The Influence of Givenness and Heaviness on OSV in Japanese

The Influence of Givenness and Heaviness on OSV in Japanese Satoshi Imamura Japanese Studies, University of Oxford 41 Wellington Square, Oxford, OX1 2JF Abstract Several studies contend that the main motivation for scrambling is heaviness. In particular, Yamashita (2002) maintains that scrambling has nothing to do with givenness and that heaviness is the primary factor for scrambling. However, her conclusions count on only 19 examples and she does not distinguish VP-internal scrambling from VP-external scrambling. Thus, it is conceivable that some types of scrambling rely on givenness. In order to see if this hypothesis is on the right track, I conducted a corpus analysis of OSV order in Japanese, largely based on the quantitative approach. Consequently, it has been revealed that both givenness and heaviness have a high explanatory power for the usage of OSV order. Furthermore, there was no correlation between givenness and heaviness, showing their independent influence on OSV order. Therefore, I conclude that both givenness and heaviness are sufficient to trigger OSV order and the phenomenon cannot be fully accounted for except with reference to both. Furthemore, based on the mapping between information structure and syntactic structure, I propose that VP-external scrambling is discourse-driven while VP-internal scrambling is not. 1. Introduction A natural language may have many kinds of options for expressing the same proposition. In Japanese, for example, the meaning of a canonical transitive sentence SOV can be expressed by a scrambled sentence OSV in which the object appears before the subject. Why do languages have many options to convey the same proposition? One explanation is that these options allow speakers to choose the way information is transmitted. They differ not in what is said about the world, but in the way it is packaged (Chafe, 1976; Lambrecht, 1996; Vallduvi and Engdahl, 1996). In other words, their differences derive from information structure, i.e. how the meaning of a sentence is conveyed. Specifically, it has long been recognized since the work of the Prague School that speakers prefer to put given information before new information. However, this description begs the question because givenness itself is not a clear-cut concept. Therefore, it is necessary to define givenness in an objective way. In this paper, givenness is defined by a quantitative approach (Givōn, 1983) and regarded as discourse-old information i.e. information mentioned in the preceding discourse. In other words, previously mentioned constituents are considered to be given information. Another explanation for variable ordering of arguments is based on heaviness. Hawkins (1994) observed that long constituents tend to be put in earlier positions than shorter ones in Japanese in order to facilitate the processing cost of heavy constituents. In this study, I am going to investigate the usages of OSV PACLIC 28 224 Japanese Studies, University of Oxford 41 Wellington Square Oxford, OX1 2JF United Kingdom Copyright 2014 by Satoshi Imamura 28th Pacific Asia Conference on Language, Information and Computation pages 224–233 order in terms of givenness and heaviness, mainly based on quantitative data from a Japanese corpus. This paper …

Content truncated to 500 words