Abstract
In this paper, we demonstrated how the particle ka in Japanese evolved its disjunctive and indeterminate uses. Ka was originally used only as a question particle until Early Middle Japanese. The disjunctive use is first found in the sentence final position in Kakuichibon Heike Monogatari and, after the 15th century, largely attested in NP-positions. Our paper shows that indeterminate use of ka emerges in the adverbial position before the appearance of ka in NP-positions. Descriptive studies on the history of the relevant particle have been concerned with the period when a particular use emerged. This study, however, shows through the analysis of highly verifiable data in which examples are classified by their morpho-syntactic patterns, that a particular use does not appear abruptly but is formed gradually. Based on these facts, we argue that the history of ka in NP-positions, including indirect-question use, is explained by the features [±N] and [±wh].
