Philosophy on Bamboo

Studies in the History of Chinese Texts

Edited by

Martin Kern, Princeton University Robert E. Hegel, Washington University, St. Louis Ding Xiang Warner, Cornell University

VOLUME 2

The titles published in this series are listed at brill.nl/hct

Philosophy on Bamboo
Text and the Production of Meaning in Early China

By

Dirk Meyer

LEIDEN BOSTON 2012

Cover illustration: Image of bamboo strips from Gudin One Z y, adapted from Hbi shng Jngmn sh bwguan (1998) by Tobias Kegler This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data Meyer, Dirk, 1975 Apr. 18 Philosophy on bamboo : text and the production of meaning in early China / by Dirk Meyer. p. cm. (Studies in the history of Chinese texts ; v. 2) Includes bibliographical references and index. ISBN 978-90-04-20762-2 (hardback : alk. paper) 1. Chinese classicsCriticism, Textual. 2. Chinese literatureCriticism, Textual. I. Title. II. Title: Text and the production of meaning in early China. III. Series. PL2461.Z7M467 2012 181.11dc23 2011034520

ISSN 1877-9425 ISBN 978 90 04 20762 2 Copyright 2012 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

Fr Claus V. Bock (Hamburg, 7 May 1926Amsterdam, 5 January 2008)

CONTENTS Acknowledgements ............................................................................ Introduction ........................................................................................

PART I

ix 1

ANALYSIS: TEXT AND STRUCTURE 1. The Zhng xn zh do The Way of Fidelity and Trustworthiness ............................ 2. The Qing d y sh Failure and Success Appear at Their Respective Times ...... 3. The Wu xng Five Aspects of Virtuous Conduct ........................................................................................ 4. The Xng z mng ch Human Nature is Brought Forth by Decree ........................
PART II

31

53

77

131

TEXT AND PHILOSOPHY 5. Text, Structure, Meaning ............................................................ 6. Applying the Methodology: Ti y shng shu The Ultimate One Gives Birth to Water and Laoz ................................................................................. 7. Writing Meaning: Material Conditions of Meaning Construction in Warring States Philosophy ............................ 8. Conclusion: Writing Philosophy ............................................... 177

209

227 245

viii

contents
PART III

TEXTS, RECONSTRUCTIONS, AND PHILOLOGICAL NOTES 9. Reconstructing the Zhng xn zh do ............. 10. Reconstructing the Qing d y sh .................. 11. Reconstructing the Wu xng ................................................. 12. Reconstructing the Xng z mng ch ............. 13. Reconstructing the Ti y shng shu .............. Bibliography ....................................................................................... Index .................................................................................................... 257 269 283 311 353 363 387

ACKNOWLEDGEMENTS Many people, scholars and friends alike, have helped to shape this book. It grew out of my doctoral dissertation at Leiden University, where I was supervised by Barend ter Haar and Joachim Gentz. Their support, patience, and rigour have had a profound impact on this project. Both my mentors were supportive, intellectually stimulating, and enthusiastic teachers. Meetings with them were mostly a pleasure, even when they resulted in me rewriting a chapter or deleting an idea. Critical, but always reassuring, Barend and Joachim made me move on when necessary and brought me back to reality when I was flying too high. Over time, my supervisors became friends, as well as colleagues, and the book owes much to the trust shown to me by these two exceptional individuals, even after I left Leiden and moved to Oxford to take up my current position. In 2008, I spent three months in Princeton. During this period, I profited from regular meetings with Martin Kern, at which we discussed the book in great detail. I thank Martin for making my visit to Princeton possible, for closely reading my manuscript, engaging with it critically, and enabling me to rethink and rework some of the ideas, which I had previously taken for granted. I also wish to thank the early China community at Princeton University for their warm welcome and our fruitful discussions, particularly Willard Petersen, Andrew Plaks and Mick Hunter. I thank the Netherlands Organization for Scientific Research (NWO) for providing financial support for this visit. More recently, the manuscript has greatly benefited from the advice of Wolfgang Behr and William Boltz, who offered extensive criticism and comments for which I am deeply grateful. Jim Reed, Professor Emeritus of German at The Queens College, Oxford, read the entire manuscript and helped me improve my English. I further profited from conversations with friends and colleagues: Michael Nylan, Rudolf Wagner, William Baxter, Chn Jin, Paul van Els, Michael Puett, and Paul Goldin. Profound thanks go to my friends and colleagues at the University of Oxford, at both the Institute for Chinese Studies and The Queens College. Laura Newby sailed the Institute safely through the unpredictable

acknowledgements

tides of academic reality and made sure that young colleagues were given the necessary room to work on their projects. Peter Ditmanson and Justin Winslett repeatedly eased the burden of teaching and examining, helping me submit the manuscript in time. The Queens College forms a friendly and warm community, assuring that I do not live in exile. My deep thanks go to Phillip Harries, Fellow in Japanese, who eased my way into Oxford college life. I thank my friend Tobias Kegler for propelling me towards Leiden in the first place, as well as for creating the cover image for this book. Parts of chapters 1 and 2 appeared in a very different form in the Bochumer Jahrbuch zur Ostasienforschung 29 (5778) and Oriens Extremus 45 (179210). Deep thanks also go to my parents, who, always supportive, understand that there are many reasons one cannot be at homesuch as having a job in Oxford or, in fact, writing a book. I further wish to thank ine McMurtry, not only for the countless conversations we had that helped me to clarify ideas that previously made good sense in my head but failed to work in explanation, but, most importantly, for bearing me even when I wanted to talk, yet again, about this book. Dirk Meyer Oxford, 21 May 2011

INTRODUCTION

PHILOSOPHY ON BAMBOO This book analyses a defined corpus of philosophical texts from the Warring States period (ca. 481222 BC). It treats texts as objects in their own right and discusses the relationship between the material conditions of text and manuscript culture, writing, techniques of meaning construction, textual communities, and philosophy in the Warring States period. Based on a close reading of these texts, my analysis distinguishes between what I call argument-based texts and context-dependent texts. This distinction is not an absolute one but a methodological necessity to order the complexity of reality by drawing attention to the shared characteristics of diverse philosophical texts from this period. In a nutshell, meaning construction in the ideal type of argument-based texts is closely linked to writing. Contextdependent texts require reference to (typically oral) commentators and participate in a triangular relationship of meaning transmission consisting of the text, a mediator of meaning, and the receiver of the message. Hence, whereas argument-based texts facilitate a stand-alone philosophical exercise, context-dependent texts serve as a platform for broader philosophical processes that largely remain outside the text itself. Text and Ideas When engaging with early thought, scholars often take texts as mere repositories of ideas. This is to ignore that the physical manifestation of a text is actually the mediatorand therefore a remnantof early thought. The usual approach hence fundamentally neglects the relation that may exist between ideas and the material carrier that conveys these ideas to the present day. Instead of thinking of texts as mere vessels of thought when investigating the practice of philosophising in early China, I take a different approach. Rather than focusing primarily on the ideas expressed in texts, my analysis starts by dealing with the texts themselves as meaningful objects in their own right. The underlying assumption is that

introduction

the text as a material object can reveal vital information about the text as a cultural phenomenon. This means that it can provide information about the purpose for which philosophical texts were used, by whom they were used, as well as how they were used in the exchange of ideas at that time. In other words, by taking a closer look at the texts and their various strategies for constructing meaning, it is possible to gain a fresh view of the practices of philosophical reasoning more than two thousand years ago in those territories which today we call China.1 Early thought is always mediated by objects. And the only reason that early philosophical activities are known today is simply because, in one way or another, they were put into writing. This calls for conscious reflection about the relatedness of text and thought as well as about text as the primary remnant of early thinking because of the implications these issues have for the study of ideas. These implications can be summarised roughly in four groups of questions. First, it is essential to think about the degree to which philosophical concepts were shaped by being put into writing. Does the written word influence the structure of reasoning? And if it does, to what extent? Is writing only the transcription of thought, and is the philosophical text, by implication, merely the mimesis of the mimesis, as Jan and Aleida Assmann suggest, following Aristotle?2 Or do written texts in fact impart to thought a degree of abstraction that would be absent in oral discourse, as David Olson assumes?3 Were the texts composed in writing, and did this at least facilitate the complex analysis of a philosophical concern? Or was the writing on bamboo undertaken only after the texts had been composed orally? Do the texts represent structurally consistent edifices of thought that can (and should) be studied individually? Or are they only fractions of a larger, ongoing, and coherent (or incoherent) discourse that, however, did not survive to the present day? If this is the case, it needs to be given expression in the analysis of the written piece of thought as only a piece of thought which in itself is insufficient for (re)constructing a coherent philosophical edifice. And then, is it possible at all to demonstrate sufficiently the incompleteness of the surviving discourse in our engagement with the written ideas? If so, to what extent?
For ease of argument, I shall henceforth refer to the territories of the broader cultural domain of the Zhu as early China. 2 A. Assmann and J. Assmann 1998. 3 Olson 1994.
1

introduction

Second, questions such as these make it necessary to disentangle the complex relation between the spoken and the written word in those texts which the modern exegete aims to understand. It is essential to reflect on the wider question of whether writing imparts independence to the ideas expressed in the texts. Does writing free ideas from any situational context? If it does, then the question becomes how far it does that. To what extent are written thoughts necessarily mediated by media other than the text alone? Is it possible to understand the (isolated) idea that was given expression (or hinted at) in a written object under scrutiny? Or is a reconstruction of some kind of mediatoror meaningful contextnecessary to understand the written ideas? Do written thoughts gain independence simply because they are written down? Or do they only refer to a spokenand, at the date of manuscript production, still activediscourse outside the written text? Is it possible at all to reconstruct an imagined oral discourse behind the written text on the basis of the written text alone? Should we, finally, postulate an oral and ongoing discourse as the background for all philosophical texts? Or only for certain ones? Or not at all? How can we discern such an oral discourse behind the written textor its absence? Third, to look at a written piece of thought also means to look at some sort of philosophical record-keeping. Whether merely the notes of a teachers words, possibly lacking formal rigour and serving only a students memory, or an elaborate and stand-alone composition, in itself a masterpiece of philosophical writing, the act of writing something down forms a conscious act of record-keeping. It is crucial to bear this in mind when evaluating early thought. The question then becomes how the formal aspects of a text and its philosophy correlateif they do so at all. Are there specific types of philosophical language? Do we see texts that point to other meaningful references? Or should they in fact be understood as structurally closed compositions that can exist only in their given (formal) arrangements? Did the authors of early philosophical texts intentionally put certain ideas in particular compositional patterns? And, if so, does this say anything about the philosophy itself? Or about its composers? Or about the different philosophical traditions in which these texts were used? Fourth, for a proper presentation of written ideas in the history of Chinese thought, it is vital that we consider how the various finds present the texts. That is, to what extent do texts that seem to have a consistent focus but come from different sites diverge from one

introduction

another? What is the relationship of a certain text to its different and differingmanifestations? To what extent are changes in texts, or their relative stability, relevant for a proper understanding of their ideas? What do the changes in the different manifestations of a text (or their stability) say about the philosophical concepts that the text conveys? Is it possible at all to speak of coherent philosophical ideas if it transpires that a certain transmitted philosophical text differs appreciably from its excavated counterpart? And what do the changesor the stabilityof texts tell us about these texts themselves? Or about their authors? Is it possible to reconstruct the history of a text? And to what extent would this be meaningful? What do deficiencies in texts imply for a modern understanding of early philosophy as expressed in these texts? Is it possible to reconstruct early philosophy in a meaningful way? And if so, should one go so far as to reconstruct coherent edifices of thought even for a period such as the Warring States? Questions such as these are addressed in the present study of the philosophical materials that come from a late Warring States tomb in the ancient Kingdom of Chu . They arose out of an attempt to study the written ideas from the Warring States in a more coherent way, aiming to do justice to both the texts studied and the ideas expressed in them, whilst trying to avoid imposing modern concepts on philosophical texts from the distant past. Tomb Gudin One This study is based on a closed corpus of texts excavated in 1994 from a tomb near Gudin , Hbi Province. I henceforth refer to this tomb as Gudin One.4 In various respects, Gudin One proves ideal for a qualitative study of text and thought in Warring States China. The texts were part of a tomb assemblage and came to light during a documented excavation. As I shall argue, they have in common the endeavour to establish stable philosophical concepts. They were part of a discipline

4 Tomb Gudin One is located only nine kilometres north of the old capital of the Kingdom of Chu at Jnn , close to the village of Gudin in the Shyng District, Sfng , Jngmn City. The excavation report was published by the Hbi Province Museum in the City of Jngmn, henceforth referred to as Hbi Province Museum (Hbi shng Jngmn sh bwguan 1997).

introduction

and may be termed broadly philosophical.5 To date, Gudin One is the only well-documented Warring States tomb to contain a variety of such texts.6 They reflect different kinds of sociopolitical philosophical reasoning and even address different audiences. Some are concerned with proper rule and discuss appropriate measures of government both from the perspective of the advisor and from the perspective of the ruler himself; other texts engage with moral self-cultivation or ponder the dichotomy of Heaven and man. As far as their form is concerned, the Gudin One texts include some that contain long and continuous disquisitions of a philosophical nature; others are only one or two statements in length. The broad variety of philosophical texts epitomises the wide range ofsometimes conflictingtextual materials and diverse philosophical activities during the Warring States period. It minimises the danger of presenting only a one-sided picture of text and thought in mid- to late Warring States philosophical discourse. In this respect, Gudin One provides a solid framework within which to work on early texts. Texts from a less well-documented environment, such as those from the so-called Shnghai collection of Chu manuscripts, about which it is known only that they were purchased at an antique market in Hong Kong, do not offer such opportunities.7

5 It is clear that, in comparison to the Western world, the intellectual realm of early China came up with a very different set of ideas concerning the cosmic order, the human being, and the ideal social environment. This has led many Sinologists to doubt whether it makes sense at all to speak of philosophy when talking about the history of thought of early China. I hold that in the ancient Chinese context there is indeed something that should be termed philosophy. It reflects the attempt at being one with the world, in either a reasoned, a spiritual, or a metaphysical way. For the most part, this attempt is strongly performance oriented rather than a mere theoretical exercise. The one theme that features perhaps most prominently in the materials under review is the issue of moral self-cultivation. For a critical engagement with the question of whether China does have philosophy, see A. Cheng 2005 for further references. See also Defoort 2001, 2006a. 6 Since the foundation of the Peoples Republic of China, thousands of tombs have been excavated in China. According to Png Ho (1999c, 23), more than five thousand tombs of Chu provenance were unearthed in modern Hbi and Hnn alone. Pin Yuqin and Dun Shn (2003) provide basic information about important textual finds between the years 1900 and 1996. Enno Giele (2001) provides a convenient overview of the various tomb finds in China (although his site needs another update). For a detailed account of archaeological evidence for early China as a whole, see Falkenhausen 2006. 7 The Shnghai collection of Chu manuscripts was acquired by the Shnghai Museum in 1994. It contains some 1,200 inscribed bamboo strips. Since 2001 the Shnghai Museum has been publishing these strips, and so far, volumes 17 have appeared. Bought from dealers at an antique market in Hong Kong, the provenance

introduction

Because these texts were found in a tomb, they can be located fairly precisely in time and space. We know, accordingly, that the textual materials from Gudin One, together as a group, formed one part of the tomb assemblage. As I have argued elsewhere, the texts came to us as one set.8 For the purpose of this argument, it is therefore irrelevantalthough deeply regrettablethat the tomb was looted at least twice before archaeologists from the Hbi Province Museum decided to carry out the rescue excavation.9 Even though the looters destroyed parts of the tomb assemblage and may even have taken an appreciable number of inscribed bamboo strips from the tomb, we can nevertheless rule out that they added furtherfakestrips to the assemblage of Warring States manuscripts. People enter tombs for material gain, not to hoodwink the historian of early Chinese thought. Methodologically, the group of texts exhumed from Gudin One can therefore be considered a closed set of manuscripts. It is a tomb corpus defined exclusively by its locus, that is, the tomb Gudin One, and not by the tomb occupant. In this light, the common term tomb library is revealed as misleading.10 Unlike the word library, tomb corpus as

of these manuscripts remains uncertain. After they were made publicly accessible, it was repeatedly assumed that the strips came from a site close to Gudin One or even from the same tomb (see, e.g., Ma Chngyun 2001, 1:2). The assumption that they might have come from Gudin One is based on three observations: first, the chronological proximity of the appearance of these strips and those from Gudin One; second, the overall style of calligraphy in which the strips are inscribed; third, the similar philosophical orientation of the texts. Despite the similarities between the strips from Gudin One and those from the Shnghai collection, I argue against the assumption that the Shnghai strips were originally taken from Gudin One. I do so on the basis of two observations. First, the strips from the Shnghai collection are exceptionally long (measuring up to 57 cm) by Gudin One standards. Second, whereas neither of the two collections of manuscripts displays an internal overlap of texts, they each yield an instantiation of the texts Z y (Black Robes) and Xng z mng ch/Xng qng ln / (Nature Derives from Heaven/Treatise on Nature and Sentiment), and it would be highly unlikely for the tomb robbers to have made such a clear-cut selection of manuscripts that were, when found, in disarray. Thus, it is most likely that the two caches of bamboo strips come from different sites (but probably from the same area). 8 See Meyer 2009, 830. The importance of the contexts provided by tombs as a referential framework has also been discussed by Kern (2002). See also the important discussion on looted artefacts by Renfrew (2000). 9 Tomb looters tried to force access to Gudin One in August 1993; they tried again in October of the same year, and this time they were successful. 10 On the problematic issue of the correlation of tomb and contents and the consistency of tomb equipment, see also Friedrich 1996; Gentz 2006a. For an interpretation of the early Chinese burial system based on Mawngdu Three, see Poo Mu-chou 1998.

introduction

here defined connotes no a priori connection of the exhumed texts with theunknowndeceased. Talking about a library inevitably brings to mind the tomb occupant. This introduces a subjective element relating to the selection of texts that is difficult to justify. Because it is still unclear why these texts were included in the tomb, it is methodologically important to define the tomb corpus exclusively by its textual contents in the context of the tomb. This approach permits evaluation of these materials and their different strategies of constructing meaning even if one were to hypothesise that the philosophical texts from Gudin One were used as mere burial objects and therefore were not read (or selected) by the unknown deceased whom they accompanied. Gudin One dates from the mid- to late Warring States period. Most scholars believe that it was sealed around 300 BC.11 This gives a fairly precise terminus ante quem for the composition of this group of exhumed texts, which date to before the institutionalising of thought during the Qn (ca. 221210 BC) and Hn (ca. 202 BCAD 8; 23220) empires.12 The palaeographic materials from Gudin One therefore give the historian of early thought a glimpse of philosophical texts before they were alteredor even suppressedby later hands. In this respect they differ from texts for which we lack a precise terminus ante

11 For a discussion of the date of burial, see Cu Rny 1997, 1998; Lu Ynhun 1999; Png Ho 1999a, 1999b, 1999c; L Xuqn 2000a, 2000c; among many others. Wng Baoxun (1999) is a rather isolated voice in that he believes that Gudin One could have been closed as late as 227 BC. The structure of Gudin One is typical of a mid- to late Warring States tomb, as suggested by comparison with other sites from this period. See, e.g., Tomb Two from Boshn , Jngmn, Hbi Province, henceforth Boshn Two (Hbi shng Jngsh til kaogu du 1991); Tomb 245 from Yutishn , Jinglng , Hbi Province (Hbi shng Jngzhu dq bwguan 1984); and Tomb Dngyng Zhojih , Hbi Province (Hbi shng Ychng dq bwguan 1992). The terminus ante quem of Gudin One is probably the conquest of Yng in 278 BC. It is generally assumed that the structure of (aristocratic) tombs changed drastically after the assault of Yng by invaders from Qn under General Bi Q (?257 BC) in 278 BC. (See also the discussion in Wng Baoxun 1999, 366367, which also summarises other scholars positions.) Because Gudin One is perhaps slightly later than Boshn Two (based on a dated inventory strip, Boshn Two was possibly sealed sometime between 323 [or 322] and 316 BC), Gudin One can be dated between 323 (or 322) and 278 BC. For a discussion of the date of burial of Boshn Two, see Png Ho 1999c, 24; L Xuqn 1999d, 13; Li Bnhu 1991. 12 For a detailed discussion of changes in intellectual climate following the Warring States period, see Petersen 1995; Kern 2000, 184ff., with further references. There are, however, also arguments claiming that the influence of imperial patronage after 221 BC may be overstated. See Nylan 2009, with further references.

introduction

quem, where there is good reason to assume editorial interference. However, because these authors adopted archaising styles, the different chronological layers can hardly ever be established with certainty. Unlike transmitted texts, the tomb corpus of Gudin One provides a rare opportunity in that it offers an immediate view into the structure of the philosophical text of the Warring States period. Unprecedented insights into the intellectual debate of a formative period of Chinese philosophical thinking, reading, and writing can accordingly be gained. In summation, Gudin One makes possible a qualitative study of text and thought of the Warring States period with unprecedented methodological coherence. To explore the habits of composing and using philosophical texts in early China calls for a methodological distinction between text and manuscript. I define text as the textual matter transmitted. It is the formulation of an idea that can take both oral and written form, and so it is abstracted from any material carrier.13 A text can therefore travel orally and so independently of material contexts, either with teachers, experts, or advisors or via trade routes or at markets, from person to person. Manuscript is the material textual representation, that is, the physical manifestation of a text on silk, bamboo, wood, or the like. To confuse themas happens so often in studies that explore newly received palaeographical materialsinevitably leads to monocausal lines of argument and hence to a distorted picture of historical reality. I shall refer to the receiver of a texts message as the recipient. This allows for the fact that his or her role is far more complex than that of a reader or (anonymous) audience. Recipient can denote an individual but also anot necessarily self-aware or well-defined group of people exposed to these texts. The Tomb Corpus of Gudin One The tomb corpus of Gudin One contains 804 bamboo strips, of which 730 are inscribed. Overall, these carry some 13,000 characters,

13 This definition corresponds to that given by Konrad Ehlich (1998; see also Ehlich 1982), who defines text in a sense that comprises the everyday mundane category but in such a way that it does not need to be (entirely) written in nature. Text can also appear in oral form or, as Martin Kern puts it, co-exist in both (2005b, 293, n. 1). Text, furthermore, does not denote any utterance but is an identifiable entity.

introduction

written in calligraphy with obvious Chu characteristics.14 Using various criteria, scholars group the materials into a varying number of individual texts. The texts are written on strips of dissimilar length. On the basis of physical evidence, six different groups can be distinguished. The first contains manuscripts comprised of strips that are 32.332.5 centimetres in length. The second group contains strips that are 30.6 centimetres long. The third contains strips that measure between 28.1 and 28.3 centimetres. The fourth group contains strips of 26.426.5 centimetres; the fifth and sixth groups contain strips that are decidedly shorter: 17.217.5 and 15.115.2 centimetres. The physical characteristics of the bamboo strips are in fact a good indicator as to which were bound together as a group and, potentially, even as one manuscript. But the physical length of the strips is no indicator of the importance of the text written on them. As will be discussed in this study, during the Warring States period there was no correlation between a text and the length of the bamboo strips on which it was written. The strips do not reflect the status of the text recorded. It is only in the imperial context of the Eastern Hn (AD 25220) that statements finally appear which mention a correlation between the status of a text and the length of the bamboo strips on which it has been written.15 It is likely that, in the Imperial Libraries, the need arose for systematic ways of storing texts and records. In such a context, texts and manuscripts almost certainly began to take on a fairly fixed form. From this resulted a new notion linking the status of a text and its material carrier, as described by the Eastern Hn authors Wng Chng (AD 2797) and Zhng Xun

14 Two texts deviate from this standard. It has been observed that the calligraphy of the so-called Zhng xn zh do and the Tng Y zh do shows a particular style that differs from that of the other materials. L Xuqn (in Allan and Williams 2000, 178) even goes so far as to assume that the calligraphy on the bamboo strips of these two texts is probably not written in Chu script at all. As will be discussed in chapter 1, this interpretation probably overestimates the differences in the calligraphy of these texts. 15 In his preface to Chnqi Zuo zhun zhngy (7a), Zhng Xun notes the length of two feet four inches for the Classics, of one foot two inches for the Xio jng , and of eight inches for the Lnyu . (All lengths refer to Hn Dynasty measurements.) Two feet four inches corresponds to 55.44 centimetres; one foot corresponds to 23.1 centimetres. See Twitchett and Loewe 1986, xxxviii. According to Wng Chng , the sayings of the ancients were written on tablets of two feet four inches. See Tsien 2004, 116.

10

introduction

(AD 127200). But the situation of the Warring States was different. The Warring States period was characterised by a gradually developing manuscript culture in which predominantly oral texts were occasionally written down on bamboo and, sporadically, on wood and even on silk.16 There are no indications of strictly organised methods of recordkeeping similar to those implemented by the Eastern Hn.17 It should therefore not come as a surprise that none of the texts reconstructed from the corpus of Gudin One were written on strips that conform to the length described by Eastern Hn authors. Whenever different manifestations of a philosophical text from the Warring States come to light, they take quite different physical forms. The physical variations among the strips in pre-imperial China, such as differences in their lengths, in how they were cut, and in the style of calligraphy, simply reflect differences in time and space in the production of the manuscripts. These variations say nothing about the texts themselves.18 In this analysis of the texts from Gudin One, two considerations will be of primary importance: the texts strategies of meaning construction and their place in transmission history. Not all the texts from Gudin One are entirely new to us. Some of the textual materials have in fact persisted in transmitted texts to the present daybe it in full, in smaller or greater fractions, or in mere quotations. In one way or another they can still be identified, at least in part, with a transmitted counterpart. Other textual units from Gudin One are familiar to the historian of Chinas intellectual past from other finds of palaeographical materials. These materials may not have survived the transmission process to the present day. However, as different instances of writing down largely identical texts suggest, it can be assumed that some of these texts had at least some importance before disappearing from the surface. Other texts and textual units from Gudin One are entirely new to us. Each of these contributes to our understanding of Warring States intellectual history. They inform us about the nature of a text in early China in terms of the stability or fluidityof certain concepts and text composition and demonstrate
16 On the use of silk for written documents, see Tsien 2004, 129ff.; see also Wnw 7 (1973) and Wnw 10 (1982). 17 I am aware of speculations that archives existed as early as the Shng dynasty (see Falkenhausen 1993, 163164), but claims proposing large-scale and systematic archiving long before the unification of China lack substantial evidence. 18 For a study analysing purpose, form, genre, and possessor of a manuscript, see H Pngshng 2000; Richter 2005, 9293.

introduction

11

that high-level standardisation of texts must have been a rather late (i.e., surely a post-Warring States) development. Those texts that do not have a known counterpart further attest to the broad diversity of texts and intellectual arguments during the Warring States. As mentioned, I basically distinguish between two different strategies of meaning construction applied in the texts from Gudin One, although without a doubt it is possible to describe more particularities and subtypes of meaning construction in these materials. The two types of meaning construction should be understood as ideal types. Ideal type in this context is devoid of any Platonic notion of a perfect thing or phenomenon. Instead, referring to the Weberian concept, it simply denotes the attempt to order the complexity of reality by highlighting certain characteristics of a given object or phenomenon.19 The two ideal types of philosophical reasoning in writing can be summarised as follows. On the one hand, there are those texts that generate meaning by advancing what I call argumentative patterns. Texts of this type aim to be persuasive by establishing their reasoning on the power of good arguments. I call them argument-based texts. On the other hand, there are those texts which I call nonargumentative texts or context-dependent texts. Unlike the other type, the context-dependent texts do not seek to establish argumentative force by virtue of reason. Instead, they largely rely on established and identified authoritiesand hence on contextsfor stating their concerns. It goes without saying that this distinction is not an absolute one, but by accentuating the common characteristics of these materials, it describes two extremes on a continuous scale of texts. Argument in this context should not be mistaken for the concept known from the classical Greek tradition. Argument-based texts do not seek to ascertain truth by applying the techniques commonly seen in Western philosophical discourse. Logical deduction and syllogism are not their characteristic features. Instead, the argument as applying to the kind of texts discussed should rather be described as a pattern that, in its use, generates argumentative force. The recipient should be persuaded to accept the philosophical position presented in the text as good and, accordingly, as something that can be put into practice.

19 For Max Webers concept of ideal types, see his Wirtschaft und Gesellschaft. This work was first published posthumously by his wife in 1921 and 1922 as volume 3 of Grundriss der Sozialkonomik.

12

introduction

Instead of advancing arguments in the sense of building upon logic and epistemology, these texts present philosophical positions of argumentative force. A study of argumentative patterns in early Chinese philosophical texts requires a detailed analysis of the formal structure underlying their makeup to illustrate how meaning is constructed in early written philosophical discourse. The aim is to cast light on the different strategies of philosophising in early China. Yet despite the contribution of a fine-grained analysis of the written remnants of thought to understanding early ideas, the study of text and writing as tools of meaning construction does not always find an equivalent commitment in the study of early philosophyand this holds true in particular for the study of Chinese philosophy.20 The present study aims to correct this picture. Scholars disregard for the manifold strategies of meaning construction as seen in early Chinese philosophical discourse is especially noteworthy since with Chinese philosophy we engage with thought that so often is postulated to be fundamentally different from our own.21 This in itself already calls for a detailed investigation of the various ways of constructing meaning in written discourse. It cannot be the task of a study like this to provide a full description of all the texts from Gudin One and their contents; nor could it be the aim to describe all the features of meaning construction in one individual text exhaustively. This would go beyond the bounds of a study such as this one and call for a project of encyclopaedic dimensions. Instead, by describing the main features of meaning construction in early philosophical texts, I shall show how far a detailed analysis of the strategies for generating meaning in early Chinese written discourse can further our understanding of the philosophical activities of that time. My contribution to the field therefore lies in generating a methodology for engaging with written ideas from early
20 See, however, Vladimir Spirins contribution to this topic in 1976. Unfortunately, this work did not receive the attention it deserved. Spirins main arguments were reiterated in French in 1991 (cf. Behr and Gentz 2005, 7). See further Wagner 1980, 2000, 2003a, 2003b; Behr 2006; Gentz 2006b, 2007a; Meyer 2006, 2007; Schwermann 2006. In September 2009, Dirk Meyer, Joachim Gentz, and Wim De Reu held a conference at The Queens College, University of Oxford, to address exactly this issue. A volume containing some of the contributions, edited by Gentz and Meyer, will be published. 21 Franois Jullien, who sets out to demonstrate the alterit of Chinese thought in many studies, is probably most representative of this position; see especially Jullien 1995. See also Jullien 1989, 1991, 1992. For a critical assessment of his work, see Billeter 2006.

introduction

13

periods, and not so much in describing all the peculiarities of written philosophical texts. The larger questions addressed by this study therefore are as follows. What are the techniques of meaning construction in early Chinese written philosophical discourse? What are the dialectical processes between social communities, on the one hand, and the philosophical text, on the other? And finally, what do texts tell us about the very activity of philosophising in early China? In what follows I provide a brief survey of the individual texts from Gudin One. Texts that are not analysed in detail in the main body of my study will be given more space here. Texts with a Transmitted Counterpart The texts from Gudin One that, in one way or another, survived transmission to the present day all fall into the category of contextdependent texts. These are the Z y (Black Robes) and the texts, written on three different bundles of bamboo strips, that closely resemble the received Laoz , or Do d jng (Classic of the Way and Virtue). Since the Gudin One texts customarily identified as Laoz (as I shall argue below, this is an untenable claim) were collected on three different bundles of strips, they are generally referred to as Laoz A , Laoz B , and Laoz C . The received Z y, for its part, has been incorporated into the compilation L j (Record of Rites). The texts from Gudin One are substantially shorter than their transmitted counterparts and differ appreciably in terms of internal organisation and structure. Those texts that are incorrectly called Gudin Laoz, as well as the Z y, also share close correspondences with other recently discovered palaeographical materials, which suggests that they had at least some significance in early times. The Z y is also part of the Shnghai collection of Chu manuscripts, and a complete copy of the Laoz was excavated in the 1970s from a tomb dating to the early Former Hn dynasty, namely Tomb Three from Mawngdu , Chngsh , Hnn Province (henceforth Mawngdu Three).22 I shall deal with these materials in more detail in part II of this study.

22 See the excavation report in Hnn shng bwguan, Zhnggu kxu yun kaogu ynji suo 1974. For an interpretative transcription, see Mawngdu Hn m bsh zhngl xiao zu 1974. The tomb contained a letter to the netherworld that can

14

introduction Texts with a Counterpart from Other Finds

Texts known from other sites are the so-called Wu xng (Five Aspects of Virtuous Conduct) and Xng z mng ch (Human Nature Is Brought Forth by Decree). A version of the Wu xng was found in Mawngdu Three; a text strikingly similar to the Xng z mng ch is part of the Shnghai collection of Chu manuscripts, in which context it is referred to as the Xng qng ln (Treatise on Human Nature and Unshaped Feelings) by modern editors. None of these texts entirely match their counterparts in Gudin One, but they bear remarkable similarities. The Gudin One Wu xng and Xng z mng ch will be discussed in part I of this study. Texts with No Counterparts The remaining texts from Gudin One were unknown to modern scholarship before the excavation of the tomb. Among these is a cosmogony written on fourteen strips; this text is customarily referred to as the Ti y shng shu (The Ultimate One Gives Birth to Water). The Ti y shng shu is written on bamboo strips that are physically identical with the Laoz C strips. For this reason it is sometimes interpreted as a lost part of an imagined Proto-Laoz. I disagree with this hypothesis. The Ti y shng shu establishes one coherent argument over the entire length of fourteen bamboo strips. The argumentative patterns and the strategies of meaning construction of this text are not in congruence with the so-called Laoz C. My analysis suggests that they should be understood as distinct texts (see chapter 6). The text that is now as a matter of course referred to as the Zhng xn zh do (The Way of Fidelity and Trustworthiness) is

be dated to 168 BC. This should also be the date of burial. Another complete Laoz from the Western Hn was reported just before this manuscript went to press. It was part of a collection of 3,346 bamboo strips purchased by Bijng University in spring 2009 (organisation and transcription work started in March 2009). Preliminary reports state that it is written on some 220 strips and contains roughly 5,300 graphs. Organised into an upper and a lower canon, Laoz shng jng and Laoz xi jng , the text has a d do order. See Bijng Dxu chtu wnxin ynji suo 2009.

introduction

15

a piece of political philosophy that addresses the ruler of a state with the request for moral government. This text, written on nine strips, will head the present study of meaning construction in early Chinese philosophical discourse. A text now called the Qing d y sh (Failure and Success Appear at Their Respective Times) is concerned with the dichotomy of Heaven and man and points out a seeming inequity caused by the power of Heaven over man. The aim of the text is to provide guidance that allows man to deal with lifes imponderables as caused by Heaven. The way in which meaning is constructed in this text will be discussed in chapter 2. Rather similar to the Zhng xn zh do in terms of its compositional structure and thought is the brief text now referred to as the Lu M gng wn Z S (Duke M of Lu Inquires of Z S). Although parts of the text are phrased in the form of dialogues, the main argument is constructed according to the formal patterns seen in the Zhng xn zh do. For this reason I shall not carry out a detailed analysis of the Lu M gng wn Z S in this study. The shape of the eight bamboo strips on which the Lu M gng wn Z S is written and its style of calligraphy are identical with those carrying the Qing d y sh. Nevertheless, the two texts differ markedly in content, structure, and tone. They are distinct texts. The physical similarities suggest that both manuscripts were produced in chronological and geographical proximity, possibly even at the same workshop. The Lu M gng wn Z S records two dialogues, the first between Z S and Duke M of Lu and the second between Duke M of Lu and the minister Chngsn Y , whose name does not appear in transmitted records. Upon a direct question by his lord, Z S offers a definition of loyal minister. The answer displeases Duke M, and he dismisses Z S. The subsequent dialogue between the duke and Chngsn Y perfectly matches what one might call an imagined teaching scene.23 Having listened to his dukes lament, Chngsn Y gives an apologia in support of Z S and his view of what defines a loyal minister. In the course of this it becomes apparent that Chngsn Y not only shares Z Ss view but, in his praise of Z Ss ideal, puts it into practice.

23

Lewis 1999, 5363.

16

introduction

The previously unknown text now referred to as Tng Y zh do (The Way of Tng [Yo] and Y [Shn]) is very similar to the Zhng xn zh do in its strategy of constructing meaning. Moreover, the physical characteristics of the twenty-nine strips on which this text appears strongly resemble those of the manuscript carrying the Zhng xn zh do. They are almost the same length (28.1 and 28.3 cm respectively) and are both cut evenly on each side, and the calligraphy on the strips is almost identical. However, as holds true for the relationship between the Lu M gng wn Z S and the Qing d y sh, the physical similarity of the two manuscripts does not say anything about an intellectual connection between the Tng Y zh do and the Zhng xn zh do; it only suggests, again, that the manuscripts were produced in chronological and spatial proximity and were possibly even manufactured at the same workshop. The Tng Y zh do basically follows the compositional patterns of the Zhng xn zh do butjust like the Lu M gng wn Z Sin a less strict manner. I shall therefore not carry out a detailed study of this text here. The Tng Y zh do promotes abdication.24 The texts programmatic argument contains three main positions. It notes a tension between proper rule and filial piety inherent in the very act of abdication. This polarity is carried to its extreme in the texts slogan of loving the related versus venerating the capable, or, in other words, filial piety versus the act of resigning voluntarily for the sake of someone more accomplished. As the Tng Y zh do puts it, abdication is a sagacious issue and a matter of rightness. Filial piety, on the other hand, is the utmost expression of benevolence. By referring to Yo and Shn the two legendary rulers who were the personification of filial piety and rightnessit annuls the ostensible conflict between the two positions. For the authors of this treatise, filial piety is in itself a fundamental aspect of a rulers unsolicited ascendancy, and only when being filial can a ruler love the entire people, which for the Tng Y zh do equates with the highest expression of a rulers

24 On the basis of three newly excavated texts, among them the Tng Y zh do, Yuri Pines (2005) elegantly reconstructs the evolution in views on abdication during the Warring States period. For a detailed discussion of the tension between family values and abdication, see Oyng Zhnrn 2002. For a discussion of the texts philosophical affiliation, see Jing Gunghu 1999 and, slightly emended, 2001; L Lng 1999, 497; Dng Sxn 2000a, 382; Wng B 2001; Png Bngbn 2004; Defoort 2006b.

introduction

17

rightness. Taking the texts position to the extreme, how could the very act of loving the entire people be expressed more than by having them ruled by an accomplished person, for whose sake one should abdicate? Seen from this perspective, filial duties and abdication are no longer contradictory positions. Ideally, the two belong on one and the same plane, and so the text holds that only by abdicating can a ruler act in accordance with the virtue of filial duty. This notion leads to a second claim: that abdication is required for a positive transformation of the people. Promoting the idea that the ruler of a state has all the means of transforming the environment by virtue of his model of proper conduct and virtuous behaviourin some sense this resembles the Weberian concept of charisma25the message of the Tng Y zh do comes close to ideas put forward in the Zhng xn zh do and bears a strong resemblance to a position set out in the Mngz that states that a great man is one who rectifies himself and the world is rectified.26 Lastly, the Tng Y zh do argues that abdication is a necessary precondition for a rulers well-being. Physical deterioration is a natural phenomenon. Every human beingincluding the ruler of a statefaces this at a certain point in life. But before reaching the point of excessive weakness, the ruler should resign. By promoting this as a standard, this piece of politico-ethical philosophy generates a general pattern of abdication27 that is quite innovative among those texts that subscribe to this politically delicate and historically problematic position.28 To the political world of those days, history had shown the disastrous outcome to which abdication might lead. In 314 BC, King Kui of Yn (r. 320314) resigned in favour of his minister Z Zh . The leaders of the various states condemned his move as an acute violation of political propriety. The ruler of Zhngshn took advantage of this incident to invade Yn and conquer part of

For a reading of d as charisma, see Unger 2000, 113. (Mngz zhng y 1992, Jn xn shng 13.19, 532). Pines 2005, 258. 27 Pines 2005, 261. 28 Other early texts, such as the Mngz and the Yo dian chapter of the Shng sh , for instance, stress the uniqueness of abdication from Yo to Shn, thus ruling out a systematic yielding as envisioned by the authors of the Tng Y zh do (Pines 2005, 271, 27475).
25 26

18

introduction

its territory.29 Therefore, a text which uncompromisingly declares the advantages of abdication might seem peculiar. Yet, as early as 1989, Angus C. Graham had already argued that the recurring topos of an abdication myth in transmitted Warring States texts can only reflect the tip of the iceberg of a larger programmatic discussion.30 Three long but highly corrupt texts fall outside the present study of meaning construction in the philosophical discourse of the mid- to late Warring States. These are the so-called D chng (The Great Constancy; also referred to as Chng zh wn zh ), Shang xng (Reward and Punishment; also referred to as Zn d y ), and D y (Virtue and Rightness; also referred to as Li d ).31 The internal organisation of the strips on which the texts are recorded is far from being resolved, and even the proper association of bamboo strips within these units is anything but clear.32 Under these circumstances, it is possible to reconstruct with confidence only bits and pieces of these units. But nothing can be said with certainty about these units as meaningful wholes. Since a study of their argumentative strategies would therefore be highly speculative, I have excluded them from this study. Also falling outside my scope are the materials generally referred to as Yu cng (Collected Sayings), which modern editors have divided into four different texts.33 The Yu cng differ in many respects from the other texts from Gudin One. To begin with, they are written on bamboo strips that are decidedly shorter. The longest strips from Gudin One are around 32.332.5 centimetres in length. Laoz A, Z y, Wu xng, Xng z mng ch, Chng zh wn zh,

29 This is recorded in a bronze inscription (see Hbi shng wnw ynji suo 1995, 1:341369). See Pines 2005, 269. 30 Graham 1989b, 293. Pines (2005) subscribes to Grahams view. Finds of similar texts from the Shnghai collection of Chu manuscripts further corroborate their observation. See the newly discovered texts Z Go and Rng Chng sh . See Ma Chngyun 2001, 2:3347, 181199 (for the scans of the strips and transcription of the Z Go), 139146, 247293 (for the scans of the strips and transcription of the Rng Chng sh). 31 See Chn Wi 2003. 32 As far as I can determine, Gu Y was the first to propose a largely new arrangement for the strips of the Chng zh wn zh (just after the bamboo-strip texts were made public in Hbi shng Jngmn sh bwguan 1998) and started the ball rolling for a discussion of the three units. See Gu Y 1998; see also Ln Sqing and Zhu Fngwu 1999; L Lng 1999; Wng B 2000; Chn Wi 2000a, 2001, 2003; L Xuqn 2000b; G Shkao [Scott Cook] 2006. 33 See Hbi shng Jngmn sh bwguan 1998.

introduction

19

Zn d y, and Li d are written on strips that fall into this category. There are 331 bamboo strips in total, which equates to approximately 45 percent of the inscribed materials from Gudin One. There are also strips of circa 30.6 centimetres in length, on which the Laoz B, with its 18 bamboo strips, is written. The thirty-eight bamboo strips bearing the Zhng xn zh do and Tng Y zh do are 28.128.3 centimetres in length. The manuscripts carrying Laoz C, Ti y shng shu, Lu M gng wn Z S, and Qing d y sh, altogether 50 bamboo strips, measure 26.426.5 centimetres. The four Yu cng texts, in contrast, are written on bamboo strips that are either 17.217.5 or 15.115.2 centimetres long, and so their physical appearance already marks a break from the rest of the tomb corpus. But not only the physical manuscripts carrying the Yu cng differ from the rest of the tomb corpus. The four units, which modern editors have arranged from the two groups of bamboo strips,34 also differ notably from the other texts in style and content. Only the particular unit now labelled Yu cng Four can in fact be thought of as something like a coherent text. Yu cng Four generates narrative patterns of the kind seen in later author texts such as, for instance, the Xnz . Yu cng Four is inscribed on twenty-seven bamboo strips that are circa 15.115.2 centimetres in length, on which 402 characters remain (the manuscript is corrupt, with possibly three graphs missing). It describes a strategy for surviving at court, as well as for gaining and using influence. It is likely that the text was addressed to advisors at court.35 The remaining texts grouped under Yu cng show no sign of meaningful narratives (as in author texts), argumentative patterns (as in argument-based texts), or authoritative references (as in context-dependent texts).36 Rather, they collect isolated

34 See ibid., 7586 (for scans of Yu cng One), 8793 (for scans of Yu cng Two), 95102 (for scans of Yu cng Three), 103107 (for scans of Yu cng Four). 35 Given that Yu cng Four is more than just a collection of sayings, T Zngli and Li Zuxn (2001, 315) give Yu cng Four a title and call it Shn yn yng xng (Cautious Words and Wise Actions). 36 Note, however, that Christoph Harbsmeier argues that Yu cng One constitutes a series of highly abbreviated results of logical analysis, very close in style to the Dialectical Chapters of the Mjng . Yu cng One should therefore be understood, not as a Yu cng (collection of speeches or sayings) at all, but rather as a truncated jng , for which the shuo explanation part remains to be reconstructed. As such, it would fall into the category of a context-dependent text. Harbsmeiers position is outlined in his forthcoming article Some Philosophical Notes on the Gudin Manuscript Yucng 1.

20

introduction

sayings lacking any coherent synthesis. Many of these sayingsor possibly brief ideasalso appear in other texts of the tomb corpus. L Xuqn assumes that the Yu cng had some kind of teaching purpose.37 Robin Yates hypothesises that the sayings address specific philosophical themes, which is why he compares them to the Sh yn chapter of the Guanz .38 But whereas the authors of the Sh yn establish consistent unitsnotwithstanding their formulaic appearancenothing like this seems to apply to the Yu cng. The philosophical or educational function of these texts remains far from evident. Since they collate different notions instead of synthesising them into coherent accounts, they may have served a predominantly mnemonic function. Authors versus Editors Not much is known about the authorship of early Chinese philosophical texts. As will be demonstrated in this study of text and meaning construction, the complex interaction between the oral and the written should make us rethink the conventional idea that standardised philosophicaltexts already existed in the pre-imperial period. The interplay of the spoken and the written, as well as the fusion of various traditions within one text, precludes the existence of a single identifiable author for the different texts. The fact that the so-called transmitted texts should in effect be regarded as rather late standardisations of previously much more fluid texts, characterised by the complex relationship between the spoken and the written, in itself reduces to absurdity any meaningful attempt to (re)construct an imagined urtext.39 It follows that attributing these rather fluid entities to an individual author is likewise highly speculative. This has led Sinologists to talk about editors instead of authors in their study of text and authorship of presumably ancient philosophical works.40 This concept of an editor of ancient works was developed

See Allan and Williams 2000, 179. Ibid. 39 Martin Kern (2002, esp. 148ff.) deals with this issue in some detail. 40 The question remains as to how far one should speak of ancient transmitted texts with regard to their contemporary editions.
38

37

introduction

21

parallel to the use of the term in related fields, such as Greek philology and biblical studies. The result of abolishing the neat distinction between author and editor (or redactor) in recent scholarship and elevating editors to the status of authors is a changed understanding of the role of an editor compared with that of an author. Even evidence for interpretative additions in ancient texts does not suffice to attribute them automatically to ancient editors. The role of editors is to identify interpolations as corruptions and to remove them.41 This clearly qualifies the Alexandrian scholars of the Hellenistic period as editors. The same is true for those scholars who were working under the guidance of Li Xing (798 BC) and Li Xn (46 BCAD 23) in the Imperial Library. After receiving the order in 26 BC from Emperor Chng of Hn (r. 327 BC) to collate the texts from the Imperial Library, groups of scholars under their direction (after Li Xings death, the editorial project was continued by his son Li Xn) compared various versions of texts, identified what they believed were interpolations, and removed thesein their eyescorrupt passages from the texts.42 Their work matches excellently the definition of to edit as to restore (something to its former condition).43 As discussed at length by B. A. van Groningen, the modern word edition derives from the Greek ekdosis (editio in Latin) and is related to the verb ekdidonai , which is used in situations in which any kind of belongings are handed over to someone elses control.44 Applied to a text this means that the author no longer has control over the text, and it is left to the whims of those who wish to abuse it.45 Thus, in a strict sense, edition (ekdosis) simply denotes the written version of an ancient text since there was no such thing as copyright or a fixed text. Differences in the various manifestations of an ancient text, such as in the Wu xng from Gudin One and
See Seters 2006, 3. Notes were made on each text that was edited and were recorded in the Bi l (Separate Records), most of which is now lost. Li Xn (46 BCAD 23) produced the Q l?e (Seven Summaries), the basis of the later Ywn zh (Records of Arts and Letters) in the Hn sh (Documents of the Hn). For the remaining parts of the Bi l, see Yo Zhnzng 1899. Zhng Shnhu and G Sh provide excellent studies on the Ywn zh. See Zhng Shnhu 1990; G Sh 1987. See also Lewis 1999, 325332. 43 Seters 2006, 14. 44 See Groningen 1963, 5. 45 Ibid., 5.
42 41

22

introduction

that from Mawngdu Three, should therefore not automatically be attributed to editors. They are not authorised versions of a given text but are isolated instances of writing down a fully coherent entity that was not yet fixed in ancient days. The fact that in antiquity written objects were subject to change is not a sign of editorial activity,46 and therefore, ascribing changes (such as additions, omissions, a different organisation) to editors is anachronistic. I shall therefore refrain from talking about editors when dealing with (the changes in) texts prior to imperial attempts at canonising history and thought.47 In this study of text and meaning construction in early China, I shall designate those who produced a text as authors. But authors in antiquity were substantially different from authors in the modern sense. Authors in antiquity were not artists. Their works were not copyrighted. Crucially, the authors of antiquity remained anonymous. Travelling between the states and being truly cosmopolitan, they were part of a productive intellectual tradition in which the ideas they worked with were the collective property of an lite community. Ideas were taken up, changed, and used purposefully according to the needs of the moment. Hence, texts from the early period should in general not be ascribed to a single author. Texts of antiquity were synchronic artefacts and the work of many authors. Leo Oppenheimer coined the expression stream of tradition in reference to literary production in cuneiform Mesopotamia.48 Karel van der Toorn applied the term to the production of the Hebrew Bible by scribes.49 Stream of tradition may also be a useful concept in the present context, whilst keeping in mind the differences between ancient scholarship in the Near East and the productive period of an evolving manuscript culture during the Warring States.

See Seters 2006, 22. Around 213 and 212 BC under the Qn (221210 BC), and again in 136 BC under the Hn , measures were takeneither condemned or celebrated by later generationsto unify ideas. On the execution of scholars and the burning of writings in 212 and 213 BC, see Bodde 1986, 95; see also Neininger 1983; Petersen 1995; Kern 2000, 183ff.; Csikszentmihalyi and Nylan 2003. For a comparison between the measures taken in 212 and 136 BC, see Kern 2000, 191ff. 48 Oppenheimer 1977, 13. 49 Toorn 2007.
47

46

introduction Odes (sh) and Documents (sh)

23

In the present study, I shall not treat pre-Qn sources as if they were well-defined, consistent bodies. This applies to works such as sh odes and sh documents, which are habitually referred to as the Book (or Classic) of Odes and the Book (or Classic) of Documents. Concepts such as classic or canonisation were absent during the Warring States.50 They evolved only gradually during the Qn and Hn dynasties, when history started to be understood from the perspective of coherent texts rather than predominantly oral traditions. A closer look at excavated sourcesbut also at quotations in transmitted texts and works such as the Ywn zh in the Hn sh , which is based on the Q l?e by Li Xn confirms this suspicion.51 There can be no doubt that the wording of sh as quoted in pre-imperial manuscripts already had some stability during the Warring States period. Martin Kern has shown this convincingly in various studies.52 But the confines of this anthology were still rather fluid. Some odes that are quoted in pre-Qn sources some of which are introduced explicitly as sh (or sh yn in an ode it is said) in the palaeographic materialsare no longer extant today.53 Despite this, during the Warring States the majority of what today are canonised as Odes were remembered and recognised songs among lite groups of the time. The songs played an important role in the identity formation of these groups. For these reasons, when referring to these songs as a group, I shall use Odesinstead of Odes because I do not want to create the impression that they were a fully fixed anthology in the Warring States period as they are today, let alone at any earlier moment in history. The capitalisation of Odes is intended to show that the odes formed a recognisable corpus of high cultural value in early Chinadespite their diffuse boundaries. The sh present an even more complicated picture. In excavated materials from the Warring States, quotations that modern scholars identify as being from Documents are never introduced as sh
50 See also Nylan (2009) on the lack of canonisation of the jng (includes todays Odes and Documents) in early China, which also acknowledges that these texts were still variable in content and form at that time. 51 On the Ywn zh from the Hn sh, see above, n. 42. 52 See Kern 2003, 2005a, 2005b. 53 Quotations of authoritative sources will be discussed in chapter 5. On citations in early literature, see Schmlz 1993.

24

introduction

(documents). Instead, quotations are, at best, referred to by different titles that are now habitually thought of as names of chapters but that may just as well have been names of speakers. The quotations in the palaeographical materials that are now customarily equated with sh also lack overall consistency. This points to a rather late standardisation of these lines. I like to think of the sh as Documentsthat is, as an invented tradition that took shape during the Eastern Zhu in an attempt at cultural self-legitimisation and that only gradually hardened into a well-circumscribed anthology.54 Having lost the Zhu capital in 771 BC, the political lite of the Zhu were forced to flee to their eastern capital at Luyng . It must have been a sudden move. Bronzes and other artefacts were left behind, and the Zhu were forced to reinvent tradition based onidealisedmemories of the past as well as on faith in their future.55 But just as in the case of Odes, I refer to this body of quotations and text fragments as Documents since they referred to a shared cultural lore of authoritative value. Certainly by the mid-Eastern Zhu period they were recognised by the lite groups of that time as standards of knowledge and had become necessary tools for participating in the cultural sphere.56 Organisation of the Book This book consists of three parts. Part I provides a detailed form analysis of the argument-based texts from Gudin One. I hope to show that approaching early texts from the perspective of their formal structure is methodologically important for dealing with the written ideas from the early periods. Part II discusses the intellectual environment of the written materials from the Gudin One corpus. The focus of the discussion is on the relationship between material conditions of text, manuscript culture, and writing, as well as on the strategies of meaning construction and philosophising, in the Warring States

54 See also Vogelsang (2002) on the Go chapters in the Shng sh. Vogelsang raises the possibility of a late Western Zhu or early Spring and Autumn date for the early layers of the Shjng (and Shjng). Martin Kern (2007b) raises similar concerns. 55 Haar 2009, 32, 38. 56 For further discussions on the sh, see Matsumoto Masaaki 1968, 520; Chn Mngji 1985, 1135; Li Qy 1997, 424; Lewis 1999, 105109; Schaberg 2001, 7280; Vogelsang 2002; Kern 2005b, 297, n. 9; among others.

introduction

25

period. Part III provides philological justification of my reconstruction of the texts considered in this study. The study of formal patterns necessarily differs from the study of grammar in early Chinese texts, as undertaken, for example, by Hans Georg von der Gabelentz (18401893);57 Angus C. Graham (1919 1991), with his many articles on grammar and logic, was a trailblazer in this and many other areas of early Chinese studies;58 and Christoph Harbsmeier, who can be seen continuing this line of research.59 A form analysis of written philosophy as conducted in this study must describe grey zones in a way that a study of grammar does not need to do. Another way of studying formal patterns is what has been termed xi c , that is, the use of proper words. This probably comes closest to the ars bene dicendi, the art of good speaking, as expressed by Quintilian.60 One could also conduct a study of parallel features in Chinese prose as was done by John Francis Davis (17951890),61 von der Gabelentz,62 Gustave Schlegel (18401903),63 Yoshikawa Kjir (19041980),64 and Rudolf G. Wagner. Wagner was arguably the first scholar to focus on the philosophical relevance of parallel features in Chinese texts.65 But whereas the study of formal features in (early) Chinese texts is largely concerned with the semantic precision of the terms used, my study of meaning construction in Warring States written philosophical discourse has different aims and goes much further than that. I focus on, not the lexicon of a text, but the larger structures underlying the
See Gabelentz 1881, 1883, among other works. See Graham 1972, 1973, 1977, 1978a, 1978b, 1989a, among other works. 59 See Harbsmeier 1981, 1998. Here one should also mention the achievements of scholars such as Wolfgang Behr, Chou Fa-kao, and Robert Gassmann, among many others. 60 Quintilian quoted from Rahn 2006, 2:14, n. 5. 61 As early as 1830, John F. Davis had already pointed out the importance of parallelism in Chinese composition. 62 See Gabentz 1878. 63 In 1896 Gustave Schlegel (18401903) published his translation of the preface to the D Tng xy j , in which he set out to demonstrate the importance of parallelism in Chinese prose. For a critical evaluation of Schlegels translation, see Zach 1902. 64 Yoshikawa Kjirs study (1953) on parallelism in the Laoz deepened the study of parallel structures in early Chinese texts. 65 See Wagner 2000, 56. Before his trilogy (2000, 2003a, 2003b), Wagner had already articulated his ideas on parallelisms in Chinese philosophical texts in several articles. See Wagner 1969, 164167; 1980; 1999b. Gentz (2006b) provides a long-overdue summary of the Western reception of parallelism in Chinese writing.
58 57

26

introduction

makeup of a philosophical text as a whole. Thus, unlike the studies that mainly focus on the analysis of what classical literature calls a zhng , or section, my study is concerned with the makeup of a philosophical text as a whole, and hence my focus is on the cultural phenomenon of the writtenphilosophicaltext as well as on the practice ofphilosophicalwriting in the larger context of Warring States intellectual history. The Three Parts of This Book The form analysis of the Zhng xn zh do introduces part I of my study (chapter 1). It describes the means by which an argument-based text like the Zhng xn zh do constructs meaning. This analysis prepares for a more general definition of argument-based text itself. As argued in this chapter, the composition of this piece of Warring States politico-philosophical written thinking should be described as a weblike structure that functions as an important constituent for conveying meaning. The formal structure not only is a technical device to present the argument in a coherent manner but also reflects its contents and thus highlights the philosophical concern of the text that the rulers sovereignty should expand everywhere, including the cosmic spheres of Heaven and Earth. The Zhng xn zh do can be considered an example par excellence of the correlation between form and content in early philosophical argument-based texts. Chapter 2 deepens the discussion of argument-based texts in early China and provides a detailed analysis of the form and content of the previously unknown Qing d y sh. As the subsequent discussion will show, even though the Qing d y sh is organised in a radically different fashion from the Zhng xn zh do, its structure nevertheless fulfils the same integrating function as is seen in the former text. Its compositional structure is a device that deliberately mimics the socalled logical structure of the argument that the text aims to transmit. It is a philosophically relevant element of meaning construction. Chapter 3 provides a detailed study of the compositional structure of the argument-based text Wu xng from Gudin One and a comparison with its manuscript counterpart from Mawngdu Three. The comparison casts light on their astonishing similarities and at the same time reveals that they still differ appreciably in terms of internal organisation. It seems to be the general consensus to attribute these

introduction

27

differences to conscious changes made by later editors. This is quite certainly wrong and neglects the nature of a text during the Warring States period. The differences in the two manifestations of the Wu xng simply reflect different instances of writing down a predominantly oral text. Instead of overemphasising the differences between the two instantiations of the Wu xng, it is more instructive to look at the analogies between the two. I contend that, in what matters, the two manifestations of the Wu xng are not so different at all but together advance one coherent system of thought, which I call the wu xng theory. Chapter 4 closes the analysis of argument-based texts with a study of the Xng z mng ch. This text has a close counterpart in the Shnghai collection of Chu manuscripts. The two manifestations share remarkable similarities in approximately the first half of the texts. The order of building blocks throughout the first half of the texts can be considered fairly fixed. It is possible to describe a coherent philosophical system behind their arrangement. I consider this part of the text its corein philosophical terms also. The latter parts of the texts present a further, more detailed elaboration of the core part and provide some sort of application for the individual of what has been outlined theoretically in the core text. As such, the latter parts, which might originally have been grafted on to the core text, leave more room for textual variation of precisely the kind seen in the two instantiations of the text. Even though the Shnghai counterpart of the Xng z mng ch was bought from antique dealers at a market in Hong Kong, which resulted in the loss of the referential framework for the analysis of this text, I argue that it is nonetheless possible to reconstruct parts of its textual history by reference to the Gudin One Xng z mng ch. Part II of this study positions the argument-based texts from Gudin One in the broader context of the intellectual activities of Warring States lite circles. This part of my study contrasts the techniques of meaning construction in argument-based texts and in nonargument-based texts. Here I generate a descriptive definition for the context-dependent texts (chapter 5). Based on the discussion of the two ideal types of texts, I examine the larger issue of authorship and writing in the intellectual world of the Warring States. On the basis of the preceding chapters, in which I have developed a methodology for dealing with the written remnants of early Chinese thought, chapter 6 discusses the disputed relationship of the materials

28

introduction

generally referred to as Ti y shng shu and Laoz C. By now it seems to be an established consensus to consider the previously unknown materials now referred to as Ti y shng shu as part of an imagined Proto-Laoz. This assumption is based on the fact that the Ti y shng shu shares the same material carrier with the socalled Laoz C. I shall show that, despite their physical relatedness, the two were nevertheless unrelated texts. Chapter 7 considers the relationship of writing and thought and discusses the material conditions for meaning construction in late Warring States philosophical discourse. Here I discuss the implications of the findings for Warring States thought overall and contrast the philosophical background of the two traditions of philosophising as manifested in the two types of meaning construction. In part III (chapters 913) I supply complete translations and discuss the philological issues of my reconstruction of the texts.

PART I

ANALYSIS TEXT AND STRUCTURE

CHAPTER ONE

THE ZHNG XN ZH DO THE WAY OF FIDELITY AND TRUSTWORTHINESS Conventional treatment of texts as mere repositories of ideas does not suffice to establish a sound and historically valid reconstruction of early thought. Hence, my approach is to look at the philosophical texts from early China as meaningful objects in their own right to further our understanding of the social practice of using and composing philosophical texts. This requires a close reading of the texts to cast light on the various strategies used to construct meaning in early Chinese philosophical communication, and thus to put on display the argumentative patterns behind the makeup of these texts. The Zhng xn zh do (The Way of Fidelity and Trustworthiness) proves ideal as a starting point for my analysis. No other text from the tomb corpus of Gudin One is similarly straightforward in organisation. In this respect, it functions as an epitome of what I define as argument-based texts and introduces my methodology for dealing with the written remnants of thought from early China, which I shall develop in the course of this and the next four chapters. Adherence to a pattern gives a text passage a distinctive rhythm and so signals its importance. The Zhng xn zh do discovered in Gudin One is organised according to a pattern of strictly parallel lines that appear in an ababc scheme. This pattern applies to both the micro- and the macrolevel of the text. I use the term microlevel to refer to stable textual units, whereas the macrolevel of composition refers to the text as a whole. By also applying a recurring ababc scheme to the macrolevel of composition, the authors of the text formally establish textual links and thus connect the different notions advanced in the text in a larger consistent scheme. This is a vital strategy to construct meaning in argument-based texts. Crucially, the formal structure behind the makeup of an argument-based text is not only an important device to establish a concise language in philosophical writing but also an essential means of promoting textual stability.

32

chapter one The Text on Bamboo

The Zhng xn zh do is a relatively short text. The manuscript consists of only nine strips, on which 251 graphs are still legible. Of the argument-based texts from Gudin One, only the Lu M gng wn Z S (Duke M of Lu Inquires of Z S) manuscript contains fewer graphs. The bamboo strips of the Zhng xn zh do are for the most part in good condition. The strips are cut evenly at both ends. Their length is 28.228.3 centimetres. As for the style of calligraphy and the shape of the strips, the Zhng xn zh do bears a strong resemblance to the Tng Y zh do (The Way of Tng and Y), which is also part of the same tomb corpus.1 It seems reasonable to assume that the Zhng xn zh do and the Tng Y zh do were copied in chronological or geographical proximity to each other. It is possible that these particular manifestations of the texts were fixed on bamboo at the same workshop. Despite calligraphic similarities, we can nevertheless be fairly certain that the two texts were fixed on different bundles of bamboo strips. This can be judged from the markings on the strips that indicate the previous positions of the two cords, which kept the individual bundles of strips together.2 The bamboo strips on which the Zhng xn zh do has been inscribed are well preserved; the calligraphy is particularly clear. Of the nine strips that constitute this manuscript, only one has broken. The missing part probably contains two graphs. This means that the manuscript as it was placed in the tomb probably consisted of a total of 253 graphs.

1 Qi Xgu and Png Ho, among others, have observed that the style of calligraphy of the Tng Y zh do and the Zhng xn zh do differs from that of the other manuscripts from Gudin One. L Xuqn goes so far as to argue that the two manuscripts are not written in Chu script at all (see Allan and Williams 2000, 178). This position overemphasises peculiarities in the calligraphy of these manuscripts, and it is better simply to note that the calligraphy of the two shows some non-Chu characteristics. 2 The Zhng xn zh do was held together by two cords that were 13.5 centimetres apart; the Tng Y zh do was held together by two cords that were 14.3 centimetres apart. See Hbi shng Jngmn sh bwguan 1998, 157, 163.

the way of fidelity and trustworthiness Thought and Contents

33

Many scholars still use traditionaland anachronistic!labels such as Confucianism or Daoism when classifying remnants of thought from the Warring States period. The Zhng xn zh do does not correspond to any of these categories, which is why scholars sometimes hesitate as to how they want to classify this text.3 This once more demonstrates that the attempt to associate a newly excavated text from the Warring States period with one of the traditional labels of philosophical affiliation is not very instructive and certainly does not further our understanding of the text itself. The problem of using such labels is that they suggest an intellectual consistency that is difficult to justify and, worse, may in fact distort historical reality. Accordingly, I shall refrain from categorisations that were used during the Latter Hn in the retrospective attempt to classify different currents of thought.4 In contrast, I approach the texts from Gudin One by describing the various ways that meaning was constructed in the texts. This necessitates discussing the social background against which these texts were produced. Hence, instead of studying what I think is a methodologically problematicin an almost Platonic senseideal(ised) world of thought by reducing ancient texts to vessels of ideas, I propose to shift the focus to the analysis of the practice of philosophising in the ancient

3 For instance, L Xuqn remarks that the Zhng xn zh do maintains a Confucian tonewhatever this may bebut that its authors should rather be labelled as vertical and horizontal strategists (2000a, 107). L Cnshn argues that the Zhng xn zh do was intended to be studied by the ruler. On this basis, he associates this work with the so-called Zhngling branch of the Confucian school (in Allan and Williams 2000, 253). Lio Mngchn ascribes the Zhng xn zh do to Zzhng , a student of Kongz. He bases this on the fact that Zzhngs sons Shnxing and Z S served Duke M , which is also why the text was found in the Chu region (in Allan and Williams 2000, 253). 4 During the Eastern Zhu (771256 BC), probably only the r and the m constituted self-conscious traditions. As Petersen (1995, 3337) has convincingly shown, the term ji as used by Sma Tn does not mean school at all. Sma Tn did not call these traditions ji but labelled them the r (r zh ) or the m (m zh ). Petersen interprets ji to be individual persons; Michael Nylan (1999, 50, n. 82) understands ji as denoting scholastic lines (not genealogical). On the problematic use of traditional labels when discussing philosophical texts from the Warring States period, see also Csikszentmihalyi and Nylan 2003; Kern 2000, 9. See further Creel 1949; Gardner 1998; Peterson 1988.

34

chapter one

world as it can be judged from material evidence. I deem this to be a much better approach to pre-imperial thought. The Zhng xn zh do is a piece of political philosophy. The text advocates moral government, and it is instructive to see that the Zhng xn zh do directly addresses the ruler of a state, who is called a jnz throughout. The text demands that the ruler act through his own model of proper behaviour. By calling him jnz, which by the time of the midWarring States had already acquired the meaning of gentleman, the Zhng xn zh do plays with the two concepts behind this term: ruler and gentleman. The term jnz as used in the text therefore denotes more than the political (or social) function of a ruler. Everyday reality made it plain that during the Warring States period, a high pedigree did not necessarily mean that a ruler also embodied the persuasive model of proper behaviour as required by the texts programmatic idea of moral government. But according to the authors of the text, being a jnz also implies that the ruler has to behave in a gentleman-like fashion. This means that he should become a person of moral superiority, not just superior by birth. As the Zhng xn zh do argues, being a morally superior person is part of the immanent nature of a real ruler. As a consequence, the ruler must attempt to realise his immanent nature and become a true jnz, that is, a person, who embodies both the high social pedigree of a ruler and the moral superiority of a gentleman.5 In a brief outline, the Zhng xn zh do prescribes how the ruler may achieve this goal. He has to model himself on the natural world, by which he is transformed into the human reflection of the cosmic elements Heaven (tin ) and Earth (d ). To achieve this, he has to establish fidelity (zhng )6 and trustworthiness (xn ) as the only principles of government. According to the authors of the Zhng xn zh do, these two principles are the essence of benevolence (rn ) and the basis of rightness (y ), and they imply a transformative power over the environment of the ruler. Accordingly, the Zhng xn

5 On the concept jnz in Warring States philosophical discourse, see Rubin 1976, 2026. 6 In the Zhng xn zh do, zhng and xn are qualities of the ruler. The concept of zhng here carries the connotation of being true to oneself, so I translate it as fidelity in this text (as opposed to the concept of zhng as used, for instance, in the Lu M gng wn Z S, where it should be translated loyalty).

the way of fidelity and trustworthiness

35

zh do advocates a political agenda that aims to ensure order through moral power. It thus corresponds to what Benjamin Schwartz terms light government,7 in which the ruler over a state acts according to the transformative power of his own example of proper conduct. Structure and Thought The construction of meaning in the Zhng xn zh do takes place not only on the semantic level but also, to a large extent, on the level of its formal structure. The formal structure of the text adds fundamentally to the communication of meaning. The entire Zhng xn zh do is made up of a set of components. On the lowest level, these can be described as a, b, and c components. Throughout the text, the a component contains notions related to fidelity (zhong ), b contains information about trustworthiness (xn ), and c formulates a conclusion based on a and b. This can best be shown by referring to the text itself:
a b c a To hold old ways in high esteem and never render [them] impure is fidelity in its culmination. b To hold antiquity in high reverence and take it as a principle is trustworthiness in its culmination. c The highest fidelity has no pretension; the highest trustworthiness is not perfidious; now this is what this is about.8

This illustrates the elementary structure of the Zhng xn zh do. In most cases, the elementary structure of a single abc scheme is doubled to create an ababc or, to be more precise, 1ab2abc scheme. The second ab group (henceforth 2ab) fulfils two functions. First, it further refines and so concludes the information given in the first ab group (1ab). As such, it functions as the c component to the first ab group. Second, it continues the argument, which then is concluded by the final component, c. We get the scheme shown in figure 1.

7 8

Schwartz 1985, 107. Strips z3/16z4/16.

36

chapter one
2ab concludes 1ab

2ab continues the argument

1a 1b 2a 2b c

Figure 1: The Microstructure of the Zhng xn zh do

Because the aspects of a and b are discussed in overlapping mode before they are combined in a conclusion, c, I refer to this scheme as overlapping structure. The text can be divided into six units according to the abc scheme. William G. Boltz has coined the felicitous term of building blocks for these textual units.9 1. Canto 1
1.1a10 [A]11 1.1b 1.2a 1.2b 1.c 2.1a 2.1b 2.2a 2.2b 2.c 3.a 3.b 3.c [B] [C] [D] [E] [F]

9 See Boltz 2005. I fully agree with Boltz that building blocks are characteristic of early Chinese texts. But for reasons discussed below, I disagree with his conclusion that this implies a composite nature that opposes integral, structurally homogeneous texts (2005, 7071). 10 The underlined number before the dot refers to the building block; the number after the dot refers to the segment; a, b, and c indicate the components. 11 The letters in brackets refer to the philological discussion in part III of this book.

the way of fidelity and trustworthiness 2. Canto 2

4.1a [G] 4.1b 4.2a 4.2b [H] 4.c [I] 5.1 5.2 5.3 5.c 6.1a 6.1b 6.2a 6.2b 6.c {} [][J] [K] [L] [M] -----12[N] [O]

37

[P]13

The following in-depth analysis of the Zhng xn zh do is divided into two parts: first, a description of the formal structure of the microlevel of the text, that is, the arrangement of each building block; second, a portrayal of the macrostructure of the text, that is, the organisation of the Zhng xn zh do as a whole. However, before going into a fine-grained analysis of the Zhng xn zh do, I think it necessary to distinguish the overlapping structure from the interlocking parallel style described by Rudolf G. Wagner.14 Interlocking Parallel Style and Overlapping Structure The overlapping structure and the interlocking parallel style correspond to some degree. Both types split up an argument into two binary matters a and b. In the interlocking parallel style the two matters a and b are joined with each other in a pattern such as 1ab2ab3ab4ab (theoretically this pattern could be continued infinitely). Conceptually,

I have added the lines to show that the jnz (gentleman-ruler) is the subject of both line 6.2a and line 6.2b, which are entirely parallel. 13 I shall give a translation of this passage below. 14 This style is most prominently discussed in Wagner 2000, ch. 3.
12

38

chapter one

the two matters run parallel to each other in the form of two strings: 1a2a3a4a (this could be continued to 5a6a . . . etc.) and 1b2b 3b4b (5b6b . . . etc.). For reasons of simplicity, I illustrate the interlocking parallel style with an example chosen by Wagner himself to discuss what he terms the open type of this pattern:15
1 2 3 4 5 1 He who interferes destroys them. 2 He who holds fast loses them. 3 That is why the Sage 4 does not interfere and thus does not destroy; 5 does not hold fast and thus does not lose.

The silent structure of this passage is as follows:

I 1a He who interferes destroys them. 2b He who holds fast loses them. 3c That is why the Sage II 4a does not interfere and thus does 5b does not hold fast and thus not destroy; does not lose.

The overlapping structure, in contrast, should be understood as a joined parallelism with the characteristic of an argumentative chain. Each step of that chain cumulatively adds to the argument before the argument is rounded off in a conclusion. The second ab unit works in two directions, as it specifies the first ab group and, simultaneously, continues the argument into the final component c. Compare the following overlapping structure:
1a: General statement about a 1b: General statement about b 2a: Specification to 1a 2b: Specification to 1b c: Conclusion to ab

(1c)

When comparing the open interlocking parallel style with the overlapping structure, the differences and similarities between the two types

15

Ibid., 62 (adjusted).

the way of fidelity and trustworthiness

39

of argument construction become clear. Both types discuss two distinct matters, a and b, in a joined fashion. However, whereas the interlocking parallel style isolates the item in separate rows, the overlapping structure assigns a double-directed function to the 2ab unit: the 2ab unit further refines the general issue of 1ab and hence functions as c to 1ab; the 2ab unit then continues the argument on a new level, and the entire unit is brought to a conclusion in the final slot of the pattern, in c. Microstructure Building Block 116
1.1a 1.1b 1.2a 1.2b 1.c 1.1a 1.1b 1.2a 1.2b 1.c [A] z2
z1 Not to be pretentious and not to be destructive is the culmination of fidelity; Not to cheat and not to be cunning is the culmination of trustworthiness. When fidelity is accumulated [by the ruler], [he] can be approached [by the people]; When trustworthiness is accumulated [by the ruler], [he] can be trusted [by the people]. That fidelity z2 and trustworthiness have been accumulated [by the ruler] and the people did not approach and trust [him]there has never been such a case. z1

The formal structure of the first building block reveals both the subject (the ruler) and the object (the people) of the conduct of fidelity (zhng ) and trustworthiness (xn ). As defined in this building block, the conduct not to be pretentious and not to be destructive (: 1a)17 and not to cheat and not to be cunning ( : 1b) equals the highest form of fidelity and trustworthiness

For a philological discussion of the reconstruction of this text and its translation, see chapter 9. 17 Most scholars identify the character hi (to harm) as d . However, the formal structure of this unit requires a negative term here. See chapter 9 under [A].

16

40

chapter one

respectively. The 2ab pair then concludes 1ab by stating: When fidelity and trustworthiness are accumulated (/), then [someone (?)] can be approached (trusted; 2b). The concluding component c reiterates parts of 2ab (accumulating fidelity and trustworthiness []) and then states the result of this conduct: to approach and to trust (/). By taking up the elements mentioned previously, c concludes the entire matter. Furthermore, this component adds another element to the discussion: the people, mn . As component c reveals, the people directly respond to the qualities of fidelity (zhng) and trustworthiness (xn) with trust and by approaching the one who acts on these principles. Accordingly, the subject who displays these qualities is the ruler (of a state), who is named in building block 2. Building Block 2
2.1a 2.1b 2.2a 2.2b 2.c [B] [C] z3 [D]

2.1a The highest fidelity is like the soil; it transforms the things but does not attack them. 2.1b The highest trustworthiness is like the seasons; [they] succeed [one another] and [the circle] does not end. 2.2a Men of fidelity have no z3 pretension; 2.2b Men of trustworthiness are not perfidious. 2.c The gentleman-ruler ( jnz) goes along with this, and therefore [he] does not cheat life, nor does [he] turn his back on death.

The second building block reiterates elements from building block 1, namely utmost fidelity (zh zhng ) and trustworthiness (zh xn ).18 Building block 2 thus explicitly continues the argument of the preceding building block. The overlapping structure relates the terms in two contiguous building blocks.19 In building block 2 fidelity and

1.1a: ; 1.1b: ; 2.1a: ; 2.1b: . 19 First, the 1ab unit of the first building block reads not to be pretentious (b ) and not to cheat (b q ); the same pattern occurs in the 2ab unit of building block two, but the second element (not to cheat) is replaced by are not perfidious (b bi ), so that the two are given equal structural significance, and either of
18

the way of fidelity and trustworthiness

41

trustworthiness are situated in the context of the natural world with the mention of the soil and the seasons. This clarifies the quality of the ruler and implies that he must model himself on nature. The actions of soil and the seasons are nothing other than the mere realisation of their nature, which is causing seeds to sprout in the case of the soil, and returning each to its appropriate time in the case of the seasons. If a ruler conduct himself in accordance with the principles of fidelity and trustworthiness, then he is comparable to the soil and the seasons. His immanent nature prescribes this conduct. Consequently, he has to enact fidelity and trustworthiness even if it seems that he cannot gain any advantage from doing so. Only then can he realise his immanent nature and become a true jnz, a gentleman-ruler. Building Block 3
3.a [E] 3.b z420[F] 3.c To hold old ways in high esteem and never render [them] impure is fidelity in its culmination. 3.b To hold antiquity in high reverence and take it as a principle is trustworthiness z4 in its culmination. 3.c The highest fidelity has no pretension; The highest trustworthiness is not perfidious. Now this is what this is about. 3.a

This building block is a collage of building blocks 1 and 2.21 It is almost entirely composed of parts taken from these units. Building block 3

them can be substituted for the other. Second, the pattern of an overlapping structure also relates terms that appear in one and the same building block. For example, in building block2, the 2ab unit reads have no pretension (w ) and are not perfidious (b bi ). The same pattern occurs in the concluding component c, but the first element (have no pretension) is replaced by does not cheat (b kung ). Conceptually, these notions are given equal structural significance, and one can be substituted for the other, so that conceptual definitions are established in the text. For a discussion of persuasive definitions and the distinction between emotive and conceptualor descriptivemeaning, see Stevenson 1938, 1945. 20 The cross is used to indicate that either the text or the strip is corrupt, or that a graph cannot be identified with certainty. 21 For a discussion of the collage-like composition of early Chinese texts, see also Schwermanns (2006) analysis of the Tngwn chapter of the Liz. Rolf Trauzettel (1979) was perhaps the first to use the term collage (Collage-Technik) in the context of early Chinese text composition.

42
Building block 1

chapter one
1.1a 1.1b 2.1a 2.1b 2.2a 2.2b 2.c 3.a 3.b 3.c [] []

Building block 2

Building block 3

Figure 2: The Collage-like Nature of Building Block 3

thereby summarises their account and concludes the first part of the text, which I call canto 1. See figure 2. Only the first subsets of the ab unit of building block 3 are new.22 This part formulates the central idea of the argument. As can be seen easily from the two elements, building block 3 situates fidelity and trustworthiness in the context of the days of old.23 Except for the formula Now this is what this is about ( f c zh wi c ) in the concluding component, the remaining elements of this argument on fidelity and trustworthiness are taken from the previous account. By combining the different building blocks with each other, the authors of the Zhng xn zh do directly connect building block 3 to the idea established in building blocks 1 and 2. The technique used in building block 3 can therefore be compared to that of developing an overall argument in building blocks 1 and 2. The statement Now this is what this is about in the concluding component c not only concludes this building block with the insight that behaving with fidelity and trustworthiness implies a veneration of the ways of the old but also sums up everything that has been stated so far. It brings the first part, canto 1 of the Zhng xn zh do, to a close.

In the figure, I have pointed this out by putting these elements in brackets. On the identification of the first characters in the b component as d gu to hold antiquity in high reverence, see chapter 9 under [F].
23

22

the way of fidelity and trustworthiness Building Block 4

4.1a [G] 4.1b 4.2a 4.2b z5 [H] 4.c [I] 4.1a 4.1b 4.2a 4.2b 4.c

43

The highest fidelity is not pleasant for [the people]. The highest trustworthiness is not restricted in time. Not pleasant for [the people] and yet providing enough to nourish, such is the Earth. Not to be restricted in time z5 and yet able to be met with, such is Heaven. To be in tune with Heaven and Earth, this is what fidelity and trustworthiness are about.

Building block 4 opens the second canto of the Zhng xn zh do. It posits the virtues of fidelity and trustworthiness within the abstract context of the cosmos and thus functions as a continuation of building block 2. We know already from building block 2 that the rulers fidelity and trustworthiness are the realisation of his immanent nature. Building block 4 now argues that fidelity and trustworthiness display the same qualities as the abstract cosmic elements Earth (d ) and Heaven (tin ), to which the building block connects two further features: to please [the people] (yu ) and the aspect of temporality (q ). The ruler follows the patterns as prescribed by Heaven and Earth. Through his active conduct of fidelity and trustworthiness he can embody their qualities too. This is how he becomes the human reflection of Heaven and Earth. Building Block 5
5.1 5.2 5.3 5.C 5.1 5.2 {} []z6 [J] [K] If [only] kind with words but in fact not acting in accordance with them, the gentleman-ruler (jnz) rather refrains from speaking so. If letting the mind loose, {and yet being}24 z6intimate in appearance, the gentleman-ruler (jnz) rather refrains from displaying this.

24 The curly parentheses indicate missing parts in the text. The reconstructed part is set in italics.

44
5.3

chapter one

If acting according to the old but pleasing the people by serving [them the special taste of] zhng, the gentleman-ruler ( jnz) is one who refrains from relying on this. 5.C As to these three [ways], the man of fidelity refrains from doing them, and the trustworthy man refrains from applying them.

This passage breaks away from the basic pattern of the Zhng xn zh do. The argument from building blocks 14 now shifts from the abstract level to a concrete directive. Building block 5 lists prescriptions for conduct that are directly addressed to the ruler of a state. It should be read as the culmination of the argument. Additionally, segment 5.3 contains elements from both building block 3 (on antiquity) and building block 4 (on the notion of pleasing and nourishing the people), which again recur in building block 6 below (see fig. 3). I have already shown that building block 3 comprises the entire argument of canto 1, but it also concludes it. I have also demonstrated that building block 4 introduces the second canto of the Zhng xn zh do, but elements from building block 4 also reoccur in building block 6. This being so, it becomes evident that building block 5 contains all the argumentative features of this treatise on governmental affairs. See figure 3. Building Block 6
6.1a 6.1b 6.2a 6.2b 6.c z7 [L] z8 [M] ----- [N] [O] z9 [P] When fidelity becomes the z7 way [in the state], then all kinds of skilled labour will not decay, and all the nourishing of the people will be sufficient. 6.1b When trustworthiness becomes the way [in the state], then all groups of things will be completed, and all kinds of good deeds will thus be established. 6.2a Thus, when the way in which the gentleman-ruler ( jnz) conducts himself is indeed z8 of fidelity, [even] the Mn barbarians will come close to and follow [him]; 6.2b [And] when his words are indeed trustworthy, [they] are sincere and can be passed on. 6.1a

the way of fidelity and trustworthiness Concluding element of canto 1 3.a 3.b 3.c 4.1a 4.1b 4.2a 4.2b 4.c 5.1 5.2 5.3 5.c Concluding element of canto 2 6.1a 6.1b 6.2a 6.2b 6.c [] [] [] [----- []

45

Introduction to canto 2

Figure 3: Culmination of the Argument in Building Block 5 6.c Fidelity is the realisation of benevolence (rn); [And] trustworthiness is the basis for rightness (y).

It was for this reason that what [the gentleman-ruler] applied to the Mn and M barbarians in the days of old z9 was something of this kind.

Building block 6 draws a utopian picture of the consequences of following the prescription given in building block 5 (i.e., when fidelity and trustworthiness serve as the leading principles for the reign of the ruler): a reign guided by these principles will lead to a moral transformation of the entire state. Macrostructure The foregoing account has shown that each of the building blocks of the Zhng xn zh do discusses one aspect related to acting with fidelity and trustworthiness:

46
Building block 1: Building block 2: Building block 3: Building block 4: Building block 5: Building block 6:

chapter one
discourse on reign discourse on nature discourse on reign discourse on nature discourse on reign conclusion

On a more structural level, we arrive at the following scheme:

Building block 1: 1a Building block 2: 1b Building block 3: 2a Building block 4: 2b Building block 5: (specification of the argument) Building block 6: c

From this it seems that the macrostructure of the Zhng xn zh do to a great extent reflects the basic pattern of the individual building blocks (microstructure). If this is true, it should be possible to identify each building block with one component of a building block. See figure 4. Based on these considerations and taking the feature of overlapping structure into account, building block 3 can be identified as the 2a component of the macrostructure of the Zhng xn zh do. As such, it functions as the explanatory specification to building block 1, which can hence be seen as the 1a component of the text. Building block 4 should, accordingly, be identified as the 2b component of the macrostructure in that it further specifies building block 2, which is the 1b component of the text. This seems to be a sound assumption because building block 1 (the 1a component of the macrostructure) indeed explains the general implications of fidelity and trustworthiness in their highest form. Building block 3 (the 2a component of the macrostructure) continues on this and then closes by explaining:
3.c [. . .] Now this is what this is about.

In the same vein, building block 2 (the 1b component of the macrostructure) opens up the discourse on nature. Building block 4 (the 2b component of the macrostructure) continues this and then closes by explaining:
4.c [. . .] this is what zhng and xn are about.

the way of fidelity and trustworthiness

47

1A 1B 2A 2B
Transformation of the argument

1.1A 1.1B 1.2A 1.2B 1.C 2.1A 2.1B 2.2A 2.2B 2.C 3.A 3.B 3.C 4.1A 4.1B 4.2A 4.2B 4.C 5.1 5.2 5.3 5.C 6.1A 6.1B 6.2A 6.2B 6.C

[] [] [] [] [] []

Figure 4: The Higher Abstraction of the Zhng xn zh do

48

chapter one

Moreover, segment 2 of building block 5 states that the mind and the appearance of the true jnz never diverge. The same issue reappears in 2a of building block 6 (see fig. 5)note that building block 6 formulates the conclusion in the macrostructure of the Zhng xn zh doto which the explanation is added that this kind of behaviour is the enactment of fidelity (zhng). Segment 1 of building block 5, in turn, addresses the issue that the words (of a jnz) must correlate with his behaviour. The 2b component of building block 6 takes this issue up again, adding to this the positive explanation that this behaviour conforms to the ideal of being trustworthy (xn). Finally, as noted, segments 1 and 2 of building block 5, which prescribe the rulers behaviour, are negatively inherent in the 2ab component of building block 6 (the concluding component of the entire treatise); building block 5, in turn, contains all argumentative features of this essay on governmental affairs. On the basis of this consideration it can be concluded that, by combining the partial definitions from previous building blocks, the following reading applies to the character fidelity (zhng) in the concluding building block 6, as defined by the Zhng xn zh do:
Not to [be] pretentious and not to [be] destructive (1.1A), that is, to hold old ways in high esteem and never counteract them (3.A) and thus never to let [ones] mind loose when displaying intimacy (5.2).

The same principle must be applied to unravel the definition of the character trustworthiness (xn) in the concluding building block 6. Accordingly, trustworthiness has the following connotation in the context of the Zhng xn zh do:
Not to cheat and not to [be] cunning (1.1B), that is, to hold antiquity in high reverence and to take it as a principle (3.B) and thus never to let [ones] words contradict [ones] actions (5.1). (See fig. 5.)

