A S T U D Y O F T H E W A R R I N G S T A T E S G R A P H S : S T R U C T U R A L D I S C R E P A N C I E S B E T W E E N T H E C H U A R E A A N D Q I N S T A T E by Y U K I K O N A G A K U R A B . A . , University of Alberta, 1995 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R OF A R T S in T H E F A C U L T Y OF G R A D U A T E STUDIES (Department of Asian Studies) We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A October 2000 © Yukiko Nagakura, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. The University of British Columbia Vancouver, Canada Date /OsJ. J . 0 ^OnS) DE-6 (2/88) Abstract Chinese graphical forms of the Warring States period have traditionally been characterized as varying by region. This thesis investigates discrepancies observable in the scripts of the Warring States Chu and Qin regions, as extant in inscriptions and epigraphy on a variety of media. Wi th the postulate that the Warring States graphical forms were part of a continuous evolution of guwen from the Shang period to the Script Reform that followed the Qin Unification, these discrepancies are treated as the accumulations o f a common diachronic process. To define this process, the two-step formation o f semanto-phonetic graphs is adopted as jiajie augmented with semantic determiners, and evolutionary modifications tending to induce graphical divergence are classified for simple and multi-element graphs, based on the work o f Boodberg, Boltz, Chen, Qiu, He, Gao and others. A table of graphs is constructed to allow for diachronic evaluation o f the process o f formation and evolutionary modification, and for synchronic comparison o f Chu and Qin forms, using the Baoshan bamboo slips as the primary source, supplemented with material from the Shuihudi bamboo slips, the antecedent Shang oracle bones and Western Zhou bronzes, and other sources. Discrepancies observable in simple and multi-element graphs are discussed in detail, placed in the appropriate category o f formation or evolutionary modification, and, i f possible, assigned the period o f genesis (Shang, Western Zhou, or Warring States). The 12,472 total Baoshan bamboo-slip graphs are placed into the 1,473 entries in this table. 154 cases of discrepancies are discussed, o f which the genesis of 33 can be dated, with 7 originating in the Warring States period and 26 in the Shang and Western Zhou periods. A high degree o f homogeneity in the Chu and Qin Warring States script is thus confirmed. Notwithstanding that the scarcity o f source material precludes definite conclusions, it is suggested as a possibility for further study that graphical discrepancies were not at all a feature o f the Warring States period exclusively, but should be traced to earlier times as accumulating throughout entire period when guwen were used. T A B L E O F C O N T E N T S Abstract i i Table o f contents i i i List o f Tables v i List o f abbreviations and notational conventions v i i i Acknowledgment ix I N T R O D U C T I O N 1 C H A P T E R O N E Terminology and Literature Review 4 1.1. Terminology 5 1.1.1 Liushu 7\ • 'the six types o f script' 5 1.1.2 Terms used in this thesis 6 1. 2. The earliest process for formation o f the Chinese script 8 1.2.1 The three stages o f the early graphical development (Boodberg and Boltz) 8 1.2.2 The three stages of the early graphical development (Chen and Qiu) 11 1. 2. 3 The features o f H Y graph or "ideographic" compound graph 15 1.2.3. 1. The existence of H Y as a multi-element graph formation principle 15 1 .2 .3 .2 . The principles o f the H Y graph formation 17 1. 2. 4 The features of xingsheng (SP) graphs 18 1. 2. 4. 1. Thejiajie and yinshen" 18 1. 2. 4. 2. The principles of the SP graph formation 19 1. 2. 4. 3. The alternate readings o f graphs 20 1. 2. 4. 4. PEs and SEs consisting o f multiple elements 21 1. 2. 4. 5. The relative positions o f a P E and an S E within a single graph 22 1.3. The old script (guwen i?X)'- From the Shang to the Warring States period 23 1.3.1 The chronological division o f the Pre-Qin writing (guwen jr_) 23 1.3.2 The Shang writing 24 i i i 1.3.3 The Western Zhou writing 26 1.3.4 Writ ing o f the Spring and Autumn period 27 1.3.5 Writ ing o f the Warring States period 27 1.3.6. The graphical evolution of guwen 30 1.3.6. 1. The orthodox and vulgar forms o f the script 30 1. 3. 6. 2. Simplification and elaboration 31 1.3.7. The graphical features of the Warring States graphs 33 1.3.8. The Script Reform at the Qin Unification 40 1. 4. Formulation o f the development of guwen adopted in this thesis 41 C H A P T E R T W O Material and Methodology 43 2. 0. Introduction 43 2 . 1 . The table 43 2 .1 .1 The arrangement o f the horizontal line of the table 43 2 . 1 . 2 . The vertical order o f the table 48 2. 1. 3. The design of the table 48 2. 2. The main material: the Baoshan bamboo-slip graphs 48 2. 3. Procedures for analysis o f structural discrepancies 49 2. 3. 1. Transcription into modern forms and treatment of variants 49 2. 3. 2. Principles for analysis of the structural discrepancies 51 C H A P T E R T H R E E Simple Graphs 54 3. 0. Introduction 54 3 .1 . Definition o f simple and multi-element graphs 54 3. 2. Types o f discrepancies between the Chu and Qin simple graphs 56 3 .3 . Tables o f simple graphs 57 3. 4. Commentaries on discrepancies between the Chu and Qin simple graphs... 68 C H A P T E R F O U R Multi-element Graphs 70 4. 0. Introduction 70 4. 1. Types of discrepancies between the Chu and Qin multi-element graphs 70 4. 2. Structural discrepancies present during the process o f formation 72 4 . 2 . 1 . SP graphs 72 4. 2. 1. 1. Divergent usage of jiajie graphs 72 4 .2 . 1. 2. Divergent usage of semantic determiners 73 iv 4. 2. 1.3. Chu augmentation of jiajie with semantic determiners 76 4. 2. 1. 4. Different graphemes used at SP formation 78 4. 2. 1. 5. Different usage of semantographs 79 4. 2. 1. 6. Different usage of semantographs and PEs 80 4. 2. 2. H Y graphs 80 4. 3. Divergent structural modifications o f SP graphs 81 4. 3. 1. Omission of SP grapheme 81 4. 3. 2. Replacement of SP grapheme 82 4 . 3 . 3 . Augmentation of SP with extra P E 88 4 . 3 . 4 . Augmentation of SP with extra SE 90 4. 3. 5. Augmentation of SP with grapheme o f unknown function 93 4. 4. Different structural modifications of H Y graphs 94 4. 4. 1. Omission o f H Y grapheme 94 4. 4. 2. Replacement o f H Y grapheme 95 4. 4. 3. Transformation of H Y to SP 97 4 . 4 . 4 . Augmentation o f H Y with extra S E 100 4. 4. 5. Augmentation o f H Y with grapheme o f unknown function 104 4. 5. Divergent evolutionary changes from and misrecognition o f earlier forms 106 4. 6. Unclassifiable structural discrepancies 114 4 .6 . 1. Cases in which ancestors cannot be determined 114 4 . 6 . 2 . Cases of unclassifiable graphs and modifications 118 4. 7. Discrepancies existing in the earlier times 119 4. 8. Summary o f modifications discussed 122 4. 9. Tables of multi-element graphs 127 C H A P T E R F I V E Conclusion and Implication 144 Bibliography 146 Appendix I Table: The Comparison o f the Chu and Q i n W S Graphs 160 Appendix II List o f Graphs from the Western Zhou Bronze Vessels 227 Appendix III List o f the Western Zhou Bronze Vessels 232 Appendix I V List o f the Chu Bronze Vessels 234 Appendix V List o f the Qin Bronze Vessels 240 v List of Tables Table 1. 1 25 Table 1.2 33 Table 3. 3. 1. 1. Omission o f strokes 58 Table 3. 3. 1. 2. Addit ion of strokes 58 Table 3. 3. 1.3. Executional discrepancies 60 Table 3. 3. 2. Simple graphs with structural discrepancies 61 Table 3 . 3 . 3 . Unclassifiable discrepancies 61 Table 3 .3 .4 . Adoption o f different oracle-bone variants 61 Table 3 . 3 . 5 . Simple graphs exhibiting no divergence 62 Table 3 .3 .6 . Graphs which do not have the Qin counterpart graphs 67 Table 4. 1. 1. Graphs discussed in Chapter 4 123 Table 4. 1.2. Graphs which underwent multiple divergent processes 124 Table 4. 2. 1. 1. Divergent usage ofjiajie graphs 127 Table 4. 2. 1. 2. Divergent usage o f semantic determiners 127 Table 4. 2. 1.3. Chu augmentation of jiajie with semantic determiners 128 Table 4. 2. 1. 4. Different graphemes used at SP formation 129 Table 4. 2. 1. 5. Different usage o f semantographs 129 Table 4. 2. 1. 6. Different usage o f semantographs and PEs 129 Table 4. 2. 2. H Y graphs 129 Table 4. 3. 1. Omission of SP grapheme (A) Chu graphs 130 Table 4. 3. 2. Replacement of SP grapheme (O) Both Chu and Q i n graphs 130 (A) Chu graphs 130 (B) Qin graphs 131 Table 4. 3. 3. Augmentation o f SP with extra P E (A) Chu graphs 132 (B) Q i n graphs 132 Table 4. 3. 4. Augmentation o f SP with extra S E (A) Chu graphs 132 (B) Qin graphs 133 v i Table 4. 3. 5. Augmentation o f SP with grapheme o f unknown function (A) Chu graphs 133 (B) Qin graphs 133 Table 4. 4. 1. Omission of H Y grapheme (O) Both Chu and Qin graphs 134 (A) Chu graphs 134 (B) Qin graphs 134 Table 4. 4. 2. Replacement of H Y grapheme (A) Chu graphs 134 (B) Qin graphs 134 Table 4. 4. 3. Transformation o f H Y to SP (A) Chu graphs 135 (B) Qin graphs 136 Table 4 . 4 . 4 . Augmentation o f H Y with extra SE (O) Both Chu and Qin graphs 136 (A) Chu graphs 137 (B) Qin graphs 137 Table 4. 4. 5. Augmentation o f H Y with grapheme o f unknown function (O) Both Chu and Qin graphs 138 (A) Chu graphs 138 (B) Qin graphs 138 Table 4. 5. Divergent evolutionary changes from and misrecognition of earlier forms 138 Table 4. 6. 1. Cases in which ancestors cannot be determined 141 Table 4. 6. 2 Cases o f unclassifiable graphs and modifications 142 Table 4. 7. Discrepancies existing in the earlier times 142 v i i List of Abbreviations Used in the Text H Y huiyi "semantic compound" (huiyi J W B Jinwen bian ^tjcWs J W G Jinwen gulin 'jkjCifcJfa J G W G Jiaguwen gulin ^ •jf' %. O B G oracle-bone graph O B I oracle-bone inscription(s) (Jiaguwen ^ # jc) P E phonetic element S A Spring and Autumn (Chunqiu :#-$0 SE semantic element SP semanto-phonetic (xingsheng $ l£) S W Shuowen jiezi 1$L%f%'¥ W S Warring States (Zhanguo |fc EH) w z Western Zhou (Xizhou^M) W Z G Western Zhou bronze graph Notational Conventions Italic : the Chinese pronunciation of a character or word. * : indicates a reconstructed O l d Chinese pronunciation. ' ' : encloses a word, and a translation o f the immediately preceding Chinese expression. « » : encloses a certain object the preceding graph originally depicted. " " : encloses a direct quote, and a word or phrase which is emphasized. v i i i Acknowledgment During my efforts to complete this thesis, which have been taken three years and ten months, I have received continuous direct and indirect support from a number o f individuals and several institutions. To all o f them I express my gratitude and appreciation. Professor Ken ' i ch i Takashima, my thesis advisor, has not only taught me the subject of palaeography, but has disciplined the fundamental attitude towards my studies, emphasizing thoroughness, carefulness and conscience. In addition to giving me generous permission to use his computer facilities, he has allowed me to use for my table his drawings of the oracle-bone graphs in his concordance, Kokotsumoji jishaku soran ? •$X%-!¥(edited with Dr . Matsumaru). For this, I am particularly grateful. Professor Edwin Pulleyblank kindly and patiently read my draft and gave me several crucial suggestions and opinions fundamental to the study of palaeography. Especially valuable was his incisive criticisms o f my faults in logic and in phonology, which have considerably helped improve the content o f my thesis: I thank him. I thank Professor Ross K i n g , who made a number o f suggestions and revisions to improve my thesis. His feedback and support helped me keep going. I also express my gratitude and respect to all the scholars, classical and modern, whose profound achievements have enabled me to pursue my humble goal. I would l ike to express my gratitude to the librarians at the Asian Library and other U B C libraries, and the secretaries in the Department o f As ian Studies, all o f whom always showed willingness to help my research and graduate life. Thanks also go to many others, who offered constant help and encouragement: M r . Zhao, Professor Yeh , Miss Suzuki, Miss Tokumoto and others, near and afar. Lastly, I would l ike to thank my husband, A l e x Shetsen, for his understanding, encouragement, and help in proofreading, given to me unfailingly al l through the last five years. ix Introduction This thesis analyzes Chinese characters used during the W a n i n g States period (5th century B C - 2 2 1 B C ) 1 i n two geographical areas: (1) the vicinity o f the ancient Chu (ft) state area, including several smaller political entities, such as Cai and Huang and (2) the Q i n # state. The primary focus w i l l be placed on the Chu area characters found on the Baoshan % iii bamboo slips that were excavated in Hubei Mit Province in 1986.2 The writing o f the Warring States period has collectively been called the characters of the Warr ing States (Zhanguo wenzi l&MX^r)- One o f the most notable epigraphic characteristics o f that time is that the scripts varied locally. During the Latter Han % era, X u Shen fJI in his postface to the Shuowen jiezi ifaXffl^- (100 A D ) described the variations in the script before the Qin period as follows. 3 After the country was broken up into seven states, farming fields were measured using various units, different widths between the wheels were used, different laws and ordinances were put into force; different styles o f clothes and hats were worn, words varied in pronunciation, and graphs varied in form. 4 After the unification o f the writing system imposed by the Founding Emperor of Q in (Qin shi huangdi Jl '$•), the non-Qin varieties previously current in other states (liuguo wenzi 1 The beginning of the Warring States period has been variously set by several old books of historical records: 481 BC (the 39th year of King Jing $c of the Zhou) by Lii Zuqian g 4 ^ in Dashiji A^ I E ; 475 BC (the first year of King Yuan % of the Zhou) by Shiji jfe. IE: Liuguo nianbiao-fc US 468 BC (the first year of King Zhending |i %i of the Zhou) by Lin Chunpu #•:#•># in Zhanguo biannian ffc H $&^; 403 BC (23rd year of King Weilie fa fi of the Zhou) by Sima Guang ^ mZizhi tongjian it'JaM.$t, according to Cihai^%. In this thesis, "5th century" is used for the beginning of the WS period. 2 Hubei Provincial Jingmen-Shashi Railway Archaeological Team 1991, 5. 3 Hereafter quoted as SW. 4 This translation is based on the Japanese translation of the postface to the SW, by Fukumoto Masakazu H ^ H - (1977,33). 1 7\ HI X "the Characters o f the Six States") vanished. 5 The generations born during Han times are said to have lost the ability to read them ( L i Xueqin 1990, 138). Unt i l the middle of the twentieth century, therefore, and the excavation o f a number o f objects that were found to bear inscriptions dating from the period o f the Warring States, tradition held that varied sets of writing had existed at that time, all o f which, except that o f the Qin , had been lost. Today, as more and more materials that carry writing o f the Warring States period are being unearthed, the characters of the Warring States, period are being intensively studied. Essentially, two features have been acknowledged by most scholars: (1) considerable regional disparities can be clearly observed; (2) the characters of all areas had ultimately descended from the Shang period, through an intermediate stage found i n the Western Zhou (Xizhou WJH) script. This thesis pursues two objectives: (1) a detailed and encompassing presentation of the Warring States graphs from the Chu and Q i n areas, and o f the graphical discrepancies between the two regional scripts during the Warring States period, in the context o f the antecedent historical development o f Chinese writing., The reasons for these graphical discrepancies are beyond the scope o f this thesis to examine in detail, but w i l l be discussed briefly when they are explicitly discernible, or when necessary for data analysis. The Chu and the Q i n graphs are appropriate choices o f topic, for they represent the greatest contrast between the regional scripts that existed in the Warring States period. The Qin graphs were the basis for the line o f Chinese writing, unbroken from the Shang dynasty to the present time, since they were the exclusive single set o f state writing used after the Qin unification o f China (221 B C ) . The Q i n graphs maintained the strongest Western Zhou graphical tradition during the Warring States period. Compared with the Q i n graphs, the graphs of the remaining states, the six major Eastern states, or liuguo 7\ HI 's ix states', which 5 Liuguo 7\ S refers to six states (Qi Chu ft, Yan #j, Han Wei f|, Zhao j$Q which are also called "the Eastern States", as opposed to Qin which was located in the west. In the Warring States period, these seven states were called qi xiong 4j$| "the Seven Powerful States". 2 Wang Guowei iE S $1 ([1921] 1956, vol . 7, 79) called collectively liuguo wenzi 7\ 'six-states graphs', appear to have diverged more from their antecedents.6 The Chu graphs, which were used by one o f the most powerful states o f the liuguo, and the state most abundant in excavated materials, is representative o f these liuguo wenzi. The Chu and the Qin graphs, therefore, constitute the most appropriate set o f the Warring States graphs for the investigation of the discrepancies between the different regional scripts. The regional graphical divergences during the Warring States period w i l l be illustrated through a comparison o f these two sets o f graphs. To this end, a table has been constructed in which the individual Chu and Qin graphs are compared in a one-to-one manner. The same table also exhibits their common antecedents, the earliest known graphs, as found on the Shang oracle-bone inscriptions and the Western Zhou bronzes, so as to demonstrate the antecedent-descendent relationship between the two sets o f regional graphs. Chapter 1 is an examination o f the fundamental nature o f Chinese writing and the historical development o f guwen (the old Chinese script prior to the Q i n Unification) as studied by Chinese and Western scholars, and attempts to set the bases for the analyses of the structural discrepancies between the two regional graphs. Chapter 2 presents the materials and methodology employed. Chapter 3 attempts to characterize the simple graphs of the Chu and Qin . Chapter 4 discusses the differences between the multi-element graphs of the Chu and Qin. The last chapter, Chapter 5, is a conclusion i n which I also consider the implications o f my analyses. 6 Already in the middle of 20th century, when excavated materials carrying the Warring State graphs were still scarce, the prominent Chinese scholar Wang Guowei (1877-1927) observed the distinction between the Qin writing and that of the other eastern states, the liuguo A IS 'six states'. Tang Lan MM ([1949] 1963), as presented later in § 1.3.1, divided the writing of the Warring States into the Qin and the eastern states graphs, 3 Chapter 1 Terminology and Literature Review The two sets o f graphs used in the Chu and Qin regions in the Warring States period belong to the category of the old graphs, commonly referred to as guwen, current in China's ancient era before the Qin Unification (221 B C ) . 7 During this period, Chinese graphs were vigorously formed, and modified rapidly. Given the scarcity o f extant materials and their remoteness from out time, a full picture o f this development is yet wanting. To gain the proper context for a study o f the variations observable in Warring States writing, we must first attempt to understand the basic principles by which the Chinese script was formed, and the resulting internal structure of multi-element graphs and then the principles by which the evolution o f the script resulted in the divergences observable in the Warring States period. After the terms are defined (Section 1. 1), we w i l l first review four theoretical studies of the formation and development o f Chinese graphs (Section 1. 2). O f the two modern formulations of Chinese graphical formation, the first was presented by Peter A . Boodberg (1937), and elaborated upon by W i l l i a m G . Bol tz (1994), and the second was espoused by Chen Mengjia (1956) and Q i u X i g u i : £ (1988 , 2000) who provided empirical support for his notions o f early graphical formation and development. We wi l l then present a historical treatment of Chinese graphs (Section 1. 3), concentrating on the following issues: (1) The chronological division of guwen from the Shang to the Warring States period adopted in our table, with a brief description of scriptural characteristics during each phase and o f the materials from each period used in this thesis; (2) The graphical evolution during the guwen period, and the graphical variation during the W S period. We conclude (Section 1. 4) with a particular view o f the development o f Chinese 7 Actually, the period and type of the script referred to by guwen or guwenzi varies slightly among scholars. In this thesis, guwen is used to refer to the graphs which were used before the Qin Script Reform, and the "guwen period" corresponds to the Pre-Qin period. 4 writing during the guwen period that has been adopted for the work presented in this thesis. 1.1. Terminology We begin with X u Shen's classification o f Chinese graphs, liushu j\ # 'the six types o f script', the origin of several terms used in this thesis.8 A series of definitions for the terms that w i l l be used follows. When a term is used mostly in studies published in a Western language, it is given in English with a corresponding Chinese term; when a term is used mostly in studies published in Chinese, it is given in Chinese with an English translation or its equivalent. 1.1.1. Liushu 7 \# 'the six types of script'9 \)Zhishi ^jf 3ft 'indicates things' 2)Xiangxing 'imitates form' 2>)Xingsheng 'form and sound' characters which symbolizes concepts or objects not depictable by a simple picture. Jl shang 'up' , "F- xia ' down ' 1 0 characters that picture objects fl yue 'the moon' , 0 ri 'the sun' characters composed o f a graph which serves as semantic and a graph which serves as phonetic indicator: jpf he, %L ji&ng ' river' , This liushu 7n # 'the six types of script' appears in the postface to the SW. Traditionally, Chinese characters have been analyzed as belonging to liushu. 9 Xu Shen's classification exclusively dominated Chinese palaeographical studies from the time it was promulgated in the first century AD to the early twentieth century. While still exerting great influence, his categories have recently been questioned and criticized. Zhuanzhu and jiajie exhibit character usage that should be accounted for separately from the other four classes, which are concerned with the internal structure of graphs. In particular, zhuanzhu has been redefined by many scholars (for example, Serruys (1957), see footnote 12), but has also been recently contested, as by Chao Yuen Ren (1968) and Qiu Xigui (1988). Whether the huiyi category is counterfactual has been debated: several of theories will be discussed in § 1.2.3. The definitions of xingsheng and huiyi graphs adopted in this thesis are given in the ensuing section. 1 0 The pronunciation is given in the modern standard Chinese pronunciation alone, unless an earlier reconstructed pronunciation is required for discussion. Li Fang-kuei's #^F# ([1971] 1980) reconstruction is used if the Old Chinese pronunciation is required for discussion. The language of the Warring States period was Old Chinese. 5 4) Huiyi itM 'combines meanings' 1 1 5) Zhuanzhu ^ y ± 'develops meaning' 6) Jiajie -fg.# ' loan' characters formed by combination, with meanings o f the constituents adding up to provide the meaning of the whole. ^ wu 'martial ' , it xin 'confidence' characters grouped by related meaning. # M o , £ lao ' o ld ' characters supplied for another word o f identical or similar sound 4^ ling 'to order'; ling 'order, command', 1 -ft: chdng ' long ' ;zAang 'elder' 1.1. 2. Terms used in this thesis Graph, character, wenzi zi a unit written as an independent entity representing a word in the context in which the graph appears. Simple graph, single-element graph, dutizi # If ^ : A graph that cannot be analyzed into subunits that can stand as independent graphs. A single graph can be used to form a compound or multi-element graph, functioning as a phonetic or a signific. 1 4 Multi-element graph or compound graph/character, hetizi ^ft^: a graph that is composed o f more than one grapheme, such as a semanto-phonetic graph or huiyi graph. A 1 1 Tang Lan ([1949] 1963, 71) denounced Xu's definition of huiyi principle as bi lei heyiyijianzhi hui thU \sJLMt$it$b 'to represent what [the graph] is designed to refer to, by juxtaposing two elements and combining their meanings' (Translation is based on Fukumoto (1977)). Tang asserted that no Pre-Qin graphs were formed according to such principle, explained by Xu with the example wu ^ denoting "valiant' by suggesting 'to stop armed force' with ge 'dagger-axe' and jh zhi 'to stop'. Only graphs postdating the Warring States period have been thus elucidated. 1 2 Because Xu's definition of zhuanzhu was totally obscure, numerous scholars have tried to redefine it. Having examined previous studies aa zhuanzhu, Serruys (1957, 151-152) proposes that zhuanzhu graphs consists of a pair of graphs with a common graphic element that are similar in sound and one of which is "a natural 'continuation or extension', a semantic 'development' of the other": for example, kao 4f" and lao The translation of this term is made by me, based on the interpretations of Atsuji Tetsuji HxtHf (1985). 1 3 These two characters given in the SW have been denounced by scholars, such as Tang ([1949] 1963, 72), Qiu (2000, 161) or Chen Shihui Bfc-irjl? and Tang Yuhui (1988, 54), as incorrect examples not actually used as phonetic loans. For example, the graph 4s which was used to write ling 'to order' was not used as a loan for the word ling 'order', but was employed to write two etymologically related meanings. 1 4 As we will examine in Chapter 4, there are constituents whose functions are unknown to us. 6 multi-element graph can also be used to form a compound or multi-element graph, functioning as a phonetic or a signifie. Element: refers to part o f a simple graph, or a simple graph within a multi-element graph. Constituent: refers to a simple graph within a multi-element graph. Grapheme: refers to a part o f a multi-element graph. It is either a unit which can stand as a graph or that which originally depicted an independent concrete object but which could not stand alone as a graph. 1 5 Phonetic, phonetic element (PE), y/n/w shengfu ^ $ F : a constituent that carries the sound of the word represented by the graph o f which the given constituent is a part. Signifie, semantic element (SE), yifu %$,yifu $,$': a constituent that carries the semantic category to which the meaning o f the word (represented by the graph o f which the given constituent is a part) belongs. Determiner, determinative: a graph added to another graph that represents several words to specify the word intended. It can be phonetic or signifie. Xiangxingzi "Pictograph": a graph that depicts the visual shape o f the object. Semanto-phonetic (SP) graph, xingshengzi ffi graph that includes both (1) an element indicating the sound of the word and (2) an element indicating the semantic category o f the word. Its meaning is carried by the whole, not the phonetic or signifie alone. "Ideographic" graph : A s applied to Chinese characters, originally referred to a graph that was believed to render an object or thing, not the word, to the mind o f the reader.16 Later it has been used to refer to a graph whose constituent(s) is (are all) related to the meaning o f the word represented, for example, huiyi graph. 1 7 Huiyizi itM%-, huiyi ( H Y ) graph, "semantic compound": a graph composed o f several semantic elements that denote the meaning of the word represented.18 1 5 The definition of a grapheme is also presented in § 3.1. 1 6 This concept had prevailed in the West in the 18th century, but was later denounced, originally by Du Ponceau (1838). 1 7 Herrlee Glessner Creel (1936) used "ideographic" to refer to huiyi graphs, the constituents of which do not indicate the sound of the word represented. 1 8 In some cases of this class, constituent parts may not be able to stand alone as a simple graph. For example, an upper object which is held by the hand in the graph f or \ (shi ^ 'to serve') was not used as a 7 Biaoyizi "semantograph": a graph the form of which is related to the meaning of the word represented. Polyphony, yixing duoyong "use one graph for many [words]": the phenomenon o f a single graph being used for several words related in meaning but different in sound. Polysemy: (1) the phenomenon of a single graph being used to express several words unrelated in meaning but identical or nearly identical in sound, namely phonetic loan. In Chinese this phenomenon is associated with the usage o f a loan graph, jiajie i&'fik "loan". 1 9 (2) the phenomenon o f a single graph being adopted from one word to another that is semantically related and has similar pronunciation to it. Paronomastic writing, the rebus, rebus writing, jiajie M ' f t : the usage of a graph representing one word to write another word that does not have a written form. Typical ly both words are (nearly) homophonous, but not related semantically. Zhouwen refers to a form of Chinese script traditionally believed to have been compiled in the fifteen-volume Shi Zhou pian jfc. If ^ (now lost) by the W Z official Zhou $ | during K i n g Xuan's (%) reign (827-782 B C ) . S W guwen refers to a form of script which X u Shen included in the SW, traditionally believed to have been taken from the books; immured i n the walls of the house in which Confucius had lived. The term "SWguwen" is used in this thesis to distinguish this script from guwen or Pre-Qin writing in general. Xiaozhuan /\s | £ "small seal": refers to the formal script form as used during the Qin dynasty. 1.2. The earliest process for formation of the Chinese script 1. 2.1. The three stages of the early graphical development (Boodberg and Boltz) Peter A . Boodberg ([1937] 1979, 363-406) proposed a systematic formulation of the earliest stages in the formation and development o f the Chinese script, simple graph. 1 9 Traditionally, the term jiajie has also been used to refer to a usage of graphs in which one graph is used to write another homophonous word that has its own writing form. In this thesis, the term jiajie only refers to usage in which one graph is employed to write another homophonous word that does not have its own graph. 8 Boodberg's study illuminated the relation between graph and sound. He speculated that the true moment o f genesis of the writing system, given a period o f conventional and habitual association between graphs on the one hand, and their associated semantic and phonetic values on the other, the graphs conjoin to become elements in a written language. When this association has been firmly cemented by long-term usage, Boodberg asserted, the conventional associations o f graph and phoneme are so tight that their abandonment is difficult, although not necessarily impossible. 2 0 A s is characteristic as " i n all logographic writing, particularly in its first pictographic-symbolic stage" (Boodberg [1937] 1979, 367), the usage o f graphs was enlarged as follows. 1. Graph A stands for basic Semanteme (S) 1: Phoneme (P) 1. E.g., ' two' 2. Graph A stands for dialectical, morphological derivative, S1:P2. E.g., ' twain' 3. Graph A stands for a cognate word, S2:P2. E.g., ' twin ' 4. Graph A stands for S2:P1. (the rebus usage) E.g., ' too' 5. Graph B stands for two completely different words due to its graphic ambiguity. E.g., 'head' (S1:P1) and 'hair ' (S2:P2) 6. Graph A and Graph C both stand for S1 :P 1.21 The third step of the graphical formation, he asserted, proceeded as follows. 