抄録
This paper studies the functions of reversed word order phrases in Japanese. Six transcribed TV talk shows are used for the data base.
Based on the assumption that reversed word order in Japanese could be viewed as a speaker's rectification of mistaken ellipsis, I first test the eligibility for ellipsis of every reversed referent by examining the speaker/hearer's cognitive status in the discourse. The study suggests that 75.16% of reversed referents are ineligible for ellipsis in the discourse while 24.84% are eligible.
Then the speaker's purpose of using reversed word order is examined for both eligible and ineligible referents for ellipsis from the pragmatic point of view. The study indicates that: 1) 38.51% of reversed constructions could be viewed as cases where the speaker mistakenly/carelessly drops some element and the element in postverbal position rectifies the ambiguity caused by his/her mistaken ellipsis, 2) 36.64% of reversed word order could be viewed as cases where the speaker asserts first a pragmatically marked part, 3) 24.85% are instances where the speaker adds the referent in postverbal position in order to confirm the referent or get the hearer's attention to the referent.
By adopting reversed word order, the speaker helps the hearer understand the discourse, relate the information to the hearer's knowledge network, and confirm the important points in the discourse. This study also shows that the speaker participates in conversations by assessing the hearer's cognitive situation.