Conclusion To approach early philosophical texts from the perspective of their formal structure is a powerful strategy for dealing with early thinking. From the philological perspective, turning the strategy by which a text constructs its argument into a strategy for reading the text enables sense to be made of passages that would otherwise be unclear. In

the way of fidelity and trustworthiness

49

1A 1B 2A 2B

1.1a 1.1b 1.2a 1.2b 1.c 2.1a 2.1b 2.2a 2.2b 2.c 3.a 3.b 3.c

[] [] [] [] [] [----- []

4.1a 4.1b 4.2a 4.2b 4.c 5.1 5.2 5.3 5.C

6.1a 6.1b 6.2a 6.2b 6.c

Figure 5: Overlapping Structure on the Macrolevel of the Zhng xn zh do

50

chapter one

saying this, I do not suggest that reading a text like the Zhng xn zh do through its formal structure should be seen only as a strategy for corroborating the palaeographical reconstruction of corrupted or missing passages. Paying close attention to the formal structure of a texts composition provides evidence about the proper sequence of the strips, as will be discussed in the next chapters. It is clear that the formal structure of a text like the Zhng xn zh do was a vital element for generating meaning beyond the level of the lexicon. By interrelating the various building blocks, the text connects the different notions advanced and so defines the meaning of certain philosophical concepts. Thus, the formal structure of an argument-based text such as the Zhng xn zh do helps to generate definitions peculiar to the text. This applies both to the texts technique of introducing new terms by relating them to previously defined ones and to the modus operandi by which the text constructs a relatedness among various terms that we had not otherwise perceived. When looking at philosophical writings not only as repositories of ideas but as meaningful objects in their own right, the texts can be seen in a new light. Argument construction in the Zhng xn zh do works in a fundamentally different way from an argument in the style of a syllogism, where A and B of necessity lead to C, such as in a treatise like the Organon of Aristotle. The Zhng xn zh do also does not generate meaning by telling stories, as, for instance, the Mngz does. In contrast, the authors of the text construct the overall argument of the Zhng xn zh do by relating the highly distinctive building blocks to one another. By means of the formal structure, a web is woven into which the text places the message. The various notions and ideas are thereby connected to positive classifications and settings as provided in the text. It furthermore is important to notice that the text as a whole basically works in the same fashion as the individual building block. The pattern of an overlapping structure not only applies to the individual building block but is the very principle underlying the makeup of the text overall. Note in this context that neither the highly distinct building blocks nor the overlapping structure depend on the materiality of the text. As will be discussed in further detail, in the Warring States period, text was an idea abstracted from its material carrier. When exploring the habits of early reading and writing, it is therefore essential to distinguish clearly between text as the matter to be transmitted and manuscript as the physical instantiation (or representative) of the text.

the way of fidelity and trustworthiness

51

Neither the individual components that constitute a building block nor the individual building blocks of the Zhng xn zh do correspond to the length of one bamboo strip of this manuscript. That is, even when one building block contains the number of graphs that would fit on one bamboo strip of this manuscript, we do not find one building block written on one strip.25 Meaning construction in the Zhng xn zh do and the other texts that exhibit this type of philosophical reasoning is not the direct product of certain conditions of writing; that is, it is not the result of the materiality of the items used, such as the length of the bamboo strips on which the text was written. In contrast, we see from this a particular mode of philosophising in early China for which tradition has left us unprepared, as it would not outlast the centuries to come. As discussed, the overlapping structure applies both to the microlevel of composition (i.e., the individual building blocks of the Zhng xn zh do) and also to the macrolevel of composition (i.e., the text as a whole). This means that not only are the various building blocks of the text designed in a consistent fashion but also the macrostructure of the Zhng xn zh do itself largely mirrors the structure of the individual units. Because the individual building block is composed in a strict 1ab2abc scheme, it achieves a particular rhythm and, hence, stability.26 And because the text as a whole emulates the makeup of the individual building blocks, the text as a whole gains stability. This is not to say just that the Zhng xn zh do and its strategy of meaning construction would not work if one of its building blocks were removed; indeed, argument construction in the Zhng xn zh do would also fail if other elements were added, or even if a building block was moved within the text. By implication, on the level of its formal structure, the Zhng xn zh do is fixed.27 As the formal structure of the Zhng xn zh do is a crucial device for conveying meaning, a text of this kind can exist only in a given form. The formal structure crucially adds to the philosophical message by reflecting the message of the text: the Zhng xn zh do describes sovereignty over a political space that extends universally, embracing the cosmic elements Heaven and Earth. The texts structure, in which
This will be discussed in more detail in chapter 3. See Jahandarie 1999, 311, for a discussion of the important mnemonic functions of rhythm and structure in early literature (and also for further references). 27 Note that this does not imply that the lexicon of the text is also fixed.
26 25

52

chapter one

the macrostructure mimics the pattern of the individual building block, reflects this. The formal perfection of the account spans the entire text, just as the rulers transformative power of proper conduct (we might call this his d ) also spans the entire world. Form and content of the Zhng xn zh do are in perfect harmony with each other. The fact, then, that one particular unit of the Zhng xn zh do, namely building block 5, breaks the otherwise consistent pattern of the text does not in the slightest contradict my conclusion that the formal arrangement is fixed. Quite the contrary, building block 5 transforms the abstract argument of the text into a concrete directive addressed to the ruler of a state. By breaking away from the otherwise consistent pattern of the Zhng xn zh do, it technically marks off the central point of the text. As I shall show later on, this technique of conscious violations of expectation was a common device of argument construction in texts of this type from the Warring States period. Note that even in the case of this particular unit, its overall position in the text is fixed. Because meaning is constructed by relating the different building blocks with each other, this particular unit had to be placed after what I have identified as the introduction of the second canto of the Zhng xn zh do, that is, building block 4, and before the overall conclusion of the text, that is, building block 6. The Zhng xn zh do can exist only in this particular form.

CHAPTER TWO

THE QING D Y SH FAILURE AND SUCCESS APPEAR AT THEIR RESPECTIVE TIMES The preceding discussion has described the makeup of an argumentbased text, the Zhng xn zh do, in which the overlapping structure endows the text with a distinctive rhythm. The repetitive use of this feature supports memorability and stability of the building block, and the micro-macro correspondence of the formal composition defines the confines of this textual unit. The text as a whole gains a stability comparable to that of the individual building block. Moreover, due to the fact that the macrostructure mimics the microstructure, the formal structure of the composition compounds the thought central to the text. Form and matter are in congruence with each other. The Qing d y sh (Failure and Success Appear at Their Respective Times) is organised in a radically different fashion. As the following discussion will show, its structure, however, fulfils the same integrating function in the process of constructing meaning. The Text on Bamboo The editors of the Hbi Province Museum report that the Qing d y sh consists of fifteen bamboo strips.1 The complete bamboo strips have a length of circa 26.4 centimetres each, and they are tapered towards both ends. As can be judged from the marks of the binding strips that remain visible on the bamboo to the present day, two cords, 9.49.6 centimetres apart, previously held the bamboo strips together.2 As for the shape of the strips and the style of calligraphy, the Qing d y sh closely resembles the Lu M gng wn Z S (Duke M of Lu Inquires of Z S).3 It can be assumed

See Hbi shng Jngmn sh bwguan 1998, 2728, 145146. See ibid., 145. 3 The Lu M gng wn Z S consists of eight bamboo strips. The text is composed in the form of a dialogue between Duke M of Lu and Z S and between Duke M with the otherwise-unknown minister Chngsn Y . The text ends with a
2

54

chapter two

that the two manuscripts were produced at about the same time and place. Despite the physical similarity of the two manuscripts, the Qing d y sh and the Lu M gng wn Z S share no textual relationship, and the argumentative patterns of the two texts differ totally. They have to be regarded as different texts. The manuscript of the Qing d y sh is relatively well preserved. Of the fifteen strips, two have both ends broken off. The missing parts probably contain some 17 graphs. The extant number of graphs is 289. This means that the manuscript as it was placed in the tomb probably consisted of a total of circa 306 graphs. Despite the small scale of textual loss, its reconstruction proves extremely difficult.4 So far, no consensus has been reached among modern scholars about how to arrange the bamboo strips. Ikeda Tomohisa and others hold that the Qing d y sh in its current form is incomplete. They assume that some strips are missing, probably those between extant strips 8 and 9.5 Chn Wi and Chn Jin , among others, think that the Qing d y sh is complete.6 However, both scholars disagree with the sequence of strips as suggested by the editors of the Hbi Province Museum, and each proposes a substantially new arrangement of the text.7 As argued elsewhere,8 I contend that the Qing d y sh in its current organisation9 is indeed complete, andexcept for interchanging strips q7 and q8the sequence of strips is largely correct.

long apologia by Chngsn Y in support of Z S, so as to convince Duke M that Z S should be regarded as a loyal minister par excellence, despite his seemingly unsatisfying answer to Duke Ms enquiry. Chngsn Ys apologia in support of Z S is phrased in the pattern of an overlapping structure. 4 For a detailed philological discussion of the Qing d y sh, see chapter 10. 5 See Ikeda 2000a, 152. Zho Pngn (2002, 20) and others share Ikedas assumption that the Qing d y sh is incomplete. 6 See Chn Wi 2003; Chn Jin 2004, esp. 316322. 7 Chn Wi suggests the following sequence of strips: 16, 8, 7, 14, 913, 15 to end. Chn Jin suggests the following arrangement: 18, 14, 913, 15 to end. Both scholars proposals pose problems in relation to the structure and the grammar of the text (except for Chn Wis proposal to exchange strips 7 and 8 with each other, which I also put forward). I shall refer to these positions in chapter 10, [J]. 8 See Meyer 2007. 9 As presented in Hbi shng Jngmn sh bwguan 1998, 2728.

failure and success appear at their respective times Thought and Contents

55

The dichotomy of Heaven and [M]an is one of the constants of Chinese thought, Graham points out.10 His remark applies well to the Qing d y sh, whose precise dating and philosophical affiliation are an issue of continuous debate among modern scholars.11 The aim of the Qing d y sh is not to explain or solve the apparent conflict between Heaven and Man but rather to provide Man with a guide to dealing with lifes imponderables as caused by Heaven. The Qing d y sh was unknown to modern readers prior to the excavation of Gudin One. It dwells on the intrinsic tension between Heaven and Man, which the authors of the Qing d y sh assert is innate in their relationship with each other. The text argues that even though Man has full control of his actions, he has no means to control their outcome. This lies with Heaven. In other words, to succeed, Man needs the opportunity offered by Heaven; otherwise, he fails. As a result, he can become a gentleman ( jnz )12 only if he does not direct his actions towards a certain goal, which he cannot predict he

10 Graham 1989b, 107. I have capitalised Man to stress the classical dichotomy of the two. 11 Core notions of the Qing d y sh, such as qing to fail and d to succeed as a pair, also appear in the Mngz (see Jn xn shng ) and the Lnyu (see Wi Lng gng ). Zhng Lwn (1999b, 218, 220) argues accordingly that the date of the Qing d y sh is close to that of the late Mngz. Monolinear attempts such as this to date a text and explain its philosophical orientation greatly underestimate a texts complexity. The Mngz, like most early texts, is not one single entity but rather an anthology of all kinds of ideas with different chronological origins and editorial engagement. (Note that in his preface to the Mngz, Zho Q [d. AD 201] already pointed out the difference between the inner and outer chapters [see Lau 1993, 331, 332].) L Cnshn (2002) demonstrates similarities of the Qing d y sh with the transmitted texts of the Mngz and the Xnz . Wng Zhpng (2004), for his part, holds that the ideas of the Qing d y sh correspond closely to those of the Xnz, for which reason he conjectures that the Qing d y sh could be a work of a postulated Xnz school. In this regard his position is not so far from that articulated by Paul Goldin in 2000, in which he assumes that the Gudin One manuscriptsincluding the Qing d y shcan be imagined to be what Xnz learned in school because Xnzs positions appear more systematically argued than anything to be found in the Gudin manuscripts. According to Goldin, there is little question that Xnz was influenced by the same doctrinal set (2000, 146). 12 Whereas gentleman ( jnz) in the Zhng xn zh do denotes the ruler, in the Qing d y sh, this designation refers to the group of ministers, advisors, and so forththat is, the next layer beneath the ruler in the social stratification of a state.

56

chapter two

will achieve anyway. In contrast, the true gentleman as presented by the authors of the Qing d y sh should value only his personal worthiness, which the text calls virtue (d ); worries about the failure or success of his actions should not trouble his mind. The exclusive focus on his own worthiness allows him to liberate himself from the vicissitudes of (political) life, in which failure and success may come in unpredictable, and often undeserved, ways.13 Structure and Thought The intrinsic tension between Heaven and Man as voiced in the Qing d y sh is also given expression in the formal structure of the text. At different levels, the formal structure renders a duality, such as Heaven versus Man, success against failure, have ( you ) opposed to have not (w ), and it seems that the text is fashioned in such a way that the formal structure of the Qing d y sh mirrors the logical structure of the argument.14 This is more than a mere device for patterning the text. I hold accordingly that the compositional structure of the Qing d y sh is consciously designed to mimic the logical structure of the argument that the text aims to transmit. As such, it is a philosophically relevant element of meaning construction in that it compounds the thought central to this text. This has a direct implication for our analysis. The Qing d y sh consists of two parts, or cantos. These can furthermore be divided into six units, which I refer to as subcantos, a term commonly used in biblical studies.15 My analysis of the strategy

13 The theme of the slandered statesman or neglected worthy is ubiquitous in late Warring States Chinese prose writing, and it becomes a recurrent topos in Hn dynasty f literature. For a discussion of this theme, see Hightower 1954; Wilhelm 1957; Pankenier 1990. 14 Here and below in my discussion of the Qing d y sh, I do not use the word logic in its rigorous sense: it is not meant as a formal device for the study of the principles of valid inference, that is, the process of drawing a conclusion based on what one knows, on the one hand, and its demonstration, on the other. Instead, in this context it is used purely to contrast the reasoned level of the argument with the formal level of the compositional structure. 15 See, e.g., Korpel 2000. Just like the building block, the subcanto is a unit that should be elaborated with respect to the individual text. I want to avoid using section as one would probably do in the analysis of modern compositions, and so I use subcanto to indicate a self-contained unit at a level of meaning construction above the building block.

failure and success appear at their respective times

57

of argument construction in the Qing d y sh proceeds along two lines. The first is a horizontal analysis of argument construction. This is a description of how the text generates a gradually rising tension from one subcanto to the next, thus developing a linear argument. The second is a vertical analysis of argument construction. This is a description of the hierarchical structure of the texts argument. As will become obvious from the discussion, the two types of analysis point to different conclusions concerning the relationship between Heaven and Man, but the apparent conflict is in fact part of the texts argumentative strategy and is resolved in an overall synthesis. First Part: HeavenMans Failure Subcanto 1: Introduction to the Problem
q2
q1 q1

Theme Development of the theme

There is Heaven and there is Man, and there are distinctions between Heaven and Man. When investigating the distinction between Heaven and Man, one will know what they act upon. There might be the right man, but no right q2 times for him. Even if he was a worthy one, he would not enact it (his worthiness). However, if only there were the right times, what difficulties could there then be?

Subcanto 1 reads like a general introduction to the concern of the text. This unit is patterned in a dualistic fashion and can be divided into theme and development. As will become obvious, dualistic patterns not only are characteristic of individual building blocks or subcantos of the text but are the predominant characteristics of the compositional structure of the Qing d y sh as a whole. The theme states the self-evident truth that two distinct entities exist: Heaven and Man. The final sentence of the theme concludes that when investigating this truism, one will, by definition, know what Heaven and Man (can) act upon. At the same time, this statement also raises an expectation and thus serves as a bridge to the next passage of

58

chapter two

the text. I tentatively refer to this feature as double-directed function of a text segment:
q1

concludes continues

The double-directed feature links up two parts of a text. It can appear on different levels of text composition. The unit quoted likewise is not an isolated instance of this feature. The development of the introduction deepens the discussion of the differences between Heaven and Man by naming the two preconditions that must coincide so that Man can drive his actions to success: first, he must be a worthy man;16 and second, the times must be right. The theme of the introduction has already pointed to the polarity of Man and Heaven. The development further opens up a polarity between the positions have (you ) and have not (w ), which are used in two strictly parallel sentences in the text: There might be the right man, but no right times (). When comparing this statement with the first line of the theme (There is Heaven and there is Man ), it becomes obvious that the development of the introduction brings into play the concept the right times (sh )17 wheregiven the parallelism of the first line of the introductionone might expect Heaven (tin ). Interchanging Heaven with times, the authors of the Qing d y sh conceptually relate the two concepts. Heaven and the times are given equal structural significance, and either of them can be substituted for the other. This technique of defining newly introduced terms by relating them to well-defined ones by way of reiterating parallel patterns was an established standard in the Warring States period.18

This passage terms him xinrn worthy person. The term sh generation, epoch; world will be read as the right times in the sense that the worthy one may or may not live in an epoch (or generation) that will allow him to flourish in the world. 18 I have discussed this technique of establishing persuasive definitions in early Chinese philosophical texts in Meyer 2006 and also 2007. See also chapter 1 for persuasive definitions established in the Zhng xn zh do.
16 17

failure and success appear at their respective times Subcanto 2: Legendary Materials and the Crucial Matter of Y

59

Having stated the basic concern of the Qing d y sh, subcanto 2 assembles narrative material to deepen the discussion of the right times. The authors of the text discuss Mans dependence on circumstances: (even) the worthy one depends on someone in power:
(1) [A] q3 [B] Shn ploughed [the fields] at Mount L, and he made pottery q3on the banks of the Yellow River. The reason he became established as Son of Heaven was his encounter with Yo.
q4

(2) [C] [D] Sho Yo wore a hemp coverlet, covered [his head] with a hemp hat, and swathed himself in a [protecting] scarf.19 q4 The reason he became the assistant of the Son of Heaven when he escaped the wooden barriers for building earthen walls was his encounter with Wudng. (3) q5[E] L Wng was a slave at the ford of J, and trembling he q5watched the gates of the territory of J. Seventy years went by during which he slaughtered oxen at Zhog. The reason he was elevated to become the tutor of the Son of Heaven was his encounter with [King] Wn of Zhu. (4) q6[F]
q6 Guan Ywu (Guan Zhng) was detained in prison, where he was bound and tied up. The reason he became minister of the feudal lords when he escaped the weapons and his prisoners cage was his encounter with [Duke] Hun of Q.

19

His attire indicates his poverty.

60

chapter two
(5) q8[G]
q8

Snsh (Snsh o) thrice declined [the position of ] vice minister of war at Qs. The reason he became the senior official when he came out [of seclusion] was his encounter with [King] Zhung of Chu.

(6) q7[][H] [?][I] q7 Bail [X] was sold for the price of five rams and became the elder of the oxherders. The reason he became minister at the court when he escaped [?] was his encounter with [Duke] M of Qn.

Subcanto 1 already made it clear that the worthy one has to live during the right times for letting his worth work in the world so that his talents become widely apparent. Subcanto 2 continues this insight. Drawing upon a pool of cases that reflect the cultural memory of a particular period and a particular group (for simplicitys sake, I shall henceforth refer to this pool of cases as legendary materials),20 the account describes specific encounters of a humble person with someone in power. By correlating subcanto two with subcanto 1, the authors of the text equalor correlatethe right times (sh ) with the encounter (y ) of a worthy one with the right powerful man. Thus, in the logic of the Qing d y sh, Heaven (tin ) brings forth the right times (sh ); or the right times are an aspect of Heaven. The right times are in turn either characterised by the hierarchically superior man who recognises the worthy one as worthy and thus employs him, or he is a manifestation of the right times. It could also be that the authors regard the worthy ones encounter with the superior man as dependent on the right times. This is the aspect described as y .21

20 Note that for the contemporaneous recipient, these cases did not belong to the realm of legends. Yet I hope that anachronistically calling them legendary materials is justified as shorthand for referring to oral or written stories that together makeup a groups remembered history. For the term cultural memory (kulturelles Gedchtnis), see J. Assmann 1999, esp. 1924. 21 Compare the statement that appears in the first chapter of the Lnhng by Wng Chng (1996, 1:1). It seems that in the later philosophical discourse y became a technical term, and it may well be that at the time of

failure and success appear at their respective times

61

The paradigmatic examples, which quote the lore of legendary materials, are presented in a highly structured way. The six examples fall into two alternating patterns, which I call type A and type B. Type A is structured as follows: line 1 introduces the name and background of the worthy person (e.g., ); line 2 describes the worthy ones deprived situation (e.g., ); line 3 resolves the case (e.g., ).22 Type B differs from type A mainly in the second line, which always states that the worthy one X was freed from his deprived situation by meeting person Y.23
(A) q3
q4

(B)

(A) q5 (B)q6 (A) q8 (B) q7 [?]

The repeating pattern in which the legendary materials are presented fulfils three functions. First, the repetition highlights the basic principle that underlies these stories and myths: a person of humble social origin realises the highest merits through his encounter with the right counterpart, who understands the persons worth and employs him. It is reasonable to assume that the stories are tailored to fit the message.24 Second, the formal perfection of the account adds to the credibility of the stories themselves.25 Third, the strict scheme of alternating A and B types provides this account with a distinguishing rhythm, and thus
the composition of the texts from Gudin One, it had already achieved philosophical relevance. 22 With a slight derivation of this pattern in 3A in that line 1 is longer and contains all the information needed to advance the example. 23 Note that I have interchanged strip q8 ( ) with q7 ([?]), for which there is good evidence besides the highly structured scheme of alternating type A and type B. See my discussion in chapter 10, [G]. 24 This claim is further substantiated by the fact that certain stories (e.g., the story of L Wng) represent a traditional representation rather than historical truth. See also chapter 10, [C]; Allan 1981, 325. 25 This corresponds to analysis by Wolfgang Behr who claims that certain structures present intricate arguments by integrating them into a persuasive aesthetic environment (2006, 26).

62

chapter two

with a certain stability. This unit must thus be regarded as a stable element in its own right. Accordingly, the formal structure of this passage stresses the common truth underlying these stories and highlights the crucial notion presented therein. Subcanto 3: Deriving a Principle from the Legendary Materials The legendary materials presented in subcanto 2 makes clear that the worthy one needs to encounter a superior man so that he can be raised to an appropriate position that allows him to work in the world. Subcanto 3 uses this insight to construe a general principle from this observation. It asserts that the changing fate of a worthy one as described in subcanto 2 does not derive from an alteration of his essential traits, which the text identifies as virtue (d ) and wisdom (zh ):
q9

[J] [K] q10

q9

[Thus, the fact that] in the beginnings [these worthies] were of little value and in obscurity, [and yet] their names were later praised, is not because their virtue (d) has been added to. [That Wu] Zx was very meritorious at first, [and yet] he later fell into disgrace and was put to death, is not because his wisdomq10had weakened.

Subcanto 3 recognises a general truth behind the account of the legendary materials of subcanto 2. It was not due to an increase of virtue that the fate of the humble worthies changed. Additionally, subcanto 3 also calls to mind the case, well known at the time, of Wu Zx and his fall from grace. This expresses the insight that the changing fate of a worthy one works in both directions. Failure, as expressed in the humbleness of a worthy one, may indeed turn into success, but the reverse is also true: success, as expressed by Wu Zxs merits, may likewise turn into failureat any time, unpredictably, and even undeservedly. The formal structure of this passage stresses the similarity of these cases: subcanto 3 is composed of two nearly identical strings. Of these, the second appears as the exact negative image of the former. See figure 6.

failure and success appear at their respective times

[] , , (failure) (success) (not because of +) (success) (failure) (not because of )

63

[] , ,

Figure 6: The Underlying Thought of Subcanto 3

The highly structured passage with almost no divergence between the positive and negative statement communicates the idea that the particular instance of someones failure turning to success contains the same principle as that of the reverse case. But the structure of this passage also promotes the stability of this unit. In the process of argument construction, this unit fulfils two functions. First, it generates a rule about a humans change of fate, hence rounding off the previous account. Second, by showing the reverse case, it draws attention to the same principle underlying the different historical cases. Yet even though this unit adds information to the discussion, the present subcanto does not yet close the matter by explicating what the decisive factor of a worthy persons fate actually is. Accordingly, subcanto 3 raises an expectation on the part of the texts recipient. I call this form of adding new information without resolving the matter open argument. By summing up the previous account with a general rule, the unit closes the description as advanced in subcanto 2. By adding further information to the overall problematic of the text (as indicated in subcanto 1) without yet resolving the matter, it raises an expectation. This unit hence takes on a function much like that of the double-directed feature of a text segment: it acts as a bridge between the previous account and the next unit of the text. As the double-directed feature of a text segment here does not appear on the microlevel of the text (as seen in subcanto 1) but works on the macrolevel of the composition in that it connects different subcantos, I call it a distanced type of this feature. Subcanto 4: Formulating a Closed Argument Subcanto 4 takes the discussion one step further. Just like subcanto 3, the present unit contrasts failure with success. However, whereas subcanto 3 did not resolve the tension between the two contrasting

64

chapter two

aspects, subcanto 4 now synthesises the conflict by adding a concluding statement c. The unit under review thus formulates a self-contained argument. In the context of the argument construction of the Qing d y sh, I call this a closed argument. See figure 7.
A q11 [] [J]

+
conclusion

That [even] the thoroughbred horse becomes distressed at Mount Zhng, and the black-mottled grey horse halts at the thorns of Sho, is not because they have lost their physical strength.26 [But] that [they] exhaust everywhere within the four seas, reaching as far as a thousand q11 l [in each direction], is because they encountered the [excellent rider] Zo F. To encounter or notthis lies with Heaven. Figure 7: The Closed Argument of Subcanto 4

By formulating a closed argument, subcanto 4 takes on a central position for processing the overall argument of the text. It brings the first part of the Qing d y shcanto 1 of this textto its logical conclusion. The thoroughbred and fine horses serve as a metaphor for the worthy one, whilst Mount Zhng and the thorns of Sho reflect his humble situation. The use of the metaphors instead of the names of (semi-)historical figures converts the previous discussion into a general truth, applicable to the gentleman, who will be named further on in the text. The subsequent account must be read through this generality, which names Heaven as the decisive factor settling Mans failure or success. Naming the decisive entity of Mans fate, the unit rather ungently exposes the limits of Man. The message is anything

26 The thoroughbred horse and the black-mottled grey horse should be understood as fine horses.

failure and success appear at their respective times

65

but promising. Man appears as a mere plaything of Heavens will. The text could have ended here. Second Part: ManHis Success Subcanto 5: Inferences from the Conclusion
q12 []
[N]

[] [O] q13 [q13][P] (?)([]q14 [Q] [R] [Thus], to move is not [necessarily] to succeed. This is why [the worthy one] does not q12 {harbour resentment} even if becoming exhausted. {He [simply] hides and does not} go after achieving a name; And so he is without regret even if nobody knows him. {The [ flower] zhln grows in dark valleys};q13{it is not because it cannot be} smelled {by man} that it is not fragrant. The beautiful stone of jade is covered in mountain stones; it is not because {no one knows its}q14 goodness that it neglects itself.

The structure of this unit is comparable to that of subcanto 3. As discussed, subcanto 3 derives a general principle from the legendary materials presented in subcanto 2. By adding further information to it, expectations were raised so that the train of thought was continued into subcanto 4. Thus, subcanto 3 bridges the general idea of subcanto 2 and that of subcanto 4. Subcanto 5, then, infers a rule from the conclusion articulated in subcanto 4, in which the authors of the Qing d y sh have made clear that action does not necessarily yield the expected results. Subcanto 5 continues from that insight and adds to it the perception that although the individual has no control over the outcome of his actions, this nevertheless should not lead to resentment or regreteven if he fails. This conclusion leads one to expect that there must be a reason for this state of affairs. But subcanto 5 offers no clue as to what that reason could be. Just like subcanto 3, the unit draws a general rule from the preceding account without settling the issue. This creates an expectation

66

chapter two
generates a rule from the previous account (legendary materials) infers a rule from the previous account (conclusion) Subcanto 5 features as a bridge to connect the rule to the following pericope (conclusion) features as a bridge to connect the rule to the following pericope (conclusion)

Subcanto 3

Figure 8: The Double-Directed Feature of Subcantos 3 and 5

on the part of the recipient. Hence, in the same fashion in which subcanto 3 featured as a bridge to connect the train of thought of subcanto 2 to that of subcanto 4, subcanto 5 features as a bridge to connect the insight from the conclusion articulated in subcanto 4 to the account that follows. Just like subcanto 3, subcanto 5 can be classified as a double-directed feature of a text segment (see fig. 8). Accordingly, it should likewise be regarded as an open argument. Subcanto 5 compares the worthy one with the fragrant flower zhln in a dark valley27 and with a beautiful but forgotten gem of value in a mountain of stones. Since it states that even failed action should lead neither to frustration nor to the renunciation of ones value, subcanto 5 offers a more positive perspective on the vicissitudes of life, after the recipient of the text has just faced the sombre conclusion of subcanto 4. Subcanto 6: Directive for the Individual
q15 [S] Failure and success appear at their respective times. Virtue (d) and conduct [may] be one, [and yet] fame and slander stand by their side.

27 On the rhetorical function of flowers in Chinese poetic texts and the problem of Latinised translations, see Kern 1994. Other texts from the Warring States (or later) also use the zhln flower to express ideas like those presented in the Qing d y sh. It thus seems that by the late Warring States, the zhln flower was a topos denoting a remote or reclusive worthy person.

failure and success appear at their respective times

67

[But if ] acuity reaches the one mother, black and white q15need not be distinguished [anymore]. Failure and success appear at their respective times, [yet] dark and bright do not get reiterated [along with them]. It is for this reason that the gentleman esteems self-examination.

Subcanto 6 concludes the discussion about the apparent arbitrariness of the failure or success of Mans deeds with the newly introduced word times (sh ). Failure and success each appear at a given time irrespective of other circumstances. The notion that failure and success depend on more than just the worthy one is by now an established truth. In subcanto 6, this notion is applied explicitly to the gentleman ( jnz ). At this juncture, I think it necessary to discuss the structure of this particular unit in more detail. Formulating the conclusion of canto 2, subcanto 6 is vital for the overall message of the text. At the same time, this unit is remarkably difficult. It contains a number of enigmatic ideas. Yet the analysis of its structure can provide the key to resolving the matter. Subcanto 6 is composed of four strings of mostly parallel segments. However, not every segment reappears in each string of the scheme, as figure 9 shows. The sentences tentatively called B1 and B2 and C2 function as the connecting bridges within this unit. 1: 2: 3: 4: A B1 C1 B2 C2 A C2 g B3

Figure 9: The Scheme of Subcanto 6

Applied to the text, we have the following picture:

1: 2: 3: 4: A A B1 B2 B3 C1 C2 C2

Note that the in-depth structural analysis of subcanto 6 is an attempt to reconstruct the strategy of meaning construction applied in this cryptic passage. It cannot serve as a definitive reconstruction of this passage.

68

chapter two

Two main features of this scheme must be highlighted to appreciate its function. First, it seems that rhyme is involved in this unit. Except for segments B1 and C1 the entire unit rhymes in *-:
1: A[*d~*d] B1 (*lAj) C1 (*pra) 2: B2 [*m28] C2 [*r] 3: A[*d~*d] C2 [*ts-s] 4: B3 [*k(r)]

The break in the rhyme pattern in segments B1 and C1 is puzzling, but it probably can be explained satisfactorily. I discuss this feature below. Second, there are some gaps in the pattern of subcanto 6.29 These gaps are structurally consistent and have a function. Their use can be considered a meaningful feature of this passage. Contextually, the informed recipient bridges each gap by repeating the missing segment from the previous line. Segment A must hence be repeated in line 2; segment B2 in line 3. The conjunction this is why (g ) in line 4, then, signals that the argument as a whole should be reiterated here. Yet the primary function of these structurally significant gaps is not only to indicate that the previous segment should be reiterated but also to signal a turn in the reading of an otherwise continuous thought. We can assign to it something like an [even] ifthen turn. In the following table, I have indicated the segment to be added contextually in each gap by setting it in italics; I have marked the indicated turn by underscoring it:
1:A Failure and success appear at their respective times. B1 Virtue (d) and conduct [may] be one, C1 [and yet] fame and slander stand by their side.

28 The reconstruction of *m(r)o as presented in Baxter 1977, 291ff., accounts well for the Middle Chinese notation of mu , which is muwX. But Baxters *P(r)o repeatedly rhymes with *- in the Shjng, and words in *P(r)o show xishng contacts with words in *-. It is therefore likely that in some Old Chinese dialects that are not directly ancestral to the Middle Chinese language of the Qiyn, where the two remain distinct, *P(r)o has merged with *P(r). (See also the discussion in Baxter 1992, 465ff.) 29 The A component is missing before segment B2 in line 2. In line 3 segment B1 is missing. In line 4 segment A is replaced by the conjunction this is why; therefore (g ).

failure and success appear at their respective times

2:A [Failure and success appear at their respective times], (1A) 3:A Failure and success appear at their respective times, B2 [but] [if] acuity reaches the one mother, B2 [but if acuity reaches the one mother], (B2) C2 [then] black and white need not be distinguished [anymore]!

69

C2 dark and bright do not get reiterated [along with them anymore].

4:A [Failure and success B2 the gentleman appear at their [only] esteems respective times (A), self-examination! but if acuity reaches the one mother (B2), black and white need not be distinguished anymore (C2) for dark and bright do not get reiterated along with the many more (C2); it is for this reason (g) that

Accordingly, the gap in line 2 turns segment B2 into an opposite reading, so that black and white need not be distinguished [anymore] for the person whose acuity reaches the one mother (C2). The next turn appears in line 3. The recipient contextually repeats the segment B2 from above, which thus leads to the conclusion that dark and bright do not get reiterated [along with them anymore] if following this principle. The last gap appears in line 4. The statement starts with for this reason (or this is why) ( g ), clearly marking the conclusion of this part. As is typical for g statements, the information that leads to g appears in front of this marker. Accordingly, the informed (or guided) recipient bridges this last gap in front of the g by repeating the above information. This, I believe, is where the break in the pattern of the rhyme comes into play. Only those segments that do not conform to the pattern of the *- rhyme state the negative case that even though conduct and virtue [may] be one (B1), fame and slander [nevertheless] stand by the side (C1). Yet this is true only for someone who has not internalised the full potential of virtuous conduct, as slander and fame are no longer of concern to the accomplished gentleman. Thus, when reading this passage through its formal structure, it is possible to contextualise the reference of the statement that

70

chapter two

the gentleman lays all importance on self-examination, which otherwise would remain unclear. Accordingly, this last statement turns subcanto 6 into a closed argument. Like the previous units of this text, it becomes a stable element in the construction of the overall concern of the Qing d y sh. The passage under review has caused many modern commentators considerable headache. Nevertheless, I am confident that paying close attention to its formal structure can help us make sense of it. The parallel position of the term y , one, in lines 1 and 2 of the scheme suggests that they share the same referent; one mother, in turn, denotes the one principle according to which the gentleman should consistently align his behaviour.30 In a similar vein, this approach can be used to reconstruct the referent of the awkward terms that appear in the two parallel components C2 in lines 2 and 3 (black and white need not be distinguished and dark and bright do not get reiterated ). The referent of these segments nowhere appears explicitly, but the clear parallel structure confirms that segment C2 in lines 2 and 3 refers to something similar to what is expressed in C1 (slander and fame stand by the side ). As a result, black and white (C2, line 2), just like dark and bright (C2, line 3), should be understood as particular manifestations of success and failure. Just like slander and fame (C1, line 1), they may befall the individual at certain times. However, for the gentleman ( jnz ), whose focus lies exclusively on the cultivation of virtue (d )the guiding principle, which the text calls the one motherthe manifestations of success and failure become irrelevant. The consciousness of his own worthiness provides him with a steady imperturbability. Thus, failure and success are aspects that may return at any time, as the word sh at the beginning of this section plainly shows. As such, they are unpredictable even for the gentleman. However, if he is armed with the consciousness that only the cultivation of his own virtue is

30 See also the various occurrences of the term mu mother in the Laoz . See, for instance, the transmitted chapter 25 ( it can be taken for the mother of Heaven and Earth) and chapter 20 ( I alone differ from the others in that I honour this nourishing mother), the former of which also appears in the Gudin One manuscript A, strip a21/20. Most commentators agree that the term mother in the Laoz denotes the do , that is, way or principle. It seems that by the time of composition of the Qing d y sh, one mother or mother had already developed into a topos, so that the use of this term was by no means at odds with the conventions shared by certain Warring States philosophers.

failure and success appear at their respective times

71

valuable, they will cause him no harm. In a way, this mind set allows him to live with a lack of success in his political career, something that must have been quite common to users of this kind of text. The Macrostructure of Composition Horizontal Line of Analysis The analysis has shown that, from one subcanto to the next, the Qing d y sh carefully develops a linear argument. That is, after subcanto 1 (the introduction to the philosophical concern of this text), each textual unit builds on and contributes to the insight gained from the previous subcanto. The linear argument construction of the Qing d y sh can be represented as in figure 10. As shown in figure 10, each step in the construction of the overall argument corresponds to one subcanto (numbered 1 to 6 in the figure). Each of these steps builds on the argument of the subcanto that immediately precedes it. Subcanto 3 may serve as an example: this unit acts as a bridge between subcanto 2 and subcanto 4 by deducing from subcanto 2 the notion that success and failure depend on more than only a mans virtue or knowledge. Then subcanto 3 builds on the observation made in subcanto 2 and prepares for a higher level of insight that forms the basis of the conclusion in subcanto 4. The same is true throughout the text, and it follows that of the two conclusions of the Qing d y sh, namely subcantos 4 and 6, the latter (subcanto 6) must be the one that articulates the crucial insight of the text.
Conclusion (6) Bridge (5)

Conclusion (4)

Figure 10: The Linear Type of Argument Construction in the Qing d y sh

Introduction (1)

Bridge (3) Legendary materials (2)

Canto 1

Canto 2

72 Vertical Line of Analysis

chapter two

This part of the analysis deals with the hierarchical structure of argument construction in the Qing d y sh. This adds to our understanding of the philosophical message of this text. As discussed, the Qing d y sh contains two conclusions, each of which closes one canto. The decisive characteristic of these conclusions in the Qing d y sh is that they break away from the dualistic pattern that is so typical of this text. The closed argument advanced in these conclusions dissolves the dualistic contrast between positive statement (success) and negative statement (failure) by adding to it a synthesising element, c. The Qing d y sh as a whole is composed according to the same principle. Just like the closed arguments (subcantos 4 and 6), the hierarchical structure of the Qing d y sh consists of two conflicting elements (Heaven against Man) and a synthesising element that brings the entire argument to a higher level. I shall now focus on the compositional structure of the two conclusions (subcantos 4 and 6) to investigate their place in the development of meaning in the text overall. The two conclusions share structurally identical features. Despite the fact that canto 2 of the Qing d y sh is decidedly shorter than canto 1, the two cantos closely resemble each other in organisation. In each canto, the conclusion is preceded by a subcanto functioning as a double-directed feature of a text segment (subcantos 3 and 5), in which the preceding account is connected to a subsequent conclusion. Subcanto 3 thus transforms the legendary materials of subcanto 2 by validating the insight concerning the fate of a worthy one, which can change in either direction. Subcanto 4 concludes that failure and success rely entirely on Heavens will. Subcanto 5 draws on this insight but adds to it that this is no reason for frustration or the renunciation of ones value. Finally, subcanto 6 draws on the open argument of subcanto 5 and concludes that for the gentleman who esteems self-examination, the phenomena of failure and success lose their threat. The analysis of the macrostructure of the Zhng xn zh do has shown that the overall structure of an argument-based text can resemble that of the individual building blocks of which the text is made up. In the case of the Zhng xn zh do, the macrostructure reproduces the overlapping structure of the 1ab-2ab-c scheme, in which each building block can be seen as one component of that scheme. Something similar holds true for the macrostructure of the Qing d y sh. The

failure and success appear at their respective times

73

basic characteristic of this text is to express the tension between failure and success, or Heaven and Man. This tension is reflected in the texts macrostructure. The theme of canto 1 is Heaven (or Heavens triumph over Man); that of canto 2 is Man (or Mans triumph over Heaven). Each of the two conclusions advanced in the Qing d y sh is composed as a closed argument in which the conflict between two contrasting matters is synthesised. On the texts macrolevel, subcanto 4 sums up the first part of the text and brings canto 1 to its meaningful close. Canto 2 draws on the insight expressed in subcanto 4 and so develops the overall train of thought one step further. In view of this it appears that subcanto 4 not only features as a connecting link in the development of a linear argument and as the conclusion of canto 1 but also functions as a double-directed segment because it connects two greater parts of the text, namely cantos 1 and 2. On the hierarchical level of meaning construction, subcanto 4 thus functions as the synthesising element between the two otherwise contrasting cantos 1 (Heavens triumph over Man) and 2 (Mans triumph over Heaven). Thus, as in the Zhng xn zh do, where the compositional structure of the text as a whole mimics the pattern of each individual building block, the overall makeup of the Qing d y sh also reflects the structure of a closed argument as advanced in subcantos 4 and 6. See figure 11. Conclusion The Qing d y sh is a well-crafted composition, and it does not seem plausible to imagine that it was created on the spur of the moment. Its hierarchical structure reflects the structure of the individual subcanto. Like the closed arguments of subcantos 4 and 6, the macrostructure of the text advances a closed system. It resolves the tension of two contrasting entities (Heaven and Man) in a closed argument by adding a concluding element (which may be called c). Unlike the closed argument of the microlevel of the text, in which the concluding element c appears at the end of the textual unit, the synthesising element of the macrolevel argument is placed at the centre of the composition. Echoing the structure of a subcanto at the macrolevel, the Qing d y sh is developed as a closed system. The present organisation of the text enables the articulation of a sound argument, on both the micro- and the macrolevel of composition. None of the units as reconstructed here could be relocatedor rearranged internallywithout

74
Theme 1 Introduction Development 2 Legenday materials

chapter two

Examples

3 Bridge

General rule

canto 1
Double-directed segment 4 Conclusion General rule 5 Bridge Examples Preparation 6 Conclusion Concluding part Figure 11: The Hierarchical Argument Construction of the Qing d y sh

canto 2

seriously distorting the structure of a closed argument spanning the entire composition. By implication, it becomes clear that the overall structure of the Qing d y sh fulfils the same integrating function as that of the Zhng xn zh do. Even though the structure of the two texts could hardly be more different, in both cases it does facilitate a closed system that provides stability to the text overall. Just like the Zhng xn zh do, the Qing d y sh can exist only in this structural form. Concluding this chapter, I want to draw attention to an ostensible conflict between the hierarchical and linear construction of the texts argument. According to the linear line of argument construction, the second of the two conclusions (subcanto 6) must be the main thought that the Qing d y sh aims to communicate. The horizontal analysis

failure and success appear at their respective times

75

has shown that the linear line of constructing the argument works cumulatively, and subcanto 4 (conclusion of canto 1) functions as mere preparation for the main argument finally concluded in subcanto 6. The vertical analysis, however, suggests a pivotal function of subcanto 4 for the hierarchical structure of the texts argument: subcanto 4 connects canto 1 with canto 2 and so functions as the synthesising element allowing the text to become a closed system. This suggests that the main thought of the text is articulated in subcanto 4. Creating a closed argument on the macrolevel, subcanto 4 highlights Heaven as the decisive factor, deciding the failure and success of Mans deeds. The fact that subcanto 4 is placed at the centre of the composition further stresses the importance of Heaven. This seems to point to a contradiction between the horizontal and the vertical line of analysis, where the main point is put forward either in subcanto 4 (vertical, or hierarchical, structure of the texts argument) or in subcanto 6 (linear structure). Or is this seeming contradiction simply a misinterpretation of the case? The closed argument of the macrostructure does indeed state that Heaven is a greater power than Man. By locating subcanto 4 at the centre of the composition, the structure further accentuates the central position of Heaven for deciding a humans fate. Despite this, subcanto 6 concludes on the horizontal level that by making ones virtue the only concern, the gentleman is no longer troubled by issues such as failure and success or slander and fame. From this it follows that the gentleman does not direct any of his actions towards a defined goal, as he cannot ensure that he will achieve it. Conversely, he merely values conduct for itself, not its outcome. This is the perfection of virtue. As a result, by focusing on his virtue alone, he frees himself from the threat that success might turn into failure, bright into dark, fame into slander. The gentleman thereby not only emancipates himself from the seemingly absolute impact of these issues but also frees himself from the imponderables of life itself. In other words, by making his virtue his only concern, the gentleman emancipates himself from the decisive power of Heaven over Man! Thus, the vertical analysis of the texts formal structure has demonstrated the way in which the Qing d y sh expresses on the formal level of composition the idea that Heaven is the decisive entity that determines the failure or success of Mans deeds. The horizontal analysis of the texts formal structure has shown how the formal level

76

chapter two

of composition expresses the idea that Man can nonetheless overcome this potential threat. Taken together, the apparent contradiction between both types of meaning construction in the text is resolved. The structure of the Qing d y sh thus fulfils the same integrating function of correlating thought and content harmoniously in a structurally closed system that can also be shown for the Zhng xn zh do.

CHAPTER THREE

THE WU XNG FIVE ASPECTS OF VIRTUOUS CONDUCT Since the publication of the Gudin One manuscripts, the Wu xng (Five Aspects of Virtuous Conduct) has attracted close attention from modern scholars. Numerous articles and books have appeared that discuss the position of the Wu xng in the intellectual world of early China and compare it with its manuscript counterpart from Mawngdu Three. It has been argued repeatedly that the Wu xng is the key to contextualising the criticism of Z S and Mng K (Mencius, traditionally ca. 380290 BC) that was pronounced so harshly in the Fi shr z (Against the Twelve Masters) chapter of the Xnz (traditionally ascribed to Xn Qng ; ca. 310238 BC). The Xnz attacked the two disputers for setting up a theory for which they claimed old ancestry but which was, in fact, heavily flawed, self-contradictory, lacking proper categories. According to the Xnz, they called it wu xng , five types of conduct.1 This

1 (Xnz jji, 9495). [Men like these] scrappily follow the [way of ] the former kings without understanding their guiding threads; nonetheless, [they behaved] as if [their] abilities were manifold, [their] will were great, and [their] experience were broad and profound. [They] set up a theory for which [they] claim old ancestry, calling it the wu xng theory. [This theory] is heavily flawed, self-contradictory, and it lacks [proper] categories; [it is] mysterious and enigmatic and it lacks [proper] theory. Esoteric and laconic in its statements, it lacks adequate explanations. For ornamenting their propositions and to win respect and veneration, they claim: These doctrines are the genuine words of the gentlemen of former times. Z S provided the tune for them, and Mng K harmonised them. The stupid, indecisive, deluded present-day r are enthusiastic about them, not recognizing in what they fail. Passing on what they have received, [they] believe that Kongz and Zgng were highly esteemed for this by later generations. It is here that the fault of Z S and Mng K lies (after Knoblock 19881994, 1:224, emended). Because the critique in the Xnz was articulated so harshly, scholars such as Homer H. Dubs (1928, 7980, n. 4) subscribe to the view that the passage in question is most likely a later insertion. This view was initially expressed by Hn Yng

78

chapter three

theory, however, is mentioned nowhere explicitly in either the received Mngz (probably compiled by groups of disciples of Mng K),2 or any of the works attributed to Z S. As a result, the target of the critique remained obscure for centuries.3 The Wu xng might be the key to this riddle, for it develops a theory explicitly called five aspects of virtuous conduct (wu xng ), and it elaborates upon it in great detail. The Wu xng further displays a close affinity with the Zhng yngtraditionally attributed to Z S4and so it is often seen as the source par excellence for (re-)constructing the Z S lineage, of which otherwise not much is known.5 Furthermore, the Wu xng shares much of its technical terminology with the Mngz.6 As a consequence, it is often (mis-)taken as a missing link for reconstructing r ideasoften referred to anachronistically as Confucianismin an imagined linear development from Kongz (Confucius) to the Xnz.7 For students of the relationship between text and writing in early China, the Wu xng is noteworthy for its numerous quotations. Many of these originate from the collection of songs known as Odes (sh ), which makes this particular text, datable to the mid- to late Warring States period, a significant source for the study of Odes during this period. The only version of this anthology that has survived to the present day, thelater authoritativeMo tradition,
(ca. 200120 BC), the compiler of the Hn sh wi zhun . See also Csikszentmihalyi 2004, 59ff.; Png P 1980, 7188. 2 For theories about the nature of this work, see the controversial study by Brooks (1994). 3 See Png P 2000, 97. See also S. Cook 2000, 135. 4 See Png P 1977; S. Cook 2000, 135. Note, however, that there are considerable doubts about this conclusion. See S. Cook 2000, 130, n. 42. 5 On the tradition of Z S, see Csikszentmihalyi 2004, 257276. 6 Ibid., 110. 7 Scholars such as Chn Guyng (1992b, 394, n. 1) regard the Wu xng as a further development of the Mencian theory of the four sprouts s dun . Mark Csikszentmihalyi explains the Wu xng as a work from the fourth century that develops a systematic moral psychology of the virtues against the background of philosophers positions. Instead of simply deflecting the critiques found in the Mz, Zhungz, and Hn Fiz, Csikszentmihalyi argues that it adopted elements of outside systems, that is non-r critiques, in order to construct some of the basic tenets of the theory of material virtue (2004, 58). As for the term r itself, it is better to avoid the anachronistic denotation Confucianism in the context of the Warring States period. David Schaberg (2001, 9) employs the term ritualist for what is in fact only a loosely defined group of ritual experts and teachers. The association of r with Confucianism is a much later development which took final shape probably not much before the Western Hn. See also Eno 1990, 190ff.; Lewis 1999, 75ff.

five aspects of virtuous conduct

79

stabilised as late as the Hn dynasty and so significantly postdates the Wu xng and its quotations.8 Hence, the Wu xng is important for studies focusing on the development of a Chinese textuality.9 Even though the Wu xng does not belong to the corpus of transmitted literature, we are nevertheless fortunate in having another close variant of this text. Some twenty years before Gudin One was opened, a different manifestation of the Wu xng was excavated from the renowned Hn dynasty tomb Mawngdu Three.10 The fact that a copy of this text was excavated from a tomb which dates some 150 years later than Gudin One suggests that the Wu xng must once have enjoyed at least some popularity before it slipped entirely

In addition to the two versions of the Wu xng, fragments of the Odes are to be found in two excavated manifestations of the Z y (from Gudin One and the Shnghai collection of Chu manuscripts), the Kongz sh ln , and, to a lesser extent, the Mn zh fmu (both of which are part of the Shnghai collection of Chu manuscripts and are so labelled by modern editors). Moreover, we now possess a badly damaged and incomplete anthology (see Kern 2003, 28) of the Odes from a tomb in Shunggudu (nhu Province, sealed 165 BC; see Giele 2001). On the Shnghai collection, see Ma Chngyun 2001. The Mo tradition, which in the post-Hn period displaced the three interpretations of Lu , Hn , and Q all of which received imperial recognition under Emperors Wn (r. 179157 BC) and Jng (r. 156141)is first recorded at the court of Li D (r. 133 BC), Prince Xin of Hjin (see Riegel 2001, 99100). 9 Kern 2003, 30. See also Kerns most recent study of the Odes in excavated manuscripts, in which he traces the double phenomenon of a canonical text that is as stable in its wording as it is unstable in its writing (2005c, xxi). The presentation of the Odes in excavated manuscripts could hardly be more different from that of the Documents (sh ) in excavated manuscripts. Quotations of the latter are highly unstable and inconsistent. Kern concludes that the corpus later known as Documents was rather loose and heterogeneous, which, in turn, suggests a retrospective standardisation of these materials (2005c, 297, n. 9). In this respect, he subscribes to a view like that found in Chn Mngji 1985, 1135; Li Qy 1997, 424; Lewis 1999, 105109; Schaberg 2001, 7280; among others. Based on different quotations, Matsumoto Masaaki (1966, 520) argues that by the late fourth century BC there most likely existed three different versions of the sh , namely a r-, a m-, and a historians version. 10 In winter 1973, archaeologists discovered the previously undisturbed tomb of L Cng (d. 185 BC), who became the Marquis Di . The tomb was dubbed no. 3, Mawngdu (henceforth Mawngdu Three). It is located near Chngsh , Hnn Province. Due to a dated letter to the netherworld found in the tomb, the date of burial can be fixed fairly precise at 168 BC. Among other objects, the tomb contained silk manuscripts inscribed with up to 125,000 graphs. The tomb also contained another version of the Laoz, which will concern us in chapter 5, among other texts. For excavation reports, see Hnn shng bwguan and Zhnggu kxu yun kaogu ynji suo 1974, 1975; Chn Sngchng and F Juyou 1992, supplement. For comprehensive bibliographies on Mawngdu Three, see Zuo Sngcho 1989; L Mil 1992; among others.

80

chapter three

from view for roughly two thousand years. Having two excavated manifestations of the Wu xng makes it an ideal source for a case study that looks at the stability of an early Chinese philosophical text.11 A brief comparison of both manifestations of the Wu xng reveals astonishing similarities between the two texts in lexicon and phraseologythe main differences being that the Mawngdu Three version contains a textual commentary at the end of the manuscript, and that the two texts refer to the Odes differently. When quoting this body of shared knowledge, they use different formulae, and the length of these quotations differs too. It is striking that despite their overall consistency, the two excavated versions of the text differ markedly with respect to the internal arrangement of individual textual units. But despite their different sequence in the overall composition of the text, the units themselves remain surprisingly stable. This phenomenon also applies to the different versions of the Laoz and Z y (Black Robes) found in Gudin One, as well as the different versions of the Y (Changes). On the lexical level, these are relatively consistent texts.12 But the sequence of the building blocks differs substantially in their different manifestations. For discontinuous texts such as Laoz, Z y, and Y, it is not surprising to find relatively stable building blocks that do not have a fixed place in their overall organisation.13 Yet, unlike the Laoz, Z y, and Y, the Wu xng is not simply a florilegium offormallyunrelated ideas, distinct aphorisms, or separate building blocks. Instead, the Wu xng is an excellent example of an argument-based text because it develops one idea throughout the entire text, which consists of approximately 1,200 characters. This fact has stirred scholars imagination. How can it be that the Wu xng develops a single vision, and yet the organisation of the Mawngdu Three manifestation so markedly differs from that of Gudin One? The view has thus arisen that the dissimilar internal organisation of the two texts must reflect a consciously made editorial choice14 and

See, e.g., Xng Wn 1998; Csikszentmihalyi 2004; Boltz 2005. This ignores minor differences such as wording, the formulae when quoting another source, and the length of quotations. 13 I will examine this further in chapter 5. For a study of the Mawngdu Three version of the Y , of which the order of hexagrams and the associated texts differs markedly from that of the received version, see Shaughnessy 1997d. 14 Similar claims were also made for non-argument-based texts some fifteen years earlier, before the excavation of the Gudin One Wu xng copy (see Shaughnessy 1983, 139158, 257265; see also Shaughnessy 1997b, 197219). Shaughnessy claims
12

11

five aspects of virtuous conduct

81

hence reproduces different philosophical positions.15 Otherwise, consensus has it, if the arrangement does not matter, why change it?16 Understandable as the position outlined above may be, it nevertheless reflects a rather modern idea of text and composition that neglects the nature of a text during the Warring States period. Instead of overemphasising the differences between the two manifestations, one should view them as accidental. Individual building blocks in concert constitute larger meaningful and stable units, which I refer to as subcantos. For the Wu xng, these are the units that matter, as it is the subcanto in which a coherent idea is developed. In both instances of the Wu xng, that is, the one from Gudin One and the one from Mawngdu Three, the individual subcantos display a high degree of consistency. The philosophical stances developed therein do not differ in the slightest. Accordingly, I argue that instead of being blinded by the differences of these two manifestations of the text, it is more instructive first to look at their analogousness. Only then do we learn more about the real differences between the two manifestations of the Wu xng and so further our understanding of the nature of early Chinese argument-based texts overall. To arrive at a more comprehensive understanding of the Wu xng, it is therefore necessary to modify the focus of investigation. As developed in the previous chapters, this should be the analysis of the stable units of the text, beginning with the individual building blocks. Only a proper analysis of these will allow us to discover the various kinds of techniques by which larger meaningfuland stableunits are developed in a text. Having identified these larger meaningful unitsthe subcantosof the Wu xng, the next step will be to describe the means by which these larger meaningful entities relate to one another. Only against this background will it be instructive to compare the two manifestations of the text. It will then become obvious that, in what matters, they are not so different at all. As long as the well-defined meaningful units remain intact structurally and as long as these units remain in a logically sound location within the text overall, their

that the location of the Qin and Kn hexagrams at the head of the Zhuy are a logical necessity, and that their position in the Zhuy must echo conscious editorial choices. 15 This view was expressed most explicitly in Xng Wn 1998. See also Boltz 2005; Shaughnessy 2006, 43. 16 This question is posed repeatedly. See, e.g., Dng Sxn 2000a, 128.

82

chapter three

sequence does not matter too much. The argument remains intact. In view of that, I take issue with the suggestion assigning significance to the different order in which stable units occur in a given version.17 Instead of focusing on the dissimilarities between the two versions of the one text in the larger philosophical context, as is done by nearly all commentators on the Wu xng after the Gudin One manifestation of this text had been made public, an alternative (and first) approach should be to investigate the principles of text organisation underlying an argument-based text such as the Wu xng. Therefore, the present study does not so much focus on a (re)construction of postulated intellectual lineages or idealised philosophical positions. Instead, the primary interest of this analysis is to make explicit the strategies of constructing meaning as used in this particular piece of philosophical writing and so to describe the means by which a long argument-based text such as the Wu xng becomes a self-contained piece of thought. Of course, questions concerning a Chinese textuality or the stability of a written philosophical text in early China naturally feed into this discussion. The Text on Bamboo The Wu xng is a lengthy text. In the context of the tomb corpus from Gudin One, only the Xng z mng ch, with its 1,550 characters, requires more room for developing the argument.18 The physical shape of the strips and the calligraphy on them strongly resemble those of the Z y, which is also part of the same assemblage of texts. It might hence be the case that the two manuscripts were produced at the same workshop and may even have constituted one bundle of bamboo strips. An intellectual affiliation of the two texts should, however, not be inferred from this. As discussed, in early China texts survived independently of material textual contexts, and so a congruence in material carrier has no bearing on affinity of the textual units fixed on that carrier.19

Cf. the discussion in Boltz 2005, 54. On the Xng z mng ch and its manuscript counterpart Xng qng ln from the Shnghai collection of Chu manuscripts, see chapter 4. 19 I examine this issue further in my discussion of the Ti y shng shu and Laoz C in chapter 6.
18

17

five aspects of virtuous conduct

83

The philosophical position of the Gudin One Wu xng is developed over the length of fifty bamboo strips. The strips are tapered towards both ends. Judging from the marks that remain visible on most of these, we can assume that two cords, 12.913 centimetres apart, previously held the strips together.20 Unbroken strips have a length of about 32.5 centimetres. They are among the longest of the entire tomb corpus from Gudin One.21 As already discussed, the length of the strips used to record philosophical texts in the Warring States period does not say anything about the status of the text in question, as it probably does later, in imperial times.22 In the Warring States period, manuscripts reflect merely local instances of realisingwriting downwhat may have been predominantly oral texts. Different manifestations of a text have nothing in common with the consciously edited recensions of imperial times. When texts are not yet canonised, the diverse lengths of the bamboo strips on which the different texts are written simply reflect different modes of manuscript production used by different producers or workshops. Recurring patterns in the production of a manuscript, such as the cuttings and lengths of the bamboo strips, indicate the division of labour. This suggests a sociological setting in which professional copyists produced the physical manifestation of texts that otherwise exist independently of the manuscript that carries them. Perhaps this is an indication of the commodification of texts in early China. But it is unlikely to reflect the status of individual texts. Of the fifty bamboo strips used for the Wu xng manuscript from the tomb corpus of Gudin One, eleven are broken off at either the side or the centre. Thus, it is necessary to reconstruct up to thirtyeight graphs. With regard to the arrangement of the strips within that manuscript, the Wu xng is a rare example of general scholarly agreement.
See Hbi shng Jngmn sh bwguan 1998, 149. In the tomb corpus, the bamboo strips of the Wu xng, Z y, Chng zh wn zh, Zn d y, Xng z mng ch, and Li d have lengths of around 32.5 centimetres. The other manuscripts contain physically shorter strips. 22 See Zhng Xun in his preface to the Chnqi Zuo zhun zhng y (7a). He notes a length of two feet four inches for the Classics; one foot two inches for the Xio jng ; eight inches for the Lnyu . All lengths refer to Hn Dynasty measures. (Two feet four inches corresponds to 55.44 cm; one foot corresponds to 23.1 centimetres. See Twitchett and Loewe 1986, xxxviii.) According to Wng Chng , the sayings of the ancients were written on tablets of two feet four inches. See Tsien 2004, 116.
21 20

84

chapter three

Despite this general agreement, scholars have come up with a wide range of arguments and ideas according to which of the two manifestations of the Wu xng should be considered the more authentic or the original version of the text. Scholars have suggested that one of the two manifestations anticipates a conclusion, that one of the versions is superior (or inferior) in terms of logical stringency, or that one articulates a more pristine theory of self-cultivation, representing a closereven the originalversion of the so-called Z S and Mngz school. The differences in the two manifestations of the text are conceived of as the result of decisions consciously taken by later editors [sic].23 Other scholars, such as Mark Csikszentmihalyi, view the Wu xng from Gudin One as a composite of older texts.24 His line of argumentation is based on the observation that only some parts of the text treat the five virtues, that is, benevolence (rn ), rightness (y ), ritual propriety (l ), wisdom (zh ), and sagacity (shng ). Regarding the Wu xng to be an extension of an older text or texts, Csikszentmihalyi believes that these textual ancestors of the Wu xng

23 See Xng Wn 1998, 5960. Xng Wn believes that the Gudin One version is the more authentic representative of the Z S and Mngz school, and that the Mawngdu Three text simply displays the changes that were deliberately made by later editors. These changes would thus report on different philosophical positions. But as Csikszentmihalyi (2004, 86100) argues, to ascribe the Wu xng directly to Z S or his school is, at least, in part wrong (87). He reconstructs different layers of a Z S myth (257276). Xng Wn is also the first to ascribe the changes in the Mawngdu Three text directly to followers of the tradition of Sh Shu . (The Hn sh Ywn zh records that the work Shz (personal name Shu ) consists of twenty-one chapters. See Hn sh, 2:1724. Wng Chng describes Sh Shus philosophical position as holding that mans nature comprises both good and evil elements. See Wng Chng 1996, 1:131.) This opinion is now widely shared by many Chinese colleagues. Based on the fact that Sh Shu is quoted twice in the commentary attached to the Mawngdu Three Wu xng, Dng Sxn (2000a, 160172) contends that the Wu xng must be the work of a later disciple of Sh Shu. Dng shares Xngs view that the Mawngdu Three manifestation of the text was written later. Just like Xng Wn, Dng also holds that the differences between the two versions should indicate conscious editorial choices. However, contrary to Xng, he regards the Mawngdu Three version as inferior to the Gudin One version in terms of what he calls logical stringency (2000a, 131.) Ikeda Tomohisa (2000a) suggests that the Mawngdu Three version should be considered the original text. He assumes that the Gudin One manifestation anticipates a conclusion, which can only be due to later editorial choice. Png P (2000, 92) subscribes to Ikedas view that the Mawngdu Three version renders the pristine philosophy and, in comparison to the text from Gudin One, is organised in a much more logical fashion. Again, the changes are assumed to reflect conscious editorial choices. 24 Csikszentmihalyi 2004, 65ff.

five aspects of virtuous conduct

85

merely treated the relationship of the two virtues wisdom and sagacity. Following his argument, the older texts (or layers) were rewritten in such a fashion that wisdom was interchanged with either goodness or with a series of virtues that together constitute goodness, thus accounting for two of the major structural elements of the work: Firstly, the binary pairing of goodness and the more perfect condition of sagehood; and secondly, the more complex distinctions between the four human virtues and the fifth perfect virtue of sagacity.25 The Gudin One manifestation of the text is sometimes conceived of as directly ancestral to the one from tomb Mawngdu Three. Most representative of this view is Martin Kern.26 Kern defends this on the basis of, first, the high coherence between the two versions; second, their chronological separation by probably not more than about a century;27 third, their geographical proximity.28 As already noted above, the Mawngdu version of the Wu xng is equipped with a commentarythat is, an additional section that modern editors have dubbed an explanation (shu ). The version from Gudin One contains only the basic text. As Kern concludes, the additional section attached to the Mawngdu Three version most likely echoes a new set of explanations from a particular teaching tradition that developed only after the composition of the Gudin One version.29 Accordingly, Kern disagrees with assumption held implicitly by Jeffrey Riegel that the Mawngdu Wu xng as a whole (i.e., including the commentary attached to the end of the text) owes a conceptual debt to the Xnz,30 as this would make the Xnz the older of the two texts. However, since the Gudin One version of the Wu xng most likely predates the Xnz, as Martin Kern and others assume31 (a view to which I also subscribe), Riegels assumption seems unlikely, if not wrong.

Ibid., 65. See Kern 2003, 3839. 27 In the Mawngdu Three version (column 199), the graph for state (gu ) was used where Gudin One has bng , possibly avoiding the character bng , which was a taboo after 195 BC. Bng was the given name of Li Bng (r. 202195 BC), the first emperor of the Former Hn. 28 See Kern 2003, 38. 29 Ibid., 39. 30 See Riegel 1997, 145. 31 See Kern 2003.
26

25

86

chapter three

I hesitate to explain the high degree of consistency between the different manifestations of the Wu xng by subscribing to any of these positions. I believe that the presence of the commentary in the Mawngdu version, that versions use of longer quotations from the Odes, the higher frequency of their use, and, lastly, the clear identification of the Odes by name indicate that the manuscript from Mawngdu Three should be dated later than the version from Gudin One. However, the comparatively later date of the manuscript from Mawngdu Three does not in itself prove an ancestral character of the Gudin One Wu xng, and labelling it so might, however unintentionally, evoke the image of a stemma codicum of their textual relationship. As I will argue, the extraordinary degree of textual overlap results from the high stability of the meaningful textual units from which the two texts evolved. But nothing indicates the primacy of either of the two. Thought and Contents The Wu xng discusses five aspects of virtuous conduct, wu xng , and shows how these virtues relate to each other. I call this the wu xng theory. The wu xng theory is a fully coherent system, established through careful textual links and references that enable stable philosophical concepts to develop. At first glance, the Wu xng appears inscrutable in logic and organisation. But the wu xng theory as developed in the text is in fact a coherent programme of self-cultivation. The final goal of this programme is to nourish d moral force, potency, or virtue within the individual. Central to the theory of self-cultivation are five virtues, each of which describes one particular aspect of virtuous conduct. These include the four virtues named in the Mngz in the context of the notion of the four sprouts (s dun )benevolence (rn ), rightness (y ), ritual propriety (l ), and wisdom (zh )32to which the Wu xng adds sagacity (shng ). Interestingly, even though joy (l )33 is also an important aspect connected

See Mngz 2A6 and 6A6. For a detailed analysis of the term l in Warring States philosophical discourse, see Nylan 2001b. Nylan, who translates the term l as pleasure, differenti33

32

five aspects of virtuous conduct

87

to the cultivation of virtue, it is not mentioned as an individual position in the wu xng theory. Nor is music. The Wu xng advances a political agenda in which moral selfcultivation is not addressed for its own sake but is a vital aspect of good rule. Hence, whereas the Qing d y sh develops an autonomous sphere of self-cultivation which allows frustrated ministers and advisors to keep faith despite worldly imponderables, and thus addresses the group of advisors in particular, it seems that, just like the Zhng xn zh do, the Wu xng targets the ruler himself. Only by cultivating his moral force can his rule become like that of King Wn the ultimate goal of every ruler.34 In this context it is instructive to note that in the philosophy of the Wu xng, self-cultivation is not limited to men of high social pedigree. The text defends the position that becoming aware of ones own abilities is the central step to be taken for moral cultivation. This displays the idea of a common humanity, as this kind of self-consciousness can be nourished in every human being. The Wu xng is a tour de force of densely woven argument, in which the theory of self-cultivation with its application to the realm of politics and the exertion of power is developed around the concepts of clairaudience (cng ) and clairvoyance (mng ).35 Clairaudience and clairvoyance are defined as the vital preconditions for selfcultivation. They facilitate ones abilities to see ( jin ) the worthy one and to hear (wn ) the Way of the gentleman, without which self-cultivation is impossible. The wu xng theory contends that only when one is clairaudient and clairvoyant, can one obtain sagacity (shng ) and wisdom (zh ).36 And as stated elsewhere in the text, sagacity and wisdom are the necessary preconditions for developing the other three virtues, and the development of all five, in turn, is the precondition for nourishing moral force (d ). Clairaudience is the ability to hear the Way of the gentleman (jnz do ).37 The cognitive grasp, or awareness (zh ), of this ability is called sagacity (shng ). Becoming
ates between the rhetoric of pleasure as used by Warring States thinkers and the discourse of pleasure as developed to the level of a theory. 34 See strip w29/1213. 35 The proper translation of the concept mng always causes considerable headache. For a detailed discussion of mng , see Maspro 1933. 36 See strips w20/19w21/9. 37 See strips w26/7w27/14.

88

chapter three

aware that one is able to hear the Way of the gentleman enables the individual to turn this knowledge into appropriate action. Applying this awareness to appropriate action is defined as rightness (y ). To advance rightness at its proper time, finally, is what the Wu xng defines as moral force (d ), the final result of self-cultivation. The function of clairvoyance in the context of self-cultivation is elaborated in a similar way. It is equated with the ability to see the worthy person (xin rn ).38 Parallel to the above case, it is made clear that developing an awareness (zh ) of this is called wisdom (zh ). Being at ease with the awareness to see the worthy one is what the authors of the Wu xng understand as benevolence (rn ). Showing honour for the worthy one is ritual propriety (l ). Hence, in the logic of the Wu xng, the virtues wisdom and sagacity operate in a parallel way to each other. Wisdom is crucial for developing benevolence and ritual propriety, whereas sagacity develops rightness and, as a final result, moral force. It thus becomes clear that the Wu xng distinguishes between two types of virtues. These are the primary, or key, virtueswisdom and sagacity (, )and the secondary, or dependent, virtues, namely benevolence, rightness, and ritual propriety (, , ). In the logic of the Wu xng, wisdom and sagacity (, ) account totally for the development of benevolence, rightness, and ritual propriety (, , ). Hence, the five virtues of the Wu xng are arranged hierarchically: sagacity and wisdom are the key virtues that account for the formation of the remaining three virtues. The Wu xng further says that sagacity takes the lead among those virtues that must be aspired to, zh (henceforth the zh group of virtues). These virtues are all connected to moral force (d ), the ultimate result of accomplished self-cultivation and an essential aspect for ruling the state and becoming just like King Wn.39 Sagacity is thus the foremost of the two key virtues. These virtues are contrasted with those that can be acted upon, wi (henceforth the wi group of virtues), which lead to goodness (shn ). Of these, benevolence and rightness rank first, as they account for the formation of ritual propriety (l ).40 Rightness, in turn, belongs to the zh group of virtues. As such it ranks higher than benevolence. Thus, the hierarchy of virtues as defined by the wu xng

38 39 40

See strips w27/15w28/14. Strip w30. See strip w31/12w32/4.

five aspects of virtuous conduct

89

theory is headed by sagacity and wisdom. These are followed by rightness and benevolence, and, lastly, ritual propriety. Based on the hierarchy of virtues and the distinction between those that must be aspired to (zh ) and those that can be acted upon (wi ), it becomes clear that the authors of Wu xng put forward a system in which sagacity takes the lead in the formation of rightness. Together with the appropriate understanding (zh ), it ultimately leads to moral force, which equals the Way of Heaven. Wisdom (zh ), for its part, takes the lead in the formation of benevolence and ritual propriety. Together with the appropriate understanding, this ultimately leads to goodness, which equals the Way of humans. Whereas the Way of Heaven describes a state of sentiment, the Way of humans describes concrete affairs. It should be noted that if any of these virtues is lacking, sagacity cannot be developed either. Yet, as detailed, sagacity takes the lead in the formation of the other virtues, and so is the crucial factor in the entire process of self-cultivation. The wu xng theory thus expounds an idea of self-cultivation that takes a circular form and in which the final result of self-cultivation is also its point of initiation. I call this the paradox of self-cultivation.41 The paradox of self-cultivation entails that an individuals realisation of any of the five virtues already depends on the accomplished cultivation of the other virtuesand vice versa. The way out of this paradox as seen by the authors of the Wu xng lies in a humans awareness (zh ) of his innate capacity to have the five virtues within him. Structure and Thought Just like the Zhng xn zh do and the Qing d y sh, the Wu xng is composed of highly distinct textual units, the building blocks. These units remain noticeably stable in the different manifestations of the text. They are combined into greater wholes and so form larger meaningful entities, the pericope and subcanto.42 These units are integrated into a system of cross-referential links, by which the notions

Compare this with Nivisons coinage paradox of virtue (1996, 33ff.). Note that there is no universally valid definition of what a pericope actually is (just as there is no such definition for section). Instead, what the confines of a pericopeand subcantoare has to be determined for each particular text. As a rule, the pericope is a structural unit between subcanto and building block.
42

41

90

chapter three

introduced at one point in the text inform those of other units, thus connecting the various textual units into one coherent vision of the Wu xng as a whole. Characteristic of the Wu xng is the recurrent reference to the shared cultural memory of contemporaneous China. It is impossible to locate the source of all these references, but they are used in a coherent fashion in the two manifestations of the Wu xng. As such, the two instantiations of the text yield vital information as to how argumentbased texts in early China made use of a shared tradition to construct meaning. Concepts belonging to a shared cultural identity are torn from their original contexts in sources outside the textfor instance, a poem from the Odesand brought into line with the argument of the Wu xng. The authors therefore not only expound the argument of the text but also explicate the references used in the text. At times, these exegetic layers also need some contextualisation to become meaningful. This can be carried out on different levels of the text, so that one can speak of exegetical processes of different orders. It is through this net of cross-referential textual links that the Wu xng becomes a selfcontained system, and only when identifying these links can the train of thought it develops be grasped. The difficulty of following these cross-referential links is that the various exegetic processes of the second and third order may be mutually intertwined, so it is not always clear how far the different concepts relate to one another. Despite this, both manifestations of the text follow this system of cross-referential links in a coherent manner. Reading the Wu xng therefore implies that one always has to follow two lines of argument processing: first, a steady and linear development of the wu xng theory and, second, the recurrent references to earlier passages of the text. Meaning Construction Subcanto 1 The opening passage of the Wu xng consists of five meaningful units of the first order, or building blocks. It spans eight bamboo strips. It mentions the five central virtues of this textbenevolence (rn ), rightness (y ), ritual propriety (l ), wisdom (zh ), and sagacity

five aspects of virtuous conduct

91

(shng )43and introduces the vital categories used in the text. These are the fundamental distinction between the concepts five aspects of conduct (wu xng ) and four aspects of conduct (s xng ); moral force (d )which I shall translate as virtueand goodness (shn ); aspiration (zh ) and action (wi ). All these will be contextualised later on in the text.
1.1 [A] w2 w3 [] [B] [] w4 [C] w1 The five aspects of [virtuous] conduct (wu xng) [are as follows]: When benevolence is manifested internally, we call it virtuous conduct;44 When it is not manifested internally, we call it conduct [only]. When rightness is manifested internally, we call it virtuous w2 conduct; When it is not manifested internally, we call it conduct [only]. When ritual propriety is manifested internally, we call it virtuous conduct; When it is not manifested internally, we call it w3 {conduct [only]}. {When wisdom is manifested} internally, we call it virtuous conduct; When it is not manifested internally, we call it conduct [only]. When sagacity is manifested internally, we call it virtuous w4 conduct; When it is not manifested internally, we [still] call it virtuous conduct. [] w5 [D] [Hence], virtuous conduct comprises five [aspects]; when they are brought into harmony, we call it virtue; when [only] four [aspects] of conduct are brought into harmony, we call it goodness.
w1

1.2

See strip w1/1w9/2. The pronoun we does not signal general agreement. Instead, the speaker of the text here makes a prescriptive claim in the sense that something ought to be denominated in a certain way.
44

43

92

chapter three
Goodness is the Way w5 of humans. Virtue is the Way of Heaven.

2. [] w6 [][E] [] [F] When the gentleman ( jnz) lacks concern in his inner mind, then [he] will be without wisdom in his inner mind. When [he] lacks wisdom in his inner mind, then [he] will be without w6 {delight} in his inner mind. {When [he] lacks delight in his inner mind, then [he] will not} be at ease. When not at ease, then [he] will not be joyful. When not joyful, then [he] will be without virtue. 3. w7 [G] [] [H] When all aspects of conduct are manifested internally, and w7 they are conducted at their [appropriate] time, then we call [that person] a gentle {man}. The scholar-knight who sets his aspiration on the gentleman-way is whom we call aspiring scholar. 4. [I] w8 [J] [K] [] [] [] w9 . Goodnesswhen refraining from acting for it, there will be nothing for approaching [it]. Virtuewhen refraining from w8 aspiring to it, [it] will not be accomplished. Wisdomwhen refraining from thinking about it, it cannot be attained. [This is because], if thinking is not clear, one will not be investigating, [and] if thinking does not grow, it will not manifest. If [thinking] is not manifested, one cannot be at ease; If not at ease, one cannot be joyful; If not joyful, w9 one will be without virtue.45

45 It is for two reasons that I suggest a change of subject in building block 4 and the following argumentative chains. At first sight, it might seem as if the foregoing

five aspects of virtuous conduct

93

In structural terms, building blocks 1.1 and 1.2 form one meaningful unit of the second order, or pericope; building blocks 3 and 4 form another pericope. They are linked by building block 2 (see figure below). The two pericopes share the same compositional features and terminology.46 The two opening lines of building block 1.2 sum up the account presented in 1.1 by pointing to the difference between the cultivation of five virtues and the cultivation of four virtues.47 The subsequent lines of building block 1.2 continue the argument and introduce the crucial distinction between goodness and moral force (d ). The five aspects of virtuous conduct and the four aspects of virtuous conduct are thus established as fixed concepts in the Wu xng. This distinction is taken up again in the second pericopenamely in building block 4where the idea is further contextualised. The concept of five aspects of virtuous conduct is given a new spin in building block 3. Seen from this perspective, building block 1.2 functions as a doubledirected feature of a text segment. It sums up the preceding textual unit and, at the same time, points to the continuation of the argument in what follows. Building block 2 seems out of place formally. It introduces the gentleman ( jnz )taken up again in building

argumentative chain had to be continued, which would suggest reading the line if [it (i.e., thinking)] is not manifested, [it] will not be at ease. But in fact, this sequence should be understood as composed of two argumentative chains (preceded by three headings), interlinked with each other. The second argumentative chain (that in which one, or he, had to be added) is a result of the first and, crucially, is taken from building block 2, where the subject clearly is the gentleman ( jnz ). As a result, the gentleman must also be the subject of the second argumentative chain in building block 4. 46 The concept five aspects of virtuous conduct (wu xng ) is used prominently in building blocks 1.1 and 3, and the concepts of virtue and goodness feature in building blocks 1.2 and 4, both of which are correlated with the concepts of five versus four aspects of virtuous conduct (wu xng / s xng ) in building block 1.2, as well as action versus aspiration (wi / zh ) in building block 4. 47 Since the publication of the Gudin One Wu xng, scholars have debated over the last line of building block 1.1: When it is not manifested internally, [we still] call it virtuous conduct . Because this line deviates from the basic pattern of 1.1, it has been argued repeatedly that this is a scribal error for what should in fact read When it is not manifested internally, it is called conduct onlyleaving out the two graphs (the x of virtue), as seen in the Mawngdu Three Wu xng. However, because building block 1.2 takes up the distinction between the inner realisation of either four or five virtues, I hold that it is not the Gudin One representation of this line that shows a mistake but the Mawngdu Three instantiation of the Wu xng, which displays some kind of correction (or simply a scribal error), suggesting that to later readers this passage might have become unclear.

94

chapter three

block 3and highlights the importance of wisdom for self-cultivation. It formulates the central idea of this unit. As mentioned, building blocks 3 and 4 further conceptualise ideas from the first pericope. The continuity of the train of thought is thus formally established, and this unit becomes a meaningful unit of the third order, that is, a stable subcanto, in which fixed concepts are established. Just as the concept five aspects of virtuous conduct is contrasted with that of fourmoral force with goodnessthis subcanto defines gentleman as a fixed concept by contrasting him with the scholar-knight (sh ). It is made clear that the Way of the gentleman can be achieved insofar as the individual aspires to it (zh ). Referring to the distinction between moral force and goodness, and adding to it the concepts of aspiration and action, building block 4 alludes to the fact that in the logic of the wu xng theory, goodness is a concrete issue that can be acted upon practically. Moral force, in contrast, is rendered an abstract idea that can only be aspired to. By paralleling wisdom with goodness and moral force, building block 4 formally illustrates that the process of gaining wisdom through thinking, s is in practical terms subject to the same pattern as that of nourishing goodness or moral force. At the same time, moral force is the final result of moral cultivation. The cultivation of wisdom, however, is a necessary precondition for achieving moral force.48 The present subcanto differs from its counterpart in the Mawngdu Three manuscript as follows. The Mawngdu Three version introduces the five aspects listed in building block 1.1 in the order of benevolence (rn), wisdom (zh), rightness (y), ritual propriety (l), and sagacity (shng), whereas the order in the Gudin One version is benevolence, rightness, ritual propriety, wisdom, and sagacity. For the fifth of the virtues, the Mawngdu Three version reiterates the formula used in building block 1.1.49 I believe that the Mawngdu Three version errs here in the presentation of the wu xng theory. The distinction between cultivating four virtues within and cultivating five virtues within, as
48 The formal parallelism that is created here is thus of philosophic importance and must be rendered in the translation of this passage. 49 It thus reads When [sagacity] is not manifested internally, it is called conduct [only] (172/18173/2; not counting lost characters). The Gudin One version states: When [sagacity] is not manifested internally, [we still] call it virtuous conduct w3/15w4/11; not counting lost or repeated characters).

five aspects of virtuous conduct

95

introduced in building block 1.2, corroborates this suggestion. This difference may reflect an alienation from the wu xng theory, or it may simply be due to a mistake that occurred when the Mawngdu Three instantiation of the Wu xng was fixed on bamboo. Building block 2 is twice as long in the Mawngdu Three version of the text. The chain, which is the same as in the Gudin One version, is repeated except that the Mawngdu Three version substitutes sagacity for wisdom in the second row.50 Png P believes that the Gudin One version omits this part.51 That may be the case. Yet, it could likewise be that the difference is due to an insertion in the Mawngdu Three version of the textunconsciously or consciously possibly for the purpose of making it more explicit. So far, we have no grounds on which this issue can be settled. Lastly, building block 4 is slightly longer in the Mawngdu Three version of the text.52 None of these dissimilarities distorts the logical organisation of this subcanto. Despite minor dissimilarities at the level of the individual building blocks, structurally the arrangement of the five building blocks that constitute this subcanto differs not in the slightest in the two manifestations of the text. Subcanto 1 must hence be seen as a stable unit. See figure 12. References to Odes The first unit in which the Wu xng quotes the Odes connects directly to subcanto 1. It takes up roughly three and a half bamboo strips53 and is composed of two building blocks. This unit, subcanto 2

50 The Mawngdu Three version thus reads 1 2 3 4 5 6 7 8 9 10 Lines 610 are not extant in the Gudin One version. 51 See Png P 2000, 32. 52 See my discussion in chapter 11, [I]. 53 From strip w9/3 to strip w12/17.

96

chapter three

1.1.

w2 w3 [] [] w4 a b a[] w5 b concludes 1.1 continues into 3 and 4

w1

pericope

1.2.

2.

[] W6 [] [] w7 [] a b w8 c c1 c2 [] [] [] W9 . Figure 12: Subcanto 1 of the Wu xng pericope

3.

4.

five aspects of virtuous conduct

97

of the Wu xng, dwells on the triangular relationship of benevolence (rn ), wisdom (zh ), and sagacity (shng ).
5.1 [L] w10 [M] w11 [][O] [] [] [P

[N]

If not benevolent, one cannot be clear in ones thinking. If not wise, one cannot grow in ones thinking. If neither benevolent nor wise, whilst not yet having seen a gentleman, the sorrowful heart w10 cannot be disturbed. Until [I] have seen the gentleman, [my] heart cannot be delighted. [But] when I have seen him, and when I have met him, then my heart will w11 {X be delighted}. {X Now this is} what this is about. 5.2 [] w12 54 {If not} benevolent, one cannot be clear in ones thinking. If not sagacious, one cannot be effortless in ones thinking. If neither benevolent nor sagacious, w12 whilst not yet having seen a gentleman, the sorrowful heart cannot be agitated. Until [I] have seen the gentleman, [my] heart cannot be stilled.

The odes referred to in this unit are known today as Caochng (Grasshopper)55 and Ch j (Presenting the Chariots).56 They describe someones desperate need to see the gentleman ( jnz ) to still his agitated heart.57 At first glance, it might seem that the two

54 As Csikszentmihalyi (2004, 283) points out, the line when I had yet to see the gentleman often appears in the Odes quoted in the Wu xng text, emphasising the effect of the gentleman on the outside world. 55 Mo 14. 56 Mo 168. 57 Note the internal conflict in the attempt to reconstruct the original meaninglet alone something like the urtextof the Odes. The text of the Mo tradition with its correlating set of interpretations is available today and was one exegetical tradition of the Odes available in Hn times. Yet, as we now know from the Kongz sh ln (see Ma Chngyun 2001, 1:1141, 119168), in the Warring States period there also

98

chapter three

building blocks do not constitute a coherent unit, as the corresponding idea of building block 5.1 appears to be closed with the formula {Now this is} what this is about.58 However, considering the feature called a distant parallelism,59 we arrive at a different conclusion concerning the proper demarcation of this unit, its stability, and also its principal concern. The two building blocks display a high degree of textual overlap.60 Only the last lines of building block 5.1 are not extant in 5.2. See figure 13.

5.1.

w10 w11 [] [] [] [] w12 Figure 13: Subcanto 2: Quoting the Odes

5.2.

existed at least one radically different reading of the Odes. The commentary added to the Mawngdu Three instantiation of the Wu xng suggests that by 168 BCthe date by which this manuscript had been producedthe interpretation of Odes as provided by the Kongz sh ln was still circulating. Given that the commentary to the Mawngdu Three Wu xng, which, despite a chronological difference of more than ca. 150 years, is in line with the Kongz sh ln reading of the Odes, it can be assumed that this particular approach to the Odes must have been prominent at least in the Warring States periodbefore it was displaced by the Mo tradition (note in this context the rather late spread of the Mo tradition; see Riegel 2001, 100). However, whereas texts like the Kongz sh ln allow a glance at the interpretation of the Odes circulating in the Warring States period, we still lack most of the text of the Odes. This situation should be kept in mind when dealing with the Odes today. Kern (2007a) describes this state of affairs in detail. 58 [] []. 59 Structural elements can have the function of binding larger units together. Delimitation theories in biblical studies have rendered this plausible by paying attention to so-called distant parallelism (Korpel 2000, 48). I refer to such structural elements that not only pattern one individual building block but also link most parts of one pericope to each other as distant (from distant parallelism), so as to denote its special function. 60 I have shown this with a dashed rectangle in figure 13.

five aspects of virtuous conduct

99

The latter building block repeats the formula of the former, except that it replaces the discussion of wisdom with that of sagacity. This suggests a correlation between the two. As with subcanto 1, in which building block 2 cuts through an otherwise uniform entity and so structurally signals the main idea of that unit, the latter lines of building block 5.1 are not extant in 5.2 and so form a structurally alien element cutting through an otherwise uniform pattern. As in the previous subcanto, and typical of this feature, these lines both formulate and signal the fundamentals of this unit. I refer to this as a principal insertion.61 In this context, the stress of subcanto 2 thus lies in the fact that seeing the gentleman results in delight. This is a central notion in the Wu xng and reoccurs throughout. To be delighted when seeing the accomplished one is not for its own sake. Instead, as detailed in the principal insertion of subcanto 1, the ability to be delighted is a basic precondition for man to complete his process of self-cultivation and to nourish moral force within. The unit under review should therefore not be read in isolation but elaborates the notion expressed in the principal insertion of subcanto 1. Delight, that unit proclaimed, is the necessary condition for activating the resonating chain that will finally lead to the cultivation of ones moral force.62 Seeing the accomplished person, in turn, is the precondition for being delighted, the present subcanto explains. As detailed, the principal insertion relates the two parallel building blocks 5.1 and 5.2 together. It thus creates a stable unit. Placed between two building blocks, it is a distant type of this textual device. Carrying the main thought of a textual unit, it stresses the necessity of seeing the accomplished one so as to complete the process of self-cultivation. The analysis further highlights the way in which the Odes are interwoven with the argument of the text. Using the Odes in the text, the Wu xng functions as a commentary to this anthology, as already remarked by Jeffrey Riegel and Martin Kern.63 But the same process also works the other way around. By putting the central claim of this subcanto in a quotation from the Odes, the Odes also add

61 Note the similarity of the principal insertion with the double-directed parallelism as described by Gentz (2006b). Both features cut through otherwise uniform units. But unlike the principal insertion, the double-directed parallelism does not express the core notion of a unit. 62 . 63 See Riegel 1997; Kern 2005b.

100

chapter three

weight to the philosophical position of the Wu xng itself. The two texts lend each other mutual support on various levels and in highly elaborate ways. Differences between the subcanto under review and its counterpart in the Mawngdu Three Wu xng remain on the level of minor lexical or scribal variation. Moreover, the Mawngdu Three instantiation of the text introduces the Odes by name,64 and the quotation itself is slightly longer.65 Yet, the organisation and internal structure of this subcanto differ not in the slightest in the two manifestations of the text. As in subcanto 1, subcanto 2 proves stable despite some variation at the level of the individual building block. Exegetical Processes of the Second Order The next unit of the Wu xngsubcanto 3consists of three building blocks, namely 6.1, 6.2, and 6.3, or slightly more than three bamboo strips.66 Its three argumentative chains neatly explain the precise meaning of concepts used in the previous subcanto. This unit highlights the connection between clear thoughts and benevolence; growing thoughts and wisdom; and effortless thoughts and sagacityall of which are necessary preconditions for seeing the accomplished one. See figure 14.
The thinking of the benevolent one is clear. [This is because, only] if ones [thinking] is clear, w13 will one be investigating.67 If investigating, one will be at ease. If at ease, one will be gentle. If gentle, one will be delighted. If delighted, one will be intimate. If intimate, one will be affectionate. If affectionate, one will be caring.

The ode is introduced by the formula sh yu (). It adds until I have not seen a gentleman, [my] sorrowful heart cannot be disturbed before continuing with the phrase quoted in the Gudin One instantiation of the Wu xng. 66 Strips w12/9w16/2. 67 As in the argumentative chain of subcanto 1, these units have to be read with a change of subject. As detailed, the argumentative pattern of the Wu xng is circular, on both the macrolevel of the argument and the microlevel of the text; the pattern of the argument always runs: Only if X does x, xi, xii, (etc.) can he be X.
65

64

five aspects of virtuous conduct

101

6.1

Elaboration of building block 5.1 and 5.2, line 1 [] w13 [Q] [] [][R] [] [] [] [] [] [] []

6.2

w14

Elaboration of building block 5.1, line 2 [] [] [] [] [] [] [] w15

6.3.

Elaboration of building block 5.2, line 2 [] [] [] [] [] [] [] w16 Figure 14: Subcanto 3

102

chapter three
If caring, one will be of a jadelike coloration. If being of a jadelike coloration, ones [thinking] will be manifested. If ones [thinking] is manifested, one will be benevolent. The thinking of the wise one is growing. [This is because, only] if ones [thinking] grows, will one be attaining. If attaining, one will be nonforgetting. If nonforgetting, one will be clairvoyant. If clairvoyant, one will see the worthy person. If seeing the worthy person, one will be of a jadelike coloration. If being of a jadelike coloration, ones [thinking] will be manifested. If ones [thinking] is manifested, w15 one will be wise. The thinking of the sagacious one is effortless. [This is because, only] if ones [thinking] is effortless, will [it] be manifested. If ones [thinking] is manifested, one will be nonforgetting. If nonforgetting, one will be clairaudient. If clairaudient, one will hear the Way of the gentleman. If hearing the Way of the gentleman, one will be of a jadelike tone. If being of a jadelike tone, ones [thinking] will be manifested. If ones [thinking] is manifested, w16 one will be sagacious.
w14

Subcanto 6 once more shows that in argument-based texts from the Warring States period, the formal structure of a textual unit is clearly detached from its material carrier, that is, the bamboo strips. None of the three argumentative chains corresponds to the length of a strip, even though it would have been easy to place each string on one striphad there only existed the concept of a correlation between formal structure and physical carrier.68 This plainly shows that for the recipient of an argument-based text (his or her role will be discussed in further detail in part II of this study) the physical appearance of an argument, as well as of the text as a whole, had no impact on the

I therefore disagree with William Boltz (2005), who attempts to explain the stable textual units from the materiality of their physical carrier, namely by pointing to the length of the bamboo strips. Dissimilar instances can only be made out for non-argument-based texts. For instance, the fragmentary Shunggudu manuscript of the Odes seems to record each ode on one individual strip. Edward Shaughnessys approach to reconstructing the original sequence of the Z y, that is, the source text available to the editor of the L j, [sic] follows such considerations. In his study, Shaughnessy hypothesises that in this source text, each textual unit must have been written on one individual bamboo strip. He also conjectures that the source text did not record more than twenty-one to twenty-four characters per strip. For the full account of his argument, see Shaughnessy 2006, 6470. For a critique of this monolinear argumentation, see Kern 2005b, 304, n. 22.

