THE DISSERTATION AS HANDBOOK: A NEW GUIDE TO THE SHUO-WEN CHIEH-TZU.

• Author: MILLER, ROY ANDREW

Designed as a vade mecum for Sinological neophytes whose studies require them to extract lexical, epigraphical, or phonological materials from the Shuo-wen chieh-tzu of A.D. 100, Marc Winter's monograph attempts a general introduction to the problems of the text, followed by annotated translations of the lemmata for its 540 graphic-keys. Unfortunately the author's jejune bibliographical command renders the former portion of his book alternately redundant and incomplete, while his unfamiliarity with the salient problems obtaining in Chinese historical linguistics today largely vitiates the latter section, as does also his unsatisfactory philological control of his sources, both ancient and modem.

THIS MONOGRAPH, WHICH BEGAN LIFE as a Winter-semester 1995-96 Universitat Zurich dissertation, merits attention if only because it abandons the accepted parameter of that genre, and seeks not so much to inform as to instruct, specifically to help others in their use of the Shuo-wen chieh-tzu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (= SWCT), an A.D. 100 encyclopaedic compendium of Chinese script and language. [1] This makes it a thesis with no particular thesis, apart from the entirely commendable goal of helping others who may wish to consult this massive text: "Es ist fur Sinologinnen und Sinologen konzipiert, die fUr ihre Arbeit das Shuo Wen Jie Zi verwenden, ohne zu Lexikologen werden zu wollen" (back cover).

In certain respects Winter's dissertation may perhaps serve such persons well enough, especially his "Teil 1: Analyse" (pp. 12-234). Here students of the SWCT will find accounts of the text's authorship and structure, together with inconclusive discussions of its relationship to the much-mooted "Six Scripts" of Chinese orthography and to the Old Text-New Text controversy that engrossed the Han literatus-elite. W has enjoyed the advantage of six years of study in the People's Republic, reflected in his citation of a flood of recent Chinese publications whose tempting titles make one wish they were readily available to us all; unfortunately few of them are. Still, and in so far as one may judge from W's excerpts from these recent contributions, few of them live up to the promise of their titles; in sum, the recent boom in SWCT studies in China appears to have added little to what we have known about this text since the mid-1950s.

But even if the wealth of new publications that W acquired in China added substantially to our understanding of the SWCT, which does not appear to be the case, this in itself would neither justify nor explain why W neglects mentioning anywhere in this introductory section of his Handbuch the comprehensive essay by William G. Boltz on this text, in Early Chinese Texts: A Bibliographic Guide, ed. M. Loewe (Berkeley: Society for the Study of Early China, 1993), 428-42. [2] With that essay in hand, the would-be-user of the SWCT might safely overlook much of W's "Teil 1"; conversely, if W himself had consulted it, he too would have been spared needless duplication of effort.

Over and above this major bibliographical lapsus, the first half of W's monograph is not without its other problems. His concern for terminology, expressed in more than one passage, is admirable; but it does not always accomplish its goal. Formulations such as "[d]er graphische Aspekt eines Graphems heisst Graph" (p. 28) hardly achieve either precision or clarity. Troubled by what he terms the "Ungenauigkeit" of the designations "'Orakelinschriften' oder 'oracle bone inscriptions"' (p. 25, n. 8) he opts instead for the abbreviation "SPI" for "Skapulaund Plastroneninschriften," but then usually goes on further to gloss the "SP" of this designation with Chin. chiaku wen [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (e.g., pp. 253, 304, 528); this seems like terminological over-kill, to say the least. He discusses the many reasons why the SWCT's term for its graphic-keys, pu-shou [CHINESE CAHRACTER NOT REPRODUCIBLE IN ASCII TEXT], should neither be understood nor rendered as "radical" (pp. 58 ff.); but th en proceeds yet once more to perpetuate this misleading term "radical" throughout the six hundred pages of his book. His terminology for earlier reconstructed linguistic entities in the history of the Chinese language contradicts accepted Sinological usage at the same time that it throws into question W's own understanding of his materials; we read of the "antike Lesung" of this or that character (p. 398), alongside references to "rekonstruierte Aussprache" (p. 306), "Angabe zur antikchinesischen Lautung" (p. 307), "rekonstruierten Aussagen" (p. 464), and "Gemass der rekonstruierten antiken Lautungen" (p. 519; similarly, p. 521). [3] Will the Handbuch-user be able to guess that all these mean much the same? Probably not, because none of them is terminologically precise.

