The structure of the Japanese inferential system
This paper provides an analysis of Japanese inferential auxiliaries (<i>daroo</i>, <i>rashii</i>, <i>soo</i>-<i>da</i>, <i>yooda</i>) that enables us to explain all occurrences of these words in discourse. Despite extensive research on these forms, grammatical analysis has not yet successfully explained their distribution. This paper demonstrates that the prevailing view of these forms as <i>evidentials </i>is insufficient. Instead, we postulate a new hypothesis in which the speaker’s choice of auxiliary is based on his or her desire either to strongly present inferential information or to draw attention away from it. In this analysis, the language user is seen as more active in the interpretation of messages than is traditionally assumed.