68

five aspects of virtuous conduct

103

comprehensibility of the text. It seems, however, that rhythm played a vital role. As already detailed, subcanto 3 explicates notions that were introduced in the previous subcanto. As such it naturally follows the passage to which it refers (w9/3w12/16) in order to be meaningful. The cross-referential links thus provide the Wu xng with textual stability. The Mawngdu Three text is organised in the same fashion. The fact, then, that this subcanto explicates individual lines from another subcanto furthermore suggests that the notions to be explained must have been issues of high importance in contemporary China. For the recipient of the Wu xng, they must have been concepts that were also recognised as conceptsotherwise, the authors of the Wu xng would have had no need to elaborate them in further detail. Only their meaning had become an issue over timeor the Wu xng targets an audience to which these concepts were potentially unclear. For us, the modern interpreters of the Wu xng, the exact origin of these concepts is no longer known. It is fairly certain, however, that they were taken from the pool of shared knowledge.69 It is furthermore instructive to note that the argumentative chains that dwell on notions from subcanto 2 are themselves highly enigmatic; especially the reference to jadelike coloration (y s ) or jadelike tone (y yn ). They remain unexplained here, but the Wu xng dwells upon these concepts further below. Comparing the two instantiations of the Wu xng, the first of the three argumentative chains differs slightly.70 This might be due to a conscious act of rewriting this passage, or it may simply be a

69 This is corroborated by the fact that the concepts jadelike coloration (y s ), jadelike tone (y yn ), and sound of bronze bells ( jn shng ) are used repeatedly in the text and also appear in transmitted texts from the late Warring Statesfor instance the L j (13:3/33) and Mngz ( 5B:1/10)and early Western Hnfor instance the Hn sh wi zhun (1:16/4; 1:16/4), among other. These concepts had a political connotation and were used as attributes of accomplished rulers. As Holloway states, they describe an external manifestation of a persons internal virtue (2002, 144). 70 After strip w13/6 ([]), the Mawngdu Three version continues: will not lack concern. When not lacking concern, then he will be of a jadelike coloration. When being of a jadelike coloration, [his thinking] will be manifested. When [his thinking] is manifested, he can be benevolent . Note that from the fact that this passage refers to a ready quoted instance, one can deduce that the missing part reads something like he knows about the Way of the gentleman. Knowing the Way of the gentleman, . . ..

104

chapter three

change that crept in unconsciously in the course of the transmission of the wu xng theory. But regardless of how it came about, the difference does not affect the structure of the subcanto. In terms of textual organisation, subcanto 3 remains stable in both manifestations of the Wu xng. As above, the formal structure of the subcanto proves more stable than that of the individual building blocks of which it is composed. Direct Reference to Other Sources The Wu xng refers not only to the Odes but also to other sources that were shared knowledge in China at the time. The next textual unit is interwoven with quotations from both the Odes and passages whose origin remain unclear to the modern interpreter. Subcanto 4 spans four and a half bamboo strips and consists of four building blocks.71
7.1
[S]

The polite man, the gentleman, unified in his deportment he is. Only after one is able to be unified [in ones deportment] will one be able to become a gentleman. [Hence, the gentleman] is mindful of his singularity.72 7.2
w17 [] w18 [] [U] w17 [T]

{I looked after her yet I could not reach her}, and my tears fell like rain. Only after one is able to disarray the wings will one be able to develop utmost grieving. [Hence], the gentleman is mindful of his w18 {singularity}.

71 These are building blocks 7.19. The whole unit runs from strip w16/3 to w20/17, not counting the graphs that are no longer extant on the strips and that had to be reconstructed. 72 See also Riegel 1997, 160. Riegel assumes that what he translates as innermost self (my singularity) points to the self-mindfulness of the gentleman, in that he heeds his own thoughts and not the world around him, a crucial element of the Wu xng (161). See also the discussion by Li Xnfng (2000, 315326).

five aspects of virtuous conduct

8. [] w19 [] [V]

105

In his acting for goodness, there is [always] something with which the {gentle}man begins and something with which he ends. In his acting for virtue, {there is [always] something with which} the gentleman {begins, but there is nothing with which} he ends. 9. [W] w20 [] [X] The sound of bronze and jade [stones] resonating along with itthis is a person possessing virtue. The sound of bronze is goodness. The tone of jade is sagaciousness. Goodness is the w20 Way of humans. Virtue is the {Way} of Heaven. Only after there is a virtuous person can there be the sound of bronze and jade [stones] resonating along with it.

The different building blocks that constitute subcanto 4 are linked by a discussion of the concept of gentleman ( jnz ). Each building block highlights a different aspect of this concept. Compared with other subcantos, this unit seems to be rather loosely structured. This contrasts with the rigorous makeup of the individual building blocks that constitute it. Building blocks 7.1 and 7.2 quote the Odes.73 They share the same structure. Together they constitute one pericope. The stress in the pericope is on the uniform deportment of the gentleman. Building blocks 8 and 9 constitute another pericope. By referring to the distinction between goodness (shn ) and moral force (virtue) (d ) in connection with the gentlemanas introduced in subcanto 1the unit finalises the first part of the Wu xng by summing up an idea central to the text.

73

Mo 152 and 28.

106

chapter three

8. a [] b a b w19 [] 9. 1A 1B 2A w20 2B [] C Quotation

Referential Explanation

Figure 14: Explanation of Quotation in Subcanto 9, line 1

Building block 9 consists of two parts. It is headed by a rather technical introduction of unknown origin that also appears in the Mngz.74 The different use of this line in the two texts suggests that it originated from yet another source, from which it was quoted independently. Building block 9 conceptualises this quotation. The referential explanation of this line is phrased in the pattern of an overlapping structure, as figure 14 shows. Despite the fact that the introductory statement is somewhat enigmatic, it serves as an authoritative peg for the argument of building block 9. This suggests that it belonged to the cultural knowledge of the time. Quoted here, it must have had authoritative value. Yet, just like the Mngz, the Wu xng feels the urge to explain this statement.75 The passage under review once more displays the high degree of (inter-)textual referentiality that characterises the Wu xng. The passage phrased in the pattern of an overlapping structure explains the introductory statement of this building block in the context of the argument. This statement, in turn, is a quotation from yet another source. But the explanation of this quotation is itself, nearly in its entirety, made up of elements (or notions) taken from other building blocks

See Mngz 5B1. The Mngz, however, offers a different interpretation of this sequence. This suggests that the two texts were copying a different source, and they must have followed different scholastic traditions of how this line should be understood. Compare this line also with a passage from the Odes, Xiao ya: Bi j (Mo 186): that man is like jade; do not treat like gold and jade your sounds (Karlgren 1950, 129). This passage, however, is situated in a different context and has been understood quite differently.
75

74

five aspects of virtuous conduct

107

of the Wu xng: its first ab group consists of terms taken from the introductory statement, and these elements also appear in subcanto 3; the second ab group is identical with the second ab group of building block 1.2 (subcanto 1). The explanatory passage of building block 9 rephrases the technical terminology of the introductory statement, by which the concepts used therein are highlighted. This once more corroborates the assumption that the introductory statement is a quotation from yet another source. The authors of the Wu xng substitute the word yn sound for zhn to resonate (written as () in the text), by which the two are given equal structural significance so that either of the two can be substituted for the other. When quoting from a body of shared cultural knowledge, the authors reproduce the original wording (introductory statement). In the explanatory parallel passage, they paraphrase it.76 The conclusion (c) of the explanatory passage again reproduces the terminology of the opening line (), trusting that the recipient will now understand the quotation through the elaborating parallelism from above. The same is true also for the pair sagacity and virtue. The first ab group has sagacity where according to the parallelism of this unitthe word virtue could be expected. This suggests the interchangeability of these concepts. It is hence made clear that being sagacious is itself the full realisation of moral force (or virtue). Thus, building block 9 is more than just the exegetic effort to contextualise the authoritative quotation in accordance with the philosophy of the Wu xng. Like the extracts from the Odes, the line quoted also formulates a central insight of the text. Parallel to the sound of bronze, which is brought about after the sound of resonating jade stones, goodness can only be brought about with the presence of the sagacious one. Whether this is through the mere presence of his model77 or through direct acts of initiation, the Wu xng remains silent about this. Building block 9 reads like a collage, made up of a quotation from an authoritative source and an explanatory passage composed of elements that are taken from other building blocks of the Wu xng. Strictly speaking, building block 9 is a further variety of overlapping structure:

Otherwise, the introductory statement would also have had to read instead of ; or the explanatory parallel passage would have had to read instead of . 77 Compare this with the notion expressed in Mngz 6A7.
76

108

chapter three

the argument, put in the form of an overlapping structure, adjoins an introductory statement, which is familiar to the target recipient. This statement serves as an authoritative peg for this passage, and the concluding line of the explanatory overlapping structure repeats the wording of the authoritative introductory statement. I call this feature of an overlapping structure authority-framed. As detailed earlier, the feature that I call overlapping structure signals a core juncture in the development of an argument. By connecting various aspects from previous building blocks and correlating these to newly introduced references, the passage under review brings the vital concern of the first part (canto 1, the meaningful unit of the fourth order) to its logical close. This is the distinction between goodness and virtue, on the one hand, and between the Way of humans and the Way of Heaven, on the other. The differences between subcanto 4 and its counterpart from Mawngdu Three remain on the level of minor lexical variation of certain graphs and a longer quotation of Odes in building blocks 7.1 and 7.2. As discussed, the first part of the Wu xngcanto 1is made up of a system of references that tightly connect different layers of the text. Some textual units comment on other, more cryptic parts. But the explanatory units can also contain elements that themselves remain obscure so that they require further contextualisation in yet other passages of the text. The Wu xng thus establishes a system of cross-referential links that span different text-referential layers. What at first sight might appear inscrutable in organisation does in fact lead to textual stability. Structurally, canto 1 of the Wu xng is stable in both manifestations of the text. Xng Wn closes canto 1 after building block 9.78 Because building block 9 finalises the argument concerning goodness and moral force (or virtue), this seems to be a sound assumption. However, compared with the other argument-based texts that can be reconstructed from the tomb corpus of Gudin One, the partition of the Wu xng into two cantos is more fluid. In fact, subcanto 5 should as a whole be considered the element of partition. It consists of four building blocks, two of which summarise the previous account. The remaining two building blocks introduce the second canto of the Wu xng. In this respect,

78

See Xng Wn 1998.

five aspects of virtuous conduct

109

subcanto 5 can be considered a distant type of double-directed text segment. For ease of argument I shall nevertheless begin my analysis of the second canto of the Wu xng with subcanto 5, whilst keeping in mind the bridging function of this unit. Canto 2: Structurally Different? The second part of the Wu xng is considerably longer than the first part but requires less discussion. Canto 2 is written on more than thirty strips.79 It contains twenty-one building blocks, or five subcantos. Differences between the two manifestations of the Wu xng mainly apply to this half of the text.80 The first stable unit of canto 2subcanto 5spans two bamboo strips.81 Of the four building blocks that constitute this unit, three are in the form of argumentative chains (namely 11, 12, 13). Each of these dwells on one virtue. Taken together, these building blocks constitute one pericope. Building block 10 differs. It does not focus on one virtue in particular but summarises the discussion of the virtues from canto 1 by relating the different notions to each other. One could even go so far as to identify building block 10 as an individual subcanto.
10. w21 [Y] [] [] [] [] [Z]

79 Canto 2 of the Wu xng runs from strip w20/17 (not counting the reconstructed graphs on strip w20) neither clairaudient nor clairvoyant to the last graph on strip w50. 80 Dng Sxn (2000a, 127ff.) divides the second part of the Wu xng into three larger units. His first unit runs from w16/3, beginning with the quotation from the Odes, to w22/18 []- []. His second unit runs from w22/19 <> w23 to w44/20 . His third unit runs from w45/1 w45 to the end of the text on w50. Xng Wn (1998) divides it into four units. His first unit runs from w20/18 - w21 - to w22/18 []- []. His second unit runs from w22/19 <> w23 to w32/4 w32 -[] . His third unit runs from w32/6 - to w49/15 () . His last unit runs from w49/16 to the end of the text on strip w50. Both suggestions disfigure stable entities of the text. 81 Subcanto 5 runs from strip w20/18 to strip w22/18.

110

chapter three
If neither clairaudient nor clairvoyant, one can be neither sagacious nor w21 wise. [This is because,] if not wise, one will not be benevolent [either]. If not benevolent, one will not be at ease [either]. If not at ease, one will not be joyful [either]. If not joyful, one will lack virtue.

11. [Aa] [] [] [] [] If not [having feelings] of affection, one will not be delighted [either]. If not delighted, one will not be intimate [either]. If not intimate, one will not be affectionate [either]. If not affectionate, one will not be caring [either]. If not caring, one will not be benevolent. 12. [Ab] [] [] w22 [] 82 [] If not upright, one will not be resistant [either]. If not resistant, one will not be decisive [either]. If not decisive, one w22 will not be dignified [either]. If not dignified, one will not carry out [ones tasks either]. If not carrying out [ones tasks], one will not be righteous. 13. [] [] [] [] If not keeping [appropriate] distance, one will not be respectful [either]. If not respectful, one will not be stern [either]. If not stern, one will not show honour [either]. If not showing honour, one will not be reverent [either]. If not reverent, one will lack ritual propriety.

82 The combination of xng with jian may have been a topos in early Chinese philosophical discourse. The D Di L j: Xiaobin has as for the Way, if one is not of grave demeanour, [it] cannot be carried out; if it is not carried out, one will not enjoy it (11.1/65/3).

five aspects of virtuous conduct

111

As discussed, the virtues of sagacity, wisdom, and benevolence are given central importance in canto 1 of the Wu xng. They are the focal concepts of building blocks 2 and 4 (subcanto 1), 5.1 and 5.2 (subcanto 2), and 6.16.3 (subcanto 3). Building block 10 correlates these virtues with the concepts of clairaudience and clairvoyance from building blocks 6.2 and 6.3 by reproducing a chain of mutually resonating virtues from building blocks 2 and 4.83 Building block 11 further dwells on benevolence, already introduced at an earlier point in the Wu xng. It thereby continues the argumentative chain of mutually resonating feelings from building block 6.1.84 This chain is a constant in the wu xng theory and most likely also in the intellectual world at that time. It can be assumed that it was a unit that the recipient of the Wu xng would have recognised. Building blocks 12 and 13 are entirely parallel. Together they constitute one pericope. Each of them discusses one virtue that has not yet been considered in canto 1 of the Wu xng. They introduce new argumentative chains on which the text will comment further below. Strictly speaking, these two building blocks introduce canto 2 of the Wu xng. Comparing the two manifestations of the Wu xng, the four building blocks of this subcanto prove extremely stableapart from some minor graphical differences in the first line of building block 11.85 Different, however, is the location of building block 10 within this

83

2. 4. [] w6 [] []

W8 [] [] [] w9

84 85

. Instead of (w21/11) as appearing in the Gudin One manifestation of the

(188/19). Most commentators transcribe this with text, the Mawngdu Three uses bin to change (see, however, my discussion in chap. 11, [Aa]). Also, whereas building block 10 of the Gudin One Wu xng reads w21 . . . If neither clairaudient nor clairvoyant, one can be neither sagacious nor w21 wise, the Mawngdu Three version repeats the latter part of this chain, thus reading . . . If neither clairaudient nor clairvoyant, one will be neither sagacious nor wise; if not sagacious not wise, . . .. The overall structure of this unit shows a clear preference for the Gudin One version and so it is likely that this difference results from a transmission error.

112

chapter three

subcanto. The Mawngdu Three manifestation of the Wu xng locates it at the end of subcanto 5, thus following building blocks 1113, whereas the Gudin One manifestation of the text places it at the top of this subcanto. This difference does not alter the reading of the text. As seen, building block 10 differs structurally from the other building blocks of this unit, so it does not need a fixed location in subcanto 5, as long as it is not placed between the other building blocks. Each of building blocks 1113 dwells on one of the virtues introduced in subcanto 1 (1.1), in the same order as is used there! And the Wu xng will further elaborate each of these building blocks below in the textagain, in the same order as presented in subcanto 5. Hence, building blocks 1113 together form a stable unit within subcanto 5. Placing building block 10 either at the beginning or at the end of subcanto 5 makes no difference. Only locating it between building blocks 1113 would have disfigured the consistent structure of subcanto 5. Ideas Central to the Wu xng Theory Subcanto 6 Nine building blocks constitute subcanto 6. They are written on nine bamboo strips.86 Central ideas of this subcanto are hearing (wn ) and seeing ( jin ) the Way of the gentleman ( jnz do ) and the worthy person (xin rn ). These notions are related to the concepts of clairaudience and clairvoyance, which refer to the virtues of sagacity and wisdom respectively (discussed in some detail in canto 1) but also to rightness and ritual propriety as detailed further below in the Wu xng.
14.1 w23 [Ac] w24 Never w23 having heard of the Way of the gentleman is what we call not clairaudient. Never having seen a worthy person is what we call not clairvoyant.

86

Subcanto 6 runs from strip w22/18 to strip w32/4.

five aspects of virtuous conduct

113

Having heard of the Way of the gentleman and yet not understanding w24 that it was the Way of the gentleman is what we call not sagacious. Having seen a worthy person and yet not understanding that he possessed virtue is what we call not wise. 14.2.
w25

[] [] [] [] w26
w25

87

[But] to see and understand him [as a worthy one] is wisdom. To hear and understand it [as the Way of the gentleman] is sagacity. To illuminate the luminous is wisdom; To hold in awe the awe-inspiring is sagacity. To illuminate the luminous refers to below88 [and] to hold in awe the awe-inspiring w26 refers to above89 are what this is about. 15.1. w27 [Ad] To hear the Way of the gentleman is clairaudience. To hear and understand it is sagacity. The sagacious one understands the w27 Way of Heaven. To understand and enact it is rightness. To enact it according to its time is virtue. 15.2. w28 [Ae] To see a worthy person is clairvoyance. To see and understand him w28 is wisdom. To understand and be at ease with him is benevolence. To be at ease and show respect to him is ritual propriety.

Cf. Mo 236. This means that to illuminate the luminous is a process that works from above to below. 89 This means that to hold in awe the awe-inspiring is a process that works from below to above.
88

87

114

chapter three

15.3. w29 [] [Af ] [] [] [Ag] w30 [] 90 [Ah] Sagacity and wisdom are those [virtues] from which ritual propriety and music originate w29 {and by which} the five [aspects] {of [virtuous] conduct will be harmonised}. When harmonised, one will be joyful. When joyful, there will be virtue. When there is virtue, states and families will revive. The way King Wn appeared was like this. When {King} Wn w30 {was on high, he was illuminated by} Heaven. That is what this is about. 16.1 w31 To see and to understand [the worthy one] is wisdom. To understand and be at ease with him is benevolence. To be at ease w31 and use him is rightness. To use and show reverence to him is ritual propriety. 16.2 w32 [] Benevolence and rightness are those [virtues] from which ritual propriety derives and by which the four [aspects] of [virtuous] conduct are harmonised. When harmonised, w32 there will be congruence. When there is congruence, there will be goodness.

Building blocks 14.1 and 14.2 constitute one pericope. They refer to earlier passages of the text, namely to subcantos 3 (in particular, building blocks 6.2 and 6.3) and 5 (in particular, building block 10). It was stated in these units that clairaudience and clairvoyance are the necessary preconditions for seeing the worthy one and hearing the Way of

90

Cf. Mo 235.

five aspects of virtuous conduct

115

the gentleman respectively. The pericope under review further elaborates this. It is made clear that being clairaudient and clairvoyant are the necessary preconditions for wisdom and sagacity correspondingly. Figure 15 demonstrates how building blocks 15.1 and 15.2 relate to building blocks 14.1 and 14.2. The different virtues are correlated with one another, and so argumentative chains are developed elaborating the five virtues central to the Wu xng in further detail. Structurally, building blocks 15.1 and 15.2 are almost entirely parallel in that they distinguish between two groups of virtues: those virtues that result from being clairaudient and those that result from being clairvoyant. Having defined the meaning of clairaudience as the ability to hear the Way of the gentleman,91 the recipient is now informed that sagacity is the result of realising this ability. It is imperative to develop a cognitive grasp, that is, an understanding, of ones ability to hear the Way of the gentleman.92 To apply this understanding, that is, to undertake appropriate action on the basis of this, is what the Wu xng defines as rightness.93 The ability to advance rightness at the appropriate time, finally, is defined as virtue (or moral force).94 Parallel to this, clairvoyance is defined as the ability to see the worthy one.95 To understand this ability is wisdom.96 Being at ease with the consciousness of seeing a worthy one is defined as benevolence.97 Showing reverence for the worthy one is what the Wu xng defines as ritual propriety.98 By implication, wisdom is given the same pivotal function for engendering benevolence and ritual propriety that sagacity has for developing rightness and virtue. The two building blocks hence define sagacity and wisdom as necessary preconditions for establishing the five virtues introduced in building block 1.1 without which nourishing virtue () is rendered impossible, and so they detail the distinction made in the Wu xng between key (, ) and dependent virtues (, , ).

91

The definition is in building block 15.1, thus explaining building blocks 6.3 and

14.1. See building block 15.1, line 2. See building block 15.1, line 4. 94 See building block 15.1, line 5. 95 The definition is in building block 15.2 (first line), thus explaining building blocks 6.2 and 14.1. 96 See building block 15.2, line 2. 97 See building block 15.2, line 3. 98 See building block 15.2, line 4.
93 92

116

chapter three
14.1. A B A B 14.2. B w25 A B [] A [] C [] [] w26 15.1. A1 A2 w27 A3 A4 15.2. B1 B2 w28 B3 B4 Figure 15: The Pericopes of Subcanto 6

If we now compare the distinction made in this unit between key and dependent virtues with the distinction made in canto 1 (building blocks 3 and 4) between acting for (wi ) goodness and aspiring to (zh ) virtue (or moral force), it is possible to further contextualise the different concepts of the Wu xng: building block 15.1 defines an abstract realm, whereas 15.2 details a concrete undertaking. Figure 15 shows that building block 15.1 displays the feature of a principal insertion. The third of the five linesthe sagacious one understands the Way of Heaven <>appears to be an alien element in an otherwise coherent argumentative chain. It has accordingly been explained as a misplaced commentary, because it interrupts the flow of the passage.99 A careful analysis of this passage leads to a different conclusion. The same statement appears at the

99

See Csikszentmihalyi 2004, 297.

five aspects of virtuous conduct

117

same location in the Mawngdu Three Wu xng, thus suggesting a conscious choice for patterning the argument. It is unlikely that both manifestations of the Wu xng which we possess today reflect the very same mistakeunless we assume that they were copied directly from one another or from the sameunknownthird source. Yet, given the structural differences in the organisation of the two manifestations of the textdifferences that cannot be explained by assuming a slip of the eye by the copyist or blamed on the materiality of the manuscriptsassumptions of this kind are not very plausible. It is also unlikely that a (written) commentary to the Wu xng existed when the Gudin One version was placed in the ground and that it was ignored by the two manifestations of the Wu xng, especially since the external commentary of the Mawngdu Three Wu xng does not identify this line as misplaced. Hence, instead of constructing scenarios for the way this mistake might have occurred in both manifestations of the text by conjecturing that the Gudin One Wu xng created an environment that inspired the same error in the Mawngdu Three text, it is more fruitful to try to solve this problem on the basis of the structure of the text, which is in so many ways consistent. As discussed, the feature that I have called a principal insertion is a vital device for patterning thought in early Chinese argument-based texts, and it is also used repeatedly in the Wu xng. Since the notion that the sagacious one understands the Way of Heaven proves to be a main idea of building block 15.1 (and of the Wu xng as a whole), it is reasonable to assume that it featured here as a conscious device to interrupt the flow of the passage and so to stress the idea central to this unit. Especially considering the fact that this building block dwells on the virtue (d ) of the sagacious person (shng rn ), this is clearly a much more productive approach to this line. Building block 15.3 applies the wu xng theory to the realm of politics. Only if the five virtues are harmonised will joy (l ) arise; and only if there is joy can there be virtue (d ). Virtue, in turn, is an essential quality of the ruler. The ode quoted corroborates this notion.100

100 Building block 15.3 draws on the Dya: Wn Wng (Mo 235), which reads when King Wn was on high, oh, he illuminated Heaven.

118

chapter three

Subcanto 6 takes a foremost position in the articulation of the philosophy behind the wu xng theory. Referring to the previous units, it differentiates between the key and dependent virtues, namely sagacity and wisdom versus rightness, benevolence, and ritual propriety.101 The subcanto further highlights how virtue () and goodness () can be achieved. But it also applies the wu xng theory to the realm of politics. By elevating virtue to the foremost principle of good government, the authors of the Wu xng display a political ideal not very different from that of the other argument-based texts discussed so far. Differences between the two instantiations of subcanto 6 are fairly straightforward. First, the Mawngdu Three Wu xng sometimes omits the particle y after definitions.102 Second, the Mawngdu Three manifestation displays minor lexical differences. These echo scribal idiosyncrasies and have no structural significance.103 Third, the Mawngdu Three version introduces the quotation of the Odes in building block 15.3 by name (sh yu odes state). Except for these rather insignificant dissimilarities, subcanto 6 proves exceptionally stable. Structurally, the two manifestations of the Wu xng are identical. Even on the level of the lexicon the two prove decidedly stable, and so is the sequence of building blocks within this subcanto overall. But the location of subcanto 6 in the two manifestations deserves attention. The order of subcantos 6 and 7 in the Gudin One manuscript is reversed in the Mawngdu Three version. This can

101 The dual treatment of different types of virtues therefore does not suggest that the Wu xng had a composite nature (as proposed by Csikszentmihalyi 2004, 65ff.) but is a vital device for articulating the wu xng theory. 102 In line 4 of building block 14.2; in line 4 of building block 16.1; in line 2 of building block 16.2. It has repeatedly been denied that y was originally a particle, and instead, it has been suggested that originally it was a full copula, to be. The fact that it is omitted in one manuscript might therefore hint at a diachronic phenomenon. I thank Wolfgang Behr for alerting me to this. 103 First, instead of reading to understand and enact it is rightness in line 4 of building block 15.1, which is also corroborated by the structure of this passage, the Mawngdu Three Wu xng reads to understand and enact it is sagacity . Because the commentary of the Mawngdu Three Wu xng refers to this line with rightness, not with sagacity, this seems to be a scribal mistake. Second, in line 2 of building block 16.1, the Mawngdu Three Wu xng mistakes n for the adverbalisation marker r . Third, in line 1 of building block 16.2, the Mawngdu Three Wu xng reads benevolence and rightness, those are from which ritual propriety and wisdom derive, instead of benevolence and rightness, those are from which ritual propriety derives. The analysis of the structure suggests that this is a transmission error. This is corroborated also by the fact that the commentary of the Mawngdu Three Wu xng repeats only two virtues, not three.

five aspects of virtuous conduct

119

be explained, and I will come back to this after discussing the next subcanto. Subcanto 7 Subcanto 7 spans five and a half bamboo strips.104 They constitute three building blocks. Central concepts of this subcanto are the inner mind (zhng xn ) and the outer mind (wi xn ). The subcanto also continues the discussion of the dependent virtues from earlier units of the text.
17. w33 [Aj] [] [Ak] When facial coloration, look, manner, and appearance are gentle, this is love. When using the inner mind to interact with others, this is delight. When the inner mind is delighted by this and transferred w33 on to elder and younger brothers, this is intimacy. To feel closeness and to extend this feeling is affection. To be affectionate and to be genuine about it is caring. To care for ones father and, secondarily to this, to care for others is benevolence. 18. w34 [Al] w35 [] When the inner mind w34 rightly enacts [something] by being discriminative, this is uprightness. To [be] upright and perpetuate this is resistance. To [be] resistant and not to fear the strong and powerful is decisiveness.105 Not w35 to harm the great Way for the sake of the petty ways is

These run from strip w32/5 to w37/15. Being decisive in this context means arriving at a decision and daring to carry it out. Png P 2000 notes that this combination also occurs in the Dya: Zhng mn (Mo 260): [Zhngshn F], he did not intimidate the widow and the poor and did not fear the strong and powerful.
105

104

120

chapter three
grave demeanour. To punish severely if there is a great crime is carrying out [the Way properly]. To venerate [the noble ones] and to honour the worthy ones according to their level is rightness.

19.

w36

[Am] w37

To use the outer mind when interacting with others is keeping [appropriate] distance. To keep [appropriate] distance while being dignified is being respectful.106 To be respectful and not remitting is being stern. To be stern and fearful w37 is showing honour. To show honour without being arrogant is being reverent. To be reverent when widely interacting with others is ritual propriety.

w36

The three dependent virtues of benevolence, rightness, and ritual propriety were already discussed in subcanto 5.107 That unit introduced the series of mental states without which the development of these virtues is impossible. But neither the mental states nor the virtues themselves were properly defined. Subcanto 7 undertakes to do this (see fig. 16). Structurally subcanto 7 bears a close resemblance to subcanto 6. Both units detail the meaning of (technical) terms that were used earlier in the discussionexcept that they do so on dissimilar levels: subcanto 7 explains the dependent virtues of the wu xng theory discussed first in building blocks 1113 (subcanto 5), whereas subcanto 6 explains the key virtues discussed first in building block 10 (subcanto 5). However, it is irrelevant for the development of the wu xng theory whether the discussion of the key virtues comes first or that of the dependent virtues. What matters for the construction of meaning is that the proper sequence of the building blocks within these subcantos

In the terminology of the wu xng theory, the concept of the worthy person (xin rn ) is connected to the notion of ritual propriety and the corresponding sentiments (see, in particular, subcanto 6, building block 15.2). As a result, the worthy person has to be added to the present train of thought. 107 See building blocks 1113.

106

five aspects of virtuous conduct

17. W33 []

121

explains

11. [] [] [] []

18. W34 explains W35 []

W36

12. [] [] W22 [] [] 13. [] [] [] []

19. W37

explains

Figure 16: Internal Exegetical Efforts

is maintainedand this is indeed the case in both manifestations of the text. It is very unlikely that this dissimilarity reflects a difference in philosophical agenda. Subcanto 7 displays extraordinary coherence in the two versions of the Wu xng. The only difference is a slightly dissimilar use of the particle y .108 The Application of Virtue to Worldly Affairs The remaining two subcantos of the Wu xngsubcantos 8 and 9together span twelve bamboo strips. Subcanto 8 is written on six and a half bamboo strips.109 It contains three building blocks. The focal

108 In building block 18, line 6, and in building block 19, line 3, y is omitted in the Mawngdu Three text. In building block 17, line 4, the Gudin One Wu xng omits y , but the Mawngdu Three Wu xng does use it. 109 Subcanto 8 runs from strip w37/16 to the end of strip w44.

122

chapter three

point of the discussion is the gentleman ( jnz ) and the application of the wu xng theory to law and punishment.
20. w38 [An] w39 [] [Ao] 1A w40 [Ap] 1B [] [Aq] 2A 2B w41 C [Ar] 110 w42 What is not succinct should not be enacted; what is concealed cannot be discriminated by the w38 Way; When there is a serious transgression, and it is punished severely, this is to be succinct. When there is a minor transgression, and it is pardoned, this is concealment. When there is a serious transgression and one refrains from w39 punishing it severely, this is not enacting [the Way properly]. When there is a minor transgression and one refrains from pardoning it, this is not being discriminated according to the Way. To be succinct as a term is something like to soften raw silk by boiling;w40 it is great but rare. To conceal as a term is something like to hide what is to be hidden; it is small but numerous. To be succinct is an aspect of rightness. To conceal is an aspect of benevolence. [Just as] hardness is an aspect of rightness, softness is an aspect of benevolence. To be neither forceful nor pressing, neither hard nor [too] soft is what this is w42 about. 21.1. 111 [As] 1A 1B 2A w43

110 111

Mo 304. Cf. Mngz, Wnzhng xi ( 5B1). In Mngz zhng y 1992, 397.

five aspects of virtuous conduct

2B C ()-[]( )
[At]

123

The gentleman gathers great achievements. When able to advance in this, one can become a gentleman. Those unable to advance in this should all remain in their hamlets. As for what is great but rare, w43 [the gentleman] is able to take from it. As for what is small but numerous, [the gentleman] is able to take from it. He who shines brightly when reaching the Way of the gentleman is whom we call a worthy person. 21.2. w44 [Au] When the gentleman w44 understands [the worthy one] and lifts him up, this we call honouring the worthy person. When [the gentleman] understands and serves him, this we call one who honours the worthy person. The latter means that a scholar-knight honours the worthy one.

As can be seen from figure 18, building block 20 consists of two longer sequences. The latter sequence is patterned as overlapping structure and, as a whole, serves as the c component of the preceding sequence. I call this form of overlapping structure interleaved. The appearance of the concepts dignified ( jian ) and circumspection (n ) in this unit is surprising, especially because the statement itself is quite enigmatic. The interleaved overlapping structure that elaborates these terms, together with a quotation from the Odes, further signals their importance. The two terms are correlated with enacting (or use in reference to the worthy one) (xng ) and the Way [of the gentleman] ([ jnz] do []) in the first sequence of the building block and with rightness and benevolence in the second one. These concepts are central to the philosophical terminology of the wu xng theory and were accordingly discussed earlier in the text. By this means, [being] dignified and circumspection, as termini technici, are integrated into the greater argumentative line of the text, and the Wu xng establishes a consistent use of its technical terminology. This is especially obvious when considering that rightness and enacting (or using) correspond systematically over the entire Wu xng, and that [being] dignified ( jian ) is here describedand thus explainedby being related to these two concepts. The use of these concepts to apply the wu xng theory to ideas of law and punishment

124

w38

chapter three

1A 1B 2A w39 [] 2B 1A w40 1B [] 2A 2B w41 C w42 Figure 18: Building Block 20 interleaved overlapping structure

reinforces the fact that they are technical terms in the context of executive force and, thus, commonly known to the contemporary lite, even though their precise meanings may have become unclear to the intended recipients of the Wu xng. Subcantos 6 and 7 state that the precondition for engendering both sagacity and wisdom is to understand (zh ) that one can become clairaudient (and so able to hear the Way of the gentlemanand thus become sagacious) and clairvoyant (and so able to see the worthy one thus become wise).112 Building block 21.2 further applies the concept of understanding (zh ) to the concrete issue of recognising the worthy one. The last statement of this building block further mentions the scholar-knight, introduced earlier in subcanto 1 (building block 3), where he was described as someone aspiring (zh ) to the Way of the gentleman. As systematised in the Wu xng, the zh group of virtues and the sentiments attached to them promote virtue (as opposed to the wi group of virtuesacting upon somethingthat enable one to develop goodness). To be respectful ( jng ) belongs to the sentiments accounting for developing ritual propriety. Ritual propriety belongs to the wi group of virtues, which promote goodness. Building block 21.2 hence combines the discussion of virtue and goodness and applies it to the realm of politics. Only the accomplished gentleman, it is stated, can understand the worthy one, lift him up,

112

See building blocks 15.1 and 15.2.

five aspects of virtuous conduct

125

and show him honour. Thus, to understand the Way of the gentleman also has a practical aspect, as shown in building block 21.2. It not only describes the theoretical issue of understanding the path that leads to moral force (or virtue) but also demands of the gentleman (ruler) that he combine the theoretical understanding of the inner cultivation of the heavenly Way with the practical aspect of elevating the worthy one. The unit under review thus unites concrete undertakingsthe four aspects of conduct (s xng ), which describe the Way of humanswith the abstract notion of cultivating all five virtues (wu xng ) and thus nourishing moral force within. Hence, the wu xng theory defends the notion that virtue (moral force) (d ) is an abstract matter. It equals mans inner cultivation of the Way of Heaven (tin do ). Goodness (shn ), in turn, denotes the application of this inner state to worldly affairs. Accordingly, what the text labels the Way of humans (rn do ) is the conversion of moral force into worldly affairs. The construction of the argument in subcanto 8 is the same in the two manifestations of the Wu xng. This is true also for the building blocks that constitute this unit. Slight differences remain on the level of the lexicon113 and the formulae with which the Mawngdu Three Wu xng introduces the Odes.114 Only building block 21.2 differs. Instead of saying when [the gentleman] understands and serves him (), the Mawngdu Three Wu xng reads the gentleman follows and serves him () and adds another line.115 These dissimilarities do not affect the structural consistency of the two manifestations of the text and so do not bear on the issue at hand. The final subcanto of the Wu xng runs from strip w45/1 to the end of strip w50. Four building blocks (2225) constitute this unit. Structurally, this subcanto is the same as that of the Mawngdu Three manifestation. There are minor lexical or graphical dissimilarities, and the Mawngdu Three version uses different formulae to introduce the Odes.116 But none of these dissimilarities bear on the issue at hand.

113 This implies the structural variation of certain graphs. Also, at times one of the versions omits the particle y or r where the other version has it. 114 Sh yu an ode states. 115 Preceding the line The latter is honouring the worthy by the [aspiring] scholar-knight, the Mawngdu Three version adds the statement The former is honouring the worthy by kings and dukes. 116 Here and there a graph that appears in one version of the text is missing in the other: in building block 22, the Mawngdu Three version repeats the formula

126

chapter three

Subcanto 9 as a whole is more stable than the individual building blocks of which it is composed, This is common in the Wu xng. Conclusion The Wu xng works only as an integrated whole. Every building block connects to argumentative strings that develop larger textual units, the pericope and the subcanto. Each subcanto of the Wu xng features as an agency that presents one idea of the wu xng theory. But just like the individual building blocks, none of these subcantos work in isolation. Related to one another through an elaborate system of crossreferential links by which the different textual units further explicate one another, the various subcantos need the reference to other textual units of the Wu xng to be fully meaningful. This referential system works on different textual levels. One unit quotes a generally known source and explains it in line with the argument of the wu xng theory. Other textual units supply further details and are themselves referred to by other parts of the text. And so forth. The different argumentative strings are conjoined at some point in the text, and the Wu xng, as a whole, presents a coherent vision. This holds true also for the manifestation of the Wu xng excavated from Mawngdu Three. This materialisation of the wu xng theory contains the same stable, meaningful units of first, second, and third order (building block, pericope, subcanto). Differences basically are due only to scribal idiosyncrasies or the way in which Odes are quoted. The various agencies of the wu xng theory, however, remain stable in both manifestations. This is true not only for the subcantos

[if ] the heart says throughout; in building block 24, the Mawngdu Three version adds the graph gives birth to after the first character. In the last line of the final building block (25), the Mawngdu Three version has have () instead of to be fond of (). The Mawngdu Three Wu xng also adds another line to building block 24: that person bestows this on others; not to find the right man is to deviate from the pattern. And it quotes the Odes slightly differently: line w48/714 quotes the Dya: D mng (Mo 236), which reads God on High looks down on you; do not be unfaithful in your heart (Karlgren 1950, 188, emended). The Mawngdu Three follows the Mo text and reads ln , whereas the Gudin One Wu xng reads xin worthy [one] (w48/9 ).

five aspects of virtuous conduct

127

but also for the system of cross-referential links that connect these agencies into a coherent theory of self-cultivation. The main differences between the Gudin One and the Mawngdu Three Wu xng are the different position of building block 10 in subcanto 5 and the reversed order of subcantos 6 and 7. However, neither of these differences affects the wu xng theory. Structurally, building block 10 differs from the other segments of subcanto 5, and so it has no fixed position in this unit. It sums up ideas on the key virtues of the wu xng theory, whereas building blocks 1113 develop argumentative chains for the dependent virtues, which are specified in later textual units of the Wu xng. It is instructive to note that the order in which the dependent virtues are discussed remains stable in the different textual unitsin both manifestations of the Wu xng. Thus, as long as building block 10 does not appear between the other elements of subcanto 5, the different position of this building block within that subcanto does not alter the reading of this unit, and so it can either head (Gudin One) or follow (Mawngdu Three) this sequence. Subcantos 6 and 7, for their part, express those insights that allow the wu xng theory to be applied to worldly affairs. Subcanto 6 details the meaning of sagacity and wisdom, the key virtues of the wu xng theory. Connected to this, subcanto 6 also finalises those references which help to contextualise the differentiation of the five virtues of the wu xng theory into two groups, namely those that should be aspired to (the zh group of virtues) and those that can be acted upon (the wi group of virtues). This distinction is of utmost importance for the philosophy of the Wu xng. Finally, subcanto 6 applies the wu xng theory to the realm of good rule and so is the key to identifying the political agenda of the text. Subcanto 7, in turn, finalises the links concerning the dependent virtues: benevolence, rightness, and ritual propriety. This unit provides the full breadth of ideas behind these virtues as defined by the authors of the Wu xng. But despite the vital role which the two subcantos play for developing the practical aspect of the philosophy behind the wu xng theory, none of them presents a conclusion. Instead, the formal arrangement of the Wu xng takes a circular form, and so neither of the two manifestations of the Wu xng can be said to anticipate a conclusion, regardless of what has been argued recently. It does not matter which of the two subcantos (6 or 7) comes firstas long as they remain in a location that does not destroy the texts elaborate system of cross-referential links. Neither of

128

chapter three

these differences reflects a different philosophical agenda in the two manifestations of the Wu xng. The two manuscripts are simply two instances of communicating the very same idea: a coherent wu xng theory, which can be considered a closed philosophical system. The wu xng theory details a theory of moral self-cultivation. The ultimate goal is to nourish moral force (d ). Moral force is achieved after cultivating the five virtues by which one has achieved sagacity. Like the compositional pattern of the text, self-cultivation has circular elements. None of the five virtues can exist in isolation from the others. If any of the virtues described is lacking, sagacity cannot be developed. But sagacity takes the lead in the formation of the other virtues and is the crucial factor for the entire process of self-cultivation. Hence, the wu xng theory expounds a philosophy of self-cultivation in which the end result of self-cultivation is also its point of initiation. I call this the paradox of self-cultivation. The paradox of self-cultivation is that realising any of the five virtues simultaneously depends on the accomplished cultivation of the other virtues. The way out of this paradox of self-cultivation, then, lies in a humans awareness (zh ) of possessing the five virtues within him, which is the vital moment in the process of moral self-cultivation. Thus, self-cultivation as presented here takes on a circular form, but its crucial element lies in the ability to step out of this circle withoutso to speakleaving it. The process of self-cultivation also determines the formal organisation of the text. As discussed, just like the theory of self-cultivation, the composition of the Wu xng takes on a circular form in which the final result is simultaneously the starting point. Right from the start, the Wu xng confronts the recipient with ideas that can be understood only after having made ones way through later parts of the text. None of the subcantos can be read in isolation since each unit of the Wu xng comments on other parts of the text. The recipient thus finds himself totally absorbed in the text. This is reinforced even further by the mantra-like language of the Wu xng. And whereas the other argument-based texts analysed so far prudently develop one overall argument in a rather linear fashion, the Wu xng, instead, directly confronts the recipient in the first building block with ideas central to its theory. Building block 1 could just as well be the introduction to the wu xng theory as the summary of the entire philosophy. The structure of composition hence is to be explained as a formal device that, mirroring the wu xng theory itself, compounds the thought central to this text.

five aspects of virtuous conduct

129

This makes it unlikely that the formal structure of the Wu xng results from being composed of even older texts. The system of referential links accounts for the fact that the Wu xng presents a coherent theory of self-cultivation, the wu xng theory. That some parts of the text discuss only wisdom or sagacity, whereas other units discuss only benevolence, rightness, and ritual propriety has nothing to do with different chronological text layers that had been rewritten later on, as has been assumed. Instead, it features the hierarchisation of virtues, a crucial aspect of the wu xng theory, and the necessity of applying an abstract theory to worldly affairs. Wisdom and sagacity are the key virtues in the long-term process of moral cultivation. Benevolence, rightness, and ritual propriety are dependent virtues and, hence, are the expression of having realised wisdom and sagacity. In other words, benevolence, rightness, and ritual propriety are the particular manifestations of what the accomplished personthe individual who has cultivated wisdom and sagacitypractically brings forth when involved in worldly affairs. To approach the Gudin One Wu xng as directly ancestral to the Mawngdu Three manifestation of the text is to pose the wrong questions. As shown in the analysis, the two manifestations of the text are made up of stable subcantos, rendered coherent by various cross-referential links. Each of these represents one aspect of the wu xng theory. These units matter. They are the agents of different ideas. When systematised, they together constitute the wu xng theory. One may thus equate the individual subcanto with what has been dubbed a unit of thought in another context.117 As agencies of ideas, and so the relevant units of the text, the subcantos are likely to be the very units that had once been remembered before the wu xng theory materialised on bamboo. This would also explain why each of the subcantos of the Wu xng displays such an exceptional stabilityan even higher level of stability than that achieved by the particular building blocks which constitute those subcantos. Since the subcantos are constituted of remembered agencies of different aspects of the wu xng theory which are fully stable in the two manifestations of the text and joined by a complex but coherent system of cross-referential links, we have no grounds on which to identify either of the two manifestations as ancestral to the othereven though the Gudin One manuscript

117

See Wagner 1999a.

130

chapter three

presumably is the older one. Not the physical manuscripts but the stable units constituting a closed theory to be realised in, potentially, different forms must therefore be the focus of investigation.118 Thus, in opposition to what can be found in the literature, neither manifestation of the Wu xng can be considered to be the more original version or to express a more genuine philosophy. They simply reflect two instances of communicating the very same ideain unfixed but written form. At the same time, the preceding analysis suggests that the Wu xng is not a creation ex nihilo. The many quotations, references, and stable textual units plainly show that the authors of the text worked with a mine of materials, from which they used stock phrases, concepts, and technical terminology. These materials were contextualised and systematised in such a way that together they form a consistent wu xng theory of moral cultivation.

118 See in this context Zumthor 1983, where the concept of text mouvance is introduced. For a discussion of the stability of texts in the Chinese context, see Kern 2001, 65, where the distinction is made between fluid texts, that is, demarcated entities that can be recognised, respected, and transmitted, and works of canonical status, that is, closed texts.

CHAPTER FOUR

THE XNG Z MNG CH HUMAN NATURE IS BROUGHT FORTH BY DECREE The analyses carried out in the previous chapters has shown that the macrostructure of argument-based texts can be rigorously composed. The Zhng xn zh do and the Qing d y sh are good examples. They are texts with a fixed formal structure, but this does not mean that they are also closed texts. Both texts might have remained open to some changesat least on the level of their lexicon. But their structure of composition is fixed. For a much longer argument-based text such as the Wu xng, no such rigorous, and hence definite, macrostructure can be postulated. But it is possible to describe certain features that aid the transmission of a coherent argument, which in this case is the wu xng theory. As discussed, the textual structure of the Wu xng features as a formal device for the wu xng theory that, mirroring the wu xng theory itself, intensifies the ideas central to this text. Thus, we have a closed philosophical system paired with the somewhat flexible composition structure of an argument-based text. The same holds true for the argument-based text Xng z mng ch (Human Nature Is Brought Forth by Decree), which is the focus of the present chapter. This chapter concludes my analysis of the argument-based texts from Gudin One and the first part of this study on the construction of meaning in the written philosophical discourse of the Warring States. The Xng z mng ch is the longest argument-based text from Gudin One. The manuscript contains about 1,550 characters written on sixty-seven strips. Prior to the excavation, this text was lost from memory. Nowadays there are scholars who conceive of it as part of an imagined Confucian tradition1 simply because it shares ideas attributed to Z S , Gngsn N , or Sh Shu .2 In this chapter, I discuss the Xng z mng ch and its textual relation

1 Characteristic of this is the discussion of the Xng z mng ch by Chn Nng (1998). 2 See, e.g., Gu Qyong 2001, 24; L Tinhng 2003, 125.

132

chapter four

to its counterpart as anthologised in the Shnghai collection of Chu manuscripts. Since publication, the Xng z mng ch has attracted close attention because it discusses the concept of human nature (xng ) in a way previously unseen. It provides a multilayered analysis of human nature by relating it to the human mind (xn ) and unshaped feelings (qng ),3 but it also examines human natures relationship to matters existing outside the subject self, namely the phenomenological world as manifested in the things [in the world] (w ) or music ( yu ). In so doing, the Xng z mng ch considers the impact of these matters on human nature and offers a phenomenological account of how ones unshaped feelings relate to human nature and external stimuli.4 Whereas other early texts that have come down to us provide only isolated statements on human nature in passing, the Xng z mng ch is the earliest example of a text that provides a fully developed analysis of this topic. Many Chinese colleagues celebrate the Xng z mng ch as an intermediate evolutionary stage of an imagined Confucian discourse on this concept. Intellectually and chronologically, this text is considered a missing link between the Lnyu (Analects) and the ideas on human nature discussed in the Mngz.5 The danger involved in such monolinear approaches to early thought will have to be considered elsewhere.6 Suffice it to say that such descriptions tend to treat philosophical amalgams as if they were coherent and systematic edifices of thought. Intellectual positions are extracted from rather heterogeneous materials and ascribed to two thinkers (Kongz and Mencius).

3 The specific denotation of this concept has long been disputed. Translations for qng range from what is essential or genuine (see Graham 1986) to reality feedback (Hansen 1995). According to Angus Graham, the meaning passion develops in the Sng period, but Michael Puett (2004, 37) argues that this shift is already hinted at in parts of the Xnz and the L j, where, according to Graham, qng already came to be imbued with emotional connotations (1986). As Puett argues convincingly, it is because of its broad semantic range that this word came to be so important to the philosophical discourse of the Warring States. For the semantic range of the term qng in pre-Buddhist Chinese, see Harbsmeier 2004. In this context, I translate qng as unshaped feelings. 4 Brindley 2006, 19. Note that Brindley translates qng in this quotation as emotions. 5 See, e.g., Png P 1998; L Wiwu 2000, 310; Y Zhpng 2000, 355; Wng Xngpng 2004; L Ru 2005. 6 See, however, chapter 3, where I discuss similar claims made for the Wu xng.

human nature is brought forth by decree

133

Neither chronological layers nor agendas of authorship are taken into account.7 The Xng z mng ch is also interesting for modern scholarship because of its use of terminology. The Xng z mng ch provides insights into the semantic and philosophical breadth of terms that were used in the philosophical discourse of that time and whose particular meanings have long been the subject of modern debate. The specific use of the concept qng , for which so many different translations exist, is an example. The Xng z mng ch draws on what tradition calls sh Odes to substantiate its ideas about human nature, and some scholars assume that it even alludes to concepts from the sh , Documents.8 Even more important than hidden allusions to shared concepts is the explicit reference to odes (sh), documents (sh), rites (l ), and music ( yu ) as one group. The Xng z mng ch discusses their respective functions for the individual in the process of moral cultivation. Moreover, it also explains the role which sagacious people, the shng rn , play in turning these mainstays of Chinese culture into tools for educating lite groups.9 This has fuelled scholarly discussions. Scholars tend to equate the above labels with the so-called Confucian Classics.10 And the sagacious personor personsto which the Xng z mng ch refers when discussing the educational impact of these tools of Chinese culture is generally identified unequivocally with

7 Tsuda Skichi (1946) regards the Lnyu as full of contradictions and anachronisms. For this reason, he considers it an unsuitable tool for analysing the philosophy of Kongz . For a short discussion of this view, see Schwartz 1985, 61ff. On the chronological layering and later interpolations of the Lnyu , see the highly controversial contribution by E. Brooks and A. Brooks (1998). For the same approach to other texts, including the Mngz, see also E. Brooks1994, which reproduces central aspects of the discussion. 8 Hung Zhnyn and Hung Wi (2003, 81) argue that on strips x89 the Xng z mng ch refers to the so-called sn d concept from the Hng fn (Glorious Plan) chapter of the canonised Documents. That the two texts draw on these concepts, however, is no proof that the Xng z mng ch refers to, let alone quotes, this body of texts. Note further that the dating of the Hng fn chapter is highly problematic. According to Michael Nylan, the early layers of the Hng fn probably date to the late fourth century BC when it was used as a manual for administration. During the Hn it became used as a canon on cosmology. (1992, 1344; 105148). 9 For different views on the relation of the Xng z mng ch to Odes, Documents, and other sources of Chinese culture, see L Tinhng 2000a, 2000b; Hung Zhnyn and Hung Wi 2003; Gu Qyong 2001; among others. 10 For a discussion of this term see Nylan 2001a.

134

chapter four

Kongz himself.11 The identification of odes (sh), documents (sh), rites (or ritual propriety) (l), and music ( yu) with the Chinese Classicsand of the shng rn with Kongz may seem straightforward, but for various reasons, these identifications are problematic. Transmitted texts such as the Lnyu , the Mngz , and the Xnz that modern scholars associate with what they call Confucian ideas do not use the appellation sagacious person in a consistent fashion.12 Also, even though it is true that traditionally Kongz is often presented as the compiler of the Odes13 (his name is even closely associated with the Kongz sh ln [Kongz s Discussion of Odes], a text from the Shnghai collection of Chu manuscripts),14 J Xshng nevertheless concludes from an account in the Zuo zhun that the process of compiling the Odes had already been completed when Kongz was a child. He contends that the sagacious person in the Xng z mng ch generally denotes cultural hero(es) of the past, not (only) Kongz .15 Moreover, the mention of the four constituents of Chinese culture in the Xng z mng ch is not evidence of a set of well-defined classics as early as the mid- to late Warring States period. Mentioning these constituents of Chinese culture explicitly as one group does not imply that odes, documents, rites (or ritual propriety), and music were well-defined textual bodies, let alone books, at the time when the Xng z mng ch was composed. Music and rites especially could also be general denotations of music and rites. Also, quotations from excavated manuscripts do not supSee, e.g., L Tinhng 2000a, 2000b; Gu Qyong 2001, 25; Puett 2004, 50. In the Lnyu , Kongz is generally called master ( f z ). Yet in book 9 (Z han ) he is referred to as sagacious person (shng rn). The Mngz consistently calls him sagacious personand so does the Xnz. Yet the Xnz also calls people like B Y sagacious. 13 The Kongz shji chapter of the Sh j notes that, from the 3,000 ancient odes, Kongz removed those that were mere repetitions of others and selected those that could be used to serve ritual propriety and rightness (Takikawa 1989, Memoir 17, 69ff. [3307ff.]). 14 See Ma Chngyun 2001, vol. 1. The Kongz sh ln received its title from the modern editors. The identification of Kongz in this manuscript was not uncontested. However, P Mozuo seems to have resolved the issue in favour of reading the graph in question as Kongz (instead of Buz or Bu Shng , a disciple of Kongz known as Z Xi , allegedly born in 508/7 BC). See P Mozuo 2001, 1314. See also L Lng 2000. For an overview of the discussions about the Kongz sh ln, see Li Xnfng 2002; Chn Tngshng 2004, 315ff.; Shaughnessy 2006, 19ff.; Xing Wen 2008, 3ff. among others. 15 See J Xshng 2004, 169. J bases his argument on a passage in the Zuo zhun, Xinggng , year 29.
11 12

human nature is brought forth by decree

135

port the idea of a well-defined corpus of Documents by the Warring States period. And although recent analyses have convincingly argued for a more or less consistent body of Odes in that period, probably the only well-defined and distinguishable corpus among the four, comparison of the different records nevertheless suggests that in those days the Odes were still highly unstable in writing16in parts even in phraseology. In fact, it seems that the Xng z mng ch refers to these constituents of Chinese culture as traditions, not as written texts. And Kongz ? The question must remain unanswered. Neither the opinion held by L Tinhng and others that the Xng z mng ch refers to him when describing the making of these cultural constituents, nor the one defended by J Xshng that in this context shng rn should be understood as a general reference to cultural hero(es) of the past, can be either verified or proven wrong. Of special interest for this chapter is that the Xng z mng ch has a closely corresponding counterpart in the Xng qng ln (Discourse on Human Nature and Unshaped Feelings)so named by modern editorsfound in the Shnghai collection of Chu manuscripts.17 Just like the different manifestations of the Wu xng, the Xng z mng ch and the Shnghai Xng qng ln share a remarkable overlap throughout the first thirty-five bamboo strips of the Gudin One manuscript18in both textual organisation and phraseology.19 The largely analogous part of the two texts is highly elaborate and cohesive. Even though the language makes use of brief and sometimes highly mnemonic statements, it is not as formulaic and enigmatic as that of the Wu xng. However, as with the two manifestations of the Wu xng, the remarkable similarity of the Gudin One Xng z mng ch and the Shnghai Xng qng ln in the first part of the texts dissolves further on. This loss of overall coherence is furthermore
See Kern 2005a. On the Xng qng ln from the Shnghai corpus (henceforth Shnghai Xng qng ln), see below. 18 Unless otherwise noted, the counting of the bamboo strips refers to the Gudin One Xng z mng ch. 19 The similarity of the two texts has given rise to the suspicion (see Ma Chngyun 2001, 1:2) that the two manuscripts, Xng z mng ch and Xng qng ln, may come from the same geographic area (Hbi) or even from the same site (Gudin). We should bear in mind that because the Shnghai Xng qng ln was found, not in a scientific excavation, but by tomb looters, who sold it to an antiquities dealer in Hong Kong, its place of origin can never be confirmed. See also my discussion below.
17 16

136

chapter four

accompanied by the lack of a concise organisation of the second part of the texts. The cohesive macrostructure of the Wu xng allowed me to provide an in-depth analysis of the relevant parameters that account for the fact that the two manifestations of the text (from Mawngdu Three and Gudin One) present the elaborate wu xng theory without substantial difference, let alone distortion, even despite some compositional dissimilarities. Accordingly, even though building block 10 from subcanto 5 of the Wu xng appears at a different location in the two manuscripts20 and subcantos 6 and 7 appear in reverse order in the two texts, these dissimilarities do not influence the communication of a coherent wu xng theory. In this chapter, I provide no such detailed form analysis. The Xng z mng ch shares many formal features with the other texts from Gudin One, which have been dealt with sufficiently in the previous chapters. But it also presents some considerable philological problems that cannot always be solved by referring to its counterpart from the Shnghai collection of Chu manuscripts. The Shnghai manuscript is not as well preserved as the one excavated from Gudin One, and many graphs on the bamboo strips are faded. Many strips survive only as fragments, and the graphs on these are frequently hardly legible. This would make a detailed form analysis of the entire text problematic. Nonetheless, the compositional structure of the Xng z mng ch does contain some features that are worth describing, and I refer to them briefly in the present chapter. The fact that we possess two manifestations of the text that are so similar overall but still differ substantially from each other in places calls for a closer analysis. To anticipate my conclusion, I believe that the order of building blocks throughout the first part of the Gudin One Xng z mng ch and the Shnghai Xng qng ln has helped to create a stablebut not totally fixedtext, and it is possible to discern a coherent system behind their arrangement. I feel justified in treating this part of the two texts as one canto, that is, a coherent and structurally closed part of the text. I consider this to be the core text, the theoretical framework of the philosophy of what we today call Xng z mng ch or Xng qng ln. The second part of each text seems to present a further, more detailed elaboration of the core

20 In the Gudin One Wu xng, building block 10 appears on top of subcanto 5, whereas it appears at the end of the same subcanto in the Mawngdu Three text.

human nature is brought forth by decree

137

part, providing the application of what has been outlined theoretically in the core text.21 As such, the second parts, which might have been grafted on to the core text, leave more room for textual variation of precisely the kind we see in the two instantiations of the text. The Text on Bamboo The Gudin One Xng z mng ch is written on sixty-seven bamboo strips that are 32.5 centimetres long. The strips are tapered towards both ends. They bear marks of two binding straps at a distance of 17.5 centimetres. Physically, the Xng z mng ch bears close resemblance to the Chng zh wn zh , the Zn d y , and the Li d . This suggests chronological and spatial proximity of manuscript production, that is, the preparation of the strips and the fixation of a text on them. The four manuscripts were most likely produced in the same workshop. However, as repeatedly mentioned, the physical similarity of these manuscripts provides no information about the intellectual orientation of the texts they carry. Of the sixty-seven strips that constitute the manuscript carrying the Xng z mng ch, nine have broken. One of these strips has broken off at both of its sides. The missing parts probably contain approximately twenty-nine graphs. The Xng z mng ch is a difficult text. Numerous graphs still await conclusive identification, and the proper sequence of the strips is also still an issue. Even the find of the Shnghai Xng qng ln has not resolved all these matters. Prior to the publication of the photographs and the transcription of the Shnghai Xng qng ln,22 scholars had proposed different arrangements of the sixty-seven bamboo strips that constitute the Xng z mng ch manuscript.23 Briefly, for the first thirty-five to thirty-six strips, mainly rather moderate emendations of the reconstruction proposed initially in 1998 have been proposed.24

21 This observation corresponds with Gu Ys (2004, 1) remarks about the distribution of ideas in the Xng z mng ch. 22 See Ma Chngyun 2001, 1:69115, 215301. 23 See, among others, L Lng 1999; L Xuqn 1999b; Zhu Fngwu and Ln Sqng 1999; Qin Xn 2000; Li Xnfang 2000; Lio Mngchn 2000b; L Tinhng 2000a, 2003, 613; Chn Wi 2000a, 2003, 173207. 24 For the photographs of the strips, see Hbi shng Jngmn sh bwguan 1998, 5066; for the reconstruction see ibid., 177185.

138

chapter four

Scholars came up with far more interventions for the last twenty-one or twenty-two strips. These proposals included the insertion of some strips at the head of the lower part of the Xng z mng ch, which by now have been identified as belonging to another text, the Lu d (Six Virtues).25After the Shnghai Xng qng ln was made public, Lio Mngchn was probably the first to defend the arrangement of bamboo strips as originally suggested by the editors of the Hbi Province Museum.26 Text Division and Overall Organisation The question of whether the Xng z mng ch should be considered as one integral text or be divided into two or maybe even three individual texts is also an issue of fierce dispute.27 Objections to viewing the Xng z mng ch as one integral text are mainly based on three interconnected considerations. The first objection derives from the fact that different parts of the text dwell on different subjects. Based on this, L Xuqn , as representative of the proponents who regard the Xng z mng ch as two different texts,28 divides it after strip x36. According to L , the former half of what modern editors call the Xng z mng ch (strips x136)29 mainly deals with the effect of music on moral cultivation. He proposes to call this part Discourse on Music, or Yu shu . The latter part of the text (strips x37end) mainly discusses unshaped human feelings (qng ). He calls it accordingly Human Nature and Unshaped Feelings, or Xng qng .30 Advocates of the idea that the Xng z mng ch should be divided into two texts find further confirmation from physical features of the manuscript, as it displays dissimilar characteristics in its calligraphy. L Tinhng notes that the two parts (x135 and x3767) as distinguished by L Xuqn are written in two different hands.31 As L
25 Chn Wi (2000a, see esp. 6566) proposed placing the initial five strips of the Gudin One Li d at the top of the second part of the Xng z mng ch. 26 See Lio Mngchn 2000a. 27 See L Xuqn 1999b; L Tinhng 2003; Gu Y 2004. 28 See L Xuqn 1999b. 29 L Xuqn interchanges the sequence of strips and so considers x35 as closing this unit. 30 See L Xuqn 1999b, 23, 27. 31 See L Tinhng 2003, 1112.

human nature is brought forth by decree

139

Tinhng observes, the style of the graphs on strips x135 is rather crude. The graphs are written at some distance from each other, resulting in a far lower number of graphs per strip in x135. L Tinhng counts twenty to twenty-five per strip. The characters on the bamboo strips subsequent to strip x36 are much finer in style and written less far apart. Accordingly, the latter thirty-one strips of the manuscript have, on average, twenty-four to thirty graphs each (on unbroken strips). L concludes from this that the two parts were not only copied by different hands but also fixed on bamboo at different times. As a result, L contends that they should be considered different texts. The assumption that the Xng z mng ch should be divided into two individual texts seems to be further corroborated by the fact that the two strips x35 and x67the two strips are considered the final strips of the units described by L Tinhng and L Xuqncarry the so-called tadpole symbol ( ). In the Gudin One corpus, this symbol is also seen on the final strips of the manuscripts now called Laoz A32 and Chng zh wn zh.33 Not much is known about the early practice of interpunctuation,34 but it seems that this mark signals the end of a self-contained text (or unit). Just as is seen in the other manuscripts, the two strips of the Xng z mng ch that carry the tadpole symbol bear no further writing after this marking. About half of each strip is left blank. Despite such indications for the division into two individual texts, the idea of a partition soon met with strong reservations. First of all, in an influential article published in 2000, Lio Mngchn noticed a third marking in the Gudin One manuscript.35 Squeezed at the very end of strip x49, there is another mark ( ), which, however, is no longer clearly visible. Lio interprets this as a further instance of the tadpole symbol, except that this one is much smaller, probably because there was not much room left on the strip to make the mark. The publication of the Shnghai Xng qng ln in 2001 corroborated Lios observation to a remarkable extent.36 Strip xq40 of the Shnghai manuscript, which closely corresponds to strip x49 of the
On strip a32 and on strip a39. On strip ch40. 34 See, however, Guan Xhu 2002. See also Richter, forthcoming. 35 Lio Mngchn 2000a, 19. 36 For the Shnghai Xng qng ln, see Ma Chngyun 2001, 1:69115 (for the photos of the strips), 217301 (for transcription and commentaries).
33 32

140

chapter four

Gudin bamboo-strip text, also carries a tadpole symbol. As in the Xng z mng ch, the mark follows the exclamation [this] truly is the case (). After this, the strip of the Shnghai Xng qng ln carries no further graphs. Nearly the entire strip is left blank. But in the Shnghai Xng qng ln this exclamation appears on the final strip of the manuscript, closing the account of human nature; in the Gudin One manuscript, it appears right before the latter third of the text. The fact that the Shnghai Xng qng ln corresponds so closely with the Gudin One Xng z mng ch suggests the two units that I tentatively call the core text and its application indeed belong together and should not be regarded as two individual textsat least at this stage of textual development. Hence, the Gudin One Xng z mng ch contains three marks of division (the tadpole symbol, seen on strips x35, x49, x67), suggesting that it has been organised into three distinct parts. The Shnghai Xng qng ln confirms two of these marks. Following what corresponds to the core text of the Gudin One Xng z mng ch, the Shnghai Xng qng ln carries a big square mark ( ) that occupies the entire width of this particular bamboo strip (xq21). This seems to confirm the division of the materials into core text and its application. Moreover, the Shnghai Xng qng ln also confirms the demarcation after the exclamation this truly is the case from the Gudin One Xng z mng chexcept that in the Shnghai manuscript the exclamation signals the end of the entire text, whereas it appears in the latter third of the Gudin One manuscript. Both texts use the tadpole symbol here.37 The similarities between the Gudin One Xng z mng ch and the Shnghai Xng qng ln are striking. The differences between these texts simply exemplify what happens writing down a fully developed and coherent philosophyhere on the conflict between human nature and the phenomenological worldin two independent (!) instances. I will take this up in more detail below. The remarkable physical similarity of the two manuscripts (e.g., the analogous use of punctuation marks) has at least two more implications. To begin with, the analogous use of the tadpole symbol in two individual copies suggests that it does not necessarily signal the end of

37 As far as I am aware, Gu Y (2004) is the only scholar who still defends the partition of the Xng z mng ch (and likewise the Xng qng ln) into two individual texts. For a critique of his views, see my discussion below.

human nature is brought forth by decree

141

a text. It seems more probable to assume that it indicates the end of a self-contained textual unit within a coherent text. The analogous use of markings in the texts further suggests that the Gudin One Xng z mng ch (and also the Shnghai Xng qng ln) should not be divided into smaller individual texts. Accordingly, each manifestation should rather be understood as one text composed of different parts, and these parts were marked off accordingly. Whether these parts may also have circulated individually, or whether they existed only in combination with one another, as Chn Wi poses the question,38 cannot be answered with certainty at this point. Furthermore, the overlap of formal markings in two individual manifestations of a text strongly suggests that what modern editors call the Xng z mng ch (or the Xng qng ln) was a fairly stable text and had been in circulation in written form by the time that Gudin One was closed. As will be discussed below, the scribes of the Gudin One Xng z mng ch and the Shnghai Xng qng ln must have copied these markings separately from a third, written source text (Vorlage).39 Two Manifestations of the Treatise on Human Nature and Feelings Based on the similarity of the two texts, I hold that the Gudin One Xng z mng ch and the Shnghai Xng qng ln represent two independent instances of writing down one consistent theory of human nature and human unshaped feelings.40 The Shnghai Xng qng ln is badly preserved. The remaining text is written on about forty bamboo strips.41 These were cut evenly at both ends. Of these strips, only seven have remained intact.42 Intact strips carry between thirty-one and thirty-four characters, many of

See Chn Wi 2003, 176. Vorlage is a well-established term in biblical studies, where it commonly refers to a written model or source text for a new manifestation of that text. Vorlage as used here should not be misunderstood as a master copy of a text, or urtext. 40 Just like the Gudin One Xng z mng ch, the Shnghai Xng qng ln bore no title. Their present appellations have been chosen by modern editors. Virtually nothing is known about their original designations. 41 Five strip fragments that have not yet been clearly identified probably belonged to the Shnghai Xng qng ln. 42 Strips xq1, xq8, xq9, xq10, xq20, xq24, and xq28.
39

38

142

chapter four

which are faded. The exception is the initial strip, xq1. It has been inscribed with forty-one characters. Intact bamboo strips of the Shnghai Xng qng ln are circa 57 centimetres in length and so substantially longer than those of the Gudin One Xng z mng ch. In fact, they are the longest of the entire Shnghai collection of Chu manuscripts. That the two manifestations of this text had been fixed on strips of dissimilar length once more corroborates my contention that the physical length of the bamboo strips that constitute a Warring States manuscript does not reflect the status of the text it carries. There were no unified standards for the length of the bamboo strips used for texts during the Warring States. Accordingly, we should be cautious when inferring sociopolitical standing of Warring States manuscripts only on the basis of their material properties. Comparing the Gudin One Xng z mng ch with the Shnghai Xng qng ln, we observe two things. I have already mentioned the extraordinary consistency of the two texts: except for a few sentences that occur only in the Gudin One manuscript, the phraseology of the two texts is very similar. Moreover, the first thirty-five strips of the two texts share virtually the same organisation. This correspondence, however, substantially diminishes in the second half of the texts. Because we now possess two largely analogous instantiations of the same theory of human nature and moral self-cultivation, issues concerning the overall organisation of the text can be clarifiedespecially since the Shnghai strips are considerably longer than those of the Xng z mng ch, and thus, the sentences with which they end differ from the sentences that end the Gudin One strips. Some scholars even go so far as to hold that the Shnghai manuscript resolves all matters of textual organisation.43 Sadly, this is not the case. Even though the Shnghai Xng qng ln provides evidence for some strip clusters in the Xng z mng ch, we still have to explain why the individual building blocks by and large remain stable in the two instantiations of the text whereas the overall sequence of some of these differs strikingly.44 This is especially noteworthy since the Xng
As articulated lately by Lio Mngchn (2000a). Stable clusters of strips are as follows. The text recorded on strips x133 from the Xng z mng ch largely corresponds to that on strips xq121 from the Shnghai Xng qng ln. The text on strips x3649 largely covers xq3241 from the Shnghai manuscript. The text on strips x5059 largely corresponds to that on xq2128; the texton strips x5962 by and large overlaps with that on xq3032; the text on strips
44 43

human nature is brought forth by decree

143

z mng ch and Xng qng ln correspond so closely in their formal makeup. Briefly, the second half of the two texts (x3667 of the Gudin One Xng z mng ch) consists of two larger corresponding text sequences. However, they are located at different places in the two texts. The first sequence contains units 13 from the Gudin One Xng z mng ch, but it appears much later in the Shnghai Xng qng ln (it appears there as units 79). The second corresponding sequence consists of units 58 from the Gudin One Xng z mng ch. It appears as units 36 in the Shnghai Xng qng ln. Units 4 and 9 from the Gudin One manifestation constitute one stable cluster in the Shnghai Xng qng ln.45 Thought and Contents Two basic assumptions guide the discussion of human nature and the phenomenal world. First, the Xng z mng ch46 states on the initial strip that it is the common feature of men to have a human nature,

x6267 corresponds to that on xq2830. Moreover, the clusters on strips x6365 and x6566 can also be confirmed from the Shnghai manuscript, although the phraseology of the two is not the closest match (see also Lio Mngchn 2000b, esp. 15ff.). Chn Wi (2000a, 64ff.) suggested inserting the initial five strips from what has now been identified as an independent text labelled Li d (Six Virtues) at the head of the second half of the Xng z mng ch. The Shnghai Xng qng ln did not substantiate this idea. That the Shnghai manuscript does not contain these materials clearly does not prove that they were also missing from the Gudin One Xng z mng ch. But it provides strong evidence against the assertion that they were, and Chn corrected his earlier assumption in a later publication (Chn Wi 2003, 96). 45 The following table might help to visualise this different arrangements of the second halves of the Xng z mng ch and of the Xng qng ln: Gudin One Xng z mng ch Shnghai Xng qng ln 1. x36 2. 3. 4. x50 5. 6. 7. ?? 8.x62 9. (?)
46

1. 2. 3. 4. 5. 6. 7. . . . 8. . . . 9.

Unless stated otherwise, in the following discussion I refer to both texts when mentioning the Xng z mng ch.

144

chapter four

called xng . Second, the proposition that human nature is universal is defended. This is put as follows:
Generally speaking, even though all humans have a nature, they have no settled will in their minds. 47

and
Human nature is one within the four seas, [yet] in the application of mind each man differs. 48

For the authors the origin of human nature is undisputed:

Human nature is brought forth by decree; decree [in turn] down from Heaven. x349
x3

is sent

The authors do not attempt to justify this belief. The Xng z mng ch connects to a larger discourse in which the idea that humankind shares a universal nature, which each individual receives from Heaven, enjoys unanimous consent. The Xng z mng ch points out an inconsistency between human nature and human conduct. Even though humans share a common heavenly endoweddisposition, people realise different degrees of moral cultivation. This obviously is a problem, if not a contradiction, and so features as a pivotal concern of the text. If the primary supposition of the text is true and humankind shares a common humanity, then the implicit question is, how can it be that individuals realise different degrees of moral cultivation (or even lack it entirely)? This uncertainty is nowhere mentioned explicitly. However, it triggers the authors efforts and underlies the entire discussion of the treatise. When comparing this uncertainty with the impulse of the Qing d y sh, a significant overlap between the two becomes obvious. The motive of the Qing d y sh lies in the fact that despite his degree of moral cultivation, man nevertheless faces an uncertain fate as decided by Heaven. He remains a mere plaything of Heavens will. But man

47 48 49

Strip x1/19. Strip x9/619. Strips x2/20x3/3.

human nature is brought forth by decree

145

can overcome this vulnerability by making the cultivation of moral forceor virtue(d ) his only concern and so create a realm in which his own virtue is all that matters to him. In this autonomous sphere of the individual, he no longer depends on Heavens goodwill. The Xng z mng ch deals with the same problem but approaches it from the opposite side. What drives the Xng z mng ch is not the uncertain end of the individual as decided by Heaven and regardless of his morality but the fact that all men start off with the same potentiala universal nature endowed by Heavenand yet develop so differently. Accordingly, there must exist an intermediate stage between peoples common humanity and their later development. This intermediate stage bears a close resemblance to the autonomous sphere of the individual as detailed in the Qing d y sh. It is an area in which Heaven has no direct control over man. Only man himself (or the environment he creates) sets the course for moral cultivation. This self-sufficient room for moral cultivation is the focus of investigation of the Xng z mng ch. A Theory of Moral Cultivation A human beings mind (xn ) has a will (zh ), so the authors of the Xng z mng ch say. But in accordance with the development of human nature, the will is not predetermined.50 Alien factors such as the phenomenological world of things (w ), delight ( yu ), and practice (x ) are responsible for shaping the will of the mind. The will of the mind (xnzh ) determines human nature.51 Accordingly, self-cultivation inevitably starts with the cultivation of human will. This calls for the social environment to be improved. The fundamental nature of a goose is to stretch its neck and that of an ox is to grow big, the Xng z mng ch points out.52 The individual, in contrast, responds to his (social) environment.53 The key here is

See strip x1. See strip x6. [] {This is just like} even though {humans} have a nature, if [their] minds fail to accept it, it will not be manifested. 52 Strip x7/11 to end states: [ . . .] After the ox is born, it grows [big]; after the goose is born, it stretches [its neck]. Their nature {causes them to be like this . . .}. 53 The text states: Emotions such as rejoicing (x ), anger (n ), grief (a ), and sadness (be ) are human nature. Yet when it comes to them appearing on
51

50

146

chapter four

education, in which the Odes, the Documents, rituals, and music play vital roles. The Xng z mng ch thus allocates an educational role to society in shaping the human mind and recognises the impact of culture on individuals education. That the text puts forward the idea that the will of the mind closely relies on stimuli that it receives from the phenomenological world may suggest that the Xng z mng ch defends the view that human nature equals a blank slate.54 Depending on how he is inscribed, man will develop. But this is not the case. As noted, the text asserts that human nature derives from Heaven.55 The do , which in the Xng z mng ch clearly is a concept that transcends the phenomenological world and has the flavour of a true ethical code, nevertheless originates with the unshaped feelings of the individual. These are, in turn, part of his nature.56 Thus, the Xng z mng ch describes an integrated relationship of human feelings, true ethical code, and human nature. In this model, the unshaped feelings, which stimulate the do, are an integral element of man. Pursuing this line of thought to its logical conclusion, the true ethical code must already be part of man. It needs only to be activated. This, then, is why man can respond to the positive stimulus of education. He is attracted by sincerity as expressed in the resources of culture, such as the Odes, the Documents, rites, andin particularmusic. It is against this background that the notion continued on that same bamboo strip can be fully appreciated. In the beginning [the do ] approximates unshaped feelings; in the end [it] approximates rightness.57 This implies that the true ethical code (do) which lies within man proceeds without disruption from unshaped feelingsor the initial (raw) disposition of human nature to rightness. As Erica Brindley rightly states, this suggests wholesome, organic change that does not violate human nature.58 Thus, {he who} understands {unshaped feelings can}x4 let them out; he who understands rightness can internalise it.59 In other words, once the individual understands that unshaped feelings are the incentive to
the outside, it is due to the things [in the world] having brought this about (strip x2/219). 54 See also Brindley 2006, 21. 55 x3 (strips x2/20x3/3). 56 [] (strip x3/410). 57 (strip x3/1118). 58 See Brindley 2006, 23. 59 [] x4 (strips x3/194/8).

human nature is brought forth by decree

147

moral cultivationand so an intrinsic element of the true ethical code (do), which is in turn an integral part of man himselfhe can allow the full expression of his unshaped feelings without fearing transgression.60 And likewise, only he who understands that rightness is a logical position of the true ethical code, which is in turn an integral part of the individual self, can internalise it. Rightness, accordingly, is no longer alien to the individual. Conversely, it becomes an intrinsic part of himjust like the do.61 Thus, the Xng z mng ch claims that the everlasting (or repeated) practice of a virtue leads to the embodiment of its spirit. In sum, the Xng z mng ch draws a picture in which man is not intrinsically good. But he does have an innate affinity for the good. Hence, like the Mngz and the Xnz, the Xng z mng ch does not have a static concept of human nature. It is precisely due to his affinity for the good that the moral cultivation of man is possible in the long run. Because the true ethical code (do) lies within man himself, he tends to respond to sincere feelings, not to artifice.62 Sincerity, in turn, is an important constituent of culture, especially of music. These resources of a Warring States lite culture are necessary tools for the cultivation of man. Educating the individual with these will influence him and finally shape his mind. Accordingly, the Xng z mng ch implicitly claims that the practice of a certain behaviour leads to the internalisation of its spirit. Similar to the habitus described by Aristotle in the Nicomachean Ethics, the steady influence of positive information that intrudes into the mind will necessarily lead to its internalisation. To have internalised the spirit of Odes, Documents, rites, and music (and education in general) thus marks the final stage in the cultivation of man.

60 Compare this notion with Kongz s renowned statement: At the age of fifteen, I set [my mind] upon learning; at the age of thirty, I took my stance; at the age of forty, I was no [longer] uncertain; at the age of fifty, I knew the heavenly decree; at the age of sixty, my ears were compliant; at the age of seventy, I could follow what my mind desires without transgressing the right proportions (Lnyu, 2:4). 61 Note that the stress here is on zh , to understand, which is an important concept also in the Wu xng. 62 If seeking the [right] mind has something artificial to it, one will fall short of obtaining it (strip x37/1318).

148

chapter four Structure and Thought

The Xng z mng ch is patterned by the recurrent use of the particle fn . It summarises general knowledge on the basis of which a new idea can be introduced. In this context, it should be translated as in sum, as a rule, or generally speaking. Each unit introduced by this particle can be regarded as one pericope.63 The texts contain twenty pericopes.64 Canto 1 (strips x135 of the Xng z mng ch; strips xq121 of the Xng qng ln) is made up of twelve pericopes (112). Only pericope 3 does not appear in the Shnghai Xng qng ln. But this loss has nothing to do with a structural instability of the text. It results from the physically poor condition of the Shnghai manuscript.65 In both manifestations, the sequence of pericopes 112 is stable.66 This stability even extends to the phraseology, which is highly consistent throughout canto 1. Slight differences between the two manuscripts remain only on the level of writing forms.67 In most cases,

In many cases, the length of a pericope in these texts equals that of a building block. 64 L Lng (1999, 505) splits up pericope 8 into two units (8 and 9), even though the particle fn indicates that this is one unit (he thus counts twenty-one units overall ). He later revokes this suggestion (see L Lng 2002, 106). The consistency of this unit is further corroborated by the Shnghai Xng qng ln (see strip xq8). It should be noted in this context that the two texts also contain some pericopes in which every single statement is preceded by the particle fn. These instances do not indicate individual pericopes but exclamations of special importance. The texts also contain some pericopes in which the new idea precedes the particle fn. I refer to these units below. 65 Following pericope 2 (on strip xq3 of the Shnghai Xng qng ln) is the latter part of pericope 4 (on strip xq4 in the Shnghai Xng qng ln). All in all some sixtysix graphs are missing from the Shnghai Xng qng ln: about forty-two graphs for pericope 3 and about twenty-four graphs for pericope 4. Given the average number of graphs with which the Shnghai strips are inscribed (between thirty-one and thirtyfour each), these absences can be explained by the loss of two entire bamboo strips in the Shnghai manuscript. The loss of the pericope thus results from the poor condition of the Shnghai manuscript, not from a corruption of the text itself. 66 Sentences missing in the Shnghai Xng qng ln (in pericopes 2, 4, 7, 8, 9, 10, 11, 12) do not reflect a corruption of the text but the preservation of the manuscript, that is, broken bamboo strips. 67 Examples for this are as follows. The Shnghai Xng qng ln at times writes shng (OC *sre) where the Gudin One Xng z mng ch writes xng (OC *se-s) (e.g., in pericope 1). At times, the Shnghai Xng qng ln adds the signific to graphs, not present in the Gudin One manuscript (e.g., characters jio and l in pericope 5). Also, the two manuscripts display structural differences in the writing of the character do and the Shnghai manuscript uses zhng (OC

63

human nature is brought forth by decree

149

however, the dissimilarities between the two manuscripts reflect what Martin Kern would term a stable phraseology in relatively unstable writing,68 with the exception of only a few words.69 Also, at times the character y is omitted in either of the two texts.70 Despite these differences, canto 1 of the Gudin One Xng z mng ch and the Shnghai Xng qng ln is a highly stable entity. Of this long unit, the latter part of pericope 12all in all, thirty charactersis missing from the Shnghai Xng qng ln. Following the last statement of that row, the Shnghai Xng qng ln continues with what is pericope 16 of the Gudin One manuscript. This loss cannot be explained by the poor preservation of the Shnghai manuscript, because the next unit of the Shnghai Xng qng ln continues on that very same stripstrip xq21and the tail of this strip is well preserved. The missing fragment of the Shnghai text equals two building blocks of the Gudin One text. In the Gudin One manuscript, many words (graphs) are given expression only through signs for repetition of the previous graph. This is a common feature of argumentative chains in excavated manuscripts. By implication, even though the text of these units is appreciably longer, only thirty characters are missing in the Shnghai manuscript. This might equal the length of one bamboo strip of another (third) text. Proponents of the theory that a written text featured as Vorlage for the Shnghai Xng qng ln might therefore explain this loss with a slip of the eye by the scribe, who might simply have forgotten to copy the entire last bamboo strip of canto 1 of that imagined source text. Or they might postulate a corruption in that Vorlage itself. Advocates of the theory that the text was transmitted orally, in contrast, have to explain this loss by postulating the instability of text transmission in oral environments. I take this up further below. The differences in the two manifestations of canto 1 can be summarised as follows. First, despite its overall (structural ) stability, some

*te-s) upright where the Gudin One manuscript has dng (OC *m-te[n]-s) to determine (or *m-te-s set forth) (pericope 1). 68 In his approach to the Odes, Kern (2005a) talks about a verbally highly coherent text which was as stable in its phraseology as it was unstable in its writing. 69 One example for this is the use of the character yu (OC *[g]w]k or *[g]w] k-s) in the Shnghai Xng qng ln where the Gudin One version has kui (OC *kwhre[t]-s; or maybe even OC *kwhra[t]-s) (in pericope 7). 70 For instance, in pericope 7 the Gudin One version reads ; the Shnghai manuscript most likely reads [].

150

chapter four

differences apply to the level of scribal idiosyncrasies. Most of these differences fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components.71 This can be explained. When producing a new copy of a written manuscript, the scribe worked from the sound he heard, not from the graph he saw. By analogy to early European manuscript production, it can be assumed reasonably safely that a scribe dictated the text to himself even when he had a written Vorlage at hand. As a result, when producing a new copy, the scribe would not write the graph he saw but would write the sound he heard. Differences in the lexicon may be explained accordingly. Just as in the European Middle Ages and antiquity, the process of reading in early China can be described as primarily oral reading.72 That means that the auditory component of the word had to be established so as to catch its meaning. Jean Leclercq (19111993) has called this hearing the voices of the pages.73 Second, short text passages are omitted here and there in the manuscripts. Not all of these correspond to the length of a bamboo strip of either of the two manuscripts. This shows that neither of the two manuscripts could have served as Vorlage for the other, and so it can be ruled out that the Shnghai Xng qng ln was copied from the Gudin One Xng z mng ch, and vice versa. By implication, canto 1 of the two manifestations must have been transmitted orallyand thus independently of each other. Or it was copied from an imagined third source text, which, however, should not be mistaken for the urtext or Urschrift of the Xng z mng ch. In any case, as canto 1 of the two manifestations shows, the two manuscripts must be con-

71 In general, the criteria for phonetic similarity for loan characters and phonetic components in Old Chinese are as follows. (1) The main vowel should be the same. (2) The coda should be the same. (3) Initials should have the same position of articulation (but not necessarily the same manner of articulation). (4) One may be A-type; one may be B-type. (5) One may have *-r- and the other not. (6) The tone category can be different (i.e., final * and final *-s can be ignored). These rules are sometimes relaxed, as evidenced, for example, by png zhuan phenomena, in which open and closed syllables are substituted for each other. 72 See, however, the discussion in Behr and Fhrer 2005. 73 See Leclercq 1961, 1819; also for the quotation oral reading. On the early scribal custom of dictating to oneself the text one saw and thus recording the sound heard rather than the graph seen, see also Illich 1991 (esp. ch. 4). Michael Clanchy (1979, 218) points to the same observation in his description of medieval writing. He mentions the eleventh-century clerk Eadmer of Canterbury, who thought of composing in writing as dictating to himself .

human nature is brought forth by decree

151

sidered individual (!) instantiations of writing down a highly coherent predecessor. Subsequent to canto 1, the striking analogy of the two texts breaks down, even though the second half of the texts still shows some stable text clusters. These are pericopes 1315 (i.e., pericopes 1921 in the Shnghai Xng qng ln) and pericopes 1720 (i.e., pericopes 1518 in the Shnghai Xng qng ln). The former cluster of pericopes can be considered a subcanto, connecting directly to canto 1 in the Gudin One manuscript. In the Shnghai Xng qng ln, this unit appears at the very end and so closes the discussion of human nature. Except for some minor dissimilarities, the phraseology of this subcanto proves largely stable in both manifestations of the text. In most cases, the graphs used differently fulfil the criteria for phonetic similarity of loan characters and phonetic components in Old Chinese.74 Only occasionally a missing character has to be explained in some other way than by blaming it on the poor preservation of the bamboo strips of the Shnghai Xng qng ln.75 But, compared to canto 1, this unit shows a stronger tendency towards idiosyncrasies, and these idiosyncrasies cannot be explained only on phonological
74 At times the signific heart is added to characters in the Shnghai Xng qng ln where it does not appear in the Gudin One text, and vice versa. See,

for instance, the character du ( ) (x39/11: pericope 13) of the Gudin One Xng z mng ch, to which the Shnghai Xng qng ln adds the signific heart at the bottom of the character; the character zhng , to which the Gudin One versioncontrary to the Shnghai Xng qng lnalways adds the heart signific. Instances such as these correspond to Imre Galamboss (2006) important observation of the relatively stable use of a phonophoric paired with the relatively unstable use of the signific in early Chinese writing. Also, the graph so (OC *[s]aw-s), read as zo (OC *[ts]aw-s) quick-tempered (pericope 14 of the Gudin One manuscript, strip x42/8), is written with the phonophoric cho (OC *[dz]raw) in the Shnghai Xng qng ln (strip xq35/10). The two graphs fulfil the criteria for phonetic similarity of loan characters and phonetic components in Old Chinese. In the same pericope, the Gudin One manuscript writes hun ( ) (OC *[g]ron-s) (x42/18) where the Shnghai Xng qng ln writes a graph consisting of the phonophoric juan (OC *[k]ro[n]) on top of the signific heart . 75 In the second building block of pericope 14 of the Gudin One text (pericope 20 of the Shnghai Xng qng ln) there appears such a dissimilarity, which does not result from broken strips. Whereas the Xng z mng ch reads people easily sacrifice themselves for it, the Shnghai Xng qng ln records only [] easily die for it, thus leaving out the word rn person. Omissions of this kind can easily be explained as transmission errors. Instead of closing the first building block of pericope 15 (pericope 21 of the Shnghai Xng qng ln) by using the particle y as seen in the Gudin One manuscript, the Shnghai Xng qng ln adds a big mark on the strip: .

