A MARVS Analysis of Two Chinese Near-Synonymous Verbs of Jumping Based on Chinese Corpora

A MARVS Analysis of Two Chinese Near-Synonymous Verbs of Jumping Based on Chinese Corpora Yingying Peng Chinese and bilingual studies Faculty of Humanities The Hong Kong Polytechnic University 20113471r@polyu.edu.hk Abstract The Module-Attribute Representation of Ver- bal Semantics (MARVS) has been served as an excellent tool for the analysis and differentiation of Chinese near synonymous verbs. In the current study, two Chinese near- synonymous verbs 蹦 (b`eng; to jump) and 跳 (ti`ao; to jump) which have not been investi- gated in previous studies are analyzed within the framework of MARVS with their occur- rence data and observed distributional differ- ences obtained from two large Chinese cor- pora, GigaWord2 and the CCL corpus. The re- sults show that compared to b`eng, ti`ao is much more flexible in terms of directionality and lo- cations, which could ultimately be accounted for with the fact that ti`ao emphasizes more on the displacement of the subject caused by the action. 1 Introduction Within a semantic domain, words that can be used in a similar fashion to describe similar events are con- sidered as members of the same class of words and further named as near-synonyms(Gao et al., 2016). Near-synonyms are notoriously difficult to be differ- entiated from each other with explicit explanations even for native speakers. It thus presents great chal- lenges for second languages learners and also for the designation of cross-linguistic dictionaries and do- mains such as machine translation(Hong, 2014). In view of the significance of distinguishing near synonyms especially near synonymous verbs and in- sufficiency or even inadequacy of previous theoreti- cal frameworks (e.g., (Levin, 1993)) designated for the analysis of English to be extended for the differ- entiation of Chinese near synonyms, Huang and col- leagues (Huang et al., 2000) proposed the represen- tational model called the Module-Attribute Repre- sentation of Verbal Semantics (MARVS). The model has been proven to be more adequate in analyz- ing a broad span of Chinese near-synonymous verbs including pi2juan4 (tired) and lei4 (tired), ge1 (to slice) and qie3 (to cut)(Tsai et al., 1998), and ben1 (to run) and pao3 (to run)(Wang, 2009). Neverthe- less, beng4 (to jump) and tiao4 (to jump) as two Chinese near-synonymous verbs denoting the mean- ing of jumping with frequent usage have never been thoroughly studied. Thus, the current study aims to adopt the MARVS framework to analyze b`eng (to jump) and ti`ao (to jump) and distinguish them based on their distri- butional differences obtained from two large Chi- nese Corpora, namely Gigaword2(Hong and Huang, 2006) and CCL(Zhan et al., 2019) . 2 Literature Review This session firstly introduces the basics of MARVS. Then insights gained from previous studies on near- synonymous verbs that adopted this model are elab- orated in order to lay out a clearer picture of how the current study would be conducted. Then, there will be a brief introduction of these two Chinese verbs followed by the research questions the current study attempts to answer. 2.1 The Module-Attribute Representation of Verbal Semantics (MARVS) In order to better capture …

Content truncated to 500 words