| This dissertation explores the notion of subject in Japanese conversation, taking a discourse-functional approach. The most serious problem with earlier works on this controversial topic is that claims have been made entirely based upon constructed sentences. This dissertation contributes to the study of the question of a 'subject' in Japanese by offering a quantitative analysis based upon naturally occurring, everyday conversational data. This study reveals how different the invented data is from naturally occurring conversational data. The dissertation explores two specific issues: (1) does argument realization in Japanese support the existence of a 'subject'?; and (2) does the particle ga mark 'subjects'?.;This study proposes three main findings. First, making use of what is known as the most dependable syntactic tests to identify the subject in Japanese, two specific syntactic tests, reflexivization and honorification, were applied. The findings are that neither syntactic test provides strong evidence for a syntactic category that one can legitimately call a 'subject' in spoken Japanese. Second, applying the concept of 'privileged syntactic argument' (Van Valin and La Polla 1997), both unrealized PSAs and realized PSAs were examined. Concerning the unrealized PSAs, two possible interpretations are possible and resolving questions here will remain as most crucial in the exploration of Japanese grammatical relations. Regarding the realized PSAs, it is argued that they are in fact pragmatically motivated rather than syntactically controlled. Also, Japanese speakers seem to try not to express [+ human] PSAs as much as possible in their daily conversations. A close examination of realized PSAs does not seem to support the claim that there is a syntactic subject in spoken Japanese. Third, based on a distributional study of the particle ga, findings are that ga is not a grammatical marker of a 'subject'. It is, rather, a pragmatic marker and marks a referent as referential/specific. The particle ga can also function as an 'explicit contrastive focus' marker within certain pragmatic environments. These findings are supported by evidence not only from the analysis of 6255 predicates in adult conversation but also from an examination of children's conversational data. Based upon these findings, it is difficult to find decisive evidence for the existence of a grammatical subject in spoken Japanese. |