152

chapter four

grounds.76 However, as concluded for canto 1 (pericopes 112), the present subcanto can still be considered a verbally coherent unit which was not yet totally stable in writing. Only one instance has to be pointed out where the two manuscripts display a grave dissimilarity. In pericope 13, a unit consisting of, again, thirty characters is missing from the Shnghai Xng qng ln. As above, this absence cannot be explained by referring to the physically bad condition of the bamboo strips of that manuscript.77 The passage might have been lost in the process of an oral textual transmission. It might have been simply forgotten. Or it might be explained by imagining a scribe who copied a thirdnow lostwritten source text (but certainly not the Gudin One text) when producing the Shnghai Xng qng ln. The imagined source text must have consisted of bamboo strips carrying about thirty characters each. This would also correspond to the observations described above. The scribe might have skipped one of these strips accidentally when copying his Vorlageor the source text itself was no longer complete. The last cluster of pericopes that appears partly in both manuscripts (pericope 17 to the first part of pericope 20 in Gudin One; 1518 in the Shnghai manuscript) is framed by pericopes 16 and 21 in the Gudin One manuscript. In the Shnghai manifestation, these two pericopes precede pericopes 1518. Even though this unit also represents a vocally largely stable text, it still contains written idiosyncrasies.
76 Examples that cannot be explained on phonological grounds include the word xu (OC *m-kruk) to learn on strip x36/2 in the Gudin One Xng z mng ch (pericope 13) where the Shnghai Xng qng ln writes jio () (OC *[k] raw(-s)) to teach. As for the graph shn (OC *[d]i[n]-s) carefully (pericope 15 of the Gudin One manuscript; pericope 21 of the Shnghai Xng qng ln), the Shnghai Xng qng ln writes a graph consisting of sh (OC *[g][i]p) and yn () (OC *hni[n]) benevolence in pericope 15 of the Gudin (*a[n]). Instead of One manuscript, the Shnghai manuscript writes l (OC *[r]a-s) to think; ponder (xq39/18) (pericope 21). Note that in this particular subcantoas in canto 1the two texts always write the word do differently. Instead of sh (OC *[g][i]p) ten (w38/5) in canto 2, the Shnghai Xng qng ln reads zh (OC *N-trek) (xq32/21). Also, there seems to be no phonological connection between the graph di (OX *l) in jeopardy as used in the Gudin One text (x45/21) and graph xq37/31 of the Shnghai Xng qng ln ( ) with the phonophoric jn (OC *[k]r) (unless *thek was its phonophoric at a time when laterals were already hardening). 77 Next to the first three graphs of pericope 13 (pericope 19 in the Shnghai Xng qng ln), a passage consisting of thirty characters is missing in the Shnghai manuscript. The unit, which also appears in the Gudin One Xng z mng ch, directly continues on the very same bamboo strip, thus ruling out the possibility that the loss reflects a missingor corruptedbamboo strip in the Shnghai Xng qng ln.

human nature is brought forth by decree

153

Again, most of these can be explained on phonological grounds.78 Yet it also contains an appreciable number of peculiarities which cannot be explained by postulating a phonologic similarity in Old Chinese. Lastly, some graphs of this unit are missing in one or the other of the two texts.79 Following strip x62/11 of the Gudin One Xng z mng ch, the Shnghai Xng qng ln continues with a passage that is located in a different spot in the Gudin One text, where it differs substantially. The stability of the two texts thus dissolves right in the middle of pericope 20. This is surprising inasmuch as the pericope tends to be a stable unit in argument-based texts. I come back to this below. It can be considered a general rule that the standardisation of texts is a rather late phenomenon. In ancient times, the transmission of a text over several generations easily led to changes in it. Even if the text transmits a coherent idea or a highly elaborate philosophical systemwhich would enhance the stability of the accountthe text nevertheless is not immune to variation. I have demonstrated this in my discussion of the Wu xng. The Xng z mng ch is no exception of this rule. By implication, changes in a given text do not necessarily imply changes in its philosophy. In detailed studies, Jack Goody80 and Rosalind Thomas,81 among many others, illustrate convincingly for the Western context how texts may change over several generationseven in the case of poetry. Martin Kern discusses the complex interaction between the oral and the written text in the Chinese context, confirming

78 For instance, in pericope 17 of the Gudin One manuscript (pericope 15 of the Shnghai Xng qng ln), the two texts write the character that is identified with ln (OC *(m-)r[n]-s) (x59/15; xq29/33), consisting of two mouth graphs on top of wn (OC *m[n]). Whereas the Gudin One version adds to it the signific heart at the bottom of the character, the Shnghai Xng qng ln adds the signific earth ; the Shnghai Xng qng ln writes the character cng (x59/19; xq30/3)

is composed of the to follow, consisting of on top of tu ; character x60/2 phonophoric yu (OC *[l]a) on top of the signific tu earth. The Shnghai Xng qng ln writes it as composed of the phonophoric yu (OC *[l]a) on top of the signific earth. For hun (*[g]ro[n]-s) (x62/3) in pericope 20 of the Gudin One manuscript, the Shnghai Xng qng ln writes a graph consisting of juan (*[k] ro[n]) with the signific heart beneath (xq31/22). All these instances fulfil the criteria for phonetic similarity of loan characters and phonetic components in Old Chinese. 79 Instead of d hi as in the Gudin One manuscript (pericope 19), the Shnghai Xng qng ln writes only hi . The Shnghai bamboo strip is unbroken here. 80 See Goody 1987. 81 See Thomas 1992.

154

chapter four

(especially for the Odes) that texts were not yet standardised in ancient times.82 John van Seters exemplifies the degree of variation of texts in the biblical context, pointing out that the standardisation of the Bible (and other early texts) is a rather late phenomenon. Thus, variation in early texts is the norm. Long sequences that are very similar (not to mention sequences that are entirely the same) are unusual. My analysis of the Xng z mng ch reflects this observation. Accordingly, it is not the fact that the two manifestations of the text differ to a certain degree that is puzzling. Instead, what is rather extraordinary and should receive more attention is that long sequences in the two texts that span pericopes are very similar. So the question is not so much Why do the texts differ? but rather, Is it possible to make out formal patterns that could account for the stability of reoccurring clusters in the different manuscripts? The Core Text: Canto 1 Canto 1 contains twelve pericopes. Pericopes 212 can be grouped into three subcantos, headed by the introductory pericope 1. Pericope 1: Introduction
1.
x1 [A] [B] x2[C]

[D] x3[E] [][F] [G] []x4[H] [I] [][J] x5 [K]

82

See Kern 2002.

human nature is brought forth by decree

x1

155

Generally speaking, even though all humans have a nature, they have no settled will in their minds. It requires the things; only thereafter can it rise. It requires delight; only thereafter can it evolve. It requires repetitive practice; only thereafter can it x2settle. Emotions such as rejoicing, anger, grief, and sadness are human nature. When they appear on the outside, then this is caused by the things [in the world]. Human nature is brought forth by decree; decree [in turn] x3is sent down from Heaven; The true ethical code (do) begins with the unshaped feelings; unshaped feelings [in turn] are begotten by human nature. In the beginning [the true ethical code (do)] approximates the unshaped feelings; in the end [it] approximates rightness. {He who} understands {the unshaped feelings may} x4manifest them [at the outside]; he who understands rightness may take it in. To love and to hate is human nature. That which one loves and hates are the things [in the world]. To be good or not { good is human nature}. x5 That which one considers good or not good is [determined by] conditional forces.

SubCanto 1
2. [L] [ x6] [ ] [N]
[M]

Generally speaking, that which becomes the dominating aspect of human nature is caused by the things [in the world]: Bronze and stone have a sound, [yet] {if they are not struck, they will not x6sing}. {This is just like with humans}. Even though {they} have a nature, if [their] minds fail to accept it, it will not be manifested. 3. [] []x7[O] []x8[P] Generally speaking, mind has a will[yet] if not nourished, it will not {be settled}. {That mind cannot}x7 evolve on its own is just like the inability of the mouth to speak on its own. After oxen are born they grow [big]; after geese are born they stretch [their necks].

156

chapter four

It is their nature {which makes them be like thisas for humans,} x8in contrast, it is learning that eventually shapes them. 4. [Q] [R] x9 [S] [T] Generally speaking, of the things, there is none that is not different. That the hard becomes straight is because hardness takes it up. That x9 the soft becomes flexible is because softness takes it up. That human nature is one within the four seas, [yet] in the application of the mind each [man] differs, is something brought forth by education.

SubCanto 2
5. x10[U] [V] Generally speaking, as for human nature x10there is something that moves it, something that conforms to it; Something that interacts with it, something that grinds it; Something that makes it manifest [at the outside], something that nourishes it; [And] something that makes it grow. 6. x11[W] [X] [Y] x12[Z] Generally speaking, that which moves human nature x11are the things [in the world]. That which conforms to human nature is delight. That which interacts with human nature are the causes. That which grinds human nature is rightness. That which makes human nature manifest [at the outside] are the conditional forces. That which nourishes x12human nature is repetitive practice. That which makes human nature grow is the true ethical code (do).

human nature is brought forth by decree

7. [Aa] [Ab] x13[Ac] x14[Ad] [Ae]

157

Generally speaking, that which can be seen is what we call the things [in the world]. That which generates satisfaction in oneself is what we call delight. That which is the power of x13 the things is what we call conditional forces. That which makes something happen is what we call causes. Rightnessthis is the status indicator for the [different] groups of goodness. Practicex14this exists for practicing ones human nature. The true ethical code (do)this is the Way (do) of the [different] groups of things.

SubCanto 3: Digression
8. x15 [Af ] [Ag] x16[Ah] [Ai] [Aj] x17 [Ak] [Al] x18 [Am] x19 [An] [Ao] ?x20 (?)
x21

[][Ap] [Aq] [Ar]

158

chapter four

[ X]x22[As] [At] [Au] [Av] x23 Generally speaking, as for the true ethical code (do), the skills of the mind are the dominant [features]. The true ethical code (do) diverges into four skills, [but] only x15 the Way of humans can be followed.83 As for the [other] three skills, one can only talk of them. Odes, Documents, rites, and musicin every case their first appearing was given birth by x16man. Odes came into being by acting them out. Documents came into being by speaking them. Rites and music came into being by exalting them. The sagacious persons [then] juxtaposed them (Odes?) according to x17 their categories, collated and joined them together. [They] beheld them (Documents?) in their temporal sequence to arrange them in their proper sequence. [They] shaped their (rites?) meaning to regulate and pattern them. [They] ordered x18 the feelings [expressed in them (music?)] to be manifested [at the outside] and to be internalised. Only when this was achieved, did they turn toward teaching. Teaching is that by which [the sagacious persons] generate moral force (d) in [their] centres. As for rites, they were created on the basis of unshaped feelingsx19[but] eventually, they [also] might stimulate them. Relying upon methods, according to each casethe proper sequence of what comes first and what comes last thus befitted the Way (do). Once the proper order of actions was x20regulated, they were culturally refined. To extend this to ones manner and appearance, that was the reason for cultural refinement and regulations. [Hence], the gentleman finds beauty in their unshaped feelings and values {their rightness}. [He] x21 considers their regulation as good and esteems their manners. [He] finds joy in their true ethical code (do) and delights in their teaching. Therefore, he shows respect for them.

83 Cf. this with a line in the received Laoz: The do that can be spoken of is not the constant do (Laoz jio sh, 3).

human nature is brought forth by decree

159

Bending [his hands] {is how respect for X is expressed}; x22 the repetition of it is [cultivated] pattern. Offering presents of silk is how trustworthiness and evidence are established. Declining them [should be done in a way] befitting the true ethical code (do). Laughter is the shallow march (= surface aspect) of rejoicing. x23 Music is the deep march (= the underlying aspect) of rejoicing. 9. [Aw] x24 [Ax] x25[Ay] <<>><<>> <<>><<>>x26 [Az] x27[Ba] [Bb] <<>><<>>[Bc] Generally speaking, all sounds emanating from unshaped feelings are trustworthy; only when they enter and agitate the heart of man, do they become profound.84 x24 [Therefore,] to hear the sound of laughter is preciouswhen it comes to it, then there will be rejoicing. To hear the sound of chanted songs is highly gratifyingwhen it comes to it, one will become elated. To listen to the sound of lute and zither is x25 excitingwhen it comes to it, one will have to sigh. [And likewise], to watch the [performance] of the [ritual dances] Li and Wu [makes one] solemnwhen it comes to it, then one will be stirred.85 To watch the [performance] of the [ritual dances] Sho and Xi is incitingwhen it comes to it, x26then one will become humble.86

The character hu thick is read as profound. The Li (Mo 295) and Wu (Mo 285) are part of the Zhu hymns that praise King Wu s victory over Shng. The odes were part of a larger performance that included ritual dances. Accordingly, the passage says watching. 86 The songs of Sho and Xi are the music of Shn and D Xi (or Yu ). They were both performed together with ritual dances. Following Lio Mngchn (2002a), the lyrics of the songs also state that Wu and Yu have their own deficiencies; accordingly, the character jian should be read as humble, modest, or selfdeprecating.
84 85

160

chapter four

When moving the mind by cantillating aloud a thoughtthe sighingsound wei will follow. It is long that it occupies the second position, [and] when by returning to the good and starting from the beginning one x27has to be sincere, [and] when by entering and coming out it is compliantthis is something which initiates ones moral force. The music of Zhng and Wi, in contrast, is not of this sound; but one [easily] indulges in it. 10.
x28 [Bd] <<>><<>><<>><<>>[Be]

Generally speaking, x28 music of old (= the music of Shn and Y) highly exalts ones mind; the latter music [of King Wu of Zhu] exalts ones ambitions. They are both for educating the people. [The latter] music [of King Wu ] Li and Wu is that of grasping [ambitions]; [the former] music [of Kings Yo and Shn] Sho and Xi is that of unshaped feelings.87 11. 88 [Bf] [Bg] x30: [Bh] x31 [Bi] Generally speaking, most refined music is necessarily sadcrying is also [an expression of being] sad. They both reach out to their corresponding feelings. The corresponding natures of grief and joy are close to each other. It is for this reason that their minds x30 do not deviate far: The way in which crying moves the heart is encroaching and shattering [but in] its blazing fierceness [it] is all-consuming, and one remains grievous till the end.
x29 x29

See the discussion of these two lines in Lio Mngchn 2001, 14950. Zh means the high point. Yu could be interpreted as either music or pleasure. However, since the following part clearly talks about waning and grief, yu at this place should be understood as pleasure, the counterpart to grief.
88

87

human nature is brought forth by decree

161

The way in which music moves the heart x31is profound and deep, dense and delightful[but in] its blazing fierceness [it] is like a steady flow that leads to grief, and one becomes mournful in thoughts. 12. x32 [Bj] [][Bk]
x33 [Bl] [][Bm] ()[Bn]

Generally speaking, to have thoughts of sorrow results in sadness there after. x32 Generally speaking, to have thoughts of joy results in delightedness thereafter.89 Generally speaking, the impact of ones thoughts on the mind is extreme. Sighing, this is a mode of [expressing] thoughts. When the sound of it changes, then {the mind will follow along as well}. x33 [And] when the mind changes, then the corresponding sound of it also [changes] accordingly: Sighing proceeds from grief. Chirping proceeds from joy. Murmurs proceed from the tones [of music]. Singing out loud proceeds from the mind.
[end of overlap of canto 1]
x34

[Bo] [Bp] [Bq]

89 In these lines, the particle fn does not introduce new pericopes, but it summarises ideas presented above.

162
x35
x34

chapter four

[Br]

When there is rejoicing, then there will be delight. When there is delight, then there will be enthusiasm. When there is enthusiasm, then there will be chanting. When there is chanting, then there will be waving [of ones hands]. When there is waving [of ones hands], then there will be dancing. [Hence], dancing is the end result of rejoicing. When there is exasperation, then there will be sorrow. When there is sorrow, then there will be grief. When there is grief, x35 then there will be sighing. When there is sighing, then there will be the beating of ones breast.90 When there is the beating of ones breast, then there will be jumping up and down [as an expression of ones grief ]. [Hence], jumping up and down is the end [result] of being exasperated.

Pericope 1 consists of four building blocks. The first of these announces the distinction between human nature (xng ) and the mind of the individual (xn ). The recipient is informed that mind needs the impact of the things [in the world] (w )in this context the overall denotation of the phenomenological world outside the individual delight ( yu ), as well as repetitive practice (x ), so that it can rise (zu ), evolve (xng ), and settle (dng ).91 The subsequent building blocks substantiate this concept of human nature in its relation to the phenomenological world (). The authors of the core text then continue the discussion in the pattern of a distant abc scheme. Each of pericopes 24 (subcanto 1) dwells on one of the three entities identified in pericope 192 to specify the relationship of human nature (; pericope 2) and mind (; pericope 3) with the phenomenological world (; pericope 4). It may be due to the distant abc scheme that this unit achieves a considerable

90 For the translation of p as beating ones breast, see the ode B zhu (Mo 26) in the quietude I brood over it, awake I knock and beat my breast (Karlgren 1950, 16). 91 See strips x12/1 (xq11/24 of the Shnghai Xng qng ln). 92 See strips x5/89/22 (xq3/174/15 of the Shnghai Xng qng ln).

human nature is brought forth by decree

163

stability, and it is reasonable to consider it as one subcanto. The subcanto not only discusses the relationship between human nature (), the mind (), and the phenomenological world () but also presents the different steps the human mind is believed to undergo so as to be determined as described initially in the first building block of the text (to rise, evolve, and finally settle).93 Pericope 2 describes the means by which the mind is inspired by the phenomenological world so that it manifests (ch ); this is perhaps equivalent to the notion of taking effect () in the introduction. Pericope 3 describes what the mind requires to be stirred to action (), hence reproducing the second position of the introduction. Lastly, pericope 4 describes how the quality of being hard (or soft) causes something else to be hard (or soft). This possibly functions as allegory to the third statement of the introduction. Subcanto 2 (pericopes 57) establishes text-intern crossreferences. These allow the detailed explanation of concepts introduced earlier on. I have detailed this formal device of establishing a consistent terminology in my discussion of the Wu xng. Briefly, pericope 694 explicates the concepts used in pericope 5.95 Pericope 796 further defines those from pericope 6. Accordingly, the subcanto takes on the pattern of a distant abc scheme, too, only that in this case no tripartite relationship has been established on the horizontal level as in the case above.97 In its place, a hierarchical relationship of type is highlighted. Again, as in the previous subcanto, the formal structure of this unit introduces a stable subcanto in the two manifestations of the treatise on human nature. Pericope 8 differs.98 It is not patterned in a similarly straightforward manner. But this unit is very compelling in its language. Elaborating the last concept discussed abovenamely the do it connects smoothly with the previous account. Pericope 8 discusses the educational curriculum of Warring States lite culture, namely Odes (sh ), Documents (sh ), rites (l ), and music ( yu ).
()()x2

93

()()()

();

()()();

(). Strips x11/2212/8 (xq5/46/10 of the Shnghai Xng qng ln). Strips x9/2310/21 (xq4/155/3 of the Shnghai Xng qng ln). 96 Strips x12/914/13 (xq6/117/break of the Shnghai Xng qng ln). 97 See the previous subcanto (strips x5/89/22; xq3/174/14 of the Shnghai Xng qng ln). 98 Strips x14/1422/18 (xq813 of the Shnghai Xng qng ln).
95 94

164

chapter four

Pericopes 91299 show some similarities to pericope 8. These units dwell on the impact that music and dance have on the individual. As in pericope 8, the style of these units facilitates easy memorisation. But again, it does not contain a straightforward structure. Only here and there are recurring parallelisms or argumentative chains. Pericopes 812 appear to be long digressions about lite cultures educational impact on the individual. They are composed in a mature, nearly prosaic style of writing. The Application of the Theory: Canto 2 Canto 2 contains two clusters of pericopes that remain stable in both manifestations of the text: pericopes 1315 (1921 in the Shnghai Xng qng ln) and pericopes 17 to the first part of 20 (or 1518 in the Shnghai Xng qng ln).
13. [Bs] ; x37 ; [Bt] x38[Bu] [] [Bv] [Bw] [Bx] x39 [By] [Bz] x40 [Ca] [Cb] x41[Cc]
x36

99

Strips x22/1934 (xq1421/12 of the Shnghai Xng qng ln).

human nature is brought forth by decree

x42[Cd]
x36

165

Generally speaking, whenever one learns, to seek for the [right] mind is difficult; by following what is brought about by it, one will already have come close to obtaining the [right] mind. But this is not as good as inviting [the right mind] with music. x37 Even if one were capable in ones tasks, if one were unable in ones mind, this was not to be honoured.

If seeking the [right] mind has something artificial to it, one will fall short of obtaining it. That humans cannot employ artificial means [for seeking the right mind] x38 is something that can be ascertained. Before having tried ten times, ones mind will necessarily be exposed through it. When investigating what can be seen of it, how could one possibly miss the unshaped feelings? To be fair is a mode of rightness. x39 Rightness is a mode of respect. Respect is the regulation of the things [in the world]. Genuineness is a mode of benevolence. Benevolence is a mode of human nature. Human nature is what eventually gives birth to it (i.e., benevolence). Fidelity (zhng) is a mode of trustworthiness (xn). Trustworthiness is a mode of unshaped feelings. Unshaped feelings originate from human nature. Of the categories of love there are seven[yet] only the love of human nature is close to benevolence. Of the categories of wisdom there are five[yet] only x41 the Way (do) of rightness is close to fidelity. Of the categories of hate there are three[yet] only the hate of the non humane is close to rightness. [Finally], what constitutes the true ethical code (do) is fourfold[yet] only the Way (do) of humansx42can be followed. 14. [Ce] [Cf] x43[Cg] [Ch] [Ci] x44[Cj] [Ck]

166

chapter four
[Cl] x45 [Cm] , [Cn] x46 100[Co] [Cp] x47 [Cq] [Cr] x48[Cs] Generally speaking, impatience in the application of mind will be accompanied by excessive thinking. Haste in the application of knowledge will be accompanied by excessive worries. Extremes in the application of unshaped feelings x43 will be accompanied by excessive grief and pleasure. Privileging the role of the body will be accompanied by excessive [seeking for] pleasure. The exhausting application of strength will be accompanied by excessive [seeking for] profit. The eyes affection for x44 female beauty, and the ears joy in sounds, these are the airs of pent-up delightednessman would easily die for them. If in ones comportment as a human being one is acting as very restrained x45 but does not in fact have a very simple mind, then this is [only] ornamentation. [And] if it happens that in the very simple comportment as a human being one does not in fact intend to make the perennial endangerment nonexistent, then this is boundless. As for a mans skilful x46 words and advantageous speeches, if one does not [also] have a very unornamented mind, then this will be ephemeral. As for a mans state of being delighted about getting along and being comfortable with himself, if this is not actually coupled to the x47 unshaped feelings of striving to create, then it will be delusion. If it happens that in ones comportment as a human being one acts as if satisfied, failing to be taken care of is impermissible. If it happens that in ones comportment as a human being one acts as if recondite, x48failing to be corrected will not suffice.

100

Li functions as a loan for f .

human nature is brought forth by decree

15. [Ct] [Cu] x49[Cv] [Cw] [Cx] [Cy]

167

Generally speaking, a humans artificial activities are hateful. When one is artificial, then one is bound to regret it. When one regrets, then one is bound to be cunning. When one is cunning, no one will have [friendly] x49 relations with you. Caution, [in contrast], is a mode of benevolenceyet if it is flawed, one will not be hated. Hastiness, [however], is a mode of contrivanceif it has flaws, one will be blamed. It is truly the case that if men are not cautious, flaws are bound to occur.
[strips x50 through x59/9][Cz]

17. [Da] x60 [Db] Generally speaking, in delighting others, one ought not to be pettyminded [towards them]as a person, one has to go along with them. When words reach [the things] (i.e., they are to the point), they x60are illuminating. Raise them and be without artificiality! 18. [Dc] Generally speaking, do not be fierce in relationsit will necessarily lead to triviality. 19. [Dd] x61101[De] 102[Df ] Generally speaking, when summoned, be without fear, [but also] do not raise a solitary voice. When x61 dwelling alone, then one should practice what father and elder brother have found pleasure in.

101 102

Reading x in the sense of to follow. Reading x61/10 w do not with w have no.

168

chapter four
If only there are no greater calamities, when minor irregularities occur, this can be endured. Once they are over, refrain from speaking about them again.

20.

x62

[Dg]

[end of overlap]
x62

Generally speaking, concerning affairs of sorrow and calamity, one should wish to tackle them; concerning affairs of joy, one should wish to postpone [them].

The substantial cluster of pericopes 1315 appears further along in the Shnghai Xng qng ln (1921).103 This stretched subcanto contains many declarative sentences ending in y , parallel enumerations, and argumentative chains, resulting in compact and straightforward building blocks. Summarising remarks round off many of these rigorously constructed units. Compared with the compact structure of the individual building blocks that constitute this subcanto, the cluster of pericopes itself contains no such rigorous structure. It is surprisingly loose. The final stable cluster of pericopespericopes 17 to the first part of 20 in the Gudin One Xng z mng ch; 1518 in the Shnghai Xng qng lnspans three bamboo strips (as against two and a half in the Shnghai Xng qng ln).104 The continuous use of the negative w do not is the only recurring element of the pericopes. The consistency of this unit in both manifestations dissolves right in the middle of pericope 20 of the Gudin One manuscript. This is surprising. Pericope 20 continues in the same mode as the passages above and should thus be, at least theoretically, a steady module. The missing part appears at another location in the Shnghai Xng qng ln. But it differs substantially from the one in the Gudin One manifestation of the text. Proponents of the assumption that the source texts of the Shnghai Xng qng ln existed only in oral form might want to explain the dissimilarity by the relative fluidity of texts in that period. But to blame the corruption of the passage in question on the instability of spoken texts in contemporary China is a non sequitur and thus not helpful at all, as it would explain virtually nothingand the same is true if one wanted to explain the dissimilarity by editorial choice. The break
103 104

Strips x36x49 (xq31/31xq40 of the Shnghai Xng qng ln). Strips x59/962/11 (xq29/631/30 of the Shnghai Xng qng ln).

human nature is brought forth by decree

169

appears in astructurallyconsistent (and thus theoretically stable) unit. My analysis has shown that units of this kind are steady modules. These were the units that were once remembered and so, especially in oral texts, were units of high stability. Even if early texts were prone to change in their incessant process of transmission, so that every manifestation of a text naturally differed to some degree from previous or subsequent ones, units of this kind would remain structurally stable entities. This calls for an explanation of the break at this point. In this context it is instructive to take a closer look at the materiality of texts of that period. The passage that is displaced has twenty-nine characters. This is similar to the length of the other gaps in the Shnghai Xng qng ln, which always number around thirty or, when doubled, sixty characters. Given that these gaps appear in theoretically consistent units and given also their uniform character, the differences between the two manuscripts can be explained by assuming that the Shnghai Xng qng ln was copied from a written Vorlage that was itself different from the Gudin One manuscript. This source text of the Shnghai Xng qng ln, I presume, was written on bamboo strips carrying circa thirty characters each. The imagined Vorlage was in disarray and the strips were rearrangedor even lostbefore the Shnghai Xng qng ln was copied from it. Conclusion The Xng z mng ch is in many respects an important text (and the same is true for the Shnghai Xng qng ln). It is the earliest surviving copy of a detailed analysis of human nature. Without proposing the concept of a blank slate, its authors devote much attention to the impact of the phenomenological world in shaping the mind of the individual. Mind is identified as open to adjustments. It accounts for the moral cultivation of manor his degeneration. The authors of the text postulate a common humanity. That individuals nevertheless realise different degrees of moral cultivation calls for a detailed analysis of an intermediate stage between human nature and human behaviour. If the individuals mind is educated appropriately during this intermediate stage, the individual will set his will on moral cultivation. The authors postulate a relationship between human unshaped feelings, nature, and the true ethical code, which lies within man himself. From this it follows that, when activating this innate ethical code

170

chapter four

by receiving stimulation from unshaped feelingswhich are part of his naturethe individual can allow a full expression of these feelings without fear of departing from rightness, which, again, is close to the do. It thus becomes obvious that the text defends the idea that the repetitive practice of a virtue necessarily leads to the internalisation of its spirit. This further assumes that every human being carries within himself the potential to become a real gentleman, if educated properly. Just as important as the fact that the text allows another hithertounknown insight into the public discussion that took place in the mid- to late Warring-States intellectual world is the mention of Odes, Documents, rites, and music. The mention of these constituents of Warring States lite culture as one group is frequently seen as evidence for the assumption that they were already clearly distinguishable and identifiablewrittenentities during the time when the Xng z mng ch and the Xng qng ln were fixed on bamboo. I doubt this. Nowhere in the two manuscripts is there any explicit mention of these as written entities, let alone fixed texts (or books). Instead, they are referred to only as traditions, that is, something being performed:
x16 Odes, documents, rites, and musicin every case their first appearing was given birth by x16man. Odes came into being by acting them out. Documents came into being by speaking them. Rites and music came into being by exalting them.

None of these references describe something textual. Instead, they all have an oral, behavioural colouring. In this context, it is also illustrative to look at the preceding unit:
x15 Generally speaking, as for the true ethical code (do), the skills of the mind are its dominant [features]. The true ethical code (do) diverges into four skills, [but] only x15 the Way of humans can be do-ed. As for the [other] three skills, one can only do them.

human nature is brought forth by decree

171

This passage is anything but unambiguous, which is why I leave the latter two mentions of do without a translation here. It nevertheless becomes clear from this passage that the four constituents of a Warring States lite culture are referred to as ways of humans. As such, they are something that can be followed and, accordingly, practised. Subsequent to these passages is a description of the impact of these constituents of culture on human beings. The two manifestations of the Xng z mng ch speak of sounds, ritual patterns, dancesbut nowhere of texts. This does not, of course, rule out that the authors also had texts in mind when composing this passage; yet from this passage it seems rather unlikely that they did. I want to close my discussion by referring to the history of the two manifestations of this text on moral cultivation. The Shnghai Xng qng ln was bought from tomb looters at an antique market in Hong Kong. This violently cut the manuscript off from all references to its previous environment, leaving much room for speculation about its origin and its relationship to the Gudin One Xng z mng ch.105 Due to the striking similarity of the two manifestations, they are often considered as one text. It is continually stressed that the two manuscripts must have originated in close proximity to each otherchronologically and spatiallyand that they may even have been copied one from the other. The last assumption can be refuted. Despite the lack of a nontextual referential framework that would help to contextualise the Shnghai Xng qng ln, the history of this manuscript can still be reconstructed, to some extent at least. As detailed, the two manifestations of the text by and large share a highly consistent phraseology. Nevertheless, they cannot be considered stable in writing. Quite the contrary: in many instances, the manuscripts use different forms for presumably the same graph, here and there a signific is either added to the character or is left out entirely, or different graphs or words are chosen. Many of these differences conform to the criteria for phonetic similarity of loan characters and phonetic components in Old Chinese. Applying Martin Kerns

105 Cf. Kerns remark that manuscripts are integrated in a non-textual environment (2002, 145). This contextualisation is inevitably lost for all texts that are now part of the Shnghai corpus. It is therefore essential for the academic world to reassess the means by which it procures primary research materials in acknowledgement of how it promotes illegal trade in manuscripts. See also Renfrew (2000) for a discussion of this problem.

172

chapter four

approach to Odes in excavated manuscripts, this might suggest a verbally coherent text in fairly unstable writing. For the two instantiations of the text under review this would point to an orally based transmission. Despite Kerns findings, for philosophical writings dissimilarities of this kind do not exclude the possibility that either of the two manuscripts might have been copied directly from the other. When copying a text from a written Vorlage, the scribe in question did not work from the graphs he saw, but from the sound he heard, as he was dictating the text to himself. Silent reading was not yet common, and I hold that the same is true for silent copying. Thus, even though I fully agree with Martin Kerns findings concerning verbally consistent Odes and the orally based transmission of this anthology, his results cannot be applied directly to philosophical texts from that period because of structural differences between the two genres. Odes were often composed in an archaic language. Every single act of transmitting an ode also implies its translation at the same time.106 These texts were discontinuous since every single act of transmission was also an act of reinvention. This explains the kind of variation seen in the different quotations from the Odes in philosophical texts. The broad diffusion of Odes, then, further strengthens Kerns findings on the oral nature of this early anthology. Odes survived fairly consistently in quotations in all kinds of different texts as far as phonology goes. But no single manuscript of this anthology has been found to date. Philosophical texts, in turn, render the converse picture. They are composed in a more contemporaneous tongue. Artless speech was a conscious choice. These texts define themselves by their continuity with the present. Furthermore, the transmission of philosophical texts was predominantly accomplished on a written basis, but the diffusion of these textsin most cases this means the manuscriptswas fairly limited. Further, quotations of philosophical texts show greater variation. Applied to the philosophical texts from the Warring States, the phenomenon of a verbally coherent text in unstable writing thus does not necessarily point to its oral transmission. The fact, then, that the two manifestations under review at times use entirely different graphs which do not fulfil the criteria for phonetic similarity of loan characters and phonetic components in Old Chinese

106 One may think of the common use of reduplicates in the Odes, which prove to be particularly prone to change, as shown in Kerns study.

human nature is brought forth by decree

173

argues against the assumption that they could have been copied one from the other. In contrast, it can be assumed that each version was copied from a different Vorlage. This assumption is strengthened by the fact that even though the two manuscripts possess markings at the same junctures, the type of these markings varies. A closer scrutiny of the differences between the two manifestations of the text allows inferences on their imagined Vorlagen. Again, this should not be mistaken for the urtext or Urschrift of the texts at hand. As detailed, the gaps in the Shnghai Xng qng ln prove not to be systematic. None of these gaps allows the conclusion that they reflect passages that were left out on purpose by the authors of the Shnghai Xng qng ln. Instead, the form analysis developed in this book suggests for the Xng z mng ch that these gaps appear within otherwise cohesive and structurally stable units, thus ruling out the systematic omission of a text passage. Looking at the missing passages that do not result from the natural loss of broken bamboo strips in the Shnghai Xng qng ln, it becomes clear that the sequences in question always number around thirty characters or, doubled, sixty characters. The same holds true also for the sequences that reappear at different junctures in the manuscript. All of this is further evidence that the Shnghai Xng qng ln was copied from a writtenVorlage fixed on bamboo strips that were inscribed with about thirty characters each. The hypothetical Vorlage was already in disarray when it served as the source text for the Shnghai Xng qng ln. The Gudin One Xng z mng ch did not draw from the same Vorlage, at least not at the time when that Vorlage was in disarray. Figure 21 presents possible models of the history of the two manifestations. The arrows indicate their relationship with the imagined Vorlage. As argued, it is highly unlikely that the extant manuscripts both stem from the same Vorlage (V). Model 1 accounts for this: the stemmata describe two individual lines of text history. That no direct contact of the two manuscripts can be reconstructed suggests a paralleland separatehistory of these manuscripts.107 But whereas it is possible

107 Martin West would call these different recensions. He distinguishes between closed recensions and open recensions. According to his definition, recensions are linked in stemmata that can be traced back to archetypes (which does not need to mean an urtext). See West 1973, 14. For a critique of his terminology, see the introduction.

174
Model 1

chapter four
Model 2 V2 V1 ? V1 Model 3

V1

SH

GD

SH

GD

SH

GD

Figure 21: Three Models of Text Development

to make an informed guess as to the nature of the direct Vorlage of the Shnghai Xng qng ln (V1), this is impossible for the Gudin One Xng z mng ch. There can, of course, be no question that the two manifestations of the text had some sort of shared source that accounts for their similarity. Model 2 accounts for this. Whereas it is relatively clear that the two manuscripts did not result from the samedirectVorlage (V1), it can be assumed nevertheless that the imagined Vorlage (V1) must have had some common ancestor (V2) with the Gudin One Xng z mng ch (or its predecessors). The hypothesised shared ancestral text (V2) might have been either written or oral. This cannot be reconstructed. Lastly, it is also possible that the Gudin One Xng z mng ch and the Shnghai Xng qng ln shared the same Vorlage (V1) but the Gudin One manuscript drew on that source at a much earlier date. This would imply that the Shnghai Xng qng ln was copied from that source text at a stage when the imagined Vorlage was already in disarray. It is clear that these three models oversimplify historical reality. A text is always a synchronic artefact. It may thus incorporate different diachronic text layers. These intermediate stages may account for an asynchrony of textual transmission. None of these idealised models can account for this. In contrast, they should be understood as attempts to order the complexity of reality by highlighting certain characteristics of a given phenomenon.

PART II

TEXT AND PHILOSOPHY

CHAPTER FIVE

TEXT, STRUCTURE, MEANING In Chapters 14 I analysed the way in which argument-based texts are made up of stable and distinct units, or building blocks. In most cases, they contain a decidedly formulaic language. The different building blocks are linked up with one another by a system of cross-referential links, so that they constitute larger entities. Various notions are thereby connected into greater schemesand finally into a coherent whole. The individual building block of argument-based texts is therefore not an isolated unit but a feature of a larger, integrated entity, through which the texts become truly systematic treatments of philosophical concerns.1 The formal structure of argument-based texts that integrates the various building blocks into larger consistent wholes thus is a vital element in generating meaning beyond the level of the lexicon and syntax. It fulfils a semiotic function in that it opens up a meaningful level beyond the verbatim content of the text and so proves philosophically relevant as a mode of meaning creation. Conceptual definitions are established, and fully fledged philosophical reflections on a topic are made possible. As a result, argument-based texts facilitate systematic treatments of philosophical concerns, which reduces the degree of contextualisation necessary to grasp their ideas. Obviously, there is no unified argumentative pattern that is used to present an organised discussion of a philosophical concern and that applies consistently to all argument-based texts. However, some patterns do recur: for example, overlapping structure, double-directed parallelism of a text segment, principal insertion, micro-macrostructure correspondences, and systematic referenceseither direct or indirect to the intellectual environment in which these texts were produced.

1 Whilst agreeing with William Boltz that building blocks are characteristic of early Chinese texts, my analysis suggests that we should not conclude from this that the feature of the building block results in a composite nature of these texts that opposes integral, structurally homogeneous texts (see Boltz 2005, 7071).

178

chapter five

That argument-based texts develop organised reflections on philosophical issues has important implications when evaluating the written philosophy of the period. On the level of macrostructure, these texts present methodically consistent treatments of certain philosophical concerns. This suggests that the concept of a philosophical text as a coherent and self-contained unit was already well established in those circles that were participating in the philosophical discourse of the period and that were producing this type of philosophical text. This will be made explicit below. In what follows, I shall place the argument-based texts from Gudin One into the broader context of the intellectual activities of Warring States lite circles. To highlight their distinctiveness, I contrast the techniques of meaning construction advanced in these texts with those of texts that have predominantly non-argument-based features. I shall further generate a descriptive definition for the non-argument-based texts and discuss the issues of authorship and writing in the intellectual world of the Warring States. This will ultimately lead me to explore different ways of philosophising in early China, as well as to examine the groups that were working with the different philosophical texts. I refer to these groups as the textual communities of the Warring States period.2 The Deliberate Production of Texts As discussed in chapter 1, the Zhng xn zh do is composed of strictly organised building blocks. They contain a recurring scheme of ab-ab-c correspondences. In this scheme, the second ab group further specifies the contents of the first ab group. At the same time, it continues the argument presented and so forms an overlapping structure. The information in the building block is concluded in its final c component. Strikingly, this pattern is also realised on the level of the macrostructure of the text. By relating the various building blocks to one another according to that scheme, the macrostructure of the text is able to specify the conceptual dimensions of the rulerin this context called gentleman ( jnz )as pictured by the authors

2 On the phenomenon of communities that coalesce around particular texts and so constitute textual communities, see the discussion by Stock (1983) of textual communities in medieval England.

text, structure, meaning

179

of the text. The formal structure makes clear that only by realising his immanent nature as specified in the text, namely by behaving with fidelity (zhng ) and trustworthiness (xn ), will the ruler become a human reflection of the cosmic elements Heaven (tin ) and Earth (d ); and this should be the goal of every ruler. The penultimate building block of the text (building block 5) breaks away from this uniform pattern. This unit spells out the rules that define a norm of behaviour, to be fulfilled by the ideal(ised) ruler. The interruption of the mantra-like rhythm of the text at this juncture is a device for stating the argument of the text. It transforms the abstract account into a concrete request addressed to the ruler of a state, stating that if the ruler accepts the position as presented in the text as true, he cannot but follow the request as articulated. Building block 5 thus marks the culmination of the argument. It contains all the argumentative elements of the entire text. This must be specified in the context of the argumentative purpose of the text overall. As the analysis of the textual materials from the tomb corpus of Gudin One suggests, this break in the pattern is a recurring feature in argument-based texts. It spells out the overall concern of the text. I have defined it as principal insertion. Since, in this case, the principal insertion applies to the texts macrostructure, it is a distanced type of this argumentative feature. The brief review of the argumentative pattern of the Zhng xn zh do calls to mind the highly systematic way in which the text is organised. The philosophical position is established with great care, and it is by means of its formal structure that it achieves compelling argumentative force. The Zhng xn zh do can only exist in this structuralform. This is not a trivial remark. The fixed macrostructure that organises the argument of the text as a whole demonstrates that the very concept of text as coherent composition3 is already fully realised in a text of this type. It makes the Zhng xn zh do a self-contained unit. The highly balanced, coherent, and structurally closed nature of the texts composition suggests deliberate production by proactive authors. These authors are characterised by their control

3 For an interesting comparative perspective on this, revolving around the idea of text as texture in the Greco-Roman world, see Scheid and Svenbro 1994. I thank Wolfgang Behr for alerting me to this publication.

180

chapter five

over the materials they use.4 Intentionally, they craft a highly balanced, coherent, and, on the level of its compositional structure, ultimately closed text. I come back to this below. In its degree of systematic organisation, the Zhng xn zh do proves to be unexceptional. As the comparison with other texts from Gudin One shows beyond doubt, a high degree of systematisation is characteristic of argument-based texts. The Qing d y sh, discussed in chapter 2, may serve as an example. It deals with the tension between Heaven and Man. The formal structure of the text accentuates this dichotomy. At every level of composition, the tension described semantically is also made visible through the texts formal structure. By embedding its concern in a formal structure that also mirrors the thought of the text itself, the Qing d y sh creates a persuasive environment for its uncomfortable concernif it does not indeed achieve a mimesis of the philosophical idea on the level of its formal structure!5 As with the Zhng xn zh do, the Qing d y sh as a whole takes on the formal structure of one individual building block. The macrostructure of this unit therefore fulfils the same integrating function as that of the Zhng xn zh do and so compounds the philosophy of these texts. The ways in which the Qing d y sh processes two individual argumentative linesI have described them as lineal and hierarchicalat first sight may seem to be contradictory. However, a detailed study of the argumentative features of the text shows that the tension created is in fact an integral part of the texts strategy to emphasise its concern and to formulate a request addressed to the lite of that time. The formal structure of the text is highly balanced. Just as for the Zhng xn zh do, any additions to or omissions from the text on the structural levelor even any relocation of the texts building blocks would disrupt the organisational coherence. Again, what results is a highly crafted text that works only in its given form. Argumentative lines are designed with great care over the length of the entire text, leading to well-balancedand structurally closedunits: a further

4 Similar remarks were made by Wagner (1980, 37) concerning the Wng B Laoz. Based on the structure of the text, Wagner also postulates that it was deliberately authored. Wagners conclusion was criticised harshly by LaFargue (1994, 308309, 596, n. 35). 5 See also Schaberg (2001, 51ff.) on literary patterning that is conceived of as an image of the order that cultural achievements would bring about.

text, structure, meaning

181

piece of evidence that at this time authors consciously composed philosophical texts as coherent units. Authorial self-consciousness and the intention to develop proper formal structures so as to process argumentative lines with compelling force can further be recognised when we look at those materials which stem from third sources but have been integrated into the argument-based texts under review. The Qing d y sh incorporates stories and myths taken from a pool of shared cultural memory into its own argument. The stories are tailored in alternating parallel style so that they fit the overall tone of the Qing d y sh. The parallel form in which these stories were cast stresses the basic principle that underlies these six examples in an identical way, and the formal perfection of the account adds to the credibility of the stories themselves. Accordingly, the strict form of this passage accentuates the common truth underlying these stories. The coherent structure furthermore accounts for the fact that the materials assembled turn into a stable subcanto and can therefore function as a medium which presents a unified notion for processing the argument of the text overall. Not only can evidence for the systematic composition of a philosophical argument be seen in the rather brief argument-based texts Zhng xn zh do and Qing d y sh, but a high degree of authorial self-consciousness can also be recognised in the two long and highly complex argument-based texts Wu xng and Xng z mng ch, which I discussed in chapters 3 and 4. These two texts are often considered to be highly problematic. The Wu xng targets the ruler of a state. It establishes a programme of self-cultivation according to which the imagined ruler will become like King Wn. The Xng z mng ch presents a multilayered analysis of the relationship between human nature and the phenomenological world around the subject. Without postulating the notion of the mind as a blank slate, much attention is paid to the impact of the outside world, that is, the phenomenological world as a whole and society at large, in shaping the human mind. The authors of the Wu xng established a complex systematisation of ideas by constructing highly convoluted cross-referential links throughout the text. These links span various layers, which, at times, make it highly challenging to follow the train of thought. But the process of self-cultivation as described in the Wu xng has the same underlying pattern as the formal structure of the text. Both take on a circular form. Just as the inner cultivation of one particular virtue

182

chapter five

is shown to depend on the cultivation of the other ones, the various subcantos of the Wu xng make full sense only with reference to the other units of the text. As seen from the other argument-based texts of this tomb corpus, form and thought are brought into perfect harmony. The Xng z mng ch also establishes a hierarchical system of referential links throughout the text. To pattern its argument and attain stability for processing the different trains of thought, a multilayered analysis of mind, human nature, and the world around the individual self is carried out by making use of various devices on the formal level of the text. Parallel a-b-c schemes, realised on different levels of the text and spanning various building blocks, authority-framed units, and other formal patterns, allow the vertical organisation of an argument as opposed to only linear patterns. As with the Wu xng, some parts of the formal compositional structure of the Xng z mng ch even reproduce the pattern underlying the formation of the mind (xn ), as imagined by the authors of this text. Just as seen in the Wu xng, the Xng z mng ch describes a circular movement underlying the process of moral cultivation. The true ethical code, as I translate the concept of do in this context, always lies within the individual self. As the authors of the text see it, the individual simply has to become aware of the moral force (d ), which already is an integral element of the individual, to achieve moral cultivation. The Wu xng establishes a highly layered system of ideas. The five virtues which the text discusses are split up twice into two groups. A horizontal and a vertical differentiation of the five virtues is thereby established. The first distinction of the five virtues is that between those virtues that can only be aspired to (zh ) and those that can be acted upon (wi ). The zh group of virtues concern the inner cultivation of the heavenly Way (tin do ). The wi group of virtues concern the transformation of this inner state into the Way of humans (rn do ). The second differentiation distinguishes between primary and secondary virtues. Primary virtues are sagacity (shng ) and wisdom (zh ). They account for the formation of the secondary virtues. The wu xng theory establishes a strict hierarchy of the five virtues. Of the primary virtues, sagacity is considered the foremost one. Second to the primary virtues are benevolence and rightness because they account for the formation of ritual propriety (l ). Rightness belongs to the zh group of virtues, which ranks higher than the wi group. Benevolence belongs to the latter group but is corre-

text, structure, meaning

183

lated with the primary virtues of the wu xng theory. It seems that the authors of the text saw these two virtues as equally important. Accordingly, the wu xng theory presents the following hierarchy of virtues: sagacity, wisdom, rightness/benevolence, and, lastly, ritual propriety. By relating various trains of thought in a highly complex fashion, a coherent theory of self-cultivation is established in the text. Sagacity takes the lead in the formation of rightness. If paired with the right cognitive grasp, this ultimately leads to moral force (d ). Wisdom takes the lead in the formation of benevolence and ritual propriety. If paired with the right cognitive grasp, this ultimately leads to goodness (shn ). Moral force describes an abstract matter. It equals mans inner cultivation of the Way of Heaven (tin do ). Goodness, for its part, results from applying this inner state to ones actions in worldly affairs. Thus, what the text labels as the Way of humans (rn do ) describes the conversion of mans moral force into action in the worldly sphere. The Wu xng thus presents a multilayered, coherent, and meaningful system, the wu xng theory (and so does the Xng z mng ch in its analysis of human nature). Different argumentative trains of thought are processed in parallel fashion vis--vis one another. At times, these lines are linked together, so that the text establishes a coherent whole in which all elements find their proper place and thus construct a meaningful system. That the two manifestations of the Wu xngthe Mawngdu Three Wu xng and that from Gudin Oneare organised differently does not contradict the conclusion that the two texts present one coherent and highly developed wu xng theory. However, it does say something about the nature of texts in the Warring States period. Texts were subject to change even when presenting elaborate systems of thought. But the larger meaningful units remain consistent in both texts, and the argumentative lines remain fully coherent. The wu xng theory remains unaltered. The overall structure of the Wu xngand of the Xng z mng ch (see below)without a doubt is quite different from that seen in the shorter argument-based texts Zhng xn zh do or Qing d y sh. The Wu xng cannot be considered a closed system. It is not immune to change on the structural level of composition. But the present wu xng theory (and the concern of the Xng z mng ch) is nonetheless a highly reasoned system. Without ruling out that a text such as the Wu xng may have evolved from the accumulation of different sources and traditions, which bears witness to the hybrid form of texts at that time (something that can be argued to apply to all

184

chapter five

argument-based texts from Gudin One), I hold that the wu xng theory, which we see from both the Mawngdu Three and the Gudin One Wu xng, is a conscious, reasoned, and well-designed approach to a prevalent concern in early China. The texts discussed all carefully establish different types of convoluted trains of thoughts. Argumentative lines can thus be processed on different levels. These texts not only present linear types of argument and thought processing but also take up philosophical concerns that were widespread and hotly debated in lite circles of the time and that they develop into systematic and reasoned systems of thought. Important for the present discussion is that in the Chinese world of those days, the notion of coherent texts and highly balanced argumentative lines was already well established. It is against this background that I postulate authors behind consciously crafted texts and theories. Mastering their arts, these authors were intentionally composing wellbalanced and, structurally, even closed texts. Philosophy has found its way into self-contained writing. Non-argument-based Texts of Warring States China The building block forms the basic constituent of the argument-based text. The authors of these texts develop a complex discussion of a given philosophical topic by combining the different building blocks into larger meaningful entities of the pericope, the subcanto, canto, and finally the text overall. Hence, the building block of the argumentbased text is not an isolated unit but literally a building block of a larger whole. Non-argument-based texts differ in this respect. The idea to be transmitted does not reach beyond the level of the individual building block. Different building blocks are not related to one another on the formal level of the text. Taken together, they do not generate greater meaningful wholes. Formally, the building block of non-argumentbased texts stands alone and is the one and only unit for communicating a concern. It represents what Rudolf G. Wagner has called a unit of thought.6 As this unit forms the only and ultimate level of

6 See Wagner 1999a. Note that Wagners concept is problematic insofar as it lacks a definition of what is actually meant by thought. I use unit of thought in this context to denote a textual unit that puts forward one isolated concern.

text, structure, meaning

185

textual communication (for the moment I am not taking into account the face-to-face communication within a group that may have existed with respect to the written text but am simply concentrating on the written object unearthed), it also spells out the entire philosophy of the text. Every new unit reflects a different concern and must be conceived as an individual text in its own right. By implication, it seems that, for instance, what we call the Laoz and the Z y today are each composed of different discontinuous texts, resulting in florilegia ofat least formallyunrelated units. The denotation non-argument-based text thus does not indicate an integrated, coherent whole. Calling the ultimate units of textual communication in non-argumentbased texts building blocks is therefore misleading. Remaining isolated and, as will be shown, responsive, they do not build or contribute to larger and coherent wholes on the formal level of such anthologies. Because of the conceptual difference between textual units in argument-based and in non-argument-based texts, I refer to the building block in non-argument-based texts temporarily simply as unit (or, with Wagner, unit of thought). Due to the isolatedand therefore ultimatecharacter of the unit of thought in non-argumentbased texts, systematic and highly layered treatments of ideas of the kind developed in argument-based texts are absent in these anthologies. The units in non-argument-based texts provide only isolated, and therefore situational, statements about or responses to given concerns. It is reasonable to describe these isolated pronouncements as occasional reactions to prevalent issues. In the absence of argumentative force, meaning had to be constructed differently, and I shall describe those strategies in what follows. Laoz and Z y The florilegia of ideas that have been identifiedor at least conventionally labelled soas Laoz and Z y are non-argument-based texts par excellence. They are made up of distinct units of thought, which, in both anthologies, show no apparent relation to one another on the formal level of composition.7 Rather than serving as a building block of a larger coherent whole that facilitates a systematic analysis of

Pace Spirin (1976), who set out to demonstrate the opposite.

186

chapter five

a given philosophical concern, each unit provides only a short statement about a given situation. The two anthologies belong to the canon of transmitted texts. This makes them rather exceptional in the context of the corpus of texts under review. The other texts from Gudin One at some point fell out of the transmission process. Argument-based texts such as the Wu xng, Xng z mng ch, and Zhng xn zh do survived only in the confined space of a tomb. The Laoz , in contrast, survived as the Classic of the Way and Virtue, or Do d jng . The origin of this anthology is anything but clear. Charming legends, as LaFargue calls them so tellingly, ascribe the work as a wholethe so-called Classic of Five Thousand Words (Wu qin z jng )to the legendary figure of Laoz .8 The Z y has survived as one chapter of the Record of Rites (L j ), a ritualists compilation of controversial provenance.9 The Z y is generally thought to have originated in the circle around Z S ,10 and so it is often believed to belong to the otherwise-lost work of the so-called Z Sz , the nature of which is almost completely unclear.11 The two anthologies Laoz and Z y have been handed down to the present day. Moreover, besides the Gudin One versions, we have a Laoz (organised very differently from its transmitted counterpart) from the tomb Mawngdu Three and a Z y among the Shnghai collection of Chu manuscripts. These finds provide invaluable insights into non-argument-based texts during the Warring States period and

8 LaFargue 1994, 301. For a detailed description of some of the traditional views on the historical figure Laoz and the Do d jng, see Seidel 1969. See also Chan Wingtsit 1963. It is doubtful whether the text ever contained five thousand words. 9 The compilation of the L j most likely was completed only in the Western Hn (206 BCAD 23) by specialists at the court of the Hn, but it seems that some parts of this anthology of independent texts (Shaughnessy 2006, 51) closely reflect ideas of ritualists from the pre-Hn (Nylan 2001a, 175). 10 Z Sz , named Kong J (ca. 483402 BC), was the grandson of Kongz . 11 Most explicit about the identification of the Z y with the work of Z Sz is probably the Ynyu zh of the Su sh , which quotes the words of Shn Yu (441513), saying: The Zhng yng, Biao j, Fng j, Z y were all taken from the master Z S . Cited in L Xuqn 1999e, 76. Since the excavation of the strips from Gudin One, Chinese scholars especially have attempted to reconstruct the work of Z Sz. Publications on this issue are too numerous to be listed here. For a critical discussion of the Z Sz myth, see Csikszentmihalyi 2004.

text, structure, meaning

187

help us to distinguish the special features of argument-based texts by comparison. No attempt to establish a watertight account of the nature of text and writing in the Warring States would have been possible without these finds. Not surprisingly, none of the excavatedor otherwise illegally procuredtexts is identical to its supposed counterpart. Differences are found in phraseology, the individual units of thought, the formulae used when drawing on other sources, and the sequence of the units of thought in the anthology overall; and certain materials are absent in one or other of the manifestations. The Laoz from Gudin One What is generally referred to as the Laoz from Gudin One must in fact be described as unrelated materials found on three different bundles of bamboo strips. Consensus refers to these as Laoz A ( ), Laoz B (), and Laoz C (). The three bundles differ in the physical appearance of the strips and in the handwriting with which the various strips are inscribed. Because later copies of the Laoz beginning with Mawngdu Threecontain materials which closely match those seen on the three bundles from Gudin One, the textual materials from Gudin One are customarily viewed as an earlier version of the transmitted Laoz. Hence, the materials on the three bundles of Gudin One bamboo strips are habitually considered to belong together. However, the tomb and its textual contents do not substantiate such claims. Bundles A, B, and C Bundle A contains thirty-nine strips. Unbroken strips have a length of ca. 32.3 centimetres and are tapered towards both ends. Most are well preserved; only one is in a fragmentary condition. Judging from the marks on the strips, two cords, about 13 centimetres apart, previously connected them.12 Bundle B consists of eighteen strips. Unlike bundle A, the bamboo strips are cut evenly on both sides. With a length of about 30.6 centimetres, these strips are slightly shorter than those of bundle A. As

12

Hbi shng Jngmn sh bwguan 1998, 111.

188

chapter five

in bundle A, two cords, 13 centimetres apart, connected the strips.13 However, of the eighteen strips that constitute bundle B, only eight are complete. The bamboo strips of bundle C are remarkably shorter than those of the other two bundles: they measure only ca. 26.5 centimetres. The strips of bundle C are cut evenly on both sides. Two cords, 10.8 centimetres apart, were used to connect the strips. Bundle C contains twenty-eight bamboo strips, twelve of which are complete. Only fourteen strips can be identified clearly as belonging to the so-called Gudin Laoz . The remaining fourteen strips carry previously unknown materials. The editors of the Hbi Province Museum assume that these materials constitute yet another individual text. Originally carrying no title, these materials are now customarily referred to as The Ultimate One Generates Water, or Ti y shng shu .14 The Ti y shng shu describes the continuous process of how the world is created, starting with the Ultimate One, ti y ,15 and reflects on the implications of this process for moral government. Bundle A contains twenty units of thought. These can be identified with the following zhng of the transmitted Laoz: 19, 66, 46 (middle and end), 30 (beginning and end), 15, 64 (end), 37, 63, 2, 32, 25, 5 (middle), 16 (beginning), 56, 57, 55, 44, 40, and 9in this order. Bundle B contains eight individual units of thought. These can be identified with the following zhng of the transmitted Laoz: 59, 48, 20 (beginning), 13, 41, 52 (middle), 45, and 54. Of the twenty-eight strips that constitute bundle C, fourteen strips together contain five units of thought, all of which can be identified with parts of the transmitted Laoz: 17, 18, 35, 31 (middle and end), 64 (end). Taken as a whole, the three bundles contain about two-fifths of the materials known to us from the transmitted versions of the Laoz. A division of the materials into d and do parts is not even remotely present. The three bundles contain different kinds of markings on the strips. Black dots of different sizes mark off some of the units. Yet this is
Ibid. For the Ti y shng shu , see ibid., 13 (for a photograph of the strips), 123126 (for text and notes). 15 Since the excavation of the Ti y shng shu , studies on this manuscript run into the hundreds. For a detailed analysis of this text and further references, see Allan 2003. For a review of (mostly Chinese) scholarship on the Ti y shng shu , see Chn Lgu 2005.
14 13

text, structure, meaning

189

not done consistently.16 In bundle A, space has been left after some of these signs, thus marking the start of a new unit of thought. Bundle A also contains two tadpole symbols ( ),17on strips a32 and a39. After these symbols, the strips carry no further graphs. The strips of the two other bundles also contain markings. But only those of bundle C are used consistently: every single unit of thought in bundle C is followed by a black mark ( ).18 One might be inclined to assume some kind of internal relation between units that are not distinguished from one another by such markings. One might even want to go so far as to propose that those instances that are not marked off as individual units of thought should not be conceived of as distinct units at all even though they are treated as such by othertransmittedversions.19 Nonetheless, when scrutinising more closely what has been written on these strips, the different units show no inherent relation with one another, and they differ appreciably in tone and style. Hence, it appears that most of these instances should indeed be conceived of as individual units of thoughteven if not signalled as such. In most cases this accords with the transmitted versions of the Laoz. The marks on the strips must therefore be explained in a different way. Did they function as indicators of the importance of some of these units? Or did they only represent unusual breaks (or changes) in the rhythm?we simply do not know.20 In the same vein, the tadpole symbol used in bundle A might further suggest a distinction between the different materials recorded on bundle A. But, looking at these units, the reasons for the tadpole symbols on strips a32 and a39 are in no way self-evident. One might feel inclined to understand the tadpole symbols as used in bundle A to signal the different origin of those units assembled in that anthology. Different procedures can be imagined. One may think of a scribe
16 A mark that signals the end of the unit appears after the following units: A 1, 3, 712, 14, 15, 1719. In unit A 4 such a mark appears right before the last character of that unit. This seems to be a mistake. 17 Here taken from strip a32. 18 Here taken from strip c3. 19 Units that are not marked off by a black square on the strips are A 2, 56, 13, and 16. One could therefore argue that the following clusters are related: A 2 and 3; 47; 13 and 14; 16 and 17. 20 Knowledge about the use of interpunctuation in early manuscripts is fairly limited. See, however, the survey by Png Ho (2000, 34ff.) on punctuation marks in excavated manuscripts. See also Guan Xhu 2002.

190

chapter five

who had to assemble different Vorlagen to complete his collection of Laoz and indicated the different sources with the tadpole symbols. Or one could argue that there was no such compilation as Laoz , and it could then be the case that the tadpole symbols signal different nonargument-based texts as collected on one bundle of bamboo strips. If this were the case, the materials on bundle A would have to be seen as three different collections of various unrelated units, all of which were anthologised and fixed on one and the same bundle of bamboo strips. This assumption is not implausible given the custom of assembling different texts on the same bundle of strips,21 which is why it is essential to keep text and manuscript separate when analysing written philosophy of that time. We should beware of imposing modern ideas of text and writing on the production of text and manuscript in early China. That different textual materials are collected in one bundle of bamboo strips does not per se point to the mutual relatedness of these materials. Units A 2 and A 3 provide one example ofpresumablydifferent units of thought that are not marked as such. These are the first units in bundle A that are not demarcated by any sort of markings on the strips. The two units also appear in the transmitted versions, where they are clearly distinguished as individual zhng .22 No obvious textual or rhetorical relation between the two units can be recognised neither in the Gudin One text nor in transmitted versions. It thus appears that different units that were anthologised in that bundle were not necessarily signalled by signs on the strips:
A2 a3 a4

The Qing d y sh, for instance, shares a bundle of strips with the Lu M gng wn Z S; the Zhng xn zh do shares a bundle of strips with the Tng Y zh do. 22 Zhng 66 and the middle and end of zhng 46.
21

text, structure, meaning

a5

191

The reason rivers and seas are kings of the many valley streams is because of their a3 ability to be below the many valley streams. That is why they are kings of the many valley streams. The wise man stands in front of the people because he puts them behind through his person. He stands above the people because he puts a4 them below in his words. He stands above the people, and yet the people do not regard [him] as heavy. He stands in front of the people, and yet the people do not harm [him]. The reason the entire world is delighted to advance him and avoids criticising him a5 is because he never competes; consequently, nobody under Heaven has the ability to compete with him. A3 a6 Of all fault, none is heavier than excessive greed, Of all blemish, none leads to more grief than craving gain, a6 Of all disasters, none is greater than not knowing when you have enough. To understand that [having] enough is enoughthis is to have enough eternally.

The example shows that the two units did not share any close relation with each other in terms of a formally coherent approach to a given issue. Even though to a certain degree the two units have a common concernthey adopt a position against greed and human striving for superiorityformally they do not share any related features. The common use of the negative distributive pronoun m none, nobody,23 is not a symmetrical grammatical feature of the two units.24 The two units of thought should therefore not be understood as parts of a larger
That is, in the last line of unit A 2 (strip a5) and the first part of unit A 3 (strip a6). Whereas the negative distributive pronoun m of A 2 negates an auxiliary verb, that of A 3 negates a full verb.
24 23

192

chapter five

consistent composition of the kind seen in the argument-based texts discussed above. For non-argument-based texts it can be concluded that different units of thought need not be signalled by markings on the strips. The lack of these signs does not imply an internal relation of the different units so that together they form a coherent whole. Each unit of thought should be understood as presenting an isolatedand therefore ultimateanswer to a particular concern, regardless of the physical organisation of the bamboo strips. The end of an individual unit therefore marks the end of the engagement with a certain concern. It follows that each unit is to be considered a text in its own right. Given that the different units remain unconnected and hence isolated standpoints on a certain issue, they cannot develop a systematic, or multilayered, analysis of a philosophical concern since that would imply an investigation of one matter from different perspectives. So whereas argument-based texts connect different ideas into greater meaningful wholes, the units under review state their point in a different way: they either use analogies of the kind seen in A 2, in which the good ruler is compared to rivers and seas, from which a lesson is drawn about the behaviour of theidealisedruler. Or they simply contrast right behaviour with wrong behaviour, as in unit A 3. In any case, these two units do not establish a complex approach. An organised analysis of a dominant issue from different perspectives is totally absent. The two units do not embody argumentative force. They rely on something else to make their point. For example, in unit A 3 the addressee simply has to accept as true the position Of all fault, none is heavier than excessive greed in order to believe that To understand that [having] enough is enoughthis is to have enough eternally. The recipient of the message is not guided through any kind of a balanced discussion. Confronting the imagined recipient directly with an intellectual position was considered fully sufficient to make the case. It can be assumed that meaning construction in these texts relied on the authoritative force of their statements rather than on an argumentative force attained by leading the recipient through a thought process. This becomes evident from two observations. First, despite their brief and at times even enigmatic nature, these rather simplistic statements were nonetheless considered important enough to make the effort of fixing them on bamboo. And this is true not only for the immediate intellectual environment of Gudin Oneboth spatially and chronologicallybut for that of Mawngdu Three also. Second,

text, structure, meaning

193

the units of thought collected in bundle A (and also in B and C) were handed down for generations. They all persisted in the later canonised compilation familiar to us as the Classic of the Way and Virtue. Despite the fact that most of these units took on an appreciably different form (certain words may differ, the internal structure of these units is not the same, some units are significantly shorter than their transmitted counterpart, and so on), the thought and tone in most cases nevertheless remain largely unchanged in the received classic.25

25 Differences from the Gudin One materials can nonetheless be seen in zhng 17 and, especially, zhng 18 of the transmitted versions. (In the following I refer to the Wng B edition.) The beginning sentences of zhng 18 of the transmitted text read Cut short sagacity, discard wisdom, and the people will benefit a hundredfold; cut short benevolence, discard rightness, and the people will return to filiality and [parental] compassion; cut short artistry, discard profit, and robbers and thieves will have no existence . This implies a disregard for values such as sagacity, benevolence, and rightness. The text on bundle A, in contrast, reads: Cut short wisdom, discard distinctions, and the people will benefit a hundredfold; cut short artistry, discard profit, and robbers and thieves will have no existence; cut short activity, discard deliberation, and the people will return [to the purity] of [unlearned] youngsters . This text apparently shows no disrespect for sagacity, benevolence, and rightness. This has triggered a debate over which of the versions renders more closely the original spirit of the Laoza debate that is renewed whenever there exists a close counterpart to an excavated text, as I have discussed at length in chapters 3 and 4. Briefly, there are two main currents in the way modern scholars explain the differences. One is to argue that the Gudin Laoz [sic] represents the original spirit of the Laoz; the transmitted versions represent later emendations, presumably made around the time of the appearance of the Zhungz. Most representative for this view probably is Qi Xgu (2004a, 2006). The other view is that the Gudin Laoz [sic] has been Confucianised and thus represents a later emendation. Scholars making this assumption believe that the transmitted versions of the Laoz render its original spirit. Most representative of this view probably are Zhu Fngwu (2000), L Lng (2002), and L Xuqn (2002, 2005). Both types of reasoning represent monocausal lines of argument which, implicitly, assume straightforward stemmata for early text development. Such an approach neglects the complex processes of textual development. If we were to draw a stemma codicum for the development of early manifestations of a text, we should rather draw the complicated structure of a shrub, also referred to as a rhizome (per Deleuze) by literary scientists. (Cf. Wests [1973, 14] use of recensions, in which he distinguishes between closed and open recensions. West defines recensions as stemmata that can be traced back to archetypes.) L Lng (2002) even assumes that the tomb occupant himself changed the text. This is just as unlikely. It cannot be assumed that the occupant wrote any of these texts at all. What is more, we cannot even be certain of any kind of relation between the tomb occupant and the texts under review. Hence, I have chosen to call the materials a tomb corpus instead of a tomb library, as the latter denotation inevitably invites us to assumegroundlesslythat the texts were selected by the tomb occupant himself. For the concept of rhizome, see Deleuze and Guattari 1980.

194

chapter five

Because these units of thought were considered important enough to fix on bamboo and all units persist to the present day, the units anthologised in bundle A (and likewise in bundles B and C) must have been commonly known and also highly respected, at least in some circles of intellectual activity in the Warring States period. That they were written down, anthologised, and, to some extent, remained unchanged casts light on the authoritative character of these statements. Groups (this certainly does not denote only a narrowly confined circle of people but rather refers to an abstract cultural identity) were constituted around these authoritative pronouncements. This further stabilised their authoritative character. However, the fact that these units have already taken on the shape of authoritative ideas prevalent in some lite circles does not necessarily imply the inverse conclusion that the authoritative character of these statements results from the existence of a prevalent concept of one authoritative Laoz behind them. Without a doubt, such a concept would connect these units of thought to one philosophically prevalent current, lending a group identity to those circles. However, as William G. Boltz has put it, we should beware of labelling a late fourth-century BC manuscript with a name, for which our first evidence is a century or more later.26 The same applies, it seems to me, to concepts. To conclude, meaning in this type of text relies on reference to some kind of authority. The nature of this authority, its origin, and its dialectical processes so far remain unclear. Whether derived from masters, wisdom, or antiquity, it is certainly something coming from outside these texts and thus imposed on the different units. It can, nevertheless, be asserted that the authority underlying these units of thought is dependent on context, and so too is the successful construction of the meaning of these units. As will be discussed below, the authority underlying these texts therefore was not static but changed over time. The transmission of these isolated units of thought to the present day furthermore shows that texts of this type were highly successful in their approach to meaning construction. To differentiate texts of this type from stand-alone argument-based texts, I shall refer to them as context-dependent texts.27

26 27

See Boltz 1999a, 596. In previous publications I have called these texts authority-based.

text, structure, meaning The Z y from Gudin One

195

Another example of a context-dependent text from the tomb corpus of Gudin One is the Z y. Just like the units collected in bundles A, B, and C, the Z y does not create a reasoned set of ideas by blending different units of thought into greater meaningful wholes of pericope, subcanto, canto, and finally the text as a whole. In contrast, as with the materials of bundles A, B, and C, meaning is constructed by reference to authority. The Z y from Gudin One is written on forty-seven strips. These are each about 32.5 centimetres long. The strips are tapered towards both ends and are well preserved. Not a single strip is in a fragmentary condition. As can be judged from the marks on the strips, two cords, 12.813 centimetres apart, previously bound the strips together.28 The text of the Gudin One Z y is complete. Each unit is followed by a heavy black marking, dividing the entire work into twentythree units of thought. This number is also mentioned explicitly at the end of the text. Somebody has closed it. In contrast to the materials assembled on bundles A, B, and C, the Gudin One Z y was apparently considered one entity during the Warring States period, at least by certain communities. The Gudin One Z y contains 1,156 characters. As such, it is notably shorter than the transmitted version of the text, now a chapter of the Record of Rites, L j , which contains about 1,549 characters.29 Compared with the received version, the exhumed units are organised more tightly. Twomaybe threeunits of the received version of the L j are not extant in the excavated text. Interestingly, the formal organisation of these units differs markedly from the rest of the work. Due to the otherwise consistent scheme of the individual units and the fact that they are not extant in either of the exhumed contextdependent texts (a version is also extant in the Shnghai collection of Chu manuscripts), one might be tempted to understand these passages of the received L j as later intrusions. Lastly, similar to the materials collected on bundles A, B, and C, the units in the Gudin One Z y also appear in a very different sequence from that of the transmitted counterpart.

28 29

Hbi shng Jngmn sh bwguan 1998, 129. See Shaughnessy 2006, 66.

196

chapter five

Z y: Two Manifestations on Bamboo and the Transmitted Text The two manifestations of the Z y on bambooone from the tomb corpus of Gudin One and the other from the Shnghai collection of Chu manuscriptsare strikingly similar. Ignoring for a moment the fact that the Shnghai manifestation of the Z y is badly preserved, the two manuscript versions of the text apparently share the same length, the same contents, and even the same textual order of the units of thought assembled in them.30 Notwithstanding the extraordinary overlap of the two, which in this form is exceptional, it nevertheless seems that neither of the two manuscripts was directly used as Vorlage for the other, because many of the characters of the two texts differ substantially, both structurally and phonetically. This needs to be explored in more detail. Most of the different characters can be explained by graphic variation, as defined by William G. Boltz.31 That means that in the two manuscripts different graphs were used for expressing essentially the same word. Many of the variations were obviously phonetically orientated.32 This phenomenon might point to a text that was relatively

30 The Shnghai instantiation of the Z y is written on twenty-four strips that are about 54.3 centimetres long when complete (today only eight strips remain intact). Originally, the strips were connected by three binding straps. For the Z y from the Shnghai collection of Chu manuscripts, see Ma Chngyun 2001, 1:4368, 169213. 31 See Boltz 1997, 258. 32 In many cases, the phonophoric remains consistent in the two versions, whereas the signific changes or is left out entirely. For instance, the Gudin One text writes y (OC *rak-s) where the Shnghai text reads w to hate (OC *ak-s) (see strip zy6/18 of the Gudin One text; Zy4/3 of the Shnghai manuscript). The two graphs are in the same xishng series and share the same phonophoric. They fulfil the criteria for loans in Old Chinese. Another example for a phonetically based graphic (strip variation is mi beautiful (*mrj): the Gudin One manuscript writes zy1/5), whereas the Shnghai text replaces the signific leaf with woman (strip Zy1/4). The phonophoric remains the same in both texts. They belong to the same xishng series and have the same Old Chinese reading: mi beautiful (*mrj); wi (OC *mj). (Note that Gassmann [2005] connects variations such as in the case of mi to grammatical functions. I consider this scenario rather unlikely. Scribes and philosophers were not identical during the Warring States period, and there is no good reason to assume that such variations reflect strict grammatical features rather than mere scribal idiosyncrasies.) On strip zy2/2, the Gudin One manuscript writes y model (OC *(r)aj) with the phonophoric wo (OC *aj) and the signific heart, whereas the Shnghai manuscript uses the signific earth (strip Zy1/24). In unit 2 of the manuscript versions (what would be zhng 11 of the received Z y from the L j [strips zy2/103/12 of the Gudin One text; strips Zy1/322/16 of the Shnghai text]; by received Z y , I refer to the version in the L j zhng y

text, structure, meaning

197

stable in wording but unstable in writing. It could, however, also point to a situation in which a scribe was reading the text aloud to himself while he was in the process of producing another copy of it. As discussed in the analysis of the Xng z mng ch, this situation would ultimately result in a phonetically based copy of the text rather than a text that reproduced the graphs of the Vorlage. This would explain the graphic variations in thesestructurallyvery stable texts. Graphic variations that are stable phonetically are therefore of no use in determining whether the scribes used one of the two manuscripts or a third unknown but commonly shared manuscript as Vorlage for producing the Gudin One and Shnghai manuscripts of the Z y. Accordingly, the assumption that a scribe was pronouncing aloud the text before him while he was producing another copy of it might seem most appealing, as this would explain the extraordinary similarity of

, which is part of the Ruan Yun (17641849) edition of the Thirteen Classics of 1815]) the Gudin One text writes jin (most likely for sh ) for what the Shnghai text expresses with sh , both of which mean to show. (See Shuwn jiz, 132.) Sh has the Old Chinese reading *[g]ij-s or *[g]ij-s (and jin has the Old Chinese reading *ke[n]-s or *N- ke[n]-s). The reconstruction of sh is perhaps more problematic. According to Zhngzhng Shngfng, sh can be reconstructed as *gij-sas can many other items in the series. However, according to Baxter and Sagart, the Old Chinese reconstruction for sh reads *[d]i[j]()-s. If that were true, it could either imply a palatalisation of velars before front vowels, possibly a Chu dialect feature, or indicate that a scribe mixed up velars and dentals when dictating to himself, an error known as coronal substitution. (Other examples include *[g]e for *[d]e; *tet as phonophoric in place of *ket-s). Examples of a mere phonetic borrowing of otherwise probably unrelated words include the use of shu to hunt (wild) animals (OC *nuk-s) in the Gudin One manuscript for shou to maintain, keep, guard (OC *[s-t]u) in the Shnghai text (strip zy38/21 of the Gudin One manuscript; Zy19/36 of the Shnghai manuscript). This often occurs in palaeographic materials. This loan becomes an option only after OC *n- and *s-t- had merged, which had happened at least by the time of Middle Chinese, when both become sy-, namely shu < syuwH and shou <syuwX. The use of shu and shou seems to indicate that the merger of *n- and *s-t- had already occurred by the late fourth or early third century BC (Baxter, personal communication, May 2008). Moreover, in a quotation from a now lost chapter of Documents, namely the Proclamation of Y n (Yn go ) (zy5/9 of the Gudin One manuscript; Zy3/15 of the Shnghai manuscript), the Shnghai manifestation of the Z y has kng (OC *k-l a) for tng (OC () It was only Y n who truly reached *(k-)la) in the line: () Tng [OC *la] []. Both roots begin with a lateral, but since the one in is voiceless, they are homorganic. In unit 21, we see the use of b robust horse (OC *[b]i[t] or *[b]ri[t]) for p mate; adversary, enemy (OC *phi[t]) (zy42/13 of the Gudin One manuscript; Zy21/41 of the Shnghai manuscript), although this might also be lexically related, since p was used as a classifier for horses since the earliest inscriptions. Loans of this kind do not rule out the possibility that the scribe might have dictated the text to himself.

198

chapter five

the two manifestations of the Z y on bamboo, despite the variations in their lexicons. But there are also other types of variation in the lexicon which yield yet further information about the textual relationship of the two manifestations of the text. These are, in particular, those graphs that in principle express the same concept but do not share the same phonetic value. One example appears in a quotation from the Documents, namely from the Lord Y (Y jn ), where the Gudin One version writes cng (OC *[ts]a) for hn (OC *[g]a[n]).33 The main vowel of the two words is identical, but they have a different position of articulation of the initial, and they do not share the same coda. The two words basically have the same (broader) meaning but a different phonetic value. Variations of this kind suggest that we should not assume that the scribes produced either of the two manuscripts by copyingand dictating to himself !one manuscript directly from the other one. Thus, in spite of the similarity attested, each of the two manuscript versions of the Z y should be considered an isolated manifestation of that text, as further examples of similar changes suggest.34 This observation demands an explanation for why the two isolated manifestations of the Z y, from the tomb corpus of Gudin One and the Shnghai collection of Chu manuscripts, display such a neat structural textual coherenceand stabilitybut, at the same time, differ substantially from the transmitted version of the text. A model is needed that accounts for the coherence of the two exhumed, fully stable manifestations of the context-dependent text Z y but that, at the same time, acknowledges that the two were isolated instances of fixing the Z y on bamboo. The Z y, as we see from the two isolated manifestations on bamboo, seems to have enjoyed the status of a stable work already during the mid- to late Warring Statesat least in the confines of the contemporaneous textual communities in

33 See strip zy10/8 of the Gudin One manuscript; Zy6/20 of the Shnghai manuscript: (Zy7/20 ) In the sharp cold of the extreme winter . . . That these changes apply predominantly to the quotations from the Documents bears further witness to the open nature of itstruly unstablebody during the Warring States period. 34 Other differences include variations in which the graph used has both a different reading and a different meaning, such as, for instance, seen in a quotation from the Odes, namely King Wn (Wn wng ; Mo 235) (zy34/6 of the Gudin One manuscript; Zy17/29 of the Shnghai manuscript), in which the Shnghai version writes y (OC *(r)aj-s) for jng (OC *kre-s).

text, structure, meaning

199

which the two manifestations of the Z y were circulating. The fact that the Gudin One text explicitly names the number of units at the end of the text (rsh yu sn ) further strengthens this assumption.35 The two manifestations of the Z y on bamboo are organised in the following fashion. The opening unit can be identified with what would be zhng 2 of the transmitted Z y from the L j. After what would be zhng 2 of the transmitted Z y come units 11, 10, 12, 17, 6, 5, 4, 9, 15, 14, 3, 13, 7 (which is split up into two units in the manuscript versions), 8, 23, 18, 22, 21, 19, 20, 24in that order. What would be zhng 1, 16, and the first part of 18 of the transmitted text are not extant in either of the two excavated manifestations of the Z y. The general structure of these units can be described as follows. Every unitboth in the transmitted text as well as in the versions on bamboois introduced by quoting the masters words. In most cases, this statement is headed by the formula The master says (z yu ). Only the first unit in the manuscript versions introduces the masters words with f z yu . This formula basically carries the same meaning as z yu (if is not read as a particle f), signifying and now (the master says), suggesting that this unit directly connects to a preceding discourse generally known to the texts addressee.36 It seems that this formula was used to signal the first unit of the context-dependent text of twenty-three units of thought.37 The introductory statement of each unit is in general relatively brief. It ranges from eleven graphs (in unit 20) to fifty-nine graphs (in unit 11) in the text of the manuscripts and is only slightly longer in the canonised Z y. The masters words of the introductory statements are always followed by reference to another authoritative source. In the manifestations on bamboo this can be either from the Odes or a quotation from what consensus identifies as the Documentsor

35 Strip zy47/58. The number has been written on the middle of the strip with blank areas above and below. 36 This would be in line with the use of the particle f in the Huinnz, the first word of the text. For a discussion of the particle f in the Huinnz, see Ess 2007. For a study on the rhetorical function of the initial f as a phrase status marker in Chinese philosophy, see Wagner, forthcoming. 37 In the transmitted version of the Z y, units 1 and 7 have different introductory formulae.

200

chapter five

by reference to both.38 The authors of the Z y themselves do not appear in the text with their own voicesneither in the text of the manuscripts nor in the transmitted Z y. Ideas are given expression only through reference to the pool of a shared cultural identity. Just as seen from the various units assembled in bundles A, B, and C, the various units of thought that constitute the Z y rely on the authoritative force implied in these statements to convey meaning:
Gudin One Z y unit 1 39 zy2 Now the master said: Love beauty as [I] love the Black Robes,40 hate wickedness as [I] hate the Senior Palace Eunuch41and the people will then all submit [to you], and [your] model will not fray away. In an ode zy2 it is said: A model of decorum, that was King Wnthe ten thousand states [all] acted sincerely.42
zy1 zy1

Zhng 18 of the transmitted Z y strongly differs in style from both the other units of the manuscript manifestations and the canonised Z y. It has twodifferentintroductory statements from unspecified master(s), commonly understood to be Kongz. Only the second of these statements is followed by the reference to Odes or Documents. The first statement does not appear in the manuscript versions of the text. Units 1 and 16 of the transmitted textboth of which are not extant in the manuscript versions of the Z yalso differ from the overall style of the work. Unit 1 quotes the masters words without any further reference to an ode or a passage from the Documents. Based on this, Shaughnessy (2006, 75, 77) believes that unit 1 belongs to the Biao j, which heads the canonised Z y of the L j. Unit 16, in contrast, lists four different quotations from the Documents, otherwise unseen in the Z y. It seems that this is a later insertion. Also, the transmitted Z y once quotes the Changes, Y , namely in the concluding line of the final unit, 24, which also is not seen in the manuscripts. The manuscript versions of the Z y make no reference to the Changes at all. 39 I follow the suggestion of the editors of the Shnghai Z y (see Ma Chngyun 2001, 1:175) and the editors of the Gudin One Z y (see Hbi shng Jngmn sh bwguan 1998, 131, n. 4) and read zy1/17 with xin all instead of zng good. For the graph zy1/18 f to submit ( in the manuscript) I follow Shaughnessy (2006, 94, n. 39), as the archaic forms of and (the last word of the ode cited) are cognate. For the graphs zy1/21 (model) and 23 (crumble) I also follow Shaughnessy 2006, 7274 (see also Shaughnessy [Xi Hny ] 2004, 294295). 40 Black robes were used as court dress for high-ranking ministers during the Zhu dynasty (ca. 1025256 BC). Moreover, Black Robes is also a song in the Odes (Mo 75). 41 Xing b Senior Palace Eunuch was a title used in the Zhu court. It is also the title of an ode (Mo 200). 42 Quoting the ode D ya: Wn wng (Mo 235).

38

text, structure, meaning

Gudin One Z y unit 2 zy3 []

201

The master said: If he who possesses a state displays what he loves and displays what he hates so as to show the common people what he values as important, then the zy3 sentiments of the people will not be flawed. In an ode it is said: Be thoughtful and deferential in your position; love those who are upright and righteous.43

As is characteristic of context-dependent texts, the units under review rely on authority to construct meaning.44 This may either be odes or the words of an unspecified master (or masters), traditionally taken to be Kongz . None of these units attempts to convey meaning by employing a systematic approach to the issue. Different units of thought are not integrated into a larger and coherent wholeat least not on the formal level of composition. Instead, as seen from the materials assembled in bundles A, B, and C, the different units remain formally unrelated and fully isolated. Even though the two units under review share the same focus, namely proper rule, and refer to authoritative sources similarly, they do not advance an integrated treatment of the subject in the fashion of argument-based texts. Instead, as seen from the materials collected in bundles A, B, and C, in which what LaFargue calls Laoist sayings corroborate a generally shared belief, meaning construction relies on the reference to authority, whether to unspecified masters or to other sources of cultural authority.45 The recipient simply has to trust that If he who possesses a state displays what he loves and displays what he hates, then the feelings of the common people will not be flawed. Or the recipient has to accept the analogy drawn in the second unit quoted, namely that loving beauty and hating wickednessas the masters love the Black Robes and hate the Senior Palace Eunuchwill result in the peoples submission to the ruler. He will be like King Wn, the model of proper rule and the ideal of every ruler. As seen from the materials collected in bundles A, B, and C, the recipient is supposed to accept the positions advanced in

Quoting the ode Xiao ya: Xiao mng (Mo 207). I have randomly chosen the first two units of thought to discuss, but other units present the same picture. 45 For LaFargues ideas on the Laoz , see LaFargue 1992, 1994. On his notion of sayings from the oral tradition of a small Laoist community, see LaFargue 1992, 190ff.
43 44

202

chapter five

the different units of thought as true simply because they are authoritative. The texts do not aim to persuade the recipient by providing a balanced treatment of certain ideas. The recipient either subscribes to the positions advanced or he or she does not. This does not imply that the individual units of the manuscript versions of the Z y lack a consistent outlook. They consistently state the matter by quoting the masters insights, which function as introductory formulae marking off the beginning of each unit of thought. The masters words are followed by a reference to authoritative and well-known sources, namely ancient odes and, to a lesser extent, passages from the Documents. Despite the fact that the different units are not blended into a unified whole structurally (formally, they remain fully isolated, and hence specific answers to a given concern), the uniform pattern of each of these units nevertheless creates a sense of consistency and allows the recipient to identify with the work as a whole. The pattern and the thought of the work harmonise. Moreover, as Martin Kern has already pointed out, the macrolevel consistency of the excavated Z y is also an effective tool for ensuring textual stabilityboth for the individual unit of thought and for the work as a whole.46 It follows that even though the individual units of thought of the Z y remain separate, the macrolevel consistency of the work nevertheless allows the conclusion that the Z y was already conceived of as one work during the Warring States period. The kind of variation seen in the lexicon of the two isolated manifestations of the manuscript versions of the Z y further suggests that the text was predominantly transmitted orally rather than in writing. At the same time, this context-dependent text was a structurally closed entity already in the pre-unification Kingdom of Chu, and it was also recognised as such. The structural consistency between units of thought ensured textual stability and at the same time also made them easy to recognise as part of the Z y. Before closing this chapter, I want to draw attention to the observation that the first unit provides the hermeneutical key (Leseanleitung) for each particular unit of thought as assembled in this work, as well as for the Z y as a whole. Masters state that one should love beauty as they themselves love the Black Robes, and that one should hate wickedness as they hate the Senior Palace Eunuch. Clearly, more than just

46

See Kern 2005b, 300301.

text, structure, meaning

203

the literal meaning is involved here. Black Robes and Senior Palace Eunuch cannot simply refer to the physical court dress of high ministers at the court of the Zhu and its ritual meaning and to the office of the Senior Palace Eunuch and his potential influence on the ruler. Instead, the two references allude to odes and trust that the recipient understood this reference to cultural authority. But they do not draw on what is articulated in the odes themselves. Instead, the authors of the Z y use the phrases Black Robes and Senior Palace Eunuch to refer to what certain cultural communities agreed was the masters particular interpretation of these odes. As will be discussed below, the authors of the Z y could safely assume that within the confines of certain textual communities, knowing Black Robes implies both knowledge of a particular cultural interpretation of the ode by the textual community which the authors had in mind when composing the Z y, and also the identification of it with theunstatedinterpretation of these odes. Knowledge alone would not suffice to convey meaning. And meaning construction in the Z y was successful, at least with respect to certain textual communities. We know this because the effort was made to fix the different units on bamboo (at least twice) and because the work as a whole maintained considerable stability during the Warring States so that the units survived to the present day in canonised form. To understand the allusion to Black Robes one would have to know about the virtuous behaviour of Duke Hun of Zhng (r. 806771 BC) and of his son Duke Wu (r. 770744 BC), as, for instance, suggested by the Mo reading of the ode, and, at the same time, one would have to identify this knowledge with the unit of thought in question.47 In the same vein, the Senior Palace Eunuch refers to the wickedness of a senior palace eunuch during the reign of King Yu of Zhu (r. 781771 BC), if one again follows the Mo reading at this point. When this culturaland thus group-basedinterpretation is borne in mind, this unit appears in a completely different light. The references can be appreciated only by understanding that mentioning these odes does not describe masters bias toward or reservations about the odes themselves but refers to the complex cultural interpretation of these odes as agreed on by

47 It is, of course, conceivable that the authors of the Z y bore an interpretation in mind that differed radically from the, later prevalent, reading of the Odes as given by the Mo tradition.

