A Corpus-based Lexical Semantic Study of Mandarin Verbs of “Tui” and “La”

A Corpus-based Lexical Semantic Study of Mandarin Verbs of “Tui” and “La” Han SUN Chu-Ren HUANG Faculties of Humanities / Room GH801, 8/F, Faculties of Humanities / Room GH801, 8/F, Wing GH, The Hong Kong Polytechnic Wing GH, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong University, Hung Hom, Kowloon, Hong Kong 20092425g@connect.polyu.hk churen.huang@polyu.edu.hk Abstract This paper investigates the differences between Tui and La under the framework of MARVS. Using Gigaword2 Corpus and Sinica Corpus as corpora and Sketch Engine as the query tool, the study is a conducted following three procedures. First, preliminary observation on morphosyntactic differences in these two verbs is obtained by searching the explanation of the two terms Tui and La in the online Chinese English Dictionary of Modern Usage. Then, two data corpora are used in searching for statistical evidence, after which the results will be represented in the framework of MARVS. In conclusion, Tui is inherently a single occurrence that cannot be measured by time duration and can be identified with a temporal point, whereas La is an activity that has a time course that can be measured based on duration and has a clear boundary as the starting point. Also, distinctions in role module lie in that La indicates abstract process without clear direction or an endpoint while Tui is more concrete with one direction and an endpoint. 1 Introduction Verbs provides the key to study the nature of lexical knowledge and sentence process (Liu et al., 2000). It is believed that the “syntactic behaviors of verbs are semantically determined”(Chief et al., 2000, p. 57), so it is an effective way to explore sematic differences between verbs by comparing the syntactic differences of synonym verbs (Liu et al., 2000). And since a more semantically constrained system is needed for natural language processing purpose, Module-Attribute Representation of Verbal Semantics (MARVS) theory has been shown as a new, powerful approach to represent semantic representations of near synonyms (Liu et al., 2000). In the previous literature, Liu et al. (2000) explored semantic differences of pivotal near-synonymous members of “tou”, “zhi”, “diu”, and “reng” by using two kinds of endpoint and the observed distinctions are then incorporated into a representational paradigm MARVS. Wang and Huang (2018) examined the stance differences of “guli” and “mianli” in terms of implication of interpersonal relation variations and accounted for the results the in MARVS. However, little studies have been conducted on the near synonym “tui” and “la”. Thus, the focus of this study is to establish semantic-syntactic interdependences by observing the morphosyntactic behaviors of the verbs “tui” and “la” displayed in a large corpus and representing the results in the framework of MARVS. 2 Research questions This paper investigates the following two questions: (1). What are the differences between Tui and La? (2). How are these differences represented in the framework of MARVS?. 3 Theoretical Framework The Module-Attribute Representation of Verbal Semantics (MARVS) theory (Huang et al., 2000; Huang & Hsieh, 2015) is taken as the theoretical framework. It …

Content truncated to 500 words