Chinese has been written exclusively using characters since the inception of the writing system, which can be traced back to about 3500 B.C.E. It began as incisions on bone and turtle shell for the purpose of divination ceremonies, politics, and warfare. Many inscribed pieces dating back to 1200 to 1500 B.C.E. (during the Yin Dynasty) were found toward the end of the 19th century. Two to three thousand characters, or about a half of those characters discovered, were still in use at the time of the discovery.
By the time the Japanese scholars and bureaucrats who went to China were reading documents in Chinese, the Chinese system of writing had long been developed. It had a comprehensive dictionary of characters. A lexicographer named Kyoshin (Ch. 許慎, A.D. 30-124) compiled a dictionary of characters (whose title is read as Setsumonkaiji (説文解字) in Japanese), which contained more than 9,300 characters. By this time, a large number of character compounds had also developed. Calligraphy was also gaining status as a high form of art. (The reader may think the number of characters is staggering, but consider this: In 1716 the Chinese Emperor Kōki 康煕 (1662-1722) commissioned a dictionary of characters. When completed, it listed an impressive 47,035 characters.)
At the beginning, Chinese characters were not used for writing Japanese. Instead, those who could write—scholars, government bureaucrats, and monks, for instance—simply wrote their documents in Chinese. The first Japanese constitution, the Seventeen Article Constitution promulgated in A.D. 604, for instance, was written by Prince Shōtoku (574-622) in Chinese—using Chinese characters, vocabulary, and grammar.
As early as the fifth or sixth century, however, scholars realized that Japanese could be written or transcribed using just the sound value of Chinese characters. This method of writing was possible because Japanese had only a small set of distinct sound combinations and Chinese had a large inventory of sounds. Thus a method was devised to write Japanese using Chinese characters. That is, to write a Japanese sound ("syllable" or, more properly, mora), a character having the same or similar sound was chosen from among the Chinese characters. Note that when characters were used in this way, the original character's meaning was often disregarded. This system of writing, using Chinese characters for their phonetic value only, is called man'yōgana.
Man'yōgana, along with Chinese proper, was used for Japan's oldest poetic anthology Man'yōshū, a collection of more than 4,500 poems, which is thought to have been compiled around A.D. 760. During the Nara period (710-784) documents were written in styles ranging from purely Chinese (examples include the first constitution of Japan mentioned above, the Nihonshoki, and an anthology of poems written in the Chinese language by Japanese called Kaifūsō) to a mixture of Chinese and Japanese. Much of the Man'yōshū was written in this mixture. For instance, the following Man'yōshū poem illustrates this interesting blending of the two languages. The poem Man'yōshū 1:8, attributed to Nukata no Ōkimi (Princess Nukata) reads as follows (in romaji flanked by Japanese as written in this blended language, followed by an English translation):
Romaji: Original Japanese/Chinese blend:
Nikitatsu ni 熟田津爾
Funanori-semu to 船乗世武登
Tsuki mate ba 月待者
Shiho mo kanahinu 潮毛可奈比沼
Ima wa kogiide na 今者許芸乞菜
At the harbor of Nikitatsu
To travel on a boat
We waited for the moon to rise
The tide just came in
Let us now embark!