204

chapter five

certain textual communities which consistently identify a corpus of texts, written or oral, as authoritative and agree on one consistent interpretation of these texts. In summation, meaning construction in the Z y can be successful only if the recipient is equipped with special cultural training. Mentioning Odes in exactly this way requires at least one hermeneutical step to be taken before the unit of thought becomes meaningful. To make complete sense, the different quotations in the various units of thought must be taken as integrated cultural wholes. Reference to an ode always implies the reference to a particular cultural understanding of it, as agreed on by certain textual communities. By demanding a particular strategy to generate meaning, the first unit of thought reveals that the other units of this work ought to be read in exactly this fashion. Functioning as a hermeneutical key to the Z y at large, it is not surprising that this unit begins the Z yand that it is marked accordingly.48 The fact that the construction of meaning in a text like the one under review largely lies in its reference to authority and, crucially, itsimplicitcultural interpretations, but not in the written text itself, makes it plain that meaning is generated by pointing to the world beyond the written text. These texts rely on the philosophical contextualisation of their units, and so they crucially remain bound to the triangular relationship between the text, the mediator of meaning, and the recipient of the message that characterises the structure of meaning conveyance underlying context-dependent texts. This casts light on another feature of context-dependent texts: the ambiguity of the given text and, accordingly, the suitability of the units of thought for use in different contexts. They become widely applicable modules. Here I should point out that it has been argued that the Z y developed out of a connected discourse and became a sequence of isolated units only at a later stage of text development.49 According to this hypothesis, the primary text layers were reflections on rulership to which elements such as the formula z yu (or f z yu ), [and now] master(s) said and lines from the Odes were added. Consequently, the connected discourse was broken up and the text

48 As noted, only the first unit of thought in the Z y is introduced by the formula fz yu instead of simply z yu . 49 See Kalinowski 20002001.

text, structure, meaning

205

assumed its modular form. In a third step, the individual units were reshuffled and additional use of authoritative references was made, so that the received version finally lost the meaningful progression of statements present in the primary text layer(s). Although we can discern dominant themes in the Z y, it is probably too early to substantiate this hypothesis with conclusive evidence. Even if one were to accept this hypothesis of text development, however, it would not alter the reading of the Gudin One Z y as a context-dependent text. In the course of the pictured placement of the authoritative quotations, the units clearly gained primacy over previous hypothesised text layers. Already by the time of the Gudin One manifestation of the text it had lost the form of a connected discourse. The number of units given at the end of the text, and the fact that they were rearranged in the received version make this plain. Whether certain textual communities might have made sense of the Z y primarily against the background of earlier text layers or, in fact, in the context of a mediated, cultural knowledge is thereby irrelevant. In the textual communities in which the Z y as manifested on bamboo was circulating, meaning was generated by reference to evocative structures that lay in the intellectual environment beyond the actual writtentext but not in the modular Z y itself. Conclusion Unlike the argument-based texts, the context-dependent texts from the same corpus of texts do not articulate any concern in writing beyond the level of the particular unit. The individual unit remains formally isolated and must be approached in its own right. No formal elements blend the different units into larger integrated wholes of pericope, subcanto, canto, and finally the text as a whole. On the formal level of the text, context-dependent texts should be taken as anthologies of diverse units of thought. Despite this, the example of the Z y as materialised on bamboo shows that a context-dependent text can nevertheless have a consistent outlook. Even more importantly, a discontinuous context-dependent text may even have been composedand understoodas an entity already in the pre-unification Kingdom of Chu . As discussed, the initial unit of the Z y features as a Leseanleitung (hermeneutical instruction) for the anthology of ideas as a whole, and it is marked

206

chapter five

accordingly. That the entire collection of thoughts was taken as one fixed corpus in those textual communities in which the Z y under review was circulating is also shown by the number appearing at the end of the last unit, signalling that the text is complete. The macrolevel consistency that applies to the text as a whole further highlights this. The macrocoherence of the Z y furthers the recognition of the various units. It allows the identification of the recipient with the work at large. In a reverse process, this also furthers the texts stability, which, again, reinforces a greater acceptance of what is expressed in the various units of thought, leading again to additional textual stability. The macrolevel consistency of the Z y suggests that it is unlikely that this anthology is an accidental collection of otherwise-unrelated materials. This is relevant insofar as it reveals that the compositional difference between argument-based and context-dependent texts is not sufficiently explained by postulating a teleological model of text generation that would point to an earlier date for the arguably less mature context-dependent texts. Argument-based texts seek to establish their point by achieving argumentative force. The recipient is persuaded to accept the philosophical position presented in the text as good and, accordingly, as something that can be put into practice. Context-dependent texts, in contrast, rely on authority to state their concern. Reliance on authority that is not contextualised in the given text implies ambiguity and susceptibility to use in different contexts. Whether sayings of masters or references to Odes, Documents, or the kulturelles Gedchtnis of lite groups at that time,50 the acceptance and identification of the text with these sources of cultural identity are necessary preconditions for any successful conveyance of ideas. And meaning construction in these texts was successful, as the different manifestations of this text suggest, thus pointing to the agreement of the recipient with the ideas transported and highlighting the predominantly oral contexts underlying meaning construction in these texts. The structure underlying the process of meaning conveyance in context-dependent texts, as well as that of philosophising on the basis of these texts, was that of a triangular relationship between text, the mediator of meaning, and the recipient of the message.

50

For cultural memory, kulturelles Gedchtnis, see J. Assmann 1999, esp. 1924.

text, structure, meaning

207

It follows that meaning in context-dependent texts was imposed on the written units from the outside. Ideas are not given expression in writing but rely on a group-based expatiating and contextualising of the various units of thought. Without mediated efforts of clarification, the different units remain unspecific and ambiguous. But the explanatory element of the text is never really part of the context-dependent text itself. It remains a feature that is external to the given text. Since none of the context-dependent texts yield such an explanatory element in writing (e.g., in the form of a commentary attached to the different units of thought), it becomes clear that expatiating and contextualising the various units of thought had to be a predominantly oral exercise at a time when these texts had already taken material form on bamboo. For context-dependent texts, it thus remains a groups decision how to understand thein itself unspecific and enigmatic, sometimes epigrammaticunits of thought. The different units therefore remain open to a wide range of interpretations. In each case, the particular textual community around a context-dependent text decides on the best way to make sense of it. Later exegetical efforts concerned with the context-dependent text Laoz, and the consequent appropriation of that text by different groups according to different needs, make this plain.51 Textual communities around context-dependent texts were not static but changed over time. In contrast to context-dependent texts, the argumentative nature of argument-based texts facilitated the accessibility of these texts to a potentially wider audience. By reasoning on a certain philosophical topic, these texts advance systematic approaches in order to be persuasive. Even though they may also cite the lore of a shared lite groups identity, the construction of meaning goes beyond the mere recognition of authority. Meaning becomes accessible to everyone who is able to read and has access to these texts. It can safely be assumed that texts of this type were detached from oral contexts and circulated independently in writing. Ironically, it is exactly the texts that worked in oral contexts which finally survive in the written textual tradition, while the early written texts all drop out. This needs to be explained and I come back to this further below in this study.

51

See in this context the informative contribution by Robinet 1998.

CHAPTER SIX

APPLYING THE METHODOLOGY: TI Y SHNG SHUI THE ULTIMATE ONE GIVES BIRTH TO WATER AND LA OZI Three bundles of bamboo strips from tomb Gudin Onetentatively called bundles A, B, and Ccontain textual materials that closely resemble roughly two-fifths of the transmitted Laoz. These materials naturally generate many questions. On the one hand, they resemble the transmitted Laoz closely in spirit and tone. On the other hand, they differ considerably from the transmitted text in style and content. Despite the fact that they have been placed in the same tomb, the texts are fixed on different bundles of bamboo strips, and they also seem to be inscribed by different hands.1 As for the internal organisation of these texts on the different bundles of strips, the sequence of the individual units of thought differs significantly from that of the transmitted Laoz, and they furthermore often deviate substantially from the transmitted text in lexicon, length, as well as their internal structure and composition. In addition to this, some fourteen strips of bundle C also contain textual materials otherwise unknown and thus not part of the transmitted Laoz. Present-day scholarship refers to these as Ti y shng shu (The Ultimate One Gives Birth to Water), a title based on the first four characters of the text. The finding of the textual materials on the three bundles once more calls for a reevaluation of the nature of what we today call the Laoz. In this chapter, I shall apply the methodology outlined in the preceding chapters to these materials and elaborate on their textual relationship. Due to the fact that the textual materials of the context-dependent texts2 collected in bundles A, B, and C display such grave differences from the transmitted Laoz, a great many scholars also understand the

See Boltz 1999a, 592. Note that the materials were not only collected on different bundles of bamboo strips but also distinguished by different markings on the strips, such as, for instance, the two tadpole symbols in bundle A. This might suggest different texts or sources collected in one bundle. Thus, the fact that Mawngdu Three contained a complete Laozeven though organised in a substantially different way than the transmitted
2

210

chapter six

otherwise unknown materials collected on bundle C as surviving elements of a so-called Warring States Proto-Laoz which, however, did not make it into the transmitted text. In the following I shall outline an alternative view. Briefly, I hold that the text of the Ti y shng shu is not an integral part of an imagined coherent composition that would include the texts recorded on bundles A, B, and Can envisioned Gudin Laoz or Ur-Laoz . Instead, the Ti y shng shu should be understood as an argument-based text in its own right. Even though it shares some notions that are given expression also in the transmitted Laoz, a closer look at the Ti y shng shu nevertheless suggests that both the formal structure of composition of the text and the main ideas presented in it are not in congruence with the context-dependent texts of bundle C (or, in fact, A, B, and C). Thus, the fact that the Ti y shng shu is collected in one bunch of bamboo strips together with other context-dependent texts does not necessarily suggest their close relationship. Instead, during the Warring States period it was common for a text to share a material carrier with unrelated textsfor reasons that so far remain beyond full comprehension.3 As a general tendency, however, the explanation for this custom seems to lie in the physical availability of certain materials rather than an internal relationship of the texts collected. During the Warring States, a manuscript culture gradually came into existence. The increase in the number of written texts brought about changes in the production of manuscripts, and it may be assumed that texts increasingly became at tradable commodity. This might have led to the professionalisation of manuscript production, including the division of labour. It is likely that professional scribes copied texts onto those suitable objects that were available for writing, probably prepared by other hands. Even if we assume that a manuscript was produced by the same person, the different steps in the production of a manuscript were quite certainly streamlined. It can thus be assumed that, usually, the physical objects were not produced per text, but the texts were fitted on the physical objects prepared in advance and made available for manuscript production. Despite this,

versionsdoes not substantiate the assumption that the materials from Gudin One should likewise be seen as one text, only because later anthologies do so. 3 The Qing d y sh and the Lu M gng wn Z S, for instance, were also collected in the same bundle of bamboo strips even though the two share no close relationship with each other.

applying the methodology: ti y shng shui

211

it still seems rather unlikely to find competing materials on one and the same material carrier. But this has more to do with the customs of collecting available textual materials (possibly those with which the collector wanted to be identified) rather than with the abstract objective of writing down related texts on a shared material carrier. Whatever the precise process of manuscript production might have been, a text can be defined as the formulation of an idea abstracted from its material carrier. As such, it could travel independently of material contexts. Hence, we need to draw a clear methodological distinction between text as matter to be transmitted and the manuscript as the physical instantiation of a text. For the written philosophical texts from the Warring States it follows that it is methodologically illfounded to hypothesise about their intellectual orientation based primarily on their material condition such as the length or the cutting of the bamboo strips or the observation that texts were grouped together on the same material object. Such an approach would inevitably lead to a preconditioned understanding of the texts. Ti y shng shu : An Argument-Based Text Bundles A, B, C differ in length,4 modes of manufacture,5 and calligraphy. The texts were written on bamboo by at least three different hands. The textual materials were probably collected at different places, maybe even at different times, or possibly copied from different Vorlagen. How do we explain this state of affairs? Was there no single coherent Vorlage to consult? And if there was no such Vorlage, did one Laoz exist at all during the Warring States? And if it did, why were only certain units selected instead of the text being copied in its entirety? What happened to the remaining three-fifths of the transmitted Laoz that did not find its way on to at least one of the three bundles of bamboo strips, now referred to as A, B, and C? And if the units chosen were indeed taken from an already complete text, then why would

Bundle A contains strips that are 32.3 centimetres long; B, 30.6 centimetres; and C, 26.5 centimetres. 5 The strips of bundle A are tapered towards both ends, whereas those of bundles B and C are cut evenly at both sides.

212

chapter six

only parts of it be copied? Why would even structurally coherent units be shortened, since some of the modules seen in the three bundles are appreciably shorter than those of the transmitted text? As shown by the analysis, these would be theoretically stable units of thought. Why was the tadpole symbol inserted twice in bundle A? Other materials from Gudin One suggest that this symbol signals the end of either an individual text or of self-contained parts of it. These questions all point to the dispute about whether the transmitted Laoz was written by one historical figuremaybe around the late sixth or fifth century BC6or whether it should rather be understood as an anthologised collection of different ideas, composed by different people in different periods.7 As a result of losing confidence in the existence of one composition identified with the Laoz in the Warring States period (or even in the existence of the entire concept of a Laoz itself ), a number of scholars consider the Ti y shng shu an integral part of materials collected on bundle C. This would be the final blow against the concept of a coherent Laoz , since, as mentioned, none of these materials found their way into the later versions of the text. I fully subscribe to these hesitations concerning the notion of one Laoz composed in its entirety by one person, whenever that is proposed to have taken place. However, I do so for different reasons. The problem with considering the Ti y shng shu an integral part of the materials collected on bundle Cthe so-called Gudin Laoz Cwhich, at first sight, might seem to be the logical step in a methodologically accurate text critique, is the fact that the materials now referred to as Ti y shng shu are organised in a completely different fashion from the other materials collected in that bundle of bamboo strips. The Ti y shng shu contains materials that extend over fourteen strips. In contrast, none of the other units of thought collected on bundles A, B, and C need more space than three to four bamboo stripsat the very utmost. Thus, the sheer length of more

As, for instance, Chn Guyng (1992a) argues provocatively. Shaughnessy notes that this view has by now received some consensus status in China (2005, 443). 7 This was first argued by G Jigng (1932). Representative of Western views on the Laoz is the highly influential study and translation by Lau (1963). In several monographs and articles, LaFargue (1992, 1994, 1998) has deepened this understanding of the Laoz.

applying the methodology: ti y shng shui

213

than four times the average unit of thought in the discontinuous context-dependent texts collected in bundles A, B, and C makes it rather unlikely that the Ti y shng shu is an inherent part of these collections of thoughts, if one were to accept the supposed unity of the text. But the unity of the Ti y shng shu itself is not undisputed. Subsequent to what is generally perceived as a cosmogony (strips 18), the text changes its focus and shifts to a politico-philosophical level (strips 914). As a result, the materials are sometimes split into two individual texts and interpreted as lost parts of a Warring States Proto-Laoz . Plausible as this assumption might seem at first sight, it is illfounded for being ultimately shaped by the mere observation that the textual materials now called Ti y shng shu were collected on one and the same bundle of strips with the individual units of thought assembled in bundle C, which, based only on much later findings, are correlated with the ideas anthologised in bundles A and B. The Ti y shng shu is given no room to be understood in its own right. Yet, as detailed, collecting different texts in one and the same bundle of bamboo strips was a common practice during the Warring States. It does not imply that they were ever considered as integral parts. As mentioned, the Ti y shng shu consists of two parts, a cosmogony and a politico-philosophical discussion. But these two parts are not individual texts. Instead, they were two parts of a consistent whole. By implication, when applying the methodology of textual analysis as developed in this study, it becomes clear that the Ti y shng shu as a whole establishes one coherent argument. Unlike the remaining units anthologised in bundle C, the Ti y shng shu is, as I will show, an argument-based text. The politico-philosophical discussion is an integral part of the Ti y shng shu and is based on the preceding cosmogony and has to be contextualised accordingly. The politico-philosophical discussion of the Ti y shng shu should thus be seen as the concrete application of the insights gained from the cosmogony presented in the text. The Ti y shng shu addresses the source of political power of a state. The text can be understood fully only when taken in its entirety. The concern of the text is to explain the conceptual meaning of the Way (do) as pictured by the authors of the text, and from this to draw the relevant conclusions for proper rule.

214 Part 1: Cosmogony

chapter six

[A] []ty2 [B] ty3 [] ty4 The Ultimate One gives birth to water, water returns and joins with the Ultimate Onethis is how it completes Heaven. Heaven returns and joins with the Ultimate Onethis is how it completes Earth. Heaven and Earth {repeatedly join with each other} ty2this is how they complete the spirits and the illuminated.8 The spirits and the illuminated repeatedly join with each otherthis is how they complete the shadowy and the sunny. The shadowy and the sunny repeatedly join with each otherthis is how they complete the four seasons. The four seasons ty3repeatedly join {with each other}this is how they complete coldness and heat. Coldness and heat repeatedly join with each otherthis is how they complete moisture and dryness. When moisture and dryness repeatedly join with each other, completing the year, ty4[the circle] stops. []; [C] []ty5 ty6 Therefore, the year is begotten by moisture and dryness. Moisture and dryness are begotten by coldness and heat. Coldness and heat {are begotten} by the four seasons. The ty5 {four seasons} are begotten by the shadowy and the sunny. The shadowy and the sunny are begotten by the spirits and the illuminated. The spirits and the illuminated are begotten by Heaven and Earth. ty6 Heaven and Earth are begotten by the Ultimate One.
ty1

ty1

On the concept of mng , see Maspro 1933.

applying the methodology: ti y shng shui

[] [][][] ty7 ty8[E] [][] [F]

215
[D]

From this it follows that the Ultimate One is stored in the water, moves with the [four] seasons, [finishes] a circle, and then {starts over again}. . . . {Hence the Ultimate One is} ty7the mother of the myriad things; at times lacking, at times full, it takes itself to be the alignment of the myriad things. This is what Heaven is unable to kill, Earth ty8is unable to smother, the shadowy and the sunny are unable to complete. The gentleman (jnz) who grasps this is one whom we call a {sagacious person} . . .

Part 2: Application
. . . ty10[G] [H] . . . . ty10Below is soil, yet we call it earth. Above is vapour, yet we call it heaven. Do likewise is [only] a style name for itMay I [thus] ask for its [real] name? ty11[I] ty12 He who ty11carries out his service according to the do necessarily consigns himself to its [real] name; this is why [his] task is completed and his person can endure. In carrying out his service, the sagacious person also consigns himself to itsty12[real] name; this is why [his] achievements are completed and his person will not suffer harm. [J] [] ty13 [][. . .]

[K]

As for Heaven and Earth, their name and style name are established simultaneously; as a result, once moving beyond these realms, one does not think [of them] appropriately . . . . . . {If Heaven does not suffice} ty13in the northwest, what is below it rises in strength. If Earth does not suffice in the southeast, what is above it {rises in strength} . . .

216
ty9

chapter six
[L] [X . . .] [M] [] ty14 The Way of Heaven is to value weaknessit reduces what is completed so as to add to what is living. Cutting back on strength, requesting from {. . .}. {This is why what does not suffice above}, . . ty14will have a surplus below. [And] what does not suffice below will have a surplus above.
ty9

As with nearly every new text that comes to light, the Ti y shng shu generates more questions than one can answer. Its philosophical affiliation is debated with great passion, and some of its concepts remain obscure.9 Even the proper order of the strips is still an open question. Especially the position of strip ty9 remains a focus of dispute, as this particular strip does not connect to continuous sentences above or below.10 For the moment, I tentatively place it right before strip ty14. In this I follow Qi Xgu , who sees a coherence of the three clusters ty18, ty1013, and ty9 and 14.11 Based on the fact that the Ti y shng shu has been fixed on the same material carrier as those units which consensus refers to as Gudin Laoz C, the majority of studies discussing the Ti y shng shu focus on its supposed relation to these units.12 The fore-

9 L Xuqn (1998a, 1999c) sees a close affiliation of this text to ideas of the socalled Gun Yn Daoists; his viewpoint has received much support, especially in China. Hung Zho (2000), for his part, sees a close affiliation of the Ti y shng shu with Jxi scholarship. Most scholars follow Dng Sxn (2002) in dating the Ti y shng shu to the late Warring States period. On Warring States concepts of the Ultimate (or Great) One, see Allan 2003. 10 As Qi Xgu notes, the editors of Hbi shng Jngmn sh bwguan 1998 originally placed strip ty9 before strip ty13 and after ty12. It was only due to Qis demur that the editors finally located it following strip ty8. See Qi 2000b, 219220. Later Qi withdrew his earlier proposal to insert strip ty9 following ty8 and before ty10. According to his current view, strip ty9 should be placed right before the final strip ty14 and after ty13, thus arriving at the following order: 113, 9, 14. See Qi 2000b, 220. As is so often the case, once made public, his view enjoyed much agreement among scholars. Whereas Chn Wi (1999, 2000b) suggested placing strip ty9 following ty12 and before ty13a view shared by Cu Rny (1998, 37) and Li Xnfng (1999, 76, 78)he subscribed to Qis view after having seen his article; see Chn Wi 2003, 24. 11 See Qi Xgu 2000b, 221. According to this order, strips ty18 outline the cosmology of the Ti y shng shu . The cluster ty1013 discusses the importance of proper naming, the cluster ty9, 14 discusses the fact that the Ultimate One in itself values weakness. 12 See Cu Rny 1998; L Xuqn 1998a; William Boltz 1999a; Chn Wi 1999, 2000b; Dng Sxn 2002; Chn Lgu 2005; among others.

applying the methodology: ti y shng shui

217

most goal of my brief discussion of the Ti y shng shu is to show that it did not belong to any of the context-dependent texts assembled in bundles A, B, or C, let alone to an imagined text spanning these three bundles of bamboo strips. In what follows below, I will discuss how the Ti y shng shu works as an argument-based text in its own right. The Structure of the Ti y shng shu It has been repeatedly argued that the Ti y shng shu displays a considerable overlap of ideas with some units of thought of the transmitted Laoz (and with some units collected in bundles A, B, and C).13 It is held accordingly that the Ti y shng shu should be understood as a hermeneutical device that explores the concerns expressed in some of those units of thoughtas an early commentary to the anthology of ideas called Laoz, so to speak. This would nicely explain, as argued, why it shares the same bundle of bamboo strips with these units.14 It is against this background that some scholars split the Ti y shng shu up into separate units, each of which is considered to relate exclusively to one individual zhng of the transmitted Laoz.15 It is even suggested that the units of the Ti y shng shu were set apart from one another by black markings on the stripsas seen in the context-dependent text of bundle C and to a lesser extent also seen in the context-dependent texts from bundles A and B.16 By implication,

13 William Boltz (1999a, 595), for instance, notes that the Ti y shng shu contains passages corresponding to zhng 17, 18, 35, 31, and 64 of the transmitted Laoz. See also Qi Xgu 2000b. 14 L Xuqn (1998a), for instance, reads the first eight strips of the Ti y shng shu the part that lays out a cosmogonyas a complementary elaboration of zhng 42 of the transmitted (Wng B) Laoz. See also Qi Xgu 2000b. 15 Chn Wi (1999, 2000b) divides the Ti y shng shu into three units. As Chn hypothesises, these are concerned with zhng 42, 25, and 77 of the transmitted Laoz and must, by implication, be understood as a further elaboration of these. If this were indeed the case, it would imply that the context-dependent texts collected in bundles A, B, and C represent a selection from a, by that time already complete, Laoz. I doubt this. 16 See Qi Xgu 2000b, 220ff. According to Qi, each of the units ty18; 1013; 9 and 14 were followed by black markings on the strips. Just like Chn Wi (1999, 2000b, 2003), Qi also believes that the Ti y shng shu is a further elaboration of the Laoz. Yet, unlike Chn, he does not see a one-to-one relation between the units from the Ti y shng shu and zhng 42, 25, and 77 of the transmitted Laoz. At the end of the text, the Ti y shng shu was indeed closed by a black sign on the final

218

chapter six

the Ti y shng shu is made into a compilation of otherwise unrelated units of thought, whether part of the materials collected on bundles A, B, and C or a hermeneutical tool elaborating them. The problem with this approach is that it is ultimately determined by the transmitted Laoz. I propose instead to test a reading which considers whether the fourteen strips that constitute the Ti y shng shu may be understood as a coherent whole despite the fact that they share a material carrier with the context-dependent texts of bundle C. I shall hence investigate whether it is possible to detect a coherent system behind the composition of the text as a whole.17 One of the keys for testing the overall coherence of the Ti y shng shu is a close reading of the third building block of this text (strips ty6/8ty8/end). This unit closes the cosmogony of the text by focusing on the characteristics of the Ultimate One. It also brings into play the gentleman (jnz ).
[] [][][] ty7 [D] ty8[E] [][] [F] From this it follows that the Ultimate One is stored in the water, moves with the [four] seasons, [finishes] a circle, and then {starts over again}. . . . {Hence the Ultimate One is} ty7the mother of the myriad things; at times lacking, at times full, it takes itself to be the alignment of the myriad things. This is what Heaven is unable to kill, Earth ty8is unable to smother, the shadowy and the sunny are unable to complete. The gentleman (jnz) who grasps this is one whom we call a {sagacious person} . . .

The final sentence that mentions the gentleman reads as follows:

()()[] The gentleman (jnz ) who grasps this is one whom we call a [].

Unfortunately, strip ty8 breaks off exactly at this juncture. However, we can still reconstruct that the Ti y shng shu is about to provide
strip. But since many bamboo strips are fragmentary, we cannot say with certainty whether such markings also divided the different textual units. 17 Note that a negative outcome of this test would not per se allow the reverse conclusion that the Ti y shng shu was part of something like a Warring States Proto-Laoz .

applying the methodology: ti y shng shui

219

a positive definition for gentleman ( jnz ): as someone who is characterised by having grasped the essence of what has been outlined in the cosmogony of the text. After this passage, that is, after having provided a definition for gentleman (jnz ) as someone who has grasped the essence of the cosmogony, the Ti y shng shu no longer speaks of a gentleman. Instead, the subsequent passages only speak of the sagacious person (shng rn ). According to the Ti y shng shu , the sagacious person is someone who carries out his service according to the do . I come back to this below. Because a crucial part of the Ti y shng shu is missing, nothing can be said about this text with absolute certainty. Despite this, by paying close attention to its compositional features, it is still possible to gain a satisfactory picture of this defective text. I am confident that these features help to explain the missing parts of the text. The characteristic formal device of composition of the Ti y shng shu is the repeated contrast of different concepts with each other in order to define their conceptual meaning. This formal device is the decisive characteristic of the text. It applies to the different levels of the composition. Assuming for the moment the unity of the Ti y shng shu which can be demonstrated only by describing the different compositional features of the textit can be shown that the concept sagacious person (shng rn ) of the second part of the text is in fact correlated to the concept gentleman ( jnz ) in canto 1 (the first part of the text). The sagacious person is described as having grasped the essence of cosmogony, which is then defined. Due to the damage to the bamboo strip, this definition of the essence of cosmogony is now lost. But if it holds true that sagacious person is indeed the qualitative definition for the gentleman, namely as someone who grasps the essence of the cosmogony as described in the text, and it is reasonable to assume that it is, then it becomes clear that the unit under review fulfils two functions. It concludes the first part of the text and, simultaneously, leads the argument over to the next part of the text. This can be described as follows. First, this passage concludes the first part of the text by summing up the concrete characteristics of the Ultimate One, the ti y . As it is stated therein, the ti y commences the process that generates the cosmos and, crucially, also imbues the cosmos with all of its characteristics. Second, the unit under review further provides the information that it is necessary for a gentleman to comprehend the cosmogonic process described andeven

220

chapter six

more importantlyappreciate the essence of the Ultimate One and what results from it. This makes a gentleman (jnz ) become a sagacious person (shng rn ). By implication, the unit articulates an implicit request addressed to the gentleman. If he fulfils this request, the gentleman can be redefined as a sagacious person. By so doing, the unit under review (building block 3) leads the argument of the text from the cosmogony to a politico-philosophical level. Seen from the postulated macro perspective, it is another instance of a doubledirected feature of a text segment in that this building block combines two parts of that text. Connecting two cantos of a consistent textas will be shown belowit is a distanced type of this feature. Accepting these preconditions for the moment, the following becomes clear. The sagacious person of canto 2 relates to the gentleman of canto 1. But sagacious person should now be read as a concept meaning a gentleman who has grasped the very nature of the Ultimate One (and, by implication, of the cosmos at large), as described in canto 1 of the text. In the same way in which the Ti y shng shu no longer uses the appellation gentleman after having provided a more substantive definition for the person who has grasped the essence of the Ultimate One and the cosmos at large (and so what results from it), the text also redefines the other important concepts. On the whole, the Ti y shng shu is all about providing conceptual definitions for the things and concepts that surround us. But it is concerned, not with proper naming, but rather with defining the actuality of these things. As such, it comes close to the Aristotelian idea of defining the what it is (to tiesti ) of a thing. In particular, it is concerned about the actuality of the Way, the do .18 The Actuality of the Way (Do) Canto 2 of the Ti y shng shu proceeds to describe the sagacious person in further detail, namely in his relation to the do . The do, it may seem, is an otherwise unspecific and highly abstract concept. It is noteworthy that the two concepts which feature so importantly in canto 1 of the Ti y shng shu are so prominently absent in canto 2

18 In this the Ti y shng shu differs greatly from what is sometimes referred to as a r-ist discourse of names and appellations. For a good discussion of the dispute of Warring States philosophers on names and concepts, see Makeham 1994.

applying the methodology: ti y shng shui

221

of thesupposedly coherenttext. These two concepts are the Ultimate One (ti y ), defined as the force commencing the process that generates the cosmos and, simultaneously, pertaining the cosmos in its entirety; and gentleman. By taking the next step in testing the argumentative nature of the Ti y shng shu as a whole, it will become apparent that the correlation of Ultimate One (ti y ) with the Way (do ) is structurally the same as that of sagacious person with gentleman. The Ti y shng shu implicitly defines these concepts through its formal structure. Ideas are correlated with one another and on different levels of the text, and so their conceptual meaning is made accessible. Canto 2 of the Ti y shng shu articulates the philosophical concern of this text. This becomes obvious from a close reading of the fourth building block (strip ty10/123). Building block 4 puts forward the idea that only by knowing the [real] name (mng ) of a thing can one grasp the essence of it. [Real] name discloses the actuality of a thing, whilst style name (z ) refers only to a mere denotation that cannot grasp the full meaning of a thing. See the following figure of building block 4:
. . . ty10 [G] [H]

. . . . y10 Below is soil, yet we call it earth; Above is vapour, yet we call it heaven. Do likewise is [only] a style name for itMay I [thus] ask for its [real] name? Figure 22a: Building Block 4

Building block 4 argues that heaven and earth are only style names of what vapour and soil (the real names) describe in their entirety. The Ti y shng shu contrasts this with the do . As it is put, do is only a style name, too.19 As such, it remains an abstract appellation that cannot grasp the essencethat is, the actual meaning of a thing. As this passage makes plain, this issue must have given many philosophers of that period a considerable headache. Despite the fact that do is only a style name, it nevertheless remains the do

19

Strip ty10/1523.

222

chapter six

according to which one should perform service.20 But only by knowing the real name of it, and thus grasping this concept in its entirety, can one endure. Hence, we have the blunt question posed by the authors of the Ti y shng shu , in which they ask for the real name of the concept, hoping to provide it with more substantial contents:
Do likewise is [only] a style name for itMay I [thus] ask for its [real] name?21

If rendered in structural terms, meaning formation in this building block works as shown in figure 22b. As shown in this figure, building block 4 contrasts real namesthe essence of the thingswith denotations that are, as can be assumed, commonly used, yet are devoid of actual meaning. By so doing, the Ti y shng shu establishes a correlation between real names and their mere denotations. Heaven, by implication, is nothing other than the style name for what vapour describes in its entirety. Real name, it turns out, is the phenomenological actuality behind an otherwise rather abstract concept. In the Ti y shng shu , the phenomenological actuality behind the concepts used is fully disclosed by means of contrasting their style name and real name with each other and so establishing a correlation between them. As I will show, this technique is the device of meaning construction of the Ti y shng shu . It
(real name) . . . ty10 (style name) (real name)

(style name)

Figure 22b: Structural Representation of Building Block 4

Strips ty10/2412/14. For the reading of () () as may I ask its name, see the seminal article by Qi (2000b).
21

20

applying the methodology: ti y shng shui

223

applies not only to building block 4 but also to the macrostructure of the text. Building block 4 can thus be taken as a hermeneutical key to the Ti y shng shu in that it reveals how to approach the text as a whole. If only looking at this passage, that is, when reading the text in a merely linear fashion and not in the context of the whole text, the question posed in building block 4 remains an open one. The phenomenological actuality behind the otherwise abstract denotation do would be left unspecified. Interpreting this particular passage as an isolated unit of thought would thus be to postulate a text devoid of meaning. However, when applying the same strategy underlying building block 4 to the Ti y shng shu as a whole, that is, reading thepostulatedmacrostructure of the text as the deliberate attempt to contrast style name and real name with each other in order to establish a correlation between name and actuality, and thus disclose the phenomenological actuality behind the various concepts of the text, it becomes clear that the Ti y shng shu does indeed provide a positive reading of the otherwise abstract concept do. The two parts of the Ti y shng shu are related on formal grounds. They mirror each other structurally. Canto 2 elaborates on the insight of the cosmogony outlined in canto 1. Just as postulated for the hierarchical argumentative line of the macrocomposition of the Qing d y sh, which I discussed in chapter 2, the two parts of the Ti y shng shu show a similar, complementary relationship (see fig. 23).22 Just as in building block 4, the Ti y shng shu as a whole contrasts real names with their style names. The macrostructure establishes a correlation between the essence of a thing and its commonly used denotation. Isolated from the actual thing, these denotations are unable to describe the thing itself. Just as heaven and earth prove to be nothing other than style names for what vapour and soil describe in their entirety (thus describing the phenomenological actuality of a thing and so presenting their real names), the Ultimate One (ti y ) is considered to be the phenomenological actuality behind the style name do.23
Cf. chapter 2, figure 12. As mentioned, the Ti y shng shu defines, not words, but the phenomenological actuality of things. In this respect, the definition of do (style name; z ) as cosmos-generating principle that simultaneously pertains cosmos in its entirety the phenomenological actuality which the text calls real name, mng does not tell us the meaning of the word do. Instead, it tells us what do is said to be with respect to itself.
23 22

224

chapter six

23a: The Macrostructure of the Ti y shng shu

Canto 2 (application) real name Canto 1 (cosmogony)

Ultimate One

style name

Do

Heaven
Soil style name

Earth
vapour

real name

23b: Microstructure of Building Block 4

(real name) (style name)

. . . ty10

(style name)

(real name)

Figure 23: The Correlation of the Two Parts of the Ti y shng shu in Comparison to Building Block 4

applying the methodology: ti y shng shui

225

Only when reading the text as a whole does it become clear that dothe style name of another concept that the text aims to grasp in a more substantial wayis in its entirety explained through the formal structure of the text. Just as gentleman and sagacious person were correlated with each other in cantos 1 and 2 to explain their conceptual meaning, building block 4 (strip ty10/110/23) also contrasts the essence of heaven and earthvapour and soilwith their abstract style names and thereby shows that the concept do is likewise nothing other than the style name of something more substantial. Then, in the same way in which the various concepts are contrastedand so correlatedwith one another on formal grounds so as to elucidate their substantial meaning and fill these concepts with more concrete contents, the Ti y shng shu defines the essence of do by relating it to the cosmogonic process pertaining to the world as a whole. Thus, when looking at the Ti y shng shu from a macroperspective, the text as a whole works in the very same way as the individual building block. In this, it is very much like the Zhng xn zh do and the Qing d y sh. But the Ti y shng shu goes one step further. The text as a whole not only mimics the structure of the individual building block as seen in the Zhng xn zh do and the Qing d y sh but is designed so that the individual building block functions as the hermeneutical key for the compositional structure as a whole. By relating the two parts of the textthat is, the cosmogony and the politico-philosophical discussionwith each other according to the same principles that also apply to the individual building block, the text further explains that the cosmogony it presents goes much further than only describing the ontological process of how the cosmos is generated. The way that it defines and interchanges different concepts also shows that the cosmogonic process described explains nothing other than the phenomenological account behind the otherwise abstract concept of the do. By implication, such as heaven and earth in their relation to vapour and soil, do is but a technical term for what underlies the whole cosmos. Meaning in the Ti y shng shu is constructed by relating different building blocks with one another. The Ti y shng shu thus works in a fundamentally different way from the imaginary contextdependent text of bundle C, of which so many scholars believe it was

226

chapter six

an inherent part. By definition, it cannot be a lost part of an imagined Proto-Laoz . Conclusion The analysis has shown that the Ti y shng shu is an argumentbased text in its own right. Regarding the Ti y shng shu as different units of thought would prevent the modern interpreter from accessing the politico-philosophical message of the text. Instead, the Ti y shng shu works as a unit spanning the entire length of fourteen bamboo strips. The attempt to interpret it as a collection of otherwise unrelated units of thought because it shares its material carrier with the context-dependent text now tentatively called C is misguided. The cosmogony outlined throughout the first eight strips of the text cannot be understood in isolation. Likewise, it is not a mere commentary or the elaborate explanation of another unit of thought of the context-dependent text transmitted as Laoz. Instead, the cosmogony of the Ti y shng shu is the necessary element for understanding the conceptual meaning of the politico-philosophical concept do as it is used in the text. It fills the concept with concrete contents. According to the Ti y shng shu , good rule should follow the principle of the do. Do, for its part, is only the style name of the Ultimate One, and the Ultimate One is what pervades the cosmos. By implication, the idea underlying the text is that ruling a state by means of the do implies nothing other than following the inherent patterns of the cosmos. Good rule thus results from a proper understanding of the cosmos. The application of the cosmogonic principles described in the first part of the text (canto 1) to good rule as provided in the second part of the text (canto 2) therefore turns the Ti y shng shu as a whole into a consistent cosmology. Once and for all, we can dismiss all kinds of attempts to interpret the Ti y shng shu as a lost part of the context-dependent text of bundle C. Instead of construing an entirely new text only through the channels of tradition, we should allow the Ti y shng shu the necessary room to be understood on its own terms, that is, from its specific argumentative structure.

CHAPTER SEVEN

WRITING MEANING: MATERIAL CONDITIONS OF MEANING CONSTRUCTION IN WARRING STATES PHILOSOPHY Having explored the distinction between two fundamentally different strategies of meaning construction and philosophising in early China, I shall now explore the material conditions for reading and writing in the Warring States period. To contextualise the material and intellectual conditions underlying the exchange of written ideas will help to deepen our understanding of the intellectual world at that time. Focal questions are: What are the material prerequisites for meaning construction in philosophical discourse? How are these factors manifested in the texts under review? How does this affect the study of early thought? Writing and Thought Based on the differentiation between the two ideal types of texts, the discussion has cast light principally on two contrasting modes of meaning construction in the philosophical world of the Warring States. Of these, the context-dependent texts were defined as those texts that rely primarily on outside information for getting their concern across. Formally, the individual unit of context-dependent texts represents the finalmaterialisedengagement with a given concern. As such, it remains situational and bears witness to an occasional response to a perceived problem. Instead of generating a self-contained philosophical position with argumentative force the way argument-based texts do, the units of context-dependent texts rely primarily on authority to advance their positions. This feature contributed to the longevity of their use, as their ambiguity provoked an ongoing need for explanatory settings. As a result, these units came to be used in different contexts and, crucially, they became steady but moveable modules. This suggests that anow lostoral discourse underlying the process of meaning construction for the individual modules of contextdependent texts should be postulated. It connected the various modules

228

chapter seven

to identified traditions and so contextualised them within a given intellectual horizon. The structure underlying the process of meaning construction of the Z y plainly shows this. The direct reference to authorities is isolated from elaborating contexts. The modules remain ambiguous on the literary level of the text, and so context-dependent texts require both a predetermined acquaintance with and identificationand consentof the cultural, that is, the group-based application of knowledge behind the stories and quotations referred to. Since these texts point outside themselves instead of generating meaning from within, they must be embedded in a certain intellectual context in order to function as platforms for a broader philosophical engagement. This wider context of meaning construction was necessarily an oral one, and this is true even for those examples which at some point were copied on bamboo. The discussion of the modules collected in bundles A, B, and C, as well as the manifestations of the Z y from Gudin One and the Shnghai collection of Chu manuscripts, has made this plain. How to read the quite enigmatic modules of contextdependent texts was largely a group-based decision. In an environment where it is a largely oraland group-based exercise to give meaning to the modules of context-dependent texts, guidance becomes necessary; this is true especially where the preexisting consent of the recipient to some kind of cultural interpretation behind the text is required. This guidance falls back on a mediator of meaning. The mediator of meaning could be envisioned as a master (or masters). Or it could be the preexisting cultural consent of defined groups, which, in turn, had to be established by some kind of masters. Conversations with a master on the basis of the text therefore constitute the necessary referential frame in which context-dependent texts became meaningful. This casts light on the dialectical processes of meaning construction between the texts and the textual communities as something imposed on the context-dependent texts from the outside. Argument-based texts, on the contrary, are defined as philosophical texts written in a continuous mode. But argument-based texts too are made up of particularly stable units that are clearly distinct from one another, even if they may construct narrative patterns of the kind seen, for instance, in the authorial layers of the Xnz. These units are the basic constituents of this type of text. Despite the characteristics shared with the modules of context-dependent texts, meaning con-

material conditions of meaning construction

229

struction in argument-based texts works by linking up the individual units with one another. Much larger meaningful entities are generated accordingly. Notions advanced at different junctures in the texts can thus be connected into greater schemesand finally into a coherent whole. Cross-referential links between different textual units enable conceptual definitions of otherwise idiosyncratic notions to be established. Therefore, unlike the module in context-dependent texts, the units of argument-based texts are not isolated entities but building blocks of a bigger picture. By fusing them into integrated wholes, the texts advance weblike structures that make possible referential and, as a result, systematic development of a philosophical concern. This facilitates self-contained types of reasoning. By weaving the individual building blocks into larger wholes, these texts generate an additional meaningful level for advancing their philosophical agenda. Accordingly, these webs clearly fulfil a semiotic function. They broaden the lexicon and syntax of a text by providing conceptual definitions for the various concepts used in them, and so they open up a meaningful level behind the verbatim content of the individual units of the texts. I refer to them as the semiotic webs of the argument-based texts. Reference versus Self-Reference As I have argued, the semiotic webs of argument-based texts facilitate the systematic discussion of philosophical concerns. Connecting the different notions advanced in these texts into greater schemes, they facilitate precise definitions of the conceptual meaning of terms, even when used in a rather idiosyncratic way. Notions introduced at one point in the text are elaborated by other textual units, and persuasive definitions established. The triangular relationship between text, mediator, and recipient that characterises the conveyance of meaning in context-dependent texts is therefore no longer the required structure underlying the process of meaning construction. Argument-based texts relocate the intellectual effort from the oral contexts to the given text as materialised on bamboo. Here one should think of the means by which a text such as, for instance, the Zhng xn zh do plays with certain ideas and concepts, such as the immanent nature of the real gentleman, jnz, by correlating this concept repeatedly and on different levels of the text with certain characteristics of the natural

230

chapter seven

world. In the same fashion, the authors of the Zhng xn zh do have defined the concepts zhng fidelity and xn trustworthiness in a quite distinctive way. Through their semiotic webs, argument-based texts become selfcontained and, therefore, autonomous philosophical units. Referentially consistent, they turn into structurally closed compositions and stand-alone philosophical edifices. The Ti y shng shu , as discussed in chapter 6, may serve as an example. The consistent links and references advanced in the text allow it as a whole to turn into a consistent cosmology because the compositional structure of the text links the ontological discussion about the steady process of cosmos generation directly to the political sphere, by which the text finally advances a concrete directive for proper rule. The semiotic webs of argumentbased texts thus prove philosophically relevant as modes of meaning construction. They are the prerequisites for advancing a stand-alone philosophical position with argumentative force. Because in argument-based texts the different notions advanced in the individual building blocks are connected into greater evocative schemes, these texts become coherent and finally, on the level of the composition, complete (and completed) wholes. To a different degree, these texts no longer require furtheroralcontextualisation. The form analysis has shown furthermore that, structurally, texts such as the Zhng xn zh do and the Qing d y sh are fully fixed. They can exist only in their given form. It is reasonable to assume proactive authors behind their makeup, who produced coherent systematic philosophical arguments. Argument-based texts, it can be concluded, are the self-contained products of deliberate philosophising. Structure and Meaning Remaining outside the given texts, the elaborating structures of context-dependent texts are what I shall call silent structures. That means they are inexplicit and so culture-dependent because the construction of meaning relies on masters or a preexisting cultural consent within specific textual communities. The semiotic webs of argument-based texts, in contrast, are voicedor explicitstructures, as they generate meaning from within the text by advancing meaningful compositional patterns. Explicitlyin the formal structure of the textsargumentbased texts achieve what may reasonably be called systematic dis-

material conditions of meaning construction

231

cussions of philosophical concerns. Without the need for cultural contextualisation from the outside, texts of this type facilitate the individuals engagement with thought. Since the formal patterns of argument-based texts are the modes by which meaning is constructed, they make argument-based texts meaningful objects in and of themselves. As a result, the philosophical actuality no longer lies in the triangular relationship between master, student, and text which determined the successful communication of thought in context-dependent texts. Instead, reflecting the structure of the argument, reading argument-based texts becomes itself a philosophical performance. As a result, the philosophical idea becomes accessible to whoever had access to these texts and was able to read. Since the compositional structure of some of these texts constitutes a formal device that, as argued, compounds the thought central to these texts,1 it may be argued that the philosophical reality, within some intellectual traditions, even lies in the texts themselves.2 These observations have vital implications for the study of early thought. On the one hand, argument-based texts represent structurally closed entities. The relevant references are established within the texts themselves, by which they become direct mediatorspossibly even manifestationsof the argument. But this is not to say that argument-based texts eliminate different readings and varying interpretations. No text can ever be definite in the interpretations it facilitates.3
1 Texts such as the Zhng xn zh do, Qing d y sh, and Wu xng may serve as examples. 2 It may be noted in this context that the European history of ideas witnessed a hiatus between the seventeenth and eighteenth centuries, on the one hand, and the eighteenth and nineteenth centuries, on the other, in which attempts to explain the world on the basis of the text, that is, the Bible, shifted to attempts to explain the text on the basis of the world. Scriptural religions such as Judaism and Protestantism take scripture itself to contain the truth (see, e.g., Psalm 119), and something similar might also be the case for argument-based texts. The Renaissance thinker Pico believed that the structure of texts contained the structure of truth (see Farmer 1998, 34), and at least within some intellectual traditions of the Warring States, argument-based texts may similarly have been taken as an embodiment of the philosophical truth. These texts would therefore represent the unity of practice and thought and, as a result, facilitate the neutralisation of the distinction between philosophy and its performance. 3 Even though the argument-based texts establish semiotic webs that in and of themselves guide their reading, they nevertheless still allow certain degrees of difference as to how certain aspects are to be taken. As for these texts, one might not want to go as far as the German writer and experimental physicist Georg Christoph Lichtenberg (17421799) in his famous dictum on the variety of interpretations as advanced in his Sudelbcher (Waste Books): Ein Buch ist ein Spiegel, wenn ein Affe hinein sieht,

232

chapter seven

Nonetheless, they represent premediated, and philosophically coherent, concepts which, usually, require no third mediator besides text and exegete. Context-dependent texts, in contrast, were generatedand meaningfulagainst the background of oral dialogue. The oral discourse underlying context-dependent texts not only constitutes a vital element of the individual unit of thought but also connects the various ideas to certain traditions and, by implication, to integrated, culturally meaningful wholes. It is imperative to keep this difference in mind when dealing with philosophical texts from the Warring States because of its implications for reconstructing early thought. Whereas it is possible to reconstruct something like a philosophical edifice behind argument-based texts like Zhng xn zh do, Qing d y sh, Wu xng, Xng z mng ch, and Ti y shng shu , no such claim can be made when engaging with context-dependent texts such as, for instance, an imagined Warring States Laoz . The connection with authorities of whatever kind guaranteed the continuing importance of these texts and ensured their transmission. This might have applied to some of the argument-based texts as well, before they, for reasons discussed below, nonetheless slipped entirely from view for nearly twenty-five hundred years. But unlike the contained argument-based texts, the modules of context-dependent texts were easily adapted to different settings and interpretations. It follows that the contexts in which they were used changed over time, just as is true of the textual communities. Different textual communities decide on the best interpretation of these textual units: for their purposes and according to their reading of Odes, Documents, or any other authority linked to these modules. Since different textual communities could come up with different readings, the authority shifted gradually from textual authorityauthorities mentioned in the texts, as exemplified by formulae such as z yu master(s) say or sh yn in an ode it is saidto that of different masters/groups and their interpretations. It shifted from the texts to the textual communities. In other words, because they generate new contexts, the modules of context-dependent texts move among textual communities and interpretations. This

so kann kein Apostel heraus gucken (A book is like a mirror, when an ape peeks in, no apostle can peep out). Individual readings must, nonetheless, also be kept in mind when discussing written thought.

material conditions of meaning construction

233

renders the reconstruction of a Gegentext to these modules impossible. The productive environment against which these texts were produced is lost for good. (Re)constructing Early Thought The foregoing analysis has made clear that the concept of a Laoz from the Warring States is in fact compounded in a number of unrelated modules, but no coherent, let alone fixed, text. Even if there was an actual text known as Laoz during the Warring Statesjust as there was one during the Hn, as archaeological records suggest4the triangular relationship of meaning conveyance that determined the successful communication of thought in texts of this type nevertheless was still irretrievably lost. Because of changing contexts and different textual communities, the reconstruction of a Gegentext to these enigmatic, isolated, and ambiguous modules is, as argued, unattainable. Any attempt to (re)construct an imagined Warring States Laoz therefore lacks a coherent and well-defined text on the basis of which such an analysis could be carried out. Even more importantly, the referential system behind the makeup of such an imagined text remains irretrievably lost. It follows that even if one were to regard the unrelated materials from bundles A, B, and C as (one) particular instantiation of a Warring States Laoz , the philosophical system behind its makeup would still be missing. Since the philosophical concerns of context-dependent texts were expressed, not in the texts themselves, but in the oral discourse surrounding their various modules, they can be thought of only in highly tentative and hypothetical terms. This dilemma of text and thought in context-dependent texts is opposite that of text and meaning in the Odes of the Warring States period. Quotations in text fragments from excavated manuscripts give us a glimpse of the different interpretations of the Odes that were circulating during the Warring States period before they were finally ousted by thenow predominantset of interpretations as determined by the Mo tradition. But apart from the bits and pieces available from different excavated texts, the text of the Odes circulating during the Warring States is largely missing. It remains open

I have in mind the copy of the Laoz from Mawngdu Three.

234

chapter seven

to interpretation, and so every single quotation from the Odes is in fact a reinvention of its text. For the context-dependent texts from the Warring States the opposite is true. Different excavated manuscripts provide the lexicon of a well-defined Z y; and others might even want to consider the materials in bundles A, B, and C as a Warring States instantiation of Laoz . But the cultural context and the set of philosophical interpretations against which these texts were meaningful to the textual communities of those days are irrecoverable. Any history of thought has to take this into account. To present the ideas of, say, Laoz , can mean only to provide an idiosyncratic interpretation of the bits and pieces we have, but, nota bene, not their referential framework. In a similar vein, a correct, let alone complete, picture of the philosophy of Confucius can only be guesswork. One may possibly reconstruct a reading of Kongz as exemplified in the Mngz. The so-called Hshng gng 5 interpretation of Laoz that of Wng B (226249)has already been successfully reconstructed.6 But any attempts to reconstruct coherent, let alone closed, systems of thought of those intellectual traditions that are transmitted only in context-dependent texts are methodologically ultimately illfounded. Context-dependent texts cannot be read in the fashion of the contained argument-based texts. Such an approach would neglect the thirdthat is, the oral (and vital)component of meaning construction in context-dependent traditions; and in the case of a so-called Laoz there even is no such text. In contrast, the semiotic webs of argument-based texts that turn these texts into structurally closed entities have largely replaced the secondary contextualisation from the outside necessary to engage with their ideas. These texts render coherent systems that enable the individual to engage with their ideas. Able to be read, they can be approached and their philosophical message unlocked. As argued, it can be safely assumed that argument-based texts were already detached from oral contexts at an early point in time and circulated independently in writing.

5 According to Wagner, the Hshng gng Commentary was written in the fifth century. See Wagner 2003a, 15. 6 For a successful approach in making explicit one particular reading of the Laoz, see Wagners trilogy on Wng Bs commentary on and reading of the Laoz: Wagner 2000, 2003a, 2003b.

material conditions of meaning construction Travelling Concepts and the Fusion of Ideas

235

The voicedor explicitstructures of the argument-based texts and, accordingly, their independence from local contexts link these texts closely to writing. Another sign of written discourse as manifested in the texts of this type is their syncretic approach and the fusion of concepts they show. Syncretic in this respect has no normativelet alone teleologicalconnotation. It is meant purely as a descriptive concept for the tendency in those texts to reconcile a multitude of ideas and traditions and incorporate them in a new philosophical entity. The fact that argument-based texts from Gudin One represent a written mode of reasoning that has produced structurally closed and self-contained entities does not mean that the representatives of this type of text do not also incorporate alien sources in their attempt at argument construction, which may even be of oral origin. Orality and literacy are no clear-cut matters. Early texts are synchronic artefacts and include a wide range of different sources and traditions. The philosophical texts were part of a larger intertextuality. Without reducing these texts to a mosaic of quotations,7 it becomes plain that they were produced against some kind of productive environment that clearly left its stamp on them. Authors of philosophical texts knew about all kinds of travelling concepts8 that were around at that timeorally passed from person to person or, in fact, in writingwhich they could and did refer to. But the way in which alien concepts were utilised in the contained argument-based texts proves fundamentally different from how they were employed in context-dependent texts. Context-dependent texts did not explicate the quotations used. The philosophical enterprise was not made part of the text as materialised on bamboo. Argument-based texts, in contrast, grafted the philosophic discourse from the source referred to into the quoting text itself. Quotations in these texts no longer only refer to contexts as they do in context-dependent texts. Argument-based texts construct their contexts and so fix them in a certain way. Argument-based texts detach the quoted statements from their original contexts, explicate them, and integrate them into their own argument. The Qing d y sh can serve as an example. What could be defined as the second stable subcanto of this

7 8

Kristeva 1980, 66. Bal 2001.

236

chapter seven

text9 is in fact one coherent module of references to travelling concepts taken from the pool of shared memory of contemporary lite groups. Different stories about humble worthies meeting the enlightened ruler so that they can act in the world are presented in a highly structured way, and so the stories are made to fit the overall tone of the text. The formal perfection of the obviously modified account adds to the credibility of the stories themselves. Unlike the modules of contextdependent texts, these stories are not isolated. The Qing d y sh integrates them meaningfully into the texts philosophical framework. Taken together, they become a stable component in the overall argument of the Qing d y sh. A particular group-based reading of these instances plays no role in the communication of the message of the text. A similarly straightforward use of quotations can be found in the long, multi layered, and decidedly complicated, yet largely consistent, programme of self-cultivation of the Wu xng . The Wu xng quotes abundantly from the Odes. Various techniques used in the text indicate that it refers to a widely known source.10 But unlike the use of quotations in context-dependent texts, such as, for instance, the Warring States manifestations of the Z y, the discussion of the Odes takes place within the argument-based text itself, and not in an oral discourse surrounding the text that must be assumed but that cannot be reconstructed with certainty. All the necessary references to how a quotation is to be taken are given in the argument-based text itself. The quotation itself therefore does not implant a group-defined interpretation of the Odes into the quoting text. On the contrary, the argument-based text explicitly establishes one particular interpretation of the quoted source within itself, and so the reference to widely known sources takes the opposite direction to that of the context-dependent texts. Whether stock phrases, technical terminology, or whole units of thought, argument-based texts integrate the quotation into the argument and so provide a particular reading for it. Notice in this context that not only does the Wu xng comment on the lines from the anthology called Odes, but the quotations themselves are also a vital part of the texts strategy of argument construc-

See my discussion in chapter 2. Whereas the Mawngdu Three version of the text explicitly introduces Odes, the Gudin One manifestation uses formulae such as Now this is what this is about []() [] (strip w11) or () (strip w30).
10

material conditions of meaning construction

237

tion. A brief look at strips w912 may help to clarify this. This unit dwells on the mutual relationship of the virtues benevolence (rn ) with wisdom (zh ) and that of benevolence (rn ) with sagacity (shng ):
w10 w11 [] [] [] If not benevolent, one cannot be clear in ones thinking. If not wise, one cannot grow in ones thinking. If neither benevolent nor wise, whilst not yet having seen a gentleman, the sorrowful heart w10 cannot be disturbed. Until [I] have seen the gentleman, [my] heart cannot be delighted.11 [But] when I have seen him, and when I have met him, then my heart will w11{X be delighted}. 12 {X Now this is} what this is about. [] w12 {If not} benevolent, one cannot be clear in ones thinking. If not sagacious, one cannot be effortless in ones thinking. If neither benevolent nor sagacious, w12 whilst not yet having seen a gentleman, the sorrowful heart cannot be agitated. Until [I] have seen the gentleman, [my] heart cannot be stilled.13

Just as was seen in the Qing d y sh, the concepts used are torn from their original context. The references are transformed by a particular set of interpretations, and they are fused into a new framework, namely into that of the argument being advanced. By integrating the concepts referred to into a new train of thought, the argument-based text clarifies its idea. But at the same time, it also provides a particular interpretation for the very concept used, simply because it is torn out

11 12 13

Adapted from Odes (Mo 14). Adapted from Odes (Mo 14). Adapted from Odes (Mo 168).

238

chapter seven

of context, correlated with other notions, systematised and integrated into the new argument as advanced in the text. Unlike the function of references to authority in context-dependent texts, the fusion of contexts as processed in argument-based texts removes the ideas referred to from their original contexts and so creates a whole new setting for the ideas advanced. What can be seen from this is the intercultural corroboration of an observation made by Eric Havelock in the early 1960s. In his attempt to explain the emergence of abstract philosophy in ancient Greece, Havelock observed the impact of writing on philosophy at large. According to Havelock, writing leads to syncretic tendencies. Unlike what was seen in context-dependent texts, where quotations simply refer to group-based traditions that remain outside the written text, writing facilitates the fusion of traditions and concepts and so leads to the systematisation of ideas. Early philosophy, in Havelocks view, made abundant use of this and was shaped by syncretic tendencies. The use of quotations in these texts should be understood as an attempt to unveil the truth hidden in widely known concepts. Thus, in his Preface to Plato Havelock stated:
The saga [i.e., the Iliad] will contain a thousand aphorisms and instances which describe what a proper and moral person is doing. But they have to be torn out of context, correlated, systematised, unified and harmonised to provide a formula for rightness. The many acts and events must somehow give way and dissolve into a single identity.14

The emergence of abstract philosophy in ancient Greece would therefore be dueat least in partto what was exegetically wrung out of the mythopoetic language of Homer.15 The intellectual leadership of early Greece revolted against the immemorial habit of self-identification with the poem, and only after the spell of the poetic tradition has already been broken did the poem become the abstracted object of knowledge.16 By destroying the original syntax of the poem17 it became a systematised encyclopaedia, unseen and abstract.18 To transform the saga into an abstract source of knowledge, aphorisms had to be torn out of context, correlated, systematised, unified and
14 15 16 17 18

Havelock 1963, 218. Quoting Havelock from Farmer 1998, 78, n. 50. Havelock 1963, 216, 219. Ibid., 218. Ibid., 217.

material conditions of meaning construction

239

harmonised to provide more abstract and universally valid formulae, a process ultimately linked to writing.19 Removing travelling concepts from original contexts and fusing them systematically into new settings so as to elaborate and explain these notions applies not only to the Greek case. Havelocks description also seems to be consistent with what can be seen from the argument-based texts of the midto late Warring States period. Bernhard Karlgren has made similar observations in his study of systematic thought in Hn China. As with Havelocks analysis of early Greek ideas, Karlgren understands the intellectual products of the Hn dynasty systematisers as being ultimately worked up from early legends and myths stemming from the age of the Zhu.20 In both cases, these processes are closely linked to writing. Writing made possible the systematising fusion of mythopoetic concepts into abstract ideas, and it facilitated the layered organisation of thought. Steve Farmer, together with John B. Henderson and Michael Witzel, has cast further light on the complex correlation between early writing and abstract thought by exploring the way the systematisation of ideas appeared in the different cultural centres of the ancient world (China, Greece, India, and the Near East). In line with Havelock and Karlgren, their analysis suggests that abstract ideas grew out of exegetical processes in the engagement with mythopoetic concepts central to those societies. Exegetical tendencies and highly referential modes of processing ideas are typical of many cultural centres in the second half of the first millennium BC. They are simultaneously signs of writing and its actual result. The Materiality of Meaning Construction The concerted efforts of Farmer, Henderson, and Witzel have deepened the study of the correlation between (early) writing and abstract thought in early civilisations around the world.21 They highlight the remarkable parallels in the appearance of philosophy and cosmology in the main cultural centres of the ancient world. In what they

Ibid., 218. See Karlgren 1946, 1968. See also Farmer 1998, 7879, n. 51. 21 See Farmer 1998, 2006; Henderson 1984, 1991, 1998; Witzel 1979, 1997, 1998; Farmer, Henderson, and Witzel 2002; Farmer, Henderson, and Robinson 2002.
20

19

240

chapter seven

call a cross-cultural framework (or cross-cultural model),22 they link these developments to exegetical processes of exactly the kind described above. As Farmer and colleagues argue, these processes were fostered by the first widespread use of lightweight writing materials, and the subsequent development of stratified textual traditions, that began simultaneously in all advanced world cultures in this period.23 Writing allowed the fusion of concepts and the highly layered organisation of thought. Just like the exegetical efforts characteristic of the Wu xng, the highly contextualised and convoluted modes of thought processing that can be witnessed in argument-based texts are both a sign of writingfrom the perspective of the modern interpreteras well as its actual result. They enable complex ideas in more intricate trains of thought to be expressed. The syncretic syntheses of travelling concepts ultimately resulted in the emergence of highly layered texts. This facilitated more sophisticated systems of thought. As has been argued, the beginning of this process was the endeavour to comment on textual authority.24 The repeated effort to harmonise widely known sources led to ever more correlative visions of reality.25 Given the structural differences in the ways quotations are used in the two ideal types of texts, the bulk of these efforts in the Chinese context can be seen in the argument-based texts. This accords with the observation made by Farmer and his colleagues that the links between the development of systematic ideas and syncretic processes is suggested by the fact that similarly structured systems emerged in China, India, the Middle East, Europe, and Mesoamerica whenever information flows increased and tendencies to harmonise traditions reached extremes.26 In their view, and I subscribe to it, the origins of abstract thought thereby lay not so much in literacy, as held, for instance, by Goodylet alone in the introduction of the alphabet, as suggested by Havelock27but rather in the broad diffusion of light writing materials,

22 See, e.g., Farmer, Henderson, and Witzel 2002; Farmer, Henderson, and Robinson 2002. 23 Farmer, 1998, 7879. 24 Ibid., 29. 25 Ibid. 26 Farmer, Henderson, and Witzel 2002, 51. See also Farmer, Sproat, and Witzel 2004. 27 Such determinist accounts are now generally discredited, as W. Johnson (2009, 3) notes. For a critique of what Brian Street (1984, 19ff.) called autonomous model, that is, the theory that the alphabetical system itself deeply affects society and

material conditions of meaning construction

241

whether bamboo strips in China, palm leaves in India, parchment or papyrus in Greece, which facilitated the more systematic integration of previously unrelated oral and written traditions.28 This process should be dated roughly to the second half of the first millennium BC,29 the approximate date of the argument-based texts from Gudin One. The fact that we lack earlier bamboo-strip texts, however, does not in itself prove that bamboo was not already used as a writing material at an earlier date. Quite the contrary, we can be rather certain that bamboo was used as a carrier for writing long before the Warring States period and thearchaeologicalappearance of argument-based texts. The question, however, is to what extent lightweight writing materials, such as bamboo and wood, were used before the wide appearance of written texts in the Warring States period. Already for the so-called nyng period,30 a fully fledged writing system can be attested that seems to have contained a repertoire of about four to five thousand characters. Robert Bagley assumes that such an elaborate writing system could not have performed well without lexical lists.31 However, no such lists or any mention of something similar have survived to the present day. Bagley suggests that this does not rule out the existence of these lists but should instead be interpreted as indicating that they were written on materialsbamboo? that did not survive.32 This is, of course, highly speculative, although perhaps not groundless. But even if there were such lists (which I very much doubt), despite the fact that none of them have survived to the present day, why should we assume that they had to be fixed on a carrier such as bamboo or wood? Given that clay, if baked, is extremely durable and the advanced use of clay in the Chinese context (e.g., for
culture, see Street 1984, 4465; Thomas 1992, 1528; Olson 1994, 120, 3644; as well as W. Johnson 2003, 1013. 28 Farmer 1998, 79, n. 52. For Goodys hypothesis, see Goody 1977. For Havelocks ideas, see Havelock 1963. On the use of palm leaves in India, see esp. Al Azharia Jahn 2006. 29 See Farmer, Henderson, and Witzel 2002, 56. 30 Y n , the last capital of the Shng period, was located in modern nyng, Hnn Province. The nyng period roughly covers the last two centuries of the second millennium BC. 31 See Bagley 2004, 222. Note that the writing system in Egypt did in fact perform well without such lists. The fact that the Sumerians did have lexical lists is probably more closely related to the fact that Sumerian was already a dying language at the time when those lists started to appear. See Van De Mierop 2004, 31. See also Krispijn 1992; Selz 1998. I thank Wolfgang Behr for alerting me to this debate. 32 Bagley 2004, 222.

242

chapter seven

moulds in the production of bronzes),33 it would not seem all that strange to assume that lexical lists might have existed also on materials other than bamboo, such as clay. Given all this, I feel inclined to call into question the existence of such lists in early China. Another possible indication that bamboo was already used as a material carrier for writing in the earliest periods of the Chinese script is based on the fact that inscriptions on bronze and bone sometimes imitate brush writing. This hints at some kind of primacy of the latter.34 The early existence of the writing brush and lampblack ink further supports this assumption.35 Indeed, it is sometimes suggested that bronze inscriptions were copies of what most likely was first written on bamboo or wood.36 The character c (and the allographs and ) to announce, to recite (the charge or bestowal )37 further corroborates these assumptions, as it appears in inscriptions that date back as far as Shng times (ca. 16001100 BC). The character c is thought to represent bamboo strips bound together with a string into one bundle.38 The fact that the word c (OC *[tsh]rek) probably was cognate with j to pile up; accumulate (OC *[ts]ek)Laurent Sagart suggests that the medial *-r- indicated an object with repetitive structures39might further indicate that any piled up object could have been used as a carrier for writing. In sum, there is much reason to assume that bamboo strips served as a carrier for script long before the second half of the first millennium BC. But even if bamboo strips already served as a writing carrier long before the Warring States period, and even if the material had some kind of primacy over bronze inscriptions, this does not imply

33 For the technological aspects of the casting of bronze vessels, see Shaughnessy 1991, ch. 2. 34 See Bagley 2004, 218. 35 Even for the late Shng period, we have evidence of writing characters on smooth surfaces such as jade with a brush. See Kaogu xubo 1981, 504; Bagley 1999; Boltz 1999b, 108. For the assumption that some oracle-bone inscriptions were written with a brush before they were incised, see Keightley 1985, 4647. The graphic structure of the character y writing brush further corroborates the assumption that the writing brush was used in Shng times. 36 See, e.g., Falkenhausen 1993, 163164. 37 The character c is conventionally rendered to write down (on bamboo strips); see, e.g., Schssler 1987, 2007. Kern (2007b, 152ff.) has, however, convincingly argued against this interpretation. 38 See also Chavannnes 1905. 39 See Sagart 1999, 214.

material conditions of meaning construction

243

that bamboo strips (or wood) were used extensively for this purpose before the Warring States period. I prefer to think that, although the technology of writing and manuscript production is certainly old, the extensive use of light writing materials was a mid- to late Warring States innovation. This innovation fostered the spread of a manuscript culture that can be assumed with relative certainty to have existed during that period. Different excavated manuscript finds indicate this. The analysis suggests that the extent of literacy that is sometimes assumed for the Western Zhu (ca. 1099/56771 BC),40 or even earlier, is much exaggerated.41 None of the indicators of a widespread manuscript culture, such as elaborate systems of correspondences and syncretic systems of ideas, can be traced to before the second half of the first millennium BCand this holds true not only for China but also for Greece and India. The widespread use of light writing materials, it is argued, fostered the systematic development of these default conditions.42 Therefore, although the textual component was necessary for the systematic development of ideas and concepts to be processed in a nonlinear fashion, it was, as archaeological data indicate, certainly not sufficient.43 The disadvantage of a model such as that laid out here is that to some extent it has to rely on assumptions rather than on concrete facts. The positive aspect is that these assumptions about the material conditions of roughly the second half of the first millennium BC are subject to disproof by new finds. As far as I am awareor as I interpret the different accountsno such evidence has been found. As a result, I feel justified in regarding the framework developed initially by Karlgren and Havelock, and much furthered by Farmer and his colleagues from different fields, as a plausible background to my conclusions concerning the written nature of the late fourth century BC argument-based texts from the Warring States tomb Gudin One.

For the dating, see Shaughnessy 1991, xix. See Farmer, Henderson, and Witzel 2002, 79. 42 Ibid., 64. 43 Space forbids me to discuss the impact of constant political and social reshapings on the development of cognitive competence and abstract thinking. For a discussion, see Otto 1947; Kirk and Raven 1957; Humphreys 1983; Naddaf 2005; among others; all of which suggesting that abstract reflections crucially result from societal changes, not the other way round.
41

40

244

chapter seven Conclusion

Similar to the distinction between argument-based and contextdependent texts, which I have drawn to highlight the different strategies of meaning construction, the distinction between oral and written texts is in no way absolute. Texts are fundamentally synchronic artefacts and may appear in hybrid form. Ancient oral texts come to us in writing. Otherwise, of course, we would have no evidence of their existence. A written text may contain oral layers through the integration of oral stock phrases, the use of lines from the Odeswhich, during the mid-to late Warring States period, were still unwritten as a corpusand reference to concepts that appear in other philosophical contexts. The analysis of the argument-based texts from Gudin One suggests that these texts were not creations ex nihilo. Instead, their authors clearly worked from a mine of materials, from which they used stock phrases, ideas, and terminology. These materials may have been available to them in written form as well as orally in the form of memorised concepts and building blocks. But unlike the texts that remain bound to the triangular relationship of meaning conveyance, the way in which argument-based texts combine these oral elements productively into a new setting to fix their meaning and advance a new idea is fundamentally a written one, so that, in the end, texts of this type become independent entities. Writing this kind of stand-alone philosophical text was facilitated by the widespread use of light writing materials such as bamboo. Even though the technology of writing is much older, texts of this type coincide with the appearance of an emerging manuscript culture during the Warring States period. They are as much the result of a rising manuscript culture as they are facilitators of it.

CHAPTER EIGHT

CONCLUSION: WRITING PHILOSOPHY Excavated philosophical materials from the Warring States period provide insights into the complex relation between texts, their material carriers, textual communities, and the intellectual activities of early China. The trigger for this study was the rich potential of these texts for gaining a better understanding of the social practice of reading and writing in the intellectual world of the Warring States. Based on a close reading of the textual materials from Gudin One, I have distinguished two ideal types of meaning construction in a diverse corpus of texts, which, for ease of argument, I have arranged into argument-based and context-dependent ones. This differentiation is not an absolute one but methodologically necessary to order the complexity of reality by highlighting the common characteristics of diverse materials. Traditional analysis customarily sees texts from the distant past as mere vessels of thought. The study of ancient Chinese philosophy is in general characterised by an overt bias for literary evidence and a relative neglect of the much more complex intellectual, as well as material, reality of meaning construction and philosophising. This has led to some serious misrepresentations of ancient realities. Artificial boundaries were imposed between textual traditions and philosophical affiliations that were in fact far more fluid than the exclusive focus on literary evidence from selected passages might suggest. Instead of following the traditional approach, which so often still attempts to retrieve an imagined urtext or hypothesises about the intellectual origins or the source texts of later textual recensions whenever a new text comes to light from supervised excavation sites (or, increasingly, by roundabout and illegal ways from tomb looters), I have proposed shifting the focus of analysis. Rather than giving priority to the ideas expressed in texts and so, implicitly, reducing the texts from the distant past to mere vessels of thought, I like to think of texts as meaningful objects. In so doing I suggest that the text when studied as a meaningful object can reveal vital information about the text as a cultural phenomenon, including the production, use, and function

246

chapter eight

of philosophical texts in China more than two thousand years ago. Hence, instead of studying what I think is a methodologically problematic, idealised world of thought, I have proposed shifting the focus to analysing the practice of philosophising in the ancient world as it can be judged from material evidence. The Production of Philosophy One can easily talk about the production of philosophy when examining a shift that occurred when philosophical texts of ancient China no longer constructed meaning in predominantly oral ways. I do not mean that text production in early China was largely a by-product of economic activity in the sense that literary superstructures are understood as essentially an effect of the developing material base.1 Instead, I have thought about material conditions in a more narrow sense. As argued in this book, I hold that with the widespread use of light writing materials during the second half of the first millennium BC in China a shift took place in philosophical productivity that enabled new forms of philosophical enterprise to occur. This shift in writing and thought was partly dependent on a rising manuscript culture, made possible in China by the first widespread use of bamboo as a writing carrier. Before the spread of light writing materials, texts, including those with a philosophical orientation, were largely part of a wider oral performance. They recorded dialogues or imagined teaching scenarios and were generally so ambiguous that further contextualisation was required to access their meaning. But this contextualisationwe may call this the illocutionary force of the textwas nowhere provided in the recorded text itself. A lost oral discourse underlying the process of meaning construction for these texts must be postulated. This oral discourse connected the various modules to identified traditions and so contextualised them within a given intellectual horizon. Texts of this type functioned predominantly as platforms for broader philosophical engagement and were only fragments of the bigger philosophical picture. As a result, the texts remained essentially bound to a triangular relationship of meaning construction between the text, the mediator of meaning, and the recipient. I have called these forms of texts contextdependent.
1

For this position, see, e.g., Caudwell 1946, 201, 130.

conclusion: writing philosophy

247

With the widespread use of bamboo as a light and readily available writing material in China there appeared forms of philosophical productivity in writing that were characterised by webs of cross-referential elaborations. Notions introduced at one point in the text were explicated by other textual units, and so these new forms of written communication were able to establish conceptual definitions of textspecific ideas. For the first time in the history of those territories which we today call China, certain texts made explicit the illocutionary force of the text. I have called these new forms of written philosophical productivity argument-based texts. Texts of this type became standalone entities that made possible the systematic discussion of philosophical ideas. Unlike context-dependent texts, they were independent of locally based interpretations. These texts were both the sign of a manuscript culture and its immediate result. However, as discussed in this study, the use of bamboo as a material carrier for writing in China can be traced to the earliest layers of written communication and so predates these new forms of philosophical production. There is evidence that some oracle-bone inscriptions were brush-written before they were incised,2 and it is sometimes suggested that bronze inscriptions were only the copies of what was first written out on bamboo or wood.3 Mere technological availability therefore was not the sole factor in the development of these new forms of text. However, as I have argued in this study of text and meaning construction in early Chinese philosophical discourse, at that time bamboo was quite certainly not used extensively, and writing for whatever purpose remained an activity restricted to rather small and narrowly defined groups, as the oracle-bone and bronze inscriptions make plain. Writing remained an aide-mmoire of what a particular textual community wished to remember, and the texts capture only the gist of a larger ritual activity.4 Prior to the second half of the first millennium BC when a manuscript culture began to develop, there existed no extensive written literature in China with stand-alone texts. This picture is given further support by a careful evaluation of excavated texts, especially when contrasting the textual history of the Odes with that of the new forms of philosophical productivity at

See Keightley 1985, 4647. As, e.g., held by Falkenhausen 1993, 163164. 4 For the function of bronze inscriptions in ritual, see Kern 2009, 152ff. See also Kern 2007b.
3

248

chapter eight

that time. As manuscript finds show, during the Warring States period philosophical texts circulated in different manifestations. This suggests a relatively wide diffusion of manuscripts at that time. Despite this, the philosophical texts are not quoted in other texts. There exists no single quotation in which one philosophical text refers verbatim to another one, to indicate either agreement or disagreement with its intellectual position. This does not mean that different philosophical texts did not present opposing positions. Quite the contrary. Also, as I have shown in this study, examples abound in which the intellectual environment in which these texts were produced can be reconstructed with some certainty, and it becomes clear that the philosophical texts from early China were in fact part of a larger Warring States period intertextuality in which travelling concepts informed a multitude of ideas. Concepts are borrowed and given new readings. The Odes from the Warring States render the reverse picture. To date, no single manuscript has been found that contains Odes alone. They exist only in quotations. Given the wide use of Odes in pre-imperial China and the importance the anthology enjoyed as both an encyclopaedia of knowledge and the lingua franca of contemporary diplomacy, this is remarkable. Verbatim quotations of Odes in philosophical texts abound. To a large degree, they are used to corroborate a certain philosophical position by lending it authority. At the same time, many texts present their own reading of the passage quoted. Hence, there exists a variety of readings of selected passages of the Odes in the Warring States period that reflect the different interpretations of this anthology at that time. Comparison of the various quotations of Odes in exhumed texts has furthermore shown that, phonologically, they prove comparatively stable. The lexicon of the cited text is, however, surprisingly loose. No single quotation is entirely uniform. Significantly, in the Warring States period, the text of the Odes remains an object of interpretation, such that every single quotation of the Odes is a reinvention of its text. As demonstrated by Martin Kern, we are informed about various readings of the Odes during the Warring States period without having their actual text.5 This phenomenon points to the oral nature of this anthology during the Warring States, and it strongly suggests that this anthology of cultural heritage and encyclopaedia of knowledge was never fixed

See the discussion in Kern 2007a.

conclusion: writing philosophy

249

in writing as one corpus in pre-imperial China. This stands in plain opposition to contemporaneous argument-based texts, which, clearly, circulated in writing independently as stand-alone entities and so without any need for contextualisation by masters of the texts. It can be concluded therefore that with the illocutionary force made explicit in philosophical texts from the Warring States period, there appeared new forms of communication that attempted to establish philosophically sound positions in exclusively written form. These new forms of philosophical productivity that turned texts into stand-alone philosophical entities were the immediate result of an evolving manuscript culture. The spread of light, easily extendable, readily used, and portable writing materials fostered new forms of writing and, hence, of philosophical production. Writing Philosophy Scholars such as Marshall McLuhan, Jack Goody, Ian Watt, David Olson, Walter Ong, and many other theoreticians of literacy and written communication have long pointed to the particular kind of internal influence which writing has on the matter to be communicated.6 Writing, consensus has it, is of special importance for the advancement of philosophical concepts. It is more than just the transcription of speech.7 Aristotles famous formula that speech can reproduce thought, whereas writing can only reproduce speech, seems to be mistaken.8 Instead, it appears, as Goody puts it, that writing has a particular kind of internal influence since it changes not only the way we communicate, but the nature of what we communicate.9 Seen as an instrument of cognitive development, writing is often held to impart a degree of abstraction to thought which is absent from oral discourse.10 In this respect, it is no wonder that text and writing as devices of meaning construction have turned into objects of research

6 McLuhan 1962; Goody 1986, 1987, 1997, 2000; Goody and Watt 1968; Olson 1980, 1994; Ong 1976, 1982. 7 See Olson 1994, 258. 8 Aristotle, Organon. According to A. Assmann and J. Assmann, writing would thus be nothing more than the mimesis of mimesis. See A. Assmann, J. Assmann, and Hardmeier 1998, 265. 9 See Goody 2000, 136. 10 Olson 1994, 7.

250

chapter eight

themselves. Although rejecting the repeatedly mentioned hypothesis that writing itself brings about an evolution in thinking,11 I agree with Lloyd that, at the very least, it advances the availability of certain types of argument construction.12 The argument-based texts from Gudin One are a good example of this. These texts, I hold, are ultimately products of written reasoning. The Hybrid Form of Texts Just as the distinction of argument-based and context-dependent text describes two extremes on a continuous scale of texts rather than absolute categories, so too does the distinction between the written and the oral text. Philosophical texts from the Warring States period are principally synchronic artefacts. As shown in this study, the philosophical texts from early China were part of a larger intellectual discourse in which travelling concepts informed a multitude of ideas in different ways. Texts that reflect an orally based activity come to us in the form of written artefacts, and written texts were not transmitted exclusively in written form. Changes in different manifestations of essentially written texts make this plain. Oral stock phrases and the like repeatedly appear in texts of the written tradition, and the repetitive nature of the mantra-like language of many of the building blocks appearing in ultimately written texts such as the argument-based text Wu xng point to oral contexts of some textual materials and layers. But the meaningful integration of these oral contexts into the new context of a consistent philosophical entity that becomes an autonomous product of thought would not have been possible without writing. And writing this type of text would have been difficult without the context of a manuscript culture, such as arose in the Warring States period. The syncretic tendencies and convoluted patterns of argument construction as seen in argument-based texts are clear signs of writing. Writing both facilitated and fuelled the generation of new types of reasoning. That coherent concepts are advanced in structurally consistent entities further demonstrates the premeditated confrontation with a problem of some kind. The systematic engagement with ideas

11 12

See the discussion by Halverson 1992. See also Falk 1990. See Lloyd 1990.

conclusion: writing philosophy

251

in structurally coherent texts casts light on deliberate attempts to find reasoned solutions for philosophical problems. The Decline of Argument-Based Texts The particular type of reasoning manifested in argument-based texts should not be taken simply as an oddity of tomb Gudin One. This is made plain by the find of argument-based texts from the Shnghai collection of Chu manuscripts. Yet, as far as we can see today, the end of the Warring States period also saw the end of this type of philosophical reasoning in writing. It is ironic that none of the texts of the written type survived outside tombs to the present day, but those texts from the same corpus that worked in oral contexts all persisted into received traditions! Whereas light writing materials continued their triumphant advance through the millennia, certain types of philosophising that, to a large extent, had been made possible by the broad diffusion of these materials ceased. This calls for an explanation. Parallel to the decline of the tradition of the argument-based texts, China witnessed great changes. We are clouded by uncertainty as to the beginning of institutionalised writing in early China. Certainly by the time of the Qn, at least, the office of erudites was established, whose position was further manifested under the Hn. It is in this context of offices and Imperial Library that the hardening of philosophical traditions took shape and that the closure of the canon advanced, as enforced particularly under Hn Wud (r. 14187 BC).13 But these changes in the intellectual climate do not in themselves explain how a whole tradition of systematised philosophising came to an end at the dawn of the imperial age. What are the reasons, then, for this decline? Did the type of philosophising as reflected in argument-based texts become obsolete? Were texts of this type transformed into new types of philosophical reasoning? Could these texts not survive the banning of writings outside official institutions and the Imperial Library because they lacked strong patronage from particular groups? Were the argument-based texts

Petersen (1995) and Kern (2000, 184ff., also for further references), among others, provide a detailed discussion of changes in the intellectual climate following the Warring States period. Michael Nylan (2009, with further references) claims that the influence of imperial patronage after 221 BC might be overstated.

13

252

chapter eight

simply too difficult to remember in those days, in which the physical possession and use of these texts was becoming a dangerous habit? The banning of writings, as allegedly demanded by L S (ca. 280208 BC), was directed in particular against uncontrolled learning. Banned texts included Odes (sh ) and Documents (sh ) circulating outside groups of officials and, in particular, uncontrolled anecdotes ( y ) of the manifold lineages (bai ji ). The subsequently implemented ban also included the various archives, except for those of the Qn.14 As Kern has demonstrated, these measures of the Qn imperial courthabitually a place of traditional ritual and classical scholarshipseem to have been aimed at controlling texts rather than at suppressing scholarship as a whole.15 As the comparison with other societies and cultures shows, this was a typical process in the establishing of a canon.16 The number of texts that fell out of the transmission process during these times must have been quite large. The destruction of texts, one may further argue, caused by the devastating fire that raged for months in the Imperial Library of the Qn, the home of the banned texts, when Xing Yu took the capital Xinyng by force in 206 BC17 furthermore erased a considerable body of texts from the memory of later generations. Yet text finds such as the manifestation of the Wu xng in the early Western Hn tomb Mawngdu Three make it plain that some argument-based text did in fact survive the early imperial age. The banning of selected texts outside the Imperial Library certainly was too short-lived for the majority of texts to be disremembered. References to the banning of writings and the destruction of the Imperial Library therefore cannot sufficiently explain the disappearance of argument-based texts as a philosophical tradition.

Sh j, 6:255, 87:25462547. See also Petersen 1995; Kern 2000, 190191. See Kern 2000, 188ff. For a competing view on the Qn, see H Jn 1999. 16 See A. Assmann and J. Assmann 1987. The appointment of a specialist in cultural memory whose task it was to comprehend the past and present (tng gu jn ) (see, e.g., Hn sh 19A, p. 726) is just one indication of the deep roots in textual and ritual traditions of the Qn. Martin Kern (2000, 184191) holds that the infamous burning of writings, now sometimes inadequately referred to as bibliocaust, was rather a move to monopolise classical learning and thus comparable to Hn Wuds appointment of erudites for the five canons and the simultaneous expulsion of competing doctrines in 136 BC. 17 An act of violent destruction such as this one certainly was no isolated instance. As Mark Lewis (2007, 101) reports, capitals were burned to the ground whenever a new dynasty took control.
15

14

conclusion: writing philosophy

253

The measures of the Qn, continued in different form under Hn Wud, can be read as directed primarily against heterogeneous and, by implication, uncontrolled ideas in written form. This may have had a direct effect on argument-based texts. Texts like the Wu xng not only quote canonical texts but also establish a particular reading of them. In many cases, these interpretations differ substantially from those singled out under imperial patronage, such as, for instance, the interpretation advanced by the Mo Odes. It is possible that the paid professionals at court felt undermined by such unorthodoxy.18 Losing their function as exegetical tools for authoritative sources, texts of this type might have slowly droppedor been forcedout of the transmission process. But what about texts like the Zhng xn zh do, Qing d y sh, and Ti y shng shu that, among other argument-based texts from Gudin One and the Shnghai collection of Chu manuscripts, do not integrateexplicitly or implicitlyreferences to later canonical works into their own argument? It seems therefore that the mere reference to imperial patronage in the attempt to establish a canon after the end of the Warring States period is also insufficient to explain the dissolution of the philosophical tradition exemplified by the argument-based texts from Gudin One. Yet another way to look at this phenomenon is to see these changes in the context of natural text development. Argument-based texts, the analysis has shown, are characterised by the way they expound all the relevant information necessary to generate meaning within the written text itself, and so these texts tend to transport largely self-contained philosophical systems. No specific contexts are necessary to access their message. This does not mean that this is achieved at every text layer or in every reference to shared knowledge. Some explanations may fail and some references may remain unidentified and unspecified, as happens from time to time especially in the manifestations of the long argument-based texts analysed in this study. But meaning is nonetheless borne by the overall coherence of the argument. Text flow and rhythm further help to bridge uncertainties. The text on the whole is relatively unfettered by unspecific passages or unsuccessful references in the construction of an argument. By becoming

18 The Mo tradition eventually displaced other interpretations of the Odes during the reign of Hn Wud (ca. 133).

254

chapter eight

independent of locally based interpretations, argument-based texts were potentially accessible to whoever had access to them and was able to read. Oral contexts became less important, and, as discussed, already at a relatively early point in time, these texts circulated independently in writing. As structurally closed entities, they left less room for interpretation and so were themselves prone to change. Argumentbased texts almost invite permutation. Ideas expressed in these texts come to fruition in other contexts, so that, in the course of time, these texts become redundant as stand-alone entities. Certain ideas and concepts expand into new textual environments and traditions, and argument-based texts dissolve as independent entities. The hypothesis concerning the text development of the modular Z y might hint at such processes of textual change. In contrast to argument-based texts, the modules of context-dependent texts locate the intellectual activity not within the written text but in the different textual communities around them. The modules of these texts function predominantly as platforms for all sorts of philosophical conversations, and so they remain crucially bound to the triangular relationship of text, mediator, and recipient. Because these modules did not establish the relevant reference within the written text but left it to be construed in the oral context of group-based communications, the modules remained ambiguous even when written down. But their intrinsic connection with authorities of whatever kind guaranteed their importance and ensured their transmission. Their ambiguity required explanatory settings, and these modules were easily adapted to the different contexts in which they were used. Context-dependent texts therefore beg for repetition. Being applied here and there but always calling for different interpretation each time, they permuted their contexts. Accordingly, authority shifted gradually from the texts to the textual communities who decidedand still do decidehow to read and interpret these modules, and so the modules of context-dependent texts could move among textual communities and interpretations. Because these texts were so open to interpretation, they could be applied to all kinds of arguments and situations. The ever-evolving act of interpretation endlessly reconstitutes the contextdependent text. In the end, their very ambiguity and need for interpretation kept them alive. As a result, it is the oral texts from the Warring States period that survived in the written textual tradition, while the early written texts dropped out of the transmission process.

PART III

TEXTS, RECONSTRUCTIONS, AND PHILOLOGICAL NOTES

CHAPTER NINE

RECONSTRUCTING THE ZHNG XN ZH DO This chapter provides the text and translation of the argument-based text Zhng xn zh do, followed by philological notes. In both the transcription and the translation of the text, a superscript letter z refers to the manuscript (Zhng xn zh do), and the number immediately after the z is the strip number. In my discussion of the text, z1, for instance, refers to strip 1 of the Zhng xn zh do, and z1/9, to the ninth graph on that strip. Some parts of the text are marked with a cross. This indicates that the text on the strips (or the strip itself ) is corrupt or that a graph cannot be identified with certainty. The translation must then be tentative, too. Missing parts that I reconstruct on the basis of the continuous argument of the text are marked by curly parentheses {}. The reconstructed passage is set in italics. The letters in brackets refer to the philological discussion at the end of this chapter. Text and Translation: Zhng xn zh do The Zhng xn zh do consists of two cantos and six building blocks. Canto 1
1.1a 1.1b 1.2a 1.2b 1.c 1.1a 1.1b
z1

[A] z2 Not to be pretentious and not to be destructive is the culmination of fidelity. Not to cheat and not to be cunning1 is the culmination of trustworthiness.
z1

1 Zh must be understood as positive knowledge that prevents one profound understanding. In the Laoz the expression zh is often used in the negative sense for artfulness, cunning.

258
1.2a 1.2b 1.c

chapter nine
When fidelity is accumulated [by the ruler],2 [he] can be approached [by the people]. When trustworthiness is accumulated [by the ruler], [he] can be trusted [by the people]. That fidelity z2 and trustworthiness have been accumulated [by the ruler] and the people did not approach and trust [him]there has never been such a case. [B] [C] z3 [D]

2.1a 2.1b 2.2a 2.2b 2.c 2.1a

The highest fidelity is like the soil; it transforms the things but does not attack them.3 2.1b The highest trustworthiness is like the seasons; [they] succeed [one another] and [the circle] does not end. 2.2a Men of fidelity have no z3 pretension; 2.2b Men of trustworthiness are not perfidious. 2.c The gentleman-ruler goes along with this, and therefore, [he] does not cheat life, nor does [he] turn his back on death. 3.a 3.b 3.c 3.a 3.b 3.c [E] z4 [F] To hold old ways in high esteem and never render [them] impure4 is fidelity in its culmination. To hold antiquity in high reverence and take it as a principle is trustworthiness z4 in its culmination. The highest fidelity has no pretension; the highest trustworthiness is not perfidious. Now this is what this is about.

