.As observed from Chinese Gigaword Corpus 2005, the five most frequently used SFPs in both SMC and MMC are ne, ma, ba, ah and la. For each SFP, samples of 100 lines containing the SFP were taken for MMC and SMC respectively. Each line was then coded for a total of six variables regarding the use of the SFP in it, including pragmatic functions of the particle (Zhang 2010), modal adverbs (Zhang 2010), genre, use of negation, sentence type and whether or not the SFP can be omitted. The statistical analysis of each variable was obtained using functions available in R and the polytomous package (Arrpe 2012; cf. Han et al. 2013; Appre 2008). The statistical results suggest that some variables serve as a differentiating factor among the five SFPs (either in SMC or in MMC). Furthermore, the distribution of the five SFPs in SMC and MMC is similar, the main difference being the correlation with the adverb variable, genre variable and the relation with declarative sentences. In short, this paper seeks to discuss the differences between modal particles in SMC and MMC and simultaneously, introduce how the functions in R and the polytomous package can be used in comparative studies.