Japanese Subjects and Information Structure : A Constraint-based Approach

Japanese Subjects and Information Structure: A Constraint-based Approach Akira OHTANI Faculty of Informatics, Osaka Gakuin University 2–36–1, Kishibe–Minami, Suita, Osaka 564-8511, Japan ohtani@utc.osaka-gu.ac.jp Yuji MATSUMOTO Graduate School of Information Science Nara Institute of Science and Technology 8916–5, Takayama-cho, Ikoma, Nara 630-0192, Japan matsu@is.naist.jp Abstract This paper is concerned with how topic/focus articulation should be optimally integrated into Japanese grammar. Based on Engdahl and Vallduv´ı’s (1996) Information Structure, we pro- pose an analysis with the following characteristics: (i) information structure is an integral part of Japanese grammar and interacts in principled ways with both syntax and phonology, (ii) the representations of topic/focus in the information structure and its interactions with the par- ticles wa/ga show one-to-many relation, and (iii) the ordering of grammatical functions and its interactions with other grammatical parts play an important role in determining the focus domain. 1 Introduction Japanese is a language in which topic and focus are identified by the use of the particles. In the case of subjects, they are either marked by wa or ga. The two particles are illustrated in the following examples: (1) a. Ken-wa Ken-TOP kaet-ta. leave-PAST b. Ken-ga Ken-NOM kaet-ta. leave-PAST ‘Ken left.’ ‘Ken left.’ The fact that (1a) and (1b) are given the same semantic interpretation, does not tell us what wa and ga really mean. These particles are said to have two uses respectively. Kuno (1973),1 for instance, distinguishes the use of wa by referring to (2a) as a thematic, and (2b) as a contrastive: (2) a. ‘As for Ken, he left.’ b. ‘As opposed to other people, Ken left.’ Much syntax-based work has been done on the use of these particles but the basic distinction between them is the same. Whatever the correct characterization of the uses of wa shown in (2) may be, the meaning of (1a) that is common to both uses is that the proposition that somebody left is true if that somebody is Ken. Whether a subject marked by wa is uttered in one use of wa or the other cannot be determined without knowledge of the context. In this paper, we demonstrate that the relation between syntactic constituency and contextual infor- mation is not a tight one-to-one mapping as purely syntax-based analysis assumes. Based on Vallduv´ı and Endahl’s Information Structure (Vallduv´ı, 1992; Engdahl and Vallduv´ı, 1996; Vallduv´ı and Engdahl, 1996), we propose a constraint-based one-to-many mapping mechanism which captures some aspects of topic/focus interpretation. The study of information structure, we argue, is essential in addressing fundamental questions regarding the multi-dimensional grammar for Japanese. 1Kuroda (1965; 1972; 1990) recognizes the same distinction but characterizes it differently. PACLIC 18, December 8th-10th, 2004, Waseda University, Tokyo - 93 - 2 The data: Distribution of Topic/Focus Before going into the detailed discussion, we first summarize how topic and focus are identified through- out this paper since their definitions vary considerably among linguists. 2.1 Test: Erteschik-Shir (1997) Erteshik-Shir studies the interface between discourse structure and syntax in Erteschik-Shir (1997) and defines a grammatical level of focus structure …

Content truncated to 500 words