Xu Shen's graphic analysis revisited
Résumé
Like other writing systems, the Chinese writing is based on a limited number of basic elements. Compared to the 26 letters of the alphabet, or the 70 hiragana and katakana signs for (modern) Japanese, the precise number of basic units is unknown in the Chinese case. It is almost never touched upon, or varies a lot, not only between specialists and corpora, but even within the same corpus. In trying to elucidate the reason for such a different treatment in the case of the Chinese script, one cannot but focus on the first author who developed a proper analysis of the script. When Xu Shen wrote the Shuo wen jie zi , the first dictionary of Chinese characters compiled at the end of the first century, he provided a systematic analysis of more than 9 000 characters. Considering that their structure was best brought out by the older style of graphs known as the small seal script, he based his analysis on this older style, which kept graphemes more clearly apart. If compound characters can include more than 10 basic graphemes, Xu Shen basically only retained 2 main constituents to which he attributed a phonetic or a semantic function. Most characters (about 86% in Shuowen) consist in the combination of a semantic and a phonetic constituent (). In these cases, Xu Shen's distinction between semantic and phonetic constituents considerably simplified the graphic analysis. Once a constituent is attributed a phonetic function, there is no need to try to understand its meaning in a character (or provide ad hoc interpretations for it). But what about semantic compound characters without identifiable phonetic constituent, included in the other 14% characters in Shuo wen jie zi? The study of semantic compound characters is more complex and, at times, more subjective, but it allows a better understanding of Xu Shen's motivation. I shall present the different strategies and formulae (X, X, X , X Y, etc.,) employed by Xu Shen in these cases, with concrete examples, and discuss the interpretation of the relative location of the constituents (such as for example in :). The latter as well as the different kinds of motivated explanations, Xu Shen often adds, show that the graphic analysis is designed to ascertain the gloss of the character. We will see that even though Xu Shen provided some graphic analyses with 3, 4 or even 5 constituents, in most cases, he basically retained 2 main semantic constituents. Systematic and sound as it may appear, Xu Shen's analysis in terms of constituents, clearly motivated by the semantic interpretation of the characters, has eluded the fundamental question concerning the basic units of the Chinese writing system.
Domaines
Linguistique
Origine : Fichiers produits par l'(les) auteur(s)