1. Graph A (S1 :P 1) remains as it had been. 2. Graph A ( S l :P2) + Graph E (phonetic indicator) = semanto-phonetic graph 3. Graph A (S2:P2) + Graph F (phonetic indicator) = semanto-phonetic graph 4. Graph A (S2:P1) + Graph G (semantic indicator) = semanto-phonetic graph 5. Graph A stands for the original word (Sl :P1) ; Graph C becomes associated with another word, or becomes a specialized or technical representation o f the original word(S3:P3). 2 0 Boodberg used the term "phoneme" to refer to a single syllable that coincided with the reading of a single graph. Boodberg did not stipulate the phase of graphical development at which this firm association between graph and semanteme/phoneme is to have been achieved. 2 1 The phenomenon by which a single word had more than one graphic representation has been widely acknowledged (for example, Todo Akiyasu (1965)). The simple cause was that characters were not necessarily created in a single place and time. 9 Boodberg's succinct scheme has furnished us with an essential framework for an account of the early development of Chinese writing; Based on Boodberg's study, W i l l i a m G . Bol tz (1994) has developed an elaborate account o f the early evolution o f the Chinese script. He delineates three stages. (1994, 59-70) Step 1. Graphs representing the names of objects were created by drawing pictures o f the objects. Each graph was associated with its pronunciation and meaning. Undepictable notions, such as 'to come', remained outside the range o f graphical representation allowed by this method. Step 2. The scope of graphical representation was extended in two directions: (1) paronomastic writing (rebus or jiajie 'fH'ft), and (2) parasemantic writing. (1) In paronomastic writing, a word representing an undepictable idea is written with an existing graph used for a homophonous or near-homophonous word. For example, the graph % xiang 'elephant' was used to represent the abstract homophonous word xiang 'image, apparition'. Such a graph, representing several words phonetically similar but semantically varying, can be called polysemic. (2) In parasemantic writing, an abstraction is rendered with an existing graph whose "depictive quality" is able to suggest the meaning o f the word intended, regardless o f the potential phonetic or cognate relations between the two words. For example, the graph P kou 'mouth' was used to represent the word abstract and semantically related word ming 'name, to ca l l ' . 2 2 Such a graph as D , representing several words semantically related but phonetically varying, can be termed polyphonic. While these two "multivalent" usages of established pictographs are efficient in making easier the enormous task of memorizing a large number of unrelated graphs, the one-to-many relationships are inherently ambiguous. Step 3. This shortcoming was solved by adding another existing graph to specify the word intended. In this way were formed the so-called semanto-phonetic graphs (xingshengzi 2 2 It should be noted that, as will be discussed in § 1.2.4.3, there is no direct evidence that the graph kou D was used for the word ming 'to call, name', and that this usage of the graph kou D has remained hypothetical. 10 (1) In the case o f the paronomastic usage (jiajie) o f the graph used for both xiang 'elephant' and xiang ' image' , the word intended can be elucidated by the addition of the required semantic value. For ' image' , the graph A ren 'man ' was added to specify its meaning. 2 3 This type o f attached graph is called a semantic determinative. (2) I f a graph is used parasemantically, the specification o f the phonetic values is more effective than that o f the semantic, for the meanings o f two words may be too nearly congruent. For example, the graph D was used for kou 'mouth' and ming 'to call, name'. To specify the usage ming 'to call , name' the graph & representing the word ming 'dark' was added, eliminating the possible representation for the word kou 'mouth ' . 2 4 Boltz calls this added graph a phonetic determinative. Traditionally, Bol tz (1994, 63-64) states, paronomastic writing (jiajie) alone had been recognized as a precept for the development of Chinese script. Parasemantic writing, another important principle, had been overlooked, but "it appears to have been widespread in the formative stages of waiting in China" (Boltz 1994, 64). 1. 2.2. The three stages of the early graphical development (Chen and Qiu) W e shall examine the other formulation o f three-stage development, as proposed by Chen Mengjia (1956, 77-80) and refined by Q i u X i g u i (1988, 2000). Chen delineated the development o f the Chinese script in the following terms: X.Xiangxingzi "pictographs", 2. Jiajiezi lUMlr ^ "phonetic-loan graphs" 3. Xingshengzi"pictophoneticgraphs"25 2 3 Although the semantic relation between the graph ren A and the word 'image' cannot be definitively elucidated, this graph can be considered a semantic determiner, because the addition of this graph distinguished the meaning intended from the other. 2 4 Boodberg ([1937] 1979, 376) pointed out that the SW implied the reading of the graph *> as the sameas^ wmg'dark' in "cong kdu congxi xizhe mingys %k U.$t&^^^-tfe" which Boltz (1994,) interprets that the graph was "derived from 'mouth' and from & "as for fr, it is ming < *ming 'dark'" (Boltz 1994, 105). 2 5 "Pictophonetic" is used for the English translation of Chen Mengjia's term "xingsheng graph". Chen used the term "pictograph" for the constituents of the xingsheng graphs. 11 Xiangxingzi H^fy^-, the 'pictographs', came into existence first. The need to record concepts inexpressible by shape alone was satisfied in two ways: (1) by using the pictograph for another word which was "semantically extended" (yinshen ? | # ) 2 6 from the word the graph represented, such as H ri 'sun' which was used to write 'day'; and (2) by using the pictograph for another word which was pronounced identically or similarly (jiajie). Ambiguities were created, however, when many such jiajiezi "phonetic loan graphs", came into use, especially when a single graph was borrowed for more than one word. Xingshengzi or "pictophonetic graphs" were therefore devised by juxtaposing in a single graph a phonetic loan graph with another pictograph that acted as the "signific" for the word being expressed. Originally, xingshengzi "pictophonetic graphs" were used to distinguish the multiple uses of a single graph, but later they came to be the main principle for the invention of new graphs. Q iu X i g u i ' s account (1988, 2000) o f the early development o f the Chinese script is more detailed, supported by his close analysis o f the graphs. Qiu ' s conception o f the development can also be divided into three stages. 1. Pictorial graphs were created first. He proposes to use the term biaoyizi 'semantographs' to include all the graphs, the constituents o f which implicit ly relate to and convey the meanings o f the words represented, — zhishi ^ jf , xiangxing fUffr, and huiyi t o f theSW(1988 , l l ) . 2 7 2. The second stage o f development (1988,4-7; 154): (1) Jiajie 'fH'ft (paronomastic writing): A n existing graph was used to write a homophonous or near-homophonous word which could not be expressed by drawing pictures. 2 8 Qiu, unlike Chen, does not stipulate the usage o f the graph for another word whose 2 6 The Chinese term "yinshen" can be literally translated "semantically extended". Whether or not the relationship between the two words or meanings that are described by the term yinshen is actually "semantic extension", however, is debatable. We will discuss this usage of graphs and attempt to determine the meaning of this terminology in § 1.2.4.1. 2 7 Qiu does not explicitly include multi-element biaoyizi in this stage, but he gives jft she 'to shoot with bow' as one of examples of this stage (Qiu 2000,4). 2 8 Qiu surmises that phonetic loan graphs (jiajie) appeared soon after the creation of biaoyizi, or indeed simultaneously (1988,4-5). 12 meaning is related (yinshen) to the word the graph represents together with the jiajie as the second stage o f early graphical development (1988, 1-8). In the third stage o f graphical development (the semanto-phonetic graph formation), however, he distinguishes between jiajie graphs and graphs used to write etymologically related words (1988, 154). It can therefore be understood that he holds the usage o f graphs for etymologically related words separate from jiajie at the second stage o f the early graphical development. (2) The usage o f the graph for the word whose meaning was "extended" (yinsheri) from the word the graph originally represents: A single graph became attached to another word somehow derived from the original word the graph represented29; for example, q& 'to take' which was used for a word #w,'to take a wife ' . (3) Yixing duoyong — fyfyffi 'use o f a graph for many [words]': A single graph represented several words in meaning related to the form o f the graph but differing in pronunciation (1988, 5). This usage is identical to Bol tz ' s "parasemantic writ ing". For example, the oracle-bone graph j\ (^ c) was earlier used to write the two words Jn 'adult male' and da ' b ig ' . During the early stage o f development, in some of these cases, for example, j\ which was used to write two wordsJu 'adult male' and da ' b ig ' , and ]) which was used to writeyue 'moon ' and xi 'evening', different forms o f the graph were used to resolve differences in meanings, as ^ ju 'adult male' and j\ da ' b ig ' ,}yue 'moon' and ]) xi 'evening' (1988, 5-7). This is calledfenhua ^M-fc 'differentiation'. Z.Xingshengzi ^ "semanto-phonetic graphs" are formed by the addition o f a semantic determiner to graphs used in jiajie and those used to write etymologically related words (yinshen), o f a phonetic element to a biaoyizi (to indicate the reading), or of a semantic element to a semantograph (to clarify or emphasize its meaning). Although Qiu surmises that semanto-phonetic graphs could also be formed by adding phonetic elements to yixing duoyong — graphs, or parasemantic writing, as in the graph # bi 'nose', he does not consider this path to the formation of xingsheng graphs to have been major. 2 9 This usage will be discussed in § 1.2.4.1. 13 4. Huiyi itM%-In addition to the formation of semanto-phonetic graphs, Qiu (1988,122-139) stipulates another multi-element graph formation principle, huiyi, which renders a meaning produced by multiple semantic elements.3 0 The four major studies o f the graphical development during the guwen period fundamentally agree that early development o f Chinese script consists o f three principal stages. The last two stages can also be considered as a 2-step process o f SP graph formation. Stage 1: Creation of pictorial graphs Stage 2: Multivalent usage o f pictorial graphs: first step o f SP graph formation (Step 1) Stage 3: Formation o f SP graphs by adding another graph: second step o f SP graph formation (Step 2) There are three points on which the four scholars do not completely agree: (1) the existence o f huiyi as a formation principle stipulated by Chen, Q i u and many other scholars was not accepted by Boodberg and Boltz, although Boltz admits the difficulty of its verification (see Section 1.2.3); (2) the existence o f "parasemantic writ ing" ("polyphony"), as advocated by Boodberg and Boltz , has not been incorporated into Chen and Qiu ' s processes o f graphical development, probably because o f the latter's strictly empirical approach to the subject (see Section 1.2.4.3); (3) Boltz does not stipulate in the stage 2 (multivalent usage of graphs) the usage o f a graph for an etymologically related word (traditionally termed a "yinshen" word by Chinese scholars) without a written form o f its own (see Section 1.2.4.1). In the ensuing sections, "Stage 2" and "Stage 3" refer to the last two stages in the three stages o f the entire early development o f Chinese script, and "Step 1" and "Step 2" refer to the 2-step process o f SP formation. Since the compound graphs are the principal material used in this thesis to investigate the nature o f structural discrepancies between the Chu and Q i n graphs, the following two 30 Huiyizi will be discussed in § 1.2.3. 14 sections examine in detail the huiyi ( H Y ) and the semanto-phonetic (SP) graphs. 1.2.3. The features of H Y graph or "ideographic" compound graph This section w i l l present: 1) the existence o f H Y as a multi-element graph formation method; 2) the principles o f the H Y graph formation. 1.2.3.1. The existence of H Y as a multi-element graph formation principle This thesis accepts huiyi as a principle of multi-element graph formation despite the difficulties inherent in this notion. Herrlee Glessner Creel (1936, 98-104), a strong advocate o f the "ideographic" nature o f Chinese writing, strongly believed in the existence of H Y ; Peter A . Boodberg (1979, 378-383) posed important counter-arguments. Creel gave the following examples o f H Y graph formation. 1. The graph jt£ xian was formed by combining two genetically similar graphs, & yu ' f ish ' and ydng 'sheep', to represent the word 'fresh'. 2. The graph jft hdolhao was formed by combining ;two graphs nfl 'woman' with -f- zi ' ch i ld ' , using the mother-child relationship to render 'to love, to like; good'. 3. The graph^. dong 'east' can be analyzed as the sun ri 0 positioned behind a tree mu to indicate its direction. Though Creel 's analysis o f semantic compounding in these three cases appears reasonable, two of his examples can be shown to be false. That xian ^ is a H Y graph was denied by Boodberg ([1937] 1979, 383) who quoted the S W ' s elucidation that the right element ^ *rang(*zjang) 'sheep' in the graph ^ *sjan (*s;an) 'fresh' is an abbreviated form of the graph j^j? *shan (*sian) 'smell o f sheeps' which serves as phonetic element.3 1 The H Y status of the graph $L dong 'east', which Creel quoted from the S W , has been invalidated by the form of the oracle-bone graph $ , which has been deciphered as $L dong 'east'. 3 2 3 1 In this section, Old Chinese pronunciations reconstructed by Li Fang-kuei are provided, followed by Boodbergs' in parentheses. 3 2 The shape of the oracle bone graph does not appear to represent the sun behind a tree. The current 15 Boodberg ([1937] 1979, 379-383) argued that the strength of the established association between a graph on one hand, and a semanteme and phoneme on the other, should be so strong that to create a new compound graph from two established graphs and to assign this new graph to a new semanteme and phoneme completely different from those of the components would be a violation of the fundamental laws of economy inherent in the development of any script. He asserted that the formation o f "ideographic" graphs (that is, the combination o f two or more existing graphs with a reassignment o f meaning and sound) cannot exist as a linguistic category:3 3 O n the contrary, Boodberg ([1937] 1979, 378-379) proposed that knowledge o f polyphony allows the identification of one of the components as a phonetic: for example, the graph Vfi *mjiang (*miwDng) 'bright', supposedly an "ideographic" compound, contains a phonetic element H ri 'sun' that as a polyphone represents the etymonic sound for 'bright' related to but distinct from the known reading *njit (*rjiiet) ' sun' . 3 4 These arguments render the status o f H Y graphs precarious. To classify a compound graph as H Y hinges on whether or not a constituent phonetic element can be found; the connection o f the meaning represented individually by every constituent to the meaning of the word represented by the compound graph tends to obscure a phonetic constituent. For two reasons, however, to accept the existence of H Y as a principle o f formation would appear legitimate: (1) proposed forgotten alternate readings o f graphs which might be serving as phonetic elements in H Y graphs remain hypothetical; (2) many graphs without any apparent phonetic element can be elucidated by the H Y principle. Q iu uses this formulation, examined in the next section. In the treatment this thesis gives to multi-element graphs, therefore, it w i l l often be the case that a given graph w i l l contain no grapheme related to its reading. explanation is that $, originally used as pictograph for the word "bag" tuo %, by representing the shape of a bag with both ends tied, was adopted for the word "east" as a phonetic loan (jiajie). Ding Shan T LU (JGWG V.4: 3010-3011) introduced Xu Zhongshu's ^"f $T explanation that the OBI graph $ dong 'east' was the original graph for ^ tuo 'bag'. Ding explained that graph % tuo (j& *tagh) 'bag', which had the same initials as dchg(%.$ *tung) 'east' (tuoyu dong weishuangsheng^&JUfa^tQ), was aphonetic loan graph for 'east'. 3 3 Boltz (1994, 64-66, 71-72), too, insists that there existed no compound graphs without phonetic element (the classic category huiyi), and that all so-called HY graphs include a phonetic the sound of which has been forgotten. 3 4 Boodberg's explanation of the graph ming ty remains hypothetical because the alternative reading *mjiang (*miwong) of the graph ri B has not been verified. 16 1.2.3.2. The principles of the H Y graph formation W e shall examine several Qiu 's empirical principles for the formation o f huiyi (Qiu 1988, 122-139). The examples he provides for H Y graphs appear to support the existence of this category. (1) Combination o f graphs to depict two objects rendering the action or situation described, undertake', and ( # gong; T: an object; B: a pair of hands) 3 6 'together'. (2) Combination o f graphs with the relative positions o f the components conveying the meaning intended, e.g., ^ ( i t zhi; T: foot; B: a line which is the position where one remains) ' to stop' (3) Combination of a man or animal with a certain organ o f the body, to render the organ's function, e.g., f f ( J i wang; T: an eye; B: a standing figure) 'to look at something distant', and ^ ( ^ xiuT: nose; B: dog) 'to sniff . (4) Duplication o f elements to denote a collection (in the mathematical sense o f the word), e.g., / jd \ (#• lin; L: tree; R: tree) 'forest' (5) A combination of two (or occasionally more) graphs as an inference o f a more general meaning, e.g., $j lie 'inferior' which is rendered by 'J? shao ' l ittle' and jj li 'strength'. 3 7 (6) A combination o f graphs which cannot be categorized into any o f the above. For example, sao # means 'to sweep' and is symbolized by a hand carrying a whisk. In this case, because the sounds o f #• sao 'whisk ' and ,# sao 'to clean' are very similar, the graph # can be regarded as huiyi jian xingsheng itM$fc3&$£ '[being] H Y and SP [at the same time]'. This thesis w i l l treat such graphs as SP . 3 8 3 5 L stands for the left element of the graph, and R for the right element of the graph. 3 6 T stands for the top element of the graph, and B for the bottom element of the graph. 3 7 This graph appear to be formed, however, after the Warring States period (Qiu 2000, 54). 38 Huiyi jian xingsheng '[being] HY and SP [at the same time]' refers to a phenomenon by which the meaning of the word expressed by the graph can be elucidated based on the semantic roles of constituent graphemes with at least one of them also acting phonetically. Qiu categorizes this example as one type of HY graphs, but also acknowledges it as SP graph. Others such as Chen and Tang (1988, 46) recognize them as SP (Sp ji ; L : food served in a container; R: a kneeling person) 3 5 'to engage in, graphs. 17 1. 2.4. The features of xingsheng (SP) graphs The following features of semanto-phonetic graph formation are relevant for the analysis of graphical discrepancies to be presented in this thesis. 1) Jiajie and "yinshen"; 2) the principles of the SP graph formation; 3) alternate readings of graphemes; 4) multi-element PEs and SEs; 5) the position of a PE and an SE within a single graph. 1. 2.4.1. The jiajie and "yinshen" As mentioned in Section 1. 2. 2, Boltz, unlike Chen and Qiu, does not include in his formulation of early development the following path of SP formation: Step 1: the usage of an established graph representing a given word (A) to write its etymologically related, or ''yinshen" word (B): Step 2: the augmentation of another graph to form a SP graph which was exclusively assigned to the word (B). We shall briefly discuss this SP formation path with Qiu's example. Step 1: f% jie 'to untie' was used for a word 'to relax (psychologically)'39, related meaning of the word 'to untie'. Step 2: The graph ffl was augmented with a semantic determiner f (i^) xw 'heart' to form the SP graphxie % which exclusively represented the word 'to relax'. This example illustrates that the usage of the graph jie %• for the word 'to relax' is not paronomastic usage if the paronomastic usage is defined as the usage of the graph to write another homophonous but semantically unrelated word (such as in the case in which the graph %. xiang 'elephant' was used to write xiang 'image'). This case in Step 1 can be considered as the multivalent usage of the graphs for the semantically or etymologically related words. Since it is difficult to define the semantic relationship between this type of 3 9 The meanings of these words are adopted from Chinese Writing (Qiu 2000, 250) in which this example appears. 18 paired words, and it is not within the scope of this thesis, we w i l l not use the term yinshen'''' or "semantically extended", but the term "etymologically related" for the relationship between such pairs o f words. In our analysis o f the processes resulting in the structural discrepancies between the Chu and Q i n graphs, cases of (1) paronomastic writing (jiajie) and (2) the usage of the graph described above w i l l be treated together as the first step o f SP graph formation, and an augmentation of semantic determiners to the graphs used in the two principles w i l l be treated together as the second step o f 2-step SP graph formation. The first category occupies the majority o f SP formation cases, but the case which is indisputably the case (2) w i l l be so labelled. 1. 2.4. 2. The principles of the SP graph formation W e w i l l adopt Qiu ' s categories o f semanto-phonetic graph formation, the most exhaustive treatment given to relation between the graph and the word (Qiu 1988, 151-156). The first four operations produce new SP graphs, and the remaining two processes are modifications, more particularly, an augmentation o f phonetic element to existing semantic graphs. (1) A semantic determiner is added to a graph which has been loaned from a homophonous word. This is the primary formation principle stipulated as the third stage (an augmentation of semantic determiners) o f the early graphical development. W yi 'w ing ' + 0 ri 'sun, day' =Q$yi 'tomorrow' (2) A signifie is added to a graph which has been used to write two semantically related words to form the graph which represent one of the two. JU Hang ' a pair'+ jfc si ' s i lk thread' = jgf Hang ' a pair o f [shoes]' (1988, 230) (3) A signifie is added to a semantograph to clarify the meaning o f the word represented.40 qidng «bed» +&.mii 'tree' =jj^. chuang 'bed' 4 0 Theoretically, graphs that consist of a semantograph and a semantic determiner are not SP graphs. The original semantographs were recognized as phonetic elements, after a semantic determiner was added to these graphs, and thus the augmented graphs were taken to be SP graphs. This phenomenon occurred with many semantographs , for example; ca i 'to pick up [from trees]' which was augmented with shou and zhengjL 'to go on an expedition' which was augmented with chi %. 19 (4) A n extra phonetic is added to a graph for the etymologically related word the graph represents. (6) A part o f the semantograph is altered into a phonetic element to indicate its reading. A graph Jr. xiu 'to offer food [in the ancestral temple]' first consisted of ydng 'sheep' and X you «to hold in hand» , but later S chou «to grasp by hand», 'the second o f the twelve Earthly Branches', which was closer to its reading, replaced you X -The first three of these principles of SP graph formation were the most commonly used (Qiu 1988, 156). 1. 2. 4.3. The alternate readings of graphs We shall briefly discuss Boltz 's "parasemantic usage of graphs". Boltz (1994, 63-64) states that parasemantic writing, another important principle, had been overlooked since the presence of such polyphony could be only temporary. For example, the fact that the graph D was used for the word ming 'to call , name' and the association between the graph P and the word ming 'to call , name' sank into obl ivion when it was compounded with the graph & to form the graph % which was exclusively assigned to the word ming to call, name'. It is possible that while the rebus usage o f graphs may have been practiced throughout the history o f the graphical development, the parasemantic usage o f a graph should be considered minor because it can be considered to have occurred for only a limited number o f simple graphs. Furthermore, as Boodberg ([1937] 1979, 367) perceived, parasemantic usage existed only at the earliest stages o f graphical development, when graphs still maintained their pictorial quality. Bol tz (1994, 103-105) has proposed the jiajie representation of words homophonous to an alternate reading of the graph, and, upon addition of a semantic determiner, development into SP. Such alternate readings o f graphs are highly supported when several graphs are found to contain a given simple graph and the readings o f those graphs all belong to the same 20 rhyme group. The following is an example: The graph D represented two words: (1) kou 'mouth'; (2) ming 'name, to ca l l ' . 4 1 The graph P standing for the word ming 'to call , name' was augmented with a phonetic determiner & , 4 2 The graphs D standing for the word ming 'bird-call ' was augmented with a semantic determiner niao The graph P was added to the graph 4" ting 'order, command' which was used to write the word ming 'fate' to specify its usage by adding P as a phonetic determiner. The resulting SP graphs are: % ming'name, to ca l l ' ; p , | ming 'bird-call ' ; $r ming 'fate'. 4 3 Pursuing this line o f investigation may yield more examples o f the parasemantic usage of graphs, and the discovery o f obliterated alternate readings o f graphs might in turn contribute to the reclassification o f some H Y graphs as SP graphs, rather than as compound graphs o f semantic elements alone. It should nevertheless be acknowledged that the proposed alternate readings o f graphs yielded by this process presently remain hypothetical, without concrete evidence. 1. 2.4.4. PEs and SEs consisting of multiple elements It has been noted that SP graphs may contain multiple SEs and/or PEs. Boltz (1994, 70-71) observes that this is the result o f the recursion of Step 2 {jiajie, or paronomastic usage of a graph) and 3 (an addition of a semantic determiner). A n SP graph that had completed its evolution through all three stages could be re-used as a loan graph for another homophonous word and could be augmented with another semantic determiner. However, Boltz (1994, 71) surmises, the multiple constituents o f a given graph sometimes result from "an aesthetic 4 1 That the graph P represented ming 'name, to call' is not attested, and its modern reading is given here as Boltz surmises. 4 2 As for the reading of the graph ft as ming, see footnote 24 above. 4 3 The SW explains ming »,| and ming % as HY graphs. Duan elucidated that ling 4" serves as PE in the graph/n/ng 4fr. The possibility of the alternative reading of the graph D in the meaning 'to utter, name' as mingwas originally proposed by Boodberg ([1937] 1979, 376). 21 sensitivity to graphic balance and elegance". 4 4 Tang Lan ([1949] 1963, 107-108) and Qiu (1988, 157-160) also hold that there should be only one SE and P E in a single character at its formation. In S W , for example, bao 'treasure' had been analyzed as consisting o f the 3 SEs midn ' roof , 5 yu 'jade' and M. bei ' cowry ' , and the 1 phonetic -& fou 'earthern vessel'; $ | bo 'beam connecting pillars' , o f the 4 SEs 7fc mil 'tree', cao 'grass', 5 shui 'water' and ~f cun 'unit of measurement', and the 1 phonetic element, ^ Jn 'to begin'. But Tang and Qiu both insist that the compounding must necessarily have used exactly one of each. In most cases of multiple SEs, Qiu argues, a huiyi graph can be detected. Thus $ bao 'treasure' possessed a single SE that combined the graphs ^ midn ' r oo f , 3. yu 'jade' and K bei ' cowry ' and is attested in the oracle-bone graph Tang sees a series o f distinct compounding steps to yield apparently multiple PEs: Jn ' b ' , as a P E o f ^ . / u 'to spread, throughout (the land)', itself used as a P E in :J$ pn 'to spread, throughout (the land)', which is then used as a P E of 3$ bo 'grass field' , which is used as a P E in $$ bo 'beam connecting pillars' . Thus there is exactly one P E and one SE used at each step. Q iu also speculates that double PEs in a single SP graph occurred in rare cases by repeated addition o f further PEs to an expanding SP graph. 1. 2. 4. 5. The relative positions of a PE and an SE within a single graph The relative positions of P E and SE within a given SP graph have been classified into 8 cases 4 5 1. left SE: right P E yX j&ng '(Yangzi) river' 2. left P E : right SE j% jfn 'dove' 3. top SE: bottom P E Jft cao 'grass' 4. top P E : bottom SE |j» po 'o ld woman' 5. periphery SE: inside P E HI pu ' f ie ld ' 6. periphery P E : inside SE $f yan '(river) to flow' 4 4 Boltz does not, however, provide an example of the use of multiple elements used for aesthetic purposes. 4 5 Tang ([1949] 1963,104) gave the first 6 types , and Qiu (1988, 166) added the last two. Examples for 1- 6 are Tang's, and those for 7 and 8 are Qiu's. 22 7. P E in the corner fang ' room' 8. SE in the corner $ zai 'to load' The "left S E : right P E " type is the most common combination o f phonetic and semantic elements i n a single graph. In analyzing the function o f the constituents o f a compound graph, however, we need not consider their position. 1. 3. The old script (guwen From the Shang to the Warring States period Having gained an understanding o f the processes observable in the formation o f graphs, we w i l l now trace their orthographic evolution during the guwen period, and the media upon which they are found. The following topics w i l l be covered. 1) Chronological divisions of Pre-Qin writing The Shang writing: oracle-bone graphs The Western Zhou writing: bronze graphs Writ ing of the Spring and Autumn period: bronze graphs Writing of the Warring States period: graphs on bronze, bamboo, stone, wood and silk manuscript 2) The graphical evolution of guwen the orthodox and vulgar forms o f the script simplification and elaboration 3) The graphical features of the Warring States graphs 4) The Script Reform at the Qin Unification 5) Formulation of the development oi guwen adopted in this thesis 1.3.1. The chronological division of the Pre-Qin writing (guwen isjQ The approximately 1100 years o f graphical evolution covered in the table o f graph given in Appendix-I have been divided into 4 historical periods based on Tang Lan's historical classification ([1935] 1965, 4-5). Instead o f the traditional classification o f old writing according to the medium on which it appears (for instance, bronze vessel, pottery, seal, and so forth), Tang proposed that guwen should be classified by their forms during the accepted periods of Chinese history. He described three historical phases for the ancient script, with 23 distinctions during the last phase, the Warring States period. His categories are as follows. 1. Shang graphs (Yin-Shangxi wenzi Shang period 2. Graphs of Western and Eastern Zhou (UangZhouxi wenzi Ji] % X ^): from Western Zhou to the end o f the Spring and Autumn period 3. Graphs of the Six States (liuguoxi wenzi js HI % X the Warring States period 4. Graphs of the Qin state (Qinxi wenzi X*¥)'- the Warring States period Despite certain criticisms, Tang's classification has gained widespread acceptance.46 We have adopted it for the construction o f our table, separating the Autumn and Spring period from the Western Zhou period. This division has been introduced to reduce the quantities o f the data from each period to a manageable level, and also because several sample writings from various states appear to indicate discrepancies among the Warring States graphs from the middle o f the Autumn and Spring era onwards (He 1989, 2). The following sections briefly discuss the features o f the writing during each phase, providing the necessary historical background. 