"Writing in Early Japan"
Anne Commons
Columbia University
Early Kokuji in Japan
The Japanese term kokuji can have a variety of meanings: in its broadest sense it can refer to the writing system of a given country, not necessarily that of Japan, and it can also refer to the kana syllabaries, Japan's indigenous script, as opposed to the introduced Chinese characters, or kanji, but I am using it here to refer specifically to "Chinese" characters made up in Japan, according to the Chinese model, which are also known as waji, or wasei kanji . The number of kokuji in common use in contemporary Japanese is very small: of the 1,945 Joyo Kanji, a mere eight are kokuji, and a further two are included in the list of characters approved for use in personal names. The total number ever produced, however, is approximately 1500, and, from an historical viewpoint, an interesting pattern seems to emerge in that the numbers of kokuji produced, period by period, was greatest earliest on, the numbers decreasing in subsequent ages. The first collection of kokuji, and the one on which I shall concentrate my discussion, is that appearing in the Shinsen Jikyo (Newly Compiled Mirror of Characters, c.901): a dictionary listing some 21,000 characters, it includes a list of approximately 400 kokuji
The kokuji which appear in the Shinsen Jikyo (and other early texts) reflect the adaptation of an initially alien writing system to the Japanese lexicon: by contrast, the kokuji formed most recently, which is to say in the Meiji period, are fundamentally different in that they are examples of an established writing system adapting itself to foreign vocabulary. While some kokuji are able to be regarded as variant forms of, or equivalents to, existing kanji, others exist which are entirely original coinages, constructed from the basic units of the writing system according to historical principles, rather than as variations on a theme. It may be noted that although kokuji may be called a subset of kanji in general, and are, broadly speaking, referred to as such (as when being discussed in relation to kana, or as part of the Joyo Kanjl), there are certain aspects in which they differ, as a group, from the kanji in use in Japan, mainly in terms of their formation, in that most kokuji are formed according to the kai'i (semantic-semantic) pattern, whereas the overwhelming majority of kanji are keisei (semantic-phonetic) characters. Kokuji tend to lack on (SinoJapanese) readings, having only kun (Japanese), but, conversely, examples do exist which have only on and lack kun, and still others have both, just like authentic kanji The formation of kokuji and the assignment of their readings according to the inherited Chinese models, where applicable, are indications of the thoroughness with which kanji have been assimilated into the Japanese language, and the huge proliferation of kokuji appearing in a text as early as the Shinsen Jikyo is an indication of the speed with which the Japanese attained a sophisticated understanding of the principles of character formation as they made the massive step from a pre-literate society to one employing an ancient and complex writing system originally invented for a language with truly profound differences to their own.
David Lurie
Columbia University
Beyond Transcription: The Uses of Writing in the Man'yôshû
The reluctance of modern linguists to examine writing in depth has often been commented upon, but recent years have seen increased discussion in linguistic circles of its nature, function, and history. One feature of this discussion has been an insistence that writing represents spoken language; given the long history of uninformed European and North American speculation about non-alphabetic scripts, such an insistence is salutary. However, due to various factors including examples provided by the decipherment of hieroglyphic writings like Mayan and Egyptian, reactions against Orientalist interpretations of Chinese writing and culture, and assumptions about the relation of writing and speech which lie at the heart of modern linguistics, the representation of spoken language has often been elevated to something of an absolute principle. In the context of scholarship on East Asia the chief proponents of this principle, which will be referred to as the transcriptive view of writing, are John DeFrancis and J. Marshall Unger.
DeFrancis and Unger argue that, in terms of graphic form and linguistic function, Chinese characters represent spoken Chinese. A less well-developed corollary is that in the Korean, Japanese, and Vietnamese context, characters have (and have had) the same relationship to speech. It is inarguable that in these cultures, as in every other known culture, the writing of natural languages reflects the oral and aural shape of those languages; the intention here is emphatically not to replace the primacy of phonography with the primacy of ideography. Rather, it is to question the usefulness of a rigid insistence on writing as transcription, and to suggest that attention paid to other aspects of writing is not always a sign of allegiance to what Unger and DeFrancis label the "ideographic myth." It is a commonplace that innovation in writing takes place at cultural and linguistic boundaries, as scripts are borrowed and adapted; East Asia is no exception. A close examination of the transformations undergone by the Chinese writing system as it moved into new, non-Sinitic language communities will demonstrate the complexity and variety of writing with characters, and of writing in general.
Attempts to argue for a more complex concept of writing often degenerate into disputes over how to define the term itself; for that reason it may be more productive to postpone more abstract considerations, and to begin with a careful examination of writing as it is practiced in a given cultural context. Chinese writing was present in Japan at least as early as the fifth century, but its use does not seem to have been widespread until the middle of the seventh, and it is in the eighth that the most diverse use of characters is apparent. The primary source for early writing is the great Nara period (710-794) anthology of vernacular poetry, the Man'yôshû. Its twenty volumes and 4500-odd poems display a stunningly varied system of writing, ranging from the so-called ryakutaika ("abbreviated-form poems"), which amount to cryptic modifications of classical Chinese, to mixed notations functionally identical to modern Japanese orthography, to syllable-by-syllable phonogram transcriptions which R. A. Miller once described as written Japanese "at its historical best, as rational and efficient as it ever would be."
In Japan and abroad, scholars often succumb to the temptation to organize these various writings into a narrative of increasing phonography. There is little doubt that, as confirmed by evidence from epigraphy and archaeology, the abbreviated poems represent the oldest stratum of vernacular writing, and it is also clear that certain lines of phonograph writing eventually develop into the later Heian kana systems. However, consideration of the various scriptal registers in the anthology suggests that many factors influenced those who inscribed the poems. In particular, a category of usages usually called gisho, or "playful writings," displays a sophisticated literary intentionality that is often at odds with straightforward transcription. Numerical, visual, and allusive puns and riddles abound, making gisho an ideal departure point for a more complex consideration of the function and purpose of writing. In particular, these practices demonstrate that many received notions of man'yôgana (Nara phonograms) rely on an impoverished conception of the cultural function of writing and language; these poetic inscriptions should be viewed as aesthetic and artistic performances in their own right, rather than transcriptions of greater or lesser linguistic efficiency.
Yukio Lippit
Princeton University
Nara Imperial Scriptoria
The presentation proposes to examine the role of the eighth-century imperial scriptoria of Nara Japan on the history of Japanese calligraphy and Japanese writing in general. The scriptoria were set up, of course, to copy sutras commissioned by the imperial government. and in general followed continental prototypes in format. The significance of these institutions of the activity of sutra copying in general, however, differed greatly between the two countries. In Japan, where in the eighth century writing was still in adolescence and calligraphy in its infancy, the scriptorium was the site where the sorting-out of T'ang calligraphic styles took place. This is not to say that every sutra produced at these writing factories looked the same, but rather that here first took root the idea of a mode of writing that emphasized speed, legibility, and the infinite repeatability of the individual strokes of the Chinese character. It was against this background that the more complicated, multivarious practice of the "art of writing" could develop in Japan.
What is the difference between writing and the art of writing? The boundary between the two has always been blurry, and nothing underlines this fact more than the fact of sutra-copying. The copying of sutras is at the same time political, religious, bureaucratic, "artistic," and anonymous. The eighth-century Nara sutras represent the full panoply of the "functions" of writing: an examination of the imperial scriptoria that produced them is simply a way of approaching these "function" by pondering one of the sites of production of ancient Japanese writing.