Segmentation And Representation Of Translocative Motion Events In English And Chinese Discourse: A Contrastive Study

This dissertation is a contrastive study of the segmentation and representation of English and Chinese translocative Motion events (TME), which encode at least two location-changes and possess the Macro-event Property (MEP). In accordance with the MEP defined in Bohnemeyer et al. (2007), we construct a parallel corpus with a total of 201 English and 224 Chinese sentences expressing TMEs. Based on this we conduct a series of statistical analyses to explore the similarities and differences between English and Chinese TMEs with the MEP.Adopting the model proposed in Bohnemeyer et al. (ibid.) and the conceptual structure theory put forth in Jackendoff (1983), and accepting that a TME conceptually consists of three subevents, namely the departure event (DE), the passing event (PE) and the arrival event (AE), we first align the English and Chinese constructions with their translations. The alignment reveals that about 70% the TME constructions are structured similarly in the source and target language. Moreover, in both English and Chinese, conceptualizations of TMEs are mapped mostly onto the DE + AE pattern and the PE + AE pattern. Patterns such as DE + PE + AE and DE + PE are also found, but of only a very limited amount.After the comparison of the segmentation of the English and Chinese TMEs with the MEP, we then examine the conceptual structures of the subevents. We find that English and Chinese users conceptualize DEs and PEs alike, but differ in construing AEs. DEs in both languages cluster around the constituent combination of <P_AS><SP>. Besides this, the English DEs have another important conceptual representation, namely <MO><P_AS><SP>. PEs in both languages are conceptualized as <P_AR><PaP>. AEs are conceptualized differently by the users of the two languages. For English users, AEs are conceptually <P_AG><EP> or <MO><P_AG><EP> in structure, whereas for Chinese users, AEs are <P_AG><MO><EP> or <MO><P_AG><MO><EP> in conceptual structure.Thirdly, based on the lexicalization patterns proposed in Talmy (1985, 2000b: Chap. 1), we explore the conceptual identities of the four TME constituents, namely, Actant, Motion, Path and Ground. It is found that about 70% of the Actants are animate in English and Chinese TMEs, and the inanimate ones are mainly from the fictive TMEs. 40% of the English DEs and AEs are found encoding Motion, and for the English PEs it is 60%. Motion-encoding in Chinese TME subevents shows an ascending tendency of DE < PE < AE, i.e., while Chinese DEs rarely encode Motion, nearly every AE requires Motion-encoding. English and Chinese users conceptualize Paths similarly: both favor <P_AS> over <P_AF>, <P_AG> over <P_AT>. The four types of Grounds, namely, starting point, passing point, endpoint and goal, are encoded in nominals in both languages.Finally, we examine the claim that conceptually English and Chinese TMEs have much in common and they are realized by more or less the same surface linguistic forms. Our findings in representation investigation partially corroborate this claim. For example, the nominal conceptualization of Actants and Grounds is realized by NP representation in the majority of the cases, and Paths are usually represented by PPs and Pvs in both languages.However, conceptualization and representation are by nature two distinct things. By ranking the overall representation types, our investigations reveal that English and Chinese TMEs are rarely represented by the same surface forms. <P_AS>PP<SP>NP (DE) + <P_AG><EP><MO>Pv (AE) is the only one we found in our corpus. In terms of overall representation, of the three subevents, only PE has similar core structure in the two languages.Our data analysis also shows that, although English and Chinese users generally segment TMEs similarly, there are still 30% of the TMEs that are segmented otherwise in the target language. A closer look at these cases reveals the following causes for the difference: 1) the reduction of the subevents, 2) the loss of the MEP, and 3) the change from dynamic to static.Furthermore, Motion in English TMEs is typically represented by one single Motion verb, which may be a Pv, an Mv or any other kind, whereas in Chinese SVCs are frequently utilized to convey TMEs, and sometimes to convey even one subevent in a TME. Such an orientation in Chinese has a chain effect upon the subsequent representations of Paths and Grounds, particularly AE Paths and Grounds. As English only permits one Motion verb in the representation of TMEs, it usually has to resort to other devices to represent Chinese verbal Paths or Grounds when they are translated into English, especially to use PPs and ADVs to represent Chinese verbal Paths and to use ADVs in addition to NPs to represent Chinese verbal Grounds. The reverse is also true in English-Chinese translation. Such asymmetry in the representation of TMEs in English and Chinese can be attributed to the differences between the two languages in Motion-encoding, iconic motivation, and lexicalization.The major theoretical implications of this study are twofold: 1) typological classification of languages on the basis of one or two criteria is not error-free, and 2) by aligning the English and Chinese TME representations, we find that, in both Chinese and English, the introduction of coordination in fictive constructions does not always lead to the loss of the MEP, as claimed by Bohnemeyer et al. (2007, 2010).