1.3.2 The Shang writing The earliest extant samples of Chinese script are graphs incised primarily on turtle plastrons and bovine scapulas.4 7 This writing is usually called jiaguwen f itX 'turtle-shell and bone inscription', and, in English, "oracle-bone inscriptions", for their content chiefly concerns divination. The majority o f such inscriptions have been found at Xiaotun Vil lage ( / j N T f c # ) , Anyang Ci ty ( ; £ | # " r f r ) , Henan Province These writings can be dated between around middle to the end o f thirteenth century B C , the beginning o f K i n g Wuding's reign, and around middle to the end o f eleventh century B C , the fall o f the Shang dynasty. 4 7 Most o f the writings have been incised with a sort of engraver (kedao M 77), but occasionally 4 6 For instance, Qiu (1988), although adopting Tang's classification, observes that characteristics of the script during each period are not strictly confined by dynastic breaks. Characters of the late Shang, for example, are similar to those of the early Western Zhou. Chen and Tang (1988) continue to accept the classification of old writings by the medium of the inscriptions, such as bronze or bamboo slips. 4 7 The beginning of the King Wuding's reign has been dated from 1339 BC (Dong 1945, vol. 2, 2) to approximately 1200 BC (Keightley 1978, 228). The fall of the Shang dynasty has been dated from 1127 BC to 1018 BC (see Shaughnessy 1991,219). Intermediate approximate dates for the beginning and fall of the Shang dynasty are given in this thesis. 24 they were written in ink with a brush. 4 8 About 5000 graphs have been found in total, o f which approximately one-third have been deciphered 4 9 Although conspicuously pictographic, the objects which the graphs depict are not always clear, even when the graphs can be deciphered and transcribed into modern equivalents, e.g., f J? , yu -f- 'to . . ' . Comparison of the early and late oracle-bone graphs shows that evolution abstracted their pictorial features (Qiu 1988, 42-43). Pictographs came to be also used as phonetic loan graphs, and there are multi-element graphs in which one pictographic element indicates sound. To obtain an overall sense o f this writing system, it is useful to introduce some statistics. L i Xiaoding (1968, 1993) has analyzed the 1225 oracle-bone graphs whose form, sound and meaning have been deciphered, and has grouped them according to the traditional Six Categories. Table 1.1. Category of Graphs Number o f Graphs ( % ) Pictographs 276 ( 22.53 % ) Zhishi Graphs 20 ( 1.63 % ) Huiyi Graphs 396 ( 32.33 % ) Phonetic Loan Graphs 129 ( 10.53 % ) Semanto-Phonetic Graphs 334 (27.27 % ) Others (Unknown) 70 ( 5.71 % ) Total 1225 ( 100 % ) A s we can see, the Shang writing had already reached the third stage o f graphical development, namely the formation of SP graphs; but semantic graphs, i.e., pictographs and huiyi graphs, were still dominant, and SP graphs had at that time barely surpassed phonetic 4 8 A small number of graphs on pottery, stone, jade, and horns have been found. 49 Jiaguwen bian f #;>t $t (1965), a major compilation of oracle-bone graphs, lists 4672 graphs (except for hewen "compounded phrase"), although Kokotsumojijishaku soran ? if ^ ^ # ^ IE (1994) points out that more than 70 graphs have multiple entries. Kokotsumojijishaku soran also lists 476 graphs not included in the Jiaguwen bian, although the number of attested graphs is somewhat less than 476, because some of them have been provided by some specialists without actually ascertaining their original sources. The number of known oracle-bone graphs is therefore about 5000. 25 loan graphs in their proportion, though they were later to form the overwhelming majority of graphs.5 0 The vocabulary is limited to specific matters o f divination: important future events, such as sacrificial ceremonies, military campaigns, hunting expeditions, agriculture, sickness, or general queries about disasters during the night or day, and during a ten-day period, etc. (Keightley 1978, 33-35) 5 1 1. 3. 3. The Western Zhou Writ ing The extant writing of the ensuing Western Zhou period (middle to end o f eleventh century -771 B C ) is found primarily on bronze vessels manufactured mostly for ancestral rites of devotion, for the commemoration of deeds, and as imperial gifts from members o f the Royal House, nobles, vassals and officials. These are usually referred to as jinwen 'the bronze-vessel inscriptions' or 'the bronze inscriptions'. 5 2 Although there is no doubt about the genetic relationship between the writing system o f the Shang era and that o f the Western Zhou period, scripts on bronze vessels do not necessarily exhibit traits o f palaeographic evolution more advanced than those found on the oracle-bones.5 3 Western Zhou writing has been generally considered uniform (Guo 1972; Tang 1986), although some graphical variation has been observed (Gao 1987, 146-180). The most notable tendencies o f evolution during this period are not structural but stylistic in nature: xiantiaohua the "linearization" of pictorial execution, and pingzhihua - f - j l ^ t , the "streamlining" o f curvilinear lines of graphs (Qiu 1988,46). 5 4 It must be noted that the Western Zhou bronzes attest only a small subset o f the 5 0 A table of comparative sample oracle-bone and SW forms from each category is given in § 1.3.6.2. 5 1 In addition to the above, Keightley enumerated weather, childbirth, distress or troubles, dreams, settlement building, orders, tribute payments, divine assistance or approval, and requests addressed to ancestral or nature powers. 5 2 Bronze products that carry inscriptions are conventionally called "bronze vessels". In this thesis, "bronze vessels" is used for them collectively, and "bronze artifact" refers individually to a bronze product not properly a vessel, such as a sword or a set of bells. 5 3 This point will be discussed later (§ 1.3.6.1). 5 4 The stylistic features will not be treated in this thesis. 26 graphs in use during the Western Zhou period. 5 5 1.3.4. Writing of the Spring and Autumn period The political power o f the Zhou gradually declined. Three hundred and fifty years after the overthrow of the Shang, K i n g Y o u & was killed by the Rong & "western barbarians" in 771 B C . The capital was moved from Hao H (south-west o f present-day X i a n city, Shanxi Pro v. ( r ^ . W ^ W ^ ^ W # ) t o Luoy i {§, (present-day Luoyang, Henan Pro v. j 5 f ^ % %• W)- One o f the powerful feudal states, the Q i n state, which was originally located in present-day Shanxi Prov., 5 6 began to expand eastward, and eventually occupied the former capital o f Zhou. The Royal House o f Zhou remained as the nominal Imperial Court, but real power alternated among several "vassal" states. This age is called the Spring and Autumn period. The surviving writing o f this period is chiefly seen on bronze vessels, bells and weapons. The production o f bronze artifacts was no longer monopolized by the central Zhou House, as political and economic power shifted to various feudal lords who began producing bronze artifacts in their own foundries. Divergence among the various feudal states during this period is recognized only in stylistic features, although some minor structural discrepancies are also observable. 5 7 1. 3. 5 Writing of the Warring States period The Warring States era no longer had the unifying central political power formerly provided by the Zhou Royal House. The various states, as independent political entities, 5 5 For example, Bernhard Karlgren (1936, 158-159) believed that many more characters must have been in use during the Western Zhou era. Jinwen bian (1985. 4th ed.), a compilation of bronze graphs, lists 1416 different graphs. 5 6 Traditionally, Xu Guang's (#• Jlc) explanation that the Qin was originally located in Huaili H JL (present-day Shanxi RfeW Prov.) has been accepted, although an alternative location in present-day Gansu Jfr Prov. has been proposed (Lin 1981, 33-34). Xu's explanation was collected in Shijijijie ji IE which was compiled by Pei Yin |||B during the Song 3fc time, 420-479 AD (Zhonghua Shuju 1972, 177). 5 7 Qiu (1988,47) speculates that variations of bronze graphs among various states were mostly stylistic rather than structural. As will be examined in § 1.3.6.2, minor structural changes, such as the addition of one or two strokes of no significance, took place during the SA period (e:g. iZ > zheng 'to go on a campaign'). 27 competed for supremacy. Wi th the rapid economic and societal developments, writing also became popularized. Writ ing from this period has been found on a wider variety of materials such as seals, stones, currency, pottery, bamboo slips, and silk cloth, in addition to bronze objects, with a greater variety o f content than that of previous periods. He L i n y i (1989, 78-169) has divided the entire "Chinese" area into five regions which can be classified according to their homogeneous graphic features o f the Warring States period. Includes Q i ($-), Lu (#), Zhu ( $ ) , Teng ( 0 ) , X u e ($i), and other small states The Y a n ( ^ ) state Includes Jin (£• ) , Han (#) , Zhao ( & ) , W e i (K0, Zhongshanguo (4" dl Hi), Eastern Zhou (JUM), Zheng (IP), W e i (flr) 4. Chu ft area Graphs: Includes Chu (ft), W u ( & ) , Y u e (ft), X u (fr), Ca i (H), Song (5jc), and other small states 5. Q in 0- area Graphs: The Qin ( # ) state Q i fl?- area Graphs: Y a n ^ area Graphs: Jin -ff- area Graphs: The Chu and Qin , the two geographical areas whose writing w i l l be examined in this thesis, were located adjacent to one another, in the south and west o f the whole area, respectively. The Records of the Historian (Shiji jfc. f£) states that, king W u of Chu (ca. 740-690 B C ) identified his state as "barbarian" (manyi & J | ) , 5 8 Nevertheless, with its Chinese writing system and presumably its spoken language, the Chu state appears to have been within the sphere o f Chinese civilization during the Warring States era. The Q i n region, originally the home o f a small western tribe, but also with Chinese language and writing, had by the fourth century B C emerged as a strong Chinese state.59 Although it is possible that 58 Shiji 40 (Yoshida 1979,412). 5 9 Although the state of Qin was described to be a western or northern barbarian tribe (for example, by the Qin prime minister Shang Yang ft f t (390-338BC) in Shiji, 68 (Mizusawa 1997, 214)), its ethnic origin is ambiguous. 28 various original native (non-Chinese) languages persisted in these regions, there is no question that the educated classes from the various areas communicated in the standard language, yayan fill" 'proper [-ly pronounced] language'. 6 0 Several historical records point to this. For example, Xunz i JUj^- (ca. 313-238 B C ) was born in Zhao went to Q i jf- to study, and lived in Chu ft where he engaged in literary and educational work. Su Q i n 0#, who was born in the Eastern Zhou capital and studied under the Q i literati, visited and successfully persuaded six states into allying together against the Q i n state.61 There were seven Chu literati who served as Qin prime ministers. 6 2 A fundamental question related to the theme o f this thesis thus arises. To what extent did the standard language used in different regions during the Warring State period vary, particularly with respect to pronunciation? The question is crucial, for deviation in pronunciation affects the efficiency o f the phonetic element of a given graph, which might then be changed; regional variations in pronunciation may lead to divisions or weakening of the phonological system. Ih addition to what the historical records mentioned above tell us, we can find another clue to the answer i n the Chuci ft ^ , a collection of a type o f poems largely composed by the Chu literati-official Q u Yuan JS M-(ca. 340-278 B C ) whose rhymes are basically confined to the standard rhymes o f the Shijing H H . 6 3 This thesis postulates that: 1) there may have been regional, dialectal, characteristics in pronunciation o f words, a matter to be kept in mind; and 2) regardless o f any regional variation in pronunciation of words, the underlying phonological system remained the same 6 0 The commentator Zheng Xuan IP (127-200 AD) interpreted yayan Sll" as 'to speak with proper pronunciation'. This annotation can be found inLunyu jijie tafffcompiled by He Yan "f»T#? during Wei H time (220-265 AD) (Zheng 1981, 112,396). 61 Shiji 69 (Mizusawa 1997,224-265). Su Qin lived around in the early third century BC. 6 2 Gan Mao # Qu Gai B £ , Xiang Shou fa #, Wei!Ran #fc#, Mi R o n g ^ , Chang Ping jun g and Li Si all served as prime minister of Qin between 309 and 208 BC (Ma Feibai 1985, 187-195). There is also an anecdote describing the quality of standard language of the Chu literati. Mencius (jfe^?-) described the Chu literati Xu Xing f r f f as a southern barbarian who speaks like a shrike (butcherbird) (Mengzi "Tengwengong zhangju shang" A). It is not clear, however, whether the implication is that yayan as spoken by the Chu literati was deviant, or whether Mencius merely offered a pungent criticism of Xu by playing up Xu's origin in the region of a different dialect. This incident appears insufficient to prove that the standard language spoken by the Chu people was different from that of the other regions. 6 3 There are not many differences between the rhymes of the Shijing and Chuci: for example, the graphs belonging to geng # rhyme group and the graphs belonging to zhen % rhyme group rarely rhyme together in the Shijing, but more often in the Chuci (Luo and Zhou 1958, 81). 29 across the regions. 1. 3. 6. The graphical evolution of guwen We now turn to the individual graphs to determine the general tendency of graphical evolution. We shall trace their development from the Shang era to the Warring States time, and then examine the characteristics o f the graphical divergences o f the Warring States period. 1. 3. 6.1. The orthodox and vulgar forms of the script A n y discussion o f the evolution o f individual graphs during the historical period considered in this thesis must take into consideration the two guwen traditions then current. Q iu (1988, 42-69; 1989, 81-120) has distinguished these traditions as the formal or orthodox form (zhengti l E H ) , and the vulgar form (suti #11), and has expounded a theory of their relationship. The orthodox form was more intricate in its shapes and more meticulous in its execution than the vulgar form. Orthodox writing was used for official or some dignified purposes, and was by its nature conservative. The vulgar form was, in contrast, simplified and rough, and tended to evolve very quickly. A t each period, the two systems existed side by side. Shang oracle-bone graphs, for example, are carved in the vulgar script; most of the bronze inscriptions are in the orthodox form, the exceptions tending to be placed during the late Shang. Throughout the guwen stage, the rapidly evolving vulgar script acted as a catalyst for the evolution of the formal one. The influence o f the vulgar script on the formal, Qiu asserts, was already present in the late Shang bronze and strengthened during the subsequent Western Zhou and Spring and Autumn periods. 6 4 During the Warring States period, the vulgar script came to supplant altogether the orthodox;form in the eastern regions, particularly in the Chu. In the Qin state, on the other hand, both the well-ordered and balanced orthodox form and the vulgar form developed from it generally maintained their distinct uses until the end o f the guwen period, although occasional vulgar writing is sometimes found on the Qin 6 4 For example, the forms of the graphs on the Jin # state bronze vessel, Luanshu fou HI?-fir, was different from that of other bronze graphs produced during the SA period, probably under the influence of the vulgar forms. 30 bronzes, for example, the Shang Yang maodun $j f t ^ ^ f c . In our table, the "vulgar" oracle-bone graphs precede the "orthodox" Western Zhou bronze graphs as the earlier o f the two forms. Orthodox writing being the more conservative, however, the oracle-bone graphs do not necessarily preserve a more archaic shape of script than the later bronze inscriptions. Is it correct, then, to treat O B I graphs as the attested ur-sample o f the Chinese writing system? I would answer this question in the affirmative. Q i u has observed that, although the forms o f the O B I script may in some ways be more evolved than those o f their orthodox Shang bronze-vessel counterparts, that two forms from the Shang period onward tended not to diverge, but indeed to converge under the influence of the vulgar form. The partitioning o f the time scale into active periods o f development approximately 250 to 300 years in length allows us to note, moreover, that the vulgar form of any given period does not appear much more evolved than the orthodox script of the period following. It is therefore valid to treat the oracle-bone graphs as being not only temporarily but systemically the earliest samples o f Chinese writing, although the coexistence of and interaction between the two traditions should always be kept in mind. 1.3. 6. 2. Simplification and elaboration Two opposing tendencies can be observed in the diachronic graphical evolution of individual simple graphs throughout the guwen phase: they are (1) primarily simplification and abstraction o f the pictorial forms, jianhua HHt, and (2) secondarily elaboration, fanhua $Ht (Qiu 1988, 28-30, 42-47; 2000, 45, 70). Simplification for the ease o f writing takes three forms: (a) xiantiaohua jS&IHt, "linearization o f pictorial forms", in which the thickness o f lines is made uniform and a solid square or round part o f a graph is replaced with a line, as in i > f wdng ' k ing ' ; (b) pingzhihua^-'M.i^, "streamlining of graphic forms" in which a curvilinear, uneven line is straightened or separate lines are connected, as in | f ma 'horse'; and (c) bihuahua %E H. It, "segmentation o f graphs into several strokes" as in %> ^ si ' s i l k ' . 6 5 The elaboration o f simple graphs adds one or two strokes to classify the graphical 6 5 The first two examples are given by Qiu (1988,46-47). The third one is found on the Baoshan bamboo slips. 31 representation and to avoid formal ambiguity, as in ^ > _h shang 'above', and J > 5 yu 'jade', "f > fi\ rbu 'meat' (Qiu 1988, 29). Unfunctional elaboration (just adding strokes without any purpose or function) can also be observed during the Spring and Autumn period: for example, ^ > ^ tian 'heaven, sky' , and lj£ > U£ zheng 'to go on a campaign' (Qiu 1988,29). 6 6 The simple graphs do on the whole become simplified, diminishing their pictorial value as well when they act as constituents in multi-element graphs. Another type o f elaboration was the augmentation o f a given simple graph with another simple graph. This type o f elaboration o f existing graphs must be distinguished from the combinational processes used to form new graphs. There are several examples o f mere graphical elaboration without a simultaneous change of narrowing o f meaning, such as ^> > J E M.feng 'phoenix' (Qiu 1988, 6). The favored principle for graphical formation also shifted during the guwen period, from semantic graphs (such as H Y ) to SP graphs. It can be surmised that initially creation was exclusively pictographic. After al l o f the basic graphical formation principles had appeared, the formation of SP graphs became more common than purely semantic compounding (Qiu 1988, 32). The proportion o f the biaoyizi & "express-meaning graphs" decreased for two reasons: (1) simplified and abstracted simple graphs no longer carried sufficiently strong visible evidence o f their meaning to function efficiently in semantic combinations; (2) existing biaoyizi were transformed into SP graphs by the addition of phonetic indicators. The shift o f the commonly used principle o f formation is manifest from the comparison of the statistics for each category.6 7 The first column below gives the numbers for the oracle-bone graphs, and the second one, for those Q i n dynasty graphs (221-202 B C ) categorized as xiaozhuan %. in the S W . 6 8 6 6 Although Qiu gives ^ tmn 'heaven, sky' as an example of unfunctional elaboration, which can also be observed during the Spring and Autumn period, this graph can be found on the oracle bone inscriptions. 6 7 One of the OBI graphs' categories, "unknown", and the ambiguous traditional category zhuanzhu used ioxxiaozhuan graphs were labeled as "other". 6 8 The OBI figures are taken from Li Xiaoding (1993,91-95); the SW figures, from Zhu Zhunsheng &%kM (1788-1858) mLiushu Yaolie is^^f\ mShuowen tongxun dingsheng IfoXftMfeM-32 Table 1.2. Category o f Graphs O B I S W Pictographs 276 (22.53 %) 364 ( 3 .84%) Zhishi Graphs 20 ( 1.63 %) 125 ( 1.32%) Huiyi Graphs 396 (32.33 % ) 1167 (12.32 % ) Phonetic Loan Graphs 129 (10.53 % ) 115 ( 1.21 % ) Semanto-phonetic Graphs 334 (27.27 % ) 7697 (81 .24%) Other 70 ( 5.71 % ) 7 ( 0.07 % ) T o t a l 1225 ( 100 % ) 9475 ( 100 % ) The above data suggest that the formation o f SP graphs became the primary method of formation toward the end of the guwen phase. 1. 3. 7. The graphical features of the Warring States graphs This section describes the classification o f the processes that contributed to regional structural graphical discrepancies. He L i n y i (1989, 184) postulates that the principles of graphical development during the Shang and W Z periods carried over to the W S period. He (1989, 184-223) focuses on three major phenomena in the evolution of W S individual graphs: simplification, elaboration and alteration. W e shall separate his examples o f simple and multi-element graphs. 1. Simple graphs Simplification of simple graphs is the omission of a stroke or strokes, and the abstraction o f form: as in W > ^ rna 'horse' . 6 9 It must be noted that each such case may be a particular scribe's momentary abbreviation, not a conventionalized form, particularly i f examples are rare. In Chapter 3, we w i l l examine complication o f simple graphs as a regional feature of the writing, although He gives no examples. Variations i n execution, such as % > 1[ chB 6 9 The former graph is from the Baoshan bamboo slips, and the latter is a Qin graph appearing on the Shuihudi Hifeiifc bamboo slips. 33 'car, are not treated in this thesis.' 2. Multi-element graphs 1) Simplification Simplification o f multi-element graphs ranges from the omission o f a stroke or strokes, to the abstraction o f a grapheme, to the omission o f an entire grapheme. Omitted graphemes may be an S E and a P E o f an SP graph, an S E o f an H Y graph, or a grapheme o f a multi-element SE or P E . 7 1 For example: \ )^ye 'leaf, generation' (PE # shl 'thirty, generation' in an SP graph 31 .ye ' l ea f is omitted) 7 2 i i ) ^ a « ' p e a c e f u l ' (SE A midn ' r o o f in a H Y graph ^ an 'peaceful' is omitted) 7 3 i i i) ^ chu ' C h u (state name)' (one o f the two identical elements mu 'tree' is omitted). 7 4 A s an example in iii) shows, duplication o f graphemes is sometimes eliminated. However, simplification must in some cases be treated cautiously. SP graphs that show apparent abbreviation, such as the loss of P E or SE, may in fact have been yet unaugmented semantographs or jiajie graphs respectively, still in the second stage of their formation. It is possible that such retarded development was specific to a given region. I f the antecedent of a given graph is unaugmented, the unaugmented graph used in the W S period cannot be called 7 0 Both graphs appear on the Zhongshanguo'f Jj H (Pingshan (ty\k) county, Hebei (jf h^) Prov.) bronze belonging to the Jin -fr script region. (He 1989, 217) The first graph can be found on xiao ding/h #(f (small cauldron),_/a«gz«o dou jjlM.3- (a ritual vessel with a square base) and others. The second graph appears onpinggai dou ty j | fi (a ritual vessel with a flat lid). 7 1 Since graphs, in their next evolutional step, were augmented with semantic or phonetic elements. The omission of semantic or phonetic elements might stem from their usage in the previous stage (He 1989, 185-191). 7 2 This graph appears on the funerary object (Zhangjia^^.) from the Zhongshanguo tomb (He 1989, 189). Although it is an example of omission of a PE from an SP graph found on a bronze vessel, He states that this type of omission was rare, and that the currenciy of the WS period mostly carries graphs of this case. 7 3 This graph appears in the bronze bell Zhedaozhong;#i?J$t (He 1989, 188). 7 4 This graph appears on the Chu bronze cauldron Yinyue ding^ $ jfjj (He 1989, 189). As this example shows, duplication of graphemes is sometimes omitted. 34 simplified, unless dating makes clear that the graph had passed through an augmented stage. 2) Elaboration Elaboration o f multi-element graphs is the addition o f a stroke or strokes, or o f a grapheme, without a change in the meaning o f the word represented by the multi-element graph. The addition of a grapheme, according to He, seems in some cases to be motivated by phonetic or semantic reasons, and in others without apparent motivation. On occasion, however, it may be quite difficult to determine the function of such graphemic augmentation, a) Non-functional augmentation The following are He's examples o f non-functional graphemic augmentation: iv) % shi ' room' : H contains two identical graphemes J ( 5 zhi 'to reach'), by duplication; 7 5 V ) $L qin 'close, parent': Ipf is augmented with the element ^ midn ' roof ) ; 7 6 vi) W bing 'the third of the ten Heavenly Stems': § is augmented with the grapheme d ( P kou 'mouth'). 7 7 He (1989, 196-198) enumerates the following 10 simple graphs that, when added to existing graphs, produced no change of meaning in the compound, and indeed assumed no apparent function (such as to indicate, the pronunciation). In parentheses are shown examples of graphs to which the grapheme in question was added, and the material on which the augmented graph appears.78 1) F hdn ' c l i f f (%_huan 'to turn round'; ZS) 2) midn ' r o o f ($§, qin 'close, parent'; Wangshan H iJj bamboo-slip graph) 3) p hu 'door' (jf fa ' l aw ' ; ZS) 4) jfc li 'to stand' qido 'h igh ' ; H M ) 7 5 This graph appears on the Chu bronze cauldron Yingan ding 'tX-Mfi (He 1989, 195). 7 6 This graph appears in the Wangshan H di bamboo-slip graphs (Re 1989, 196). 7 7 This graph appears in the Chu silk-cloth graph, Changsha boshu % # (He 1989, 197). 7 8 Some material names are abbreviated: ZS stands for Zhongshanguo + ib HI bronze vessels, HM stands for Houma mengshu ^ ,(§ Ji # (jade covenant inscriptions from Houma, Shanxi (Jj W) Prov.), and CS stands for Changsha silk cloth. 35 5) P kou 'mouth' (W bing 'table, the third of the ten Heavenly Stems'; CS) 6) 0 yue 'to speak' (jft lian 'to gather i n ' ; Z S ) 7 9 7) <^ xin 'heart' sui 'ear o f grain'; H M ) 8) X you «to hold in hand» (J& di ' land' ; H M ) 9) kpu 'to beat' (£|E 'to start'; Biaoqiangzhong J$k%^) 10) T 'pennant or streamer' ( i f tang 'hot l iquid ' ; CS) b) Functional augmentation A n extra grapheme could be functionally augmented in two ways: (1) phonetically and (2) semantically. In both cases, a simple graph (grapheme) was added to graphs o f three types: simple graphs, semantic compound graphs ( H Y graphs in He's term), and SP graphs. (1) Examples o f augmentation with a phonetic element are: ix) ifechi (*thj9gx) 'tooth' (simple graph + PE) = te> chi (*thj9gx) 'tooth' + ± zhi (*tjsgx) 'to stop' 8 0 X ) £ bao (*p9gwx) 'to protect' ( H Y graph + PE) = 1^ bao (*psgwx) 'to protect' + & jou (*pji9gw) 'earthen vessel ' 8 1 xi) $ qi (*dzid) 'place name' (SP graph + PE) = ffqi (*dzid) ' Q i 'place name' + # qi (*tshid) ' w i f e ' 8 2 The following case is not strictly an augmentation, but rather a modification o f shape of a part o f the graph into a phonetic element. However, the result is the same as that of an augmentation with a P E , so it has been included in this group. xii) W (§0 sheng (*skh?-jingh) 'sage': = ^ sheng (*skh?-jingh) 'sage' + ££. ting (*thingx) «man standing on the earth» 83 7 9 This augmented graph also appears on the Baoshan bamboo slips, and will be treated in § 4.3.3. 8 0 These augmented and unaugmented graphs appear oh the Yangtianhu $j bamboo slips (He 1989,201). 8 1 This graph appears on the Qi ^ bronze vessel Chenhou Wu dui S-lt^ Hc (He 1989, 201). 8 2 This graph appears on one ("Tianche" ffl J$-) of ten stones which are collectively named as the Stone Drum shigu %i£ of the Qin (He 1989,202). 8 3 This altered graph also appears in the Baoshan bamboo graphs (He 1989, 212). The reconstructed sound is based on Li Fang-kuei. 36 (lower part o f the H Y graph was altered into a PE) (2) Examples o f graphemic augmentation intended to make explicit the meaning o f the word represented are: 8 4 vii) "7^ xiong 'elder brother' = xiong 'elder brother' + A ren 'man' (simple graph + S E ) 8 5 viii) I t ydng 'south side o f a h i l l ' = H ydng 'south side o f a h i l l ' + ± nl 'earth' (SP graph + SE) . 8 6 3) Alteration Alteration is the replacement o f PEs or SEs, or the mistaken substitution o f one grapheme with another. (1) Replacement o f P E This can be seen, for example, in the graph )3t ding (*dingh) 'to settle' in which the original P E IE zheng (*tjingh) 'correct' is replaced with T ding (*ting) 'the fourth of the ten Heavenly Stems'. 8 7 (2) Replacement o f SEs The two SEs in the pair below are each combined with a given P E to represent the same word. Gao (1987,146) has observed that the usage of distinct but semantically congruent semantic elements in a multi-element graph was both synchronic and diachronic. A s a result, a given word could at the same time have had several representations with different semantic elements. It can be surmised, therefore, that when a semantic determiner was added to the phonetic loan graph to derive a SP graph, the choice o f the S E varied according to the scribe. He (1989, 205-207) also observes that several cases of the S E in an SP graph replaced with another graph semantically similar. O f the following 34 pairs o f SEs used interchangeably 8 4 There is no adequate example for the addition of a graph to a HY graphs (He 1989, 198-200). 8 5 This graph appears on the Chu bronze vessel, Wolingjun dou S. (H6 1989, 199). 8 6 He used this sample from the bronze inscription collection, Sandaijijin wencun 3. r£ llf & % fe (He 1989, 200). This also appears in the Baoshan bamboo-slip graphs. 8 7 This graph with its phonetic element T appears on Houma mengshu 1^ ,l§ H 1r (He 1989, 212). Qiu (1988, 57) ascribes this phenomenon to a different use of the phonetic loan graph, to which a semantic determinative was added. 37 during the guwen period, the former 27 pairs are taken from Gao (1987, 146-180) 8 8, and latter 7 pairs from He. He also specifies those observed in the Warring States graphs (marked with "WS") . 1) A ren 'man': -fc nfi 'woman' (^yu 'to produce') 2) "ft" shou 'head, neck': X ye 'head' (M ydn 'forehead, face') W S 3) D kou 'mouth': "jjf ydn 'to speak' (^yong 'to chant') W S 4) <^ xin 'heart': "g ydn 'to speak' {% de 'virtue') W S 5) -f- yin 'sound': "t" ydn 'to speak' ( H y i « 'careful') 6) f*J rbu 'meat': # gi i 'bone' (f£ 'bone') 7) i z M 'foot, to stay': M. zu ' foot ' 8 9 (IS zhong 'to follow') 8) A chub 'to go on the road': \tzhi 'foot, to stay' ( i £ m 'against') 9) A chub 'to go on the road': ^ C/H 'road' ( i ^ bian 'border') 10) A c/zwo 'to go on the road': ^ zow 'to run' ( i t qian 'to send') 11) jt /?u 'to beat': ge 'dagger-axe' (jfcjiii 'to save') W S 12) , | m a o 'b i rd ' : £ znwl bird with a short tail ' (0yi 'chicken') 13) H /neng 'tree frog': ± cnd«g 'insect' (isfc znu 'spider') 14) i f c 'grain plant': mi ' r ice ' ($j c/ao 'rice plant') 15) # mi ' r ice ' : ^ s/n 'food, to eat' (fe /* 'grain') 16) $Lyi 'clothe': ih 'cloth (long like towel)' ($£ qun 'skirt, hem') 17) & si ' s i lk thread': 'clothe' ( f t ku 'trousers') 18) £ sud 'rope': & si ' s i lk thread' (jft S«6M ' s i lk band') 19) jfc su 'white, s i l k ' : ^ s i 's i lk thread' (flkhuan 'relaxed') 20) | L ludn 'disorder': & si ' s i lk thread' (# c/wo 'indulgent') 21) ^ m/arc ' roof : T ' c l i f f (% zAdi 'house') W S 22) & /dw 'earthen vessel': X, wa 'unglazed pottery' (fe: gang ' a kind of earthenware') 23) JE min 'plate': % wa 'unglazed pottery' (|fc cing 'a kind of earthenware') 8 8 Gao enumerated 41 pairs, but those which do not appear in my tables have been excluded. 