Semantisme et classification dans l'ecriture chinoise: Les systemes de classement des caracteres par cles du Shuowen jiezi au Kangxi zidian.

• Author: BOLTZ, WILLIAM G.
• Position: Review

Semantisme et classification dans l'ecriture chinoise: Les systemes de classement des caracteres par cles du Shuowen jiezi au Kangxi zidian. By FRANCOISE BOTTERO. Memoires de 1'Institut des hautes etudes chinoises, vol. 37. Paris: COLLEGE DE FRANCE, INSTITUT DES HAUTES ETUDES CHINOISES, 1996. Pp. 224.

The Chinese script, from its first appearance in the second half of the second millennium B.C., has enjoyed more than three thousand years of uninterrupted existence down to the present. Beyond this, it continues to be used by the same people who invented it, to write the same language for which it was invented, in a form that is in most of its fundamental structural principles largely unchanged from its first creation. (Changes in secondary features affecting the outward appearance of the characters are, of course, considerable.) No other writing system in the world can make such claims.

For nearly the last two-thirds of its more than three thousand years' existence the script has been analyzed, and written words have been classified according to the perceived structure of the individual graphs that still serve as logographic or, as they are sometimes called, morphosyllabographic, "building blocks." In simple terms, these components of the writing system may be called "characters" (this is the term Bottero uses), and can be understood to have functioned as the basic units of the script throughout its history, more or less as they did when it was created, and in fact still to function that way. It is these characters that Xu Shen [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII], took as the orthographic units to be analyzed and classified in compiling the Shuowen jiezi [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII], dictionary, which he completed in A.D. 100. The Shuowen has in turn served as a model for similar lexicographical classifications down to the Qing dynasty, the most important instan ce being the Kangxi zidian [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII], compiled, as its name would suggest, under the aegis of the Kangxi Emperor (r. 1662-1722).

It is well known, of course, that the more than nine thousand individual character entries in the Shuowen are distributed among 540 semantic classes. Each of these classes is defined orthographically by a semantic "classifier" (abbr. "cl.," Fr. cle, Ch. bushou [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII]), which appears (by definition) in the graphic structure of every character entered into that class. These classifiers are usually, but not always, single graphic components, and they by and large originate in the semantic determinatives that were added as secondary components to characters that had been used paronomastically, that is, in rebus fashion, in the course of the development of the script. Later, as is also well known, the number of classifiers was reduced to 214, and this set has been fundamental since it was taken as the basis for the organization of the Kangxi zidian. What is not quite so well known is that the set of 214 was created earlier than its appearance in the Kangxi dictionary would suggest, by Mei Yingzuo [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII], as the framework for his Zihui [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII], dictionary of 1615. All the same, we are familiar with the set of 214 "Kangxi" classifiers, often imprecisely called "radicals," thanks to the Kangxi zidian, not to Mei Yingzuo's Zihui.

The rationale and logic that underlies the choice and order of the Shuowen classifiers has long been a topic of serious study, and Bottero's is the latest and certainly one of the most thorough and thoughtful of such investigations. What distinguishes her work from other studies of the Shuowen classifier system is that she carries the investigation beyond the Shuowen to examine the classification schemes that it inspired in all of the important lexicographical works that come between it and the Kangxi zidian. She focuses on the nature of this kind of orthographic classification and its semantic underpinnings, both in its first manifestation in the Shuowen and as it evolves from its Han prototype down to the Kangxi period of the Qing dynasty. Bottero divides this period into two parts, the Tang being the crucial "transition period" from one to the other.

Chapter I deals with the Shuowen, not only the first but without doubt the most important of all of the dictionaries that are arranged by semantic classifiers. Bottero reflects this clearly by giving nearly one-half of her entire study over to that one dictionary, including a fully annotated translation and carefully scrutinized study of Xu Shen's "Postface" (pp. 15-81). Her concern is not as much with the linguistic and orthographic question of why classifier Y follows X and precedes Z as it is with the thinking and beliefs that may have motivated the selection of certain graphs to serve as individual classifiers and that may underlie their...