Organizer: Mary S. Erbaugh, University of Oregon
Chair and Discussant: Victor H. Mair, University of Pennsylvania
The ideographic myth holds that Chinese characters and Japanese kanji express ideas directly without reference to spoken language. In this controversial but mistaken theory, ideographs express not words, but pictures of mostly concrete objects. Meaning is claimed to be universal and transparent, though imprecise. Languages are reduced to writing systems. Ideographs are typically discussed in isolation, with a near obsessive focus on the fascinating 3% of characters which do not include a phonetic element. Writing is equated with abstract painting, while extended texts are slighted. When ideographic writing is said to be unchanged since 3,000 BC, it becomes easier to disregard differences over time genre, and style (classical and vernacular) which span unrelated language families in China, Japan, Korea, and Vietnam.
The more sweeping reductionist claims credit ideographs with fundamental differences in brain structure and national style. Japanese and Chinese borrowed the notion of changeless non phonetic ideographs from 19th western missionaries. Modern nationalists then reified the ideograph as a badge of national uniqueness, as they rejected Confucianism, classical languages, and a linguistic tradition of rhyme table phonetics. Among westerners, the myth is tainted by orientalism. French enthusiasts tend to praise ideographs as promoting peace and poetry; Anglo German views tend to excoriate them as barriers to logic, science and capitalism.
Why should we care? Decades of careful research refute the ideographic myth, rendering it as outdated as creationism or the flat earth hypothesis. Experiments show that characters function very much like other writing systems, including alphabets and syllabaries, to convey the sounds of spoken words. Linguists are frustrated when our fellow Asianists indolently persist in discussing Asia in crude and discredited l8th century terms. And the nature of characters has serious implications for how we think about Asia, how we teach about it, and how we teach the languages written with them. Policy implications extend beyond the US classroom to millions of native speakers. A double panel will explore interdisciplinary implications for both Japanese and Chinese. The first panel is theoretical, examining the impact of the myth on history, literary theory, philosophy, and the effect of discourse genre on writing systems. The second panel examines applied research on reading and writing by Chinese and Japanese native speakers, and implications for teaching US students.
J. Marshall Unger, University of Maryland
It is frequently claimed that kanji in Japanese texts function as logograms or ideograms. A large and growing body of empirical evidence shows that such claims are false, but because of the tendency of scholars of literature and historians to neglect the findings of linguistic science, these claims continue to exert a strong and pernicious influence on how Japanese history gets written.
An egregious example of what can go wrong when otherwise knowledgeable writers act on vague or erroneous assumptions about the Japanese writing system is afforded by a passage in Toshio Nishi's 1982 book Unconditional Democracy: Education and Politics in Occupied Japan 1945 1952 (pp. 203 4) in which he quotes from a 1946 statement by George Bailey Sansom ("Education in Japan," Pacific Affairs 19:413 15) and severely criticizes the U.S. Education Mission of 1946. The Mission's report of 1946 recommended, among other things, that the Japanese consider the romanization of their script. The reaction of historians to this recommendation, typified by Nishi, has generally been negative: it is seen as the result of the misguided efforts of certain Americans-specifically, Robert King Hall-who were allegedly technocratic utopians ignorant of Japanese culture. Certain writers have even gone so far as to argue that even the modest script reforms beginning with the toyo kanji and gendai kanazukai of 1946 were unnecessary and had to be accepted by the Japanese government in order to prevent even more "damaging" measures.
A careful analysis of the key passage in Nishi reveals that he and even Sir George himself stumbled at crucial points in their respective arguments because of their uncritical acceptance of the Ideographic Myth and its more common corollaries. The distortions each introduced into his discussion of contemporary events and their background can be traced to an unarticulated but clear underlying belief that kanji function differently in Japanese writing from the way all other kinds of written symbols function in other writing systems of the world.