8 9 it zhi 'foot (below the ankle): J5L z" 'foot (between the knee and the ankle)' 38 24) I 'outer wall o f a city': ± ra 'earth' (4ft cheng 'walled city') W S 25) 0 /uin ' f ie ld ' : i ra 'earth' ( f /w 'to remain') W S 26) i ra 'earth': 4 . / & ' h i l l , ladder' (ifc gut 'to destroy') 27) 0 ri 'sun': ^ ywe 'moon' ( $ #1 'period') W S 28) @ mu 'eye': ^jidn 'to see' ($ , s/» 'to observe') W S 29) ^ /man ' roof : "X xue 'entrance of a side-cave residence' (% yu 'to lodge') W S 30) jfc 5 i ' s i lk thread': # 'to bind ' (jbfe c/zww 'pure') W S 31) Jlty cao 'grass': /MM 'tree' (5Ji /iuw 'pungent vegetables') W S 32) 77 dao 'knife' : # ren 'blade' (Ig zhdo 'to instruct') W S 33) X you 'to hold in hand': je /?u 'to beat' (fifing 'respect') W S 34) A mil 'tree': «e 'grain plant' (#. xm 'to rest') W S (3) Mistaken substitution In some cases, it seems that certain graphemes are so similar in graphic form that they were interchanged by confusion (He 1989, 208-210). Most o f the mistakenly altered graphs, according to He, were proscribed by the Script Reform. Although he does not specify the function o f the grapheme for this phenomenon, his examples seem to suggest that it occurs mostly with semantic elements. Examples are given in parentheses: the graph which was written with two different graphemes; and the material on which graphs with alternating graphemes in question appear.90 1) A ren 'man': ^ gong 'bow (fi ddi 'to substitute, generation'; H M ) 2) g mil 'eye': 0 tidn ' f ield ( # kdn 'to see'; ZS) 3) H ri 'sun': @ mil 'eye' ( $ ming 'bright'; H M ) 4) U bei 'cowrie ' : @ mil 'eye' (If shang 'reward'; ZS) 5) 0 ri 'sun': 0 tidn ' f ie ld ' ( # x l 'past'; ZS) 6) P kou 'mouth': 0 ri 'sun' (^ you 'help [from a god]'; Caihoupan |^#) 7) ib zhi 'foot, to stay': cao 'grass' ( # ben 'to run'; Stone Drum "Tianche 'V'Lingyu" 9 0 Some material names are abbreviated: ZS stands for Zhongshanguo , bamboo-slip graphs found in Sui County (pf Hubei Province (j#Mk#), appear in the 6th column. This set of material, from the early W S period, is a list of funerary objects on 240 pieces o f bamboo slips with 6,696 graphs.5 The occupant of the tomb in which these bamboo slips were found was a sovereign o f the small state o f Zeng H . 6 7. The Xinyang i t |# bamboo-slip graphs found in Xinyang County (-fg Henan Province (^T^J f i l l the 7th column. 7 They date to the early W S period. The inscriptions include a list o f funerary objects (29 bamboo slips; about 894 graphs) and the writings of a Confucian scholar (121 bamboo slips; 470 graphs). The burial has been surmised to be that o f a person belonging to iheshidafu i J^£L 'literati and officialdom' o f the Chu state. 8. The Yangtianhu #7\ iJ$ bamboo-slip graphs found at Changsha (-^nty), Hunan Province 4 A number of seal graphs are treated as WS graphs which display the variant orthographic features mentioned by He (1989). There conspicuously deviate from the standard form in the degree of simplification or fusion. Such deviations can be ascribed to the limited space available for writing, and to artistic considerations, for the seals were official or personal emblems. The graphs on pottery, most of which were produced by stamping with seals, therefore, have also been excluded. 5 Hubei sheng Bowuguan (1989, v. 1. 461-464,452-453). According to this book, the tomb from which these bamboo slips were found may have been built between 433- 400 BC. 6 This small state has not appeared by the name Zeng H in any received texts. Its identification has not been verified. 7 Henan sheng Wenwu Yanjiusuo (1986,139-140, 120-122). 45 (M18%), fiUthe 8 t h column. They date to the middle o f the W S period. 8 This set contains a list o f funerary objects written on 43 pieces o f bamboo slips and includes about 317 graphs. The identity of the occupant o f the tomb is unknown. 9. Two sets o f the Wangshan J£ di bamboo slips excavated from Jiangling County ( j £ f ^ l £ ) , Hubei Province f i l l the 9th column: records o f divination (207 bamboo slips; 1093 graphs) from "Grave N o . 1", and a list o f funerary objects (66 bamboo slips; 925 graphs) from "Grave N o . 2". They date to the middle o f the W S period. 9 The occupant o f "Grave No . 1" has been tentatively identified as the great-grandson o f K i n g Dao (reigned 401-381 B C ) . 1 0 10. The 10th column is for graphs on silk cloth allegedly found at Changsha Hunan Province (}$•$}#). They date to the middle o f the W S period. 1 1 The inscriptions, written on a silk cloth with about 900 graphs, consist of three parts, dealing with mythology, the relationship between disasters and human actions, and various calendric taboos. 11. The 11th column is filled with Baoshan j?L\h bamboo-slip graphs. 1 2 They date to the middle o f the W S period. 1 3 Information about this material w i l l be presented below (Section 2. 2) in more detail. b) The Qin graphs 5. The bronze-vessel graphs o f the Wan ing States period appear in the 5th column. 6. The graphs incised on stones, shiguwen %$X%, fill the 6th column. They date to the W S period. 1 4 The inscriptions are in verse and describe noblemen hunting. There are about 431 graphs. The stones on which the inscriptions were incised have been lost. The specimens 8 ChuWenhuaYanjiuhui (1984, 20-21). 9 Hubei sheng Wenwu Kaogu Yanjiusuo (1996, 208-210). 1 0 Hubei sheng Wenwu Kaogu Yanjiusuo, and Beijing Daxue Zhongwenxi (1995, 136). 1 1 ChuWenhuaYanjiuhui (1984, 13-15). 1 2 The information about this main material, the Baoshan bamboo slips, will be given in § 2.2. 1 3 Hubei sheng Jingsha Tielu Kaogudui. (1991, Vol. 1, 330-333). 1 4 The dating of shiguwen 'fi^nX has been debated; it ranges from the early SA period to the WS period. This material is placed in the first column of the WS graphs (Mattos (1988, 367-369), and He (1989, 159)). 46 were collected from rubbings. 1 5 8. The wooden-slip graphs found in the tomb in Qingchuan County ( # Jl| | £ ) , Sichuan Province ( 0 )\\ %), fill the 8th column. They date to the middle o f the W S period. 1 6 Six items from orders ranging from amendments to land laws to public construction projects are recorded using 146 graphs. 10. The stone-inscription graphs, ihezuChuwen M f t ^ , fill the 10th column. They date to the middle o f the W S period. 1 7 The same statement o f imprecation upon the Chu king was addressed to three different deities. 1 8 The original three sets o f stone inscriptions, about 320 graphs in all , have been lost, but an ancient rubbing has survived. 1 9 11. The Shuihudi Hifeitk bamboo-slip graphs found in Yunmeng County (9 PM), Hubei Province (Mit^), fill the 11th column. They date to the late W S period. The inscriptions include legal documents; a catechism about laws; guides for judicial officers; an administrative document issued by the governor (junshou IP ^ ) o f the Southern Commandery; a chronicle of major events (not used); and a calendar noting diurnal prohibitive and auspicious omens. 2 0 The number of bamboo slips amounts to 1155 pieces. The occupant o f the tomb from which these bamboo slips were found was a Qin local official named X i C. The modern transcription 12. The 12th column presents the Baoshan graphs with their constituents transcribed to modern form. This column, therefore, indexes the table, and we shall call its graphs the entry 1 5 MaXulun(1935). 1 6 Sichuan Bowuguan, and Qingchuan Wenhuaguan (1982, 11). 1 7 The dating has been debated, and ranges from 312 BC to sometime during the Chu King Qinxiang I f (298-263 BC). The date 312 BC, based on Chen Zhaorong's review (1993, 574-576), seems to be acceptable. 1 8 Of these three, one (yatuo$.l$fc) has been excluded as a likely fake Guo Moruo Zhuzuo Bianji Chuban Weiyuanhui 1982,284-285). 1 9 Transmitted rubbing is not the original but the hand-copy of rubbing (Guo Moruo Zhuzuo Bianji Chuban Weiyuanhui 1982, 280-284; 315-320; 327-332). 2 0 Shuihudi Qinmu Zhujian Zhengli Xiaozu (1990,2). The occupant of the tomb Xi $• lived between 262-217 BC. The chronicle ends with the events of 217 BC. Material from the chronicle has been excluded, for it is certain that the text was written after the Qin unification (221BC). I have considered the other inscriptions as having been written during the last years of the WS period. 47 graphs. 13. The 13th column supplements the table with supposed modern equivalents for each Baoshan graph. 2.1.2. The vertical order of the table The graphs, Baoshan bamboo-slip scripts as entry graphs, have been vertically arranged by their number o f strokes, beginning with the graph — yi 'one', placed in the 11th column. If no graph from other materials could be found for the entry graph, the cell o f that entry was left blank. 2.1. 3. The design of the table The first row of each entry (3 to 11) allows the overall homogeneity of Chu script to be examined. The two consecutive rows for each entry allow the differences between the Chu and Qin forms for each entry graph to be observed. The table illustrates the generic relationship between the ancestor forms (the oracle-bone and the W Z bronze graphs) and their Chu and Qin descendants, and the evolutionary divergence exhibited by the two regional W S traditions. 2. 2. The main material: the Baoshan bamboo-slip graphs The Baoshan tomb lay close to the center of the Chu state. Furthermore, the Baoshan bamboo slips attest the greatest number of distinct graphs. The Baoshan graphs have therefore been selected as the pivotal or indexing form in the tables. The Baoshan bamboo slips were found in the cemetery called Baoshan Chumu Q Jj ft S , located on a h i l l 3 k m south o f Shilipu Township ( + J l M & ) in Jingmen City (M f"J ip ), Hubei Province (Mit$), 15 k m north o f the site o f Jinancheng $£$3 4$, the Chu capital in the W S period. 2 1 278 slips in al l , bearing 12,472 graphs, were excavated from "Grave No.2", the largest of the seven Chu graves. The burial date has been determined as the twenty fifth 2 1 Hubei Sheng Jingsha Tielu Kaogudui. (1991, Vol. 1, 1-5,265). 48 day o f the six month o f the Chu calendar in 316 BC.22 It falls in the Warring States period (5th century B C - 221 B C ) . The bamboo slip inscriptions include records o f census registration and law suits, records of divination, and a list o f funerary objects. 2.3. Procedures for analysis of structural discrepancies 2.3.1. Transcription into modern forms and treatment of variants 12,472 Baoshan bamboo-slip graphs into different graphs inscribed on the Baoshan bamboo slips used in this thesis. Without access to the original bamboo slips, three references (Baoshan Chumu Jj ft Is, Baoshan Chujian Wenzi bian iiL Jj ftfflX^H, and Chuxijianbo wenzi bian ft^ $f r%%%-t§i) are useful for deciphering the various Baoshan graphs and sorting them by the corresponding modern forms. These three compilations are, however, based on subtly but essentially different principles. Having been published at different times and by different groups o f scholars, these three works occasionally vary in their modern transcription for individual Baoshan graphs. 1. Baoshan Chumu i iLilj ftS presents photographs o f al l bamboo slips and several specimens for each graph, and transcriptions o f all bamboo documents.2 3 2. Baoshan Chujian wenzi bian % ill ftfif j t groups graphs in each entry by their stylistic and structural similarities and differences. Each grouping thus formed is illustrated by a single representative graph copied by hand from the slips. 2 4 3. Chuxijianbo wenzi bian ft %. % % % ^ $S details all graphs in handcopy with the 2 2 The Chu state referred a given year by the principal event of the previous year. Seven years are individually mentioned in the Baoshan bamboo slip inscriptions, in two groups of three (earlier) and four (later) successive years. The first one of the former three has been recognized as 322 BC because the event in its statement coincides with the one in 323 BC in Chushijia fti£|t in Shiji j£. IE. The occupant of the tomb Shao Tuo gp j& was buried while Shao Gu (M &) was Dasima (A fl Mi I an official title), between 323 and 298 BC. During this period, the order of the month and day of the latter four years only fit the four years from 319 to 316 BC. Thus the funeral took place in 316 BC and these bamboo slips were produced between 322 and 316 BC. (Hubei Sheng Jingsha Tielu Kaogudui 1991, Vol. 1. 332-333.) ; 2 3 This was published in 1991. The transcriptions and translations were done by Liu Binhui #J ##, Peng Hao % Hu Yali $ JJ£ JI, Liu Zuxin $J and Li Xueqin # $ $/. 2 4 This compilation was published in 1996 by Zhang Shouzhong 5ft «f • 49 phrases in which they appear" On occasion, graphs with different constituents appear under a single entry. For example, under the graph $ | yuan 'to change the field and cultivate it ' the graphs 1S, and | | are included because the latter two, which have no modern equivalent characters, are interchangeably used w i t h ^ a w $1 as personal names (Teng 1995, 118). The author seems to consider them as variants o f the graphyudn jUL. W e now describe the procedure o f transcription followed in this thesis. A l l graphs were compared with the transcriptions given in the Baoshan Chumu. Graphs rendered illegible by the poor condition o f the bamboo slips or, more often, by unclear photography were excluded. The modern transcription for each graph was independently arrived at using Noe l Barnard's two-step transcription method (1973, 1996), with the result checked against each of the three bamboo-slip references. Barnard's first step is to transcribe archaic graphs into modem characters by replacing their constituents with their modern counterparts. The second step is the replacement o f the "modernized" graph with an actual existing modern graph. This two-step transcription principle has been adopted for two reasons: 1) the first step emphasizes any variation in the usage o f constituents (structural discrepancy) between the W S Chu and Q i n forms; 2) the second step allows Chu and Qin counterpart graphs to be properly paired even i f they are composed of different constituents. O n occasion, the first, "modernizing" step o f the transcription presents difficulties because we cannot find a modern equivalent for a Baoshan element. In these cases, it was attempted to find a modern equivalent for the graph as a whole, rather than on an element by element basis, (e.g., zhl # 'pronoun substitute for the head o f a relative clause') The graphemically "modernized" graphs (step 1) have been placed in column 12; the true modern equivalents, in column 13. There are several cases of Baoshan graphs with different constituent structure having been transcribed into a single modern equivalent, for example, the graph v t . , ^ and % are all 2 5 This concordance is published in 1995. The transcription and classification of the graphs were done by Teng Rensheng $ £ i ^ . 50 transcribed into # bit 'to walk' (Teng 1995, 125). 2 6 Such graphs were often distinct in their usage: in the example above, the first graph was used as a verb, the latter two graphs for personal names. In this thesis, graphs with grammatical or contextual differences in usage have been given separate entries. Graphs showing the following features have been treated as variations o f a single graph, and have been placed together in Column 12: 1) omission or addition of one or two strokes without loss of graphemic identity 2) omission or addition o f one or two strokes resulting in the change o f graphemic identity, but the graphs either are used in the same context or represent the same word. 3) omission o f a grapheme, but graphs either are used in the same context, or represent the same word. The variants o f a graph sharing a column have been ordered by decreasing frequency of usage. 2. 3. 2. Principles for analysis of the structural discrepancies In this thesis, the principal materials for an analysis of structural discrepancies between the Chu and Qin W S graphs have been the Baoshan and Shuihudi bamboo slips, respectively. Two fundamental conditions have been adopted. 1. When there are structural differences between the Baoshan graphs and their modern equivalents, but the Qin form is wanting, the S W form has been adopted as a replacement of the Qin when available. 2 7 2. When the Baoshan graph in question has multiple variant graphs, the criteria by which the representing Chu graph has been selected to be compared with the Qin form are as follows: 1) the one used most often; 2) i f both are used with the same frequency, the one which can be seen in other Chu 2 6 The Hanjian yf 1fif, a concordance of guwen form (ca. 960 AD), gives the graph which is the same as^.. 2 7 Although the SW is a transmitted text with accumulated errors, the forms given in the SW, xiaozhuan /JN $ E , are the closest known counterparts of the Qin graphs. 51 materials. The structural discrepancies between the Chu and Qin graphs have been grouped as follows: (1) different constituents used at formation; (2) antecedent forms graphemically modified, in the Chu or Q in variant, or both; (3) form progressively and unintentionally modified. The procedure of placing graphical discrepancies between the Chu and Qin forms o f the graph into these three groups is as follows: 1. When the antecedent form is wanting: 1) the function o f each grapheme has been analyzed; 2) Chu and Qin forms showing different PEs with the same SE are treated as divergent jiajie (SP formation step 1) augmented with the same semantic determiners (SP formation step 2). 3) Chu and Qin forms showing the same P E with the different SEs are treated as the same jiajie (SP formation step 1) augmented with divergent semantic determiners (SP formation step 2). 2. When the Shang or the Western Zhou antecedent form is extant: 1) the Chu or Q i n form is identified as divergent from antecedent (or both forms diverge); 2) the divergent grapheme and its function (SE or PE) is identified. Detailed classifications o f structural discrepancies between the Chu and Qin simple graphs and multi-element graphs w i l l be presented in Chapter 3 and 4 respectively. Simple and multi-element graphs w i l l be properly distinguished in Chapter 3. The procedure for identifying the functions o f a given grapheme is as follows. The potential that each component grapheme o f a given graph serves as phonetic element is evaluated. The readings o f the graphs are identified using the O l d Chinese phonological system established by L i Fang-kuei ([1971] 1980). We do not yet fully understand 52 phonological relationship between the graph and its P E . In this thesis, the identification o f PEs is based on two fundamental principles: (1) both the graph and its P E have the same or, at least, a homorganic initial consonant, and (2) both the graph and its P E belong to one o f the following rhyme groups: 2 8 1) both the graph and its P E belong to the same rhyme group. 2) both the graph and its P E belong to certain different rhyme groups that share the same main vowel and are called the yin-, yang-, or rw-type groups, distinguished by their distinct finals. 3) Both the graph and its P E belong to certain pairs of rhyme groups sufficiently close in sound to be used as belonging to the same group. 1. zhi and vow # rhyme group 2. zhi and jue % rhyme group 3. zhong 4" and qin 1^ rhyme group 4. wei # and zhi Jjg rhyme group 5. ge Hfc and yuan yt rhyme group 6. ge and jia H: rhyme group 7. yang |# and dong rhyme group 8. hou M and vw fa rhyme group 2. A n y constituent grapheme not identified as a P E has been taken as an SE when semantically related to the word the composed graph represents, and as an element of unknown function otherwise. We are now in a position to distinguish properly between simple and multi-element graphs, and to formulate a clear classification of the types o f formation and modification that lead to W S divergence between the Chu and Qin simple and multi-element graphs. 2 The basic phonological relationship between the graph and its PE was stipulated by Karlgren as follows: (1) both PE and the graph it composes belong to the same rhyme group, and (2) initial and final consonants of both PE and the graph it composes are articulated at the same place, i.e., homorganic (Karlgren [1940] 1957,456-459; [1954] 1963, 271; 1968, 10-17). It must be noted that these two conditions are rough initial outlines. 53 Chapter 3 Simple Graphs 3. 0. Introduction This Chapter w i l l present tables juxtaposing the Chu and Q i n simple graphs, and describe the nature o f discrepancies between the Chu and Q i n forms. The following topics w i l l be covered. 1) Definition of simple and multi-element graphs 2) Types of discrepancies between the Chu and Qin graphs 3) Tables of simple graphs 4) Commentaries on the Chu and Qin simple graphs 3.1. Definition of simple and multi-element graphs The basic definition is that a simple graph is one composed o f a single element: for example, X . you «hand, to hold in hand», zi ' ch i ld ' , yue 'moon' and @ mu 'eye', all o f which are originally a depiction o f a single concrete object. When classifying graphs into simple (single-element), and multi-element graphs, however, we encounter a certain difficulty in drawing a clear line between the two types. For example, the graph jian 'to look' , which was originally composed o f elements denoting an eye and a man, can be classified either as a multi-element graph, i f the two elements are counted separately, or as a simple graph, i f the interpretation is that o f a man whose eye is emphasized. 2 9 In this thesis, the single criterion used to categorize graphs as single or multi-element is the number of graphemes contained in a graph. Graphemes w i l l be taken as either: (1) a minimum unit that originally depicted a single object and that can stand alone as a graph, such as zi ' ch i ld ' , or 2 9 The clear definition of a simple graph and actual classification of all graphs into these two categories seem not to have been attempted. Barnard (1978, 187) classified "archaic graphs" into four groups, i.e., simple single-element, complex single-element, simple multi-element and complex multi-element graphs, but does not give a definition for each group, the example jian JL is classified as a complex single-element graph. As has been suggested by Dr. Takashima (personal communication), the graphical minimal contrast, when found in OBI or the WZ bronze inscriptions, can be a useful criterion for determining the classification of individual graphs. The phonetic function of the constituent, if known, can also be taken as factor in determining the classification of individual graphs (This has been suggested by Dr. Pulleyblank). 54 (2) a minimum unit that originally depicted a single object but that cannot stand alone as a graph, such as the upper element of & gong 'together'. 3 0 (Miriimum units are those that can only be further decomposed into their individual strokes.) When the earliest (oracle-bone graphs) form o f the graph in question is wanting, or is o f no help in determining the original form o f the graph, classical or modern elucidations have been sought to determine the original form, and thus to classify the graph as either simple or multi-element by the number o f its graphemes. For example, the oracle-bone counterpart of the graph it bhi 'north' depicts two men standing back to back. The graph bei it has therefore been classified as a multi-element graph. The graph % zhao 'omen, on the other hand, whose oracle-bone counterpart is not attested, has been classified as a simple graph given its elucidation in the S W and the Shinjigen as depiction a crack in a heated turtle shell. Wi th the objects represented thus established, the final categorization o f graphs by this criterion has been based on their W S forms. The formal definitions o f simple and multi-element graphs adopted in this thesis follow. Simple graphs are: (1) graphs composed o f a single grapheme depicting a single object, such as yue and zhao (2) graphs composed of a single grapheme and a set of strokes that in isolation would carry no meaning, such as # dan 'cinnabar'. Multi-element graphs are; (1) graphs composed o f multiple graphemes that are also themselves independent graphs, such as jian ^ and %L qu 'to take'; (2) graphs composed o f two graphemes, one o f which, though not independent, was originally used to depict an object, such as 4* in ^ ( ^ shi 'to serve'). (3) graphs originally composed o f a single grapheme that by the W S period had become multi-element by augmentation with another graph or by modification of the original form, such as zhu 'to run' and | | long 'dragon'. 3 0 As can be seen in the oracle-bone graph p\, the upper element of & gong is elucidated as an arbitrary object carried with a pair of hands. 55 3.2. Types of discrepancies between the Chu and Qin simple graphs This section presents the types of discrepancies that can be seen in the Chu and Qin forms o f the simple graphs. Recall from the first chapter that Q i u has considered simple graphs subject to the following evolutionary processes: linearization o f pictorial forms, streamlining o f curved lines and connection o f lines previously disjoint, segmentation of graphs into several strokes, and omission and addition o f strokes. These processes which produced no structural discrepancies between the two regional forms o f a graph, but minor variations only, are classified into three categories, and presented as "non-structural modifications of one graph". a) Omission of strokes b) Addit ion o f strokes c) Executional differences (first three considered together) Structural discrepancies between the Chu and Qin simple graphs which can be found in a simple graph o f type (2) as defined above w i l l be presented as "structural modification of one graph". Wi th a few exceptions, either a Shang oracle-bone or a W Z graph can be found corresponding to each of the W S graphs treated in this thesis. Comparison o f the Chu and Qin W S forms to the corresponding antecedent graphs can often show the earliest known period at which the discrepancies that came about as a result o f the processes listed above arose. The genesis o f a discrepancy may lie at any moment between the graph's creation (pre-Shang or later), and the end of the W S period: for example, during the period of the OBI . We can find such cases in which the oracle-bone graphs were written in variations which were used in the Chu and Q i n scribes, respectively. We w i l l put simple graphs o f this type into their own category, "discrepancies inherited from the Shang period by the adoption o f different oracle-bone variants for the corresponding W S graphs". There is one case in which an almost completely different form replaced the Qin W S graphs and partially the Chu W S graphs ( 0 si 'four'). This graph w i l l be presented as an "unclassifiable case". 56 3.3. Tables of simple graphs Simple graphs are grouped into the categories defined in the previous section and presented in separate tables by category. The full list o f simple-element tables is as follows: 3. 3. 1. Simple graphs with non-structural discrepancies between the Chu and Q i n forms 3 .3 . 1. 1. Omission of strokes 3. 3. 1.2. Addition of strokes 3. 3. 1.3. Executional discrepancies 3 .3 .2 . Simple graphs with structural discrepancies between the Chu and Qin forms 3 . 3 . 3 . Unclassifiable discrepancies 3 .3 .4 . Adoption of different oracle-bone variants 3 . 3 . 5 . Simple graphs exhibiting no divergence 3 .3 .6 . Simple graphs which do not have the Qin counterpart graphs 57 3. 3. Tables (Simple graphs ) &&X ®&x Chu Qin fcffi Oracle-bone H Jj W Z bronze ^ ^ W Z bronze t l f t x : S A bronze /&di /rtff W S bronze H^i t f c f t f l Zenghou Y i bamboo slip 2g, -ft <(g ^ Stone drums Xinyang bamboo slip Yangtianhu bamboo slip Qin wooden tablet Wangshan bamboo slip Silk manuscript ZuChuwen stone tablet Baoshan bamboo slip Shuihudi bamboo slip Modern grapheme Modern character 3.3.1. Simple graphs with non-structural discrepancies 3.3.1.1. Omission of strokes ft ffl !/ dj UJ 4 k \ 5 M M KHz A n 58 ft *** ## 'Siilj'tt'ffi ** ** # te** 6 7 ^ If 7T 7 ft 8 f 9 Ml * ? 10 fl 1 11 £ f f 12 13 T< in Jx 14 I t * 15 r$ t* 16 17 tt •=><. 18 *** 19 * 20 §§6 s 9 59 ft a f t # 21 4f % 22 ID) m ffij ft 23 (ii HI 24 25 ft 26 $ $ & 27 f fit t# * 111 28 3.3.1.3. Executional discrepancies ft ¥ t * a f t # W4r* l + Etfc ¥ ? ? 2 0 3 W # 4 Ban CL * 5 ? 60 ft ¥ t * & * W & * # & * $ & * #t aft aft # W & * # & * 5 f t * fflfe* maim® 6 & 7 8 If •f * 9 XX ft 10 ?* Mi 11 ft *** Mi s 3.3.2. Simple graphs with structural discrepancies ft ¥ t * & * W & * * & * $ & * ft* aft aft # W & * *&* « & * l i f t * l % * 3.3.3. Unclassifiable discrepancies ft ¥*** WJH & * W & * # & * ft* aft ** aft # W & * *&* $ & * l i f t * l — 03? •EE (ED es CTP 3.3.4. Adoption of different oracle-bone variants ft ¥ t * & * W & * *&* SLih ft* '&iJj'W"(Bf aft aft ** # W & * *&* 5 f t * l i f t * l +T ±+ tt t t + 2 tt c b C 3 t 61 ft &&X & ill ftf? aft tf* aft # &&X 3 f few t § 3.3.5. Simple graphs exhibiting no divergence ft W4r* $&x aft aft at* # &&X lift* l — — — — ' — — — — — -— — — — 2 ) 1 Lz 1 LL 3 n A A ;L X X A A A / \ 4 1 11 + + > f -H + t 4 + + 5 1 1 X 1 6 i >> 77 7 *\ A 7? 8 — — = = — — — — = = 9 tt + 4- -b - 4 -10 %J h 11 T TT 12 — — — — — - — = — -• — — 62 ft ft* W & * $&* ft* -allien »Sft aft # &* W & * 5 S i * l i f t * « • ** 11 i =fc i I D ± ± -In +± 1 A f f 1 < T TT 7 ID T =5= 1 A ft * 1 0 i 1 t 1 7 1 / 1 fi f t l o t 1 Q l i L b it • f c 1 y Oft nil JL i Z U ± 0 1 —1 e 5 a Z I c n ? ¥ zz f f f f 0 3 ? c 11 Z D EJ OzL t i l T ±^ Z 4 i I I I I 0 < 1 ih Z D 'l' OA zo 0 7 t it z / 3-. 63 ft f *z. *# ** # « * 28 etsi y H y y y 29 HN M M M 30 X % t >K * 31 % X jit I 1 32 s O © <3> e 0 0 33 K J O 9 34 * 35 4 36 i '4 37 t A" 38 w 39 St 40 f 1 > f 41 # * 42 ¥ f 3 64 ft ¥ f * &* ft* a f t aft ** # *** l i f t * 43 m f } j t 44 X n" X 21 X X 45 ft St / \ 46 I I I — 31 47 ill t 48 I * 49 ±i X. X i 50 6 e 6 e 51 1 ± 1 52. < IT # 53 t 54 © © ffl 55 $ E 56 % 57 65 ft ¥#* &* S i r * aft aft # l i f e * 58 YIP f ¥ 59 8 60 n i n i ^ s 61 A 62 t t 63 t t 64 23 o it € , 65 41 * *. 66 67 68 f H 69 * 70 X IF •w 71 I fi f 1. 72 /TV «T\ if * w f 66 ft f t * i r * W i r * * i r * l ^ i r * aft «* aft ** # W i r * * i r * $ i r * 5 ft* t i * * 73 ? T ? 5 ? 74 m § f 75 I f S ? 