2011 Volume 7 Issue 4 Pages 83-98
This work aims to verify within Ryukyuan languages and dialects the validity of the authors' kakari musubi hypothesis (including reconstruction of Proto-Japonic forms, typologies, origins, and detailed historical changes), heretofore built up through comparisons and analysis within the Okinawan and Japanese languages. In the |ga|-Type kakari musubi hypothesis that this work focuses on, we distinguished three usage subtypes, Type I (agreement of sentence-internal |ga| and the so-called mizenkei), Type II (agreement of sentence-internal |ga| and the more conventional adnominal form), and the sentence-particle use of the kakari particle. The test shows that the Type I and Type II patterns and functions in Kushi dialect of Kunigami district of Okinawa accord with the hypothesis, and that its pseudo-sentence-final usage is interpretable as a grammaticalization from the Type II pattern. In the dialects Inokawa (Tokunoshima island), Benoki and Kanna (Kunigami), and Hatoma (Yaeyama in the Sakishima group), we have presented our views concerning the voiced / voiceless allophones of |ga| in its sentence-particle usage. We have proposed a heretofore unnoticed sound change in Torishima dialect for a usage in which the construction appears to be Type II but the function Type I, and thus argued that these data do not disprove our hypothesis. In addition, with regard to a discrepancy of form and function in Nishizato dialect of Miyako, we attempted a defense of our hypothesis. Finally, we disagreed with the generally accepted explanation that Nakijin |kuse:| is the cognate of Japanese koso, and laid out our hypothesis that |ga| played a part in its emergence.
