In the early stages of Japanese writing, Chinese characters were sometimes used as phonetic symbols to represent the syllables of Old Japanese. This phonetic character writing, known as man'yôgana, has often been viewed as a highly complex and redundant system, as there were approximately 1000 different ways to write no more than 89 distinct syllables. However, a study of eighth century Japanese poetry written in man'yôgana reveals that ancient writers knew of a simple and consistent way to write Old Japanese phonetically. There are over 600 songs and poems, found in the Man'yôshû, the Kojiki, Bussokuseki-ka, various fudoki, the Shoku Nihongi and Kakyô hyôshiki, which are written entirely in man'yôgana, with a core set of about 150 characters. Despite the presence of a number of rare man'yôgana, particularly in the Man-yôshû , and occasional differences between man'yôgana usage in these texts, these songs and poems provide evidence that a highly consistent man'yôgana writing system existed for Old Japanese. The use of this man'yôgana system was not an isolated eighth century event but an important part of the development of the Japanese writing system. Many of the characters used in eighth century man'yôgana poetry are also found in the phonetic transcriptions of names in early seventh century (Suiko, period) documents. More than two-thirds of the Chinese characters from which modern hiragana and katakana are derived were already in common use as man'yôgana in the eighth century. Like most writing systems, the man'yôgana system used to write poetry has both archaic and innovative elements, indicating that it was not invented all at once in the eighth century but was the continuation of a writing tradition that dates back to at least the early seventh century. It is a pivot point between pre-eighth century phonetic name transcriptions and the development of kana writing from the ninth century onward. Although identical to Chinese characters in visual appearance, man'yôgana are similar in function to modern hiragana and katakana, and are therefore most accurately viewed as a type of kana. |