76 ^# f ¥ gtgtftf-11 1 1 1 1 J $ X 18 3.3.6. Simple graphs which do not have the Qin counterpart graphs ft T t * W i r * * i r * $ i r * -aih-itnf aft tf* aft *W i r * * i r * ® , i r * t i * * l L? $ 2 u u 3 4 tl * t t 5 ft Pi 6 /v. T 7 8 -tr 9 i * 10 11 <& 12 0 13 I i i * 14 18 t St S * * S t 67 3.4. Commentaries on discrepancies between the Chu and Qin simple graphs The tables presented in the previous section have allowed us to discern the differences between the Chu and the Qin simple graphs by inspection. Table 3.3.1.1, 3.3.1.2, 3.3.4 and 3.3.5 show that most o f the W S simple graphs manifest themselves nearly identically in the Chu and Qin scripts, although occasional stylistic differences can be noted. Although the lack of Qin counterparts in Table 3.3.6 prevents strong conclusions from being drawn, the near-identity between the Chu graphs and their oracle-bone counterparts suggests the possibility that those used in the Chu area did not diverge from those used in the Qin. Discrepancies caused by progressive and independent executional deformation of the earlier graph (Table 3.3.1.3) may result in the usage o f totally different forms in the two regions after a sufficiently long passage o f time, but the Chu and Qin graphs in Table 3.3.1.3 appear to have diverged only to a small extent. In this general tendency of identity between the Chu and Qin forms, the Baoshan graph fc WO 'end' (Table 3.3.2.1) appears as a rare case o f substantial difference between the Chu and Qin simple graphs. This single example, however, is not sufficient to show a clear discrepancy between the Chu and Qin forms of simple graphs. Although He does not include these two graphs among his examples o f pairs o f graphemes interchanged mistakenly, it is possible that their similarity in form and meaning erroneously caused their interchange. The Wangshan graph in question, another Chu W S graph, did not use the grapheme he 'grain plant' used in Baoshan, but the same grapheme (fc mu 'tree') as the one used in the Qin form. The Baoshan graph mo fc, therefore, is an uncertain case o f the use o f different simple graphs by the Chu and Qin . The graph EJ si ' four' is an example o f drastic change in form, or indeed almost replacement with a new form. The earlier form seen in the O B I was used, as can be seen, until the S A period by both Chu and Qin . The new form appears to have replaced the older form in the Qin, while undergoing transition in the Chu area. There are three cases in which both the Chu and Q i n forms can be found in the oracle-bone graphs. Although the features involving discrepancies are trivial consisting of 68 several strokes, it reminds us o f the phenomenon of graphical variations in OBI . A s already noted in the first chapter, simple graphs created as pictorial depictions of actual objects were the first to have come into existence. Because it is likely that the associations among the simple graphs and the words they represent (i . e., sound and meaning) had already become by the W S period firmly established by a long history o f continuous usage, it is probable that most simple graphs did not require structural revision better to reflect both meaning and sound. It is for this reason that simple graphs used in the two different regions naturally manifest little divergence from their common ancestors and so from one another. 69 Chapter 4 Multi-element Graphs 4. 0. Introduction This chapter w i l l discuss the Chu and Q i n multi-element graphs. This chapter is divided into the following three parts: 1) Typology of discrepancies between the Chu and Qin multi-element graphs 2) Presentation and commentary on the structural discrepancies between the Chu and Qin multi-element graphs 3) Summary o f discrepancies exhibited 4.1. Types of discrepancies between the Chu and Qin multi-element graphs We w i l l first examine structural discrepancies that may have arisen in the process o f formation o f the multi-element graphs. Secondly, we w i l l discuss the modifications the multi-element graphs had gone through after they were formed, and these separately for the SP and the H Y categories. Cases in each category are presented, being divided into the following groups. (O) Cases in which both the Chu and Qin graphs underwent the same types o f process in different ways. (A) Cases in which the Chu graphs underwent modification. (B) Cases in which the Qin graphs underwent modification. A s a third group we w i l l present the discrepancies caused by unintentional evolutionary changes and misrecognition of the graphical forms. Various discrepancies not falling into any of the types stipulated above w i l l be presented in the fourth group. Lastly, cases in which both the Chu and Q i n forms can be found in the Shang or Western Zhou graphs w i l l be presented. In the discussion, graphs and their phonetic elements (graphs) are followed by their modern pronunciation, their O C pronunciation, and the meanings of the words represented; semantic elements (graphs) are followed by their modern pronunciation, the O C rhyme group they belonged to, which is sufficient for showing their non-phonetic function, and the meanings 70 of the words they represented. The modern pronunciation is not given i f it is not known because the graph in question is not used today, or in cases where the grapheme is not itself an independent graph and does not have modern pronunciation. In the case o f H Y graphs, the way the graphemes denote the meaning o f the word a given graph represents is explained. When the meaning of the word a given graph represents is unattested or uncertain, the source o f the elucidation is provided. When the antecedent graph is not presented in the discussion, its form is shown in parentheses after the explanation o f graphical composition. O B G stands for an oracle-bone graph, and W Z G stands for a Western Zhou bronze graph. A t the end o f this chapter are attached tables o f the graphs discussed, in the order presented. The discussion o f structural discrepancies between the Chu and Q i n multi-element graphs w i l l be presented in the following order. 4. 2. Structural discrepancies present during the process of formation 4. 2. 1. SP graphs 4. 2. 1. 1. Divergent usage of jiajie graphs 4. 2. 1. 2. Divergent usage of semantic determiners 4. 2. 1.3. Chu augmentation o f jiajie with semantic determiners 4. 2. 1. 4. Different graphemes used at SP formation 4. 2. 1.5. Different usage of semantographs 4. 2. 1. 6. Different usage o f semantographs and PEs 4 . 2 . 2 . H Y graphs 4. 2. 2. 1. Different graphemes used at H Y formation 4. 3. Divergent structural modifications of SP graphs 4. 3. 1. Omission o f SP grapheme 4. 3. 2. Replacement of SP grapheme 4. 3. 3. Augmentation o f SP with extra P E 4. 3. 4. Augmentation of SP with extra SE 4. 3. 5. Augmentation of SP with grapheme o f unknown function 4. 4. Divergent structural modifications of H Y graphs 71 4. 4. 1. Omission of H Y grapheme 4. 4. 2. Replacement o f H Y grapheme 4. 4. 3. Transformation o f H Y to SP 4. 4. 4. Augmentation o f H Y with extra SE 4. 4. 5. Augmentation o f H Y with grapheme of unknown function 4. 5. Divergent evolutionary changes from and misrecognition o f earlier forms 4. 6. Unclassifiable structural discrepancies 4. 6. 1. Cases in which ancestors cannot be determined 4. 6. 2. Cases o f unclassifiable graphs and modifications 4. 7. Discrepancies existed in the earlier times 4. 2. Structural discrepancies present during the process of formation 4.2.1. SP graphs In this section, we w i l l present the cases in which structural discrepancies between the Chu and Qin occurred during the two-step SP graph formation process. 4.2.1.1. Divergent usage of jiajie graphs The following 7 cases, for which no Shang and or Western Zhou antecedent forms have been found, show different PEs with the same SE and have therefore been classified as structurally divergent at SP formation step 1 (divergent usage o f jiajie; subsequent identical augmentation with semantic determiner). 1) M y u ( f c $ *pjagx) 'dried meat': SE ft (*#) , 'meat ' Chu: % f ; P E Xfii (fc US bjagx) 'father' Qin : fl! 9$ ; P E iftju (fc $5 *pjagx) 'honorific suffix to a man's name' 2) ftjl (fc$ *kjiak) 'halberd': SEge X (Ik IP) 'dagger-axe' Chu: # $ ; P E £ (## *ibnadfe) 'bits, piece' 1 Qin: $ & ; P E f - (7C$ «flag p o l e » 2 3) ® qu&n (7fc£P *gjuanx) 'enclosure for animal': SE • wei (#$5) «enclosure» 1 According to Qiu, ji #5 can be written with + being as a PE (Qiu 1979, 33). 2 This phonetic element is elucidated by Qiu. (1979, 33) 72 Chu: ® <@/ ; P E A ( 7 C $ *gwjian) 'to make a rice ba l l ' Qin: S ^ ; P E # qudn (7GIP *gwjian) 'to bend in a circle' 4) # yi (If Ip *rak) 'to be pleased' SE rafc) 'white') SW:j? ; P E - £ ze (# IP *ra£) 'to inspect' 6) M gwan (xip *&wra«) 'barrier': SE men (X%> *mzn) 'gate' Chu: PI 1? ; P E $ guan (7GlP *Awan) 'to be accustomed' Qin:H If ; P E gwa" (7U IP *£wra«) 'k ind of hair style with two tufts' 7) * ao (f IP *ngagw) 'to parch': SE A hud (Jrlp) ' fire' Chu: g # ; P E g ao (f IP *«gag-w) ' loud sound' S W f t : P E & ad (ft IP *ngagw) 'to saunter about' 4.2.1.2. Divergent usage of semantic determiners The following 14 cases show the same P E with different SEs, and have therefore been classified as structurally divergent at SP formation Step 2 {jiajie augmented with divergent semantic determiner). 1) M qi (^ .IP *gjzg) 'period' 4: P E qi (iliP *gjzg) 'possessive pronoun', or J\ (^ .IP *kjsg) 'small table' Chu: M ^ ; SE 0 ri {% IP) 'sun, day' 3 f l and ilk can be considered to be equivalent to>>2 # ze i f respectively. The SW has the graph which is explained in the SW as dabai ye cong da bai guwen yi wei zezi A 6 -fe A 6 * 1^ ^ ^ "[the word it represents] is "big white". [It] consists of [the graph] dk '[big'] and bai ['white']. [It] is guwen [form] of the graph ze ['luster']" The Baoshan graphs in question consist of shi fc instead of dd Among the WZ bronze graphs which have been transcribed into fie du 'be tired of, ample (sound)' Shi Qiangpan and Maogong ding % jffj also consist of shi fc instead of da ^ , while Nangong Hu zhong $j % ty & consist of dd ^ instead of shi fc. The interchange of shi fc and da A appears to be caused by their similarity in shape. The graph and ^ might be two variants of the same graph. 4 The PE of this graph is qi Q in the Qin graph (a pictograph of a bamboo basket which was loaned for 'it' with an additional PE, 7> 'low table'), and 7\ in the Chu graph which abbreviated a grapheme of a pictograph of a bamboo basket. 73 Qin: M ^ ; SE fl yue (fl IP) 'moon, month' 2) 4& m'ii (#IP *nrjsgwx) 'knob': P E It C/JOM ($H$5 *hnrjsgwx) 'the second of the twelve Earthly Branches' Chu : # i f ; SE S vw ( i t #5) 'jade' 5 S W : Ift; SE&jm (ft IP) 'metal' 3) ©L(* IP*%mA; ' ha rd ' ; P E c/*e« E (£ IP *sg/m) 'subject' Chu ; SE /ro & C i l P ) 'to beat' Qin ; S E >>ow X ( i l iP) «to hold in hand» This discrepancy can be seen in the graph ft xidn (^IP *gin) 'wise ' in which the graph g& or was used as a P E . Chu: & &f-^f 6 ; SE /ro je ( J . IP) 'to beat' was used in the graphgfc. Qin : ft g ; S E you X (^IP) «to hold in hand» was used in the graph S3.. 4) IS: sha. (fl IP *sriat) 'to ' : P E ^ . (JJgip *djidh) «long-hair animal», 'to curse' 7 Chu: ^ ; SE k />h (MIP) 'to beat' Qin: | £ ^ ; S E & shu ( £ IP ) ' a kind o f lance' 5) £ kudng ( I f IP *gwjiang) 'mad': P E £ ( | # I P *gw/awg) ' f i e f 8 Chu: ow # ^ as^^. This case, therefore, should be considered as the case in which the Chu form was the one having been modified (by the replacement o f the grapheme). This has been categorized as a case of Chu replacement o f SP grapheme (§ 4.3.2.), but has been shown in this section as one o f the same pattern o f structural discrepancies. 14) Mjian (ftIP *kljam) 'sword'; P E fc qian (ftip *skhjian) ' a l l ' Chu: tf £§ 1 6 ; SE &jfn (ft tf ) 'metal' Qin: M & ; S E ij aao ( f IP) 'knife' 4.2.1.3. Chu augmentation oi jiajie with semantic determiners The second step for graphical development is the addition of a semantic determiner to a phonetic loan graph, which yielded an SP graph specified to represent the word for which the graph was borrowed. This operation did not take place for all loan graphs. The loan graph came to be firmly associated with the word for which it was loaned; thereupon, a new graph was formed by augmenting it wi th an S E , and was assigned to the word the loan graph 15 It has been noted that the grapheme s/m SL was used by the Chu instead of zhi % . (Liu Binhui et al. 1991, 378) There are twelve different Baoshan graphs which containsM five of which can find their counterpart graphs composed of zhi % in their modern characters, and four of which can find their Qin counterparts composed of zhi %. 16 The graph H appears on two pieces of the sword known as Wu king Guang's sword (Wu wang Guangy/an ^.'S.^tM) as a variant graph ofjian M. This Chu graph is also treated in § 4.3.3. 76 originally represented. For example, a graph was believed to have been a depiction of a snake, but later became exclusively associated with the word 'other' through jiajie; the word 'snake' was then written with the graph she 4& which was augmented with an SE ± hui ' insect'. Such assignments o f a loan graph to a word, being carried out differently in each region, possibly contributed to the regional discrepancies visible in the SP graphs. The following 6 loan graphs are examples of such regional assignments: the Chu used SP graphs composed o f jiajie graphs and semantic determiners, while the Q i n used unaugmented bare jiajie graphs. Both the original depiction o f the graph and the word for which the graph was loaned are given, separated by " •• > ". 1. The graph P kou ( fe iP) 'mouth' was added to the following jiajie graph, although it is unknown whether there is a semantic relationship between them. 1) W bXng (|#IP *pjiangx) 'horse buttocks [viewed from behind] ' 1 7 ••• > 'the third o f the ten Heavenly Stems ' Chu: W l i Qin: W W 2. The graph ± fu ( & IP) 'earth' was added. 1) % v/ (IH IP *rid) was written with the graph shi f. f shi (IHIP *skhrjid) 'corpse' ••• > '(eastern) barbarian' Chu: S i Qin: 3 1 ^ 3. The graph 0 ri ("If IP) 'sun, day' was added to the following graphs, which were borrowed to represent the words semantically related to time. 1 9 1) 4- dong ( + IP *tengw) 'spinning device' — > 'winter ' 2 0 1 7 Several proposals have been made for the object the graph pf depicted, i.e., [horse] rump (Takashima 1996, v. 1. 205), human buttocks (Todo 1965,439-441) and talbe (Ogawa et al. 1967,19). 1 8 There is another Baoshan bamboo-slip graph which contain W as 'fS which does not have a modern equivalent graph, which suggests that the attachment of the graph kou P to blng W was established. 1 9 The addition of the graph B to the graphs used for seasons have been recognized, for example, by Ceng Xiantong f g i | (1993. 72). 2 0 The graph which was borrowed is not yet known. The shape of the graph is explained as being a 77 C h u : ^ Qin : 4- % 2) jfc chen (;5tiP *djteri) 'bivalve' — > 'the fifth o f the twelve Earthly Branches' C h u : # ; | i ^ J * Qin: Jkjg.21 3) I xid (fc IP *gragx) 'the name of dance/music' 2 2 ••• > 'summer' Chu:^f Sf Q i n : K 41 • f l | 4. The graph J-_ zhi (^ .Ip) 'foot' appeared to be added to the following graph which represents the word semantically related to activities related to movement implied by 'foot'. 2 3 1) $L Idi (i:!P *hg) 'wheat' - > 'to come' Chu: $ 2 4 Qin: * $ $ In the W Z bronze graphs, we find (Lai Z/H #f) which is close to the Chu form and the unaugmented Q i n form ^ (Hu ding jfij). 4.2.1.4. Different graphemes used at SP formation In the following 4 cases, the Chu and Q i n corresponding SP graphs were formed from completely different graphemes. 1) Chu: BE (£ 'a kind o f a bronze vessel ' 2 5 : P E -£ gu (fc IP *kwagx) 'ancient' spinning device in the SW. 2 1 The graph ^ appears in the Shuhudi bamboo slips rishu, but was not used for the cyclic date. It is surmised to be borrowed for the word 'lips'. 2 2 The graphxj'a J. appears iaLiji (The Book of Rite). The word the graph xia J[ represents is explained as the name of dance and music (Morohashi 1986 v3. 304, 312). 2 3 The graph zhi it is equivalent of the modern character zM ib 'foot'. 2 4 This graph ^ is used as a constituent in a multi-element graph ^ . 2 5 The graph EE was equivalent to the modern character f& $ because SE is used to used in the inscription on the Chu bronze vessel Chuwang Yinyue fu fti % $ %. which is recognized as ju referring to that vessel. 78 S E C ( M ) ' b o x ' S W : jg.fi (#> IP *pjagx) ' a kind o f a bronze vessel' P E ^ Ja ( I P *pjagx) 'honorific suffix to a man's name' S E ft zAti (5ft IP) 'bamboo'and S E JL min ( I f IP) 'vessel' 2) Chu: ^ | f jian 'cocoon, s i l k ' 2 6 P E jJL jian (JG IP *£z'an/i) 'to look ' SE jfc 5i ( A l p ) 'si lk, thread' Qin : ]$[ ^ jian (7EIP fo'a>u) 'cocoon, s i lk ' : H Y with a P E rfV7 H Y &hui ( t i p ) 'insect' which produces jfc s i ( $ | p ) ' s i lk ' P E rfi (jclP *man) 'to be enough merely to cover its body ' 2 8 3) Chu: 5 | H 'breath' P E gfc7*(#IP **7'3<0 'to complete' S E X hud (#lp) 'fire' Qin: & ? i ( M **fc/au(&IP) 'feather' SW: $E71 «g (#IP *tsjing) 'flag made of an animal tail and feather' P E £ sTie/ig (# IP *sring) 'to be born' S E T (TCIP)'flag' 4.2.1.5. Different usage of semantographs 2 6 According to the SW, jian £$, is the SW guwen form equivalent to jian 2 7 The graph $f has been elucidated as an HY graph by the SW, and as an SP graph by Ogawa et al. (1967,791). According to Dr. Pulleyblank (personal communication), a graph with the initial consonant *m-might have functioned as PE for the word with the initial consonant *k-. 2 8 The meaning of the word the graph represents is based on Duan Yucai S %H (1735-1815) in the SW. 2 9 The Chu graph "ft which is augmented with a mouth element is treated in § 4.3.5. 79 As has been reviewed in Chapter 1 (§ 1.2.4.3), SP graphs were formed by augmenting a semantograph (biaoyizi) with a graph which indicated the reading of the word represented. There is one case in which the Chu and Qin used different semantographs and the same phonetic element to form an SP graph. 1) % gusui ( jc t f *kwan) 'hat, cap': PE % yuan (JE tf *ngwjari) 'head' Chu: % semantograph & 'cap', a depiction of a cap. SW: % ; semantograph (HY) which was composed of 'covering object' and-*}" cun 'thumb, unit of length' which appears to depict a cap being put on by hand. 4. 2.1. 6. Different usage of semantographs and PEs There is one case in which the Chu and Qin used different SEs consisting of several elements which appear to be related to the meaning of the word the graph represents and different PEs. Since it is difficult to determine whether this is a case of SP formation through jiajie or by augmenting PE to semantograph (HY graph), this is treated separately from the different graphemes used at SP formation (§ 4.2.1.4). 1) M. /fi (flgtf *ljidx) 'shoe, to walk' Chu: 1 | f PE X ye (% tf *git) 'head' SE # zhou (# tf) 'boat' which might be a depiction of boat-shaped shoe andihz/ii G£.tf) 'foot' Q in : J | | k PE f shi (J i t f *skhrjid) 'corpse' or | zi (||tf *sgjidh) 'sel f 3 0 SE ^ 'foot', and chi % (0 tf) 'to go' 4.2.2. H Y graphs There is only one case in which the Chu and the Qin used different H Y graphs. 1) Chu: I ^ 'face' W ( M ) 'head' and 0 jiit (#t f) 'mortar' 3 0 According to SW, the phonetic element was said to beshi f3. 80 Qin: W mian (§/ % (TGIP *mjianh)31 @ /nw ( # $ 3 ) 'eye' or W ( ^ $ 5 ) 'head' is inside of the enclosing line 4. 3. Divergent structural modifications of SP graphs Structural modifications o f SP graphs may involve the omission o f a grapheme, the alternation o f a grapheme, or the augmentation o f the original SP form with a P E , an SE or with an element the function o f which cannot be determined. The cases in which it is the Chu or the Qin form that can be recognized as having been modified w i l l be presented separately under each category. The earlier counterparts o f the graphs in question attest which graph was modified; various sources have been sought to determine the composition o f the missing antecedents. 4. 3.1. Omission of SP grapheme (A) Chu graphs There are 3 cases in which a grapheme of the Chu SP graph was omitted. 1) J $ yu (fcip *rag) 'carriage'; % yu (fcip *rjag) 'to lift with hands' and a car % cht (fc IP *kjag), both o f which are phonetically and semantically conformed to the graph they compose. Chu: IF ^ ; two hand graphemes were abbreviated.3 2 Qin: J$ ^ ; the earlier composition was preserved. 2) If gul (#lp *Jcwpd) 'to return'; originally as can be seen in the OBI $ | ; ^), it consists o f P E % (#IP *ted)33 and SE # sao (tM) 'whisk ' , and an optional grapheme LL zhi 3 1 According to the SW, this graph consists o f f (a variant graph oishou "it" 'head (above the neck)'), depicting the shape of a human face( cong sh ou xiang ren mian xing t^W^AW^)- The grapheme used in the Qin variant graph, mil @ 'eye' seems to be an abbreviated form of W • 3 2 Both # and appear to conform to the sound and the meaning of the word the graph in question represents. Originally this graph should have four hands holding a car for two reasons: (1) the old sound of this graph yu is the same as yu & 'to keep company, to ally', ;§£. 'to lift', all of which can be elucidated to be formed with % four hands signifying "to lift [an object] in cooperation" as both a PE and SE; (2) the two hands below an object signifying to hold that object were mostly used to represent 'to present offering s to the gods or deities', which was not congruent to the word the graph yu % represented. 3 3 The graph gui $ was earlier in the WZ period composed of | and sao # as compulsory elements, with chi ^ or A as optional elements. The graph § was used in the OBI to represent the word '[army] camp', 81 $3)'foot', but later, as can be seen in the W Z bronzes, it appears to have become augmented optionally with elements related to movement chl % (0%$) 'to go'. Chu: 3f % # % % % ; P E $ was omitted. Qin: I f § § 3) ^ xi (0$ *hjiak) 'fear'; as the W Z bronze graph 'ffy shows, this graph was composed of P E & (0% *khjiak) 'crevice in the wa l l ' and SE JH hu 'tiger'. Chu: ^ |f H ^ ; SE jfi hu ( M ) 'tiger' was reduced to p 'tiger's head'. S W : M (B) Qin graphs There is no known case of a Q in graph omitting a grapheme. 4.3. 2. Replacement of SP grapheme (O) Both Chu and Qin graphs There are 2 cases in which the replacement o f a grapheme took place in both Chu and Qin SP graphs. 1) i t cfo'fjM *drjid) ' s low' Chu: i% %l ; P E F shi ( M *skhrjid) 'corpse' or ' er ( J M ' two' SW: i t ; P E # x i ( ) M ***u ( J | IP) 'to beat' Qin : i§L 8) # gui (ItIP *)bv/aJA) 'noble'; PEt)3 (?) 'to hold upward' 4 2 and SE £ fee/ (&IP) 'cowrie ' ; the earlier composition can be seen in the W Z bronze graph ^ i t yi 'to lose'. Chu: J| |[ l=f ; PEE)3was replaced with an unknown element.4 3 Qin:#f f 44 This graph was used as a P E in the following Chu SP graphs, mamtaining the same composition 4 5 a) i t yi 'to lose' Chu: i t *jf" Qin: i t There are three more Baoshan graphs [J. gw* 'box ' , ^ ^ 'k ind o f tree' and | f %| ^ kui 'to present [food]', which contain the graph J| (f ^) as a constituent. These graphs do not find their counterparts in the Qin graphs. 9) |& gui (^IP *kj9gwh) 'k ind of ritual vessel'; this graph does not appear in the Baoshan bamboo-slip graphs, but was used as a P E in the graph j$tjiu 'stable'. We can find the oracle-bone antecedent g r a p h ^ In the Chu form, left grapheme |? was replaced by another graph shi ($&IP *djzk) 'to eat', and the right grapheme shu & by another grapheme & pu (it IP) 'to beat'. The graph | £ with shi & instead o f H can be 4 2 The meaning of this graph is based on Ogawa et al (1967,955). 4 3 The upper element of j | , § and ^ has been transcribed as ^ in this thesis, although there is no such modern graph. The element which is similar in shape to this element also can be found in the Baoshan graphs, =j and ^ ( j j [ which has been transcribed as 4t bian 'head dress' by Li Jiahao # (1986, 391-395)), # ( # ) a n d ^ ( # ) . 4 4 This Qin graph is also treated in§ 4.5. 4 5 In addition to the graphs given here, there is also JP which does not have a descendent in modern characters. 85 found in some WZ bronze vessel inscriptions, such as"§=^ in Jihou gui B H. %. a) ML jiil ( M *kpgwh) 'stable'; P E gfc g i i l (ft $ *kjsgwh) ' k ind o f ritual vessel' and SE f. (ft $ ) «house on the cl i f f » 4 6 Chu :Jgt$p41; PEg^was used for P E of jiu Q i n : M. fe ; P E | £ was used for P E of jiu Jf. 10) # foio ( # $ 'badger': P E 4- ge (#$ *klak) 'each' and S E & zhi ( & # ) 'to crawl, as feline beasts on prey or as reptiles'; (WZG^tjT) Chu: m ; S E z / u 4 was replaced with S E i, s7m ( & $5) 'mouse'. Qin : & ffr 11) i f / a ( ^ $ 3 *kwjapl>*pjap) 'model, l aw ' ; P E -£ #w ( & $ 5 *khjagx) 'to depart' and two SEs, S 7 W I ( M ) 'water' and H zAi (TGHO 'deer-like animal ' ; 4 8 (WZG f f ) Chu: | 0 ; i was replaced with jifv. (&3P *£/'##) 'man ' which might have served as P E . Q i n : > f ^ i f (B) Q i n graphs There are 5 cases i n which the Q i n form o f a SP graph replaced one grapheme with another. 1) # si ( £ .$ *rj9gh) 'government office'; P E ± zhi ( £ H S *tj3gx) 'foot' and SE X you ( i : $5) «to hold in hand» ; (WZ ^) Chu: ^ % 4 9 Qin : ; cun (%$>) 'thumb, unit o f length' replaced you %. 4 6 The SW explains tins graph $ as consisting of PE |£ and SE f. 4 7 This Chu graph is also treated in § 4.5. 4 8 According to the SW,/a i# consists of three SEs shui j , z/iz M and AM i . Although the SW identifies it as an HY graph and the graph qu i does not conform to the sound of/a jf, qii is acknowledged as a PE. According to Dr. Pulleyblank (personal communication), a graph which had *k- might have served as PE for the word with an initial *p-, even though these sounds were not articulated in the same place of articulation. 4 9 The Chu graph in question was also modified by an augmentation with an element. This will be treated in § 4.3.5. 86 A s a graph which consists o f vow X was used in the W Z bronze vessel Shibo gui the graph in question originally consists o f you X instead of cun ^f.50 This form is used in the following graph. a) # ding (ZM *tegx) 'rank; to classify ' ; P E # si ( i l t f *rpgh) 'government 2 ) B# sMOttf *dj&g) ' t ime ' ; P E i t z M CtiP **/agx) 'foot' and SE 0 ri (JTUO 'sun, day' Chu: # ^ -Qin : B*f ^ ; P E zhi i t was replaced with another graph # si (iliP *rpgh) 'government office [place]' It has been proposed that the graph © 0 ri 'sun, day' was used to write the word B# shi ' t ime' in the O B I . 5 1 We also can find the oracle-bone graph # & which has been deciphered as the prototype o f shi B#. In the development o f the graph for ' t ime', it can be surmised that the graph 0 ri 'sun, day' was used to write the semantically congruent word 'time' (parasemantic usage) and zhi i t was added as a phonetic determiner, yielding the SP graph # to represent specifically ' t ime' . The Q i n form can be considered as a case in which zhi i t was replaced with another graph si # which also could serve as P E . 3) # fu ( f c lp *phjag) 'to spread'; P E $ ju (fcU5 *pjagx) 'honorific suffix to a man's name' and SE -f cun (JtiP) 'thumb, unit o f length'; ( O B G Originally, as can be seen in the W Z bronze graph, the graph in question was composed of Ju i f and X you (iliP) «to hold in hand», but the Q i n used cun Tf" instead of you X - 5 2 This 5 0 Tang Yuhui (1986,40) notes that there is no extant late WZ materials carrying the graph cun ->f. 5 1 TakashimaKen-ichi's work in which this usage of the graph ri 0 is discussed will appear in a web-site publication at the University of Oslo, Norway. 5 2 In addition to the graphs treated in this section, there are 4 more graphs in which the Chu used youX and the Qin used cun -f. office [place]' and SE % zhu (ft IP) 'bamboo' Chu: tr ^ Qin: ^ ; cun replaced vow X-SW :H 1) iiju 'to give' 2) )ft Ju 'treasury' Chu:Tt; Qin: ^ Chu:f ; Qin: i f 87 discrepancy between the Chu and Qin graph can be surmised in the following graphs, which usedfu # as a PE. 5 3 a) # bo *pwak) 'to arrest' Chu: ^ $ SW: # b) i t 7u ( & $ W a g ) 'to spread' Chu: % SW: ft 4) % chudng (jclp *drjuanx) 'to transmit'; PE # z/jua« (jciP *tjuan) 'exclusively' and SE Aren(%%>) 'man'; (OBG iff) Chu: Q i n : ^ A s its OBI and WZ bronze antecedent graphs show, this graph originally used the element you X> as in the Chu form. The Qin used the form whose vow X was replaced with cun TT-5) M wen *mjm) 'to hear'; PE & nun *hman) 'dusk' and SE er%-' e a r ' ; ( W Z G ^ ) Chu: # 15 Q i n : « If Although the oracle-bone graph in question was an H Y graph which was composed of an ear and a man, the WZ bronze graphs consist of a PE nun f£ and an S E er ^ . The Q i n graph replaced hun& with another PE f"J men *m$ri) 'gate'. 4.3.3. Augmentation of SP with extra PE (A) Chu graphs There are 3 cases in which the Chu usage o f an SP graph was augmented with an additional PE not present in the Qin form. 1) M yia flfclP *krar) 'good, to praise'; PE fa jia (flip *krar) 'to add' and SE £ ( M ) , the original graph ofgu i& 'drum'; (WZG ) Chu: ; an extra PE ^ he (ife$5 *gwar) 'grain plant' replaced the SE £ . 3) # 'to obtain' Chu: ; Qin: 4) # ru 'to weed out grass with a shell, disgrace' Chu: ^ ; Qin: ^ These graphs are treated in § 4.4.2. 5 3 There is another graph # W (M. £P *pjagh) 'to beat' in which/u I? is also used as a PE, but since there can be found no Qin counterpart graph nor its antecedent graph by which the earlier form can be attested, it 88 Qin: # 5 4 The following two graphs have not been found in Shang and W Z periods. Recall from Chapter 1 that one P E and one SE were used, when SP graphs were formed. Since the graph qian fc serves as P E , and the combination of qian fc and gan # cannot be found as a graph, the grapheme gan # is treated as an additional element. 2) lim (fttf *ljiam) 'to gather'; P E fc qian (fttf *tshjian) ' a l l ' and SE k pv. 01 tf) 'to beat' Chu: # § t ; an extra element if" gan (l|£tf *kam) 'sweet' which conforms to the sound o f the graph in question. S W : & 3) M jian (fttf *kljam) ' sword ' ; P E fc qian (fttf *skhjian) ' a l l ' and SE !) aao ( f tf) 'knife' or &jin (fttf) 'metal ' 5 5 Chu: tf H>$ ; an extra element # gan (tfctf *kam) 'sweet' which conforms to the sound of the graph in question. 5 6 Qin: M (B) Q i n graphs There is only case in which the Qin used an SP graph augmented with a P E not present in the Chu form. 1) ^ J>e (fc tf *riagx) 'uncultivated land'; P E ± ta (fc tf *tagx) 'earth' and SE 7fc mil (1£ tf)'tree'57; ( O B G | J J ) Chu: ^ > preserved its original components. Qin: j £ | * ; P E f yu (fc tf *ragx) 'to give' was added, is not included in this section. 5 4 This graph is from the Qin bronze artifact Shiqinian chengxiang Qizhuang + 4j # t . 5 5 The graph $ appears on two pieces of the sword known as Wu king Guang's sword (Wu wang Guangy'j'an ^i^fc j#J) as a variant graph of jian ffl. This graph is also treated in § 4.2.1.2. 5 6 There is another Chu graph which contains qian but does not include an additional grapheme gan This graph %. also appears on the Wangshan bamboo-slip inscriptions, but has no equivalent in modern characters. 5 7 Qiu elucidates this graph as an HY graph as the SW does. Since the OC pronunciation of the 89 4.3.4. Augmentation of SP with extra SE (A) Chu graphs There are 9 cases in which the Chu form was augmented with an additional SE not present in the Qin form. 1. Addit ion of graph ± ru (fc tf *tagx) 'earth' The graph tu ± appears to have been added to the graphs which represent words semantically related to 'earth' (Takada [1925] 1975; Q iu 1980). There are 4 graphs thus augmented. 1) Tjf shi ( i l t f *dji9gx) 'market'; P E & i and SE ^ ' ? ' 5 8 ; ( W Z G ' j * ) C h u : ^ # # 5 9 Qin: "ijf jh 2) ML chen ( ^ tf *c?r//n) 'state name in the Zhou time'; P E jft c/zen (^-tf *drjin) 'to display' which was reduced to % dong 'east', and SE^w | (ft tf) ' h i l l , ladder' C h u : ^ ^ ^ ^ ^ 6 0 Qin: | * P% According to Wang Guowei (1968, 7350), the form of the graph chen % evolved from ?$[ to and U- (an elaborate form of W). The earliest sample of the graph can be found in the S A bronze vessel inscriptions, such as ^ i n Chenhou gui PJc^ Ji-3) It ling (^.tf *ljsng) 'to ascend'; P E ^ ling (MM *ljzng) 'to transgress', and SE § fii (ft tf) ' h i l l , ladder' 6 1; ( W Z G \ % Chu: i f f # Qin: it ^ grapheme ni i conformed to that of the graph ye If, this graph is treated as an SP graph in this thesis. 5 8 The meaning of this grapheme is ambiguous. The SW elucidates as fence. According to the dictionary Shinjigen ff^if , it means 'even, flat'. 5 9 The graph shi Op with nl i has been pointed out by Qiu (1980). 6 0 The graph |£ was used as the state name Chen 1$. which existed between the beginning of the WZ and 479 BC in present Henan Province, (jf^f 6 1 According to the SW, this graph consists of PE ling and SE fit % . No other reliable elucidation 90 4) Mydng (|#|p *rjang) 'south side of a h i l l ' ; P E # ydng(B^ *rjang) '[sun] to rise' and SEya | (ft lP)'hill, ladder'; ( O B G {?) Chu: m l§ ft g Qin: $?r%f% It has been recognized that the grapheme/w | tends to be accompanied by / i i i (He 1989, 200). Three graphs listed above are in all extant cases composed of both fii % and nl ± . O f other Chu graphs, $\ f ^ ) , ft ( E 2) , r$ Of), (m), ® 0$), K (SS), It (S), and If (&) are not known i n any instance which includes tu ±. 