Mary S. Erbaugh, University of Oregon
When the structure of language becomes a model for analyzing both verbal and non verbal 'texts,' which may range from Freud to fashions in women's shoes, that linguistic model had best be accurate. Unfortunately, many post structural literary theorists seize on the ideographic myth as offering an indispensable vision of 'difference' and the 'other.' If characters did not exist, French theorists would almost have to invent them. Derrida's theory of logocentrism gleefully turns linguistics upside down to view writing as central, and spoken language as parasitic. Logocentrism views writing in alphabets as a force for oppression, slavery, and imperialism. But characters, mislabeled as 'non phonetic languages,' are explicitly viewed as being like women, as silent, passive, and unchanging, as closer to emotion, gesture, and the concrete objects found in nature.
Though Derrida says 'nothing exists outside the text,' his school examines astoundingly few texts in Chinese or Japanese. Once characters are viewed as paintings rather than words to be spoken, 'language' in China and Japan is somehow reduced to poetry in writing. A few poems may be examined in translated blank verse, with a single European word per character. Pronunciation, rhyme and other formal structure, of passionate concern to native speakers, are often ignored. Chanted, drummed or sung poems remain unmentioned. Narrative, drama, and other genres whose oral complexity is less easy to ignore are little discussed. Sound bound, culture bound, passionately cherished at home, longer texts such as The outlaws of the marsh (Shui hu zhuan) seldom appeal to western readers anyway.
Literary émigrés from China or Taiwan, delighted to fling off stiflingly realist criticism, can be exhilarated by post structuralist theories of difference. But many of the most brilliant adaptations of post-structuralism to Chinese use the deliberately ambiguous and oracular Dao de jing to exemplify the language as a whole (Michelle Yeh, Wai lim Yip, Zhang Longxi). The Dao de jing may be the text which fits post structuralist theory best, with its mythical author, and ancient, terse, paired paradoxes whose indeterminacy was made for divination.
But clinging to the ideographic myth reinforces western orientalism. The focus on the Dao and a few short poems reinforces the notion that because some authors chose to write with terse, elegant ambiguity, the Chinese language is incapable of precision. If meaning is subjective and history irrelevant, why should Europeans waste years studying Chinese? If characters are silent, as well as simultaneously ambiguous, unchanging and concrete, why shouldn't an English monolingual publish yet another 'translation' of Mao or the Dao? If Mao and Dao are almost the only Chinese works they know, a universe of writing remains locked outside from joyful discovery. And much potentially exciting Chinese influence on theory is blunted and dulled. Much remains to be discovered about the intricate relations between spoken and written language, between graphic, phonetic, and typographical elements in script. An enormously varied range of genres survive, both classical and modern. Each author makes a spectacular range of stylistic choices, whether in annals or 'murky' verse. Any one of them could prove invaluable to illuminating the little understood connections between speech and script.
Chi yun Chen, National Tsing Hua University, Taiwan
Philosophical debates over Chinese often argue that ideographs express ideas directly, in unchanging concrete form. In fact, varying discourse genres had an important effect on characters, as scribes developed them slowly over many centuries for texts of widely varying explicitness. Character use, wording, and syntax, all varied by discourse type. Dialogue form used in Shang dynasty oracles reflects a spoken style. The question and answer format helps decode the terse, unpunctuated sentences. Scribes used a variety of competing characters in the Shang and early Western Chou dynasties (1,600-900 BC). Some characters were sign graphs, schematized drawings of objects which symbolized extended word meanings. Other characters symbolized voice signals more directly. The sign graphs and voice graphs developed in separable but inter related ways.
As writing became increasingly widespread, by end of the Western Chou in 771 BC, scribes had extended many more old sign graphs metaphorically for a wide range of new words. I will discuss the gradual development, integration, and standardization of sign graphs and voice graphs. These are discussed with reference to divination texts such as the I Ching and the T'ai hsuan Ching, as well as Confucius' own deep concerns over characters, language and discourse during the language discourse crisis.
