The aim of this study is to examine the reasons why Man'yōgana 氣 represented two borrowed sounds /kë/ and /gë/ in Old Japanese literature. In Old Japanese Japan had no its own writing system which caused it to use Chinese characters to express its language. Therefore, Japanese had to depend on Chinese characters, one of which was 氣. 氣 was semi-voiceless in Ancient Chinese, which means voiceless and aspirated stop. However, Japanese used it not only as an unaspirated stop but as a voiced one. Preceding studies have shown that the voiced usage of 氣 was general phenomenon in Man'yōshū. But I thought there must have been certain reason or reasons different from ‘general'. So I studied on this phenomenon in Old Japanese literature.
After analyzing the usages of 氣, I have reached the following conclusions :
1) In Old Japanese literature 氣 had two borrowed sounds /kë/ and /gë/. However, the two sounds were not used through the division of historical periods concerning the Man'yōshū or A Collection of a Myriad Leaves, but during the third period (711-733 A.D.), especially the fourth period (734-759 A.D.) with the rate 96.5%, that is, 82 examples among 85.
2) We can say that generally the two usages in 氣 appeared in the fourth period, because 6 poets, except Ōtomono-Yakamochi, made use of 氣 as /kë/ and /gë/ even in the same poems.
3) The fact 氣 represented /gë/ in the Old Japanese literature means that it reflected the devoicing of initials during the Tang Dynasty(618-907 A.D.) in the North.
4) Many Man'yōgana were under the influence of Korean writing system which employed Chinese charaters to represent the Korean language, but the 氣 evolved itself in Japan without its inflence.