2. Addit ion o f other graphs 1) If gui (#lp *&W/3G0 'to return'; originally as can be seen in the O B I ^ ) , it consists o f P E $ *ted)62 and S E f sao (ft IP) 'whisk ' , and an optional grapheme ^ z/zi (^ IP) ' foot ' , but later, as can be seen i n the W Z bronzes, it appears to have become augmented optionally with elements related to movementc/z? % ($^ |p) 'to go' . Chu: i | % # # % & ; 4 chi ( $ $ S ) 'to go' was added. Qin : # § § 2) / f r /u (feIP *£/ W gx) 'treasury'; P E # / « (felP *pjugh) 'to give ' and S E T (ft IP) «house on the cliff» or ^ midn (jciP) ' r o o f 6 3 Chu: jJ X bei (^ IP) 'cowrie ' was attached. This graph was used for the graph related to treasure and money, such as bao 'treasure', cdi #t 'goods, material wealth' and hud ^ 'goods, money'. Q i n : ) ^ 6 4 The graph 6e7 M does not form an independent graph with ji fii or f~ (or ^ ) . The graph bei X should be an additional element added after / u M was formed. 3) UL bai (## *pradh) 'to defeat'; P E £ bei ( & $ *pad&) 'cowrie ' and SE jc /ra (^ IP) 'to beat' has been proposed, this graph is treated as SP in this thesis. 6 2 The graph gui If was earlier in the WZ period composed of |§ and sab #• as compulsory elements, with chi % or A as optional elements. The graph $ was used in the OBI to represent the word '[army] camp', but what the graph $ depicted is unknown. 6 3 This graph is also treated in § 4.2.1.2. 6 4 This Qin graph is also treated in § 4.4.2. 91 Chu: Jft $ f f ( 6 5 Q i n : J f c & f t i The Chu graph in question consists o f two 6e* X and one />u & , although the top Z>ez X to another grapheme mzi @ (:ft;iP) 'eye'. This is a variant graph (zhouwen $$30 included in the S W under the entry of bai The Qin form can be found in the O B I as and the Chu form in the W Z bronze Shishi gui &fc$jE. j | as 4) # wei (HIP *gwjadh) 'to protect'; P E # wei (ItIP *gwj9d) 'tanned leather' and SE ff xz>zg(|#iP) 'road, to go' ; C h u : ^ % v ; i t 'foot' was attached. Q i n : % ^ f Both the Chu and Qin form can be found in the W Z bronze, a s ^ f in W e i zun % % a n d ^ ^ in W e i gui % %. The Qin form can also be found in the OBI . 5) fc pu (MIP *buk) 'servant'; P E J | (M IP *buk) 'vexatious' and SE A ren 'man' Chu: i f ^ $ ; E c/zen 'subject' was attached. Q i m f c f f The graph c/zen E. which does not constitute an independent graph with ren A or H should be an element added later. (B) Q i n graphs There is one case in which the Qin used an SP graph augmented with an additional SE not present in the Chu form. 1) & wang (|#lp *gwjang) 'to go'; it was originally composed of P E i£ wdng (|#!p *gwjang) ' k ing ' and SE i t zhi ( i l lp *tjogx) 'foot', as j£ and £ which can be found in the OBI . Chu: f 67 Q i n : ^ ; SE ^ chi ( $ $ S ) 'to go' was added. 6 5 This graph is also treated in § 4.5. 6 6 This Qin graph is also discussed in § 4.5. 6 7 This Chu graph is also treated in § 4.3.2. 92 4.3. 5. Augmentation of SP with grapheme of unknown function (A) Chu graphs There are 2 cases in which the Chu graphs were augmented with a grapheme whose function is unknown. 1) if qmg ( # t f *tshing) 'blue'; PE £. sheng (#tf *tsjingx) 'to be born' and SE # dan (jctf) 'cinnabar, red'; ( W Z G rS) Chu: "f ^ ; D £dw ( l£ t f ) 'mouth' was attached.68 Qin:* f The graph "S" was used as a PE in the following graph, a) ft qing (#tf *tshing) 'to request' Chu: f# Ojn: ft 3f tff at $ 2) # si ( i ltf *rpgh) 'government office'; PE ±zhi ( i ltf *tpgx) 'foot' and SE X you ( i l tf) «to hold in hand»; (WZG ^) Chu: ; D kou ( M t f ) 'mouth' was attached to three graphs out of five. Qin: f 69 This form is used in the following graph. a) # ding ( i ltf *tegx) 'rank; to classify '; PE # si (£.t f *rpgh) 'government office [place]' and SE ft z/m (ft tf) 'bamboo' Chu: ft ^ Qin: (B) Qin graphs There are 2 cases of Qin SP graphs augmented with graphemes of unknown function not present in the Chu form. 1) \% shmg (Htf *sjang) 'to hurt'; PE # ydng (|#tf *rjang) '[the sun] to rise' and SE A 6 8 As was discussed in Chapter 1, Botlz (1994, 1030105) proposed the alternate reading of D *mjing). In this case, this alternate reading coincides with the reading of the graph to which it was attached, but we cannot assume that it served as PE because we do not know whether this alternate reading had yet fallen into oblivion at the time the graph D was added to the graph in question, if it had had such an alternate reading. 6 9 This graph is also treated in § 4.3.2. 93 r e / i ( £ i P ) 'man' Chu: f j §>> Qin: jjfr 1% ; the top element above the grapheme ydng # was added for reasons that are not clear to us. We do not have the earlier graph in question. The top element above the grapheme ydng # is an additional element put after the SP graph was formed wimydngjj? and ren A according to the principle of SP graph formation (one P E and one SE). Since no graph consisting of ydng # with the top element has been found, it was probably not used as P E . 2) i% de ( I tIP *te£) 'virtue'; P E % zhi (J$|p *drpk) 'straight' and S E /C? xin ( f t iP) 4.4. Divergent structural modifications of HY graphs The structural modification o f H Y graphs includes omission o f a grapheme, replacement of a grapheme, and the augmentation of a P E , an SE , and an element whose function is unclear. H Y graphs whose modifications can be recognized in the structure of the Qin and Chu graphs have been grouped together respectively. 4.4.1. Omission of HY grapheme (O) Both Chu and Qin graphs There is a graph whose graphemes are not fully preserved in both the Chu and Qin graphs. 1) # qin *dzjin) ' Q i n (name o f the state)'; a pair o f hands X you ( i l p ) holding a pestle wu (fc, IP) is pounding on grain plant ^ he (HfciP); ( O B G \ \ ) 'heart SW: ^ ; SE ^ chi (0 IP) 'to go' was added for reasons that are not clear to us. 70 Qiu (2000, 204) explains that % is the original graph for |% de 'virtue' 94 (A) Chu graphs There are 2 cases in which the Chu used abbreviated form of H Y graphs. 1) lp s A i ( J M *srjid) '[army] camp'; SE $ (#tf) '[army] camp' and SE fp (^tf) 'to surround' 7 1; ( O B G ^ , W Z o ^ ) Chu: fp ^ ^ ^ ; P E | was completely omitted. Qin : 0fp Bfl5 ff5 ; P E $ was preserved except for one case out o f five in all. 2) 0 qi (IH $5 *khjidh) 'to abandon'; two hands holding a basket containing an upside-down child about to be abandoned; ( O B G JaJ<). Chu: -f^f ; the basket was omitted. Qin: ^ ^ ; the original composition was preserved. (B) Qin graphs There is one Qin H Y graph exhibiting the omission o f a grapheme or graphemes 1) iC yi (ffctf *ngjar?) 'fitting, right'; meat in the container or stand originally denoted a certain sacrificial ceremony; ( O B G j | j ) . Chu: ^ ^ ; although the form of a container or stand was reduced, 7 2 it preserved two elements o f meat. (Jin: ; one o f two elements meaning meat was omitted. 4.4.2. Replacement of HY grapheme (A) Chu graphs There is 1 Chu H Y graph whose grapheme was replaced with another grapheme. 1) % kou ( M tf *khugh) 'to rob'; 'to rob' was rendered by midn (% tf) 'house', jt yuan ( j c t f ) «head» and & / ? u ( J i t f ) 'to beat'; ( W Z G $(1) 7 1 Although its original identity is unknown, this graph 0 appears in the OBI for the word meaning "[army] camp", and the grapheme ip (3S$5 *tsriap)1 'to surround' was added, probably as a semantic element, as seen in the Western Zhou bronze-graph. % can be the phonetic element for shi $p, because (1) the former appears as a single graph representing the word shi If '[army] camp' in the OBI, and (2) some graphs belonging to thezhi | | rhyme group and wei %. thyme group are rhyming together in the Shijing (Li Fang-kuei [1971] 1980, 64-66). 7 2 This feature is also treated in § 4.5. 95 Chu: ; pu k was replaced with ^ ge (Ufc^ P) 'dagger-axe'. Qin: ^ ; the composition which consisted of tXt, jc and k, was preserved. (B) Q i n graphs There are 3 cases in which graphemes in Qin H Y graphs were replaced. 1) tt /w (fe£P *pjugh) 'to give' ; H Y ; the S W explains it as consisting o f SE A ren 'man' and ~f cun (Xty) 'thumb, unit o f length'. W e find two variant graphs in the W Z bronze graphs as ^ (Rui ding ^ which consists o f X you «to hold in hand» and ^ (Hu ding 3^$) which consists o f cun -f. Since the graph cun which was not used in the W Z bronze graphs other than this case, the Q i n form with cun appears to have been a modified form Chu: 11 ft Qin: # ^ The Qin used the graph which was composed of a man and cun -f, and the Chu used the one composed of a man and vow X- This graph Jit tt was used as a P E in the following graph. The discrepancy between the Chu and Qin was maintained, a) foju (fe£P *pjugx) 'treasury' Chu: 7 3 Q i n : ^ ^ 2) # de ($fc$5 *tek) 'to obtain'; a shell X bei 'cowrie ' is held in hand vow X; ( O B G 1 % C h u : ^ Qin : ^ 7 4 3) # H i (Ji$P *njuk) 'to weed out grass with a shell, disgrace'; ^ chen {%%f>), a pictograph of a bivalve, is held by a hand you X; ( O B G fi^)-7 3 This Chu graph is also treated in § 4.2.1.2 and 4.3.4. 7 4 This Qin graph is also treated in§ 4.4.5. 7 5 The SW defines rfl # as 'disgrace'. This graph does not appear in the OBI, but there is a graphic which is explained as an antecedent graph of nbujjg^Xn weed out grass' by Qiu (1992), and Wen Shaofeng JS-^ I? and Yuan Tingdong (JGWG v.2: 1139). Although the graphru # itself does not appear in the excavated materials, it can be surmised to have originally consisted of chen Jfc and you X -96 C h u : ^ Q i n : f 4.4.3. Transformation of HY to SP (A) Chu graphs In these 6 cases, the Chu H Y graph became an SP graph by being augmented with a P E , having an SE replaced with a P E , or modifying the form o f an SE into a P E . 1) JTL xiong ((#$3 *hwjang) 'elder brother'; a mouth element and kneeling figure; ( O B G \ 2) H wang (!#$$ *mjangh) 'to look from afar'; a man who is standing on the ground ting (#$0 'to stick out, stretch' with the emphasized eye which was interpreted as the not occur. 3) >f Hang (|#$3 *ljangx) ' a pair o f ; the original identity appears to be a depiction o f an object in pair, for example a depiction o f two same graphs Mng ^ «horse rump» 7 8 ; ( W Z G 4) & he (M-^ *g?p) 'to jo in ' ; 'to jo in ' was denoted by a mouth-like element which depicted a container and an element on top of it which represented its l id ; ( O B G £ ) 7 6 The graph \% is the earlier form of the graph & wang 'to go', although the lower part was modified. Chu: ^ ^ ; P E wang (^ $5 *gwjiang) 'to go' was added. Q i n : < 76 C h u : ^ f ; part o f the graph was modified and changed into a graph which served as P E ^ ydng (|#£P *rang) 'sheep'. Qin : ; the original form was preserved. 77 This Qin graph in question is treated in § 4.6.2. 78 This explanation was proposed by Ling Yiguang (JWG v.9: 4852). 97 Chu: # ; P E hdn the added element possibly is ^ ^ hdn (ft IP *gsm) 'to keep in the mouth, contain' which is congruent phonetically and semantically. 7 9 Qin: the original composition was preserved. 5) 3L jmg IP *king) «warps which are put on a loom» The lower part o f this H Y graph 3. which was written as 5 hi the W Z bronze vessel inscriptions, in the Chu, was altered to ££. ting (#IP *thingx) 'to stick out, stretch' to indicate the reading. This altered Chu graph j | was used in the following SP graphs, while the Qin counterpart graph was preserved in its original form. 8 0 a) Wfmg ( # | p *kjingh) 'strong' Chu: gr j)f SW:$? b) # y'wgOM *kingh)' a kind o f tree' Chu: ^ S W : # c) M fing (#IP *kjingx) 'neck' Chu: H | ? Qin: % * f 6) I ./w {% IP *6/3Jbv) 'to return'; an unknown object and a foot ^ ; (OBG§ ) 8 1 This graph is used as a constituent o f the following graphs. In the Baoshan graphs the upper element, an unknown object, was replaced with a graph | f you (ftIP *ragwx) 'wine vessel' which is phonetically conformed to the graphs in which it is a constituent. a) fu ( f l i p *bpkw) 'to return' Chu: i | % 8 2 SW: 7 9 Qiu (2000, 218-219) surmises that the graphs, a depiction of a mouth holding some object, might originally have been used represent the word 'to hold in mouth'. Until the graph hdn & was formed to write 'sweet', the graph ^ might have been used to write both words 'to hold in mouth' and 'sweet'. It is not known whether the graph hdn was already in use in the Chu area during the WS period. 8 0 Since the form of another Baoshan graph $M seems to be equivalent to £f jmg (#$3 *king) although the right grapheme 3L 'warps' is abbreviated, it is not included in this section. There is another graph which contains the grapheme § , as 13. 8 1 The object written above a foot has been variously surmised as (1) an abbreviated form of guo fp 'outer wall of the city', (2) a kind of vessel, or (3) a cave for dwelling. (JGWG v. 1: 862-863) This graph is taken as an HY graph in the present thesis, for the grapheme which can serve as PE for this graph has not been clarified. 8 2 This graph is also treated in § 4.6. 98 b) H fit (ft $ *ppkw) 'abdomen'. Chu: i t 3f Q i n : M (B) Q i n graphs In these 4 cases, the Q i n H Y graph became an SP graph by being augmented with a P E , having an SE replaced with a P E , or modifying the form of an S E into a P E . 1) # sa n (ft^ p *s9tn) 'three'; it was evolved from the graph o o representing stars in the O B I . Chu: (51,; preserved the earlier composition of the graph in question. Qin : % ; P E ^ shan (ft£P *sr9m) 'something drawn by a brush' was added to denote the reading of the graph. 2) g? jit *krid) ' a l l ' ; two man elements A ren (£3P) and # gan 'sweet'; ( W Z G ^ ) 8 3 Chu: ^ ; preserved the earlier composition o f the graph in question. Qin: H1; the top elements are replaced with two fcfc bi (ISi^ p *pjidx) 'to compare' which was congruent to the sound of the graph. 3) it.yuan (jc^P *gwjanx) 'distant'; P E ^yudn (jc^ *gwjan) ' long gown' and SE ^ representing the movement The H Y graph yuan serving here as a P E , appears in the O B I with and without the element O, as Y and . According to Q iu (1985), the grapheme O was later added to indicate the reading o f this graph. 8 4 The Chu graph does not include this extra grapheme, while the Qin graph does. C h u : $ 8 5 8 3 The graphjie ^ is elucidated as consisting of two men and bai 6 'white', but the WZ bronze vessel graph has an element which appears asga« i f instead of bai ^. 8 4 According to Qiu, this grapheme signifies 'round' which is represented by the modern character yuan Ji 'round' (TCIP *gwjan). 8 5 This graph appears in the Baoshan bamboo slips without O, but in the Guodian fp bamboo slips produced in the Chu state during the WS period it appears withO- Although the grapheme O has a phonetic function, it seems optional for the graphs yu an 3 & and hudn i t . . Another graph hudn Ht might be the same case as hudn ii., since both are composed of the same phonetic element^ 99 The following two graphs i t . and 3g. are composed of yuan a) 4 A w a « ( 7 t t f *gwfr? ;a«) ' to return' Chu: 8 6 Qin:g| b) 3$[ hudn (7c tf *gw(r?)ari) 'ring-shape jade' Chu: SW: * 4) 4 ftwg(J^.tf *dung) 'servant; a servant who is tattooed with a kind o f needle JfL Chu: f [ f ; preserved characteristics o f the O B I graph' ^ by 'needle', an 'eye' with a man's figure standing on the ground' Qin : J § ; the middle part is modified into dbng (^.tf *tung), which serves as P E . 8 7 A s shown below, the Qin counterpart preserved its quality o f graphic representation. a) # zhong (^.tf *tjung) 'to grow (grain)' Chu: $ff Q i n : * | b) tf zhong (Jfctf *tjung) ' be l l ' Chu: f f Qin : § 4. 4. 4. Augmentation of H Y with extra S E (O) Both C h u and Q i n graphs There is one case in which an H Y graph developed into two variant graphs, each o f which was augmented with a different graph from the other. The Chu and Qin used these two different augmented graphs as shown below. 1) ifcyou (f t tf *regw) 'to s t rol l ' 8 8 Chu: ^ ; S E i _ (tf tf), an element signifying movement 8 6 The Chu graph in question was written both with and without the element O . 8 7 This addition of dong %. can be seen in the WZ bronze inscriptions. 8 8 It is unknown what the original graph consisting of -?- zi 'child' and f~ 'flag' represented, since it was used to write the place name in the OBI. (JGWG v.4: 3059-3060.) This unaugmented graph which depicts a man holding a flag is treated in the present thesis as an HY which renders 'to sway' as stipulated in Shinjigen (Ogawaetal. 1967). 100 Qin: $F SE s/*wi (Jgtf) 'water' Originally, the graph in question consisted of -J- z i ( i l t f ) ' ch i ld ' and f~ ' f lag ' , as seen in the O B I jf^ ". This graph was later augmented with a graph. The Chu form was augmented with the semantic element which was used in graphs representing actions involving movement, such as 'to go', while the Qin graph was augmented with the grapheme shul fa. (A) C h u graphs There are 6 cases o f Chu H Y graphs being augmented with an additional graph of related meaning. Graphs in which a segment o f a grapheme was modified into an independent grapheme have also been considered as having been augmented, and are treated in this section. 1) ^ zha ( f c tf *dzragh) 'to make'; originally written as % or in the O B I . The Baoshan graph in question was augmented with X you (^.tf) which is semantically related to 'to make'. Chu: ^ SW: ^ This augmented graph is also used in the following Chu graphs, while the Q i n counterpart graphs used the graph without a hand element. a) # zud (fc tf *tsagh) 'to make' Chu: j% f% f-f $ Qin:l£ % b) 0 zud (fc tf *dzagh)'sacrificial meat' Chu: $f S W : # 2) ~%L ji (£4tf *gj9p) 'to reach'; H Y graph which denoted "to reach" by a man A ren {% tf) with a hand X you ( i l t f ) catching him from the behind, as \ which can be found in the OBI . Chu: i# ^ ; two SEs % chl (tftf) 'to go' and i t zhi ( i l t f ) 'foot' were added. S W : & W e can find both forms in the W Z bronze vessels: (|p| in Fushu xu #,>fe3I, and ^ in Bao you 1 0 1 3) # C qiu (ft $ *tshpgw) 'autumn'; fc he Oft UJ) 'grain plant' and fc Awo ( # $ ) ' f i re ' 8 9 Chu:f0 ^ ; 0 rc(M) 'sun, day' was augmented and the grapheme hud fc was reduced to two strokes or omitted. 9 0 Q i n : & % ^ 4) f O M *phjingh) 'frank words ' 9 1 ; & j;ou (ft IP) 'from' and (ft IP)' ? ' 9 2 This graph was used to write the word ft />z«g 'to go to inquire' in the Baoshan bamboo slip inscriptions. Chu: | f ^ ^ : a graph P £dw (fe$5) 'mouth' was added. SW: f t ( f ) Among the W Z bronze graphs, we can find both the Chu and Q i n forms on the W Z bronze vessel inscriptions; ^ on Fansheng gui #-^. j l and IP on Maogong ding % fe^i .93 5) jfc hudng (li^P *gwang) ' ye l low' ; ramified and evolved graph from a pictograph o f an arrow or a crippled man ( O B G ft $ f )- 9 4 Chu: % ^ ^ ^ ; the lower part (the part depicting an arrow feather or the legs o f a cripple) was modified into the graph fc hud (#$P) 'fire' which can be semantically related to 'yel low' , i t Qin: ^ ; the modification of the lower part did not take place. 8 9 The OBI graph representing 'autumn' was written with a pictograph of an insect which has been surmised as being a locust with and without a fire element. It can be surmised that the graph with he and hud ji. is the product of augmentation of he fc and an abbreviation of the pictograph of the insect. 9 0 The element hud A in this graph ranges from being omitted, to being included. The one with a fire grapheme appears on the Baoshan wooden slip and the Guodian bamboo-slips (Liude no. 25). 9 1 This meaning is given by the SW. 9 2 The identification of the grapheme under & is ambiguous. It has been transcribed into-? (#&$3 *khsgwx). The explanation of-5 is given in the SW as «an appearance of vapor which is blocked its rising». Since the graph in question does not contain a grapheme which denotes the sound of the word it represents, this graph is treated as an HY graph in the present thesis. The graph % is later developed into the modern character If 'to go to inquire'. There is also a Baoshan bamboo-slip graph ~%, but it does not have the modern equivalent graph. 9 3 The graph % in the bronze inscriptions was used to write the word "to assist". 9 4 The origin of the graph hudng ^ has been proposed: (1) this graph was ramified and evolved from a pictograph of an arrow which was proposed by Zhao Cheng (1988, 75); (2) this graph was originally a depiction of a deformity which was later represented by the modern character wang jg, 'crippled', which was proposed by Qiu (1992, 218-219). 102 6) $ 1 /S ( f c t f *ljagx) 'troops'; two or three men under the flag denoted troops; (OBG f$ ft)-Chu: it # # ; SE A consisting of % chl (0 tf) 'to go' and tt zhi (iltf) 'foot' was added. Qin: ; the original composition was preserved. (B) Q i n graphs There are 3 cases of Qin H Y graphs being augmented with an additional graph of related meaning. 1) 1^ hou ( # t f *gug) 'target, feudal lord'; originally consists of r «hanging archery mark» and ft shi ( J i t f ) 'arrow'; (OBG f l ) . 9 5 Chu: Jfc ; preserved the earlier composition. Qin: ^ ; A ren (^tf) 'man' is added to denote the meaning 'feudal lord'. 2) jfc mian (7c tf *mjanx) 'to avoid'; originally consists of ^ midn ( jctf) 'head-dress' and A ren (% tf) a man who is wearing a head-dress; (WZG^p). 9 6 Chu: °^ ; preserved the earlier composition. Qin: \ 3k. ; the graphemes midn ^ and ren A fused and an additional element ren A was augmented on top. 3) Hf qidn (7Ctf *dzrianx) 'to advance'; a boat, # zhbu ( f t t f ) , with ± zhi ( i l t f ) 'foot' which signifies movement; (WZG E9 ).97 9 5 Although several explanations have been proposed, what an element f means is actually not clear. 'The hanging archery mark' is adopted from Shinjigen (Ogawa et al. 1967, 63). 9 6 SW originally did not include this graph. Duan added it later. His elucidation that it renders ' [rabbit] to free' by the form which depicts a rabbit without its legs has been denounced. The graph ^ is elucidated by Qiang Yunkai (1935) to be composed of ^ and ren A depicting a person who is wearing a head-dress. According to him, this graph is the antecedent of the modern mian j?,'head-dress'. Another elucidation of this graph is that it is an antecedent of mian jfc 'to give birth to child' rendered by the form of the a man under a buttocks. (Ogawa et al. 1967, 89) 9 7 The graph % 'to go, ahead', which is explained as huiyi graph in the SW, appears in OBI. An element which appears to be a boat might be a boat-shaped shoe, which was proposed by Lin Yiguang #.H jt (JWG v.2: 801) The graph qidn fif which was created by adding a grapheme 7J d&o to # for the word 'to cut' replaced the graph #, and another graph jian i$ was created for the word 'to cut' (JCZ: 143-144). Since both # in the Baoshan bamboo slips (no. 122,123) and qidn I? in the Shuihudi bamboo slips (Falii dawen F"1 no. 12, 15) are used for the same word, that is 'ahead' and 'before' respectively, they can be considered the two 103 Chu: % ; the earlier composition was preserved. Qin: 1 0 d\o ( f IP) 'knife' is added. 4. 4. 5. Augmentation of HY with grapheme of unknown function (O) Both Chu and Qin graphs There is one case in which both the Chu and Qin graphs were augmented with an element whose function is unknown. 1) IL luan (jcip *luanh) 'disorder'; SE an element £ ( i f IP) which was said to depict a knot or a braided string 9 8 and SEs, two hands X you (^ .IP) «to hold in hand» above and below it. Chu: 9 9 Qin: % ^ The earlier form is an H Y graph which rendered 'to subdue' with the depiction o f two pulling both sides o f a lump o f entangled thread (Ogawa et al. 1967, 27). The graph luan | L appears on the W Z bronze as ^ in Fansheng gui andg^j>in Maogong ding -^^jfr i which contains four mouths on both sides o f an element £ (other W Z bronze-vessel antecedents o f this graph do not). The Chu form in question, therefore, was not a variant graph formed during the W S period, but rather an adoption o f a variant existing during the W Z period. The Qin used a graph to which an element o f unknown identity was later added. 1 0 0 (A) Chu graphs variants representing the same word. 9 8 An object the graph £ originally imitated is not clear. Guo Moruo (JGWG v. 1: 392) surmised that £ used in the oracle-bone graph $P yu 'to drive [a carriage]' imitated a form of the knot which is a bridle (Guo Moruo, JGWG vl: 392) It is also used in the oracle-bone graph and it was surmised as imitating a kind of braid by Yu Xingwu (1979, 64-66). The graph £ , therefore, might imitated an object like a knot of string or braid. An object £ imitated in the graph can be elucidated as a braided string . 9 9 This graph can be taken as luan $L because the graph luan SIL incised on the guwen ("i" part of the Santai shijing 3. ft If is the same as this graph. The graph luan f|L appearing on the WZ bronze, Maogong ding also contains four mouths on both sides of an element £ , while another WZ bronze graphs in question do not. 1 0 0 It was explained, for example, as Z^yi 'the second of the ten Heavenly Stems' by the SW or as the end of a entangled thread by Ogawa et al (1967,27-28). 104 There are 3 cases o f Chu H Y graphs being augmented with elements o f unknown function not present in the corresponding Qin form. 1) H geng ( * t f *kzngh) 'to extend across'; originally, it was composed o f the moon fl yue(fl tf) between two horizontal lines, as can be seen in the O B I as and | ] . Chu: S 5£ f E ^ 1 0 1 ; an element 1- or t whose function is unknown was added. Qin: § I § 2) i.zhu (-f^tf *tjugx) 'master, owner'; originally a depiction of burning fire on a stand. Chu: $L ; a grapheme ^ midn (7c tf) ' r o o f was added. Q i n : ± £ £ 3) fjj jian ( 7 c t f *krian) 'space between'; the moon fl yue (fl tf) and the gate men (% ID tf) dentoe 'space between'; ( W Z G ffi). Chu: 89 $f ; a grapheme 77 dao (Hf tf) was added. Qin: PQ ffi (B) Q i n graphs There are two cases of a Q in receiving an attachment of unknown function not present in the corresponding Chu form. 1) # de (j&tf *pjagh) 'to obtain'; a shell K bei (^tf) 'cowrie ' is held in hand X you tf) «to hold in hand» ; ( O B G Q( ^ ). Chu: ^ ; preserved the original composition. Qin: ^ ; an additional graph ^ (#tf) 'to go' was added. 1 0 2 The Qin form of this graph is also be found in the oracle-bone graphs OjC^j), although vow X was replaced with cun 2) jSfc ./'we (^ tf *dzjiat) 'to cut o f f ; a knife, 77 <£o ( f tf), and £ ( f tf) «knot or braided 1 0 1 The Chu forms, 3E f|, were augmented with an element For t whose function and identity of the forms are unknown. The SW stipulates thatiE is theguwen form of a graph if. heng(%;%$ *gzng) 'constant' which is composed of PE heng Sand SE i^xvt (ft IP) 'heart'. There is a Baoshan graph ^ which contains a heart element. It might be a case in which a semantic element was added to the HY graph to emphasize or clarify the meaning of the word represented. (An augmentation of a semantograph with a semantic element has been discussed in § 1.2.4.2.) It is possible that both graphs were used during the WS period. 1 0 2 This Qin graph is also treated in § 4.4.2. 105 string» are used to denote 'to cut o f f ; ( O B G | | $ ) Chu: |$$; preserved the older composition o f the graph in question. Qin : $L ; G, «kneeling person» whose function is unclear was added. 4.5. Divergent evolutionary changes from and misrecognition of earlier forms In this section, we w i l l present the discrepancies resulted from evolutionary changes or misrecognition o f earlier forms. The criterion for this category is that change were made by unintentional reinterpretation or erroneous recognition o f the former form o f the graph. This category has the following features: a) the modified forms do not function as P E or SE; and b) the temporal succession of forms manifests a linear evolutionary change in execution; or c) the original and modified forms are similar in shape. The cases in which the Chu graphs, Q i n graphs, or both exhibit changes from the earlier forms are presented together. The graphs which exhibit changes are denoted by "! ". In the following sections, the O C sound for the graph is provided when an evolutionary change o f the graph is related to its P E . 1) ^ shen «lightning» 'the ninth o f the twelve Earthly Branches' Originally the oracle-bone graph in question Q and depicted lightning. W e also find the oracle-bone graph ^ , which has been explained as the elaborate form of the graph shen tjs (Tang 1986, 60) and the original form o f the graph dian % ( Y u 1979, 10-11). Tang surmised that the Chu form o f the graph shen ^ might be the descendent of this elaborate form of the graph shen ^ . The small squares of the elaborate form o f the graph was changed into two P kou (JJtHS) 'mouth' i n the Chu graph. The Q i n form with two A zhao (Wtf) «hand holding something*, 'claw' appeared to be evolved from the simple form of the oracle-bone graph % or 106 ! Q i n : ^ 2) fc ddi *dzgh) 'to take the place of, generation' It consists o f P E yi (1&IP *rsk) 'to shoot [with arrow and string attached]' and SE A ren 'man ' . The graph yi \ in the Chu form was replaced with \ gc (HII[IP) 'dagger-axe'. ! C h u : # ? t Qin: ft T\ 3) & ting (#IP *ding) 'court of palace' The explanation o f this graph varies: (1) SP graph which consists o f P E ££. ting (#IP *thingx) 'to stick out, stretch' and SE (the SW); (2) H Y graph which consists of A ren IP) 'man' standing in front o f a staircase (Ogawa et al. 1967); (3) SP graph which consist o f P E ^(^IP) 'thick hair' and an left element depicting the court in the palace. The W Z bronze graph in question varies as in Y u ding j £ $ in Shiyou gui £ S p | f % and ^ in Song ding £f jff}. The Chu graph adopted theh^, and the Qin adopted(J[. The right element of the Chu graph became different from the earlier form which can be transcribed as ren A- The left element o f the Qin graph became closer to the modern grapheme 5 _ . ! C h u : £ ^ ! Q i n : * £ 4) & Hang 'good' It is unknown what this graph originally depicted. The upper and bottom parts evolved differently in the Chu and Qin graphs from the oracle-bone graph \ and the W Z bronze graph \ (Gebo gui felb 1) 5) OLyi' fitting, right' Meat in the container or stand originally denoted a certain sacrificial ceremony. The oracle-bone graph in question was written as j | . The form of the container in the Chu graph was reduced to a grapheme *** midn ' roof . 107 Qin: HQ 6) H qi 'wife ' Originally, the oracle-bone graph in question \ \ or ^ was drawn as a woman whose hair was emphasized and ybu X - The hair part of this graph graphs evolved differently as shown below. ! Chu:^T 7) J l hudng 'august' Originally, it was a depiction o f flames on the stand. The upper part which is a depiction of light from a fire on a stand evolved from the W Z bronze graph ^ differently. ! C h u : ^ ! Q i n : | l 8) # rub 'to approve' In the O B I , this graph depicted a person with both hands raised towards his hair as ^ and Y. Later, a mouth element was added, as can be seen in W Z bronze inscriptions. The Chu and Qin used forms evolved differently. ! C h u : ^ : ^ ^ ! Q i n : # -This graph is used in the following graph as a P E . a) g ni 'to hide' ! Chu: S 9) & cang 'storehouse' Originally, it was a depiction of a roof (a top element) and a walled structure (a bottom element) which resembles a mouth element with a door (a middle element) as can be seen in a oracle-bone antecedent • ! Chu: Ji" ; the mouth l ike element became one or two horizontal lines with additional short horizontal lines in its middle, resulting in the shape under the roof resembling the graph for the word 'horse' ^ . Qin : % ; the form of the original composition (a roof, door and mouth-like 108 structure under the roof) is preserved. 10) §tsha 'to k i l l ' It consists of P E ^ «long-hair animal», 'to curse' and SE & shu 'k ind o f lance'. The left grapheme evolved differently in the Chu and Qin. ! Q i n : f e ^ 11) * mu 'dusk' Originally, it was a depiction o f the sun ri B setting behind some grass which was expressed by four grass elements drawn on its top and bottom, as can be seen in a i t oracle-bone graph The Chu graph preserved the structure o f its oracle-bone antecedent. In the Q i n form, the two pieces o f grass at the bottom became the element consisting of two hands. Chu: ;ji ; ! Qin: M I 12) f da/(IfeHJ)'to borrow' It consists o f P E yi (K&IP) 'to shoot [with arrow and string attached]' and SE X bei (HIP) 'cowrie ' . In the Chu graph, the P E yi gained an extra stroke, producing a non-PE grapheme 3t gc (HfciP) 'dagger-axe'. ! C h u : £ # SW: ft 13) # lou 'name o f constellation' The W Z graph was written as . Two hands, and a middle upper element whose significance is not clear had been transformed into the upper element in the Qin graph as shown, while the Chu graph in question preserved more o f its earlier constituents. Chu: ^ ! Q i n : £ £ £ 14) ^ wei 'to make, to be' Originally, as the oracle-bone form ^ shows, this graph was composed o f an elephant 1 0 3 This graph is also treated in § 4.2.1.2. 