On the So-Called Thematic Use of Wa: Reconsideration and Reconciliation

On the So-Called Thematic Use of Wa: Reconsideration and Reconciliation⋆ David Y. Oshima International Student Center, Ibaraki University, 2-1-1 Bunkyo, Mito 310-8512, Japan oshima@mx.ibaraki.ac.jp Abstract. In past studies, there have been two major lines of analyses of the func- tion/meaning of the particle wa in Japanese in its thematic use. Some scholars argue that it marks a topic, while others claim that it marks ground (old, presupposed, or backgrounded information). I demonstrate, testing their predictions against empirical data, that neither the topichood-based approach nor the groundhood-based approach constitutes a complete theory of wa in its thematic use, and argue that the two approaches need to be combined in an ade- quate fashion. Namely, I propose that wa indicates topichood only when it is associated with certain grammatical functions, such as direct object, while it merely indicates groundhood when it is associated with others, including subject. Keywords: Japanese discourse, information structure, topic, focus 1 Introduction This paper develops a semantic analysis of (some major functions of) the particle wa in Japanese, which integrates and reconciles two major lines of analyses in the existing literature: the topichood-based approach and the groundhood-based approach. There have been a great deal of studies of information packaging strategies in Japanese, in- cluding and especially the use of the putative ‘topic-marker’ wa (see Noda 1996, Fry 2003, and Heycock 2008 for literature surveys). Most previous studies of wa start from the assumption that, although wa is often (and indeed, most frequently, according to Fry’s (2003) corpus-based survey) attached to a subject, wa is not a subject-marker. This assumption is based on the fact that wa can be attached to constituents other than subjects, such as (direct or indirect) objects, (temporal or locative) modifiers, and even modifiers of nominal constituents; the sentences in (1) are adapted from Noda (1996:2), and the one in (2) from Mikami (1963:13).1 (1) a. Kodomo-tachi-wa child-Pl-wa karˆe-o curry-Acc tsukutte-i-mas-u. ‘The kids are making curry.’make-Asp-Polite-Pres (wa occurring on a subject) ⋆ Part of the present paper was presented at the Semantics Research Group meeting held at Hirosaki Gakuin University on October 26th, 2007 and the Ibaraki University Linguistics Research Group meeting held at Ibaraki University on May 30th, 2008. I would like to thank the audiences for valuable feedback. I benefited greatly from the insight- ful comments of Makoto Kanazawa, Christopher Tancredi, Ikumi Imani, Kiyomi Kusumoto, Fumio Fujii, Kunio Nishiyama, Takuya Goro, and Takayasu Namiki. Copyright 2009 by David Y. Oshima 1 Throughout the paper, the English translations of Japanese examples are meant to be rough approximations, rather than propositional and information-structural equivalents. The abbreviations used in glosses are: Acc = accusative, Asp = aspect, Aux = auxiliary, Ben = benefactive, Cl = classifier, Dat = dative, Gen = genitive, Hon = honorific, Imp = imperative, Nom = nominative, Pl = plural, Pres = present, Pro = pronoun, Q = question marker, SFP = sentence final particle. 405 23rd Pacific Asia Conference on Language, Information and Computation, pages 405–414 b. Karˆe-wa curry-wa kodomo-tachi-ga child-Pl-Nom …

Content truncated to 500 words