The user of this volume will also be less than well served by W's desultory introduction to the historicallinguistic problems presented by the SWCTs punning definitions, many (or most) of which may be read as paronomastic etymologies. Some of these are perhaps original with Hsu Shen [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (ca. 55-ca. 149), [4] the compiler of the text; but in the main they are important because they preserve substantial fragments of what must have been an even more extensive body of arcane scholastic lore. W's introduction to these passages remains essentially uninformative, mainly because it fails to relate the phenomenon of Han paronomastic etymological speculation to Chinese cosmological assumptions and to Chinese society's view concerning the nature of language in general. [5]

Unfortunately, this largely unresolved problem of the Han paronomastic word-hoard also plagues W's "Teil 2: Ubersetzung" (pp. 235-586); much of what he has to say there too will only bewilder and mislead rather than aid the Handbuch-user.

In this second half, W attempts an ambitious and potentially valuable project, the translation of Hsu's lemmata for his 540 graphic-keys (W's "radicals"), annotated and documented with references to the entire range of secondary literature, Chinese and Western, old and new. But in this daunting task W's reach has clearly exceeded his grasp. Many of the SWCT's entries concerning its 540 graphic-keys bristle with philological conundra that cry out for treatment at greater length, and at a greater depth of familiarity with the Sinological literature, than W has been able to bring to bear upon them. It is hardly ever simply, a question of merely translating an entry and then commenting upon it. For W's translations of Hsu's lemmata to have been genuinely useful they would have had to have been preceded by intensive investigations of the texts in the light of wide reading in the Ch'ing commentators, one or another of whom probably unearthed every possible (or at least, every surviving) philological ancillary for our understanding of this all-too-frequently gnomic source.

But this W has time and time again failed to do. Chiefly he contents himself with a glance at the commentary of Tuan Yu-ts'ai [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (1735-1815). To be sure, this is the single most useful of the Ch'ing sources. But more than once, if he had read further on among the other commentators, he would have found more solutions, including many for problems that he does not appear even to recognize as such.

To illustrate this particular variety of philological problem, the discussion of one specific example must suffice; many others could be explored.

Hsu's gloss for yu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] "have," his graphic-key 238, begins with a truly cryptic phrase: [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]. This W translates correctly enough, "Was man nicht haben sollte" (p. 387); and one begins to wonder if here we have to reckon with a case of lucus a non lucendo. But then even more enigmatically, Hsu adds, [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] again, W: "Im Chun Qiu und den Kommentaren heisst es [immer wieder]: Es gab Sonnen- oder Mondfinsternisse." But he cannot, or at least does not, explain what the first part of this lemma has to do with the second; even more importantly, he is silent on the question of why Hsu appears to explain "have" with its putative opposite "ought not to have." Tuan's commentary on this puzzling lemma is curiously (and uncharacteristically) unhelpful; and W has not gone beyond its resources. But two Ch'ing commentaries by Wang Yun [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (1784-1854), both reproduced in extenso in the SWCT ku-lin [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT], 3004, clear up the difficulty. They show that the first part of Hsu's gloss is neither original with him nor actually a lexical gloss; rather it is a grammatical note that may be traced back to a now mostly lost commentary on the Tso chuan by the colorful scholar-general Th Yu [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (A.D. (222-84). [6] There Tu had himself cited a still earlier grammatical gloss on two Tso chuan passages (Legge, tsl., pp. 42, 707) by the Han scholars Liu Hsin [CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT] (fl. ca. A.D. 1) and Chia K'uei (A.D. 30-101). They had interpreted yu in both these passages as indicating the occurrence of prodigies or other events not properly expected (e.g., good years [[CHINESE CHARACTERS NOT REPRODUCIBLE IN ASCII TEXT]] of harvest during the reigns of evil rulers, or grackles appearing in the...