109 and a hand grabbing the trunk of the elephant. Both the Chu and Qin graphs diverged. ! Chu:fe%fe^ 15) ^ chun 'spring' The O B I graphs in question s j* , B j - and 1$ consist o f two SEs, trees and the sun, or either one o f the two, and a P E "fe tun 'to accumulate, station'. Whi l e the Chu descendent graph preserved all elements separately, the Q i n fused grass and tun fe as shown. C h u : § L ! Q i n : # % 16) | | long 'dragon'; The O B I graph for 'dragon' ^ or ^ was a pictograph o f an imaginary l iving thing. The right part had differently evolved. ! Chu: fE ! Qin: II f | H The evolved forms of this graph were faithfully maintained when used as a P E in the following two graphs. a) $ft gong 'to show respect 1 0 4 ! Chu: f£ f £ ! Qin: $ # b) ft, chong 'favor' ! Chu: ! Qin : 17) 1fr g"* 'noble It consists o f PE*l3 'to hold upward' and SE K bei 'cowrie ' . In the Qin graph, it was changed into a different form from its precious one. Chu: A f f 1 0 5 ! Qini-fr-frf 1 0 4 The graph H has been elucidated as the older form of the modern character gong $$ 'to offer'. 1 0 5 This graph is also treated in § 4.3.2. 110 This graph was used as a P E in the following SP graphs, maintaining the same composition, a) i t yi 'to leave behind' Chu: i| $ ! Qin: it t f There are three more Baoshan graphs H gui 'box ' , 'k ind o f tree' and H "f^ fz (HIP) 'wall ' It consists o f P E # W (HIP) 'ruler' and SE ± ru (&IP) 'earth'. The Chu graph's lower part of ^ xi>j (^-IP) 'bitter' was changed into # / i n g (#1)3) ' w e l l ' , while the Qin graph used-I- fii (#lp) ' h i l l , ladder' instead o f p /ze (If IP) «kneeling person» and O . 1 1 1 ! Chu: §F ! Qin: If g 25) jjfc. 'to offer': It consists o f P E | (TCIP) ' k ind o f tripod with hollow legs' and S E fc quart (7cIP) 'dog ' . The graph I consists o f P E (&IP) 'tiger's head' and S E ffi li ( $ $ ) 'tripod with hollow legs'. The graph xidn Jtfc. with a grapheme either f^rf ding (#lp), as ( ? bo g w z ' J l f t J i ) , or K ^ez ( £ $ ) 'cowrie ' U 2 , asj| | j (Zhaobo gui -9 46 It), instead o f /i ifj, as ^$ (Ke xzz jf ), can be seen in the W Z bronze vessel inscriptions. The Chu adopted two variants, with bei H and li if), misrecognized the form o f the grapheme li ifj and changed into the two graphs as shown in (b). Chu : (a) Jft $ L ; SE bei M. ! C h u : ( b ) f t f ? $ ; S E / i i f Q in : I t | t ; SE / i i 26) #z>zg 'reward, to be happy' The bottom part o f the W Z bronze graph |1, a tail o f a deer-like animal, was interpreted by the Q i n scribe as the foot element and by the Chu scribe as the grapheme A/zwi ( t ip) ' insect ' 1 1 3 ! Chu: 1 % ! Q i n : ^ f 1 1 1 The Old Chinese pronunciation of an element O is unknown, but it can be surmised as being close to the sound of the graph ft yuan (76 $5 *gwjari) 'round' which carries an element O as a PE. 1 1 2 For example, the two WZ bronze vessels named Zhaobo gui $ J£ each carries one of the graphemes for the graphxj'an £t (JWB, 683) 113 ^ ^Z-There are variant graphs, >^ and which do not contain the element chong A . 1 1 3 27) flj wei ( £ t f *gwjadh) 'to protect' It consists o f P E % wei *gw/9d) 'tanned leather' and SE f t *ing(Bfr3P) 'road, to go' . The graph # was written in the O B I a s ^ ^ a n d ^ j r The Chu graph used the former form with an added element ifc zhi ( i l t f ) 'foot' as discussed above in § 4. 3. 4. On the other hand, the latter form was used by the Q i n with modifications probably caused by misrecognition. C h u : ^ ^ ! Q i n : ^ | f ^ ; the lower part o f the grapheme % was written similar to the graph f (%%) 'to surround'. 4. 6. Unclassifiable structural discrepancies There are 15 cases that cannot be classified into any o f the categories used above. Cases in this category are divided into two groups: (1) cases in which both the Chu and Qin graphs which differ from each other are found in the earlier times, and it cannot be determined which was the ancestral form and modified form; (2) cases in which graphs that were modified cannot be determined to which category o f graphs (SP or H Y ) they belonged, and it cannot be determined which category o f processes listed above the modification graphs underwent belong to. 4. 6 .1 . Cases i n which ancestors cannot be determined There are 12 cases in which both the Chu and Q i n graphs were found used in the earlier times, and the relationship o f the original and modified cannot be determined. 1) ± zud (Jfctf) 'left'; S E X you ( i l t f ) «to hold in hand» and S E P kou (Mtf) 'mouth' in the Chu graph or X gong (^.tf) 'chisel ' in the Qin graph. The Chu graph was composed of kou P , while the Qin graph was composed of gong X . Both forms can be found in the W Z bronze inscriptions, as ^ in X i n g zhongj|& M and ^ i n $ S . C h u : ^ ^ Qin: £ f£ 114 2) % cheng (^lp *7-jsng) 'to assist'; in the O B I , it was drawn as a person with a pair of hands, with its optional element depicting a hole in the ground, as (r\( and 1 .^ The Chu graph appears to be a descendant o f the original graph without a 'hole in the ground', while the Qin graph descended from one with a hole. C h u : # Qin:>& 3) jf!| li (J$$ *ljidh) 'favorable'; P E a pictograph of a plow and SE fc he 'crop' The grapheme )) li (liilp *ljid) ' p low ' appears in the O B I graph in question with and without two dots signifying lumps o f earth, as ty and J^. 1 1 4 The Chu used the grapheme ^ 'p low' with two dots, and the Qin adopted it without two dots, which would evolve into the modern grapheme l] dao ( f IP) 'knife ' . Chu: fh%^> Qin: $\ 9fd 4) # hou (felP *gugx) 'behind, after'; according to the S W , this H Y graph consists o f an element £ which renders 'to l ink ' or 'to bind ' , a foot facing downward ^ , and chi %, a grapheme which denotes activities relating to movement. We can find both the Chu and Qin forms on the W Z bronze vessel inscriptions: ^ on ? ding^ fi, and -1^ on L i n g gui 4"JI- Since it a n d ^ are used together as an grapheme & which was used for the word related to action o f moving, it is hard to determine which was the original composition. Chu: J/)%.; the grapheme it zhi 'foot' which often appears with an element chi % with which signifies human activities involved in movement. Qin : ^ ; the grapheme it zhi 'foot' was not included. 5) ft zai *dzagh) 'to plant'; P E ^ z a z GtlP *dzggh) 'disaster' and fc mu (felpj 'tree' We can find two graphs for the word 'disaster' in the O B I , ^ f a n d ^ , differing in their left upper element. The Chu and Qin forms appear to have derived respectively from these two graphs. 1 1 4 Qu Wanli Jg % JL (1984,460) surmised dots to be a depiction of lumps of earth. 115 C h u : $ Q i n : ^ . 6) 1 ydn (ft tf *grjam) 'salt' This graph cannot be found in the earlier materials produced before the W Z period. According to the SW, the graph in question consists of P E j& jian (ft tf *kram) 'to inspect' and SE | » | (fc tf) «salt». The Chu graph might have abbreviated the upper part of jian I£ and used only JE min (|#tf) 'vessel'. Chu: Hi H Qin: 1 ^ 7) itk di ( J i t f *djidh) 'earth'; P E •£ fa /tub ( ^ t f *thar) 'other', an SE g ./« (ft tf) ' h i l l , ladder' o r ± tu (fctf) 'earth' It is hard to determine which o f the following graphs was closer to the earlier composition o f the graph for 'earth. 1 1 5 Chu: El fi£ ±^ Qin: it ^ ; SE fit $ was not included. 8) H ji (iHtf *dzjsp) 'to gather, collect '; 'b ird with a short ta i l ' zhui (#tf) came to perch on the 'tree' mu (/JI tf); ( O B G ^ a n d Shang bronze g r a p h c ^ f on Xiaojimuyi zhi A^%^z,m * Since both forms appeared during the Shang period, we cannot determine which one was original or modified form. Chu: % f Qin: 9) ^ ning (#tf *ning) 'tranquil' The S W lists both graphs ning ^ 'peaceful' (which consists o f S E midn (jc tf) ' r o o f and SE xm (fttf) 'heart' and S E J L min (Htf) 'vessel') and ning ^ 'would that' 1 1 5 According to Takada ([1925] 1975, 518-519), the graph representing the word 'earth' was possibly written as l i , &, % , during the WZ period when it came into existence. Originally, Takada surmised, 'to fall from the heights' (SW: cong gao duiye %t$Hj Duan explained that 1$. and|S£ were the orthodox and vulgar graph for the word 'to fall') was used for 'earth', andgfc was mistakenly written this way because of the similarity in the graphical form. The graph |$$. was a variation with a grapheme tu i which was commonly added to the grapheme ju % . The graph pfc 'to fall', too, used for the word 'earth', and later a was created for the word 'earth'. 116 (which includes an extra grapheme "7 (#1)5) ' ? ' 1 1 6 ) . Duan explained that the graph ning & and ning are gujinzi The original form o f this graph appear in the O B I in three variations as , Q\. It might be that Chu was the descendent o f the graph consisting without the bottom element, while the Qin was that o f the graph with it. Chu: H Q i n : f f 10) £ ding (SIP *teng) 'to ascend, to offer' In the O B I , there are graphs which appear to be original graphs o f the modern graph % *1 W as f\ and 5. • The graphs without two foot elements are used for the sacrificial ceremony, and could be interpreted as 'to offer', but the graphs with two foot elements are used for personal names. It is not known when and how the latter composition of graphs came to be used for 'to offer'. The Chu used the graph with a pair o f hands, while the Qin used the one without a pair of hands. C h u : # « S Q i n r g 11) ft. fi (f l ip *bpkw) 'to return'; P E £ fi (ftIf *bpkw) 'to return' and SE , either % chi (0IP) 'to go' ( W Z G }f a n d j j ) According to the SW, this graph consists o f P E M. fi (ft IP *bpkw) 'to return' and SE, either % chi (#IP) 'to go' , but, as the W Z bronze graph & in San pan fj[# shows, the graph which consists oi ^ was also used during the W Z period. Gao (1987, 158) explains that the grapheme chi % was an abbreviation o f the grapheme ^ which is a combination o f chi % and i h z « i ( ^ . i P ) 'foot'. Because it is difficult to determine, however, whether the graph i n question was originally composed o f and the grapheme zhi i t was abbreviated later, this case is treated in this section. Chu: i | ^ 1 1 7 ; i t zhi ( i l i P ) 'foot' was attached. S W : $ 12) $ (TtiP *lan) 'disorder'; it consists o f two SEs t ydn (76UJ) and & mi (1£IP) ' s i lk 1 1 6 The identification of the grapheme -5 is ambiguous. According to the SW, it denotes "an appearance of vapor which is blocked its rising". 1 1 7 The modification of the form of the graph %_ is discussed in § 4.4.3. 117 thread' or £ (If IP) «knot or braided string». According to the S W , this graph consists o f two SEs "g ydn (76IP) and & mi (HIP) ' s i lk thread', but, as the W Z bronze graph fj? in Song ding flgjff} a n d i n Doubi gm' j l R! H shows, both forms were already used in the W Z times. Chu: gyg S W : M 4. 6. 2. Cases of unclassifiable graphs and modifications There are 3 cases in which graphs in question and modification cannot be classified into the categories used in this thesis. 1) % qi (^IP *fd9g) ' it, its'; a pictograph o f bamboo basket 0£_lp *kj^g) ^ A pictograph o f bamboo basket ( i l l p *kpg) ^ w h i c h had been loaned for a third pronoun in the O B I , and additional P E 7\ ( i l iP *kjsg) ' l ow stand to offer an object' as can be seen in the W Z bronze graph int. 1 1 8 The graph representing a 'bamboo basket', originally a loaned graph, was abbreviated from the Chu bamboo-slip form. The graph qi however, appears in the W S Chu bronze inscriptions. The graph qi Q used as a grapheme in the graph ji 'foundation' was also written in an abbreviated form A in the Baoshan bamboo slip inscriptions. Chu: A JT7X Qin : £ 1 1 a) i / i ( i L i P *kj3g) 'foundation' Chu: 2£ S W : * 2) H wang (#IP) 'to look from afar'; It was a depiction o f a man who is standing on the ground f£ ting (#IP) 'to stick out, stretch'. Its emphasized eye became the grapheme £ chen (^ip) 'subject ' . 1 1 9 In the Chu graph, P E iz wang (|#IP *mjang) 'to disappear' replaced the eye element, while an 118 The abbreviated grapheme is a pictograph of a bamboo basket which was used alone to represent qi 'it' in OBI. The abbreviated graph in the next case is also this grapheme. 119 Chen Hi was also a vertical depiction of an eye in the OBI. 118 element fl yue (fl tf) 'moon ' was added to the original graph ( \ or \ ) in the Qin graph. 1 2 0 while the Chu form remained in its original form. The Qin form can be found in the W Z 4. 7. Discrepancies existing in the earlier times The Chu and Qin graphs in the following 21 cases appear on the Shang oracle-bone or the W Z bronze vessel inscriptions. The discrepancies between the Chu and Qin forms are therefore the result of the adoption of different pre-WS variants. Both the Chu and Qin graphs are presented below followed by the antecedent graphs and their identity. 1) iiju 'to give' Chu: ft / f ; ft (Rui ding^ Q i n : # ^ ; ^ (Huding^) 2) %. zud 'left' Chw.fe&ife ((Bangui Q J n : & £ ; g (Xingzhong$S-lt) The graphyue f=\ was added to create a graph for the word 'full moon' (JWG v. 10: 5178-5179). This graph is from the Stone Drum:Qianyi. bronze inscriptions as (Shenzi gui i f c ^ l i j | ) . Chu: F§ W) Qin: ft121 3) % cheng 'to assist' Chu: # ; ^ (OBG) Q i n : > & ; ® (OBG) 4) M li 'favorable' 119 C h u : f h % ^ ;V ( ° B G ) Q i n : f i j f c -,y> (OBG) 5) Jk. ji 'to reach' Chu: i £ g (Fushu xu h &£) SW: A ^ ( B a o y o w & f S ) 6) M yuan *«win) 'abyss' Chu: J$JB};§ (OBG) Q i n : ^ (Shenzigu/ i t * t & ) 7) # c?e 'to obtain' C h u : ^ ; ^ ( (OBG) Qin: (OBG) 8) # hou 'behind, after' C h u : # ; ^ i ( ? r f m ^ H ) Q i n : ^ ; ^ (Ling gu/41) 9) & za/ 'to plant' C h u : $ ; ^ ( O B G ) Oin: jj& ; ^ t ( O B G ) 10) & ji 'to gather, collect' Chu: ; ^ (Xiaojimuyi z/*/ /Ml#£$) Qin: 4 $ (Maogong dihg € ^ ) 11) ^ mng'peaceful ' Chu: H;1? (OBG) Qin:f ^ ; [ y l ( O B G ) 12) vwa« 'distant' Chu: § ; V (OBG) Qin: £1 ; y (OBG) 13) $ii 'disorder' C h u : ^ ^ ; $ ( S o n g d i / i g 4 g | | ) sw: m jjjj (Doubi g«i a. M1) 120 14) ft ju 'to return' Chu: i | 3t ; ^ (San pan ft jfc) SW: ft ; ) f (Xiaochen ?gwz / J N E j& tS) 15) 4 *»«g 'servant, chi ld ' Chu: f f ; ? (OBG) Qin: 11 ; | f (Fansheng gui 1) 16) IS: gwi (ft IP *kj9gwh) 'k ind of ritual vessel' in ft/iii (ft IP *kj3gwh) 'stable' Chu : M # ; ^ (Jihou gui & £ 1 (sAi * was used in gwi ft) Q in : J f c f e ; $ ^ (Ling g«i4 - f (g was used in gwi ft) 17) # ,zW 'to hold ' ' Chu: & ft $| f# H ; (Duoyou rfmg # * ih) Qin : (OBG) 18) £ dercg (SIP *fc»g) 'to offer' Chu: | ^ #r H ; $ (OBG) Qin: f ; ^ (OBG) 19) ft tow 'to defeat; to be defeated' Chu: ft $ #X ; S3 (Shishi gui Qin:ftfeftl;0l;(OBG) 20) x/a/i 'to offer' Chu : (a) ft %. \ (Zhaobo gwi # # 1) Qin : ft ft ; %((Zhaobo gwz # # %) 21) # we/ 'to protect' C h u : ^ 1 g ;ifo"f (OBG) and^(Weizun%%) S W : # ; ^ f ( O B G ) 121 4.8. Summary of modifications discussed Table of the graphs discussed in Chapter 4 List o f abbreviations * : This mark following the graph indicates that the graph shown in the list was also used in other SP graphs as a constituent, as presented in Chapter 4. S W G : The graph in the S W is used as a substitute of the Qin graph in question. C W S : From the table (Appendix I) the graphemic change of the graph can be considered to have happened during the W S period. C W Z : From the table (Appendix I) the graphemic change of the graph can be considered to have happened during the W Z period. O B I : Both the Chu and Qin divergent forms can be found in the Shang OBI . W Z : Both the Chu and Qin divergent forms can be found in the W Z bronze inscriptions. 122 Table 4. 1. 1. Graphs discussed in Chapter 4 Modification Evolutionary Formation Type of SP Modification HY Modification Unclass. Process Chu Oin Chu Qin Chu Oin JB # ifJ CWZ Unspecifi. Omission If If J[ CWS ft Ancestor Step ffl jj SWG & ± wz 1 # SWG total 3 0 3 2 ft ft % OBI jiajie SWG ^ SWG # wz # >f»J-*wz 3pJ OBI £. wz SWG jf CWS ^ CWS $ OBI total 7 j& c w s Si itfc SWG % OBI & & Replacen « f * ^ OBI -g- OBI & wz * $ wz Step Jfr Jt wz * 12 2 * 11 II Unclassifi. Sd. * SWG * * Graphs total 13 7 1 3 jt $ SWG i f s rf # & & OBI wz jKH wz 3 *l P E 4 wz total 14 A It Chu u it* Step Chu g *^cws it wz 2 4 Chu m total 3 1 6 4 ft ft Sd. It X Chu Chu Chu e n t St It ft 7A wz % wz total 6 a SE It Jt SWG t Jfr | i SWG/WZ Step H i & wz CWS 1&2 & SWG o n total 4 total # 9 1 7 4 other % JR SWG Unk * ^*cws fg SWG HL wz SL # OBI total 2 WZ ± ft HY W ra total 1 total 2 2 4 3 Total 30 cases 11 cases 21 cases 16 cases 22 cases 15 cases Total 34 cases Total 41 cases 37 cases 27 cases 15 cases 123 Table 4. 1.2. Graphs which underwent multiple divergent processes Each graph underwent the processes marked with " X " . Graphs are indexed by entry number in the table presented in Appendix I. Abbreviations C: classification o f the graph G : graph H/S: modification occurred upon a H Y graph and upon an SP graph for which the H Y graph was used as a P E . No. C G Formation Modification Process Evolutionary Change Omission Replac Augmentation Stepl Step2 ;emeii P E SE Unknown Jiajie S.det Chu Qin Chu Qin Chu Qin Chu Qin Chu Qin Chu Qin 1458 SP M X X 1067 H/S ft X X X 521 SP fix. X X X 1087 SP X X 389 SP <# X X 283 SP X X 957 SP * X X 1224 SP ft X X 1356 SP # X X 332 HY X X 1074 HY # X X 781 SP X X 431 HY # X X 392 SP i t X X 1083 SP # X X 923 HY X X 1407 HY ft X X 124 In this chapter, we have examined 105 graphs for which the Chu and Qin forms differ in graphical structure. There are 34 cases in which structural discrepancies originated in the two-step SP and one-step H Y formation process. The discrepancies o f 7 SP graphs resulted from different usage o f jiajie with the same semantic determiner to remove overload from homophonous or near-homophonous representation. There are 14 SP cases in which different semantic determiners were added to the same phonetic elements to represent a given word. 4 SP graphs exhibit entirely different usage o f both jiajie and semantic determiners. There is 1 SP case in which different semantographs are used with the same phonetic element. There is also another SP graph i n which different semantographs are used with the different phonetic elements. We find only 1 case o f an H Y graph formed from different elements. For these 28 cases in all, no antecedents from any period can be found, and so it can only be proposed that their formation in the two regions was independent or partially so. 1 2 2 The six remaining cases show Chu-area only augmentation with semantic determiners to jiajie graphs. It can be suggested that the augmentation of one graph Idi 'to come') took place during the W Z period. For the rest, the time at which augmentation occurred is unknown. There are 69 graphs whose structural discrepancies were the result of various types of graphemic modifications: 36 SP graphs (Chu 30 cases; Q in 11 cases) and 33 H Y graphs (Chu 21 cases; Q i n 16 cases). 1 2 3 The numbers for SP and H Y modification can be summarized as follows: Omission Replacement A d d / P E A d d / SE A d d / unknown SP: Chu 3 13 3 9 2 Qin 0 7 1 1 2 H Y : Chu 3 1 6 7 4 Qin 2 3 4 4 3 1 2 2 Todo (1965, 13) has proposed that a Chinese graph to represent a given word could have been independently formed by different people, at different locations. Gao (1987, 146) has proposed that the usage of divergent but semantically congruent semantic constituents in multi-element graphs was not necessarily synchronic but diachronic. As a result, a given word could at the same time have several representations with different semantic elements. 1 2 3 The number of graphs which have been treated in Chapter 4 and the number of cases in which the Chu or Qin graphs underwent modification do not match. Some graphs must be counted more than once. See Table 4.1.2. 125 There are 4 Chu graphs (2 SP graph # j | and 2 H Y graphs and one Qin graph (1 H Y graph J [ ) for which the modification can in the table be put during the W S period. The forms of 6 graphs used in the Chu area (3 SP graphs WM # and 3 H Y graphs can be found in the oracle-bone or W Z bronze graphs; 4 H Y graphs (##^.4) used in the Qin state can similarly be traced to earlier antecedents. There are 27 graphs whose structural discrepancies between the Chu and Q i n were caused by unintentional evolutionary changes: 10 cases in which both the Chu and Q i n evolved differently; 12 cases in which only the Chu graphs underwent changes; and 5 cases of Q i n only. In the table, we find that both the Chu and Qin modified forms o f the graph Ji first appeared during the W S period. O f the 22 Chu modified forms, there are 3 graphs (tf> whose evolved forms can be found in the W Z period. There is only Qin modified form (#) which can be found in the W Z graphs. There are 15 graphs which cannot be classified into the categories above. Both the Chu and Qin forms o f 10 graphs are found on either the Shang oracle-bones or W Z bronzes, and the original composition o f 2 graphs are ambiguous, and so it has been impossible to determine the nature of the modification that caused the discrepancy; 3 graphs have not been classified because the processes of modification they underwent are not assignable to any one o f the categories. From the extant materials, as traced in the table, it has been possible to date the discrepancy in the Chu and Qin forms as originating in the Warring States period in 7 cases, and in the earlier Shang and Western Zhou periods, to 26. Scarcity o f material precludes dating the rest. 126 4. 9. Tables (Multi-element graphs) ft Chu P 3 L M Yangtianhu bamboo slip # Qin KM Qin wooden tablet Oracle-bone SLLJ Wangshan bamboo slip W Z bronze *# Silk manuscript ~$SjkX W Z bronze lift* ZuChuwen stone tablet &&X S A bronze Baoshan bamboo slip m&x W S bronze Shuihudi bamboo slip * f c i Zenghou Y i bamboo slip Modem grapheme Stone drums Modem character 1BB Xinyang bamboo slip 4.2. Structural discrepancies present during the process of formation 4.2.1. SP graphs 4.2.1.1. Divergent usage of jiajie graphs ft # 1 it 2 ail T T 3 a • 4 tl # 5 ft if 6 11 f RD M wn 7 # H 4.2.1.2. Divergent usage of semantic determiners ft ** # &&X 1 © % 2 I T 3 ft 127 &* a f t # t l * * 4 * f % ft* 5 l i 6 a. is 7 (It * 8 V Iff • 9 w 10 ft *» 1* 11 If te I f 12 ft n f i 13 §8 4i 14 n if $ * M 4.2.1.3. Chu augmentation of jiajie with semantic determiners ft &* &&X *t a f t * * # &&X l p q PI 2 3 A f] A A ft * 0 4 J S 128 ft ffl <& WiS W i r * * i r * $ i r * adj # T 1A i r * W i r * * i r * ** 5 $ I 0 4.2.1.4. Different graphemes used at SP formation ft ¥#* WiS i r * W i r * * i r * $ i r * ** # W i r * * i r * $ i r * 1 H d E It 2 It i l But PR 3 ft 4 * f & 4.2.1.5. Different usage of semantographs ft ft* WiS i r * W i r * *&* « i r * # W i r * * i r * « i r * l i f t * 1 T T 1=1 7TJ 4.2.1.6. Different usage of semantographs and PEs ft ¥t* WJS i r * W i r * * i r * $ i r * -flit # Wir* * i r * $ i r * l i f t * l ^ # * I t 4.2.2. H Y graphs ft ¥t* WiS W i r * * i r * $ i r * £ih ftt aft # W i r * * i r * m i r * tf«t* l i f t * l I iff i i 129 4.3. Divergent structural modification of SP graphs 4.3.1. Omission of SP grapheme (A) Chu graphs ft ¥ t * WM &* W & * *&* $&* aft a f t # W & * #&* $&* t l * * 1 ft 2 Ik i *» 3 n ^ -T\ >T\ 4.3.2. Replacement of SP grapheme (O) Both Chu and Qin graphs ft ¥ t * ISM W & * #&* $&* a f t a f t # W & * *&* H & * t i * * 1 i t 2 i 1 II fl III II ti 1 (A) Chu graphs ft f t * &* W & * «&* a f t a f t # W & * $&* t i * * l 7= i-2 % i t£ 3 f?l JT 4 ii f£ «: i ft 130 ft &* &&X ** aft aft # &&X tmx 5 Si % 6 * 7 He 8 $ * f t it i t It 9 if m% 10 m 11 * 1 (B) Qin graphs ft ¥# X W&X adj aft aft # &x *** %&X lift* m 1 * f tf a t . f * 3 tic A Hi it 4 c % M D •» 131 4.3.3. Augmentation of SP with extra P E (A) Chu graphs ft f t * WiS &* W i r * * i r * $ i r * 'S.Jj'HrlBf aft aft # W i r * *&* $ i r * *r&* l i f t * 1 * 1 2 i t 3 if $ * (B) Qin graphs ft f t * WJH i r * W i r * # i r * $ i r * f M i JLiii aft aft # W i r * * i r * $ i r * ^^ * l i f t * l *±* S? ^ if 111 4.3.4. Augmentation of SP with extra SE (A) Chu graphs ft f t * WiS i r * Wir;* * i r * l ^ i r * #t aft *» aft # W i r * *&* $ i r * l i f t * l f ft 2 3 ft a s IS n « 4 f? I m 5 6* I* & s *» it 6 fit 132 ¥ t * &* &&X &&X ft* a f t «ft ** # W & * &&X Sk&X t i t * 7 l» ft fk 8 ffl * 9 # ft m ( B ) Qin graphs ft ¥** mM &* m&x *&* «,&* ft* a f t # m-tx &&X &&X 5 t i * I E * * l If 4 f ft 4.3.5. Augmentation of SP with grapheme of unknown function (A) Chu graphs ft f t * £* W & * *&* S $ & * ft* a f t m a f t # W4r* *&* $&* ^ &* i l * * 1 t t w i t 2 a t JlL f i f f ( B ) Qin graphs ft ¥ t * mM &X m&x *&* $&* S U J ft* a f t a f t * * # m&x #&* &&* ^&* !&** l m * 2 i 4.4. Divergent structural modification of H Y graphs 4.4.1. Omission of H Y grapheme (O) Both Chu and Qin graphs ft ¥ t * WiS i r * W i r * # i r * a i r * &djt fr f f i a f t a f t # W i r * * i r * l ^ i r * 5 I A * t i * * 1 ft « ft i t # ft $ (A) Chu graphs ft ¥** WiS •kX W i r 3 * i r * K i r * adj a f t a f t # W i r * * i r * ftir* # t £ * u s * l tf HP Bp ff56ff} 2 % >* # * (B) Qin graphs ft ¥ t * WiS W i r * * i r * a i r * %9tL adj a f t a f t # i r * W i r * * i r * a i r * %$£X t i * * mmm® 1 A PI s & HI m 4.4.2. Replacement of H Y grapheme (A) Chu graphs ft ¥ t * WiS i r * W i r * * i r * a i r * # » adj «# ^djitffi a f t a f t # W i r * * i r * a i r * t i * * 1 f% (B) Qin graphs ft ¥ t * WiS i r * W i r * *&* a i r * adj a * ^.djitffi a f t a f t * * # W i r 3 *&* a i r * ^fe* fc® i i * * mM.ik%® l ft # of f 1 134 ft ¥ t * W$ Wir* * i r * ftir* ^iljItlSf aft aft # W i r * * i r * a i r * l i f t * f # Z # 3 f t 4.4.3. Transformation of H Y to SP (A) Chu graphs ft ¥ f * W$ i r * W i r * * i r * a i r * f -feih-Hrff aft a « * * # W i r * * i r * a i r * i i * * 1 rl S 2 3 ran fa r f 4 A y # 5 & if II 6 t i Si t tt tt t 135 (B) Qin graphs ft f t * mM m&x *&* $&* ft* lift 0$ aft # m&x *&* $&* *ft* U S * 1 % O OO A AA 2 fi & s 3 i f mi & • 4 f l I t I i ft nn If If fl It 4.4.4. Augmentation of H Y with extra SE (0) Both Chu and Qin graphs ft f t * mM &* W & * *&* $&* ft* aft aft # W & * *&* $&* t i * * l 9 136 (A) Chu graphs ft ¥ t * WJS wi-* * i r * a i r * a f t as * * # Wir* * i r * ftir* i l * * 1 k IdE °< lk & ft # 2 3 1 * 4 * 5 1 I I n n * t * f f i t 6 & (B) Qin graphs ft W$ i r * Wir* ^ i r * a i r * SLih a f t aft * * # Wir* * i r * a i r * 5ft* fc® 11** l f It If i t 2 3 1 137 4.4.5. Augmentation of H Y with grapheme of unknown function (O) Both Chu and Qin graphs ft ¥ t * mM &x m&x &&X &&X HB 4.OJ ft* aft af t * * # m&x %&x &&X ti** 1 $ * I f l It (A) Chu graphs ft ¥ * * mM &x m&x &&X &&X HB ft* aft m aft m&x &&X ^ si* ti** mtim® l I f ? F 2 T T 3 H 9 v> W Sl (B) Qin graphs ft ¥ * * &* m&x &&X &&X HB aoj ft* aft aft m&x &&X ?k1Bf ti** l ^ # 2 (22 ft* 4.5. Divergent evolutionary changes from and misrecognition of earlier forms ft ¥ • * mM m&x &&X &&X HB adj ft* a * # m&x %&x &&X ?kiBf ii** l a i % % ¥ % <¥ 2 ft ft ft ft 138 ft f t * Wit* * i r * l ^ i r * aft ## a f t # W i r * # i r * a i r * 5 ft* fc® l i f t * 3 ID % 1 4 M \ 1 i t ft 5 1 A S I 6 fl$f< f * T f t f 7 a i f i m i 8 # I Pi-9 t t f 10 ** /f ix. ft* 11 t t e> Y t MI 12 13 * * i f * 14 H m iff « 25 /Tift 15 *3* O T <^ * f i 139 ft f t * &* ®4r^ «» ** a f t a f t # W4r* l i f t * 16 * f? f t f t IE II tin If % H « ** 17 * f t i t it 18 Jgc 19 JS is 9m. 20 fe If 21 JS //I 22 ! * m # * 23 ft Be % 24 m 25 ?* It fe % lit A ft ft fc ft. 26 1 it 27 ifr 4.6. Unclassifiable structural discrepancies 4.6.1. Cases in which ancestors cannot be determined ft ¥ t * * i r * a i r * «« a f t # W i r * * i r * a i r * 5 f t * l i f t * 1 % % & f i 2 A 3 % J f i l % ft 4 Ik i i t ft & 5 6 Ik S 7 ft * 8 OBX BR if * 9 jm. 10 t # * 11 t i t i t 12 $ 141 4.6.2. Cases of unclassifiable graphs and modifications ft ¥ t * ® $ £* W & * *£* $&* ft* aft *« aft sin # W & * *&* 5 S i * t l * * 1 M M Tv TT rcTl TT: y N W | / \ 1st 7T it 2 I t. 3 1 1# m tf 4.7 Discrepancies existing in the earlier times ft ¥#* W & * *&* $&* ft* aft # W & * #&* K i r * t £ * * 1 ft tt 2 fx % 3 ©A 4 ft 5 6 1 HI tf 7 A # 8 ft 9 & 10 ! * 142 ft ¥ t * WJS i r * W i r * * i r * a i r * f M i aft aft # Wir3 # i r * a i r * l i f t * 11 A tm. 12 13 il tt A 14 fr i 15 f 11 % i * I i 16 m I t JR: Ji 17 tt ft 18 * * 19 ft 20 ?t fe s ft ft * ft ft 21 % S 143 Chapter 5 Conclusion The characterization that Warring States "graphs varied in form" given in the S W had since the first century A D had become traditionally accepted, but scientific study o f the Warring States writing has been made possible only with the abundant archaeological discoveries o f the recent past. He L i n y i ' s study o f the Warring States graphs, among other scholars, presents wel l the state o f contemporary knowledge o f Warring States writing. B y today, regional discrepancies have been thoroughly examined; the observable tendencies governing the regional variations have been categorized: notions such as graphemic augmentation, omission, and replacement have been introduced. Geographic divisions based on internal homogeneity o f the graphical forms have been widely accepted. Scholars have in the main reached two essential conclusions: (1) considerable regional discrepancies existed in the writing, which at the same time was (2) uniformly descended from the Shang oracle-bone and Western Zhou graphs. Based on several previous studies, this thesis has attempted to draw a detailed and encompassing picture o f two aspects of the regional discrepancies: (1) the degree to which the Chu and Q i n writing diverged during the Warring States period; and (2) the nature of the observable discrepancies between the two regional traditions. The table o f graphs constructed sufficiently shows the degree o f heterogeneity and homogeneity between the Chu and Qin graphs. Inspection confirms that both the Chu and the Q i n graphs indeed descended from common oracle-bone and Western Zhou ancestors: the uniformity observable between the two regional traditions i n simple graphs overwhelms the observable discrepancies; the consdierable structural discrepancies in multi-element graphs have been observed, of which some originated before the Warring States period. That the graphical situation o f the Warring States period was the outcome of processes continuous since the Shang period has therefore been the basis for attempting to characterize the observable discrepancies between the two regional traditions. I f Chu and Q i n writing descended from a common ancestor, and i f Warring States graphs formed an unbroken evolutionary development from the Shang period (He 1989), then any regional Warring States structural discrepancy was the partial outcome o f individual modifications or events that occurred in exactly one o f the regions before or during the Warring States period, and in total the outcome o f all these events in both regions. We have traced (in Chapter 4) various observable discrepancies between the Chu and Qin writing to the level o f such events and o f the region in 144 which these events occurred. Given the standard hypothesis that the Qin writing was the most conservative o f the five regional traditions of the Warring States period in preserving the earlier graphical forms, one would expect that the modifications o f the Qin graphs should be minimal. The examples presented in Chapter 4, however, show that structural graphical modifications were not confined to the Chu area only. Although the number of Chu modifications tabulated in this thesis exceeds the number o f Q i n modifications, cases o f Q i n deviation from antecedent forms preserved in the Chu tradition have also been found. It is possible that future archaeological discoveries w i l l lead to a reassessment o f the relative degrees o f conservatism present in the five regional traditions. There is a scholarly tradition that graphical divergence was confined to the Warring States period. In the table we can trace the changes in graphical form o f several Chu and Qin graphs. Several cases o f divergence are found to have originated during the Warring States period: for example, the Chu modified form o f # hudng ' ye l low ' (Table entry 1014), the Shang and Western Zhou forms o f which are unmodified. For most cases, however, the exact moment at which the processes of modification first produced the discrepancies o f the Warring States graphs cannot be determined for lack of data. In the extant materials we find only a few cases of structural discrepancy born in the Warring States period. On the other hand, there are more cases o f both the Chu and Q i n forms, one modified and the other not, already in use during the Western Zhou period: or indeed, in a few cases, as early as the Shang. Graphical variations occur during the Western Zhou period (Tang 1986; Q i u 1988), but the prevailing opinion has held that Western Zhou writing remained on the whole uniform (Guo 1972; Tang 1986), a single tradition under the strong central power o f the Zhou Royal House. That both Chu and Q i n forms can be seen in the earlier Western Zhou graphs suggests, however, another possibility: the basic graphs such as pictographs having been created, Chinese writing did not remain uniform throughout, but began to diverge toward the graphical situation o f the Warring States period. Whether the observable graphical variants can be traced through a definite structural tradition, or appeared, spread and been transmitted by random chance, the graphical situation o f the Warring States period definitely appears to be the accumulated result o f the earlier vigorous graphical development of the guwen period. A s more materials bearing earlier graphs are brought into the light, we shall better understand the Warring States graphs in relation to the early graphical development of Chinese script. 