Timothy J. Vance, Connecticut College
The Japanese went through a long history of orthographic contortions to adapt the Chinese writing system to their language, and the result, in the words of John DeFrancis, is "one of the worst overall systems of writing ever created." A particularly lamentable feature of the system is the now deeply entrenched practice of using the same kanji (Chinese character) both to represent elements borrowed from Chinese and to represent native Japanese elements with similar meanings. An example is the kanji used to write both Sino Japanese juu 'ten' and native Japanese too 'ten' (Nelson #768). The association with juu is a simple reflection of the fact that a substantial amount of Chinese vocabulary was borrowed into Japanese along with the writing system. The association with too, on the other hand, required the perception of a semantic connection between the imported Chinese element and the native Japanese element. The ampersand provides an analogous example in the English orthographic tradition. In books & papers it represents the native English word and, whereas in books &c. it represents et, a borrowing from Latin.
In everyday parlance, the Sino Japanese elements associated with a kanji are called its on'yomi, and the native Japanese elements associated with it are called its kun'yomi. When such a native element is an inflected word, the kanji is typically used to represent only the constant portion, with the variable remainder spelled out in the hiragana "syllabary." For example, a kanji (Nelson #4284) is used to write the first syllable of the inflectional forms of the native Japanese verb meaning 'see,' such as the nonpast affirmative mi ru, the past negative mi nakatta, and the conditional mi reba, etc. It is not uncommon for a kanji to have more than one kun'yomi; an extreme example is the kanji which is used to write nouns such as ue 'upper part' and kami 'top' and verbs such as agaru 'rise,' ageru 'raise', and noboru 'climb' (Nelson #798).
It is tempting to see the existence of kun'yomi as evidence for the claim that kanji are ideographs, but such a conclusion is unwarranted if ideographs are understood to be symbols that represent meaning directly without mediation by a formal unit such as a word or a morpheme. The meaning associated with a formal unit can, and usually does, change over time, and it should come as no surprise that an on'yomi and a kun'yomi associated with the same kanji can diverge semantically. For example, there is a kanji (Nelson #775) with the on'yomi jiki and choku, and the semantic range of these Sino Japanese morphemes includes meanings such as 'straightness,' 'directness,' and 'honesty.' The same kanji is used to write the native Japanese verb naosu, which has a range of meanings including 'repair' and 'correct.' According to the entry in Nihon Kokugo Daijiten, the meanings of the verb developed from an older meaning 'make straight.' This semantic shift is unremarkable in itself, but one of its consequences is a conspicuous discrepancy between meanings associated with the on'yomi and the kun'yomi. If kanji were truly ideographs, semantic divergences of this kind would presumably be impossible.
On the other hand, there is at least one clear case in which an on'yomi seems to have taken on the divergent meaning of a kun'yomi associated with the same kanji. The primary modern meaning of the native Japanese adjective yasui is 'inexpensive,' but this meaning represents a shift from earlier 'peaceful, calm.' The kanji used to write yasui is (Nelson #1283) has the on'yomi an generally carries the older meaning, but in the Sino Japanese compound anka 'low price' it clearly carries a meaning close to that of yasui. The earliest citation for yasui with the meaning 'inexpensive' in Nihon Kokugo Daijiten is 909, and the earliest citation for the word anka is 1906, so it seems to have taken the on'yomi about a millennium to catch up with the kun'yomi, but the coining of anka makes no sense without reference to the kanji linking an to yasui. Examples of this kind, like the invention of kun'yomi itself, can certainly be cited in support of the claim that kanji have an ideographic aspect, at least potentially, but this modest claim is a far cry from simply calling kanji ideographs. The arbitrariness of the connection between sound and meaning is regarded as a fundamental feature of human language, although the existence of onomatopoeia makes it necessary to equivocate. In the same way, the basic principle that writing (even Japanese writing) represents linguistic forms is not invalidated by the existence of examples like anka.
Would you like to return to the Interarea, Library, & Teaching Table of Contents? Choose another area?