| Abstract: |
In this paper I discuss the historical change of the Japanese question particle ka and argue that its development goes in the opposite direction to the one assumed under the view named ‘subjectification’. ka has been used as a direct question marker since Old Japanese, but it evolved an indirect question use in Middle Japanese. This change is characterized as a loss of speaker-oriented meanings since direct questions are more speaker-oriented than indirect questions, as can be shown by scope relations. The loss of speaker-orientedness can also be observed in the development of indirect question use of ka. In its early stage, ka entails the speaker's uncertainty, inherited from direct questions. However, it does not exhibit that uncertainty in its later stage, used in contexts where the speaker knows the answer of the embedded question. Since speaker-orientedness is a defining property of ‘subjectivity’ and the changes exhibited by ka are considered to be a natural process of language change, those changes of ka constitute a significant piece of counter-evidence to the hypothesis of ‘subjectification’. |