2 The concept of accumulating ( j ) moral conduct is also prominent in the Zhungz . In the Xioyo yu it is, for instance, said: When the accumulation of water is not deep, then it will lack the strength to carry a large boat (see Zhungz zuanjin, 1993, 2). 3 These natural forces do not create the thing itself but merely make it work. The idea behind this is probably that of changing the form (edos or morph ) instead of the matter (hl ) of a thing. 4 The Shuwn jiz (1998, 571) explains y () as y means to become impure. The ode Goqi (Mo 80) reads: That man there, he is steadfast unto death. See Karlgren 1950, 54.

reconstructing the zhng xn zh do Canto 2

4.1a 4.1b 4.2a 4.2b 4.c 4.1a 4.1b 4.2a 4.2b 4.c 5.1 5.2 5.3 5.C 5.1 5.2 5.3 5.C 6.1a 6.1b 6.2a 6.2b 6.c [G] z5 [H] [I]

259

The highest fidelity is not pleasant for [the people]. The highest trustworthiness is not restricted in time.5 Not pleasant for [the people] and yet providing enough to nourish, such is the Earth. Not to be restricted in time z5 and yet able to be met with, such is Heaven. To be in tune with Heaven and Earth, this is what fidelity and trustworthiness are about. {} []z6 [J] [K] If kind with words but in fact not acting in accordance with them, the gentleman-ruler rather refrains from speaking so. If letting the mind loose, {and yet being} z6 intimate in appearance, the gentleman-ruler rather refrains from displaying this. If acting according to the old but pleasing the people by serving [them the special taste of ] zhng,6 the gentleman-ruler is one who refrains to relying on this. As to these three [ways], the man of fidelity refrains from doing them, and the trustworthy man refrains from applying them. z7 [L] z8 [M] ----- [N] [O]

z9 [P]
5 The Zhungz, Z Yng reads: If today the number of objects was calculated, one would not stop at ten thousand. But the fact that one is limited and calls [them] ten thousand is due to the high number [of them], so [they] are called out aloud that way. The Chng Xunyng subcommentary explains: q means limited. See Zhungz zuanjin 1993, 218. 6 Zhng is probably a dish of fish and meat mixed together, as in medieval times. See Xjng zj (1991, 2:7374).

260

chapter nine

6.1A When fidelity becomes the z7 way [in the state], then all kinds of skilled labour will not decay, and the nourishing of the people will all be sufficient.7 6.1b When trustworthiness becomes the way [in the state], then all groups of things will be completed, and all kinds of good deeds will thus be established. 6.2a Thus, when the gentleman-ruler in the way he conducts himself is indeed z8 of fidelity, [even] the Mn barbarians will come close to and follow [him]. 6.2b [And] when his words are indeed trustworthy, [they] are sincere and can be passed on. 6.c Fidelity is the realisation of benevolence (rn); [And] trustworthiness is the basis for rightness ( y). It was for this reason that what [the gentleman-ruler] applied to the Mn and M barbarians in the days of old z9 was something of this kind.

Notes on the Text and Translation [A]: Zhu Fngwu identifies the character z1/4 as d to arrive at. He explains z1/4 on the basis of its similarity to the graph d as seen in manuscript Laoz A, strip a8/14: .8 The identification of z1/4 with the character d is, however, problematic. The contexts in which the two graphs were used differ greatly, and the two graphs differ enough structurally that I think it unreasonable to identify z1/4 as d to arrive at. Reading the character as d as proposed by Zhu would also distort the overall pattern of this passage, which reads as follows:
A: no [x] and no [y]; that is zhng in its culmination; B: no [c] and no [d]; that is xn in its culmination.

According to the underlying structure of the Zhng xn zh do, this passage clearly demands a negative term. H Lny proposes to read the graph as to to be weary about something.9 Chn Wi accepts this reading.10 On the basis of similarities to a graph written

7 The different kinds of skilled labour (baigng ) is also a prominent concept in the Mz (see, e.g., Mz, Ji yng zhng, 255). 8 See Zhu Fngwu 1998, 121ff. 9 See H Lny 2001, 164. 10 See Chn Wi 2003, 75.

reconstructing the zhng xn zh do

261

in the same style appearing in the Shuhu d (8.1),11 T Zngli and Li Zuxn identify the character as hi to harm.12 I accept their reading. [B]: Qi Xgu reads the character z2/17 identified as gu by the Hbi Province Museum editorsas hu (to develop).13 The character gu consists of the graph and the phonophoric wi (Old Chinese *[g]w(r)aj). Wi (*[g]w(r)aj) is phonetically close to hu (OC *qwhraj-s), and I follow Qi. Qi Xgu identifies the character z2/21 ( ) as f to attack.14 This perfectly corresponds with the structure of the text, and I follow Qi. [C]: T Zngli and Li Zu xn identify the character z2/26 ( ), which consists of ci and the phonophoric b (OC *pij), as b (OC *pi[t]) necessarily.15 The loan characters are not a perfect match but are possibly close enough phonetically in the underlying Chu dialect. Another suggestion is provided by Li Zho , who proposes to read the graph as b close together; successive.16 The Old Chinese reading of b is *[b]ij-s, and so it is the more likely reading of this graph. [D]: Throughout the Zhng xn zh do cases of ligature (hwn ) appear that are not marked explicitlyfor instance, by a stroke. The concept jnz gentleman, for instance, is written throughout as one graph ( ).

(here identified as d ji [E]: L Lng reads graphs z3/1718 17 ) as ti jiu very old. Li Zho follows his reading.18 The problem with this reading is that it neglects the parallel structure of the passage under review, which I discuss under [F].
11 12 13 14 15 16 17 18

Cf. Shuhud Qn jian wnz bin 1994, 117. See T Zngli and Li Zu xn 2001, 66. See Hbi shng Jngmn sh bwguan 1998, 163, n. 3. Ibid. See T Zngli and Li Zu xn 2001, 67. See Li Zho 2005, 162163. See L Lng 1999, 501. See Li Zho 2005, 163.

262

chapter nine

[F]: The Hbi Province Museum editors identify z3/26 as to a kiln for burning pottery or earthenware.19 This reading makes it difficult to construct any sense out of this line. The editors admit that this sentence might be corrupt, and most editors struggle to reconstruct any meaningful reading. I suggest reading z3/26 as a ligature writing a putative predicate-object construction, namely d gu to hold antiquity in high reverence, which can be corroborated by reference to the structure of this passage (3. A read as d ji ). The two characters are written close together and are therefore hardly legible. One may be inclined to interpret this by saying that d gu to hold antiquity in high reverence was a standard concept in certain traditions and so the graphs were purposely written close together (just like the compound jnz ; see [D] above). However, given the conditions of manuscript production in the Warring States period, I feel that we simply cannot be sure whether this is a scribal idiosyncrasy or rather a graphic hint at an implied meaning, although the existence of similar examples from different manuscripts seem to favour the view that it is a scribal idiosyncrasy.20 The Hbi Province Museum editors identify the characters z3/28 29 (here read as zh chng to take as a principle) as zh shng .21 Zh (*tA) and zh (*ta) belong to the same rhyme group and xishng series. The same holds true for chng (*[d] a) and shng (*[d]a-s), which also belong to the same rhyme group. The verbalization of zh (= ,*t + []a) is perhaps a Chu-specific usage. [G]: The editors of the Hbi Province Museum identify the character z4/20 as du to snatch away (OC *-lot).22 Qi Xgu suggests

See Hbi shng Jngmn sh bwguan 1998, 163, n. 6. One instance in which two graphs are written so close together that they are easily mistaken for one is on strip Zy17 of the Shnghai Z y. These graphs (Zy17/
20

19

2930) are generally read as j (see, e.g, Shaughnessy 2006, 113). It is, however, likely that ligature instead concatenates two graphs () so closely together that they are easily mistaken for one word. This reading also seems to be the better option because the quotation of Odes in the Shnghai Z y matches the length of that in the Gudin One text and also the received version. 21 See Hbi shng Jngmn sh bwguan 1998, 163. 22 Ibid.

reconstructing the zhng xn zh do

263

the reading shu to speak (OC *l ot-s).23 The same character, however, reappears on strip z6/10, and Qi reads it there as yu to be pleased (OC *lot). Belonging to the same rhyme group, the characters shu and yu are interchangeable. I consistently read the graph as yu to be pleased in the Zhng xn zh do. [H]: The editors of the Hbi Province Museum leave graph z5/3 unexplained. Qi Xgu reads it as yo to restrain.24 Chn Jin thinks that the upper part of the graph is similar to xi (OC *[g]ra), as based on bronze inscriptions. He suggests accordingly the for .25 The graph could then be read as y direct transcription (OC *[]rak-s) to meet; to receive. After the obvious end of the sentence ([such] is Heaven), and therefore the end of 2B, the Hbi Province Museum editors continue the sentence up to what I define as the end of the first sentence of building block 5. However, the character y here serves as a marker to end the sentence. The sentence above (4.2A) talks about the Earth, followed by the sentence about Heaven. The following line (4.C), to be in tune with Heaven and Earth, . . . (discussed under [I]), obviously concludes the information given in building block 4 on zhng and xn. is problematic. [I]: The correct identification of the character z5/7 Zhu Fngwu identifies it as xn to follow.26 This graph was originally transcribed as ji to regulate by the Hbi Province Museum editors.27 L Lng reads it as s to resemble.28 Chn Jin suggests reading it as pi to be in tune with, to match, as commonly used in the bronzes and Odes in this meaning.29 As for the right part of the graph , Chn Jin sees some justification for reading it as b , just as in fi , as a comparison with other manuscripts suggests.30
Ibid. See ibid., 164, n. 10. 25 See Chn Jin 2002b, 56. 26 See Zhu Fngwu 1998, 125. 27 See Hbi shng Jngmn sh bwguan 1998, 163. 28 See L Lng 1999, 502. 29 See Chn Jin 2002b, 3. 30 Cf. (Boshn strip 203); (Boshn strip 250); 116). Quoted from Chn Jin 2002a, 225.
24 23

(Wngshn no. 1, strip

264

chapter nine

[J]: After the character xn mind, strip 5 breaks off, leading to [X][X][X] on the remaining strip. The Qi Xgu reconstructs [][][X][/] letting the mind loose (), and yet to be intimate in [ones] appearance, . . ..31 The shape of the character identified in this Gudin One manuscript as shn to state, to express, strongly resembles the character which the Shuwn jiz explains is the ancient script variant (gu wn ) of shn .32 [K]: Most scholars read the character z6/6 gu old (, *ka) as g therefore (, *ka()-s), which would be the concluding particle after a chain of arguments. This, however, runs counter to the formal structure of this passage, which works as follows. The unit is an enumeration. I have marked this with the numbers 13 in the translation, and it is followed by a concluding remark. This becomes apparent, first, due to the strict usage of [x]/ then the jnz would rather not x; second, the concluding remark (C) states these three [above-stated] fallacies. The structure of this section is the following: To do [x] (=positive) but () thereby to do [y] (=negative), the jnz would rather not [x]. By implication, the character g here must be read as gu old, since the chain of argumentation still continues. The phonophoric of the character z6/9 is *s.re (or *[ts] he?). On this basis, Li Zho suggests reading the graph as zhng (OC *[ts]()re) to compete; to struggle. Another reading would be to take its direct reading, namely zhng , which describes a special dish of fish and meat mixed together as mentioned in the medieval source Xjng zj .33 Neither of the two possibilities would alter the structure or the message of the present statement. [L]: The editors of the Hbi Province Museum read graph z7/6 as gu old (*ka). Qi Xgu proposes reading it as k to decay.34 Gu old (*ka) and k to decay (*[g]a) share the same phono-

31 32 33 34

See Hbi shng Jngmn sh bwguan 1998, 164, n. 13. See Shuwn jiz, 753. See also Hbi shng Jngmn sh bwguan 1998, 164, n. 14. See Xjng zj 1991, 2:7374. See Hbi shng Jngmn sh bwguan 1998, 164, n. 16.

reconstructing the zhng xn zh do

265

phoric and fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. I follow Qis proposal. ( ) as lin to feel persistent [M]: L Lng reads graph z8/3 35 attachment. Zhu Fngwu states that the character is mn (OC (OC *mo[n]), which should thus be interchangeable with lun *m.ro[n]), which he reads as mn , the name of a non-Zhu tribe (OC *m-ro[n]).36 I follow Zhu. The editors of the Hbi Province Museum transcribe the character z8/5 f (OC *ph(r)a)as f to adhere to (OC *N-p(r)o-s).37 Phonologically, this is unlikely. Li Zho reads it as f , which shares the phonophoric with the graph on the strips.38As he notes, f is also used in the sense of to adhere to; to follow.39 The fact that the Hn sh uses the two interchangeably suggests that in Hn times, at the latest, their reading was close enough (in some dialects) to be interchangeable. The Middle Chinese reading for f is bjuH; that of f is pjuH. It is possible that in someChu ?dialects, this change occurred quite early.40 Hence, although *-a and *-o clearly were distinct in (early) Old Chinese, they had merged by Hn times after

See L Lng 1999, 503. Zhu Fngwu 1998, 127. 37 See Hbi shng Jngmn sh bwguan 1998, 163. 38 See Li Zho 2005, 166. 39 Ibid. 40 This is a reasonable suggestion because something similar can also be seen in the use of graphs in other Warring States manuscripts. Note in this context that the reconstruction of Old Chinese is not that of one language with sudden drastic changes but one that has to deal with a variety of dialects and different gradual changes. An example of the early change to the Middle Chinese reading already in the Zhu period can be seen in the Shnghai manuscript Kongz sh ln . The transmitted Odes read wi ch wi x I make fine cloth and coarse cloth (Mo 2) (after
35 36

at this juncKarlgren 1950, 3). The Shnghai manuscript displays the characters ture (strip kz24/23), for which Chn Jin proposes the direct transcription (the left part of the first character is missing) (see Chn Jin 2002a, 222). The phonophoric of the former of the two graphs (ch in the transmitted version) is d . D has the Old Chinese reading *tij. The transmitted graph ch fine cloth should be reconstructed with *t-qhrj. This does not fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components as defined above. The Middle Chinese reading for ch , however, is trhij, and it is reasonable to assume that the change of *t-qhrj already occurred in the Zhu period. (As a side note to this, it is thus plausible to assume that ch is the earlier use, not d .)

266

chapter nine

labial initials like *p in some rhyme classes, and they may already have merged in the dialect of the strips. Accordingly, I follow Lis reading until a better solution is found. [N]: I follow Qi Xgus suggestion to read graph z8/12 as tn sincere.41 [O]: Graph z8/23 ( ) is a particular writing for y rightness in the Zhng xn zh do manuscript. in the sense of Chn Wi proposes reading graph z8/25 j basis (OC *k(r)) instead of q temporality (OC *[g](r)). He does so on the basis of similarities between this passage and the Qin f ln , W bn .42 According to him, the line should also be understood as the parallel counterpart to (). On this basis this line can be read as a parallel counterpart to the previous statement, which makes basis (j ; OC *k(r)) the counterpart to nucleus or, better, realisation (sh ; OC *[g](r)). Phonologically this would be a sound assumption, because the two words fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. They have the same position of initial articulation (a velar as initial consonant) and share the same main vowel and the same coda. I know of no other passage in which a notion appears similar to Trustworthiness is the limitation for rightness. But as suggested by the Qin f ln, there were notions similar to that of () Rightness is the basis for trustworthiness. Accordingly, I follow Chns suggestion.43 [P]: The use of sh clan, lineage for sh this can be seen in many texts from the Warring States. Sh has the Old Chinese reading *[d]e; sh can be reconstructed as *[g]e. This does not fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components since the initials of the two do not have the same position of articulation (sh has a velar initial; sh has a
See Hbi shng Jngmn sh bwguan 1998, 164. See Chn Wi 2003, 8182. 43 Note the similarity of the phrase fidelity is the realisation of benevolence (6.C) to D Di L j 9.4/54/26: Sagacity is the blossom of wisdom. Wisdom is the fruition of benevolence. Benevolence is the vessel of trustworthiness. Trustworthiness is the replication of rightness. Rightness is the root of benefit.
42 41

reconstructing the zhng xn zh do

267

dental initial). The common use of sh clan, lineage for sh this may therefore be explained by diachrony, by suggesting that this use reflects Chu dialect, or by reconstructing a prefix with a velar, as lately suggested by William Baxter and Laurent Sagart: *[g]e~ *k.de ~ *k. de. Evidence for this can be found in both the xishng rhyme groups and borrowings in other languages (zh paper *[k.t]e). The reconstruction of the additional statement is problematic. The character below h (z9/4: ) consists of the graphs yu (OC *gw-s) and mn (OC *m[n]). The direct transcription of it is . Zhu Fngwu reads it as mn (OC *m-ro[n], southern barbarian).44 Obviously, this is problematic on phonological grounds. Furthermore, Zhu considers the character z9/5 to be m (OC*mrak), the name of a northern barbarian tribe.45 The editors of the Hbi Province Museum read graph z9/5 as lu (OC*[r]o) to chatter, to mutter.46 L Lng only hesitatingly transcribes the characters in quesand with below.47 T Zngli and tion (9/45) as Li Zuxn suggest reading the two graphs as kilu to expound and promote. This would give the reading it was for this reason that in days of old [the sovereign] enacted expounding and promoting followed this [principle of zhng and xn].48 For the time being this must remain an open question. I refer to Zhus transcription, even though it seems that it is, from a phonological perspective, problematic.

44 45 46 47 48

Zhu Fngwu 1998, 128. Ibid. See Hbi shng Jngmn sh bwguan 1998, 163. See L Lng 1999, 503. T Zngli and Li Zuxn 2001, 7576.

CHAPTER TEN

RECONSTRUCTING THE QING D YI SH This chapter provides the text and translation of the argument-based text Qing d y sh, followed by philological notes. In both the transcription and the translation, a superscript letter q refers to the manuscript (Qing d y sh), and the number immediately after the q is the strip number. Text and Translation Canto1
q2 There is Heaven and there is Man, and there are distinctions between Heaven and Man. When investigating the distinction between Heaven and Man, [one] will know what they act upon. There might be the right man, but no right q2 times for him. Even if he was a worthy one, he would not enact it (his worthiness). However, if only there were the right times, what difficulties could there then be? (1) [A] q3 [B] Shn ploughed [the fields] at Mount L, and he made pottery q3 on the banks of the Yellow River. The reason he became established as Son of Heaven was his encounter with Yo.
q4 q1 q1

(2) [C] [D]

270

chapter ten
Sho Yo wore a hemp coverlet, covered [his head] with a hemp hat, and swathed himself in a [protecting] scarf. q4 The reason he became the assistant of the Son of Heaven when he escaped the wooden barriers for building earthen walls was his encounter with Wudng. (3) q5 [E] L Wng was a slave at the ford of J, and trembling he q5 watched the gates of the territory of J. Seventy years went by during which he slaughtered oxen at Zhog. The reason he was elevated to become the tutor of the Son of Heaven was his encounter with [King] Wn of Zhu. (4)q6[F]
q6

Guan Ywu (Guan Zhng) was detained in prison, where he was bound and tied up. The reason he became minister of the feudal lords when he escaped the weapons and his prisoners cage was his encounter with [Duke] Hun of Q.

(5) q8[G]
q8

Snsh (Snsh o) thrice declined [the position of ] vice minister of war at Qs. The reason he became the senior official when he came out [of seclusion] was his encounter with [King] Zhung of Chu .

(6) q7[][H] [?][I]

q7

Bil [X] was sold for the price of five rams and became the elder of the ox herders. The reason he became minister at the court when he escaped [?] was his encounter with [Duke] M of Qn.

q9

[J]

reconstructing the qing d yi sh

[K] q10

271

[Thus, the fact that] in the beginnings [these worthies] were of little value and in obscurity, [and yet] their names were later praised, is not because their virtue (d) has been added to. [That Wu ] Z x was very meritorious at first, [and yet] he later fell into disgrace and was put to death, is not because his wisdom q10 had weakened. [L] q11 [][M] That [even] the thoroughbred horse becomes distressed at Mount Zhng, and the black-mottled grey horse halts at the thorns of Sho, is not because they have lost their physical strength. [But] that [they] exhaust everywhere within the four seas, reaching as far as a thousand q11 l [in each direction], is because they encountered the [excellent rider] Zo F. To encounter or notthis lies with Heaven.

q9

Canto 2
q12 [] [N]
q13

[] [O] [q13][P] (?)([]q14 [Q] [R]1

1 This difficult passage has yet to be confirmed. The Hbi Province Museum editors transcribe it as follows: q14 ) (Hbi shng Jngmn sh bwguan 1998, 145). L Lngs transcription reads: [] q14 (L Lng 1999, 494). T and Li read: [] (T Zngli and Li Zu xn 2001). I suggest [] because this would logically connect to the passage above and to the conclusion below. See also my discussion below.

272

chapter ten
[Thus], to move is not [necessarily] to succeed. This is why [the worthy one] does not q12 {harbour resentment} even if becoming exhausted. {He [simply] hides and does not} go after achieving a name. And so he is without regret even if nobody knows [him]. {The [ flower] zhln grows in dark valleys}; q13 {it is not because it cannot be} smelled {by man} that it is not fragrant. The beautiful stone of jade is covered in mountain stones; it is not because {no one knows its}q14 goodness that it neglects itself.2 q15 [S] Failure and success appear at their respective times. Virtue (d) and conduct [may] be one, [and yet] fame and slander stand by their side. [But if ] acuity reaches the one mother, black and white need q15 not be distinguished [anymore]. Failure and success appear at their respective times, [yet] dark and bright do not get reiterated [along with them]. It is for this reason that the gentleman esteems self-examination.

Notes on Text and Translation [A]: The identity of Mount L (Lshn ) cannot be determined with certainty.3 Most scholars follow Qin M, who identifies Mount L with Mount Li Shou , which is near the confluence of the Fn and Yellow rivers in modern southwest Shnx .4 [B]: The character q3/3 appears as hu? ( ) on the bamboo strips. L Lng transcribes it as h bank of a river.5 T Zngli and Li Zuxn suspect that the character is a loan for

2 Although it is quite impossible to know the original reading of this passage since too many characters are simply absent, an approximate reading may still be reconstructed on the basis of the lines even if nobody knows [him], he is without regret), and (not fragrant since no [X] smell [it]). See my discussion below. 3 Cf. Zhnggu lsh dmng d cdian 1995, 145. 4 See Qin M 1962, 42. 5 L Lng 1999, 493; 2002, 87. See also J Xshng 2001, 118.

reconstructing the qing d yi sh

273

g (OC *ka). This would be the name of a river, thus making ,6 which would perfectly correlate with . Li Zho reads it as pu (OC *pa), banks of a stream.7 A transcription of the character as h (OC *qa) bank of a river is also conceivable as it correlates with the story in Sh j 1 Wu d bn j (Basic Annals 1), which notes that Shn made pottery on the banks of the Yellow River (). [C]: The name Sho Yo does not appear in transmitted texts. Accordingly, L Lng does not combine the two characters to generate the reading of a personal name but explains q3/13 ( yo , OC *law) as yo (OC *law-s) and considers it to be an attribute of y clothing, thus making shabby clothing.8 He further assumes that q3/12 () is either a mistaken character or a variant of a personal name. L s interpretation of reading yo (OC *law) as yo (OC *law-s) and thus taking only q3/12 as a personal name would also be a valid interpretation. None of these readings would change the overall interpretation of this passage. Despite this, the structure of this enumeration of semihistorical stories in the present subcanto is consistent in that a predicate immediately follows the name of the person. Therefore, it is more plausible reading q3/1213 as a personal name, followed by the predicate y to cover oneself , to wear. I follow L s interpretation and read q3/14 as x male nettle hemp.9 [D]: Legendary materials often refer to Wu dng (whose temple name was Go Zng ; Wu dng was the first of the nine historic rulers of the Shng dynasty; Robert Bagley puts the reign of Wu dng at around 1200 BC)10 in connection with F Yu , whose situation is described in similar ways to that of Sho Yo in the Qing d y sh.11

See T Zngli and Li Zuxn 2001, 29. See Li Zho 2005, 170. 8 See L Lng 1999, 493, as well as 2002, 87. 9 See L Lng 2002, 86. 10 See Bagley 1999, 181. 11 See also the Mz Shng xin zhng (), 57ff.:
7

274

chapter ten

) zng j jn [E]: I follow Qi in his reading of graphs q4/1517 ( .12 Tales about L Wng mention the place name Jjn () in context with L Wng.13 The ford of J (Jjn) is situated in modern Hnn.14 The phonologically unproblematic reading of zng (ewe; OC *[ts]a) as zng (slave; *[ts]a) further suits the context of the tale. L Wng, also known as L Shng, was one of King Wu s advisors. He received the titles Tigng wng Our Great Ancestors Hope and Tutor Shngf . Sarah Allan has compared the historical evidence of L Wngs service as a minister of Kings Wn and Wu of the Zhu dynasty with the many references to him in Zhu and Hn literature and concluded that these accounts are very contradictory. L Wng was a nobleman of the Jing clan, which traditionally intermarried with the Zhu royal family. It is possible that he was the uncle of King Chng of Zhu.15 However, legendary material always describes himjust as in the Qing d y shas a humble man, who was raised up from obscurity by King Wn and subsequently made minister.16 Zhog is the former capital of Yn located northeast of Q County, in modern Hnn. Western Zhu period Wi established its capital there, while Zhog belonged to Wi throughout the Warring States. [F]: The character q6/5 appears as yo (instead of , ) (OC *l[aw] or *lu) on the strips. Qi Xgu, however, argues that this character should be read qi (prison, OC *[s-m-l]u).
In antiquity, Shn ploughed [the fields] at Mount L; he made pottery on the banks of the [Yellow] River and fished in the lake of Li (lake of Ho , Zzhu ). Yo discovered him at F Z (uncertain name), made [him] the Son of Heaven, handed him the government of all under Heaven so that [he] should rule the entire people under Heaven. Y Zh (Y Y n) used to be the private counsellor of the daughter of the clan You Xn, [and then] was employed as a cook. Tng discovered him [and] made [him] his personal senior official so that [he] should rule the entire people under Heaven. F Yu wore coarse hemp cloth and belted it with a rope; his labour was to build earthen walls at F Yn. Wudng discovered him, made [him] one of the Three Dukes, handed him the government of all under Heaven so that [he] should rule the entire people under Heaven. 12 See Hbi shng Jngmn sh bwguan 1998, 146, n. 6. 13 See Allan 19721973, 74. 14 See Tn Qxing 1991, 1:2425. 15 See Allan 19721973, 6072. 16 Cf. Allan 1981, 21.

reconstructing the qing d yi sh

275

The story of Guan Zhng can be found in Sh j 62 Guan Yn li zhun (Memoir 2). [G]: Chn Wi proposes to interchange strips q7 and q8. Strip q7 records the story of Bil X (see [X] below). In the organisation of the text as presented by the Hbi Province Museum editors, this was the fifth of the six cases of men who rose from destitution to fame. Strip q8 carries the story of Snsh, originally the last of the six examples.17 Chn bases his reorganisation of the text on two considerations, the first of which is chronological evidence. Bil X served Duke M of Qn (r. 656621 BC); Snsh o served King Zhung of Chu (r. 613591). Bil X thus preceded Snsh o by at least three decades. The second is physical evidence. At the bottom . Chn Wi argues that this marks the of strip q7 is a black stroke end of the list of examples.18 To follow Chn in placing q8 before q7, which I do, does not influence the reading of the present subcanto. If the subcanto remains intactthat is, as long as no strips are added or removed from the group of strips q2/16q8/endand as long as none of the six examples are distorted by changing the order of strips q2q5, the order of strips q6q8 does not alter the content of this passage. They are stable components that may be placed in any position within this subcanto. Despite this, it is plausible that the list of anecdotes was continued in chronological order, since strips q2/16q5/end of the present subcantotheir sequence is beyond doubtclearly listed them in chronological order: Shn, served Yo (traditionally, r. 2366 2356); Sho Yo served King Wu dng (r. ?1189); L Wng served King Wn of Zhu (r. 1099/561050).19 Unnoticed by Chn Wi, strip q7 furthermore is the only instance in which the example given does not end with a concluding y but carries the black mark instead. So, when interchanging q7 with q8, the black stroke could feature as the ending mark of this section, which would, furthermore, be in perfect chronological order. Also left unnoticed by other scholars, this change is corroborated by the parallel form of what I have termed the A and B pattern of the present account. I follow Qi, who reads graph q8/4 as ch to decline.20
17 18 19 20

See Chn Wi 2003, 47. Ibid. See Shaughnessy 1999, 2526. See Hbi shng Jngmn sh bwguan 1998, 146, n. 11.

276

chapter ten

Graph q8/5 was originally transcribed as []. As Chn Wi correctly adds, throughout the Gudin One manuscripts, the graph is often exchanged with .21 The two graphs (OC *k(r)k and *kh(r)k-s) and (OC *[g](r)) are similar in sound. Together with q8/6 it would be read Qs , a place-name in Chu . This accords with a record in Xnz 5.1 Fi xing (Contra Physiognomy).22 Since at the time of these texts, *kh(r)k-s might possibly have changed to *kh(r)-s, I follow Chn in his reading. The title Sho Sma is a variant of Xio Sma vice minister of war.23 In Chu the senior functionary was termed lngyn .24 However, in various sources, Snsh o was consistently termed chancellor.25 [H]: On interchanging strips q7 and q8, see [G]. Qi Xgu reconstructs this part as Bail X was sold for the price of five rams,26 based on the Huinnz Xi w xn (Endeavour and Duty), where the phrase appears as follows: Bail X was sold.27 The story of Bail X can be found in Sh j 5 Qn bnj (Basic Annals 5). [I]: The character q7/13 remains obscure. According to the context of the story, it is clear that it must have to do with ox and sheep tending. From early antiquity, the term qng was used for eminent officials, sometimes particularised with further qualifications. Throughout the Zhu it meant minister, denoting the highest category of official serving the king and feudal lords.28

See Chn Wi 2003, 46. [A]s for Snsh o from Chu , he was a native of the small hamlet Qs. [He] was partly bald and [his] left foot was too long; so short [was he] that [he could] go under the poles of a state carriage. Despite this [he] made [the sovereign] of Chu protector over the states. 23 See Hucker 1995, 416. 24 See Loewe 1999, 1018. 25 Cf. Sh j 83 Lu Zhnglin Zu Yng lizhun (Memoir 23). 26 See Hbi shng Jngmn sh bwguan 1998, 146, n. 9. 27 See Huinnz 19/203/6. 28 See Hucker 1995, 173.
22

21

reconstructing the qing d yi sh

277

[J]: Strip q9 is one of the most controversial passages of the Qing d y sh. Neither Chn Wi nor Chn Jin connect strips q9 and q8 (both suggest placing q9 between q14 and q10). Moreover, the proper reading of some of these graphs is as yet an open issue. Originally, were transcribed as to and hai the two graphs q9/23 (also written ). L Lng transcribes the character q9/2 as to to sheathe. As he states, to (OC *l u) is derived from yao (OC *[l] u), which, throughout early manuscripts, is commonly confused with as hu (OC (OC *l u). He furthermore transcribes graph q9/3 *m k-s) obscure, dark.29 T Zngli and Li Zuxn follow the suggestion made by the Hbi Province Museum editors and transcribe the graph in question as to overflow, a torrent, rushing water. In its borrowed meaning of n shu muddy waters it also carries the meaning of little value.30 Zho Pngn approaches the two characters rather differently.31 He believes that the two graphs should be read tan hai a boneless brine of pickled minced meat, which, with reference to the Chu c, he reads in a borrowed meaning that denotes a certain kind of torture. As Zho presumes, this torture should refer to the historical figure of B gn ,32 who, in various sources, is often named with Wu Z x.33 For this explanation to work (the name never appears on the strips), Zho proposes adding an unknown number of strips between q8 and q9, thus arguing that the Qing d y sh must be incomplete. I doubt this. As demonstrated in my discussion of the text,34 the Qing d y sh not only is complete but also has a very rigid and argumentatively concise composition. That neither Chn Jin nor Chn Wi connects strip q9 with q8, but with q14, is problematic for various reasons. Despite the fact that strip q9 structurally marks a break after subcanto 2 and introduces a new aspectcrucially, this leads to the open argument of subcanto 335both scholars read the whole passage in one go. When focusing

See L Lng 1999, 495496; 2002, 88. See T Zngli and Li Zu xn 2001, 32, 33. 31 See Zho Pngn 2002, 1819. 32 Prince B gn served under King Zhu of Yn , by whom he was killed. The story of Prince B gn can be read in Sh j 3 Yn bnj (Basic Annals 3). 33 See Zho Pngn 2002, 20. 34 See chapter 2. 35 See my discussion in chapter 2.
29 30

278

chapter ten

on the structure of subcanto 2, the pattern of two parallel passages that run in the form of 3-3-4, 3-3-4 [] becomes apparent. This parallel sequence would be destroyed by reading the sentences continued in q14 and q8 ( q14[]36 . q9 . .).37 We would thus have thevery unlikely structure of 4-4-4-4-9(!)-3-4-5-3-5. For the identification of character q9/2 , I follow L Lngs interpretation.38 For phonetic reasons I read graph q9/3 as hu (OC *mk-s) obscure, dark, as a phonetic loan for hai . The only received reading of is yu (MC <hjuwH, which gives the Old Chinese reading *[g]w-s. Interpreting hu (OC *mk-s) as hai works only if the *m - initial in *mk-s and the *qwh- in had already merged. Although entirely conceivable, it nevertheless would besurprising if this merge had occurred at such an early date. [K]: Compare the similarity of this account with the story in the Hn sh wi zhun (ch. 7):39 []; The fact that in the beginning Wu Z x was very meritorious, [and yet] he later fell into disgrace and was put to death, was not because his wisdom had [periods of ] flourishing and decline [but] because he first met Hl 40 and later Fchi 41. The account in the Shu yun 17 Z yn (Miscellaneous Sayings) records the case by saying: It was not that his wisdom either increased or decreased.42 The story of Wu Z x, including not only his success in taking revenge for his fathers and brothers unjust deaths but also his own execution, appears repeatedly in early texts.43 Briefly, after Wu Z xs father and elder brother were murdered by the king of Chu, Wu Z x fled to the State of Wu .

[] read as one character. Transcription follows the one chosen by Chn Jin 2004, 317. 38 See L Lng 1999, 495496; 2002, 88. 39 See Hn sh wi zhun 7.6/50/24. 40 Hl (?496 BC), also Wuz Gung , Gongz Gung (or simply Gung ), or Hl , ruled the State of Wu from 514 to 469 BC. He was one of the Five Hegemons. 41 Fchi (?473 BC) ruled the state of Wu from 495473. He was son of Hl. 42 See Shu yun 17.17/144/10. 43 See Lewis 1990, 84. On the various accounts, see D. Johnson 1981; 1980a; 1980b.
37

36

reconstructing the qing d yi sh

279

In the State of Wu he gained the favour of the king, whom he helped to defeat Chu. Later, however, Wu Z x fell into disgrace and was executed by order of the kings successor. [L]: The editors of the Hbi Province Museum transcribe the character q10/4 as d (OC *[t]ewk) horse of good quality. L Lng identifies it as (OC *rek) in difficulty, distressed.44 He justifies this by referring to the Shu yun 17 Z yn, in which the graph also appears in combination with j a thorough bred horse ( ).45 This reading is confirmed by the fact, unnoticed by L , that in the subsequentand entirely parallelsentence, the second graph also is a verb, which describes the difficulties of a good horse when facing a seemingly insuperable difficulty. The character q10/5 zhng can be read as nervous, in tension, as mentioned also by T Zngli and Li Zuxn.46 However, referring to the pattern of the subsequent sentence (fine horseverb describing difficultiesplace-name), I rather view zhng as a place-name which can be identified with the name of a place in Jn during the Spring and Autumn period.47 Reading zhng as nervous would destroy the parallel structure of the passage. The editors of the Hbi Province Musem transcribe the character q10/7 as q a piebald horse.48 Q is further read as q (OC *[g](r)). L Lng identified it as jn a fine horse.49 However, the combination of j and jn is unattested in transmitted records (as opposed to that of ).50 Later, L corrected his transcription to

See L Lng 1999, 496; 2002, 88. See Shu yun 17.17/144/11. 46 See L j zhsh 21 Z j xi (Miscellaneous Records), 5:751: strained and yet without relaxation, Kings Wn and Wu were incapable of this; relaxed and without tension, Kings Wn and Wu would refrain from acting so. Once strained, once relaxed, that is the way of Kings Wn and Wu. See also T Zngli and Li Zuxn 2001, 34. 47 See Zhuo zhun, Xinggng, year 23, 604; Psho y , also referred to as Psho , Shotng , or simply Sho ; the place in Jn is located in Hnn, west of modern Jyun Municipality (see Zhnggu lsh dmng d cdian1995, 926). 48 See Hbi shng Jngmn sh bwguan 1998, 145. 49 See L Lng 1999, 496. 50 See Xnz 23 Xng (Nature Is Bad): 23/117/14. See also Zhungz 17 Qi shu (Autumn Floods). In Zhungz zuanjin 1993, 131.
44 45

280

chapter ten

q black-mottled grey horse, understood in the sense of j a thoroughbred horse and also referring to fine horses.51 [M]: Qi Xgu remarks that the story of the excellent rider Zo Fu also appears in chapter 7 of the Hn sh wi zhun and in Shu yun 17 Z yn. On this basis, Qi argues that the character q10/15 should be read as zhung ; q10/20 as zh ; and q11/3 as zo , referring to the rider Zo F. Qi further argues for adding the character fu after q11/3 and so completing the name of the famous horse rider.52 [N]: Strip q12 is broken at both ends. At the top of strip q12, presumably three graphs are missing. L Lng reconstructs .53 According to the underlying structure, I suggest reconstructing the present passage as follows. The topic of subcanto five is action that bears no result. The pattern runs: failure, and yet no X [something negative, intercepted by preceding no].54 The next sentence contains two elements, each bolstering up the other: (A-element) although A, (B-element) B to achieve a name. In this case the following sentence reveals the reading of the previous statement: (A-element) although nobody knows him, (B-element) he is without regret.55 Thus, even though it is impossible to reconstruct the exact reading of this unit, an approximate reading can still be reconstructed. The pattern applies to the entire passage and reoccurs on strips q11q14. It always reads: X (something negative), and yet Y .56 [O]: L Lng argues that at the bottom of q12, presumably six graphs should be added. Based on the Xnz 28 Yu zu and the Hn sh wi zhun (ch. 7), both of which record a similar story, he adds the

See L Lng 2002, 89. See Hbi shng Jngmn sh bwguan 1998, 146, n. 13. 53 See L Lng 1999, 496; 2002, 88. 54 Consequently, the reading yun , resentment. 55 Accordingly the reading of the A-element [he] hides and of the B-element [and yet] is without regret. 56 This is another indication that this passage is a unit and should not be disrupted, contrary to what both Chn Wi and Chn Jin argue.
51 52

reconstructing the qing d yi sh

281

graphs .57 This suggestion does indeed fit the context of the passage. First, the top of the firstcorruptedgraph is still visible. This seems to be the grass-component . Second, this reading can easily be brought into line with the next sentence [not] not fragrant because [x] smells it.58 This perfectly fits the pattern I have described above (strips q1112:A- and B-element). Since this reading is corroborated by transmitted texts (Xnz 28 Yu zu and Hn sh wi zhun, ch. 7), I follow L s suggestion. [P]: Presumably four graphs are missing at the top of strip q13. According to the pattern of the present subcantodescribed under [N], aboveit is in fact possible to come up with an informed guess on the approximate reading of the passage by completing the pattern X (something negative), and yet Y , based on the subsequent four characters ( [x] not fragrant since no [y] smell [it]).59 [Q]: Presumably four graphs are also missing at the bottom of strip q13. Since the subsequent line is still visible on strip q14 ( goodness, neglect itself ), an informed guess can be made as to the approximate reading of this passage based on the remaining graphs. Taking the A- and B-elements of the pattern X (something negative), and yet Y into account, the approximate reading should be { } {nobody knows its} goodness, this leads to:[x] does not do [yneglect itself ] because nobody knows of its goodness, or something along those lines. have caused commentaThe visible characters q13/59 tors considerable headaches. My reading of beautiful stone of jade and contain follows the reading of Li Lxin ,60 Li Zho ,61 Yn Shxun ,62 and Chn Jin .63 [R]: On the position of strip q14 in the Qing d y sh, see my discussion under [J].

57 58 59 60 61 62 63

See L Lng 1999, 496; 2002, 88. Compare the pattern I have described for this passage. Hence as approximate reconstruction. See Li Lxin 2000. See Li Zho 2000, 2005, 175. See Yn Shxun 2000. See Chn Jin 2004, 316.

282

chapter ten

[S]: The character q14/19 is transcribed as y (OC *lk) a dart; to shoot with bow and arrow, and it is difficult to make any sense of that reading.64 I suggest a parallel reading with q14/11 , which should be instead of y (OC *lk). My suggestion to directly transcribed as read the graph as y (one; OC *[]i[t]) offers a simple and sound reading, as it can be justified phonetically and perfectly connects to the explanation of ones behaviour in accordance to the one d from above. The reading of this graph as y one is well attested.65 Even though phonetically somewhat problematic, I tentatively follow L Lng in reading zh (a particle; OC *t) as z black66 (OC *[ts] r), for it connects best to the pairing of yu/mng below (q15/7, 8). T Zngli and Li Zu xn suggest reading q15/2 as l small, . It minute; to regulate.67 The direct transcription of the graph is seems to consist of two phonophorics, namely l , which can be reconstructed as OC *(m.)r, and li , which can be reconstructed as OC *(m.)r (<*m.rk). These phonophorics match the Old Chinese reconstruction for l (OC *[r]), and I follow the suggestion made by T and Li.

64 65 66 67

See T Zngli and Li Zu xn 2001, 37. See Kern 2005a, 187188, n. 43. See also H Lny 1998, 1080. See L Lng 1999, 496; 2002, 88. See T Zngli and Li Zu xn 2001, 37.

CHAPTER ELEVEN

RECONSTRUCTING THE WU XNG This chapter provides the text and translation of the argument-based text Wu xng, followed by philological notes. In both the transcription and the translation, a superscript letter w refers to the manuscript (Wu xng), and the number immediately after the w is the strip number. Text and Translation Subcanto 1
1.1. [A] w2 w3 [] [B] [] w4 [C] w1 The five aspects of [virtuous] conduct (wu xng) [are as follows]: When benevolence is manifested internally, we call it virtuous conduct; When it is not manifested internally, we call it conduct [only]. When rightness is manifested internally, we call it virtuous w2 conduct; When it is not manifested internally, we call it conduct [only]. When ritual propriety is manifested internally, we call it virtuous conduct; When it is not manifested internally, we call it w3 {conduct [only]}. {When wisdom is manifested} internally, we call it virtuous conduct; When it is not manifested internally, we call it conduct [only]. When sagacity is manifested internally, we call it virtuous w4 conduct; When it is not manifested internally, we [still] call it virtuous conduct.
w1

284

chapter eleven

1.2. [] w5 [D] [Hence], virtuous conduct comprises five [aspects]; when they are brought into harmony, we call it virtue; When [only] four [aspects] of conduct are brought into harmony, we call it goodness. Goodness, this is the Way w5 of humans. Virtue, this is the Way of Heaven. 2. [] w6 [][E] [] [F] When the gentleman ( jnz) lacks concern in his inner mind, then [he] will be without wisdom in his inner mind. When [he] lacks wisdom in his inner mind, then [he] will be without w6 {delight} in his inner mind. {When [he] lacks delight in his inner mind, then [he] will not} be at ease. When not at ease, then [he] will not be joyful. When not joyful, then [he] will be without virtue. 3. w7 [G] [] [H] When all aspects of conduct are manifested internally, and w7 they are conducted at their [appropriate] time, then we call [him] a gentle{man}. The scholar-knight who sets his aspiration on the gentleman-way is whom we call aspiring scholar. 4. [I] w8 [J] [K] ]

[] [] w9
Goodnesswhen refraining from acting for it, there will be nothing for approaching [it].

reconstructing the wu xng

285

Virtuewhen refraining from w8 aspiring to it, [it] will not be accomplished. Wisdomwhen refraining from thinking about it, [it] cannot be attained. [This is because], if thinking is not clear, one will not be investigating, [and] if thinking does not grow, it will not manifest. If [thinking] is not manifested, one cannot be at ease; If not at ease, one cannot be joyful; If not joyful, w9 one will be without virtue.

Subcanto 2
5.1. [L] w10 [M] [N] w11 [] [O] [] [] [P] If not benevolent, one cannot be clear in ones thinking. If not wise, one cannot grow in ones thinking. If neither benevolent nor wise, whilst not yet having seen a gentleman, the sorrowful heart w10 cannot be disturbed. Until [I] have seen the gentleman, [my] heart cannot be delighted. [But] when I have seen him, and when I have met him, then my heart will w11 {X be delighted}. 1 {X Now this is} what this is about. 5.2 [] w12 {If not} benevolent, one cannot be clear in ones thinking. If not sagacious, one cannot be effortless in ones thinking. If neither benevolent, nor sagacious, w12 whilst not yet having seen a gentleman, the sorrowful heart cannot be agitated. Until [I] have seen the gentleman, [my] heart cannot be stilled. 2

1 The lines [] quote Odes, Shonn: Caochng (Mo 14). As Csikszentmihalyi (2004, 283) notes, when a before/after structure is usedsuch as while I have not yet seen a gentlemanthe gentlemans effect on others is emphasised. 2 This passage appears in Xiaoy: Ch j (Mo 168): when I have not yet seen my lord, my grieved heart

286 Subcanto 3

chapter eleven

6.1. [] w13 [Q] [] [][R] [] [] [] [] [] [] [] The thinking of the benevolent one is clear. [This is because, only] if ones [thinking] is clear, w13 will one be investigating. If investigating, one will be at ease. If at ease, one will be gentle. If gentle, one will be delightful. If delightful, one will be intimate. If intimate, one will be affectionate. If affectionate, one will be caring. If caring, one will be of a jadelike coloration. If being of a jadelike coloration, ones [thinking] will be manifested. If ones [thinking] is manifested, one will be benevolent. 6.2.
w14

[] [] [] [] [] [] [] w15

w14

The thinking of the wise one is growing. [This is because, only] if [ones thinking] grows, will one be attaining. If attaining, one will be nonforgetting. If nonforgetting, one will be clairvoyant. If clairvoyant, one will see the worthy person.

is agitated. When I have seen my lord, then my heart will calm down (Karlgren 1950, 113, emended); and in Shonn: Caochng (Mo 14): when I have not yet seen my lord, my grieved heart is agitated. Let me have seen my lord, let me have met my lord, and my heart will then calm down (Karlgren 1950, 9, emended).

reconstructing the wu xng

287

If seeing the worthy person, one will be of a jadelike coloration. If being of a jadelike coloration, ones [thinking] will be manifested, If ones [thinking] is manifested, w15 one will be wise. 6.3. [] [] [] [] [] [] [] w16 The thinking of the sagacious one is effortless. [This is because, only] if ones [thinking] is effortless, will [it] be manifested. If [ones thinking] is manifested, one will be nonforgetting. If nonforgetting, one will be clairaudient. If clairaudient, one will hear the Way of the gentleman. If hearing the Way of the gentleman, one will be of a jadelike tone. If being of a jadelike tone, ones [thinking] will be manifested. If ones [thinking] is manifested, w16 one will be sagacious.

Subcanto 4
7.1. [S] The polite man, the gentleman, unified in his deportment he is.3 Only after one is able to be unified [in ones deportment] will one be able to become a gentleman. [Hence, the gentleman] is mindful of his singularity. 7.2.
w17 [] [T] w18 [] [U] w17

{I looked after her yet I could not reach her}, and my tears fell like rain.4

Odes, Gufng: Shji (Mo 152). This passage also appears in the Gufng: Ynyn (Mo 28): I looked after her, and yet I could not see her; my tears fell like rain. The Mawngdu Three Wu xng quotes the same ode but cites two more lines.
3 4

288

chapter eleven
Only after one is able to disarray the wings will one be able to develop utmost grieving.5 [Hence], the gentleman is mindful of his w18 {singularity}.

8. [] w19 [] [V] In his acting for goodness, there is [always] something with which the { gentle}man begins and something with which [he] ends. In his acting for virtue, {there is [always] something with which} the gentleman {begins, but there is nothing with which} [he] ends.6 9. [W] w20 [] [X] The sound of bronze and jade [stones] resonating along with it,7 this is a person possessing virtue. The sound of bronze is goodness; The tone of jade is sagaciousness. Goodness is the w20 Way of humans; Virtue is the {Way} of Heaven. Only after there is a virtuous person can there be the sound of bronze and jade [stones] resonating along with it.

5 This passage also appears in the ode quoted above: the swallows in the air, their wings are unruly. The Mawngdu Three Wu xng quotes the same ode but includes two more lines. 6 The Mawngdu version of this sequence reads: (8.12). 7 Mngz 5B1 reads: Ko ngz is said to have gathered great achievements; gathering great achievements is the sound of bronze and jade [stones] resonating along with it. The sound of bronze is the beginning of an inherent pattern; jade stones resonating along with it is the end of an inherent pattern. Beginning an inherent pattern is a matter of the wise one; ending an inherent pattern is the matter of the sagacious one. Mngz offers a rather different interpretation of this sequence than the Gudin One Wu xng does. This suggests that the two did not copy this line from one another but drew on different sources. Compare this line with Odes, Xiao ya: Bi j (Mo 186): , which was used in another context and has been understood differently.

reconstructing the wu xng Subcanto 5 (Canto 2)

10. w21 [Y] [] [] [] [] [Z]

289

If neither clairaudient nor clairvoyant, one can be neither sagacious nor w21 wise. [This is because,] if not wise, one will not be benevolent [either]. If not benevolent, one will not be at ease [either]. If not at ease, one will not be joyful [either]. If not joyful, one will lack virtue. 11. [Aa] [] [] [] [] If not [having feelings] of love, one will not be delighted [either]. If not delighted, one will not be intimate [either]. If not intimate, one will not be affectionate [either]. If not affectionate, one will not be caring [either]. If not caring, one will not be benevolent. 12. [Ab] [] [] w22 [] [] If not upright, one will not be resistant [either]. If not resistant, one will not be decisive [either]. If not decisive, one w22 will not be dignified [either]. If not dignified, one will not carry out [ones tasks either]. If not carrying out [ones tasks], one will not be righteous. 13. [] [] [] [] If not keeping [appropriate] distance, one will not be respectful [either]. If not respectful, one will not be stern [either].

290

chapter eleven
If not stern, one will not show honour [either]. If not showing honour, one will not be reverent [either]. If not reverent, one will lack ritual propriety.

Subcanto 6
14.1. w23 [Ac] w24 Never w23 having heard of the Way of the gentleman is what we call not clairaudient. Never having seen a worthy one is what we call not clairvoyant. Having heard of the Way of the gentleman and yet not understanding w24 that it was the Way of the gentleman is what we call not sagacious. Having seen a worthy person and yet not understanding that he possessed virtue is what we call not wise. 14.2.
w25 [] [] [] [] w26 w25

[But] to see and understand him [as a worthy one] is wisdom. To hear and understand it [as the Way of the gentleman] is sagacity. To illuminate the luminous is wisdom. To hold in awe the awe-inspiring is sagacity. To illuminate the luminous refers to below [and] to hold in awe the awe-inspiring w26 refers to above are what this is about.8

15.1. ( <> ) w27 [Ad] To hear the Way of the gentleman is clairaudience. To hear and understand it is sagacity. The sagacious one understands the Way of w27 Heaven.

8 This passage also appears in Odes, Dya: Dmng (Mo 236): shedding light on the bright is below, awing the awe-inspiring is on high.

reconstructing the wu xng

To understand and enact it is rightness. To enact it according to its time is virtue. 15.2. w28 [Ae] To see a worthy one is clairvoyance. To see and understand him w28 is wisdom. To understand and be at ease with him is benevolence. To be at ease and show respect to him is ritual propriety. 15.3. w29 [] [Af ] [] [] [Ag] w30 [] [Ah]

291

Sagacity and wisdom are those [virtues] from which ritual propriety and music originate, w29 {and by which} the five [aspects] {of [virtuous] conduct will be harmonised}. When harmonised, one will be joyful. When joyful, there will be virtue. When there is virtue, states and families will revive. The way King Wn appeared was like this. When {King} Wn w30 {was on high, he was illuminated by} Heaven.9 That is what this is about. 16.1. w31 To see and to understand [the worthy one] is wisdom. To understand and be at ease with him is benevolence. To be at ease w31 and use is rightness. To use and show reverence to him is ritual propriety. 16.2. w32 []

Here the Wu xng draws on Odes, Dya: Wn Wng (Mo 235).

292

chapter eleven
Benevolence and rightness are those [virtues] from which ritual propriety derives and by which the four [aspects] of [virtuous] conduct are harmonised. When harmonised, w32 there will be congruence. When there is congruence, there will be goodness.

Subcanto 7
17. w33 [Aj] [] [Ak] When facial coloration, look, manner, and appearance are gentle, this is love. When using the inner mind to interact with others, this is delight. When the inner mind is delighted by this and transferred w33 on to elder and younger brothers, this is intimacy.10 To feel closeness and to extend this feeling is affection. To be affectionate and to be genuine about it is caring. To care for ones father and, secondarily to this, to care for others is benevolence. 18. w34 [Al] w35 [] When the inner mind w34 rightly enacts something by being discriminative, this is uprightness. To [be] upright and to perpetuate this is resistance. To [be] resistant and not to fear the strong and powerful is decisiveness. Not w35 to harm the great Way for the sake of the petty ways is grave demeanour. To punish severely if there is a great crime is carrying out [the Way properly]. To venerate [the noble ones] and, to their level, to honour the worthy ones according to their level is rightness. 11

Reading the graph xn as to extend, to spread out. Ikeda Tomohisa (2003, 164ff.) explains graph w35/20 (zh ) as s , which he reads as c (secondarily); this would make this sentence read: []
10 11

reconstructing the wu xng

19.
w36

293

[Am] w37 To use the outer mind when interacting with others is keeping [appropriate] distance. To keep [appropriate] distance while being dignified is being respectful. To be respectful and not remitting is being stern. To be stern and fearful w37 is showing honour. To show honour without being arrogant is being reverent. To be reverent when widely interacting with others is ritual propriety.

w36

Subcanto 8
20. w38 [An] w39 [] [Ao] 1A w40 [Ap] 1B [] [Aq] 2A 2B w41 C [Ar] 12 w42 What is not succinct should not be enacted; what is concealed cannot be discriminated by the w38 Way. When there is a serious transgression, and it is punished severely, this is to be succinct. When there is a minor transgression, and it is pardoned, this is concealment.

to venerate the noble one and, secondarily, to honour the worthy one, this is rightness. This reading implies a clear preference for the nobility above the worthy one, who, despite this, must still be honoured (yet in secondary position only). The reading according to Hbi shng Jngmn sh bwguan 1998 instead suggests that honouring the worthy person is just on one level with venerating the worthy one to his level. Hence, in his search for appropriate men, the ruler is not necessarily urged to choose a man of high social standing. 12 Mo 304.

294

chapter eleven
When there is a serious transgression, and one refrains from w39 punishing it severely, this is not enacting [the Way properly]. When there is a minor transgression, and one refrains from pardoning it, this is not being discriminated according to the Way. To be succinct as a term is something like to soften raw silk by boiling w40 ; it is great but rare. To conceal as a term is something like to hide the what is to be hidden; it is small but numerous. To be succinct is an aspect of rightness. To conceal is an aspect of benevolence. [Just as] hardness is an aspect of rightness, softness is an aspect of benevolence. To be neither forceful nor pressing, neither hard nor [too] soft is what this is w42 about.13

21.1. [As] 1A 1B 2A w43 2B C ()-[]( ) () [At] The gentleman gathers great achievements;14 When able to advance in this, one can become a gentleman. Those unable to advance in this [should] all remain in their hamlets. As for what is great but rare, w43 [the gentleman] is able to take from it. As for what is small but numerous, [the gentleman] is able to take from it. He who shines brightly when reaching the Way of the gentleman is whom we call a worthy one. 21.2. w44 [Au] When the gentleman w44 understands [the worthy one] and lifts him up, this we call honouring the worthy one. When [the gentleman] understands and serves him, this we call one who honours the worthy person. The latter means that a scholar-knight honours the worthy one.
13 This sentence also appears in Odes, Shngsng: Chng f (Mo 304), which reads: [he was] neither forceful nor pressing; [he was] neither hard nor [too] soft (Karlgren 1950, 265, emended). The Gudin One passage reads qi (the fruit of a chestnut-leaved oak) in place of the odes qi hasty. I believe that this is a mere writing variant, referring to qi hasty. 14 Cf. Mngz, Wnzhng xi ( 5B1). In Mngz zhng y 1992, 397.

reconstructing the wu xng Subcanto 9

22.
w45 [Av] w46

295

Ears, eyes, nose, mouth, hands, and feet, these six [parts of the body] are slaves to the mind. When the mind says, So be it, none dares not to be with it.15 When [it says,] Agree, none dares not to agree. w46 When [it says,] Advance, none dares not to advance. When [it says,] Retreat, none dares not to retreat. When [it says,] Profound, none dares not to [take it as] profound. When [it says,] Shallow, none dares not to [take it as] shallow. When [parts of the body] are harmonised [through mind], then there will be congruence. When there is congruence, then there will be goodness. 23. w48 [Aw]
w47 w47

w45

When understanding something by the eye, we call it to advance in it. When understanding something by analogy, we call it to advance in it. When understanding something by example, we call it to advance in it. w48 [But] when understanding something [only] by an omen, it is [through] Heaven. When God on high regards you as worthy, be not duplicitous in your heart.16 This is what this is about. 24. [Ax] w49 [Ay]

Uttering the sound of wi is used to show agreement. Line w48/714 quotes Odes, Dya: D mng (Mo 236). The line reads: God on high looks down on you; do not be unfaithful in your heart (Karlgren 1950, 188, emended).
16

15

296

chapter eleven
When Heaven is bestowing on its people, this is [through] Heaven. If w49 people are bestowing on others, this is reliance.

25. w50 He who hears the Way and is delighted upon this is someone who is fond of benevolence. He who hears the Way and is fearful upon this is someone who is fond w50 of rightness. He who hears the Way and is reverent upon this is someone who is fond of ritual propriety. He who hears the Way and is joyful upon this is someone who is fond of virtue.

Notes on Text and Translation [A]: Following each line in this building block is a mark: ( ). Its precise function remains unclear, but it probably signals the end of the different positions stated. [B]: The top of strip w3 has broken off. Presumably three to four characters are missing. Due to the recurring pattern of the statements in this building block, it can be reconstructed as . . . virtuous conduct. The manifestation of the Wu xng from Mawngdu Three corroborates this assumption.17 [C]: In the last line of the first building block a change in the formal pattern appears. The structure of this passage suggests we call it conduct [only] instead of we call it virtuous conduct . This might be a scribal error, for the other lines in this building block read When it is not manifested internally, we call it conduct [only]. According to this reading, d would not be connected to the negation of the process. Yet, it can also be assumed that the author(s) of the text regard shng as a virtue that, in any case, is an expression of virtuous conduct (d zh xng ),
17 For the transcription of the Mawngdu Three Wu xng, see Mawngdu Hn m bsh zhngl xiao zu 1980, vol. 1. The Wu xng is written in columns 170351 of the so-called Laoz A manuscript.

reconstructing the wu xng

297

excluding the negation of the case per se. I regard this assumption as more likely to be true. In the Mawngdu Three manifestation of the Wu xng, the text consistently holds to the pattern of this building block and reads When it is not given shape internally, we call it conduct [only].18 The Mawngdu Three version lists the qualities rn () benev() ritual propriety, zh () wisolence, y rightness, l () sagacity from the Gudin One text in the dom, and shng following sequence: rn, zh, y, l, shng. [D]: The Mawngdu Three manifestation of the Wu xng exhibits the same sequence as the Gudin One text. [E]: The top of strip w6 has broken off. Presumably four graphs are missing. The editors of the Hbi Province Museum reconstruct this part as -()--, adding a sign for repetition after each of the three graphs as seen also in the preceding lines. Most interpreters follow this suggestion. This reconstruction is based on the Mawngdu Three text, which reads [] 175 () ()=== ()==== ( == () ) == ()== When the gentleman ( jnz ) lacks concern in his inner mind, then [he] will be without sagacity in his inner mind. When [he] lacks sagacity in his inner mind, then [he] will be without delight in his inner mind. When [he] lacks delight in his inner mind, then [he] will not be at ease. When [he] is not at ease, then [he] will not be joyful.19 The two passages differ in that the Mawngdu Three version speaks of sagacity (shng ), and the Gudin version of wisdom (zh ). [F]: The Mawngdu Three text is identical to this passage in the Gudin One Wu xng except that it has w (OC *m(r)o) where the Gudin One text has wng () (OC *m(r)a). Furthermore, the Mawngdu Three repeats this passage but replaces zh wisdom with shng sagacity in the second row.

18 19

Column 172. Columns 174175.

298

chapter eleven (see below). Graph w6/15 sh

[G]: Graph w6/12 y appears as

appears as . This line differs slightly in wording from the Mawngdu Three version. [H]: In the line as , the graph y is written () and ( ).

(), and the two graphs for zh as

[I]: The editors of the Hbi Province Museum suggest reading graph w7/20 as jn (OC *[g]r-s) to come close, to approach.20 Chn Wi offers a different explanation. He suggests reading it as xn (OC *qhr) delight, to be joyful. Quoting the Shuwn jiz, the line in question refers to the Sma fa. It reads he who is good takes joy in the goodness of the people.21 Reference to the Boshn corpus, however, suggests reading the graph in question as jn to approach.22 I could find no single instance where jn was used in the meaning of xn , and I follow the reading as suggested originally by the Hbi Province Museum editors. [J]: The Mawngdu Three version has ch ( 177) where the

Gudin One strips have (w8/13). The editors of the Hbi Province Museum follow the Mawngdu Three reading.23 This graph also appears in the Boshn corpus, where it is read ch throughout. But this reading cannot be verified with certainty. The structure of the subsequent lines is as follows: when X is not a, then it will not be b; when it is not b, it will not be c; etc. Hence, it seems to make more sense to read graph w8/13 as to grow or the like, just as in the next sentence. The same graph also appears on strip w13/2. Most interpretations follow L Lngs suggestion to read it as ch .24 [K]: Following the line () - ()- (w8/1418), the Mawngdu Three text adds ()== (177).

20 21 22 23 24

See Hbi shng Jngmn sh bwguan 1998, 147. See Chn Wi 2003, 52. Tng Rnshng 1995, 794. See Hbi shng Jngmn sh bwguan 1998, 147. See also under [Q].

reconstructing the wu xng

299

[L]: At line w9/38 () () the Mawngdu Three version has jng essential (178) instead of qng clear. [M]: Graph w10/2 (nng ), which also appears in the quotation of the Odes, is written differently in the previous instances. as b .25 Ikeda reads it as [N]: Most editors read graph w10/9 b .26 The graph differs markedly from other instances of the characters b and b, and the passage also differs from that appearing in the Odes (Mo 14), and so these are of no help in identifying the present graph. [O]: The top of strip w11 has broken off. Presumably one character is missing. The Mawngdu Three manifestation of the Wu xng has yu at this place (180). This passage also appears in the Odes, where it differs noticeably from the Gudin One version.27 The Mawngdu Three version introduces the quotation (strips w911) with sh yu in an ode it is said (179). [P]: Strip w11 is broken. The editors of the Hbi Province Museum assume that two characters are missing. In accordance with the Mawngdu Three version, they add the graphs y to the previous sentence and b to the subsequent one.28 However, when looking at the strips, it looks as if the missing part contains at least three characters. Accordingly, I add the character f and a mark that signals the end of the passage in my reconstruction.
Image of the missing part of strip w11 compared with strip w10

Cf. L Lng 1999, 488; T Zngli and Li Zuxn 2001, 386. See Ikeda Tomohisa 1999, 23. 27 See Mo 14: when I have not yet seen the lord, my grieved heart is sad; but when I have seen him, when I have met him, my heart will then be pleased (Karlgren 1950, 9, emended). 28 See Hbi shng Jngmn sh bwguan 1998, 149.
26

25

300

chapter eleven as ch to investi-

[Q]: Qi Xgu suggests reading graph w13/2 gate. I follow his suggestion.

[R]: Graph w13/6 has been transcribed as 29 (OC *[n]). Most scholars read in the sense of wn (OC *un) gentle. Chn Wi suggests that this graph consists of the phonophoric (OC *[g] m or *[g]r[]m; as the latter is a place-name, it might be the more likely reading) and the signific xn heart.30 H Lny states that the graph has as its phonophoric.31 Phonologically, this is somewhat unlikely. The reconstruction of is *un and that of is *[g]r[]m. () (w16/310) also () () [S]: The passage appears in the Gufng: Shji (Mo 152): the Shji birds offspring is seven; the polite man, the gentleman, unified is he in his deportment. The Mawngdu Three version also has the following two lines preceding: (184) the Shji birds offspring is seven. [T]: The top of strip w18 has broken off. Based on the ode quoted, it can be reconstructed as {I looked after her, yet I could not see her}. [U]: The top of strip w18 has broken off. Judging from the manuscript counterpart Mawngdu Three (186), it seems that the missing passage originally contained two graphs, which can be reconstructed as d y . Due to the length of the lost part, it can be assumed that these graphs were followed by markings on the strip. [V]: The top of strip w19 has broken off. The missing passage probably contained five graphs. Taking the Mawngdu Three version into account (186), we might reconstruct [] {there is [always] something with which} the gentleman {begins, but there is nothing with which} [he] ends.

29 30 31

Ibid. See also Zhng Gungy [Cheung Kwong-yue] 1999, 539. See Chn Wi 2003, 5253. See H Lny 2001, 1130.

reconstructing the wu xng The shapes of graphs w18/11 ( ) and w19/1 edly. They are both read in the sense of zhng . (

301 ) differ mark-

(+) (*l e-s) is presumably used to express [W]: Graph w19/4 the word <shng > (*l e). Mark Csikszentmihalyi hypothesises that this is more than just a phonetic loan. He assumes instead that the connection is more like the relationship between l joy and yu music, and he considers this to be an unstated philosophical argument.32 I doubt this.33 Phonetic loans appear throughout the entire Gudin One corpus, so that one may well speak of scribal tendencies but not of well-defined standards. Scribal mistakes throughout the texts of the Gudin One corpus in fact suggest that the scribe(s) did not necessarily fully understand what they were copying. Philosophy and writing must be conceived of as different activities, and the philosopher most likely was not the one who fixed a particular manifestation of a philosophical text on bamboo. To assume that the scribe(s) in question was punning with loans would imply that he fully grasped the philosophical subtleties of this text and also had the intellectual ability to play with it, if author and scribe were not indeed considered one and the same person. [X]: Following w20/4, the bamboo strip is broken. In accordance with the Mawngdu Three Wu xng (188), the missing part can be reconstructed as do y . The graphs r and tin are well-nigh indistinguishable in early Chinese manuscripts. Following the structure of the argument as well as the Mawngdu Three text of this passage, graph w20/3 be read as tin heaven. should

[Y]: Ikeda Tomohisa argues that this line () ()w21 ()reflects a scribal mistake. Following the Mawngdu Three version, he argues that marks for repetition should be inserted into the line . Ikeda further assumes that itself is not a causal series.34 I doubt Ikedas line of reasoning.

32 33 34

See Csikszentmihalyi 2004, 169. Cf., however, Behr 2006 on the etymology of this word family. See Ikeda Tomohisa 1999.

302

chapter eleven

The first line of this passage does not form an argumentative chain but works parallel to the two features introduced earlier on. The text identifies these features as the characteristics of either the sagacious or the worthy one. [Z]: This passage does not occur at the same place in the text in the Mawngdu Three manifestation of the Wu xng. In the Mawngdu Three Wu xng it follows the discussion of the three charismatic aspects rn, y, l. [Aa]: The Hbi Province Museum editors read graph w21/11 ( ) as bin (OC *pro[n]-s).35 As Csikszentmihalyi states, both Ikeda and Png P connect this graph to a complex of graphs carrying a meaning similar to lin (OC *(m.)ron-s) feeling, affection.36 Phonetically this is close enough to fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. I follow their suggestion. (d ) is read as zh in the Mawngdu [Ab]: Graph w21/23 Three version. It is therefore sometimes understood in the sense of rightness in carrying out official duties; a core aspect of rightness.37 L Lng and Wi Qpng suggest reading w21/25 as s .38 S should be understood in opposition to zh, above. It thus refers to statements such as the following one from the transmitted Laoz (ch. 58): this is why the sagacious person makes square but does not trim [others], makes clean but does not injure [others]. [He] straightens but does not intimidate [others], enlightens but does not investigate [others].39 Unlike the statement in the Laoz ( x y), this passage does not oppose terms directly but forms an argumentative chain. Accordingly, I hesitate to read a term with a negative connotation here.

See Hbi shng Jngmn sh bwguan 1998, 151, n. 26. See Csikszentmihalyi 2004, 292. See also Ikeda Tomohisa 1999, 32; Png P 2000, 27. 37 See Csikszentmihalyi 2004, 293. 38 See L Lng 1999, 489; Wi Qpng 2000, 28. 39 Wagner 2003a, 318320, emended.
36

35

reconstructing the wu xng

303

Ikeda Tomohisa, Wi Qpng, and T Zngli and Li Zuxn read the graph as xi to disperse, to release.40 The editors of the Hbi Province Museum explain the graph, in accordance with the Mawngdu Three version, as l to intercept, to which reading Li Xnfng also refers.41 Mark Csikszentmihalyi notes that this graph rarely appears in early transmitted texts. The commentator on the Hnsh, Jn Zhu (fl. ca. AD 208), recognised l as an older form of li to block, to obstruct, for which reason Csikszentmihalyi translates it as resistant.42 I follow his suggestion. [Ac]: This unit (14.1) comes significantly later in the Mawngdu Three Wu xng. See my discussion in chapter 3. as tin [Ad]: Ikeda Tomohisa has proposed reading w26/22 heaven instead of r .43 The editors of the Hbi Province Museum and most commentators follow this suggestion. Mark Csikszentmihalyi suggests that the short sentence from w26/19 () () <> w27 () is a misplaced comto w27/2 mentary because it seems to interrupt the flow of this passage.44 As discussed in chapter 3, this is rather unlikely. [Ae]: Mark Csikszentmihalyi notes that the phrase () () appears reversed in the Xnz, Jndo.45 The passage reads: therefore, the gentleman is respectful in his attitude to ritual propriety and at peace with it. [Af ]: The top of strip w30 is broken. Judging from the broken piece, three to four characters are missing. According to the underlying structure, either or should be reconstructed here.46 [Ag]: For graph w29/11 in the Gudin One Wu xng, the Mawngdu Three version reads yu (200). T Zngli and
40 See Ikeda Tomohisa 1999, 33, n. 36; Wi Qpng 2000, 28; T Zngli and Li Zuxn 2001, 400. 41 See Hbi shng Jngmn sh bwguan 1998, 152, n. 27. 42 See Csikszentmihalyi 2004, 293. 43 See Ikeda Tomohisa 1993, 364. 44 See Csikszentmihalyi 2004, 397. 45 Ibid., 297. 46 See Hbi shng Jngmn sh bwguan 1998, 153, n. 35.

304

chapter eleven

Li Zuxn read w29/11 as ju , here to revive.47 The two graphs share the same phonophoric. Instead of graph w29/9 ( bng country), the Mawngdu Three version reads gu (column 199). This seems to be due to a taboo against writing the personal name of the founder of the Hn dynasty, Li Bng (256195 BC).48 [Ah]: The top of strip w30 is broken. Here the Wu xng draws on Odes, Dya: Wn Wng (Mo 235), which reads: when King Wn was on high, he illuminated Heaven. Accordingly, the missing part can be reconstructed as . [Ai]: Instead of the line () () () as in the Gudin One text of the Wu xng (w31/1218), the Mawngdu Three version reads () (202), thus adding the character zh and so creating a different chain of development. as zhn [Aj]: Qi Xgu identifies the two graphs w32/2324 qin ; the Mawngdu version reads yn qin (191; zhn is an allegro form of and ). The Gufng: Zhh (Mo 110) reads: May he be cautious, that he may come back and does not remain [there]. With the exception of Wi Qpng,49 most interpreters follow Qis suggestion.50 [Ak]: The Hbi Province Museum editors suggest reading w33/18 as yu to follow. Qi Xgu interprets it as a variant of j , which should be read j to follow. as [Al]: The Hbi Province Museum editors read w34/1920 qing y strong and powerful. Most interpreters follow this reading.51 Png P notes that this combination also occurs in the Odes, Dya: Zhng mn (Mo 260):

See T Zngli and Li Zuxn 2001, 410. On the custom of taboo observances during the Hn, see Mansvelt Beck 1987. 49 See Wi Qpng 2000, 37. 50 See, e.g., L Lng 1999, 489, 491492; Ikeda 1999, 39; T Zngli and Li Zuxn 2001, 412414. 51 See, e.g., L Lng 1999, 490; Png P 2000, 56; T and Li 2001, 415417.
48

47

reconstructing the wu xng

305

[Zhongshan Fu], he did not intimidate the widow and the poor and did not fear the strong and powerful. [Am]: The identity of graph w36/19 is still uncertain. In the

Mawngdu Three version it is read as xi to remit ( 194). The Hbi Province Museum editors suggest reading it as ji , but Qi remains hesitant about this suggestion.52 L Lng follows the Mawngdu Three reading for this graph.53 Wi Qpng proposes reading the graph as gng to till.54 T and Li suggest reading it as gu go to, transgress.55 In the Gudin One Wu xng, the combination (such as in w2324 w34 ) always appears in the construction to be/do X do/be Y; that is Z. Thus, what is named subsequent to always seems to be another entity but never the result of X. By implication, the suggestion made by T and Li to read the graph as gu to indulge, transgress can be ruled out. I follow the Mawngdu Three version at this point. [An]: Graph w37/17 appears as dng . Apparently it is written for jian (OC *kr[a]n). Most interpreters read jian as a loan for jian (OC *kre[n]) dignified, sedate; grave demeanour. This is the case throughout the Wu xng (see, e.g., w35/7 ). The word jian is used structurally identically to the word n to conceal; what is concealed to construct the argument of this passage (both words appear twice and combined with a definition of terms). It seems rather unconvincing to read graph w37/17 as jian and graph w38/10 as jian <jian >, as suggested by the editors of the Hbi Province Museum.56 Note that the second appearance of the word jian is written , thus similarly to w37/11, yet markedly different from other instances of the word <jian >.

52 See Hbi shng Jngmn sh bwguan 1998, 153, n. 48. The editors assume that the graph in question is corrupt. 53 See L Lng 1999, 490, 492. 54 See Wi Qpng 2000, 42. 55 See T Zngli and Li Zuxn 2001, 417419. 56 See Hbi shng Jngmn sh bwguan 1998, 150.

306

chapter eleven

The Hbi Province Museum editors read graph w37/23 (the last graph on strip w37) as ch . Qi remains hesitant about this suggestion.57 L Lng follows their reading.58 The Mawngdu Three version has either bin (OC *[b]ren) to discriminate or bin (OC *[b]ren) to dispute (204), and I follow the first of those two suggestions here. [Ao]: The upper part of strip w39/4 is missing. Based on the Mawngdu Three Wu xng (204) as well as the remaining fragment of the graph, it can be reconstructed as xng . [Ap]: According to the editors of the Hbi Province Museum, w39/23 ( lin; OC *[r]en-s) should be understood as a borrowing for jin (OC *kre[n]). The Mawngdu version reads h (OC *m-kaj-s), which T Zngli and Li Zuxn assume was used instead of hng (OC *[g]ra) balance,59 and which Li Xnfng (2000, p. 131) believes to be a homophone for h (OC *[g]aj).60 It is read here in the sense of dn to carry (which makes it h [OC *[g]aj]). Based on the Shuwn jiz tngxn dngshng, T and Li suggest that lin (OC *[r]en-s) was used for jian (*kr[a]n).61 The Hbi Province Museum editors state that as in the Mawngdu Three version ( 204), graph w40/4 (yn ; OC **ra[n]-s) late should be read as han (OC *qha[r]) scarce, few.62 [Aq]: The character following w40/13 only appears as a repetition of the former graph. Nonethelessor because of thisthe correct reading of this graph is as yet uncertain. Wi Qpng reads the second character as n intimate; to approach. He sees justification for this in the Zuo zhun, year 25 of the reign of Duke Xing , which reads and know to conceal the familiar.63 The Mawngdu Three text reads n (205). Png P argues that the sec57 58 59 60 61 62 63

Ibid., 154, n. 50. See L Lng 1999, 490. See T Zngli and Li Zuxn 2001, 421. Li Xnfng 2000, 131. T Zngli and Li Zuxn 2001, 421. See Hbi shng Jngmn sh bwguan 1998, 154, n. 53. See Wi Qpng 2000, 45.

reconstructing the wu xng

307

ond character should be read as t deprived; to do evil, thus referring to the earlier seen xiao zu minor crime.64 If we accept the reconstruction of the character as different from the previous graph, it implies that the latter graph must be a homophone, and the scribe, writing from dictation or memory, did not know that he should write a different graphor the scribe saw another written master copy with a repetition mark instead of an individual character. As detailed in chapter 5, the latter scenario is unlikely. Since we still lack information about the process of copying, reading, and writing in ancient China, I shall read the second character as identical with the first. Chn Wi reads this line with two identical graphs. He justifies this by noting a similar use of doubling terms in the Gudin One manuscript Li d (strips l3233), in which rng rng can be viewed as analogous to n n in the Wu xng.65 The editors of the Hbi Province Museum state that graph w41/17 must be a variant of zhn (OC *t[n]), which in turn must be a borrowing of zhn (OC *t[n]) numerous. This is substantiated by the Mawngdu Three Wu xng, which reads zhn (205), as well as by the recurring sentence on slip w43, which reads () (strip w43/811). [Ar]: It seems that w41/5 (qing ; OC *[g]a) must be reconstructed as gng (*ka) hard. The two fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. This assumption is further substantiated by the fact that the sentence w41/1821 appears identically also in Odes, Shngsng: Chng f (Mo 304): [he was] neither forceful nor pressing; [he was] neither hard nor [too] soft.66 The Mawngdu Three version is nearly identical with this line. The same passage is quoted, but the introductory formula it is said in an ode is added (204205). [As]: The line w42/26 () () also occurs in similar fashion in the Mngz , Wnzhng xi (5B1): Master Kong is said to have gathered great achievements.

64 65 66

See Png P 2000, 71. See Chn Wi 2003, 5658. Karlgren 1950, 265, emended.

308

chapter eleven

[At]: For w43/1516 p f ( f) - () the Mawngdu Three version reads suo l (l) = () (207). L Lng reads graph w43/16 as x ; for w43/15 he suggests p foot.67 Ikeda reads sou to seek for this graph. He also regards the repeated graph l (thus, l l ) as onomatopoeia for barking.68 Png P reads the line in accordance with the Mawngdu Three version. He argues that is a variant of the name Suol Shn , a thief who later became a disciple of Mz.69 As Mark Csikszentmihalyi notes, Suol was indeed a surname in Warring States China, but he thinks that it is unlikely that the characters here indicate a proper name.70 I agree with Csikszentmihalyi. In this instance, a proper name would be out of place. Wi Q png identifies graph w43/15 as ya (OC *[] ra), which he reads as h brilliant (qhrak). At least phonologically this is plausible (the brackets around *[] indicate that a uvular initial with a nasal prefix might have been involved; so the connection with *qhrak would be plausible). Wi further explains that if this graph is combined with l a variant of w43/15, l we get shining brightly, which can be understood as a description of the worthy one.71 I follow his suggestion. Qi Xgu suggests reading w43/18 as zh .72 Also, the Mawngdu Three version reads y (207) for what appears as zh in the Gudin One copy (w43/18). [Au]: Following the sentence w44/1419 () () (), the Mawngdu Three version adds . This line does not appear in the Gudin One version. is still unresolved. The [Av]: The proper reading of graph w45/11 Mawngdu Three Wu xng has y servant (209). L Lng follows this reading.73 Ikeda Tomohisa notes that the graph is written differently in the Mawngdu version. The reading of y servant for the

67 68 69 70 71 72 73

See L Lng 1999, 490. See Ikeda Tomohisa 1999, 49. See Png P 2000, 74. See Csikszentmihalyi 2004, 306. See Wi Qpng 2000, 49. See Hbi shng Jngmn sh bwguan 1998, 154, n. 57. See L Lng 1999, 490.

reconstructing the wu xng

309

Gudin One Wu xng therefore cannot be justified on palaeographical or phonological grounds.74 T and Li read it as fan to return.75 I read graph w45/11 in the meaning of the Mawngdu Three text. [Aw]: The editors of the Hbi Province Museum point out that graph w48/9 ( ) bears a strong resemblance to an inscribed graph seen on a vessel of King Cu of Zhngshn.76 The Mawngdu Three version reads ln to look down upon (favourably). Qi Xgu is in doubt whether this graph should be seen as a scribal error for ln .77 This reading makes sense, as the ode cited reads ln , too.78 [Ax]: L Lng and Wi Q png read w48/19 as tin and not as d as proposed instead by the editors of the Hbi Province Museum.79 The Mawngdu Three passage (212) has , which corroborates L Lngs and Wi Qpngs suggestion. Moreover, according to Wi, is a combination seen throughout early texts.80 Also, reading tin instead of d offers a clear and simple interpretation of this passage. Hence, I follow their suggestion and read this graph as tin . [Ay]: How to read w49/5 is still unresolved. L Lng suggests reading it as xi (OC *[g]r[a]p). According to L, this should be close in sound and meaning to x (OC *s-m-l[]p). But this is not the case.81 Png P adds that L Lngs suggestion would in fact help to contrast the two issues natural versus learned.82 Despite this observation, the suggestion made is problematic on phonological grounds. Li Xnfng reads the graph as j to depend, to rely on,83 and I follow Li.

See Ikeda Tomohisa 1999, 46. See T Zngli and Li Zuxn 2001, 426, 7. 76 See Hbi shng Jngmn sh bwguan 1998, 154, n. 64. 77 Ibid. 78 Line w48/714 is a quotation of Odes, Dya: D mng (Mo 236), which reads: God on high looks down on you; do not be unfaithful in your heart (Karlgren 1950, 188, emended). 79 See Hbi shng Jngmn sh bwguan 1998, 151. See also L Lng 1999, 492; Wi Qpng 2000, 54. 80 See Wi Qpng 2000, 54. 81 See L Lng 1999, 492. 82 See Png P 2000, 85. 83 See Li Xnfng 2000, 167.
75

74

CHAPTER TWELVE

RECONSTRUCTING THE XNG Z MNG CH This chapter provides the text and translation of the argument-based text Xng z mng ch, followed by philological notes. The philological discussion not only refers to the text as reconstructed but also points to differences and similarities between the two manifestations of the text: from Gudin One and from the Shnghai collection of Chu manuscripts. In both the transcription and the translation, a superscript letter x refers to the manuscript Xng z mng ch. Discussing the Shnghai Xng qng ln I use xq to refer to the various bamboo strips. Text and Translation The Core Text: Canto 1 Canto 1 contains twelve pericopes. Pericopes 212 can be grouped into three subcantos, headed by the introductory pericope 1. Pericope 1: Introduction
1.
x1 [A] [B] x2 [C]

[D] x3 [E] [][F ] [G] [] x4 [H] [I] [] [J] x5 [K]

312
x1

chapter twelve

Generally speaking, even though all humans have a nature, they have no settled will in their minds. It requires the things; only thereafter can it rise. It requires delight; only thereafter can it evolve. It requires repetitive practice; only thereafter can it x2 settle.

Emotions such as rejoicing, anger, grief, and sadness are human nature. When they appear on the outside, then this is caused by the things [in the world]. Human nature is brought forth by decree; decree [in turn] x3 is sent down from Heaven. The true ethical code (do) begins with the unshaped feelings; unshaped feelings [in turn] are begotten by human nature. In the beginning [the true ethical code (do)] approximates the unshaped feelings; in the end [it] approximates rightness. {He who} understands {the unshaped feelings may} x4 manifest them [at the outside]; he who understands rightness may take it in. To love and to hate is human nature. That which one loves and hates are the things [in the world]. To be good or not { good is human nature}. x5 That which one considers good or not good is [determined by] conditional forces.

Subcanto 1
2. [L] [ x6 ] [M] [] [N] Generally speaking, that which becomes the dominating aspect of human nature is caused by the things [in the world]: Bronze and stone have a sound, [yet] {if they are not struck, they will not x6 sing}. {This is just like with humans}. Even though {they} have a nature, if [their] minds fail to accept it, it will not be manifested. 3. [] [] x7 [O] [] x8 [P] Generally speaking, mind has a will[yet] if not nourished, it will not {be settled}. {That mind cannot} x7 evolve on its own is just like the inability of the mouth to speak on its own.

reconstructing the xng z mng ch

313

After oxen are born they grow [big]; after geese are born they stretch [their necks]. It is their nature {which makes them be like thisas for humans,} x8 in contrast, it is learning that eventually shapes them. 4. [Q] [R] x9 [S] [T] Generally speaking, of the things, there is none that is not different. That the hard becomes straight is because hardness takes it up. That x9 the soft becomes flexible is because softness takes it up. That human nature is one within the four seas, [yet] in the application of the mind each [man] differs, is something brought forth by education.

Subcanto 2
5. x10 [U] [V] Generally speaking, as for human nature x10 there is something that moves it, something that conforms to it; Something that interacts with it, something that grinds it; Something that makes it manifest [at the outside], something that nourishes it; [And] something that makes it grow. 6. x11 [W] [X] [Y] x12 [Z] Generally speaking, that which moves human nature x11 are the things [in the world]. That which conforms to human nature is delight. That which interacts with human nature are the causes. That which grinds human nature is rightness. That which makes human nature manifest [at the outside] are the conditional forces. That which nourishes x12 human nature is repetitive practice. That which makes human nature grow is the true ethical code (do).

314

chapter twelve

7. [Aa] [Ab] x13 [Ac] x14 [Ad] [Ae] Generally speaking, that which can be seen is what we call the things [in the world]. That which generates satisfaction in oneself is what we call delight. That which is the power of x13 the things is what we call conditional forces. That which makes [something] happen is what we call causes. Rightnessthis is the status indicator for the [different] types of goodness. Practice x14 this exists for practising ones human nature. The true ethical code (do)this is the Way (do) of the [different] groups of things.

Subcanto 3: Digression
8. x15 [Af ] [Ag] x16 [Ah] [Ai] [Aj] x17 [Ak] [Al] x18 [Am] x19 [An] [Ao] ? x20 (?)
x21

[] [Ap] [Aq]

reconstructing the xng z mng ch

[Ar] [X] x22 [As] [At] [Au]
x23

315

[Av]

Generally speaking, as for the true ethical code (do), the skills of the mind are the dominant [feature]. The true ethical code (do) diverges into four skills, [but] only x15 the Way of humans is the one that can be followed. As for the [other] three skills, one can only talk of them. Odes, Documents, rites, and musicin every case their first appearing was given birth by x16 man. Odes came into being by acting them out. Documents came into being by speaking them. Rites and music came into being by exalting them. The sagacious persons [then] juxtaposed them according to x17 their categories, collated and joined them together. [They] beheld them in their temporal sequence to arrange them in their proper sequence. [They] shaped their meaning to regulate and pattern them. [They] ordered x18 the feelings [expressed in them] to be manifested [at the outside] and to be internalised. Only when this was achieved, did they turn toward teaching. Teaching is that by which [the sagacious persons] generate moral force (d) in [their] centres. As for rites, they were created on the basis of unshaped feelings x19 [but] eventually, they [also] might stimulate them. Relying upon methods according to each casethe proper sequence of what comes first and what comes last thus befitted the Way (do). Once the proper order of actions was x20 regulated, they were culturally refined. To extend this to ones manner and appearance, that was the reason for cultural refinement and regulations. [Hence], the gentleman finds beauty in their unshaped feelings and values {their rightness}. [He] x21 considers their regulation as good and esteems their manners. [He] finds joy in their true ethical code (do) and delights in their teaching. Therefore, he shows respect for them.

316

chapter twelve

Bending [his hands] {is how respect for X is expressed}; x22 the repetition of it is [cultivated] pattern. Offering presents of silk is how trustworthiness and evidence are established. Declining them [should be done in a way] befitting the true ethical code (do).
x23

Laughter is the shallow march (= surface aspect) of rejoicing. Music is the deep march (= the underlying aspect) of rejoicing.

9. [Aw] x24 [Ax] x25 [Ay] <<>><<>> <<>><<>> x26

1 [Az] x27 [Ba] [Bb] <<>><<>>[Bc]

Generally speaking, all sounds emanating from unshaped feelings are trustworthy; only when they enter and agitate the heart of man do they become profound. x24 [Therefore,] to hear the sound of laughter is preciouswhen it comes to it, then there will be rejoicing. To hear the sound of chanted songs is highly gratifyingwhen it comes to it, one will become elated. To listen to the sound of lute and zither is x25 excitingwhen it comes to it, one will have to sigh. [And likewise,] to watch the [performance] of the [ritual dances] Li and Wu [makes one] solemnwhen it comes to it, then one will be stirred.2 To watch the [performance] of the [ritual dances] Sho and Xi is incitingwhen it comes to it, x26 then one will become humble.3

The character s seems to be a particle for emphasis here. The Li (Mo 295) and Wu (Mo 285) are part of the Zhu hymns that praise King Wus victory over Shng. The odes were part of a larger performance together with ritual dances. Accordingly, the passage says watching. 3 The songs of Sho and Xi are the music of Shn and D Xi (or Yu ). They were performed together with ritual dances. According to Lio Mngchn (2002b), the lyrics of the songs also state that Wu and Yu have their own deficiencies; accordingly, the character jian should be read as humble, modest or self-deprecating.
1 2

reconstructing the xng z mng ch

317

When moving the mind by cantillating aloud a thoughtthe sighing sound wei will follow. It is long that it occupies the second position, [and] when by returning to the good and starting from the beginning, one x27 has to be sincere, [and] when by entering and coming out it is compliantthis is something which initiates ones moral force. The music of Zhng and Wi, in contrast, is not of this sound; but one [easily] indulges in it. 10. [Bd] <<>><<>><<>><<>>[Be] Generally speaking, x28 music of old (= the music of Shn and Y) highly exalts ones mind; the latter music [of King Wu of Zhu] exalts ones ambitions. They are both for educating the people. [The latter] music [of King Wu] Li and Wu is that of grasping [ambitions]; [the former] music [of Kings Yo and Shn] Sho and Xi is that of unshaped feelings. 11.
x29 4 [Bf ] [Bg] x30 : x28

[Bh] x31 [Bi]

x29

Generally speaking, most refined music certainly is necessarily sad crying is also [an expression of being] sad. They both reach out to their corresponding feelings. The corresponding natures of grief and joy are close to each other. It is for this reason that their minds x30 do not deviate far:

The way in which crying moves the heart is encroaching and shattering [but in] its blazing fierceness [it] is all-consuming, and one remains grieving till the end.

4 Zh means the high point. Yu could be explained as both music and pleasure. However, since the following part clearly talks about waning and grief, yu at this point should be understood as pleasure, the counterpart to grief.

318

chapter twelve
The way in which music moves the heart x31 is profound and deep, dense and delightful[but in] its blazing fierceness [it] is like a steady flow that leads to grief, and one becomes mournful in thoughts.

12. x32 [Bj] [] [Bk]

x33 [Bl] [] [Bm] ()[Bn]

Generally speaking, to have thoughts of sorrow results in sadness thereafter. x32 Generally speaking, to have thoughts of joy results in delightedness thereafter. Generally speaking, the impact of ones thoughts on mind is extreme. Sighing, this is a mode of [expressing] thoughts. When the sound of it changes, then {the mind will follow along as well}. x33 [And] when the mind changes, then the corresponding sound of it also [changes] accordingly: Sighing proceeds from grief. Chirping proceeds from joy. Murmurs proceed from the tones [of music]. Singing out loud proceeds from the mind.
[end of overlap of canto 1]

[Bo] [Bp] [Bq] x35 () [Br]

x34

reconstructing the xng z mng ch

x34

319

When there is rejoicing, then there will be delight. When there is delight, then there will be enthusiasm. When there is enthusiasm, then there will be chanting. When there is chanting, then there will be waving [of ones hands]. When there is waving [of ones hands,] then there will be dancing. [Hence,] dancing is the end result of rejoicing. When there is exasperation, then there will be sorrow. When there is sorrow, then there will be grief. When there is grief, x35 then there will be sighing. When there is sighing, then there will be the beating of ones breast.5 When there is the beating of ones breast, then there will be jumping up and down [as an expression of ones grief ]. [Hence], jumping up and down is the end [result] of being exasperated.

The Application of the Theory: Canto 2 Canto 2 contains two clusters of pericopes that remain stable in both manifestations of the text: pericopes 1315 (1921 in the Shnghai Xng qng ln) and pericopes 17 to the first part of 20 (or 1518 in the Shnghai Xng qng ln). Subcanto
13.
x36 [Bs] ; x37 ;

[Bt] x38 [Bu] [ ] [Bv] [Bw] [Bx] x39 [By] [Bz]

5 For the translation of p as beating ones breast, see the ode B zhu (Mo 26): in the quietude I brood over it, awake I knock and beat my breast (Karlgren 1950, 16).

320
x40 [Ca]

chapter twelve

[Cb] x41 [Cc] x42 [Cd]

x36

Generally speaking, whenever one learns, to seek for the [right] mind is difficult; by following what is brought about by it, one will already have come close to obtaining the [right] mind. But this is not as good as inviting [the right mind] with music. x37 Even if one were capable in ones tasks, if one were unable in ones mind, this was not to be honoured.

If seeking the [right] mind has something artificial to it, one will fall short of obtaining it. That humans cannot employ artificial means [for seeking the right mind] x38 is something that can be ascertained. Before having tried ten times, ones mind will necessarily be exposed through it. When investigating what can be seen of it, how could one possibly miss the unshaped feelings? To be fair is a mode of rightness. x39 Rightness is a mode of respect. Respect is the regulation of the things [in the world]. Genuineness is a mode of benevolence. Benevolence is a mode of human nature. Human nature is what eventually gives birth to it (i.e., benevolence). Fidelity (zhng) is a mode of trustworthiness (xn). Trustworthiness is a mode of unshaped feelings. Unshaped feelings originate from human nature. Of the categories of love there are seven[yet] only the love of human nature is close to benevolence. Of the categories of wisdom there are five[yet] only x41 the Way (do) of rightness is close to fidelity. Of the categories of hate there are three[yet] only the hate of the nonhumane is close to rightness. [Finally], what constitutes the true ethical code (do) is fourfold[yet] only the Way (do) of humans x42 can be followed. 14. [Ce] [Cf ]

reconstructing the xng z mng ch

x43 [Cg] [Ch] [Ci] x44 [Cj] [Ck] [Cl] x45 [Cm] [Cn] x46 6 [Co] [Cp] x47 [Cq] [Cr] x48 [Cs]

321

Generally speaking, impatience in the application of mind will be accompanied by excessive thinking. Haste in the application of knowledge will be accompanied by excessive worries. Extremes in the application of unshaped feelings x43 will be accompanied by excessive grief and pleasure. Privileging the role of the body will be accompanied by excessive [seeking for] pleasure. The exhausting application of strength will be accompanied by excessive [seeking for] profit. The eyes affection for x44 female beauty, and the ears joy in sounds, these are the airs of pent-up delightednessman would easily die for them. If in ones comportment as a human being one is acting as very restrained x45 but does not in fact have a very simple mind, then this is [only] ornamentation. [And] if it happens that in the very simple comportment as a human being one does not in fact intend to make the perennial endangerment nonexistent, then this is boundless. As for a mans skilful x46 words and advantageous speeches, if one does not [also] have a very unornamented mind, then this will be ephemeral.

Li functions as a loan for f .

322

chapter twelve
As for a mans state of being delighted about getting along and being comfortable with himself, if this is not actually coupled to the x47 unshaped feelings of striving to create, then it will be delusion. If it happens that in ones comportment as a human being one acts as if satisfied, failing to be taken care of is impermissible. If it happens that in ones comportment as a human being one acts as if recondite, x48 failing to be corrected will not suffice.

15. [Ct] [Cu] x49 [Cv] [Cw] [Cx] [Cy] Generally speaking, a humans artificial activities are hateful. When one is artificial, then one is bound to regret it. When one regrets, then one is bound to be cunning. When one is cunning, no one will have [friendly] x49 relations with you. Caution, [in contrast], is a mode of benevolenceyet if it is flawed, one will not be hated. Hastiness, [however], is a mode of contrivanceif it has flaws, one will be blamed. It is truly the case that, if men are not cautious, flaws are bound to occur.
[strips x50 through x59/9] [Cz]

17. [Da] x60 [Db] Generally speaking, in delighting others, one ought not to be pettyminded [towards them]as a person, one has to go along with them. When words reach [the things] (i.e., they are to the point), they x60 are illuminating. Raise them and be without artificiality! 18.
[Dc]

Generally speaking, do not be fierce in relationsit will necessarily lead to triviality.

reconstructing the xng z mng ch

19. [Dd] x61 7 [De] 8 [Df ]

323

Generally speaking, when summoned, be without fear, [but also] do not raise a solitary voice. When x61 dwelling alone, then one should practise what father and elder brother have found pleasure in. If only there are no greater calamities, when minor irregularities occur, this can be endured. Once they are over, refrain from speaking about them again.
[end of overlap]

20.

x62 x62

[Dg] Generally speaking, concerning affairs of sorrow and calamity, one should wish to tackle them; concerning affairs of joy, one should wish to postpone [them].

Notes on Text and Translation The Xng z mng ch can be split up into 20 pericopes. Canto 1 (strips x135 of the Xng z mng ch; strips xq121 of the Xng qng ln) contains pericopes 112. Core Text: Canto 1 [A]: On the graph xng nature (OC *[s]e-s) (x1/5) written with the phonophoric shng (OC *sr) and the signific m ( ) (x1/5), see Henri Maspro 1933. For xng nature, the Shnghai Xng qng ln predominantly uses the phonophoric shng (OC *sr) (xq1/5). Shng is a common loan for xng in texts from the Warring States period. Qi Xgu suggests reading x1/8 din to put down (also read dng) as dng to determine.9 Dng is interchangeable with dng , so there is basically no need to replace one word

7 8 9

Reading x in the sense of to follow. Reading x61/10 w do not as w have no. See Hbi shng Jngmn sh bwguan 1998, 182, n. 1.

324

chapter twelve

by the other. For dng , the Shnghai Xng qng ln reads zhng righteous (OC *[t]e-s) (xq1/8). Dng can be reconstructed as *m-te[n]-s or *m-te-s set forth; *-n is found instead of *- very frequently after the front vowels *i and *e. Apparently, *-i changed to *-in in many dialects, and in some dialects *-e also changed to *-en. Following the last character of this line, zh intention (OC *ts), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [B]: Following the last character of this line, xng (OC *-[g]ra) to take effect, a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [C]: Following the last character of this line, dng , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. At this point, the Shnghai Xng qng ln also reads dng instead of zhng , as used in the first line. Given the fact that the building block under review comes full circle, it seems reasonable to assume that the previous use of (OC *te-s) in the Shnghai Xng qng ln should also be regarded as a borrowing for dng (OC *m-te-s). [D]: The similarity of this line with that from the Shnghai Xng qng ln, including in terms of writing peculiars, is striking. Only the graphs a (Gudin One: ; Shnghai Xng qng ln: ) and q

(Gudin One: ; Shnghai Xng qng ln: ) are written differently in the two manuscripts. As for the character xng (*[s]e-s) human nature, the Shnghai Xng qng ln (xq1/31) also has the phonophoric shng (OC *sr) and the signific m (), just like the Gudin One manuscript (x2/8). This might be evidence that the same word (xng nature) was also expressed in the previously where the Shnghai Xng qng ln only used the phonophoric shng . Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [E]: Jing descend; to send down in the Gudin One version appears as ; the Shnghai Xng qng ln has (xq2/7).

reconstructing the xng z mng ch

325

Following the last character of this line, jing , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [F ]: The character for qng in the Gudin One manifestation is written as ( x3/7). In the Shnghai Xng qng ln it appears as

( xq2/11).10 In the two versions, a (different) mark for repetition follows this character. Both texts write s to govern (OC *[s]) for sh to begin (OC *l ) (x3/5).11 It may have been the case that the graphical form for to begin was introduced fairly late in the Kingdom of Chu. To my knowledge, the first appearance of in Chu occurs in the Ynmng manuscripts from the Shuhud tomb, which dates to late Warring States/beginning of Qn. Following the last character of this line, xng , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [G]: In the Gudin One manifestation of the Xng z mng ch, the character for zhng is written as (x3/15). In the Shnghai Xng

qng ln, it appears as ( xq2/19). Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [H]: The end of strip x3/19 of the Gudin One Xng z mng ch has broken off. Judging from the lost part, presumably three or four characters are missing. The top of the first missing graph is still visible and suggests the reading qng feelings. Qi Xgu proposes reconstructing as zh nng , which accords with the overall structure of the passage.12 If one looks only at the structure of the following line, Qis proposal seems correct. However, when comparing the length of the missing part with that of the other strips, one might

The two strokes at the bottom are a sign for repetition. Note that relationships such as this one are why Zhngzhng Shngfng (2003) reconstructs s as *sl. 12 See Hbi shng Jngmn sh bwguan 1998, 182, n. 2.
11

10

326

chapter twelve

be inclined to reconstruct four graphs instead of just three.13 But the Shnghai Xng qng ln seems to corroborate Qis suggestion. It reads: ()()() [] He who understands the unshaped feelings may manifest them [on the outside]; he who understands rightness may take {it} in (xq2/23).14 The Gudin One text writes y (OC *(r)aj) for y (OC *(r) aj-s). This is commonly seen in Chu manuscripts. [I]: Instead of () () that which one loves and hates are the things [in the world] (x3/1318), the Shnghai Xng qng ln reads: () () to love and to hate is [determined by] the things [in the world] (xq3/14). [J]: After x4/20 the strip of the Gudin One Xng z mng ch has broken off. Presumably three characters are missing (the upper part of x4/20 is still visible on the strip and suggests a reconstruction as b .) Qi Xgu reconstructs the entire passage as X.15 Many modern editors accept his suggestion. L Lng, however, suggests reading the line as , which is also accepted by many editors.16 L Lngs proposal is supported by strip 3 of the Shnghai text Xng qng ln .17 [K]: As for the graph sh (OC * ret-s) (x5/6), the Gudin One Xng z mng ch only writes the phonophoric sh (OC *et-s setting; or *ret-s) to sow, plant. The Shnghai Xng qng ln adds the signific n to this character (xq3/15). [L]: As for the graph zhu (OC *to), the Gudin One version writes ( ) (x5/11). The Shnghai Xng qng ln adds another ( ) (xq3/20).

stroke to it

13 14 15 16 17

For comparing the length of the missing part with the other strips, see ibid., 61. The end of strip xq2 has broken off, so the last character of this line is missing. See ibid., 182, n. 4. See L Lng 1999, 504. See Ma Chngyun 2001, 1:73.

reconstructing the xng z mng ch

327

Following the last character of this line, y (OC *lAj) (x5/15), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [M]: Following x5/20 ( ), which is the upper part of the character ( <>), the bamboo strip of the Gudin One Xng z mng ch has broken off. L Lng suggests reconstructing the line as () () [ x6 ].18 The Shnghai Xng qng ln should, instead, be reconstructed as follows: . Gu (OC *[k](r)o) might also be read as ku (OC *k(r)o-s) to knock, strike. Compare this passage with that of the Wu xng, strips w19/320/17: w20 [] The sound of bronze and jade [stones] resonating along with itthis is a person possessing virtue. The sound of bronze is goodness. The tone of jade is sagaciousness. Goodness is the w20 Way of humans. Virtue is the {Way} of Heaven. Only after there is a virtuous person can there be the sound of bronze and jade [stones] resonating along with it . The Mngz 5B1 reads: Kongz is said to have gathered great achievements; gathering great achievements, this is the sound of bronze and jade stones resonating along with it. The sound of bronze is the beginning of an inherent pattern; jade stones resonating along with it is the end of an inherent pattern. Beginning an inherent pattern is a matter of the wise; ending an inherent pattern is the matter of the sagacious one. [N]: This line is missing in the Shnghai Xng qng ln, and this omission does not result from a broken strip in that text. This is interesting insofar as the missing statement explains the technical quotation concerning the sound of bronze bells and resonating jade stones and thus is similar to that of Wu xng. [O]: The bottom of strip x6 has broken off. Four or five characters are missing. Different editors come up with different reconstructions. L

18

See L Lng 1999, 504.

328

chapter twelve

Lng, for instance, reconstructs: [] x7 .19 T Zngli and Li Zuxn reconstruct: [] x7 .20 Li Zho reconstructs: [] x7 .21 I suggest reading the passage as follows: [ ] x7 . As for the structure of this passage, the reconstruction put forward by T Zngli and Li Zuxn is reasonable, for yu is connected to xng. However, it is the final aim to determine ones xn mind. Therefore, to read xn simply in connection with one level of development is not a very felicitous choice. [P]: After x7/19, the strip broke off. The upper part of x7/20, however, is still visible ( ), and it makes sense to reconstruct it as xng

nature ( , as in x1/5). T Zngli and Li Zuxn suggest the following reconstruction of this passage: [ ] x8 .22 L Lng suggests the following reading: [] x8 .23 My reconstruction follows L. The character (here: x8/4 ) often causes confusion. Generally, it is identified either as sh cause to or as bin to change. In the context of the present text, J Xshng suggests reading it as bin.24 The entire building block x6/9x8/6 is missing in the Shnghai Xng qng ln. But this unit is an important step in the process of argument construction in the Xng z mng ch. The process explained in building block 3 is analogous to that of the previous building block. What mind fulfils in the process of developing ones nature (previous building block) is analogous to what intention fulfils in the process of developing ones mind. The building block under review therefore directly corresponds to the last line of the previous building block. This is interesting insofar as the omission in the Shnghai Xng qng ln must therefore be a structural one. This once more suggests two
19 20 21 22 23 24

See L Lng 2002, 105. See T Zngli and Li Zuxn 2001, 148. Li Zho 2005, 88. See T Zngli and Li Zuxn 2001, 148. L Lng 2002, 105. See L Xshng 2004, 159.

reconstructing the xng z mng ch

329

isolated instantiations of writing down a verbally coherent text and excludes the possibility that the differences result from a defective transmission process of one and the same manuscript Vorlage. [Q]: The character x8/11 y appears on the strips as ( ). Chn Lnqng , Zhng Yshn , and Zu Jnzh identify it as q (OC *[g]), which they read as q (OC *g(r)) temporarily.25 The parallel structure of the present unit makes this an unlikely reconstruction. Instead, I follow the reading of the editors of the Hbi Province Museum and read it as y different, although it is written with a different graph from character x9/18, namely as ( ).26

[R]: The identification of character x8/16 () is problematic, and different editors read it differently. The general trend is to read it either as sh cause to stand27 or as zh pillar, to support.28 Li Zho supports his reading of the graph as zh by quoting a line from the Xnz Qun xu that reads: what is [too] strong causes itself to be broken [easily]; what is flexible causes itself to be bound.29 The Qng scholar Wng Ynzh (1766 1834) argues with reference to the Chnqi Gngyng zhun (Duke A , year 14) that zh (OC *m-tro) pillar should be read as zh (OC *tuk) to break.30 If this were true, the line in the Xng z mng ch may be read accordingly. [S]: The first part of this building block is missing in the Shnghai Xng qng ln. This is due to the bad condition of the Shnghai strips and says nothing about the different manifestations of the text. The Shnghai Xng qng ln continues after x9/9 ni . On the graph sh under [S]. y one, see my discussion of the Qing d y

See ibid., 158. See Hbi shng Jngmn sh bwguan 1998, 182, n. 5. 27 E.g., L Lng 1999, 504. 28 E.g., Li Zho 2005, 9394. 29 See Xnz j ji 1997, vol. 1 (1:5/2). 30 Z l died. Kongz [thereupon] called out: Ai! Heaven breaks me! See Chnqi Gngyng zhun zh sh (chap. 28), 357.
26

25

330

chapter twelve

Compare the form of the graph x9/18 y ( ) with that of x8/11 y ( ). The graph for y , instead, appears as in the Shnghai Xng qng ln. Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [T]: In the Gudin One text the character for jio to teach ( x9/19) is written with the signific yn words: . In the Shnghai (xq4/11). Xng qng ln it is written with the signific z son: The end of this building block is not formally marked off in either manuscript. ( ) causes considerable problems. The editors [U]: Graph x10/5 of the Hbi Province Museum hesitatingly read it as fng to meet with.31 L Lng disagrees. He reads it as n go to meet; go against because it is very similar to the graph ch32/13 of the Chng zh wn zh , where it is read as the antonym of shn . In the Shnghai Xng qng ln the same graphit appears as on the stripsis also rendered as n hence my reading of the graph as n go to meet. [V]: In the Gudin One Xng z mng ch, the character for jio to communicate with is written only with the phonophoric jio (OC *k(w)raw) x10/8 . In the Shnghai Xng qng ln, the signific xn is added to it (xq4/24). The editors of the Hbi Province Museum read the graph x10/11 wn as l .32 Whereas it is written only with the phonophoric (OC *ma[n]-s) in the Gudin One Xng z mng ch, in the Shnghai Xng qng ln the signific xn is added beneath the same phonophoric (xq4/27). (y ), the last graph of this line, a small [W]: Following x11/3 reading mark is added to the manuscript of the Shnghai Xng qng ln.

31 32

See Hbi shng Jngmn sh bwguan 1998, 168. See ibid., 179.

reconstructing the xng z mng ch

331

[X]: For the character x11/4 (here read as n to meet), see the discussion of the graph x10/5 in building block 5 (under [U]). In the two manifestations, the graph is written only with the phonophoric du (OC *lot-s). I always read it to be pleased in this context. as yu (OC *lot-s)

[Y]: As seen above, the signific n woman is added to the word sh (OC *het-s) force (of circumstance); condition in the Shnghai Xng qng ln (strip xq5/28), whereas it is written with only the phonophoric sh (OC *et-s setting; or *ret-s to sow, plant) in the Gudin One text.33 J Xshng suggests reading it as y skilly has the same phonophoric sh (OC *et-s or *ret-s) and is etymologically related to sh because, he contends, Confucian texts commonly draw on this notion.34 [Z]: Following the last graph of this line, y (x12/3), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. The same is true in the next line, x12/9, after do y . [Aa]: Following the last graph of this line, w () (x12/15), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Ab]: Instead of () () () That which generates satisfaction in oneself is what we call delight as in the Gudin One text, the Shnghai Xng qng ln reads: () () (). The character yu (and also ) can be reconstructed as OC *[g]w]k or *[g]w]k-s; the character kui is to be reconstructed as OC *kwhre[t]-s. The character j in the Gudin One version has the OC reading *k(r), while q from the Shnghai Xng qng ln has the OC reading *[g](r). The difference between the use of kui (OC *kwhre[t]-s) and yu (OC *[g]w]k or *[g]w]k-s) seems difficult to explain because these graphs neither match the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components nor are they graphically related. The different use of j (OC *k(r)) and q (OC *[g](r)), for their part, can be explained easily on phonetic grounds since these graphs match the criteria for
33 34

See under [K] for the graph sh (x5/6). See J Xshng 2004, 163.

332

chapter twelve

phonetic similarity in Old Chinese for loan characters and phonetic components. The difference at this point can thus be explained, for instance, by postulating a text that was stable in wording but rather unstable in writing or by ascribing it to changes that may have occurred in the process of copying the text by dictation. [Ac]: After the last character of this line, sh () (x13/6), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. The same holds true below after x13/13 (). [Ad]: Following the last character of this line, y (x14/7), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Ae]: Instead of do zh , the Shnghai Xng qng ln reads do y . . . (or maybe ). Due to broken strips, the following part is missing in the Shnghai Xng qng ln. [Af ]: Following the last character of this line, y (x15/14), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Ag]: Whereas in the Gudin One manifestation the graph for sh is written () (x15/15), it is written with the components in

the Shnghai Xng qng ln ( ) (xq8/22). The word sh writing, however, is written identically in the two texts, namely as (). [Ah]: For the line () x16 in every case their first appearing was given birth by x16 man in the Gudin One Xng z mng ch, the Shnghai Xng qng ln reads: () . . . their first appearing was brought about side by side from . . .. [Ai]: Following the graph x16/5 wi appears a mark for repetition ;

in both texts. The Gudin One Xng z mng ch uses(as in: the Shnghai Xng qng ln uses = (very unclear on strip xq9/2).

[Aj]: Qi Xgu argues that the graph x16/18 should be read as ju to rise; surge. It is composed of the two elements and ( ).

reconstructing the xng z mng ch

333

In the Shnghai Xng qng ln, the same graph consists of the two elements and ( ) (xq9/15). Following the last character of this line, y (x16/20), a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Ak]: The character x17/6 ( gun here: to behold) is written

( ) in the Shnghai Xng qng ln (xq9/27). The character x17/8 zh (OC *t) should in fact be read xin (OC *sr), as may be seen from the Shnghai Xng qng ln, where xin is used as well. Qi Xgu has already pointed this out.35 L Lng suggests reading the graphs x17/11, 12 as n shn . The Shnghai Xng qng ln also reads n shn . Li Zho argues that the combination n shn should be read shn n in proper sequence.36 [Al]: The character x17/14 t is written with a different signific, namely ru ( ) in the Shnghai Xng qng ln. The characters x17/1819 are read as ji d by Qi Xgu.37 L Lng and also T Zngli and Li Zuxn follow this suggestion.38 Most scholars nowadays read x17/19 graph as wn (OC *m[n]). The direct transcription of the graph in question is ; its phonophoric is mn (OC *mi[n]). The reason the graph is constantly identified as d in the literature is simply that it was transcribed incorrectly, without the phonophoric mn (OC *mi[n]).39 William H. Baxters approach to dating the Laoz on the basis of rhyme distinctions can be used for the passage under review. Based on rhyme distinctions, Baxter concludes that it is linguistically quite plausible to date the bulk of

See Hbi shng Jngmn sh bwguan 1998, 182, n. 9. See Li Zho 2005, 95. 37 See Hbi shng Jngmn sh bwguan 1998, 182, n. 10. 38 See L Lng 1999, 505. Later, however, he corrects his earlier view and reads it as ji wn, as most commentators now do. See L Lng 2002, 106. See also T Zngli and Li Zuxn 2001, 153. 39 Chn Jin, personal communication, 01 September 2007.
36

35

334

chapter twelve

the Laoz to the mid or early fourth century.40 For instance, he demonstrates that the Laoz rhymes xun (OC *[g]wi[n]) with mn (OC *m[n]), which reveals that a substantial part of the Laoz dates from a time when the merger of *-in and *-n had already occurred (or they were already close enough that they were a good rhyme).41 In the same vein, it can also be assumed that the use of the phonophoric mn for wn must reflect either a rather late use (possibly midto late fourth century BC) or a geographical use (possibly the Chu region). [Am]: As already seen, the Gudin One Xng z mng ch and the Shnghai Xng qng ln use a different signific for the word jio to teach. Whereas the Gudin One text uses the signific yn ( ) words, the Shnghai Xng qng ln uses the signific z son ( ). Following the graph jio to teach, both texts use a mark for repetition (the Gudin One text uses; the Shnghai Xng qng ln, =). [An]: I follow Qi Xgus suggestion to read the graph x19/2 ( ) as xng to raise; start; rise.42 Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. (zh [Ao]: The word zh (OC *ket-s) is written with the graph ; OC *[t]et) (x19/10) in the Gudin One Xng z mng ch. The Shnghai Xng qng ln writes zh (xq11/11). It should be reconstructed as OC *ket-s.43 Following the last character of this line, zh , a small reading mark is added to the manuscript of the Shnghai Xng qng ln.

See Baxter 1998, 249. It is also possible that there was a more limited mergerfor instance, that *-in had changed to *n only after labialised initials like *[g]w-. 42 See Hbi shng Jngmn sh bwguan 1998, 182, n. 11. 43 The reconstruction of zh MC < tsyejH < OC *ket-s cut out, prepare is based on the fact that the word j < kjejH < *kret-s mad (dog) in the same series. William Baxter has alerted me to the fact that if the reconstruction of the initial is correct, it would be parallel to the use of sh < dzyeX < *[g]e clan to write sh < dzyeX < *[d]e this. This could reflect a Jn/Chu dialect characteristic. It anticipates the merger (palatalisation of nonpharyngealisedtype Bvelars before front vowels) that had happened by Middle Chinese times.
41

40

reconstructing the xng z mng ch

335

( ) as x Qi Xgu suggests reading the character x19/16 44 (OC *[s-m-l]a). The Shnghai Xng qng ln also writes this graph (xq11). J Xshng suggests reading the graph as x as (OC *[s-m-l]a) regulate; put in order; arrange (here: become regulated, come into the proper order).45 I follow Qi. Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Ap]: The Gudin One Xng z mng ch writes the graph for mi (OC *mrj) beautiful with the signific n , ( (x20/15). The Shnghai Xng qng ln writes it without a signific (xq12/8). After x20/17 () the strip breaks off. The upper part of x20/18 is still and may reasonably be reconstructed as gu to honour. visible In the reconstruction of the followingpresumablytwo characters, I follow Qi Xgu.46 The Shnghai Xng qng ln corroborates his suggestion. [Aq]: Both manifestations of the text write sng (OC *s-[g]o-s) for rng (OC *[g](r)o) (x21/6). As Qi Xgu remarks, sng (OC *s-[g]o-s) is commonly used for rng (OC *[g](r)o).47 Only in the Gudin One version is there a small reading mark after the graph (). [Ar]: Following the last character of this line, yn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [As]: Due to broken strips, the lastpresumably threecharacters on strip x21 of the Gudin One Xng z mng ch are missing. The Shnghai Xng qng ln is of no help here, as the corresponding strip has broken off too. About five graphs are missing in the Shnghai Xng qng ln. Because bending ones hands is an expression of respect according to the rites, I reconstruct the passage parallel to the subsequent line as follows: () [ X]. J Xshng

44 45 46 47

See Hbi shng Jngmn sh bwguan 1998, 182, n. 12. See J Xshng 2004, 172. See Hbi shng Jngmn sh bwguan 1998, 182, n. 15. See ibid., n. 17.

336

chapter twelve

recognizes a graphic similarity between x22/2 and shu in the Zhngshn Wng dng .48 Accordingly, I follow J Xshng in reading the graph as shu frequently. [At]: Qi Xgu suggests reading the character x22/11 as zhng (OC *tr). The Shnghai Xng qng ln has dng (OC *t) (xq13/12). The two graphs fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. [Au]: Graph x22/14 consists of the signific yn words and the phonophoric s (OC *[s]). The editors of the Hbi Province Museum read it as c (OC *[N-s] (?)) word; speech.49 The editors of the Shnghai Xng qng ln read it as zh (OC *lr-s) to regulate.50 Chn Lnqng , Zhng Yshn , and Zu Jnzh explain the graph as y (OC *l), which they read as y (OC *l) to hand over; gift.51 For the time being, I follow the suggestion made by the Hbi Province Museum editors because graphically and phonologically this is the easiest explanation and contrasts best with the shallowness of laughter, about which the Xng z mng ch continues to talk in the next building block. [Av]: The editors of the Hbi Province Museum wrongly read the graph x22/20 as l .52 Instead, the graph should be transcribed as (), as Qi Xgu has pointed out.53 The Shnghai Xng qng ln also reads x rejoice (xq13/20). [Aw]: The editors of the Hbi Province Museum read the graph x23/19 as b (OC *-[b]rot) to pull up; uproot.54 Qi Xgu suggests reading it as b (OC *pat) to stir up.55 I also follow Qis suggestion to read x23/24
48 49 50 51 52 53 54 55 56

( ) as hu thick.56

See J Xshng 2004, 174. See Hbi shng Jngmn sh bwguan 1998, 180. See Ma Chngyun 2001, 1:238. See Chn Lnqng, Zhng Yshn, and Zu Jnzh 2004, 174. See Hbi shng Jngmn sh bwguan 1998, 180. Quoting Dng Yunzh 2002, 117, n. 1. See Hbi shng Jngmn sh bwguan 1998, 180. Ibid., 182, n. 20. Ibid.

reconstructing the xng z mng ch

337

[Ax]: In this instance, the two texts have the same graph for wn (OC *mu[n]) to hear, namely (OC *m un). The two texts also write the word for g (OC *[k]aj) song identically, namely h (OC *qhaj). Chn Lnqng notes that h (OC *qhaj) is the ancient writing for g (OC *[k]aj).57 The use of the word for yo to sing in the two manuscripts is more revealing. Whereas the (), the Shnghai Xng qng ln Gudin One manuscript has has the graph yo forpresumablythe same word (xq14/29). Both characters belong to the xio rhyme group, so most editors take this to be a good phonetic loan.58 However, from rhyme evidence from the Odes, William H. Baxter convincingly argues that the xio rhyme group has to be split into an *-aw rhyme (xio 1) and an *-ew rhyme (xio 2).59 The graph yo belongs to the xio 1 rhyme (*-aw); yo belongs to the xio 2 rhyme (*-ew). Thus, it seems that the distinctions between *-aw and *-ew were not well preserved when the Shnghai Xng qng ln was fixed on bamboo. Other fourth-century BC texts such as the Laoz also preserve this rhyme,60 as Baxter has convincingly argued, and so do the Odes. Baxter also notes that the Inner Chapters of the Zhungz do show examples where *-et rhymes with *-at, and he assumes that *-et had changed to something like *-iat by this time.61 Examples like these may provide some clues for dating the Shnghai Xng qng ln (late fourth, early third century BC?), and provide information on traces of dialects in early Chinese texts. We cannot exclude the possibility that the merger of *-ew and *-aw might be a Chu-specific dialect phenomenon. [Ay]: The editors of the Hbi Province Museum read the graph x25/6 as nn to refute. The graph is written with the signific heart beneath (). Qi Xgu suggests reading it tn to sigh.62 I follow Qi.

See Chn Lnqng, Zhng Yshn, and Zu Jnzh 2004, 177. See ibid. 59 See Baxter 1992, 526532. 60 Chapter 56 of the Wng B version of the Laoz , however, has one stanza rhymed with *-at and *-et. Yet, as William Boltz has already shown, the irregular rhyme was a late insertion that does not appear in the Mawngdu Three Laoz (nor in the so-called Gudin One Laoz ). See Boltz 1984, 220224; 1985 (quoting from Baxter 1998, 246). 61 See Baxter 1998, 246. 62 See Hbi shng Jngmn sh bwguan 1998, 183, n. 23.
58

57

338

chapter twelve

[Az]: The editors of the Hbi Province Museum transcribe x26/3 as yang (OC *g(r)a-s) river. L Lng reads it as yong (OC *[g] w ra-s) to draw out; to chant.63 P Mozuo notes that yang (OC *g(r)a-s) is the ancient form of yong (*[g]wra),64 which fully justifies L Lngs suggestion to read x26/3 as yong to chant, and I follow L. I follow Li Zho , who suggests reading the graph x26/8 ( ) as kui (OC *[k]whr[t]-s) to sigh loudly.65 In the Shnghai Xng qng ln this word is written with the phonophoric wi (OC *quj-s) ( ) (xq16/6).

( ), read as sh (OC *l ) to begin, the [Ba]: As for x26/19 Gudin One version has the phonophoric (s ) (OC *[s]) together with the signific yn words. In the Shnghai Xng qng ln only the phonophoric s (OC *[s]) is used for this graph (xq6/18). The last graph of this line, x27/1, is ( ), shn (OC *[d]i[n]-s), which in the Shnghai Xng qng ln consists of the elements sh , yn , and jn . A small reading mark is added to the manuscript of the Shnghai Xng qng ln after this graph. [Bb]: Some editors explain the graph x27/7 as the abbreviated form of si to control (OC *[s]). Yet, comparing this graph with x26/20 ( ),66 it transpires that the direct transcription should in fact be . In the Shnghai Xng qng ln, the graph is written with the signific m , which P Mozuo reads as zh (OC *lr-s).67 Chn Lnqng , Zhng Yshn , and Zu Jnzh read the graph di to be remiss (OC *l]).68 Based on the previous line (x26/20, [ *l ) to begin. ] sh , OC *l ), I read x27/7 as sh (OC

63 64 65 66 67 68

See L Lng 2002, 109. See Ma Chngyun 2001, 1:243. See Li Zho 2000, 91. See my note under [Ba] above. See Ma Chngyun 2001, 1:243. See Chn Lnqng, Zhng Yshn, and Zu Jnzh 2004, 181.

reconstructing the xng z mng ch

339

[Bc]: Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Bd]: The editors of the Hbi Province Museum transcribe the character x28/4 (and 8) as lng (OC *m-ro). In his earlier reading of the text, L Lng suggested reading it as dng (OC *[d]o),69 which, however, he changed to lng in his later edition of the text.70 In the Shnghai Xng qng ln the graph also appears as lng , which P Mozuo reads as lng (OC *[r]u) to become extremely exalted.71 I follow this reading. Chn Lnqng notes that gu yu music of old and y yu latter music are complementary.72 According to him, gu yu denotes the music of Sho and Xi , whereas y yu means latter musicbut without the bad connotation of new musicxn yu , as discussed in the L j , Yu j chapter. [Be]: Following the last character of this line, qng , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Bf ]: Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Bg]: The Gudin One version writes the word a (*[]j) as ( )

(x29/15). The Shnghai Xng qng ln writes it as () (xq18/11). Because both graphs share the same phonophoric, they fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Bh]: Both texts write the word dng (OC *-[d]o) as composed of the phonophoric tng (OC *-[d]o) and the signific of a hand holding a stick ( ) (x30/5).
69 70 71 72

See L Lng 1999, 505. See L Lng 2002, 109. See Ma Chngyun 2001, 1:256. See Chn Lnqng, Zhng Yshn, and Zu Jnzh 2004, 182.

340

chapter twelve

When looking at strips x2530 of the Xng z mng ch it becomes particularly obvious from the spacing of the characters that in the course of manuscript production the strips were first bound together and only then inscribed with characters. Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. as jn sh in the L Lng transcribes character x30/89 meaning of to drown gradually.73 The Shnghai Xng qng ln at this instance reads () () (xq18/29, 30). ( ) as li

I follow Li Zho and read the graph x30/11 fierce.74

[Bi]: The character x31/13 ( ) is generally read as yu (OC *liw) melancholic, sad or long lasting, drawn out.75 The Shnghai Xng qng ln has the character yu (OC *liw) flowing water (xq29/24). [Bj]: The editors of the Hbi Province Museum read x32/14 () as nn (OC *nar). Li Xnln suggests reading it as tn (OC *nar-(s)) to sigh.76 The two graphs share the same phonophoric (), and I follow Lis suggestion. [Bk]: Graph x32/21 ( ) seems to have been only one form for writing the words sh (*s-r) to cause and bin (OC *[b]ro[n]-s) cap, but the graph was also used for bin (OC *pro[n]-s) to alter, to change. Probably at a later date, when the distinction between *-on and *-en was softened (early third century BC?)77 or due to dialect differences (Chu?), it was also used for bin (OC *[b]re[n]) to discriminate. In the context of the present passage I believe we can be fairly confident in reading the graph as bin (OC *pro[n]-s) to alter.

See Li Zho 2005, 98. Ibid. 75 See, e.g., L Lng 2002, 109; Li Zho 2005, 98. 76 See Li Xnln 2000, 341. 77 It seems that in the Laoz the distinction between *-on and *-en is kept fairly strictly, whereas in the Zhungz *-on and *-en did occasionally rhyme.
74

73

reconstructing the xng z mng ch

341

After x32/22 the strip has broken off. Presumably, three characters are missing. Qi Xgu suggests reconstructing the missing passage as follows: () () () [].78 Based on the Shnghai Xng qng ln (strip xq20), the following reconstruction might be () () [] when the sound of preferable: () it changes, then the mind { follows it [too]}. [Bl]: Character x33/9 has bleached out, but it seems to consist of the elements . L Lng suggests reading this as yn to sigh.79 Most editors follow his suggestion.80 The strips from the Shnghai Xng qng ln have broken off at this junction, so they are of no help. For the time being this line cannot be reconstructed with certainty. As for the graph x33/10 (and x33/14, 18, 22) (), it is written as in the Shnghai Xng qng ln (xq21/2, 6, 10). [Bm]: According to the parallel pattern of the sentences under review, it seems that a character, namely ye , is missing after x33/19 (). For the character x33/17 (here transcribed as ji murmurs of singing) there exist different readings. My reconstruction follows that of Li Zho.81 [Bn]: Following the last character of this line, y , a large mark is added to the Shnghai manuscript, possibly to subdivide the text. After this mark, the overall coherence of the two texts dissolves. Even though bleached out, we can discern that the character x33/20 ( ) contains the elements , of which du (OC *N.to(k)-s) is likely to be a phonophoric. L Lng suggests reading the graph as o (here: the sound of singing),82 which should be reconstructed as Old Chinese *q(r)o. Du (OC *N.to(k)-s) and o (OC *q(r)o) have the same main vowel and coda (the medial *-r- and the tone category can be ignored). Yet the two do not share the same initial:

See Hbi shng Jngmn sh bwguan 183, 32. See L Lng 2002, 106. 80 See Dng Yunzh 2002, 161; Chn Lnqng, Zhng Yshn, and Zu Jnzh 2004, 189; Li Zho 2005, 99; among others. 81 See Li Zho 2005, 99. 82 See L Lng 2002, 110.
79

78

342

chapter twelve

o (OC *q(r)o) has a uvular, and du (OC *N.to(k)-s) has a dental initial, so they do not share the same position of articulation. This shows that L Lngs suggestion of reading the character x33/20 as o isat least phonologicallyhighly problematic. T Zngli and Li Zuxn suggest reading the graph as x quip (OC *(r)ar-s).83 Li Zho reads it as x () (OC *(r)ar-s) to sigh (this graph does not seem to occur otherwise in pre-Qn texts). Even though this also seems to be difficult to justify phonophorically, it appears to be a more reasonable solution as far as the content of this passage is concerned. If I am not mistaken about the reconstruction of this passage, then it seems that due to either chronological or dialectal reasons the distinction between *-a and *-o was not well preserved in the Xng z mng ch. The Shnghai Xng qng ln has (xq21/9). Following the last statement of this row ( ) the Shnghai Xng qng ln continues with what would be pericope 16 in the Gudin One manuscript. This loss cannot be explained by reference to the poor preservation of the Shnghai manuscript, because the next unit of the text continues on the same stripthat is, strip xq21. Two building blocks are absent. This loss in the Shnghai Xng qng ln is discussed in chapter 7. [Bo]: L Lng reads the character x34/3 () as to , which Li Zho explains as the appearance of being happy or cheerful.84 [Bp]: Lio Mngchn seems to have been the first to suggest reading x34/9 (OC *[g(r)]u (?)) as yo (OC *l[a]w) to shake. His reading is based on a commentary from Zhng Xan (AD 127200).85 Most scholars follow this suggestion. [Bq]: In his first edition of the Gudin One manuscripts, L Lng read as zu to get up.86 In his later edition, howthe graph x34/11 ever, he corrects this suggestion and reads the graph as wu dance.87 This seems to be the generally accepted reading now.
83 84 85 86 87

See T Zngli and Li Zuxn 2001, 162163. See L Lng 2002, 106. See also Li Zho 2005, 99; Dng Yunzh 2000, 78. Quoted from Dng Yunzh 2002, 164. See L Lng 1999, 506. See L Lng 2002, 106, 113, 114.

reconstructing the xng z mng ch

343

[Br]: Following the last character of this line, y (x35/10), a mark to subdivide the text is added to the Gudin One Xng z mng ch. Following the mark, the strip bears no more writing. Such a mark, although distinct in form, also appears on strip xq21 of the Shnghai Xng qng ln. It therefore seems that both manifestations of the text are subdivided at this junction. The Application: Canto 2 [Bs]: Qi Xgu suggests reading the graph x36/4 ( ) as qi to seek for.88 Unlike the Gudin One Xng z mng ch, the Shnghai Xng qng ln does not use the concept xu (OC *m-kruk) learn ( ) (x36/2) but rather has jio (OC *[k]raw(-s)) to teach ( ) (xq31/32). It is not certain whether this is a miswritten xu which indeed is similar in writing to jio ( / )or whether this is meant to express another idea. Except for the first six graphs of this building block, namely () () , the entire passage () () () () () () () () () () ()in all twentyseven charactersis missing in the Shnghai Xng qng ln. This loss cannot be explained by the poor preservation of the Shnghai bamboo strips, because the subsequent part that also appears in the Gudin One manuscript continues on the same strip of the Shnghai Xng qng ln (xq32/2). It could be that this loss results from an accident occurring when quoting a writtenthirdsource text (but again, not the Gudin One Xng z mng ch because the loss is not one entire strip there) and, by mistake, leaving out one entire strip carrying about 27 characters. () [Bt]: Given the context of this passage, the character x37/15 should be read as wi artificial, as was also suggested by Qi Xgu.89 The same holds true for x37/26 below.

88 89

See Hbi shng Jngmn sh bwguan 1998, 183, n. 35. See ibid., n. 38.

344

chapter twelve

[Bu]: The Shnghai Xng qng ln has () with the mind signific beneath (xq32/14). Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Bv]: The first graph in this line (following x38/3) is missing. Qi Xgu is comsuggests reconstructing it as q .90 The character x38/6 posed of and ( ). The editors of the Hbi Province Museum read it as ju . I follow their suggestion.91 Instead of sh (OC *[g][i]p) ten, the Shnghai Xng qng ln reads zh (OC *N-trek) (xq32/21). [Bw]: I follow the suggestion made by Qi Xgu to read x38/12 ( ) as ch .92 [Bx]: Graph x38/20 is problematic. The left part of the graph is yn words, speech. The right part remains unclear. Chn Li (1999) transcribes the graph as jian because of the line in the Wu xng (w40/2024). But another reference could be the line (w41/58). As the right part of the character cannot be identified with certainty, it is impossible to judge whether yn should be considered the signific or the phonophoric here. Li Zho holds that the right part of this character is a deviant form of nu woman. Scholars have also suggested that the graph might be sh to forgive, as the two would share the same phonophoric.93 P Mozuo reads the graph in the Shnghai Xng qng ln as q bent. However, as the graph is nearly invisible now on the Shnghai strip, there are no grounds on which to judge this suggestion. For the time being, I follow Li Zho. Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [By]: Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln.

90 91 92 93

See ibid., n. 39. Ibid., 180. Ibid., 183, n. 40. See Li Zho 2005, 100.

reconstructing the xng z mng ch

345

[Bz]: As for the graph x39/11 ( ), which is understood as du genuine, the Shnghai Xng qng ln adds the signific heart at the bottom of the character (xq33/16 ).

[Ca]: Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Cb]: Following the last character of this line, rn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Cc]: Note that the two texts always write the character do true ethical code, Way differently: as in the Gudin One Xng z mng ch and as in the Shnghai Xng qng ln. The Shnghai Xng qng ln also writes zhng fidelity without the heart signific. [Cd]: Following the last character of this line, y , there is a big mark on the strip of the Shnghai Xng qng ln, probably signalling the end of this unit. [Ce]: The graph so (OC *[s]aw-s), generally understood as zo (OC *[ts]aw-s) quick-tempered, is written with the phonophoric cho (OC *[dz]raw) in the Shnghai Xng qng ln xq35/10 ( ). The two graphs fulfil the criteria for phonetic similarity in Old Chinese of loan characters and phonetic components Following the last character of this line, shn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Cf ]: Instead of hun (OC *[g]ron-s) (x42/18) as used in the Gudin One Xng z mng ch, the Shnghai Xng qng ln has a graph consisting of the phonophoric juan (OC *[k]ro[n]) on top of the signific heart ( xq35/20). The two fulfil the criteria for phonetic similarity in Old Chinese of loan characters and phonetic components. [Cg]: Following the last character of this line, shn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln.

346

chapter twelve

[Ch]: The editors of the Hbi Province Museum read the graph x43/10 ( ) as bin (OC *[b]ro[n]-s) hurry, hasty.94 Qi Xgu suggests reading it as bin (OC *pro[n]-s) to alter.95 The Shnghai Xng qng ln has , which can be read as either sh (OC *s-r) or bin (OC *[b]ro[n]-s)/bin (OC *pro[n]-s).96 Based on the context of this passage, I follow the Hbi Province Museum editors and read it as bin (OC *[b]ro[n]-s) hurry, hasty. [Ci]: Following the last character of this line, shn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. As for the character x43/19 (), the Shnghai Xng qng ln writes y (xq36/15). The two share the same phonophoric. [Cj]: Both texts use shng (OC *l e-s) sage to express the word shng (OC *l e) sound. Although they belong to the same word family, this is a pure phonetic loan and should not be interpreted. [Ck]: Graph x44/6 seems to consist of the phonophoric hu (OC *[g]wk) and the signific ru ( ). L Lng reads the combination of graphs x44/67 as y to smouldering melancholy.97 In the Shnghai Xng qng ln the graphs are faint, so it is of no help in determining the proper reading of this passage. I suggest reading x44/6 as y (OC *qwk). It carries the same meaning as y , which also has the phonophoric (OC *[g]wk) like . In this line, the two graphs rn person and nn difficult are absent in the Shnghai Xng qng ln. The loss of the graph nn can be explained by the fact that the top of Shnghai strip xq36 has broken off. The loss of the word rn , however, constitutes a dissimilarity in the text itself. Thus, instead of men would easily die for them as in the Gudin One version, the Shnghai Xng qng ln just states [] one would {easily} die for them. [Cl]: Both texts use a mark of repetition after the word, which the Gudin One text has as ji to restrain. In this instance, two differ94 95 96 97

See Hbi shng Jngmn sh bwguan 1998, 180. Ibid., 183, n. 42. See my disussion under [Bk], above. See L Lng 2002, 107.

reconstructing the xng z mng ch

347

ent graphs appear in the texts. In the Gudin One text, the character x44/22 appears as ( ). In the Shnghai Xng qng ln, the char-

acter xq37/9 is written as ( ). The mark for repetition is written differently in the two manuscripts. Whereas the Gudin One Xng z mng ch indicates repetition with a single stroke, the Shnghai Xng qng ln has a double stroke =. [Cm]: After x45/4 there is a mark for repetition. The character itself is transcribed as jian (OC *kr[a]n) condensed, little, which is read as jian (OC *kre[n]) righteous, tough and honest.98 L Lng suggests reading it in the meaning of jian (OC *kra[n]) frank, righteous.99 [Cn]: The editors of the Hbi Province Museum read the graph x45/21 ( ) as y cheerful. I follow Li Zho and read it as di (OX *l) in jeopardy.100 There seem to be no phonological connection between the graph di jeopardy, endangerment (x45/21) and ( ), which has the

graph xq37/31 of the Shnghai Xng qng ln phonophoric jn (OC *[k]r).

[Co]: Following the last character of this line, li , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Cp]: Graph x46/15 ( ) yu pleasure is written appreciably dif( ).

ferently in the Shnghai Xng qng ln (xq38/15):

[Cq]: I follow Li Zho and read x47/5 ( ) as mo to have troubled eyesight, here understood as confused, demented.101 Following the last character of this line, mo , a small reading mark is added to the manuscript of the Shnghai Xng qng ln.

98 99 100 101

See, e.g., Li Zho 2005, 102; Dng Yunzh 2002, 242. See L Lng 2002, 110. See Li Zho 2005, 102. Ibid.

348

chapter twelve

[Cr]: Instead of kui (OC *[k]whret-s) to be satisfied as on the Gudin One strip (x47/11), the Shnghai Xng qng ln has hu ([g]wh[e][t]-s) intelligent (xq38/36). Following the last character of this line, k , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. ( ) Chn Wi states that this character [Cs]: As for graph x47/23 allows two distinct readings: yun or yun sincere and yun deep; silent.102 As the top of Shnghai strip xq39 has broken off, the Shnghai Xng qng ln proves of no help for this passage. I tentatively follow Li Zho and read the graph as yun deep; silent.103 As for graph x48/2 ( ), I follow Li Zho in his identification of the graph as bu (OC *(m)-pa) to correct (note, however, that Li reads it in the sense of to nourish).104 The graph consists of the signific m wood and the phonophoric f (OC *[b](r)a). Following the last character of this line, z , the Shnghai Xng qng ln adds a large mark on the strips . [Ct]: Both texts have the graph for wi artificial, fake (strip x48/7 and x48/12; xq39/8) as consisting of the signific heart and the phonophoric wi (OC *[g]w(r)aj) ( ). In the Shnghai Xng qng ln the graph is doubled by a sign for repetition =. Also, in both texts the graph for is written without the signific heart (). [Cu]: On graph x48/14 ( ) read as ln (OC *(m-)r[n]-s) (here in the sense of to have the sense of shame), see the article by Chn Jin on the Kongz sh ln.105 The graph is composed of two mouths and the phonophoric wn (OC *m[n]). The Shnghai Xng qng ln also adds the signific heart beneath the phonophoric wn (xq39/14). Following the last character of this line, y , a small reading mark is added to the manuscript of the Shnghai Xng qng ln.

102 103 104 105

See Chn Wi 2003, 206. See Li Zho 2005, 102. Ibid. See Chn Jin 2002a.

reconstructing the xng z mng ch

349

[Cv]: Instead of closing this building block with the particle y as in the Gudin One manuscript, there is a large mark on the strips of the Shnghai Xng qng ln. ( ) shn (OC *[d]i[n]-s) carefully, [Cw]: As for graph x49/3 the Shnghai Xng qng ln has a graph consisting of sh (OC *[g] ( ).106 [i]p) and yn (*a[n]): xq39/26 Instead of () (OC *ni[n]), Shnghai manuscript has l (OC *[r]a-s) (xq39/18). [Cx]: Following the last character of this line, ji , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Cy]: As for graph x49/25 see my discussion in [Cw]. ( ) shn (OC *[d]i[n]-s) carefully,

() (xq40/4 in the Shnghai Xng qng Following graph x49/29 ln), xn y ()() this truly is the case, both texts have the tadpole symbol. In the Gudin One manuscript, however, the mark is bleached out. In the Shnghai Xng qng ln the entire bamboo strip carries no further graphs after the tadpole symbol. [Cz]: The beginning lines of strip x51 [x51()() () ()() ()()]in the passage x50/1 59/9are missing in the Shnghai Xng qng ln, even though the strips are not broken at this point. This must be explained as a difference in the particular manifestation of the Shnghai Xng qng ln. ( ) ln (here in the [Da]: For a discussion of graph x59/15 sense of to feel shame), see [Cu]. Note that instead of adding the signific heart xn beneath the phonophoric wn (OC *m[n]), the Shnghai Xng qng ln has the signific earth tu (xq29/33). [Db]: On graph x60/2 under [Aj]. ( ), see my discussion of graph x16/18

106

See also my discussion under [Ba].

350

chapter twelve

As above (x48/7 and x48/12), I read gu as we artificial. The two graphs share the same phonophoric and thus fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. The Shnghai Xng qng ln also has gu (xq30/13). [Dc]: On graph x60/10 ( ), see my discussion of x30/11 fierce (and x31/6) under [Bh]. On graph x60/12 ( ( ) li

) sh to cause, see my discussion of x8/4

under [P]. Following the last character of this line, m , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [Dd]: Graph x60/17 is problematic. L Lng holds that the transcription by the Hbi Province Museum editors is incorrect.107 In his earlier publication, he proposed reading the graph as zhng (here: to draft the capable and virtuous for public service).108 Later he reads it as l street instead.109 He finds corroboration for this in the Shnghai Xng qng ln ( ) (xq30/25). Whereas the Gudin One Xng z mng <> <>, in the Shnghai Xng qng ln

ch has

graph xq30/25 ( ) is used in combination with do . Most editors regard this as the correct reading. Both texts write the graph for whatas most editors suggestshould be read as s to ponder. Following the last character of this line, yn , a small reading mark is added to the manuscript of the Shnghai Xng qng ln. [De]: As for ch to dwell, the Gudin One text has and the Shnghai Xng qng ln has (xq30/32). (x61/1),

[Df ]: Instead of d hi great calamity as in the Gudin One version, the Shnghai Xng qng ln has only hi .

107 108 109

See Hbi shng Jngmn sh bwguan 1998, 181. See L Lng 1999, 507. See L Lng 2002, 111.

reconstructing the xng z mng ch

351

[Dg]: For the graph x62/3 , hun (*[g]ro[n]-s), the Shnghai Xng qng ln has a graph consisting of the phonophoric juan ( ) (xq31/22). The (*[k]ro[n]) with the signific heart beneath two forms clearly fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. Following the last character of this line, hu , a huge mark is added on the strip of the Shnghai Xng qng ln. The Gudin One manuscript continues with pericope 20. The Shnghai Xng qng ln continues with what is pericope 13 in the Gudin One manuscript. The part with which the Gudin One manuscript continues appearsin substantially different formon strips xq2728 in the Shnghai Xng qng ln.

CHAPTER THIRTEEN

RECONSTRUCTING THE TI Y SHNG SHUI This chapter provides the text and translation of the Ti y shng shu, followed by philological notes. Text and Translation The Ti y shng shu consists of two cantos, the Cosmogony and its Application. Cosmology: Canto 1
ty1

[A] [] ty2 [B] ty3 [] ty4 The Ultimate One gives birth to water, water returns and joins with the Ultimate Onethis is how it completes Heaven. Heaven returns and joins with the Ultimate Onethis is how it completes Earth. Heaven and Earth {repeatedly join with each other} ty2 this is how they complete the spirits and the illuminated. The spirits and the illuminated repeatedly join with each otherthis is how they complete the shadowy (yn) and the sunny (yng). The shadowy and the sunny repeatedly join with each otherthis is how they complete the four seasons. The four seasons ty3 repeatedly join {with each other}this is how they complete coldness and heat. Coldness and heat repeatedly join with each otherthis is how they complete moisture and dryness. When moisture and dryness repeatedly join with each other, completing the year, ty4 [the circle] stops.
ty1

354

chapter thirteen

[]; [C] [] ty5 ty6 Therefore, the year is begotten by moisture and dryness; Moisture and dryness are begotten by coldness and heat; Coldness and heat {are begotten} by the four seasons; The ty5 { four seasons} are begotten by the shadowy and the sunny; The shadowy and the sunny are begotten by the spirits and the illuminated; The spirits and the illuminated are begotten by Heaven and Earth; ty6 Heaven and Earth are begotten by the Ultimate One. [] [][][] ty7 [D] ty8 [E] [][] [F] From this it follows that the Ultimate One is stored in the water, moves with the [four] seasons, [finishes] a circle, and then {starts over again}. . . . {Hence the Ultimate One is} ty7 the mother of the myriad things; at times lacking, at times full, it takes itself to be the alignment of the myriad things. This is what Heaven is unable to kill, Earth ty8 is unable to smother, the shadowy and the sunny are unable to complete. The gentleman ( jnz) who grasps this is the one whom we call a {sagacious person} . . .

Application: Canto 2
. . . ty10 [G] [H] . . . .ty10 Below is soil, yet we call it earth. Above is vapour, yet we call it heaven. Do likewise is [only] a style name for itMay I [thus] ask for its [real] name? ty11 [I] ty12 He who ty11 carries out his service according to the do necessarily consigns himself to its [real] name; this is why [his] task is completed and his person can endure.

reconstructing the ti y shng shu

355

In carrying out his service, the sagacious person also consigns himself to its ty12 [real] name; this is why [his] achievements are completed and his person will not suffer harm. [J] [] ty13 [][. . .] [K] As for Heaven and Earth, their name and style name are established simultaneously; as a result, once moving beyond these realms, [one] does not think [of them] appropriately . . . . . . {If Heaven does not suffice} ty13 in the northwest, what is below it rises in strength. If Earth does not suffice in the southeast, what is above it {rises in strength} . . .
ty9

[L] [X] [M] [] ty14 The Way of Heaven is to value weaknessit reduces what is completed so as to add to what is living. Cutting back on strength, requesting from {. . .}. {This is why what does not suffice above}, . . . ty14 will have a surplus below. [And] what does not suffice below will have a surplus above.
ty9

Notes on Text and Translation Cosmology: Canto 1 d (OC *[d]a[t]-s) great is read as ti [A]: Graph ty1/1 (OC *tha[t]-s) ultimate. The term d y the great one (or ultimate one) also appears in the L j, L yn , where it denotes something that exists prior to Heaven and Earth, similar to what is described in the Ti y shng shu.1 Chn Wi remarks that throughout the Warring States period, the expression ti y (OC *tha[t]-s i[t]),

1 The L j L Yn reads: That is why the eminent rites by necessity are rooted in the Ultimate One: It split up and became Heaven and Earth. It turned around and became the shadowy and the sunny. It altered and became the four seasons. It distributed and became the ghosts and the spirits. See L j zh sh 1997, 438.

356

chapter thirteen

not d y *[d]a[t]-s i[t]), was used predominantly.2 I agree with his observation and read the graph as ti (*tha[t]-s) throughout.3 The editors of the Hbi Province Museum read graph ty1/6 ( ) as fu (OC *b(r)a) to assist.4 Chn Wi proposes reading it as b (OC *(C.)[b]a) to make contact with each other; to join each other.5 I follow Chn. Throughout the text, the Ti y shng shu uses marks for repetition .

[B]: After graph ty1/21 the strip is broken. The upper part of graph ty1/22 is still visible . Based on the repetitive structure of this passage and the part remaining on the strips, the graph should be reconstructed as d earth. Comparing the length of strip ty1 with the lengths of the remaining strips, we can estimate that three graphs are missing subsequent to the reconstructed graph (d ) earth. Based on the repetitive structure of the passage, it can be reconstructed as follows: [ ()] () () Earth returns and joins with . . ..
Comparison of the missing part of strip t1 with the end of strip t2

[C]: Strip ty4 ends with the two graphs s sh () four seasons. Strip ty5whose position in the sequence is beyond doubtstarts with () () (). Hence, there is a clear break in the pattern. If we compare it with the previous sentence, the line in question should run as follows: [] . It is unlikely that the pattern has been broken deliberately in order to signal a new idea. The four seasons, just like the shadowy (yn) and sunny (yang), would then have been created by coldness and heat, which goes against the pattern. I therefore understand this

2 3 4 5

See Chn Wi 2003, 25. On the concept Ultimate One (or Great One), see also Allan 2003. See Hbi shng Jngmn sh bwguan 1998, 125, n. 2. See Chn Wi 2003, 26.

reconstructing the ti y shng shu

357

break, not as a systematic gap, but a mistake that occurred in the process of copying the text. [D]: I follow Qi Xgu, who reads graph ty6/17 ( ) as zhu to

hu should be read in the sense encircle, circle.6 Graph ty6/19 of yu again. After ty6/19 the strip is broken. Based on the length of the missing part, presumably four or five graphs are missing. Qi Xgu reconstructs the missing passage as to make itself into.7 However, considering the preceding sentence, it is more likely that the present line should continue with sh to start with, to begin with [again]. Li Zho reconstructs four graphs. He also begins the reconstructed passage with to start with, to begin with [again] and then continues with Qis reconstruction.8 This seems to be corroborated by the subsequent statement, which reads () () ()() it takes itself to be the alignment of the myriad things (ty7/813). Nevertheless, I suggest reconstructing the passage as follows:
[ ] [] [] ty7 From this it follows that the Ultimate One is stored in the water, moves with the [four] seasons, [finishes] a circle, and then {starts over again}. {Hence, the Ultimate One is} ty7 the mother of the myriad things; at times lacking, at times full, it takes itself to be the alignment of the myriad things.

I take the first line as purely descriptive, ending with and then {starts over again}. The second line is a clarification and definition of the Ultimate One based on the description: Hence, the Ultimate One is . . .. The fact that the phrase is a rhyme (: OC *le; : OC *ke), whereas the preceding line has no rhymed passage, seems to corroborate my assumption that only the last line is some type of defining statement and that the preceding line should stop after start over [again].

6 7 8

See Hbi shng Jngmn sh bwguan 1998, 126, n. 12. Ibid. See Li Zho 2005, 45.

358

chapter thirteen
( ) is read y one.9

The character ty7/4 (and ty7/6)

[E]: The editors of the Hbi Province Museum read graph ty8/3 ( ) as l diminish.10 As L Lng already noted,11 in the Xnz R xio there appears an identical line, which reads: Heaven is unable to kill [it]; Earth is unable to smother it.12 Other sources also suggest that this notion was current during the Warring States period.13 Accordingly, most scholars follow L Lng in his reading of ty8/3 as mi to smother.14 At first sight, this reconstruction seems to be difficult to defend on phonological grounds since () (OC *(m.)r) has an initial *r-, and so the phonophoric in does l (OC *[r]) diminish (both have the Middle Chinese li initial [OC *(m.)r < *m.rk]), whereas mi (OC *m.r) has the Middle Chinese mng initial (OC *mra), which is a bilabial one. One of the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components is that the initial of two words should have the same position of articulation. This is clearly not the case here. However, it cannot be ruled out that the bilabial initial *m- of mi to smother just reflects a prefix, and the word should indeed be reconstructed as *m-r, which would make this a good phonetic loan. [F]: The latter part of strip ty8 is broken off. Presumably seven or eight graphs are missing. The last graph on the strip is only partly visible , and what remains suggests that it should be read as wi () to call; be called. For my reconstruction of the first missing graphs as shng rn sagacious person, see my discussion in chapter 6. Application: Canto 2 [G]: In the present edition of the Ti y shng shu , the editors of the Hbi Province Museum place strip ty9 after ty8 and before ty10. I have, instead, placed it immediately before strip ty14.15
9 For a detailed discussion of this graph, see Ln Hngyng, Murase Nozomu, and Furuya Akihira 2004. 10 See Hbi shng Jngmn sh bwguan 1998, 126, n. 14. 11 See L Lng 2002, 33. 12 See Xnz j ji 1997, 1:138. 13 See Chn Wi 2003, 29. 14 See, e.g., ibid.; Li Zuxn 2004, 254; Li Zho 2005, 45. 15 See Hbi shng Jngmn sh bwguan 1998, 125.

reconstructing the ti y shng shu

359

The location of strip ty9 is still an issue. Originally, the editors of the Hbi Province Museum placed it before strip ty13 and after ty12. It was only due to a suggestion by Qi Xgu that they finally decided to locate it following strip ty8 and before ty10.16 Later, however, Qi withdrew his proposal. He now holds that strip ty9 should be placed right before the final strip, ty14, and after ty13.17 Qis latest suggestion proposes the following order of strips: 113, 9, 14. This gives the following clusters: strips ty18 outline the cosmology of the Ti y shng shu, strips ty1013 discuss the importance of proper naming, and strips ty9 and 14 discuss the fact that the Ultimate One in itself values weakness.18 Most editors of the text now follow this organisation.19 I consider strips ty18 as composing canto 1 and strips ty1013, 9, and 14 as composing canto 2. [H]: I follow Qi Xgu in reading graph ty10/18
20

) as z style

name. Following ty10/18 is some type of marking on the strip . Its function remains unclear. The Hbi Province Museum editors read the line ty10/2023 () () as may I ask for its name? Many editors () () follow this suggestion.21 However some argue for reading ty10/21 () in its original form, that is, as hn dark.22 Qi Xgu strongly disapproves of this,23 despite the fact that on merely palaeographical or phonological grounds both readings are possible. I follow Qis suggestion and read ty10/2021 as () ()may I ask [for], which, however, I base not merely on the line ty11/58 but on the argumentative structure of the Ti y shng shu overall.24 [I]: The editors of the Hbi Province Museum read ty11/6 ( ) as tuo to entrust.25 Qi Xgu and Chn Wi read ty11/58 as () () must [also] consign [himself ] to its name.26 ()
16 See Qi Xgu 2000b, 219220. See also my discussion in chapter 7 of the present study. 17 Qi Xgu 2000b, 220. 18 Ibid., 221. 19 See, e.g., Chn Wi 2003, 24; Li Zuxn 2004, 254; Li Zho 2005, 46. 20 See Hbi shng Jngmn sh bwguan 1998, 126, n. 16. 21 See, e.g., Wi Qpng 1999; Li Zho 2005, 46. 22 See, e.g., L Lng 2002, 32; Chn Wi 2003, 30; Li Zuxn 2004, 254. 23 See Qi Xgu 2000b, 222ff. 24 See my discussion of the Ti y shng shu in chapter 7. 25 See Hbi shng Jngmn sh bwguan 1998, 125. 26 See Qi Xgu 2000b, 223224; Chn Wi 2003, 29.

360

chapter thirteen

[J]: The end of strip ty12 broken off. Presumably three or four graphs are missing. The last graph on the strip is partly visible: . Qi Xgu notes that this could be either shng or a graph written with the phonophoric shng (OC *[d]a-s), here read as dng (*ta).27 I follow Qis reconstruction of this passage: [] ty13 () . . . {[If ] Heaven does not suffice} ty13 in the northwest.28 [K]: The end of strip ty13 is broken off. Presumably seven graphs are missing. Following the suggestion of Li Zho, I reconstruct the first three graphs parallel to those on strip ty12.29 [L]: For the position of strip ty9, see [G] above. () The editors of the Hbi Province Museum read graph ty9/5 as ju dignity.30 I follow Qi Xgus suggested reading as xu to cut, reduce.31 Qu can be reconstructed as Old Chinese *[ts]ewk; xu as Old Chinese *[s]ewk. The two fulfil the criteria for phonetic similarity in Old Chinese for loan characters and phonetic components. [M]: The end of strip ty9 is broken off. Given the length of the lost part, presumably seven graphs are missing. Although it is absolutely impossible to reconstruct the missing part of the line () () [X] Cutting back on strength, requesting from {. . .}, it is indeed possible to reconstruct the subsequent passage with some certainty. The passage begins with the statement
ty9

() () ()() () () [X]

() ()

()

ty9

The Way of Heaven is to value weaknessit reduces what is completed so as to add to what is living.. Cutting back on strength, requesting from {. . .}.

Then about six or seven graphs are missing and the passage continues with

27 28 29 30 31

See Hbi shng Jngmn sh bwguan 1998, 126, n. 17. Ibid. See Li Zho 2005, 47. See Hbi shng Jngmn sh bwguan 1998, 125. Ibid., 124, n. 15.

reconstructing the ti y shng shu

ty14
ty9

361

()()()()

{. . .} ty14 will have a surplus below. [And] what does not suffice below will have a surplus above.

It is clear that the first few lines provide background information for the last two lines, which, in turn, are parallel and seem to complement each other:
1. the Way of Heaven values weakness 2. it reduces what is accomplished to add to life 3. it cuts back on strength and requests from 4. {. . .} 5. and what does not suffice below (X does + on ) (X does so as to +) (X does [. . .]) has a surplus below (X has + on ) has a surplus above (X has + on )

Hence, I feel justified in reading the fourth sentence as parallelbut in oppositionto the fifth sentence:
XX {that what does not suffice above} has a surplus below; [And] that what does not suffice below has a surplus above.32

32 For this reading, see also Li Zho 2005, 47: {This is why () what does not suffice on high}. . . .

BIBLIOGRAPHY
Al Azharia Jahn, Samia. 2006. Comparative Studies on Different Concepts about the Origin of Writing on Palm Leaf: BotanyTraditional TechnologiesDivine Teachers. Asiatische Studien/tudes asiatiques 60.4:921961. Allan, Sarah. 19721973. The Identities of Taigong Wang in Zhou and Han Literature. Monumenta Serica 30:5799. . 1981. The Heir and the Sage: Dynastic Legend in Early China. San Francisco: Chinese Materials Center. . 2003. The Great One, Water, and the Laozi: New Light from Guodian. Toung pao 89.45:237285. Allan, Sarah Allan, and Crispin Williams, eds. 2000. The Guodian Laozi: Proceedings of the International Conference, Dartmouth College, May 1998. Early China Special Monograph Series, no. 5. Berkeley: Society for the Study of Early China and the Institute of East Asian Studies, University of California. Ames, Roger T., and Henry Rosemont. 1998. The Analects of Confucius: A Philosophical Translation. New York: Ballantine Books. Aristotle. 1920. Peri hermenias oder Lehre vom Satz: Des Organon zweiter Teil; Neu bersetzt und mit einer Einleitung und erklrenden Anmerkungen versehen von Eug. Rolfes. Philosophische Bibliothek, vol. 9. Leipzig: Meiner. Asselin, Mark Laurent. 1997. The Lu-School Reading of Guanju as Preserved in an Eastern Han fu. Journal of the American Oriental Society 117.3:427443. Assmann, Aleida, and Jan Assmann. 1987a. Kanon und Zensur als kultursoziologische Kategorien. In J. Assmann and A. Assmann 1987, 746. . 1998b. Schrift und Gedchtnis. In A. Assmann, J. Assmann, and Hardmeier 1998, 265284. Assmann, Aleida, Jan Assmann, and Christof Hardmeier, eds. 1998. Schrift und Gedchtnis: Beitrge zur Archologie der literarischen Kommunikation. Munich: Wilhelm Fink. Assmann, Jan. 1999. Das kulturelle Gedchtnis: Schrift, Erinnerung und politische Identitt in frhen Hochkulturen. Munich: C. H. Beck. Assmann, Jan, and Aleida Assmann, eds. 1987. Kanon und Zensur. Archologie der literarischen Kommunikation 2. Munich: Wilhelm Fink. Bagley, Robert W. 1999. Shang Archaeology. In Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., 124231. Cambridge: Cambridge University Press. . 2004. Anyang Writing and the Origin of the Chinese Writing System. In Houston 2004, 190249. Bal, Mieke. 2001. Introduction: Travelling Concepts and Cultural Analysis. In Goggin and Neef 2001. Baxter, William H. 1977. Old Chinese Origins of the Middle Chinese chngniu Doublets: A Study Using Multiple Character Readings. Ph.D. diss., Cornell University. . 1992. A Handbook of Old Chinese Phonology. Berlin: Mouton de Gruyter. . 1998. Situating the Language of the Lao-tzu: The Probable Date of the Tao-teching. In Kohn and LaFargue 1998, 231253. Behr, Wolfgang. 2006. Three Sound-Correlated Text Structuring Devices in Pre-Qin Philosophical Prose. In Wolfgang Behr and Joachim Gentz, eds., Komposition und KonnotationFiguren der Kunstprosa im Alten China), 1533. Bochum: Bochumer Jahrbuch zur Ostasiatischenforschung 29. Munich: Iudicium.

364

bibliography

Behr, Wolfgang and Bernhard Fhrer. 2005. Einfhrende Notizen zum Lesen in China mit besonderer Bercksichtigung der Frhzeit. In Bernhard Fhrer, eds. Aspekte des Lesens in Vergangenheit und Gegenwart (DVS)/Aspects of Reading in China, Past and Present: Collected Papers of the 2001 Annual Meeting of the German Association for Chinese Studies (DVCS), 143. Bochum: Projekt. Bijng Dxu chtu wnxin ynjisuo . 2009. Gngzu jianbo. Bijng Dxu chtu wnxin ynjisuo no. 1 (September), 17. Berlin, Adele. 1985. The Dynamics of Biblical Parallelism. Bloomington: Indiana University Press. Berner, Ulrich. 2004. The Concept of Syncretism: An Instrument of Historical Insight/Discovery? In Anita M. Leopold and Jeppe S. Jensen, eds., Syncretism in Religion: A Reader, 295315. London: Equinox. Billeter, Jean-Franois. 2006. Contre Franois Jullien. Paris: ditions Allia. Bodde, Derk. 1986. The State and Empire of Chin. In Twitchett and Loewe 1986, 20202. Boltz, William G. 1984. Textual Criticism and the Ma-wang-tui Lao tzu. Harvard Journal of Asiatic Studies 44.1:185224. . 1985. The Lao tzu Text that Wang Pi and Ho-shang Kung Never Saw. Bulletin of the School of Oriental and African Studies 42.3:493501. . 1994. The Origin and Early Development of the Chinese Writing System. New Haven, CT: American Oriental Society. . 1997. Manuscripts with Transmitted Counterparts. In Shaughnessy 1997f, 253284. . 1999a. The Fourth-Century B.C. Guodiann Manuscripts from Chuu and the Composition of the Laotzyy. Journal of the American Oriental Society 119.4:590 608. . 1999b. Language and Writing. In Michael Loewe and Edward L. Shaughnessy, eds. The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C. Cambridge: Cambridge University Press, 74123. . 2000. The Study of Early Chinese Manuscripts: Methodological Preliminaries. In Allan and Crispin 2000, 3952. . 2002. Liijih Tzy i and the Guodiann Manuscript Matches. In Reinhard Emmerich and Hans Stumpfeld, eds., Und folge nun dem, was mein Herz begehrt: Festschrift fr Ulrich Unger zum 70. Geburtstag, 1:209222. Hamburger Sinologische Schriften 8. . 2005. The Composite Nature of Early Chinese Texts. In Kern 2005c, 5078. Brindley, Erica. 2006. Music, Cosmos, and the Development of Psychology in Early China. Toung Pao 92.13:149. Brooks, E. Bruce. 1994. The Present State and Future Prospects of Pre-Han Text Studies: A Review of Michael Loewe, ed., Early Chinese Texts: A Bibliographical Guide. Sino-Platonic Papers 46:174. Brooks, E. Bruce, and A. Taeko Brooks. 1998. The Original Analects: Sayings of Confucius and His Successors. New York: Columbia University Press. Caudwell, Christopher. 1946. Illusion and Reality: A Study of the Sources of Poetry. London: Lawrence & Wishart. Chan Wing-tsit. 1963. The Way of Lao Tzu (Tao-te ching). New York: Macmillan. Chavannes, E. 1905. Les livres chinois avant linvention du papier. Journal Asiatique 5:575. Chn Guyng . 1992a. Ln Laoz wan ch shuo zi kaozhng fangfa shng chng jin de wmijin ln Liz fei wishu . Paper presented at Laoz sixiang ynji hu, Xn, May.

bibliography

365

. 1992b. Ti y shng shu yu Xng z mng ch fawei . Doji wnhu ynji 17:393411. . 1997. Laoz jnzh jny . 2d rev. ed. Taibei: Tiwn Shngw. . 1998. Ch d jianbn Laoz . Wnw 10:5556. Chn Jin . 2002a. Kongz sh ln bu sh y z. Zhnggu zhxu 24:222225. . 2002b. Sh Zhng xn zh do de pi z . Guj jianb ynji tng xn 2:26. . 2003. Gudin jian bu sh sn pin . In Gu m xn zh: Gudin Chu jian chtu sh zhu nin lnwn zhun j : , 114131. Hong Kong: Guj Ynhung wnhu . . 2004. Gudin Chu jian Qing d y sh, Yu cng s de j ch jianx tiozh . In Sarah Allan and Xng Wn, eds., Xn ch jianb ynji: jianb guj xush yntao hu wnj, 2000 nin 8 yu, Bijng 20008, , 316 322. Bijng Dxu Zhndn Gudi Wnmng Ynji Zhngxn xush cngsh zh b . Beijing: Wnw. Chn Li . 1998. Gudin Chu jian zh Xng z mng ch pin ch tn . Kongz ynji 3:5260. Reprinted as Jngmn zhjian zh Xng z mng ch pian chutn, Zhnggu zhxu 20 (1999):293314. Chn Lgu . 2005. Ti y shng shu ynji zongsh j q yu Laoz bng de xinggun wnt. Hnxu ynji 23.2:413437. Chn Lnqng , Zhng Yshn , and Zu Jnzh . 2004. Shnghai Bwguan cng Zhngu Chu zhsh (y) dbn. Edited by J Xshng . Taibei: Wnjunlu. Chn Mngji . 1985. Shngsh tngln . Beijing: Zhnghu. Chn Nng . 1998. Gudin Chu m zhjian zhng de Rji rnxng ynln chtn. Zhnggu zhxush , 3946. Chn Sngchng and F Juyou , eds. 1992. Mawngdu Hn m wnw . Chngsh: Hnn. Chn Tngshng . 2004. Kongz sh ln ynji. Beijing: Zhnghu. Chn Wi . 1998. Gudin zhsh bish . Jing hn kaogu 4:6772. . 1999. Ti y shng shu kaosh. In Gu wnz yu gu wnxin , 6566. Taibei: Chu wnhu ynji. . 2000a. Guny Gudin Chu jian Li d zh pin binlin de tiozhng . In Gudin Chu jian guj xush yntaohu lnwnj , 6474. Wuhn: Hbi jioy Press. Originally published in Jing Hn kaogu, 2000, 1. . 2000b. Ti y shng shu jiod bng ln yu Laoz de gunx . Gu wnz ynji 22:227231. . 2000c. Yu cng y, sn zhng guny l de j tio jian wn . In Gudin Chu jian guj yntao hu huy lnwn j , 143148. Wuhn: Hbi jioy. . 2001. Gudin jian sh Zn d y jio sh . Zhnggu zhxu sh 3:108120. . 2003. Gudin zhsh bish (Xn ch jianb ynji cngsh ). Wuhn: Hbi jioy Press. First published in 2002. Cheng, Anne. 2005. Y a-t-il une philosophie chinoise? Est-ce une bonne question? Extrme-Orient, Extrme-Occident 27:512.

366

bibliography

Cheng Fong Ching, and D. C. Lau, eds. 1992. The ICS Ancient Chinese Concordance Series . Hong Kong: Commercial Press. Chnqi Gngyng zhun zh sh . 1997. Edited by Ruan Yun . In Shsn jng zh sh , vol. 7. Reprint, Taibei: Ywn ynshguan. Chnqi Zuo zhun zhng y . 1997. Edited by Ruan Yun . In Shsn jng zh sh , vol. 6. Reprint, Taibei: Ywn ynshguan. Clanchy, Michael T. 1979. From Memory to Written Record: England, 10661307. London: Edwin Arnold. Clark, Elizabeth A. 2004. History, Theory, Text: Historians and the Linguistic Turn. Cambridge, MA: Harvard University Press. Connerton, Paul. 1989. How Societies Remember. Cambridge: Cambridge University Press. Cook, Constance A., and John S. Major, eds. 1999. Defining Chu: Image and Reality in Ancient China. Honolulu: University of Hawaii Press. Cook, Scott [G Shkao ]. 2000. Consummate Artistry and Moral Virtuosity: The Wu xing Essay and Its Aesthetic Implications. Chinese Literature: Essays, Reviews, Articles 22:113146. . 2006. Gudin Chu jian Chng zh wn zh dng pin zzh . Qnghua Dxu xubo (zhxu shhu kxu ban) () 1:8092. Creel, Herrlee. 1949. Sinicisma Clarification. Journal of the History of Ideas 10:1, 135140. Crump, John Irving. 1979. Chan-kuo Tse. San Franscisco: Chinese Materials Center. . 1998. Legends of the Warring States: Persuasions, Romances, and Stories from Chan-kuo Tse. Michigan Monographs in Chinese Studies 83. Ann Arbor: Center for Chinese Studies, University of Michigan. Csikszentmihalyi, Mark. 2004. Material Virtue: Ethics and the Body in Early China. Sinica Leidensia, vol. 66. Leiden: Brill. Csikszentmihalyi, Mark, and Michael Nylan. 2003. Constructing Lineages and Inventing Traditions through Exemplary Figures in Early China. Toung Pao 89:5999. Cu Rny . 1997. Shln Jngmn zhjian Laoz de nindi . Jngmn Dxu xubo 2:3842. . 1998. Jngmn Gudin Chu jian Laoz ynji . Beijing: Kxu. Davis, John F. 1830. Poeseos sinensis comentarii: On the Poetry of the Chinese. Transactions of the Royal Asiatic Society 2:393461. Reprinted, London, 1970. . 1870. Hanwen shi jie: Poeseos Sinicae Commentarii: The Poetry of the Chinese. London: Asher. Defoort, Carine. 2001. Is There Such a Thing as Chinese Philosophy? Arguments of an Implicit Debate. Philosophy East and West 51.3:393413. . 2006a. Is Chinese Philosophy a Proper Name? A Response to Rein Raud. Philosophy East and West 56.4:625660. . 2006b. Mohist and Yangist Blood in Confucian Flesh: The Middle Position of the Guodian Text Tang Yu zhi Dao. Bulletin of the Museum of Far Eastern Antiquities 76:4470. . 2007. Argumentation and Persuasion in Ancient Chinese Texts: Introduction. Oriens Extremus 45:9198. Deleuze, Gilles, and Flix Guattari. 1980. Mille plateaux. Paris: ditions de Minuit. Dng Jinpng . 2003. Tng Y zh do li d xn sh . http://www.jianbo.org/Zzwk/2003/WUHANHUI/dengjianpeng.htm (accessed February 23, 2006). Dng Sxn . 2000a. Gudin Chu m zhjian sxiang ynji . Beijing: Dngfng.

bibliography

367

. 2000b. Gudin jiansh zh tin-rn zh bin . In Wuhn Dxu Zhnggu wnhu ynji yun, ed., Gudin Chu jian guj xush yntao hu lnwn j , 582588. Rnwn Lncng t j . Wuhn: Hbi rnmn. . 2002. Chu jian Ti y shng shu ynjijin ln dngqin Ti y shng shu ynji de zongt ppng . In Chu d chtu jianb wnxin sxiang ynji , 183249. Wuhn: Hbi jioy. Dng Yunzh . 2002. Chu jian R ji xng qng shu ynji . Chtu wnxin y zh ynx cngsh P015. Taibei: Wnjunlu. Dun Yci , ed.1998. Xntin guyn Shuwn jiz . Taibei: Hngy wnhu. Dubs, Homer H. 1928. The Works of Hsntze: Translated from the Chinese with Notes. Probsthains Oriental Series, vol. 16. London: Probsthain. Ehlich, Konrad. 1982. Zum Textbegriff. Tilburg: Mimeo. . 1998. Text und sprachliches Handeln: Die Entstehung von Texten aus dem Bedrfnis nach berlieferung. In A. Assmann, J. Assmann, and Hardmeier 1998, 2443. Eisenstadt, Shermuel, ed. 1983. Die Kulturen der Achsenzeit. Frankfurt a. M: Suhrkamp. Eno, Robert. 1990. The Confucian Creation of Heaven and the Defense of Ritual Mastery. Albany: State University of New York Press. Ess, Hans van. 2007. Argument and Persuasion in the First Chapter of Huainanzi and Its Use of Particles. Oriens Extremus 45:255270. Falk, Harry. 1990. Goodies for India: Literacy, Orality, and Vedic Culture. In Wolfgang Reible, ed., Erscheinungsformen kultureller Prozesse, 103120. Tbingen: Gunter Narr. Falkenhausen, Lothar von. 1993. Issues in Western Zhou Studies: A Review Article. Early China 18:139226. . 2006. Chinese Society in the Age of Confucius (1000250 BC): The Archaeological Evidence. Los Angeles: Cotsen Institute of Archaeology, University of California. Fn Yzhu . 2002. Gudin Chu jian Tng Y zh do de shwn, jianx yu fnzhng . http://www.jianbo. org/Wssf/2002/fanyuzhou03.htm (accessed July 19, 2007). Farmer, Steve. 1998. Syncretism in the West: Picos 900 Theses (1486); The Evolution of Traditional Religious and Philosophical Systems. Tempe, AZ: Mediaval and Renaissance Texts and Studies. . 2006. Neurobiology, Stratified Texts, and the Evolution of Thought: From Mythos to Religions and Philosophies. Expanded version of a talk given at the Harvard Peking University International Conferences on Comparative Mythology, Beijing. http://www.safarmer.com/Farmer.Beijingpaper2.pdf (accessed June 15, 2007). Farmer, Steve, John B. Henderson, and Peter Robinson. 2002. Commentary Traditions and the Evolution of Premodern Religious and Philosophical Systems: A Cross-Cultural Model. http://www.safarmer.com/model.pdf (accessed June 15, 2007). Farmer, Steve, John B. Henderson, and Michael Witzel. 2002. Neurobiology, Layered Texts, and Correlative Cosmologies: A Cross-Cultural Framework for Premodern History. Bulletin of the Museum of Far Eastern Antiquities 72, 4890. Farmer, Steve Richard Sproat, Michael Witzel. 2004. The Collapse of the Indus-Script Thesis: The Myth of a Literate Harappan Civilization. Electronic Journal of Vedic Studies 11:2, 1957. Friedrich, Michael. 1996. Zur Datierung zweier Handschriften des Daode jing. Textkritische Beitrge 2:97104.

368

bibliography

Fhrer, Bernhard, ed. 2005. Aspekte des Lesens in China in Vergangenheit und Gegenwart: Referate der Jahrestagung 2001 der Deutschen Vereinigung fr Chinastudien (DVCS)/Aspects of Reading in China, Past and Present: Collected Papers of the 2001 Annual Meeting of the German Association for Chinese Studies (DVCS). Bochum: Projekt. Gabelentz, Georg von der. 1878. Ein Probestck von chinesischem Parallelismus. Zeitschrift fr Vlkerpsychologie und Sprachwissenschaft 10:230234. . 1881. Chinesische Grammatik mit Ausschluss des niederen Stiles und der heutige Umgangssprache. Leipzig: Weigel. Reprint, Berlin: Deutscher Verlag der Wissenschaften, 1956. . 1883. Anfangsgrnde der chinesischen Grammatik, mit bungsstcken. Leipzig: Weigel. Galambos, Imre. 2006. Orthography of Early Chinese Writing: Evidence from Newly Excavated Manuscripts. Budapest Monographs in East Asian Studies 1. Budapest: Department of East Asian Studies, Etvs Lornd University. Go Xiang . 1987. Guwn z tngji hudian . Beijing: Qlu sh sh. Go Yurn . 2006. Tn Tng Y zh do yu Co m zh chn de m z . In Bamboo and silk manuscripts. http:// www.bsm.org.cn/show_article.php?id=145 (accessed February 13, 2006). Gardner, Daniel K. 1998. Confucian Commentary and Chinese Intellectual History. Journal of Asian Studies 57.2:397422. Gassmann, Robert H. 2000. Understanding Ancient Chinese Society: Approaches to Rn and Mn . Journal of the American Oriental Society 120.3:348359. . 2005. Preliminary Thoughts on the Relationship Between Lexicon and Writing in the Guodian Texts. Asiatische Studien/tudes asiatiques 59.1:233260. Gatzemeier, Matthias. 1985. Wahrheit und Allegorie: Zur Frhgeschichte der Hermeneutik von Theagenes bis Proklos. In Volker Gerhard and Norbert Herold, eds., Wahrheit und Begrndung, 2744. Wrzburg: Knigshausen & Neumann. Geertz, Clifford. 1973. Interpretation of Cultures. New York: Basic Books. Gentz, Joachim. 2001. Das Gongyang zhuan: Auslegung und Kanonisierung der Frhlings- und Herbstannalen (Chunqiu). Wiesbaden: Harrassowitz. . 2006a. Zur Deutung frher Grabbefunde: Das Renzi pian aus Shuihudi. In Michael Friedrich, ed., Han-Zeit: Festschrift fr Hans Stumpfeldt aus Anla seines 65. Geburtstages, 535553. Wiesbaden: Harrassowitz. . 2006b. Zwischen den Argumenten lessen: Doppelt gerichtete Parallelismen zwischen Argumenten als zentrale Thesen in frhen chinesischen Texten. In Wolfgang Behr and Joachim Gentz, eds., Komposition und KonnotationFiguren der Kunstprosa im Alten China, 3556. Bochumer Jahrbuch zur Ostasiatischenforschung 29. Munich: Iudicium. . 2007a. Die Architektur des Zhu Xi Kommentars: Eine Textstudie zum ersten Teil des Daxue. Oriens Extremus 46:231245. . 2007b. Zum Parallelismus in der chinesischen Literatur. In Andreas Wagner, ed., Parallelismus Membrorum, 241269. Orbis Biblicus et Orientalis, vol. 224. Gttingen: Vandenhoeck & Ruprecht; Fribourg: Academic Press. Gentz, Joachim, and Dirk Meyer, eds. Forthcoming. Literary Forms of Argument in Pre-modern China. Seattle: University of Washington Press. Gerhard, Volker and Norbert Herold, eds. 1985. Wahrheit und Begrndung. Wrzburg: Knigshausen & Neumann. Gerhardsson, Birger. 1961. Memory and Manuscript: Oral Tradition and Written Transmission in Rabbinic Judaism and Early Christianity. Uppsala: Almqvist & Wissells. Giele, Enno. 19981999. Early Chinese Manuscripts: Including Addenda and Corrigenda to New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts. Early China 2324:247337.

bibliography

369

. 2001. Database of Early Chinese Manuscripts. Web site of the Society for the Study of Early China. http://lucian.uchicago.edu/blogs/earlychina/research-andresources/databases/> (accessed June 9, 2009). Goggin, Joyce, and Sonja Neef, eds. 2001. Travelling Concepts. Vol. 1, Text, Subjectivity, Hybridity. Amsterdam: ASCA Press. , eds. 2002. Travelling Concepts. Vol. 2, Meaning, Frame and Metaphor. Amsterdam: ASCA Press. Goldin, Paul R. 2000. Xunzi in the Light of the Guodian Manuscripts. Early China 25:113146. . 2003. The Old Chinese Particles yan and an . Journal of the American Oriental Society 123.1:16973. . 2007. A Bibliography of Materials Pertaining to the Kuo-tien and Shanghai Museum Manuscripts. Web site of the Society for the Study of Early China. http:// paulrgoldin.com/db3/00258/paulrgoldin.com/_download/Kuo-tienBibliography. pdf. (accessed August 18, 2007). Goody, Jack. 1977. The Domestication of the Savage Mind. Cambridge: Cambridge University Press. . 1986. The Logic of Writing and the Organization of Society. Cambridge: Cambridge University Press. . 1987. The Interface between the Written and the Oral. Cambridge: Cambridge University Press. . 2000. The Power of the Written Tradition. Washington: Smithsonian Institution. Goody, Jack, and Ian Watt. 1968. The Consequences of Literacy. In Jack Goody, ed., Literacy in Traditional Societies, 304345. Cambridge: Cambridge University Press. Graham, Angus C. 1967. The Background of the Mencian Theory of Human Nature. Tsing hua Journal of Chinese Studies 6:214271. Reprinted in A. C. Graham, Studies in Chinese Philosophy and Philosophical Literature (Singapore: Institute of East Asian Philosophies, 1986), 766. . 1972. The Classical Chinese Topic-Marker fu . Bulletin of the School of Oriental and African Studies 35:85110. . 1973. The Terminations of the Archaic Chinese Pronouns. Bulletin of the School of Oriental and African Studies 36:293298. . 1977. The Chinese Particle Tzeng . Early China 3:3135. . 1978a. Later Mohist Logic, Ethics, and Science. Hong Kong: Chinese University Press. . 1978b. A Post-verbal Aspectual Particle in Classical Chinese: The Supposed Preposition Hu . Bulletin of the School of Oriental and African Studies 41.2:314 342. . 1980. Structure and License in Chinese Regulated Verse. Journal of Chinese Linguistics (Berkeley, CA) 8.1:4258. . 1981. Chuang-tzu: Textual Notes to a Partial Translation. London: School of Oriental and African Studies, University of London. . 1983. Yun and Yueh as Verbs and Particles. Acta Orientalia 44:3371. . 1986. Studies of Chinese Philosophy and Philosophical Literature. Singapore: Institute of East Asian Philosophies. . 1989a. Conceptual Schemes and Linguistic Relativism in Relation to Chinese. Synthesis Philosophica (Zagreb, Yugoslavia) 4:713732. . 1989b. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle, IL: Open Court. . 1991. Two Notes on the Translation of Taoist Classics. In Roger T. Ames, Sin-wai Chan, and Mau-sang, Ng, eds., Interpreting Culture through Translation: A Festschrift for D. C. Lau, 119144. Hong Kong: Chinese University Press. Granet, Marcel. 1934. La pense chinoise. Paris: Renaissance du Livre.

370

bibliography

Groningen, B. A. van. 1963. EKDOSIS. Mnemosyne 4.16:117. G Jigng . 1932. Cng L sh chnqi tuc Laoz zh chng sh nindi . Ynjing Dxu shxu xubo 4. Reprin ted in Gu sh bin 4:462520. G Sh . 1987. Hnsh Ywn zh jiang sh . Edited by Bn G . Shanghai: Shnghai guj. G Shkao . See Cook, Scott. Guan Xhu . 2002. Zhnggu gudi biaodian fho fzhan sh . Chngd: B shu shsh. Guanz jio zh . 2004. Compiled by L Xingfng and Ling Ynhu . In Xn bin zhz j chng . Beijing: Zhnghu. Gu Qyong . 2000. Gudin Chu jian Xng z mng ch de xn sh gun . nhu Dxu xubo (zhxu shhu kxu ban) () 5:4858. Reprinted in Png P , ed., Gu m xn zh: mn d Gudin zhjian . Chtu sxiang wnw yu wnxin cngsh 10 (Taibei: Tiwn guj, 2002), 139148. . 2001. Chtu jianb yu jngxu qunsh de fnsh wnt . Fjin lntnRnwn shhu kxu ban 5:2230. Gu Y . 1998. Gudin Chu jian Tin jing d chng (Chng zh wn zh) pin shzhng (). Kongz ynji 3:6168. Also published in Gudin Chu jian yu xin Qn xush sxiang (Shanghai: Shnghai jioy, 2001), 208229. . 2004. Xng qng ln, Xng z mng ch du jio ou d . Zhxu zi xin . http://www.philosophyol.com/pol04/ Article/chinese/c_ancient/200407/536.html (accessed February 12, 2007). Haar, Barend J. ter. 2006. Telling Stories: Witchcraft and Scapegoating in Chinese History. Leiden: Brill. . 2007. Literacy, Writing and Education in Chinese Culture: Bibliography by Barend ter Haar. http://website.leidenuniv.nl/~haarbjter/literacy.htm (accessed September 20, 2007). . 2009. Het Hemels Mandaat: De Geschiedenis van het Chinese Keizerrijk. Amsterdam: Amsterdam University Press. Halverson, John. 1992. Goody and the Implosion of the Literary Thesis. Man 27.2:301317. Hn sh wi zhun j sh . Compiled by Xu Wiy . Beijing: Zhnghu, 1980. Hansen, Chad. 1995. Qing (Emotions) in Pre-Buddhist Chinese Thought. In Marks and Ames 1995, 181211. Harbsmeier, Christoph. 1981. Aspects of Classical Chinese Syntax. Scandinavian Institute of Asian Studies. London: Curzon Press. . 1998. Language and Logic. Science and Civilisation in China, vol. 7. Cambridge: Cambridge University Press. . 1999. Rhetoric des klassischen Chinesisch (review article). Toung Pao 85:114 126. . 2004. The Semantics of Qng in Pre-Buddhist Chinese. In Halvor Eifring, ed., Love and Emotions in Traditional Chinese Literature, 69148. Sinica Leidensia, vol. 63. Leiden: Brill. . Forthcoming. Some Philosophical Notes on the Gudin Manuscript Yucng 1. In Pre-Han and Han Traditional Chinese Texts: A Collection of Conference PapersXin Qn Liang Hn gu j guj xush yntaohu lnwnj . Harris, William V. 1989. Ancient Literacy. Cambridge, MA: Harvard University Press.

bibliography

371

Hartung, Liselotte. 1987. Zur Syntax des Parallelismus in der ostjakischen Prosadichtung. In E. Lang and G. Sauer, eds., Parallelismus und Etymologie, 55110. Linguistische Studien 161.1. Berlin: Akademie der Wissenschaften der DDR, Zentralinstitut fr Sprachwissenschaft. Havelock, Eric. 1963. Preface to Plato. Cambridge, MA: Belknap Press of Harvard University Press. H Jn . 1999. Qn chng hu lng kao . Wn B (Shanx ), 4150. H Lny . 1998. Zhngu guwn zdian: Zhngu wnz shngx . Beijing: Zhnghu. . 2001. Gudin Chu jian xuan sh . In jianb ynji 2001 . Guln: Guangx shfn Dxu. . 2002. Gudin jian gu wn r kao . In Gu j zhngl ynji xu kn , 13. Hbi shng wnw ynji suo . 1995. Cu mZhnggu Zhngshn gu guwng zh m . Beijing: Wnw. Henderson, John B. 1984. The Development and Decline of Chinese Cosmology. New York: Columbia University Press. . 1991. Scripture, Canon, and Commentary: A Comparison of Confucian and Western Exegesis. Princeton: Princeton University Press. . 1998. The Construction of Orthodoxy and Heresy: Neo-Confucian, Islamic, Jewish, and Early Christian Patterns. Albany: State University of New York Press. Henricks, Robert, G. 2000. Lao Tzus Tao Te Ching: A Translation of the Startling New Documents Found at Guodian. New York: Columbia University Press. Hightower, James. 1954. The Fu of Tao Chien. Harvard Journal of Asiatic Studies 17:169230. Hinber, Oskar von. 1989. Der Beginn der Schrift und frhe Schriftlichkeit in Indien. Akademie der Wissenschaften und der Literatur, Abhandlungen der Geistes- und Sozialwissen- schaftlichen Klasse, Jahrgang 1989, no. 11. Mainz: Akademie der Wissenschaften und der Literatur. Reprint, Stuttgart: Franz Steiner, 1990. Hirschberger, Johannes. 1980. Geschichte der Philosophie. 12th ed. 2 vols. Frechen: Komet Lizenzausgabe; Freiburg im Breisgau: Herder. Holloway, Kenneth. 2002. The Recently Discovered Confucian Classic The Five Aspects of Conduct (China). Ph.D. diss., University of Michigan. Houston, Stephen D., ed. 2004. The First Writing: Script Invention as History and Process. Cambridge: Cambridge University Press. H Pngshng . 1988. Fyang Hn jian Shjng jianc xngzh j shxi gsh zh lc . In H Pngshng and Hn Zqing 1988, 9097. . 1998. Fyang Hn jian Zhu y gi sh . Jianb ynji 3:255266. . 2000. Jiand zhd xn tn . Wnw 3.526:6674. H Pngshng and Hn Zqing , eds. 1988. Fyang Hn jian Shjng ynji . Shanghai: Shnghai guj. Hung Dkung and X Zigu . 1998. Gudin Chu jian wnz kaosh . In Jln Dxu guj zhngl ynji suo jin suo shwu zhu nin jnin wnj , 98111. Changchun: Jln Dxu. Hung Ljun . 2004. Zhngu du shng z ynji . In Xi Wiyng and Zh Yunqng 2004, 174184. Hung Rnr 1999. Gudin Chu jian Lu Mgng wn Z S kao sh . In Bin j wiyuan hu , ed., Zhng Y rn xinshng q zh shu qng lnwn j , 1:397 404. Taibei: Xushng.

372

bibliography

Hung Zho . 2000. Zhjian Laoz yng wi Jxa doji chunbn de gao chaobn . Zhngzhu xukn, 6774. Hung Zhnyn and Hung Wi . 2003. Gudin Chu jian yn Sh kao . Nnyng Shfan Xuyun xubo 2:8182, 85. Hbi shng Jngmn sh bwguan . 1997. Jngmn Gudin y ho Chu m . Wnw 7:3548. , ed. 1998. Gudin Chu m zhjian . Beijing: Wnw. Hbi shng Jngsh til kaogu du , ed. 1991. Boshn Chu m . Beijing: Wnw. Hbi shng Jngzhu dq bwguan , ed. 1984. Jinglng Yutishn Chu m . Beijing: Wnw. Hbi shng Ychng dq bwguan , ed. 1992. Dngyng Zhojih Chu m . Beijing: Wnw. Hucker, Charles O. 1995. A Dictionary of Official Titles in Imperial China. Taibei: SMC Publishing. Humboldt, Wilhelm von. 19031936. Gesammelte Schriften. Edited by Kniglich preussische Akademie der Wissenschaften (ed.). 17 vols. Berlin: Behr. Humphreys, Sally C. 1983. Die Dynamik des griechischen Durchbruchs: Der Dialog zwischen Philosophie und Religion. In Shermuel Eisenstadt, ed., Die Kulturen der Achsenzeit, 128160. Frankfurt a. M: Suhrkamp. Hnn shng bwguan and Zhnggu kxu yun kaogu ynji suo , eds. 1974. Chngsh Mawngdu r, sn ho Hn m fju jian bo . Wnw 7:3948. , eds. 1975. Mawngdu r, sn ho Hn m fju zhuyo shuhu . Kaogu 1.136:4757, 61. Ikeda Tomohisa . 1993. Batai Kanbo hakusho Gogyhen kenky . Tokyo: Kuko shoin. . 1999. Kakuten sabo chikukan Gogy yakuji . Kakuten sokan no shisshi teki kenky. Vol. 1: 1851. Tokyo: Tky Daigaku Kakuten sokan kenkykai hen . . 2000a. Gudin Chu jian Wu xng ynji . Zhnggu zhxu 21:92133. , ed. 2000b. Kakuten sokan no shisshi teki kenky . Vol. 3. Tokyo: Tky Daigaku Kakuten sokan kenkykai hen . . 2000b. Matai Kanbo hakusho gogyhen kenky . Tokyo: Kyko shoin. , ed. 2003. Kakuten sokan juky kenky . Tokyo: Kyko shoin . Illich, Ivan. 1991. Im Weinberg des Texts: Als das Schriftbild der Moderne enstand; Ein Kommentar zu Hugos Didascalicon. Frankfurt/Main: Luchterhand Literaturverlag. Jahandarie, Khosrow. 1999. Spoken and Written Discourse: A Multi-disciplinary Perspective. Stamford, CT: Ablex. J Xshng . 2001. Gudin, Shng b jian wu t: Shn, h h, shn r y, qing you c, wan qi : . Zhnggu wnz 27:113120. , ed. 2004. Shnghai Bwguan cng Zhngu Chu zhsh (y) dbn. By Chn Lnqng , Zhng Yshn , Zu Jnzh . Taibei: Wnjunlu. Jing Gunghu . 1999. Gudin Chu jian yu Z Szjin tn Gudin Chu jian de sxiang sh yy . Zhnggu zhxu 20:8192. Originally published in Zhxu ynji, July 1998.

bibliography

373

. 2001. The Gudin Chu Slips and Early Confucianism. Contemporary Chinese Thought 32.2:638. Johnson, David. 1980a. The Wu Tzu-hss Pien-wen and Its Sources: Part I. Harvard Journal of Asiatic Studies 40.1:93156. . 1980b. The Wu Tzu-hss Pien-wen and Its Sources: Part II. Harvard Journal of Asiatic Studies 40.2:465505. . 1981. Epic and History in Early China: The Matter of Wu Tzu-hs. Journal of Asian Studies 40.2:255271. Johnson, William A. 2003. Reading Cultures and Education. In Peter C. Patrikis, ed., Reading Between the Lines: Perspectives on Foreign Language and Literacy, 923. New Haven: Yale University Press. . 2009. Introduction to Johnson and Parker 2009, 310. Johnson, William A. and Holt N. Parker, eds. 2009. Ancient Literacies: The Culture of Reading in Greece and Rome. Oxford: Oxford University Press. Jullien, Franois. 1989. Procs ou cration: Une introduction la pense des lettrs chinois. Paris: ditions du Seuil. . 1991. loge de la fadeur: partir de la pense et de lesthtique de la Chine. Paris: P. Picquier. . 1992. La propension des choses: Pour une histoire de lefficacit en Chine. Paris: ditions du Seuil. . 1995. Le dtour et laccs: Stratgies du sens en Chine. Paris: Grasset. Kalinowski, Marc M. 20002001. Systmes de croyances et de pense de monde sinis. Annuaire EPHE, Section des sciences religieuses, 109:141143. Kaogu xubo . 1981. nyng Xiaotun cn bi de liang zu Y n di m , 497508. Karlgren, Bernhard. 1946. Legend and Cults in Ancient China. Bulletin of the Museum of Far Eastern Antiquities 18:199365. . 1950. The Book of Odes: Chinese Text, Transcription and Translation. Stockholm: Museum of Far Eastern Antiquities. . 1968. Some Sacrifices in Chou China. Bulletin of the Museum of Far Eastern Antiquities 40:132. Keightley, David N. 1985. Sources of Shang History: The Oracle-Bone Inscriptions of Bronze Age China. Berkeley: University of California Press. Kern, Martin. 1994. Zum Topos Zimtbaum in der chinesischen Literatur: Rhetorische Funktion und poetischer Eigenwert des Naturbildes kuei. Stuttgart: Steiner. . 2000. The Stele Inscriptions of Chin Shih-huang: Text and Ritual in Early Chinese Representation. New Haven: American Oriental Society. . 2001. Ritual, Text, and the Formation of the Canon: Historical Transitions of Wen in Early Modern China. Toung Pao 87.13:4391. . 2002. Methodological Reflections on the Analysis of Textual Variants and the Modes of Manuscript Production in Early China. Journal of East Asian Archaeology 4.14:143181. . 2003. Early Chinese Poetics in the Light of Recently Excavated Manuscripts. In Olga Lomov, ed., Recarving the Dragon: Understanding Chinese Poetics, 2772. Prague: Charles University in Prague, Carolinum Press. . 2005a. The Odes in Excavated Manuscripts. In Kern 2005c, 149193. . 2005b. Quotation and the Confucian Canon in Early Chinese Manuscripts: The Case of Zi yi (Black Robes). Asiatische Studien/tudes asiatiques 59.1:293332. , ed. 2005c. Text and Ritual in Early China. Seattle: University of Washington Press. . 2007a. Excavated Manuscripts and Their Socratic Pleasures: Newly Discovered Challenges in Reading the Airs of the States . Asiatische Studien/tudes asiatiques 61.3:775793.

374

bibliography

. 2007b. The Performance of Writing in Western Zhou China. In Sergio La Porta and David Shulman, eds., The Poetics of Grammar and the Metaphysics of Sound and Sign, 109176. Leiden: Brill. . 2009. Bronze Inscriptions, the Shijing and the Shangshu: The Evolution of the Ancestral Sacrifice during the Western Zhou. In John Lagerwey and Marc Kalinowski, eds., Early Chinese Religion, pt. 1, Shang through Han (1250 BC220 AD), 143200. Leiden: Brill. Kirk, Geoffrey S., and John E. Raven. 1957. The Presocratic Philosophers: A Critical History with a Selection of Texts. Cambridge: Cambridge University Press. Knoblock, John. 19881994. Xunzi: A Translation and Study of the Complete Works. 3 vols. Stanford: Stanford University Press. Kohn, Livia, and Michael LaFargue, eds. 1998. Lao-tzu and the Tao-te-ching. Albany: State University of New York Press. Korpel, Marjo. 2000. Introduction to the Series Pericope. In Korpel and Oesch 2000, 150. Korpel, Marjo, and Josef Oesch, eds. 2000. Delimitation Criticism. Pericope 1, Scripture as Written and Read in Antiquity. Assen: Royal Van Gorcum. Krispijn, Theo. 1992. The Early Mesopotamian Lexical Lists and the Dawn of Linguistics. JEOL 32: 1222. Kristeva, Julia. 1980. Word, Dialogue, and Novel. In Leon S. Roudiez, ed., Desire and Language: A Semiotic Approach to Literature and Art, translated by Thomas Gora Alice Jardine and Leon S. Roudiez , 6491. New York: Columbia University Press. LaFargue, Michael. 1992. The Tao of the Tao Te Ching: A Translation and Commentary. Albany: State University of New York Press. . 1994. Tao and Method: A Reasoned Approach to the Tao Te Ching. Albany: State University of New York Press. . 1998. Recovering the Tao-te-chings Original Meaning: Some Remarks on Historical Hermeneutics. In Kohn and LaFargue 1998, 255275. Laoz jio sh . 1984. Compiled by Zh Qinzh . In Xn bin zhz j chng . Beijing: Zhnghu. Lau, D. C. 1963. Lao Tzu: Tao Te Ching. Harmondsworth: Penguin Books. . 1993. Meng tzu. In Michael Loewe, ed., Early Chinese Texts: A Bibliographical Guide, 331335. Early China Special Monograph Series, no. 2. Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California. Lawton, Thomas, ed. 1991 New Perspectives on Chu Culture during the Eastern Zhou Period. Washington, DC: Smithsonian Institution. Leclercq, Jean. 1961. The Love of Learning and the Desire for God. Translated by C. Misrahi. New York: Fordham University Press. Ledderose, Lothar. 2000. Ten Thousand Things: Module and Mass Production in Chinese Art. Bollingen Series 46. Princeton: Princeton University Press. Lewis, Mark Edward. 1990. Sanctioned Violence in Early China. Albany: State University of New York Press. . 1999. Writing and Authority in Early China. Albany: State University of New York Press. . 2007. The Early Chinese Empires: Qin and Han. Cambridge, MA: Belknap Press of Harvard University Press. Li Boqian. 2000. A Brief Account of the Origins and Development of Chu Culture. In Allan and Crispin 2000, 922. L Chngl . 1999a. Kakuten Sobo chikukan Ro Bokuk mon Shishi yakuch . Kakuten Sokan no shisshiteki kenky 1:1018.

bibliography

375

. 1999b. Kakuten Sokan Ro Bokuk mon Shishi no chshinkan ni tsuite . Kakuten Sokan no shisshiteki kenky 1, 13650. L Cnshn . 2002. . In Png P , ed., , 287292. Taibei: Tiwn guj. L j zh sh . 1997. Edited by Ruan Yun . In Shsn jng zh sh , vol. 5. Reprint, Taibei: Ywn. L Jiho . 1999. D Gudin Chu m zhjian suo y . In Gudin Chu jian ynji . Shenyang: Lionng jioy. L Lng . 1999. Gudin Chu jian jiod j , Doji wnhu ynji 17:455542. . 2000. Cnji xn ch jianb guj xush taolnhu de j dian ganxiang (http://www.jianbo.org/Wssf/liling3-01.htm) (accessed July 23, 2007). . 2002. Gudin Chu jian jiod j . Beijing: Bijng Dxu. L Mil . 1992. Mawngdu Hn m ynji ml . Changsha: Hnn. L Ru . 2005. Kong-Mng zhjin xng ln ynji: Y Gudin, Shngbjian wi jchu . Qnghu Dxu zhunmn sh bsh lnwn . L Tinhng . 2000a. Cng Xng z mng ch tn Kongz yu Sh, Sh, L, Yu . Zhnggu zhxu sh 4:4143. . 2000b. Xng z mng ch de binlin j fnpin. Qnghu jianb ynji 1:196199. . 2003. Gudin zhjian Xng z mng ch ynji . Wuhan: Hbi jioy. L Wiwu . 2000. Xng z mng ch de xu yyn chtn . In Gudin Chu jian guj xush yntao hu , 310313. Wuhan: Hbi rnmn. L Xuqn . 1987. Du gu sh de fans . In Zhnggu chuntng wnhu de zi guj . Shanghai: Rnmn. Reprinted in L Xuqn , Dngdi xuzhe zxun wnj: L Xuqn juan (Hefei: nhu jioy, 1999), 1220. . 1993. Zou ch Y gu shdi . Zhnggu wnhu 7. Reprinted in Zouch Y gu shdi , 2d rev. ed. (Shnyng: Lionng Dxu, 1997), 119. . 1995. Tn Xn gu, Y gu, Sh gu . Chuntong wnhu y xindi wnhu . Reprinted in Zouch Y gu shdi 2d rev. ed. (Shnyng: Lionng Dxu, 1997), 341349. . 1998a. Jngmn Gudin Chu jian suo jin Gun Yn y shu . Zhnggu wnw bo , 3. . 1998b. Jngmn Gudin Chu jian zhng de Z Sz Wnw Tind , March 1998. Reprinted Zhnggu zhxu 20 . (1999):7580. . 1999a. Gudin Chu jian h R ji jngj . Zhnggu zhxu 20:1821. . 1999b. Gudin jian yu Yu j . In Zhnggu zhxu de qunsh h fzhanZhng Dinin xinshng 90 shu qng jnin wnj, 2328. Beijing: Bijng Dxu. . 1999c. Ti y shng shu de shsh jish. Doji wnhu ynji 17:279300. . 1999d. Xin Qn rji zhzu de zhngd fxin . Zhnggu zhxu 20:1317.

376

bibliography

. 2000a. The Confucian Texts from Guodian Tomb Number One: Their Date and Significance. In Allan and Crispin 2000, 107111. . 2000b. Shshu Gudin jian Chng zh wn zh liang zhng . Yn Ti Dxu xubo (zhxu shhu kxu ban) () 13.4 (51):457460. . 2000c. Xn fxin jianb yj du xushsh de yngxiang . Doji wnhu ynji 18:19. . 2002. Ln Gudin jian Laoz fi Laoz bn mo . In J nin Ln Jinmng jioshu sh xu lnwn j . Zhnggu shhu kxu, 16. . 2005. Zhnggu shhu kxu yun xush weiyuan hu wnkL Xuqn wnj , 444446. Shanghai: Shnghai csh chbansh. Ling To . 2003. Zhjian Qing d y sh yu zaoq Rji tin-rn gun . Zhxu ynji 4:6570. Lio Mngchn . 1998a. Cng Jngmn Chu jian ln xin-Qn Rji yu Zhuy de gunx . Guj Yxu ynji 4: 309322. . 1998b. Gudin Chu jian Rji zhzu kao . Kongz ynji 3:6983. . 1999. Jngmn Gudin Chu jian yu xin-Qn R xu . Zhnggu zhxu 20:3674. . 2000a. Gudin Chu jian Xng z mng ch de binlin yu fnh wnt . Zhnggu zhxu sh 4:1421. . 2000b. Gudin Chu jian Xng z mng ch pin jio sh . Qnghu jianb ynji 1:2867. . 2001. Xn ch Chu jian shln . Chtu sxiang wnw yu wnxin ynji cngsh 3. Taibei: Tiwn guj. Lichtenberg, Georg Christoph. 1967. Sudelbcher: Schriften und Briefe. Munich: Hanser. Ln Hngyng , Murase Nozomu , and Furuya Akihira . 2004. Sengoku moji X ni tsuite . Ki-pin 23. Ln Lzhn . 1980. Chngjin R xu sxiang tx de Xnz . Zhnghu wnhu fxng yukn 13.11:6274. Ln Sqng and Zhu Fngwu . 1999. Gudin zhjian binx fyun ynji . Guwnz yu gu wnxin 10. Li Bnhu . 1991. Cng Boshn Chu jian jsh cilio ln j Chu gu jnin yu Chu l . In Hbi shng Jngsh til kaogu du , ed., Boshn Chu m , 533547. Beijing: Wnw. Li Lxin . 2000. Gudin Chu jian z kao (wu z) (). Gu wnz ynji 22 (July) : 205209. Li Qy . 1997. Shngsh yunli j chunbn . Shenyang: Lionng Dxu. Li Xn . 1991. Xjng zj . Collated by G Hng , ed. and ann. by Xing Xnyng and Li Krn . Shanghai: Shnghai guj. Li Xnfng . 1999. Jngmn Gudin zhjian Laoz jigu . Taibei: Ywn. . 2000. Jianb Wu xng jigu . Taibei: Ywn. . 2002. Kongz sh ln shxu . Hefei: nhu Dxu. Li Xnln . 2000. Gudin Chu jian Xng z mng ch pin jinsh . In Wuhn Dxu Zhnggu wnhu ynjiyun, ed., Gudin Chu jian guj xush yntao hu lnwnj , 330354. Rnwn Lncng t j . Wuhan: Hbi rnmn.

bibliography

377

Li Zho . 2000. D Gudin Chu jian z c zh j . In Wuhn Dxu Zhnggu wnhu ynji yun, ed., Gudin Chu jian guj xush yntao hu lnwn j , 7593. Rnwn Lncng t j . Wuhan: Hbi rnmn. . 2005. Gudin Chu jian jio sh . Fzhu: Fjin rnmn. Li Zuxn . 2004. Ti y shng shu qian sh . In A Ln and Xng Wn , eds., Xn ch jianb ynji: Xn ch jianb ynji guj xush yntao hu wn j 2000 nin 6 yu, Bijng : 2000 6 , 254256. Beijing: Wnw. Lloyd, G. E. R. 1990. Demystifying Mentalities. Cambridge: Cambridge University Press. Loewe, Michael. 1999. The Heritage Left to the Empires. In Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., 9671033. Cambridge: Cambridge University Press. Lnhng. See Wng Chng . 1996. Lu Ynhun . 1999. Jngmn Gudin Chu jian b tn: Gudin Chu jian de nindi yngt h yy : . Hbi Dxu xubo (Zhxu shhu kxu ban) () 2:113. Ma Chngyun . 2001. Shnghai bwguan cng Zhnggu Chu zh sh . Shanghai: Shnghai guj. Ma Xln . 1956. Laoz jiogu . Beijing: Guj. Mair, Victor H., ed. 2001. The Columbia History of Chinese Literature. New York: Columbia University Press. Makeham, John. 1994. Name and Actuality in Early Chinese Thought. Albany: State University of New York Press. Mansvelt Beck, B. J. 1987. The First Emperors Taboo Character and the Three Day Reign of King Xiaowen: Two Moot Points Raised by the Qin Chronicle Unearthed in Shuihudi in 1975. Toung Pao 73.13:6885. Marks, Joel, and Roger T. Ames, eds. 1995. Emotions in Asian Thought: A Dialogue in Comparative Philosophy. Albany: State University of New York Press. Maspro, Henri M. 1933. Le mot Ming . Journal asiatique 223:249296. Matsumoto Masaaki . 1968. Shunj Sengoku ni okeru Shsho no tenkai . Tokyo: Kazama shob. . 1988. Matsumoto Masaaki shosaku sh . Tokyo: Ksei shorin. Mawngdu Hn m bsh zhngl xiao zu . 1974. Mawngdu Hn m chtu Laoz shwn . Wnw 11:820. . 19801985. Mawngdu Hn m bsh . 6 vols. Beijing: Wnw. Mawngdu Hn m wnw zu . 1992. Changsha: Hnn. McKenzie, Donald Francis. 1999. Bibliography and the Sociology of Texts. Cambridge: Cambridge University Press. McLuhan, Marshall. 1962. The Gutenberg Galaxy: The Making of Typographic Man. Toronto: University of Toronto Press. Mngz zhng y . 1992. Compiled by Jio Xn . Taibei: Shji. Meyer, Dirk. 2006. A Device for Conveying Meaning: The Structure of the Gudin Tomb One Manuscript Zhng xn zh do . In Wolfgang Behr and Joachim Gentz, eds., Komposition und KonnotationFiguren der Kunstprosa im Alten China, 5778. Bochumer Jahrbuch zur Ostasiatischenforschung 29. Munich: Iudicium. . 2007. Structure as a Means of Persuasion as Seen in the Manuscript Qing d y sh from Tomb One, Gudin. Oriens Extremus 45:179210.

378

bibliography

. 2008a. Meaning-Construction in Warring States Philosophical Discourse: A Discussion of the Palaeographic Materials from Tomb Gudin One. Leiden: CNWS, 2008. . 2008b. Writing Meaning: Strategies of Meaning-Construction in Early Chinese Philosophical Discourse. Monumenta Serica 56:5595. . 2009. Texts, Textual Communities, and Meaning: The Genius Loci of the Warring States Chu Tomb Gudin One. Asiatische Studien/tudes asiatiques 63.4:827856. Mz. See W Yjing 1993. Naddaf, Gerard. 2005. The Greek Concept of Nature. Albany: State University of New York Press. Needham, Joseph. 1956. Science and Civilization in China. Vol. 2, History of Scientific Thought. Cambridge: Cambridge University Press. Neininger, Ulrich. 1983. Burying the Scholars Alive: On the Origin of a Confucian Martyrs Legend. In Wolfram Eberhard et al., eds., East Asian Civilizations: New Attempts at Understanding Traditions, vol. 2, Nations and Mythology, 121136. Munich: Simon & Magiera. Nienhauser, William H., Jr. 2003. A Note on a Textual Problem in the Shih Chi and Some Speculations concerning the Compilation of the Hereditary Houses. Toung Pao 89:3958. Nivison, David S. 1996. The Ways of Confucianism: Investigations in Chinese Philosophy. La Salle, IL: Open Court. Noegel, Scott B. 1996. Janus Parallelism in the Book of Job. Sheffield: Sheffield Academic Press. Norman, Jerry. 1994. Pharyngealization in Early Chinese. Journal of the American Oriental Society 114.3:397408. Nylan, Michael. 1999. A Problematic Model: The Han Orthodoxy Synthesis, Then and Now. In Kai-wing Chow, On-cho Ng, and John B. Henderson, eds., Imagining Boundaries: Changing Confucian Doctrines, Texts, and Hermeneutics, 1756. Albany: State University of New York Press, 1756. . 2001a. The Five Confucian Classics. New Haven: Yale University Press. . 2001b. On the Politics of Pleasure. Asia Major, 3d ser., 14.1:73124. . 2005. Toward an Archaeology of Writing: Text, Ritual, and the Culture of Public Display in the Classical Period (475 B.C.E.220 C.E.). In Kern 2005c, 349. . 2009. Classics without Canonization: Learning and Authority in Qin (221210 BC) and Han (206BCAD 220). In John Lagerwey and Marc Kalinowski, eds., Early Chinese Religion, pt. 1, Shang through Han (1250 BC220 AD), 721776. Leiden: Brill. Olson, David R. 1980. Some Social Aspects of Meaning in Oral and Written Language. In David R. Olson, ed., The Social Foundations of Language and Thought: Essays in Honor of Jerome S. Bruner, 90108. New York: W. W. Norton. . 1994. The World on Paper. Cambridge: Cambridge University Press. Ong, Walter J. 1976. The Presence of the Word: Some Prolegomena for Cultural and Religious History. New Haven: Yale University Press. . 1982. Orality and Literacy: The Technologizing of the Word. London: Methuen. Oppenheimer, A. Leo. 1977. Ancient Mesopotamia: Portrait of a Dead Civilization. Rev. ed. Chicago: University of Chicago Press. Otto, Walter F. 1947. Die Gtter Griechenlands: Das Bild des Gttlichen im Spiegel des griechischen Geistes. Frankfurt: G. Schulte-Bulmke. Oyng Zhnrn . 2002. Chu jian Tng Y zh do shn r b chun de lln zhuzng . In Dng Sxn , ed., Chu d chtu jianb wnxin sxiang ynji , 1:91102. Wuhan: Hbi jioy.

bibliography

379

Png P . 1977. Mawngdu b sh jiki le S-Mng wu xng shu zh mb sh Laoz jia bn juan hu gu y sh zh y de chb ynji . Wnw 10:6369. . 1980. Bsh Wu xng pin ynji . Shandong: Q Lu shsh. . 1998. Kong-Mng zh jinGudin Chu jian de sxiang sh dwi . Zhnggu shhu kxu (May): 8895. Reprinted as Kong-Mng zh jinGudin Chu jian zhng de Rji xn xng shu , Zhnggu zhxu 20 (1999): 2235. . 2000. Zh b Wu xng pin jiozh j ynji . Taibei: Wnjun lu. Pankenier, David W. 1990. The Scholars Frustration Reconsidered: Melancholia or Credo? Journal of the American Oriental Society 110.3:434459. Png Bngbn . 2004. . http://www.jianbo.org/Zzwk/ Jblt/Wuhanhui/tangbangben.htm (accessed February 20, 2006). Png Ho . 1999a. Gudin y ho m de nindi j xinggun de wnt . In Chn Fbn , ed., Bn shj chtu sxiang wnxin yu Zhnggu gudian zhxu ynji lnwnj , 2:357364. Taibei: Furn Dxu. . 1999b. Gudin y ho m de nindi yu jianbn Laoz de jigu . Doji wnhu ynji 17:1321. . 1999c. Wngshn, Boshn, Gudin Chu m de fju yu Chu wnhu . Transactions of the International Conference of Eastern Studies/Kokusai th gakusha kaigi kiy 44:2331. . 2000. Post-excavation Work on the Guodian Bamboo-Slip Laozi: A Few Points of Explanation. In Allan and Crispin 2000, 3338. Petersen, Jens stergrd. 1995. Which Books Did the First Emperor of Chin Burn? On the Meaning of Pai Chia in Early Chinese Sources. Monumenta Serica 43:152. Peterson, J. Willard. 1986. The Development and Decline of Chinese Cosmology by John B. Henderson: A Review Article. Harvard Journal of Asiatic Studies 46.2:657 674. . 1988. Squares and Circles: Mapping the History of Chinese Thought. Journal of the History of Ideas 49.1:4760. Pin Yuqin and Dun Shn , eds. 2003. Bn shjj yl i chtu jianb gish . Taibei: Wnjunlu. First published in 1999. Pines, Yuri. 2005. Disputers of Abdication: Zhanguo Egalitarianism and the Sovereigns Power. Toung Pao 91.45:243300. Poo Mu-chou. 1998. In Search of Personal Welfare: A View of Ancient Chinese Religion. Albany: State University of New York Press. P Mozuo . 2001. Guny Shnghai Zhngu zhjian zhng Kongz de rndng: Ln Kongz sh ln hwn sh Kongz r fi Buz Z Shng . Zhnghu wnsh lncng 67:1134. Puett, Michael. 2004. The Ethics of Responding Properly: The Notion of Qing in Early Chinese Thought. In Halvor Eifring, ed., Love and Emotions in Traditional Chinese Literature, 3768. Sinica Leidensia, vol. 63. Leiden: Brill. Pulleyblank, Edwin G. 1996. Prosody or Pharyngealization in Old Chinese? The Origin of the Distinction between Type A and Type B Syllables. Journal of the American Oriental Society 116.1:105107. Qin M . 1951. Zhnggu wnhu sh daoln . Taibei: Zhng Zhng.

380

bibliography

. 1962. Sh j dmng kao . Hong Kong: Tapng. Qin Xn . 2000. Xng z mng ch (qin bn bfn) zhj () . Qnghu jianb ynji 1:142146. Qi Xgu . 2000a. Chinese Writing . Translated by Gilbert L. Mattos and Jerry Norman. Early China Special Monograph Series, no. 4. Berkeley: Society for the Study of Early China and the Institute of East Asian Studies, University of California. . 2000b. Ti y shng shu mngz zhng jijin ln Ti y shng shu de fn zhng wnt. Gu wnz ynji 22:218225. . 2004a. Gudin Laoz jian chtn . In Qi Xgu, Zhnggu chtu gu wnxin sh jiang . Shanghai: Fdn Dxu. . 2004b. Tntn Shngb jian Z Go pin de jian x . In Shngb guan cng Zhngu Chu zhsh ynji xbin , 111. Shanghai: Shnghai shdin. Quintilianus, Marcus Fabius. 2006. Ausbildung des Redners, Zwlf Bcher: Lateinisch und Deutsch. Translated and edited by Helmut Rahn. 2 vols. Darmstadt: Wissenschaftliche Buchgesellschaft. Reible, Wolfgang, ed. 1990. Erscheinungsformen kultureller Prozesse. Tbingen: Gunter Narr. Renfrew, Colin. 2000. Loot, Legitimacy and Ownership: The Ethical Crisis in Archaeology. Duckworth Debates in Archaeology. London: Gerald Duckworth. Richter, Matthias. 2005. Handschriftenkundliche Probleme beim Lesen altchinesischer Manuskripte. In Bernhard Fhrer, ed., Aspekte des Lesens in China in Vergangenheit und Gegenwart: Referate der Jahrestagung 2001 der Deutschen Vereinigung fr Chinastudien (DVCS)/Aspects of Reading in China, Past and Present: Collected Papers of the 2001 Annual Meeting of the German Association for Chinese Studies (DVCS), 88121. Bochum: Projekt. Forthcoming. Punctuation. In Wolfgang Behr, Martin Kern, and Dirk Meyer, eds., Reading Early Chinese Manuscripts: Texts, Contexts, Methods. Leiden: Brill. Riegel, Jeffrey. 1993. Li chi . In Michael Loewe, ed., Early Chinese Texts: A Bibliographical Guide, 293297. Early China Special Monograph Series, no. 2. Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California. . 1997. Eros, Introversion, and the Beginnings of Shijing Commentary. Harvard Journal of Asiatic Studies 57.1:143177. . 2001. Chinese Literature. In Mair 2001, 97110. Robinet, Isabelle. 1998. Later Commentaries: Textual Polysemy and Syncretistic Interpretations. In Kohn and LaFargue 1998, 119142. Rllicke, Hermann-Josef. 1995. Hidden Commentary in Pre-canonical Chinese Literature. In Aufstze. Bochumer Jahrbuch zur Ostasienforschung 19:1525. Munich: Iudicium. Roth, Harold. 2000. Some Methodological Issues in the Study of the Guodian Laozi Parallels. In Allan and Crispin 2000, 7188. Rubin, Vitaly A. 1976. Individual and State in Ancient China: Essays on Four Chinese Philosophers. Translated by Steven L. Levine. New York: Columbia University Press. Sagart, Laurent. 1999. The Roots of Old Chinese. Amsterdam and Philadelphia: John Benjamins. Sato, Masayuki. 2003. The Confucian Quest for Order: The Origin and Formation of the Political Thought of Xun Zi. Sinica Leidensia, vol. 58. Leiden: Brill. Schaberg, David. 2001. A Patterned Past: Form and Thought in Early Chinese Historiography. Harvard East Asian Monographs 205. Cambridge, MA: Harvard University Asia Center.

bibliography

381

Scheid, John, and Jesper Svenbro. 1994. Le mtier de Zeus: Mythe du tissage et du tissu dans le monde grco-romain. Paris: La Dcouverte. Schlegel, Gustave. 1896. La loi du paralllisme en style chinois: Dmontre par la preface du Si-y ki; La traduction de cette preface par feu Stanislav Julien dfendue contre la nouvelle traduction du Pre A. Gueluy. Leiden: Brill, 1896. Schmlz, Andrea. 1993. Vom Lied in der Gemeinschaft zum Liedzitat im Text: Liedzitate in den Texten der Gelehrtradition der spten Chou-Zeit. New York: HnselHohenhausen. Schssler, Axel. 1987. A Dictionary of Early Zhou Chinese. Honolulu: University of Hawaii Press. . 2007. ABC Etymological Dictionary of Old Chinese. Honolulu: University of Hawaii Press. Schwartz, Benjamin I. 1985. The World of Thought in Ancient China. Cambridge, MA: Harvard University Press. Schwermann, Christian. 2006. Collage-Technik als Kompositionsprinzip klassisch chinesischer Prosa: Der Aufbau des Kapitels Tng wn (Die Fragen des Tng) im Li z . In Wolfgang Behr and Joachim Gentz, eds., Komposition und Konnotation Figuren der Kunstprosa im Alten China, 125160. Bochumer Jahrbuch fr Ostasiatische Forschung 29. Munich: Iudicium. Seidel, Anna. 1969. La divinisation de Lao Tseu dans le taoisme des Han. Paris: cole Franaise dExtrme-Orient. Selz, Gebhard J. 1998. ber Mesopotamische Herrschaftskonzepte: Zu den Ursprngen mesopotamischer Herrschaftsideologie im 3. Jahrtausend. In M. Dietrich and O. Loretz, eds., Studien zur Altorientalistik. Festschrift fr Willem H.Ph. Rmer zur Vollendung seines 70. Lebensjahres mit Beitrgen von Freunden, Schlern, und Kollegen, 281344. Mnster: Ugarit. Seters, John van. 2006. The Edited Bible: The Curious History of the Editor in Biblical Criticism. Winona Lake, Indiana: Eisenbrauns. Shaughnessy, Edward L. 1983. The Composition of the Zhouyi. Ph.D. diss., Stanford University. . 1991. Sources of Western Zhou History: Inscribed Bronze Vessels. Berkeley and Los Angeles: University of California Press. . 1997a. Before Confucius: Studies in the Creation of the Chinese Classics. Albany: State University of New York Press. . 1997b. The Composition of Qian and Kun Hexagrams of the Zhouyi. In Shaughnessy 1997a, 197219. . 1997c. From Liturgy to Literature: The Ritual Context of the Earliest Poems in the Book of Poetry. In Shaughnessy 1997a, 165195. . 1997d. I Ching: The Classic of Changes. New York: Ballantine Books. . 1997e. New Evidence of the Zhou Conquest. In Shaughnessy 1997a, 3167. , ed. 1997f. New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts. Early China Special Monograph Series, no. 3. Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California. . 1999. Calendar and Chronology. In Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., 1929. Cambridge: Cambridge University Press. [Xi Hny ]. 2004. Shln Z y cujian zhngj j q zi L j bn Z y binzuan guchng zhng de yunyn h huguo . In Xi Wiyng and Zh Yunqng , eds., Xn chtu wnxin yu gudi wnmng ynji , 287296. Shanghai: Shnghai Dxu. . 2005. The Guodian Manuscripts and Their Place in Twentieth-Century Historiography on the Laozi. Harvard Journal of Asiatic Studies 65.2:417457.

382

bibliography

. 2006. Rewriting Early Chinese Texts. Albany: State University of New York Press. Shj. See Takikawa Kametar 1989. Shuhud Qn jian wnz bin . 1994. Edited by Zhng Shouzhng . Beijing: Wnw. Shu yun j zhng . 1961. Compiled by Zuo Sngcho . Taibei: Wn shi zh. Shuwn jiz. See Dun Yci . Sivin, Nathan. 1978. On the Word Taoist as a Source of Perplexity, with Special Reference to the Relations of Science and Religion in Traditional China. History of Religions 17.34:303330. Smith, Kidder. 2003. Sima Tan and the Invention of Daoism, Legalism, et cetera. Journal of Asian Studies 62.1:129156. Spirin, Vladimir S. 1976. Postroenie drevnekitajskich tekstov. Moscow: Nauka, Glavnaja Red. Vostocnoj Literatury. . 1991. Composition des textes chinois ancien. Extrme-Orient, ExtrmeOccident 13:3157. Stevenson, Charles. 1938. Ethical Judgements and Avoidability: Persuasive Definitions. Mind 47:331350. . 1945. Ethics and Language. London: Oxford University Press. Stock, Brian. 1983. The Implications of Literacy: Written Language and Models of Interpretation in the Eleventh and Twelfth Centuries. Princeton: Princeton University Press. Street, Brian V. 1984. Literacy in Theory and Praxis. Cambridge: Cambridge University Press. Takikawa Kametar , ed. 1989. Shiki kaich ksh . 1934; reprint, Taibei: Tingng. Tn Qxing 1991. Zhnggu lsh dt j . Taibei: Xioyun. Originally published Beijing: Zhnggu dt, 1980. Tng Rnshng . 1995. Chu x jianb wnz bin . Wuhan: Hbi jioy. Thomas, Rosalind. 1992. Literacy and Orality in Ancient Greece. Cambridge: Cambridge University Press. Toorn, Karel van der. 2007. Scribal Culture and the Making of the Hebrew Bible. Cambridge, MA: Harvard University Press. Trauzettel, Rolf. 1979. Erzhlte, besprochene und evozierte Vergangenheit: Su Tungpos Essay ber Knig Wu . In Wolfgang Bauer, ed., Studia Sino-Mongolica: Festschrift fr Herbert Franke, 329344. Wiesbaden: Steiner. Tsien, Tsuen-hsuin. 2004. Written on Bamboo and Silk: The Beginnings of Chinese Books and Inscriptions. 2d ed. With an afterword by Edward L. Shaughnessy. Chicago: University of Chicago Press. Tsuda Skichi . 1946. Rongo to Kshi no shis . Tokyo: Iwanami shoten. T Zngli . 2001. Gudin Chu jian Yu cng yi, sn bin lin wnt zi skao . Jngmn zhy jsh xu yun xu bo 1:9196. T Zngli and Li Zuxn . 2001. Gudin Chu jian xinqn rji shsh jiosh Taibei: Wnjun. Twitchett, Denis, and Michael Loewe, eds. 1986. The Cambridge History of China. Vol. 1, The Chin and Han Empires, 221 B.C.A.D. 220. Cambridge: Cambridge University Press. Unger, Ulrich. 1994. Rhetorik des klassischen Chinesisch. Wiesbaden: Harrassowitz. . 1998. Abri der Literatur des chinesischen Altertums: Prodesse aut declare? Lecture presented in Mnster.

bibliography

383

. 2000. Grundbegriffe der altchinesischen Philosophie: Ein Wrterbuch fr die Klassische Periode. Darmstadt: Wissenschaftliche Buchgesellschaft. Van De Mierop, Marc. (2004). A history of the ancient Near East, ca. 3000323 BC. Oxford: Blackwell. Vogelsang, Kai. 2002. Inscriptions and Proclamations: On the Authenticity of the gao Chapters in the Book of Documents. Bulletin of the Museum of Far Eastern Antiquities 74:138209. Wagner, Rudolf G. 1969. Die Fragen Hui-yuans an Kumrajva. Ph.D. diss., LudwigMaximilians-Universitt Mnchen. . 1980. Interlocking Parallel Style: Laozi and Wang Bi. Asiatische Studien/tudes asiatiques 34.1:1858. . 1995. Der vergessene Hinweis: Wang Pi ber den Lao-tzu. In Jan Assmann and Burkhard Gladigow, eds., Text und Kommentar, 257278. Munich: Wilhelm Fink. . 1999a. The Guodian MSS and the Units of Thought in Early Chinese Philosophy. Paper presented at the International Conference on Gudin Bamboo Strips, Wuhn. . 1999b. The Impact of Conceptions of Rhetoric and Style upon the Formation of Early Laozi Editions: Evidence from Guodian, Mawangdui, and the Wang Bi Laozi. Transactions of the International Conference of Eastern Studies 44:3256. . 2000. The Craft of a Chinese Commentator: Wang Bi on the Laozi. Albany: State University of New York Press. . 2003a. A Chinese Reading of the Daodejing: Wang Bis Commentary on the Laozi with Critical Text and Translation. Albany: State University of New York Press. . 2003b. Language, Ontology, and Political Philosophy in China: Wang Bis Scholarly Exploration of the Dark (Xuanxue). Albany: State University of New York Press. . Forthcoming. A Building Block of Chinese Philosophical Rhetoric: Initial f as a Phrase Status Marker. In Gentz and Meyer, forthcoming. Wng Baoxun . 1999. Shln Gudin Chu jian g pin de zhunzu shdi j q bijngjin ln Gudin j Boshn Chu m de shdi wnt . Zhnggu zhxu 20:366390. Wng B . 2000. Guny Gudin Chu m zh jian fn pin yu linzhu de fn dian xiangfa . Zhnggu zhxu 21:247262. . 2001. Guny Tng Y zh do de jge wnt . Zhjian sxiang . Taibei: Tiwn guj, 86. Wng Chng , ed. 1996. Ln hng jiosh . Edited by Huang Hu . 4 vols. In Xn bin zhz j chng . Beijing: Zhnghu. Wng F . 1997. Qin f ln jin jio zhng . Xn bin zhz j chng . Beijing: Zhnghu. First published in 1985. Wng Xngpng . 2004. Xng z mng ch du Kongz rnxngln de fzhan . Jinghn Dxu xubo 23.3:6769. Wng Zhpng . 2004. Gudin Chu jian Qing d y sh cngkao In A Ln and Xng Wn , eds., Xn ch jianb ynji: Xn ch jianb guj yntaohu lnwnj , 290306. Beijing: Wnw. Weber, Max. 1977. Wirtschaft und Gesellschaft: Grundriss der verstehenden Soziologie. 5th rev. ed. Tbingen: Mohr. Wi Qpng . 2000. Chu jian Laoz jin sh . Taibei: Wn junlu. Wi Yhu . 2002. Sh xi Chu jian wnz zhng de X X z XX. Jing hn kaogu 2:7477.

384

bibliography

. 2004. Sh x Chu jian wnz zhng de y yu shn z . In Xi Wiyng and Zh Yunqng 2004, 157161. West, Martin L. 1973. Textual Criticism and Editorial Technique Applicable to Greek and Latin Texts. Stuttgart: Tbner. Wilhelm, Hellmut. 1957. The Scholars Frustration: Notes on a Type of Fu. In J. K. Fairbank, ed., Chinese Thought and Institutions, 310319. Chicago: University of Chicago Press. Witzel, Michael. 1979. On Magical Thought in the Veda. Leiden: Leiden Universitaire Pers. . 1997. The Development of the Vedic Canon and its Schools: The Social and Political Milieu. In Michael Witzel, ed., Inside the Texts, Beyond the Texts: New Approaches to the Study of the Vedas. Opera Minora, vol. 2. Cambridge: Harvard Oriental Series, 256345. . 1998. Macrocosm, Mesocosm, and Microcosm: The Persistent Nature of Hindu Beliefs and Symbolic Forms. International Journal of Hindu Studies 1.13:501553. W Xnchou 2002. Jian b dian j ywn ynji . Gu wnz yu chtu wnxin ynji cngsh. Guangzhou: Zhngshn Dxu. W Yjing , coll. 1993. Mz jio zh . 2 vols. In Xn bin zhz j chng . Beijing: Zhnghu. X Pnln . 2001. Ln Z S de chn do sxiang . Kongz ynji, 6473. Xi Wiyng and Zh Yunqng , eds. 2004. Xn chtu wnxin y gudi wnmng ynji . Shanghai: Shnghai Dxu. Xjng zj . See Li Xn . Xng Wn . 1998. Chu jian Wu xng sh ln . Wnw 10:5761. . 2008. Guest Editors Introduction. Contemporary Chinese Thought 39.4: 317. X Zigu . 2003. Shng b Jian Xng qng ln bu sh y z . Sh xu j kn . Reprinted in Xi Wiyng and Zh Yunqng , eds., Xn chtu wnxin y gudi wnmng ynji (Shanghai: Shnghai Dxu, 2004), 155156. Xnz j ji . 2 vols. Compiled by Wng Xinqin . In Xn bin zhz j chng . Beijing: Zhnghu 1997. Originally published 1988. Yn Shxun . 2000. Gudin Chu jian san ln . In Wuhn Dxu Zhnggu wnhu ynjiyun, ed., Gudin Chu jian guj xush yntao hu lnwnj , 100107. Rnwn Lncng t j . Wuhn: Hbi rnmn. Yo Zhnzng . 1899. Q le bi l ywn . Shi shshnfng cngsh . Reprint, Shanghai: Kamng. Yi Sung-ryl . 2001. The View of Loyal Ministers in the Chu Bamboo-Slip Lu Mu-kung wen Tzu-ssu from Kuo-tien. Acta Asiatica 80:5271. Yoshikawa Kjir . 1953. Rshi ni okeru taig no kuh ni tsuite . Nippon Chgoku Gakkaih 4. Reprinted in Yoshikawa Kjir zensh , vol. 3 (Tokyo: Chikuma shob, 1969), 490504. Y Zhpng . 2000. Zhxu bnt shy xi de xn, xng, qng, jng tnjiChu jian Xng z mng ch de bi y zhong jish . In Wuhn Dxu Zhnggu wnhu ynjiyun, ed., Gudin Chu jian guj xush yntao hu lnwnj , 355361. Rnwn Lncng t j . Wuhn: Hbi rnmn. Zach, Erwin Ritter von. 1902. Einige Worte zu Prof. Gustav Schlegels La loi du paralllisme en style chinois. Beijing: n.p.

bibliography

385

Zng Chnhai . 2005. R xu jitng lnl yu zyu shhu de gngmn lnl . http://www.ica.org.cn/content/view_content. asp?id=64 (accessed February 22, 2007). Zhn Qnhu . 2002. Du Gudin Chu jian Tng Y zh do jianx, fnzhng de zi tntao. http://www.jianbo.org/Wssf/2002/zhanqunhui01.htm (accessed May 15, 2007). Zhng Gungy [Cheung Kwong-yue] and Yun Guhu , eds. 1999. Gudin Chu jian ynji . Taibei: Ywn. Zhng Lwn . 1998. Ln Gudin Chu m zhjian de pint h tin-rn you fn sxiang . Chuntong wnhu yu xindi hu 6:513. Translated and emended in The Text Titles of the Guodian Bamboo Strips (Guodian Chu mu zhujian), Contemporary Chinese Thought 32.1 (2000): 7983. . 1999a. L ln Gudin Chu jian de rn-y sxiang . Kongz ynji 1:5669. . 1999b. Qing d y sh de sh yu y. Gudin Chu jian ynji . Zhnggu zhxu 20:217220. Zhng Shnhu . 1990. Hnsh Ywn zh tngsh . Wuhan: Hbi jioy. Zhng Xianchng . 2004. Jianb wnxinxu tng ln . Beijing: Zhnghu shj. Zhang Zhenhua. 1991. Chinesische und europische Rhetorik: Ein Vergleich in Grundzgen. Europische Hochschulschriften, Reihe XX Philosophie, vol. 333. Frankfurt am Main: Verlag Peter Lang. Zho Jinwi . 1999. Gudin zhjian Zhng xn zh do, Xng z mng ch jiosh . Zhnggu zhxu sh 2:3439. Zho Pngn . 2002. Qing d y sh d jiu ho jian kaolnjin j xinQn liang-Hn wnxin zhng Bgn gsh de ynbin . Guj Zhngl Ynji Xukn 2:1821. Zhngzhng Shngfng . 2003. Shnggu ynx . Zhnggu dngdi yuynxu cngsh . Shanghai: Shnghai jioy. Zhnggu lsh dmng d cdian . 1995. Edited by Wi Sngshn . Guangzhou: Guangdng jioy. Zhu Fngwu . 1998a. Gudin Chu jian Zhng xn zh do kao sh . Zhnggu wnz 24:121128. Reprinted in Zhnggu zhxu 21 (2000): 137145. . 1999b. Gudin Chu m zhjian Tng Y zh do xn sh . Zhngyng ynji yun lsh yuyn ynji suo jkn 70.3.10:739759. . 2000. Gudin zhjian de xngsh tzhng j q fnli yy . In Wuhn Dxu Zhnggu wnhu ynjiyun, ed., Gudin Chu jian guj xush yntao hu lnwnj , 5363. Rnwn Lncng t j . Wuhan: Hbi rnmn. . 2002. . Q Lu xukn 4:1316. Zhu Fngwu and Ln Sqing . 1999. Gudin zh jian bin x fyun ynji . In Gu wnz yu gu wnxin , 5558. Taibei: Chu wnhu ynji. Zhungz zuanjin . 1993. Edited by Qin M . Reprint, Taibei: Dng d, 1993. Zumthor, Paul. 1983. Introduction la posie orale. Paris: ditions du Seuil. Zuo Sngcho . 1989. Mawngdu Hn m ynji zlio ml suoy n . Zhnggu shm jkn 23:95115.

INDEX
abdication, 1618 abstraction, 2; of thought, 249 actuality (of a thing), 220223; phenomenological actuality, 222223; philosophical actuality, 231 advisor, 5, 8, 19, 55, 87, 274 Alexandrian scholars, 21 Allan, Sarah, 9, 61, 188, 216, 274, 356 alterit (of Chinese thought), 12 anachronism: in the Lnyu, 133; anachronistic labels (of thought), 33, 78 anthology, 2324, 55, 7879, 99, 172, 186187, 189, 194, 205206, 212213, 217, 236, 248 antique market, 5, 171 antiquity, 22, 35, 4142, 44, 48, 150, 194, 258, 262, 273, 276; authors in antiquity, 22; texts of antiquity, 22 nyng , 241, 252 aphorism, 80, 238 archaeology; archaeological (appearance of texts), 241; evidence, 5; data, 243; record, 233; archaeologist, 6, 79 archaic: form (of a character), 200; language, 172; archaising style, 8 archetype, 173, 193 archive, 10, 252 argument (in early Chinese texts), (definition) 1112; 48, 5052, 5657, 61, 7173, 75, 8182, 87, 90, 102, 108, 117, 126, 128, 131, 179182, 213, 230231, 237238; coherent, 14, 26, 27, 116, 131, 213, 253; construction, 39, 5052, 57, 6364, 7172, 7475, 125, 250, 253, 328; closed, 6364, 70, 7275; open, 63, 66, 72, 277; self-contained, 64, 227, 229230, 253; argument-based texts, 1, (definition), 11 argumentative: chain, 38, 9293, 100, 102103, 109, 111, 115116, 127, 149, 164, 168, 302; force, 1112, 179, 185, 192, 206, 227, 230; patterns, 1112, 14, 19, 31, 54, 100, 177, 179; strategy, 18, 57; structure, 226, 359 Aristotle, 2, 50, 147, 249; Aristotelian (idea), 220 ars bene dicendi (the art of good speaking), 25 assemblage: of manuscripts, 6; of narrative or textual materials, 59, 82, 181, 189190, 195196, 200202, 213, 217; of different Vorlagen, 190; tomb assemblage, 4, 6 Assmann: Aleida and Jan, 2, 249, 252; Jan, 60, 206 asynchrony (of textual transmission), 174 audience, 5, 8, 87, 103, 207 authenticity (of a text version), 84 authorship, 20, 22, 133; and writing, 27, 178; of philosophical texts, 34, 2022, 179181, 184, 230, 235, 244; and a mine of materials, 130, 244 authoritative: force, 192, 200; ideas, 194; peg (of an argument), 106, 108; quotation, 107, 205; reference, 19, 202, 205; source, 23, 107, 199, 201, 253; statement, 108, 194; text, 204; tradition, 78; value, 24, 106 authority, 194, 203; underlying a text or unit of thought, 194; and meaning construction, 195, 201, 204, 206, 227, 238; recognition of, 207; textual 232, 240, 248, 254 authority-framed: overlapping structure, 108; textual unit, 108, 182 Bi Q (?257 BC), 7 Bagley, Robert, 241242, 273 Bal, Mieke, 235 bamboo strips: fake, 6; length of, 910, 16, 1819, 32, 51, 53, 83, 102, 142, 149150, 169, 187, 209, 211, 356; preservation of, 148149, 151, 342, 343 Boshn Two (Tomb Two from Boshn ), 7, 263, 298 Baxter, William, 68, 197, 267, 333334, 337 blank slate, 146, 169, 181 Behr, Wolfgang, 12, 25, 61, 118, 150, 301 Bi l (Separate Records), 21 Boltz, William G., 36, 80, 8182, 102, 177, 194, 196, 209, 216217, 242, 337

388

index
commentator, 1, 70, 82, 111, 281, 303, 333 commodification (of texts), 83. See also text as a (tradable) commodity comprehensibility (of a text), 103 conceptual: debt, 85; definition, 41, 177, 220, 229, 247; meaning, 213, 219, 221, 225, 226, 229 conceptualising: of ideas, 94; of quotations, 106 Confucianism (Confucian), 33, 78, 131134, 193, 331. See also anachronistic labels (of thought) Confucius: 78, 234. See also Kongz consistency: of ideas, 144, 202; of tomb equipment, 6; of texts 24, 8081, 86, 125, 142, 148, 168, 202; intellectual consistency 33; structural consistency 202, 206 context-dependent texts, 1; (definition), 11 copying: in antiquity, 198, 307, 332; mistakes, 152, 301, 357; from different sources, 106; silent 172 copyright (in antiquity), 2122 coronal substitution, 197 cosmogony, 14, 213214, 217220, 223226 cosmology, 133, 216, 226, 230, 239, 253, 355, 359 cosmos, 43, 219221, 223, 225226, 230 counterpart (of a text or manuscript): excavated, 4, 193; palaeographical, 14, 2627, 77, 82, 94, 100, 108, 132, 135136, 187, 300; transmitted, 1011, 13, 186, 193, 195 cross-referential links (in texts), 8990, 103, 108, 127, 129, 177, 181, 229, 247 cultural: identity, 90, 194, 200, 206; knowledge, 106, 107, 205; memory. See kulturelles Gedchtnis cuneiform, 22 Csikszentmihalyi, Mark, 22, 33, 78, 80, 84, 97, 116, 118, 186, 285, 301303, 308 D Di L j (Elder Dais Records of Ritual), 110, 266 Do d jng (Classic of the Way and Virtue), 13, 186 Daoism, 33. See also anachronistic labels (of thought)

books (and early societies), 23, 134, 170 B Y , 134 Brindley, Erica, 132, 146 Buz (or Bu Shng ), 134 calligraphy, 6, 910, 1516, 32, 53, 82, 138, 211; calligraphic similarity, 32 canon and canonisation, 2223, 83, 133, 193, 199200, 203 categorisation (of thought), 33. See also anachronism Caudwell, Christopher, 246 c , 241 charisma, 17, 302 Chn Guyng , 78, 212 Chn Jin , 54, 263, 265, 277278, 280281, 333, 348, Chn Lgu , 188 Chn Lnqng 337, 339 Chn Wi , 18, 54, 138, 141, 143, 216217, 260, 266, 275277, 280, 298, 300, 307, 348, 355356, 358359 Chng zh wn zh , also referred to as D chng (The Great Constancy), 18, 83, 137, 139, 330, Chnqi Zuo zhun (Zuo Commentary of the Springs and Autumns Annals) 9, 83 Chnqi Gngyng zhun (Gongyang Commentary of the Springs and Autumns Annals), 329 circular: argumentative pattern, 100; form of self-cultivation, 89, 128, 181182; formal structure, 127128, 181 Clanchy, Michael, 150 closed: argument: 6364, 70, 7275; corpus of texts, 4; recension, 173, 193; set of manuscripts, 6; system (of thought), 7376, 128, 130131, 183, 230, 234; syllables, 150; text or composition, 3, 130131, 136, 141, 179180, 184, 195, 202, 217, 230231, 234235, 254 cognitive: competence and abstract thinking, 243, 249; grasp (of a concept or an idea), 87, 115, 183 collage: of building blocks, 41; composition, 41, 43, 107; Collage Technik (collage technique), 41 commentary, 80, 8486, 9899, 116118, 207, 217, 226, 234, 259, 303, 342

index
Davis, John Francis (17951890), 25 Deleuze, Gilles, 193 dialect (of Chu), 68, 197, 261, 265267, 324, 334, 337, 340, 342 dialectical: chapters of the Mjng, 19; relationship between text and textual community, 13, 194, 228 dictation: of scribe to himself, 150, 172, 197, 198; and copying, 332; and writing, 307 Dng Sxn , 16, 81, 84, 109, 216 discourse: Confucian (or r-ist), 132, 220; connected, 204205; intellectual, 250; oral, 23, 227, 232233, 236, 246, 249; philosophical, 5, 1113, 15, 18, 25, 28, 34, 60, 8687, 110, 131133, 178, 227, 235, 247; surviving, 2; written, 12, 235; connection to, 144, 199; participation in, 178; distant (text features): abc scheme, 162163; parallelism, 98; principal insertion, 99, 179; text segment, 109 division of labour, 83, 210 double-directed: function (of text segments), 39, 58, 63, 66, 7274, 93, 109, 220; parallelism, 99, 177 Dubs, Homer H., 77 Duke Hun of Zhng (r. 806771 BC), 203 Duke Wu of Zhng (r. 770744 BC), 203 Eadmer of Canterbury, 150 editor: of ancient works, 2022, 27, 102; editorial interference, 8, 55, 8081, 84, 168 Egypt, 241 Ehlich, Konrad, 8 ekdosis (edition), 21 lite: circle, 27, 178, 184, 194; community, 22; culture, 147, 163164, 170171; group, 2324, 133, 206207, 234, 236; political, 24; Emperor Chng of Hn (r. 327 BC), 21 Emperor Jng of Hn (r. 156141), 79 Emperor Wn of Hn (r. 179157 BC), 79 Emperor Wu of Hn (r. 14187), 251253 encyclopaedia: of knowledge, 248; systematised 238

389

epistemology, 12 erudites: appointment, 252; office, 251 Ess, Hans van, 199 evocative: scheme, 230; structure, 205 excavation, 5, 1314, 16, 23, 33, 55, 79, 80, 131, 186, 188; documented, 4, 245; rescue, 6; scientific, 135 exegete, 232 exegetic: effort, 107, 121, 207; layers, 90; process, 90, 100, 239240; tool, 253; tradition, 97, 107; wrung out, 238 expectation (on the part of text recipient), 57, 63, 65; violation of, 52 Falkenhausen, Lothar von, 5, 10, 242, 247 Farmer, Steve, 231, 238241, 243 florilegium: of ideas, 80, 185; of unrelated textual units, 185 form (edos or morph ), 258 form analysis (of texts), 2426, 136, 173, 230 formal perfection (of text patterns), 52, 61, 181, 236 Fhrer, Bernhard, 150 Gabelentz, Hans Georg von der (18401893), 25 Galambos, Imre, 151 Gassmann, Robert, 25, 196 Gegentext, 233 Gentz, Joachim, 6, 12, 25, 99 Giele, Enno, 5, 79 Goody, Jack, 153, 240, 241, 249 government: appropriate, 5; light, 35; moral, 15, 34, 188; principles of, 34, 118; treatise on, 44, 48 Graham, Angus C. (19191991), 18, 25, 55, 132 Greco-Roman world, 179 Greece, 238, 239, 241, 243 Greek: ideas, 239; philology, 21; tradition, 11, 239; work (verb), 21 Groningen, B. A. van, 21 Gngsn N , 131 G Jigng (18931980), 212 Guanz , 20 Guan Xhu , 139, 189 Gun Yn Daoists, 216 Gu Y , 18, 137, 138, 140 habitus, 147 Halverson, John, 250

390

index
King Kui of Yn (r. 320314), 17 King Yu of Zhu (r. 781771), 203 Kirk, Geoffrey, 243 Kongz , 33, 7778, 132133, 200201, 234, 288, 327, 329 Kongz sh ln , 79, 9798, 134, 265, 348 Korpel, Marjo, 56, 98 Kristeva, Julia, 235 kulturelles Gedchtnis (cultural memory), 60, 90, 181, 206, 252, LaFargue, Michael, 180, 186, 201, 212 language: archaic, 172; borrowing and contacts, 267; dying, 241; formulaic, 135, 177; mantra-like, 128, 250; mythopoetic (of Homer), 238; of Old Chinese, 265; philosophical, 3; and formal structure, 31, 163; Middle Chinese of the Qiyn, 68; Laoz , 1314, 25, 70, 7980, 180, 185194, 195, 201, 209219, 225226, 232, 233234, 257, 302, 333334, 337, 340; Laoz A, 18, 296; Laoz B, 18, 19; Laoz C, 14, 19, 28, 82; Proto-Laoz, 14, 28 Lau, D. C., 55, 212, Leclercq, Jean, 150 legendary: figure, 186; materials, 5962, 6566, 7172, 273274; ruler, 16 Li Shou (Mount Li), 272 Leseanleitung, 202, 205 Lewis, Mark Edward, 15, 24, 7879, 252, 278 lexical: differences, 118; dissimilarities, 125; level, 80; list, 241242; variation, 100, 108; lexically related, 197 L j (Record of Rites), 13, 102103, 355 L Lng , 16, 148, 193, 305, 326 L Tinhng , 131, 133135, 137139 L Xuqn , 7, 9, 18, 20, 3233, 137139, 186, 193, 216217 Lio Mngchn , 33, 137139, 142143, 159160, 316, 342 Lichtenberg, Georg Christoph, 231 ligature (hwn ), 261262 lingua franca, 248 literacy, 235, 240, 243, 249 literary: bias for evidence, 245; level of text, 228; patterning, 180; production,

Hn Fez , 78 Hn sh wi zhun (Hans Outer Commentary to the Songs), 78, 103, 278, 280281 Hn Yng (ca. 200120) 7778 Harbsmeier, Christoph, 19, 25, 132 Havelock, Eric, 238241, 243 H Lny , 260, 282, 300 Hebrew Bible, 22 Hellenistic period, 21 Henderson, John, 239241, 243 Hshng gng , 234, Holloway, Kenneth, 103 homophone, 306307 Huinnz , 199, 276 Humphreys, Sally C., 243 ideal type, 1, 11, 27, 227, 240, 245 identity, 238, 272; formation, 23; cultural, 90, 194, 200, 206; group, 194, 207 Ikeda Tomohisa, 54, 84, 292, 299, 301303, 308309 Iliad: 238 illocutionary force (of a text): 246247, 249 imperial, 910, 20, 2223, 34, 83, 248249, 251252; court, 252; library, 9, 21, 251252; patronage, 7, 251, 253; recognition, 79 India: 239, 240241, 243, institutionalising: of thought, 7; of writing, 251 interchangeability of concepts, 107 interlocking parallel style, 3739 intertextuality, 235, 248 inventory strip, 7, J Xshng , 134135, 272, 328, 331, 335336 ji (lineage), 33; bi (manifold), 252 Jing Gunghu , 16 Jxi , 216 Johnson, William, 240 Johnson, David, 278 Judaism, 231 Jullien, Franois, 12 Kalinowski, Mark, 204 Karlgren, Bernhard, 162, 239, 243 Keightley, David N., 242, 247 Kern, Martin, 68, 20, 2224, 33, 66, 79, 85, 9899, 102, 130, 135, 149, 153154, 171172, 202, 242, 247248, 251252, 282

index
22; scientist, 193; superstructures, 246 Li Bng (r. 202195), 85, 304 Li d , also referred to as D y (Rightness and Virtue;), 1819, 83, 137138, 143, 307 Li Xing (798 BC), 21 Li Xn (46 BCAD 23), 21, 23 Li Xnfng , 104, 134, 303, 306, 309 Li Xnln , 340 Li Zho , 264 Lloyd, G. E. R., 250 logic: and epistemology, 12; and grammar, 25; logical analysis, 19; deduction, 11; stringency, 84; structure (of text and argument), 26, 56, 64, 81, 86, 88, 9495, 108, 146, 147; syllogism, 11, 50 looting, looter, 6, 135, 171, 245 Luyng , 24 Lu M gng wn Z S (Duke M of Lu Inquires of Z S), 1516, 19, 32, 34, 5354, 189, 190, 210 Lnhng (Balanced Discourses), 60 Lnyu (Analects), 9, 55, 83, 132134, 147 Maspro, Henry, 87, 214, 323 master: copy, 141; and authority, 194, 201, 206; introductory statement in text, 199200, 202, 204, 232; and interpretation, 203, 228; and triangular relationship (between text, master, recipient), 231 material: carrier (of text and ideas), 1, 89, 28, 50, 82, 102, 210211 216, 218, 226, 242245, 247; conditions (of text and manuscript culture), 1, 24, 211, 241, 243, 246; (for meaning construction), 28, 227; materiality (of a text), 5051, 102, 117, 169; (of meaning construction), 239 McLuhan, Marshall, 249 matter (hl ), 258 memory: cultural, 60, 90, 181, 206, 252; shared, 236; memory and text (relationship), 3, 131, 252; and writing (relationship), 307 mediator: of meaning, 1, 3, 204, 206, 228, 229, 246; of thought, 1, 231232, 254 manuscript: culture, 1, 10, 22, 24, 210, 243244, 246247, 249250; illegal trade, 171, 187, 245; production, 10,

391

83, 137, 150, 190, 210211, 243, 262, 340 Marquis Di (L Cng , d. 185), 79 Matsumoto Masaaki , 24, 79 Mawngdu , 6, 1314, 22, 26, 77, 7981, 8486, 9395, 98, 100, 103, 108, 111112, 117118, 121, 125127, 129, 136, 183184, 186, 192, 209, 236, 252, 287288, 296309 Mngz, 17, 50, 55, 78, 84, 86, 103, 106107, 122, 132134, 147, 234, 284, 288, 307, 327; four sprouts (s dun ), 78, 86 Mng K (also Mencius), 7778, 132 Mesoamerica, 240 Mesopotamia, 22 Meyer, Dirk, 6, 12, 54, 58 Middle East, 240 mimesis, 180 Mn zh fmu (Parents of the People), 79 m (or m zh ), 33, 79 module: steady text module, 168169, 212; text as module, 204, 227229, 232233, 236, 246, 254 Mjng , 19 Mz , 78, 260, 273, 308 Near East, 22, 239 netherworld, 13, 79 Nicomachean Ethics (by Aristotle), 147 Nylan, Michael, 7, 2223, 33, 86, 133, 186, 251 Olson, David, 2, 240, 249 Ong, Walter, 249 open: argument, 63, 66, 72, 277; interlocking parallel style, 38; interpretation of texts, 207, 233, 254; recensions, 173, 193; syllables, 150; nature of texts, 131, 198; Oppenheimer, Leo, 22 orality: and literacy, 235; oral reading, 150; oral (also spoken) discourse, 23, 227, 232233, 236, 246, 249; oral text, 10, 27, 83, 152, 168169, 244, 250, 254 Otto, Walter, 243 overlapping structure, 3638, 4041, 46, 5051, 5354, 72, 106108, 123; in comparison to interlocking parallel style, 3839; interleaved, 123124 palatalisation (of velars), 197, 334 palm leaves (as text carrier), 241

392

index
Quintilian, 25 quotation, 99, 106107, 130, 205, 240; instability of, 2324, 86, 100, 108109, 118, 172, 198, 200, 299; of Documents (sh), 2324, 79, 197, 198; of Odes (sh), 7880, 99, 104, 106, 172, 233, 248, 262, 299, 309; and argument construction, 123, 134, 199, 204, 228, 235238; and text development, 10, 235, 248 Qing d y sh (Failure and Success Appear at their Respective Times), 1516, 19, 26, 5376, 87, 89, 131, 144, 145, 180181, 183, 189190, 210, 223, 225, 230232, 235237, 253, 269282, 329 Rahn, Helmut, 25 Raven, John E., 243 recension, 83, 173, 193, 245 recipient (of a text), 8, 11, 60, 63, 66, 6869, 102103, 107, 111, 115, 124, 128, 162, 192, 201, 202204, 206, 228229, 246, 254 record-keeping, 3 reading: in antiquity, 50, 150, 197, 227, 231, 236, 245; silent, 172; strategy, 48, 50, 6869, 90, 112, 127, 205, 218, 221, 223, 225, 234, 236; reasoning, 229, 235, 250; philosophical, 2, 5, 11, 51, 207, 251 referential framework: of tomb contexts, 6, 27, 171; of texts, 234 Renaissance, 231 repetition: signs, 149, 287, 301, 307, 325, 332, 334, 336, 357, 348, 356; characters, 306; texts and text patterns, 61, 134, 254 Reu, Wim De, 12 rhizome, 193 rhyme, 6869, 262263, 334, 337, 340, 357; class, 265; distinction, 333; evidence, 337; group, 262263, 267, 337; pattern, 68 rhythm, and importance (of a text), 31, 179; and comprehensibility (of a text), 103, 253; and stability (of a text), 61; and structure, 51, 53, 189; mnemonic function, 51 Riegel, Jeffrey, 79, 85, 9899, 104 rites (l ), 133134, 146147, 158, 163, 170, 315, 335 ritualist, 78, 186 Robinet, Isabelle, 207 Rng Chng sh , 18

Png P , 78, 84, 95, 119, 132, 302, 304, 306309 papyrus (as text carrier), 241 parchment (as text carrier), 241 parallelism (in Chinese texts), 25, 38, 58, 94, 98, 107, 164 Png Ho , 5, 7, 32, 189 pericope (definition), 89 persuasive definition, 41, 58, 229 Petersen, Jens stergrd, 7, 22, 33, 251, 252 phenomenological: actuality, 222223; world, 132, 140, 145146, 162163, 169, 181, 225 philosophising: practice of, 1, 13, 24, 26, 33, 51, 206, 245246; strategies of, 12, 178, 227, 230; traditions of, 28; types of, 251 phonetic: borrowing, 197, 267, 306307, 324; criteria for similarity in Old Chinese, 150; loan, 278, 301, 337, 346, 358; similarity, 150151, 153, 171172, 197, 265266, 302, 307, 331332, 337, 339, 345, 350351, 358, 360 phonophoric (of a Chinese character), 151153, 196197, 261, 264265, 282, 300, 304, 323324, 326, 330331, 333334, 336, 338342, 344351, 358, 360 physical appearance: of an argument, 102; of text and manuscript, 19, 187 Pico, 231 Pines, Yuri, 1618 Plato, 11, 33, 238 Platonic: notion of a perfect thing, 11; ideal world 33 principal insertion: 99, 116117, 177, 179 Protestantism, 231 P Mozuo , 134, 338339, 344 Puett, Michael, 132, 134 Q le (Seven Summaries), 21 Qin f ln (Comments of a Recluse), 266 Qin M , 272 Qn dynasty (ca. 221210 BC), 23, 252; institutionalising of thought, 7, 22, 252253; imperial court, 252; office of erudites, 251 Qi Xgu , 32, 193, 216217, 222, 261264, 266, 274, 280, 300, 304, 308309, 323, 325326, 332337, 341, 343344, 346, 357, 359360

index
r (r zh ), 33, 7779, 220 Ruan Yun (17641849), 197 ruler: accomplished, 1617, 103, 129 Sagart, Laurent: 197, 242, 267 school (philosophical), 33, 55, 84 Schaberg, David, 24, 7879, 180 Schlegel, Gustave (18401903), 25 Schwartz, Benjamin, 35, 133 scribal: custom, 150; error (also mistake), 93, 118, 296, 301, 309; idiosyncrasies, 118, 126, 150, 196, 262; tendencies, 301, variation, 100 scribe, 22, 141, 149150, 152, 172, 189, 196198, 210, 301, 307 scripture, 231; Chu script, 9, 32 Seidel Anna, 186 self-contained: argument, 64; textual unit, 56, 139, 141, 178, 184, 212, 229, 235; philosophical unit, 82, 227, 230, 253 self-cultivation (also: moral self-cultivation), 5, 84, 8689, 94, 99, 127129, 145; and formal textual arrangement, 94, 99, 127128, 181; paradox of self-cultivation: 89, 128; theory of, 84, 8687, 127129, 142, 181, 183, 236 Seters John van, 2122, 154 Shnghai collection of Chu manuscripts, 5, 1314, 18, 27, 79, 82, 132, 134136, 142, 186, 195196, 198, 228, 251, 253, 311 Shang xng (Reward and Punishment), 18 shared knowledge, 80, 103, 104, 253 Shaughnessy, Edward, 80, 102, 124, 186, 195, 200, 212, 242243, 262, 275 Shn yn yng xng (Cautious Words and Wise Actions), 19 Shn Yu (441513), 186 Sh (Odes), 2324, 7880, 86, 90, 95, 97100, 102, 104, 146147, 252253; references to, 105109, 118, 123, 125126, 158, 163, 198200, 203, 237, 262263, 285, 287288, 290291, 294295, 299, 304, 307, 315; consistency of, 133135, 149, 154, 170, 172, 244, 247248; meaning construction and interpretation, 204, 206, 232233, 236; rhyme evidence, 337; Mo tradition 7879, 9798, 253 Sh j (Archivists Records), 134, 252, 273, 276

393

Sh Shu : 84, 131 Sh (Documents), 2324, 79, 133, 135, 146147, 158, 163, 170, 197, 198200, 202, 206, 232, 252, 315; Hng fn (Glorious Plan), 133; Yn go (Proclamation of Yn), 197; sn d , 133 Shuwn jiz , 197, 258, 264, 298, 306 Shunggudu , 79, 102 Shuhud , 261, 325 Shn , 1617, 59, 159160, 269, 273, 275, 316, 317 semiotic webs (in texts), 229231, 234 signific (of a Chinese character), 148, 151, 153, 171, 196, 300, 323324, 326, 330331, 333339, 344346, 348349, 351 silk (as text carrier), 8, 10, 79 Spirin, Vladimir, 12, 185 stemma codicum, 86, 173, 193 Stock, Brian, 178 stock phrase (in early texts), 130, 236, 244, 250 stream of tradition, 22 Street, Brian, 240 structural significance (of text patterns), 4041, 58, 107, 118 subcanto (of a text), (definition), 5657 Sudelbcher, 231 Su sh , 186 Sumerian, 241 taboo (observances in writing), 85, 304 tadpole symbol, 139140, 189190, 209, 212, 349 Tng Y zh do (The Way of Tng [Yo] and Y [Shn]), 9, 1617, 19, 32, 190 Ti y shng shu (The Ultimate One Gives Birth to Water), 14, 19, 28, 82, 188, 209226, 230, 232, 253361 terminus technicus (also: technical terminology), in early texts, 60, 78, 107, 120, 123124, 130, 225, 236, 327 text: authority-based, 194; fixed, 21, 136, 170, 233; fluidity, 10, 168; oral, 10, 27, 83, 152, 169, 244, 250, 254; original, 84, 90, 97, 102, 130, 193, 238; manifestation, 1, 4, 8, 10, 21, 27, 32, 60, 7986, 8990, 95, 100, 103104, 108109, 111112, 117118, 121, 125130, 135136, 141143, 148152, 154, 163164, 168169, 171174, 183,

394

index
travelling concept, 235236, 239240, 248, 250 triangular relationship (of meaning construction), 1, 204, 206, 229, 231, 233, 244, 236, 254 Tsuda Skichi , 133 unit of thought: 129, 184185, 189, 192, 199, 202204, 213, 223, 226, 232 urtext, 20, 97, 141, 150, 173, 245 Urschrift, 150, 173 variation (of bamboo strips), 10; (in texts), 27, 137, 153154, 202; (in quotation), 172; graphic and lexical, 108, 125, 196198; scribal, 100, 172 Vogelsang, Kai, 24 Vorlage (source text), 141, 149150, 152, 168169, 172174, 190, 196197, 211, 329 Wagner, Rudolf G., 12, 25, 37, 38, 129, 180, 184, 185, 199, 234, 302 Wng Baoxun , 7 Wng B (226249), 180, 193, 217, 234, 337 Wng Chng (AD 2797), 9, 60, 83, 84 Warring States period (ca. 481222), 1, 5, 8, 78; abdication, 16; intellectual climate, 7, 3334, 78, 196, 216, 249, 250251; Documents, 198; Laoz, 212; Odes, 9798, 233, 244, 248; phonetic loan, 323, 337; strategies of argumentation, 52, 58; text and manuscript culture, 910, 23, 27, 50, 81, 83, 102, 134135, 183, 186, 194, 202, 210, 227, 239, 241245, 248, 250251, 254, 262; textual communities, 178, 195 Watt, Ian, 249 Weber, Max, 11 West, Martin, 173, 193 Witzel, Michael, 239241, 243 workshop, 1516, 32, 8283, 137 writing, 13, 8, 1012, 22, 2728, 31, 50, 56, 7879, 8283, 103, 139, 140141, 148, 150152, 171, 205, 207, 210, 234, 262, 266, 294, 301, 304, 307, 324, 329, 332, 337, 340, 343; brush, 242; carrier, 1, 242, 246247; composed in, 2; conditions of 51, 227; extensive

187, 193, 196200, 202, 205206, 228, 231, 236, 248, 250, 252253, 296297, 299, 301302, 311, 319, 325, 329, 331332, 335, 343, 349; macrostructure, 37, 4546, 48, 5153, 7173, 75, 131, 136, 177180, 223224; microstructure, 36, 39, 46, 53, 224; micro-macrostructure correspondence, 177; mouvance, 130; and material carrier, 1, 89, 28, 50, 82 102, 210211 216, 218, 226, 242, 245, 247; permutation, 254; production, 178, 189190, 246; stability, 4, 31, 53, 63, 80, 82, 86, 89, 103, 108, 130, 149, 153154, 168, 202, 206; standardisation, 11, 20, 24, 79, 153154; syncretic tendencies, 235, 238, 240, 243, 250; as aide-mmoire, 247; as cultural phenomenon, 2, 26, 245; as (tradable) commodity, 210; as material object, 2, 211; as performance, 170, 231, 246; as philosophically coherent edifice, 2, 4, 132, 230, 232; as repositories of ideas, 1, 31, 50; as synchronic artefact, 22, 174, 235, 244, 250; as vessel of thought (or ideas), 1, 33, 245; intertextuality, 235, 248 textual: ancestor, 8485, 174; bridge: 57, 63, 6569, 71, 74, 253; communities, 1, 178, 195, 198, 201207, 228, 230, 232234, 245, 247, 254; organisation, 13, 26, 31, 37, 54, 72, 73, 80, 82, 86, 95, 100, 104, 108, 117, 128, 135136, 138, 142, 180, 182, 192, 195, 209, 239240, 275, 359; relationship between texts, 16, 54, 86, 198, 209 Thirteen Classics, 197 Thomas, Rosalind, 153, 240 To ti esti ( ) the what it is, 220 Toorn, Karel van der, 22 topos (in philosophical texts), 18, 56, 66, 70, 110 tomb: corpus, 68, 1920, 3132, 8283, 108, 179, 182, 193, 195196, 198; occupant, 6, 7, 193 transmission, 13, 104, 149, 194, 232; through authority, 252254; asynchrony, 174; error, 111, 118, 151; history, 10; process (of texts), 10, 131, 152153, 169, 172, 186, 254, 329; as reinvention (of the text), 172; of philosophical texts, 172 Trauzettel Rolf, 41

index
247; influence of, 249; instability of, 135, 149, 172, 197; institutionalised, 251; materials (and carrier), 240241, 243244, 246247, 249, 251; medieval, 150; practice of 26, 246; productivity, 247; prosaic style, 164; system, 241; transmission (or circulation), 202, 207, 234235, 244, 249, 254; and abstract thought, 239; and copying, 307; and systematisation of ideas, 238241, 250 writings: banning of, 251252; burning of, 22, 252; philosophical, 50, 172, 178, 184, 190, 211, 244246, 249, 251 written: communication, 247, 249; discourse, 12, 234; entity, 170; ideas, 24; mode of reasoning, 235; object, 22; relationship with the spoken word, 20; Wu xng (Five Aspects of Virtuous Conduct), 14, 18, 21, 26, 27, 77132, 135, 136, 147, 153, 163, 181184, 186, 231232, 236237, 240, 250, 252253, 283309, 327, 344 wu xng theory, 27, 77, 8687, 8990, 9495, 104, 111112, 117118, 120, 122123, 125131, 136, 182184 Wu Zx , 62, 278279 Xng qng ln (Treatise on Nature and Sentiment), 6, 14, 82, 135136, 137143, 148153, 162164, 168171, 173174, 311, 319, 323351 Xng Wn , 8081, 84, 108109, 134 Xng z mng ch (Nature Derives from Heaven), 6, 14, 18, 27, 82, 131174, 181183, 186, 197, 232, 311352 Xinyng (capital of Qn), 252 Xing Yu (232 BC202 AD), 252 Xio jng (Classic of Filial Piety), 9, 83 xishng : contacts, 68; rhyme group, 267; series, 196, 267 xi c (use of proper words), 25 Xnz , 19, 55, 7778, 85, 132, 134, 147, 228, 276, 279281, 303, 329, 358;

395

Fi shr z (Against the Twelve Masters), 77 Xn Qng (ca. 310238 BC), 77 Yo , 16, 17, 160, 269, 274, 317 Yates, Robert, 20 Y (Changes), 80 Yn (last capital of Shng), 241 Yng (capital of Chu), 7 Ywn zh (Records of Arts and Letters), 21, 23, 84 Yoshikawa Kjir (19041980), 25 y writing brush, 242 Yu cng (Collected Sayings), 1820 Zo F (legendary figure), 64, 271, 280 zhng (textual unit), 26, 188, 190, 193, 199, 217 Zhng Xun (AD 127200), 9, 83, 342 Zhngzhng Shngfng , 197, 325 Zhngshn , 17, 302, 336 Zhng xn zh do , 9, 14, 1617, 19, 26, 3153, 55, 58, 7274, 76, 78, 87, 89, 131, 178181, 183, 186, 190, 225, 229232, 253, 257268 Zhng yng (Doctrine of the Mean) 78, 186 Zhu Fngwu , 18, 137, 193, 263, 267 Zhungz , 78, 193, 258259, 279, 337, 340 Z Go , 18 Z S , 15, 7778, 84, 131, 186 Z Zh (minister of the state of Yn ), 17 Z y (Black Robes), 6, 13, 18, 7980, 8283, 102, 185186, 195206, 228, 234, 236, 254, 262 Zumthor, Paul, 130 Zn d y (also Shang xng , Reward and Punishment), 18, 19, 137