145 Bibliography Bibliographical Abbreviations G WYJ Guwenzi yanjiu JHKG Jianghan kaogu KG Kaogu KGXB Kaogu xuebao # % ifa WW Wenwu Data Sources Baoshan Chujian wenzi bian i £ ill ftft;5t^$§ (A Compilation o f Baoshan Chu Bamboo-slip Characters). 1996. B y Zhang Shouzhong tf 1. Beij in: Wenwu Chubanshe %% fcB Baoshan Chumu ^ i l j f t j s (The Chu Tombs at Baoshan). 2 vols. 1991. Edited by Hubei sheng Jingsha Tielu Kaogudui fflik%fflty4&$&%-fr&. Beijing: Wenwu Chubanshe Changsha Yangtianhu chutu Chujian yanjiu -ft y.iMty ^ J#| t B i f t f S f ^ f 2£ ( A Study of the Chu Bamboo Slips Excavated at Yangtianhu in Changsha). 1955. B y Shi Shuqing Shanghai: Qunlian Chubanshe 3$ $ & WL^L. Chu wenwu zhanlan tulu $L%fy}>%.% (Rubbings and Drawings from the Chu Cultural Relics Exhibition) 1954. B y Beij ing L i sh i Bowuguan 4 k ^ M i t i # # l t , and Chu Wenwu Zhanlanhui ft % % fik % it- Beijing: Jinghua Yinshuguan % ^ £p # I t • Chuxijianbo wenzi bian ft^flt^^^H (A Compilation o f Chu-style Bamboo Slip and Silk Manuscript Characters). 1995. Teng Rensheng Wuhan: Hubei Jiaoyu Chubanshe t i$Jlfc*fc. Guo Moruo quanji V o l . 9: kaogu bian 1 i P i ^ ^ ^ r ^ : #"£"$S 1 (Collected works of Guo Moruo, V o l . 9: Archaeology 1) 1982. Edited by Guo Moruo Zhuzuo Bianji Chuban Weiyuanhui f P ^ ^ H H & M it- Beijing: Kexue Chubanshe & J j £ # . Jiaguwen bian f H ^ l i (Compilation o f Oracle-bone Characters). 1965. rev. ed. Compiled . by Sun Haibo 3£$Mjfc. Edited by Zhongguo Kexueyuan Kaogu Yanjiusuo + @ fcttOtfLffi. Beijing: Zhonghua Shuju + 146 Jiaguwen heji f •f (Collected Oracle-bone Inscriptions). 1979-1982. 13 vols. Edited by Guo Moruo f]5 :$l^r. Edited in chief by H u Houxuan jf^ SE. Beijing: Zhonghua Shudian + Jianghan kaogu xT-7| |#"£" (Jianghan Archaeology). Jinwen bian &XWs (Compilation o f Bronze Characters). 1985. 4th ed. B y Rong Geng Revised by Zhang Zhenlin and M a Guoquan ,lf Beij ing: Zhonghua Shuju t ^ * ^ -Kaogu (Archaeology). Kaogu yu wenwu ^"t^-^X^H (Archaeology and Cultural Relics). Kinshutsu Yin-Shu kinbun shusei ifL&MkM&X(Recently Excavated Y i n and Zhou Bronze Vessel Inscriptions) 5 vols. 1989-1996. Compiled by Urano Toshinori jflf TJ& I'J. (Nishogakusha Daigaku Toyokenkyujo —^JJt^A#Jfc##9f ^£^f.) Tokyo: Nishogakusha Daigaku z.^Q'fefcQ. LiangZhou jinwenci daxi tulu kaoxhi W M&X%frfc& M$kTfc& (Western and Eastern Zhou Bronze Vessel Inscriptions: Translations with Drawings and Rubbings). 1958. Reprint. B y Guo Moruo fp j f c ^ . Edited by Zhongguo Kexueyuan Kaogu Yanjiusuo cf3 ® Bfc#-£#ff t#f . Bei j ing: Kexue Chubanshe Original edition, Tokyo: Bunkyudo Xfct, 1932-1935. Qin mingke wenzi xuan # % % X ^ s t (Sellected Q i n Epigraphs). 1976. Edited by Shanghai Shuhuashe _b Shanghai: Shanghai Shuhuashe J i ^ H r ! ^ . Qin tongqi mingwen biannian jishi #!H##i3t$l#l l$ (Qin Bronze Inscriptions Arranged Chronologically with Commentary). 1990. Edited by Wang H u i 3.%. X i a n : Sanqin Chubanshe 5. # tB JBMfc. Qingchuan xian chutu Q i n Gengtianlu mudu # J I | H t B ± # j L 0 # ^ J t i (Wooden Tablet Inscribed with the Revised Land L a w o f Q i n Excavated at Qingchuan, Sichuan). 1982. B y Sichuan Bowuguan B )\\if # f l | , and Qingchuan Wenhuaguan Xit It- WW\\ 11. 147 Sandaijijin wencun 3. 'Pt^-^X^- (Collection o f Bronze Inscriptions from the Three Dynasties). 4 vols. B y Luo Zhenyu Reprint. Hong Kong : Longmen Shudian tl PI • j£. Original edition, N . p., n. d. (Preface dated 1936). Shierjia jijin tulu (Collection o f Shierjia Bronze Inscriptions ). 1973. Reprint. B y Shang Chengzuo Taipei: Zhongxin Shuju + f f Ir/Hf. Original edition, N . p., n. d. (Preface dated 1935). Shiguwen shuji %IiXX$Li& (General Notes on Qin Stone Drums). 1935. B y M a Xulun ,!f Shanghai: Shangwu Yinshuguan W •It-Shou xian Caihoumu chutuyiwu (Relics from a Shou County Caihou Tomb). 1956. B y Anhui sheng Bowuguan £ # # i f % $ f . Beijing: Kexue Chubanshe Shuihudi Qinjian wenzi bian ^j^L^k^-MX^M (Shuihudi Q in Bamboo-slip Characters from Shuihudi). 1994. B y Zhang Shouzhong Beijing: Wenwu Chubanshe X% tB Shunju sengoku jidai seidoki no kenkya. Jn-Shu seidoki soran 3 ^^k'&ffl&Ffiit 'WLM^MW^-%=- ( A Study of the Spring and Autumn and Warring States Bronze Vessels: Synthetic Index for Y i n Zhou Bronze Vessels, vo l . 3.). 1989. B y Hayashi Minao #. & ^ ^ c . Tokyo: Yoshikawa Kobunkan )\\ The Ch'u Silk Manuscript, Translation and Commentary. 1973. Barnard, Noe l . Cambera: Department o f Far Eastern History, Research School o f Pacific Studies, Institute of Advanced Studies, The Australian National University. Wangshan Chujian Jt di (Chu Bamboo Slips from Wanshan). 1995. Edited by Hubei sheng W e n w u K a o g u Yanjiusuo %$C%X%%~fe~tf;%ff\> and Be i j ing Daxue Zhongwenxi it% A # + X%. Beijing: Zhonghua Shuju 41 ^ • ^ f • Wenwu X% (Cultural Relics). Wenwu cankao ziliao XtffiH^^^Pc (Reference Materials for Cultural Relics). Xiaotun dierben: Yinxu wenzi: jiabian /Jst fc%ZZ. fcffcJ&X^- (Xiaotun, vo l . 2: Y i n x u Inscriptions: Part 1). 3 vols. 1948. B y Dong Zuobin jff^W- Taipei: Institute o f 148 History and Philology, Academia Sinica. Xiaotun nandijiagu /h 1ft $k ¥ # (Oracle Bones from Nandi, Xiaotun). 5 vols. 1980-1983. Edited by Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo lTtBI$fc'#"$-!fel&#'&"#Fft $ 1 . Beijing: Zhonghua Shuju + # • Xichuan Xiasi chunqiumu y#f J'l T # (The Spring and Autumn tomb at Xichuan Xiasi) . 1991. B y Henan sheng Wenwu Yanjiusuo j5T^#X$}*R%?ft, Henan sheng Danjiangku qu Kaogu Fajuedui jf $ f # # 3 l L $ £ Bit, and Xichuan xian Bowuguan fljf J l | B e i j i n g : Beijing Wenwu Chubanshe il%X^ tB E^t. Xinyang Chumu 'ftlfftS (Two Tombs o f the Chu State at Xinyang). 1986. Edited by Henan sheng Wenwu Yanjiusuo ^Is^X^S^ftBi- Beijing: Wenwu Chubanshe X Xizhou jiagu tanlun W JH ¥ (Research on the Western Zhou Oracle Bones). B y Wang Y u x i n • 1984. Beijing: Zhongguo Shehui Kexue Chubanshe + B*M>"$-# tB » • F/ng/ cuibian (Rubbings and Transcription of Y i n Bone Inscriptions). 1965. Reprint. B y Guo Moruo IPJ^^. Beij ing: Kexue Chubanshe ##iBJ8M£. Original edition, Tokyo: Bunkyudo X&lit, 1937. Fmxw shuqi qianbian |& J [ # M l t T l i . (Inscriptions o f Y i n Relics, Part 1). 2 vols. 1913. Compiled by Luo Zhenyu JHi$| 31. Taipei: Taiwan Y i w e n Yinshuguan p"7f H £p ilf yj'/ww S7IW#J xubian J i fcJ i t l rM^li tS (Inscriptions o f Y i n Relics, Part 3). 6 vols. Compiled by Luo Zhenyu ft&S. N . p., N . d. (Preface dated 1933). Yin-Zhou jinwen jicheng WLM&X%$* (Complete Col lec t ion o f Y i n Zhou Bronze Inscriptions). 1984-1994. 18 vols. B y Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo t H#*##Bfe#-4-^&0f. Shanghai: Zhonghua Shuju t * * ^ -Zenghou Yimu # it (The Tomb o f Zenghou Y i ) . 1989. 2 vols. Edited by Hubei sheng Bowuguan $Mk#i$&4t. Beijing: Wenwu Chubanshe tB Jifc .^ Zhensongtang jijin tu M.^lt'&^kM (Zhensongtang Col lect ion o f Auspicious Bronze 149 Inscription). 1978. Reprint. B y Luo Zhenyu HHiE.. Taipei : Tai l ian Guofeng Chubanshe & M 9 M, tB JBt^t. Original edition, N . p., 1935. Works Cited Atsuji , Tetsuji fflitfffc. 1986. Kanjigaku: "Setsumon kaiji" no sekai UXM ^ < D # # ( A Study o f Chinese Characters: The world o f "Shuowen Jiezi"). 2nd ed. Tokyo: Tokai Daigaku Shuppankai %^fc% tB Jt£H". Barnard, Noe l . 1973. The Ch'u Silk Manuscript, Translation and Commentary. Cambera: Department o f Far Eastern History, Research School o f Pacific Studies, Institute o f Advanced Studies, The Australian National University. . 1978. The nature o f the C h ' i n "Reform o f the Script" as reflected in archaeological documents excavated under conditions o f control. In Ancient China: Studies in Early Civilization. Ed . , David T. Roy and Tsuen-hsuin Tsien. Hong Kong : The Chinese University Press: 181-213. . in association with Cheung Kwong-yue. 1996. The Shan-fu Liang Ch 'i Kuei and Associated Inscribed Vessels. Taipei: S N C Publishing Inc. Boltz, W i l l i a m G . 1994. The Origin and Early Development of the Chinese Writing System. New Haven, Connecticut: American Oriental Society. Boodberg, Peter A . 1979. Reprint. Some Proleptical Remarks on the Evolution o f Archaic Chinese. In Selected Works of Peter A. Boodberg. Comp. A l v i n P. Cohen. Berkley, Los Angeles, London: University of California Press: 363-406. Original edition, Harvard Journal of Asiatic Studies. (1937) 2:329-372. Ceng, Xiantong f j g a t , ed. 1993. Changsha Chuboshu wenzi bian -fctyltB^X^fflb (Characters on the Changsha Chu Silk Manuscript). Beijing: Zhonghua Shuju 41 ^ • Mi. Chao, Yuen Ren. 1968. Language and Symbolic Systems. N e w York : Cambridge University Press. Chen, Chusheng comp. Ceng Xiantong f j g i t , ed. Reprint(?). N . d. (Postscript dated \9S6)Jinwen changyong cidian ^ X^ffl (Dictionary o f Frequently Used 150 Bronze Characters). Fuwen Tushu Chubanshe $LX HI # tB Original edition, Shanxi Renmin Chubanshe gfe W A & lti jtit^t, 1987(7). Chen, Mengjia 1^.^^. 1964. Reprint. 7*>IXM buci zongshu Jf£jjt h f $ ^ 3 & (Comprehensive Account o f the Divinatory Inscriptions from Yinxu) . Tokyo: Daian jy^jk. Original edition, Beijing: Kexue Chubanshe # # tB 1956. Chen, Shihui Hl i f r f f , and Tang Yuhu i jfr&ifc. 1988. Guwenzixue gaiyao -frX^^U^: (Essentials of Chinese Palaeography). Changchun: Jilin Daxue Chubanshe il? ^ A J ^ t B M : . Chen, Zhaorong EfcBg^. 1993. Cong Qinxi wenzi yanbian de guandian lun 'Zuchuwen' de zhenwa j iq ix i angguanwen t i^#^^**IIW« l f t* - " s i f t * " # 7 * ^ 7 A . £ * fa\ Jif (Authenticity and Related Issues of ZuChuwen from the Point o f V i e w of Qin Script Evolution). Zhongyang Yanjiuyuan Lishi Yuyan Yanjiusuo Jikan $k.tf\% ^ M £ § § A f ' J 62-4: 569-608. Chu Wenhua Yanjiuhui $LXVCffl%it, ed. 1984. C7iw wenhua kaogu dashiji JtXIt A ^ IE (Main Events in Chu Archaeology). Beijing: Wenwu Chubanshe X^S tB Cihai Bianji Weiyuanhui l ^ ^ j U t t ^ g1h 1979. Cihai 1ft ft (Cihai Dictionary). 4th ed. 3 vols. Shanghai: Shanghai Cihai Chubanshe J^MIftM tB Creel, Herrlee Glessner. 1936. On the nature o f Chinese Ideography. T'oungPao 32: 85-161. Ding , Foyan T # e T - 1988. Reprint. Shuowen guzhou bubu %X~&^%% (The Second supplement to Shuowen Jiezi Guwen and Zhouwen). Beijing: Zhonghua Shuju tf ^ • Ml- Original edition, 1925. Dong, Zuobin 1945. Yinlipu JSifff (Table of Y i n Calendar). 4 vols. Taipei: Zhongyang Yanjiuyuan Lish i Yuyan Yanjiusuo + jfe#f % BfeMjt t£"£#f ft^f. D u Ponceau, Peter S. 1838. A dissertation on the Nature and Character o f the Chinese System of Writ ing. Transactions of the Historical and Literary Committee of the American Philosophical Society. V o l . 2. Philadelphia: American Philosophical Society. Duan, Yuca i - ^ . S ^ , , annotator. [ca. finished writing in 1807 A D , published in 1815 A D ] 1992. Reprint. Shuowen jiezi zhu IfaXffl^yi- (The Shuowen Jiezi with Commentary). 151 Shanghai: Shanghai Shudian J i ^ f ^ r / i l i . Fukumoto, Masakazu ffifcfe—, trans. 1977. Setsumon kaiji jo HhXf%'%-%L (Postscript to the Shuowen Jiezi). In Chugoku shoron taikei 41 Wi!t"Hkfc3k, V o l . 1, ed. Nakada Yujiro t ffl Jj} fc g|5. 26-62. Tokyo: Nigensha Gao, M i n g tHj^ . 1987. Zhongguo guwenzixue tonglun C~£"$i (The Third Supplement to the Shuowen Guwen and Zhouwen). Shanghai: Shangwu Yinshuguan W$MP • It-Qiu, Xigu i £ # £ . 1988. Wenzixue Gaiyao X £ (An Outline of Chinese Palaeography). Beijing: Shangwu Yinshuguan W$MP #lt-. 1989. In-Shu kodaimoji ni okeru seitai to zokutai jft jfj - f r f t j t ^ l z felt S I E ^ < h ^ H (Orthodox and Vulgar Forms o f the Y i n - Z h o u Palaeography). In Shinpojumu-Chugoku komoji to In-Shu bunka ->>TKV£A: + M^X"¥ tW^MXi^ (The Symposium: The Chinese O ld script and the Yin -Zhou culture), 81-120. Tokyo: Toho Shoten J f c ^ * ^ . . 1992. Zhanguo wenzi zhong de ' sh i ' ft ® X% t ft'ffr' (S/u in the Warring States Characters). In Guwenzixwe /ww/7 -frX^ (Collected studies on Ancient Chinese script), 99-104. Beijing: Zhonghua Shuju + ¥ . First published in KGXB 1980.3. . 2000. Chinese Writing: Wenzixue Gaiyao X"¥%^i^:- Trans. Gilbert L . Mattos and Jerry Norman. N e w Haven, C T : Birdtrack Press. Qu, Wanli JS H i - 1984. Reprint. Yinxu wenzi jiabian kaoshi J M £ ^ ¥ i & # # (Translation of Y i n x u Characters: Jiabian) V o l . 2. Qu Wanli quanji S l l l ^ t (Collected works o f Q u Wanl i ) . V o l . 6. Taipei: Lianj ing Chuban Shiye Gonsi $ M f t B W H ^ %. Original edition, N . p. 1961. Shirakawa, Shizuka £ f j l | # . 1964-1980. Kinbun tsushaku ^ r ^ C i | # (Comprehensive Interpretations o f Bronze Inscriptions). 8 vols. Kyoto, Japan: Hakutsuru Bijutsukan 156 Shuihudi Qinmu Zhujian Zhengli Xiaozu S / ^ , i 4 # S / r t f f i ^ S ^ M . 1990. Shuihudi Qinmu zhujian S ^ , i 4 # S / f t W (Bamboo Slips from the Q i n Tomb at Shuihudi). Beijing: Wenwu Chubanshe X fa tB JK^t. Sichuan Bowuguan 0 Jl| i f # | f , and Qingchuan Wenhuaguan Jl| Xit^. 1982. Qingchuan xian chutu Qin Gengtianlu mudu #J I | | ${ i i±#X . 0#7fc . J | f c (A Qingchuan County Qin Wooden Tablet on Land Law Reform). WW 1:11. Sun, Haibo 3&$MJfc, comp. Zhongguo Kexueyuan Kaogu Yanjiusuo 4 1 S i ^ l - ^ l ^ ^ ' i ' W 2 5 #f, ed. 1965. rev. ed. Jiaguwen bian ¥ itXtM (Compilation of Oracle-bone Characters). Beijing: Zhonghua Shuju 4* ^ ^M] • Serruys, Paul L - M . 1957. The Study of the Chuan Chu in Shuo Wen. Taipei: Bulletin of the Institute of History and Philology 29:131 -195. Takada, Tadachika ffl 1975. Reprint. Guzhou pian "frff Jf (Commentaries on Guwen and Zhouwen). 5 vols. Taipei: Hongye Shuju feMltMl- Original edition, Tokyo: Kochuhen Kankokai ft Jjf fi] f f 1925. Takashima, Kein- ich Forthcoming. Commentaries to Fascicle Three of Inscriptions from the Y i n Ruins: Palaeographical and Linguistic Studies. Typescript ca. 600 pp. completed. Accepted for publication as part o f Synonyma Serica Comparata. Ed . Christoph Habsmeier. A web-site publication at the University o f Oslo, Norway. Tang, L a n J j f f l . 1963. Reprint. Zhongguo wenzixue t S ^ ^ f 1 (Study o f Chinese Palaeography). Hong Kong : Taiping Shuju A^p-lr / t^f. Original edition, Shanghai: Kaiming Shudian 0fl # 1949. . 1965. Reprint. Guwenzixue daolun -frX^-^^r^fr (Introduction to the study o f Ancient Chinese script). Hong Kong: Taiping Shuju j^^-^M}. Original edition, N . p. 1935. Tang, Yuhui 'M&M- 1986. Luelun Zhanguo wenzi xingti yanjiu zhong de jige wenti n&^ffc BX^BfliWft^ttlMMfflM ( A B r i e f Discuss ion about Several Problems Regarding the Study of the Form of the Warring States Characters). GWYJ 15: 9-95. Teng, Rensheng B£*k- 1995. Chuxijianbo wenzi bian ft^H %X%-tM (Compilation of 157 Chu-style Bamboo Slips and Si lk Manuscript Characters). Wuhan: Hubei Jiaoyu Chubanshe ftUkSfclf Hi Todo, Akiyasu f f 1965. Kanji gogen jiten ^^ifMffiM (Dictionary o f Etymolgoy o f Chinese Characters). Tokyo: Gakutosha Uchino, Kumaichiro ft ^ $ — : 1962. Mos/u(Mencius) . Tokyo: Mei j i Shoin Ijfl j £ # f t . Wang, Guowei 1956. Reprint. Zhanguoshi Q i n yong zhouwen liuguo yong guwen shuo ^M^^M^XfsM^-^X^L (The Theory that the Qin used Zhouwen and the Six States used guwen during the Warring States Period). In Guantang jilin IS, jit H V o l . 7 : 79. Taipei: Y i w e n Yinshuguan Xtf1t$- Original edition, Beij ing: Wucheng Jiangshi Mi%M ft, 1921. . 1968. Wang Guantang xiansheng quanji i | | 1 | : # ; ^ . £ | l (Collected Works o f Wang Guowei). vol . 16. Taipei: Wenhua Chuban Gongsi X ¥ Hi Wi& ^\ • Wang, H u i j£#f, ed. 1990. Qin tongqi mingwen biannian jishi #3HH#«3t (Qin Bronze Inscriptions Arranged Chronologically with Commentary). Shanxi :Sanqin Chubanshe = # Hi Jfc^ fc. Yoshida, Kenko t 0 frft. 1979. SAiJb" j t t E (The Records o f the Historian). V o l . 6. Tokyo: Mei j i Shoin ^ • f t . Y u , X i n g w u 1979. Jiaguwenzi shilin f # (Commentaries on the Oracle-Bone Characters). Beijing: Zhonghua Shuju. + ^ • Y u , X ingwu chief ed., and Yao Xiaosui ed. 1996. Jiaguwenzi Gulin ^jtX ^ft#- (Collected Commentaries on the Oracle-bone Characters). 4 vols. Beijing: Zhonghua Shuju 41 ^ • Zhang, Shouzhong 1996. Baoshan Chujian wenzi bian&di&MX^Mi ( A Compilation o f Baoshan Chu Bamboo-slip Characters). Beijing: Wenwu Chubanshe Zheng, Jingruo 1981. Lunyu Zhengshizhu jishu HHf IP ft y i f t ^ . (Account of Collected Zheng Commentaries on Analects). Taipei: Xuehai Chubanshe i^yff tB 158 Zhao, Cheng 1988. Jiaguwen jianming cidian ^ Beij ing: Zhonghua Shuju 4'^ # -^Zhonghua Shuju + ^ • Mi, ed. 1959. Shiji jfc. IE. Beijing: Zhonghua Shuju + ^ • ^ • Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo + ^#"fr^f & ed. 1984-1994. Yin-Zhou jinwen jicheng (Complete Collection o f Y i n Zhou Bronze Inscriptions). 18 vols. Shanghai: Zhonghua Shuju 41 ^IrM}. Zhou, Fagao M&'ffi, chief ed. Zhang Risheng % 0 jff-, X u Z h i y i ^ j t H , and L i n Jieming comp. 1974. Jinwen Gulin $tj(lifc$- (Collected Commentaries on the Bronze Characters). 16 vols. Hong Kong : Xianggang Zhongwen Daxue 41 % A #. Zhu, Zhunsheng 1934. Reprint. Shuowen tongshun dingsheng t J t ^ t i ! | f i | ^ [ l £ (The Shuowen Annotated and Phonologically Arranged). Shanghai: Shengwu Yinshuguan T t l P P # l t . Original edition, N . p., n. d. (Preface dated 1833). 159 Appendix I Table: The Comparison of the Chu and Qin W S Graphs (Chu graphs top, Qin graphs bottom) ft Chu Yangtianhu bamboo slip # Qin Qin wooden tablet ¥** Oracle-bone S d j Wangshan bamboo slip WM&X W Z bronze *t Silk manuscript "B&X W Z bronze l i f t * ZuChuwen stone tablet &&X S A bronze Baoshan bamboo slip m & x W S bronze Shuihudi bamboo slip Zenghou Y i bamboo slip Modern grapheme Stone drums Modern character 'ff" Xinyang bamboo slip ft f t * mM W i r * t i r * •SB* aft *« ** # W i r * l ^ i r * ^&* 7kf f l l i f t * 1 — — — — — — — — — — — 2 } i L 3 n A A A A X X A /V A 4 5 • fl + + • + 1 t+. 4 + + 6 1 1 X 1 7 71 8 U u 9 A 7? 10 = — = = ' — — — — = 11 12 +t -h + -b - 4 -160 ft ¥ t * W$ W i r * * i r * ftir* am aft ** # W i r * * i r * t i r * f^t* l i f e * 13 i h _ >l K fe 14 - T 15 — _ _L — ± 16 o— — TT? T T T T T — "F T 17 =-— — = — - — — -— 18 ± ± +± 19 ft f 20 T\ TT m / \ * 7T 7T<£ 7T 21 T TT ? ^ = J 22 tt ±+ t t t t + 23 OJ 24 t t * X ft ft 25 •k <¥ 26 OK t 27 -t (6^ 161 i ft ¥t* i r * W i r * * i r * a i r * am ft* « * a f t * * # W i r * * i r * a i r * l i f t * 28 ? ? 29 H K ft* 30 A i l ±. ± ± 31 e 32 tW.Q. f f f ? f ? f f 33 ? c 11 IZ a EJ 34 MS it f 35 i n ± T T # I I I I I EE I 36 • t t 1 ¥ i + t 37 t t 38 i v th >> >> 39 40 I y \ o ti A o A o A a 41 ¥ T 42 f 3^ 162 ft W $ W i r * ftir* t i t & d j 4 t t a f t a f t # W i r * « i r * u s * Al 44 fi AS IT H SL HD Af>. eta y yy y y y 47 TT /v. T f M M M ft 40 AO di ft 4V I % t C I i n % X i 52 l m 8» 53 C D 0 e © SA K 9 DH c c ft f t *l ft ft JO tt ^7 ¥ J / C.Q n Jo t 59 f t rt 163 ft WJH i r * $ i r * aft aft # W i r * *** l^ir* *** aft* 60 61 * 62 * 63 ft ** 64 w •ar t 65 XI i * 66 f > 67 f J 7f\ 68 * # * 69 Cx 70 m i t X * 71 X IT X E X s : 72 Aft ft M l \ _ 73 I I I 74 4 5 164 ft ¥ t * W i r * #ir* ftir* #* aft aft # W i r * *ir* air* n t * 75 Ml t A * 76 1$ I 77 I • f c * « * 78 =1 11 ft a 79 80 a "V (P i IE 81 82 6| if) 83 ss 84 85 86 87 1* H 88 89 90 f t f ¥ k f 91 1 165 ft A * & i r * aft ** # Wir;* l^ i r * ^ 4 t * n t * 92 * * * 93 sty 94 * ± i 95 i ta 96 d a o 0 -S. 97 5* ?K 98 # * 99 e 6 6 e 100 )f 101 rt or 102 n 103 MM UL It 104 51 y ^) 105 (tf e JoJB 106 4 ± i ± ± 1 166 ft ¥ t * W$ i r * W i r * * i r * m i r * a f t # Wir* * i r * $ i r * ^ « t * l i f e * 107 t ir FP # 108 109 It t # a 110 t UP t 111 ffll IB © ffl ffl © ffl 112 113 fa i 114 — — E (ID eg •P 115 +0 + © l£t± Eatfc ¥ ¥ ¥ ? 116 117 118 11 b » \ 119 tl * *** 120 1% I ¥ s •)» 121 4 -8- 4-a. 167 ft ffl <& WJU * i r * $ i r * auj # t 'feuj'tt'lBF ? l f t J l f t # T f A i r * Wir* * i r * $&* tife* 122 ± ¥ 4= ± b 123 °) ? 124 oik n 125 J L ft fr - i r J L ->r ft 126 • A 127 R tt t t 4f 128 E we 129 ft 130 * 131 132 & S 133 SISi Ti 134 135 136 ft 168 ft W$ i r * * i r * ftir* •** aft « * aft # W i r * * i r * l ^ i r * fc® l i f e * 137 f*| 138 ft if 139 140 t # 141 tT 142 ft TP 143 • % IP 144 145 & 146 % 147 f t 0 148 i \ 149 .33 ? J 0) O 150 $ $ 151 If ft T 152 i l l Lil I N rn /l~-^ N 153 154 isrl s / f t 169 ft ¥ t * W i r * * i r * a i r * ft* aft aft # i r * W i r * * i r * a i r * tf&* l i f t * 155 ft Is) 156 157 ft - f t ft 158 ft # 159 n ft 160 A V, 161 162 4 163 f * f 164 165 o o o o oo 166 $ 167 01 IT / E f t 168 ms IB IM 169 n I t 170 i t 171 i t m % 172 in m in M ))) !n 1 170 ft f t * W$ W i r * # i r * a i r * ft* a f t « t aft # W i r * * i r * a i r * tfJi* l i f t * 173 1 i 4 4 174 m 6 S 175 Si 176 Bra * (g/ 177 178 179 180 181 182 T f 183 a 184 * tl" 1 * 185 186 i t a % 187 XI X c * X o 188 1 © 189 190 171 ft w $ i r * * i r * l i r * a f t ** # W i r * * i r * a i r * * i i * t&ft* 191 * 4 * * 192 193 194 195 a 196 197 198 1 199 o 200 201 f t 202 LB fe Ik 203 * 204 * 205 * 206 207 208 6 & *# 172 ft f t * S i r * >lir* f SZ. »* aft a« at* # S i r * a i r * fc® l i f e * 209 210 f If If m fk ft 211 \ i ll. i f 212 4! 213 K 214 ® ffl 215 m K ? 216 XI o 7 T 9 0 217 I * l 218 219 220 ft 221 7$ 222 flP 223 224 0 M 225 226 227 ? 173 ft i r * * i r * ft* aft # S i r * * i r * l i f e * 228 "8 $ * 229 230 231 232 if 233 ^ 234 i i t 235 f l 236 fS 237 « 238 fE 239 f 240 tt * % 241 m t f 242 ft 243 244 245 f 246 tf 247 Tri- i r dr 174 ft ¥#* WJH W i r * * i r * ftir* aoi 'fciu'tt-fr aft aft # W i r * * i r * a i r * I E * * 248 249 250 251 Si 252 is 253 W 254 f f a 255 g E 256 X IP 257 W ti 258 S. 259 « ) 260 j£ 261 in 262 f J * 263 it 264 /ft /T\ II f 265 266 "Ms-267 7& fit 268 ? T f 3 T 175 ft f t * W i r * * i r * l ^ i r * ft* a f t « » # Wir;* * i r * a i r * l i f e * 269 w f § § S 270 f #-271 t t TP 272 f r fc 273 i f 274 275 276 IS) [a) 277 ^ ^ ^ ^ ? 278 279 m cT 280 281 282 283 i i 284 285 286 287 288 289 ft 4 * # i r * a i r * «* a« «» aft *# # W i r * #&* a i r * *&* 290 It * 291 292 n 1 1 293 it 294 295 fe 296 % f H i 297 IE * 298 U e e MI ft 299 tf 300 i l 301 1-4 * m o p j u 302 303 1 * 304 $ 1 305 4 177 ft f t * ?§m i r * * i r * t i r * ft* • B i l l - i t * a f t a f t # W i r * # i r * a i r * t i f e * ***** 306 t s f $ % 307 % 308 * 309 Hi $ 310 ! ** * t 311 ft 312 ft 313 I 314 315 316 317 318 319 320 % 321 IP 322 323 !£ gp 324 325 326 327 328 W 8 Kb Pi M 178 ft • f t * WJH i r * * i r * « i r * ft* a f t Hit a f t * * # Wir* * i r * a i r * 511* l i f e * 329 * ft 330 (Si ft) 331 T J 332 a $ ft* ir m 333 334 am 335 1ST 7TJ 336 ft If 337 ® 338 339 340 341 342 it 343 & ft 344 ft 345 fe B 346 fi fir M; 347 /% 348 fl iff 349 If * 179 ft ¥#* W$ i r * Wir3 * i r * am «ft ** # * i r * a i r * l i f e * 350 351 352 f 'IS 353 'II 354 vt ft & 355 356 1 <# l tf 357 358 It 359 ft * 360 f 361 362 m ** * * * If t f 363 364 365 P 1 t 366 m * ft 367 m 180 ft ¥#* W$ i r * S i r * ft* 'feill'frfij a« «* aft # *** a i r * tf&* l i f t * 368 Si? JS f Ik 369 ±s 370 371 f 372 # 373 374 fe 375 * t f t # £ i & I I I 376 *W!r «f 377 P # ft 378 t * 379 ft 380 0 381 w f 382 ** * MM* 383 1 384 385 a I i 386 t 9 181 ft ¥ t * i r * ffiir* * i r * ftir* ft* aft ** # * i r * l ^ i r * t i t * 387 •2/4& <») 388 & 389 at f # * f 390 391 1 i i * 1 I 392 feffc i i it 393 394 ft 395 396 a 397 398 f f f f * IS 399 1 * 400 401 402 403 A 404 405 f t 406 ft te p 182 ft ¥ t * W $ W i r * * i r * f ^ i r * ft* IjLUJ'frlBf ?lft aft * * # W i r * * i r * a i r * 5 f t * fc® l i f t * 407 * 408 * 409 » ir 410 f t 411 j T 412 413 414 w s 415 f t 416 y 417 7\ 418 4 t 419 420 421 II t 5v 2 N 422 A w # o 423 t 1 f > iv Jam 424 t St 1 8 425 ^7 f I * $ 426 % T1 183 ft ¥ t * mn i S i r * * i r * *# a f t a f t * * # i r * -S i r * * i r * a i r * £&* l i f t * 427 428 429 * 430 All 431 ^ % 517 *c 518 519 520 $ t tt 521 % m 522 523 >I>I 524 525 526 f i T 527 ** t 528 529 JK 530 531 h 532 533 * 534 ff * n 535 * 536 I* 537 f 538 fit) f 188 ft ¥ t * Will i r * Wir* # i r * « i r * ft* a f t ** # W i r * # i r * a i r * 5 i i * l i f e * 539 540 541 & (51 542 543 544 545 * 546 547 f i m * 548 549 1 550 551 552 553 t 554 1 &P 555 * 556 557 IP 558 559 IP 560 IP 561 # 562 ap 189 ft f t * W i r * * i r * ftir* a« ** # W i r * * i r * a i r * u s * 563 i § § Iff # 564 fi 565 * 566 ft 567 568 » 569 $ 570 f 571 f * 572 573 574 f t t 575 f 576 577 * 578 579 # 580 tf ft j 581 582 ft 583 584 • * * 585 * 586 4-190 ft ¥ t * WJS i r * W i r * * i r * ft* ft* a f t a f t # W i r * # i r * a i r * 5 iXX l i f e * 587 * ** 588 589 590 ? l f 591 11 592 593 to 594 595 596 597 t n f 598 II 599 600 It 601 m w At 602 603 # 604 605 % 1 606 m 1 s 607 It ML 608 fc it % tt 609 «• 191 ft f t * W$ i r * W i r * * i r * « i r * ft* 'feilj'tt'lBf a f t a f t m # W i r * # i r * a i r * ^ ft* fc® l i f t * 610 1 ¥ & 611 612 $ 613 $fc$fc 614 615 ft 616 % 617 * 618 iS 619 ft 7TT * 2£ TW 620 r*j> 621 # 622 623 t t o r t m 624 "S 625 Oft * 626 627 ft 192 ft ¥ t * •kX S i r i * i r * a i r * ** a« ## # Wir;* * i r * K i r * 628 fl *t # •± It 629 630 631 It * 632 633 all 634 635 # 636 ft 637 ft 638 f 9 f f * I 639 f f T * f #* 640 i | si 641 S lie & 642 * * 643 644 * 645 i t 646 ft* Uc 647 ft He kk 648 193 ft f t * Si: ;* ft* a « « » a f t # l i f e * 649 * W s f t 650 # 651 652 653 W f 654 # 655 ft I I 656 # 657 JK 658 i i 659 A*. 660 # 661 tt>te w 662 * * ft 663 M ff tt ff 664 * 665 666 m 667 668 # 669 i t i t 670 i g • 194 ft ¥ t * *&* «&* * t aft # *&* «&* 5 S i * 671 i i f 672 * 673 # 674 675 676 677 678 SP 679 if IP 680 # 681 as* * P 682 * 683 684 685 686 IP 687 % 688 *P 689 *P 690 1 p691 * 692 693 i s I 694 695 & 696 1% 78 195 ft *• a f t 0* a f t # 697 1ft 698 699 0 700 % 701 702 f t 703 704 * • f 705 « 706 # 707 708 709 w 710 711 712 & 713 I I 4 JB 714 Ik 715 716 $ 717 ft *** Mi 718 i 719 * 720 A 196 ft 1BB ft* 'fi.Jj'ftlBf aft aft m # Wir;* lit* 721 ** 'IS 722 #. 723 9 M 724 725 726 l i 727 ill? 728 729 f 730 * 731 732 IT ft 733 * i f f 734 735 3* ft 736 it 737 €^ 738 g IS lee.739 s 11-740 5ft 741 t* (>V\. 742 f ft Al # 743 te •T5T fe fate 197 ft f t * Wi5 I §ir3 ^ i r * & i r * 9 « r . Wr^Jft ft* » * • i r * I 5ir3 # i r * K i r * 5 f t * l i f e * 744 # 745 . ft M 746 747 748 749 III i f 750 751 752 753 754 755 756 757 758 759 i l ft 760 761 tit M ft * H * 762 & & 763 f 764 9 765 %%% 7113 766 A A f t 767 * 768 198 ft K i r * ft* aft aft ** # Wi r ; * t i t * 769 % 770 771 * 772 i l 773 4H ih * #* 774 775 _ f 5 _ m # 776 777 * 778 X- ** 779 780 at 781 Ik f I S If 782 783 ft it 784 if 785 t t ig 786 787 788 # • f t 789 790 791 a? * 792 #* ft 199 ft &x a f t 0* a f t # 5 ft* l i f e * 793 ft H * f#i 794 T t 795 l & * 25 796 * 797 798 k 799 800 801 802 803 ii 804 * 805 TK 806 i f 807 ft 808 ft #1 809 810 ft 811 812 t * 813 * 814 815 816 817 Si I * 200 ft wn Sir;* fciljttfl a f t 0* a f t # Sir;* t i t * 818 819 1 I ft i *** * Hi. 820 I; lit s % 821 i & 822 823 # 824 5 825 826 827 828 * 829 ** 830 i 831 ffl i 832 f 833 & 834 f It 835 ft 1 836 3$ IP 837 % IP 838 % » 839 % IP 840 & % * 841 f mm IP 842 IP 843 IP 201 ft ¥** Sir;* * i r * Slilj ft* a f t # Wir;* # i r * « i r * t i t * 844 # * 845 846 fe • 847 at a ($1 848 # 849 * 850 FF) tf 851 m 852 # 853 854 i i I » * f B $ 855 * 856 857 * 858 859 * 860 r 861 862 863 * 864 865 ft 866 ft 202 ft ¥t* WM &X » t a f t 0* a f t # %ikX 867 * * 868 * 869 »• 870 S t 871 f 872 873 * 874 875 876 f # 877 /^•> 878 1 ### f 879 t If 880 881 JE 882 i m ;ii 883 884 885 ft te IE It 886 H 887 888 889 890 891 203 ft ¥** i r * S i r ; * * i r * m i r * S i l l ft* aft aft # sir;* * i r * K i r * l i f e * 892 18 i f 11 w # It 893 fr * 894 i f® 895 896 # 897 # * 898 899 900 901 If UK 902 IS 903 904 § t § ± 905 JS 906 907 908 i"4> if # 909 910 M 911 ft 912 m 913 1 f a 7T) # * 914 % * 915 916 3R 917 U - J , 918 3fc 204 ft ¥t* a * a f t a f t # ^ 4 t * s i f t * 919 920 921 I 922 923 fl 924 m si 925 m 926 f m 927 928 929 * 930 931 932 H f 933 % 934 ii f t fe * Iff 935 ¥ 936 937 938 A 939 ft* * 940 * n 941 # HI 205 ft ¥#* ISM i r * * i r * K i r * am g i b *# aft « t aft # Wir;* * i r * K i r * 5 S i * aft* 942 *** % 943 % 944 945 if a. fr" 946 9 947 f M 948 949 tt 950 i t # # t 951 tt 952 w 953 n °* m f ft if 954 955 w tt 956 ft 957 a» f f * • t 958 1 ft 959 960 961 SP 962 963 f ii if 964 8* $ 965 5^ * ft 966 ft 206 ft ¥ t * mn &% ft* a f t a« ** # 5 f t * 967 * 968 969 970 971 972 1 lie 973 Jl 974 te 975 976 i t 977 978 f 979 * 980 I) ? Ti WW 981 u 982 m 983 *& m 984 i w i i i 985 ft 986 JR 987 JI JR 988 # 989 Hi 4 t 990 11 991 1 992 $ * • 4 / Jit 41$, 207 ft f t * ft* 0* a f t # l i f t * 993 H 994 995 996 tl # 997 998 999 * 1000 9 H PD 8 1001 4 \* £ 1 1002 g $ 1003 * 1004 ft 1005 ft 1006 * f M 1007 It 1008 m 1009 4* i i t& 1010 m 1011 & 1012 $ 1013 1014 * * I I n * £ * * f 1015 If ft ft It it 1016 11 ft 1017 It 1* 208 ft wn &x W&3, ft* aft ** # &&X 1018 1019 IS 1020 1021 % w Br* ff f f fr f/r it ff 1022 $$|* 1023 ft 1024 = f c , J -1025 Iff * t 1026 1027 m * f % w 1028 # 1029 /it 1030 1031 1032 U f 1033 1034 v i o l / 1035 IS 1036 1037 If 1038 1039 1040 it 1041 fc if Mi 1042 209 ¥ t * mM i r * wir3 * i r * K i r * a f t 0* a f t m # m&t * i r * K i r * 5ft* t i * * 1043 1044 tl s it 1045 IS II 1046 1047 • # 1048 fete 1049 it lie 1050 w 1051 if 1052 1053 it 1054 p,l •* 1055 1 JK 1056 " I T ^ i 1057 it n 1058 i f 1059 ¥ 1060 1061 4T I 1062 1063 A h 1064 Jul, 1065 ft 1066 210 fe ¥ t * §ir;* #&* K i r * I t * 1ft-• rift I * # S i r ; * # i r * K i r * £&* 1^* 1067 f it 1068 IP 1069 W 1070 IP 1071 SP IP 1072 HI 1073 *v t'r ** * 1074 ffl $ * It 1075 1076 1077 7 ^ 1078 Ii 1079 # 1080 h ft ft % 1081 1082 % 1083 II iff* Ii ft-1084 ft 1085 & 1086 ft 1087 iff it fc* 1088 • ffi 108< f'-& 211 ft f t * &* *&* «&* ** a f t a f t # *&* $&* *&* l i f t * 1090 1091 1092 tt m 1093 fe if 1094 f i f I $ t 4 I I 1095 % "s 1096 s 1097 1098 it 1099 f 1100 1101 f f i i * H t 1102 ft 1103 ft * 1104 It % If i f 1105 H 1106 1107 1108 rtrr 1109 1110 l l l l 1112 1113 212 it ¥t* i r * Bir3 ^ i r * K i r * ft* aft aft # Bir3 * i r * K i r * 5 i i * ma* m$M® 1114 fi 1115 m ft 1116 a* ft 1117 1118 II 1119 1120 It 1121 1122 * 1123 1124 it 1125 # f 1126 ft it 1127 1128 f 1129 §8 Ii 1130 \ If % ft * 1131 it ft 1132 Be 1133 JK 1134 8 1135 ft 1136 1137 1138 2 213 fe ¥#* i r * Sir;* * i r * K i r * am ft* aft aft # ffii:;* * i r * K i r * 5 ft* fc® l i f e * 1139 If If 1140 1141 * 1142 f 1143 IT 1144 t# 1145 i l 1146 I $ % * ft 1147 7yz y&~- -Ay* 1148 ! * ** * * « 1149 1150 1151 f f 1152 % 1153 1154 to 1155 ffi -Suu 1156 # 1157 1158 5 1159 * 5 1160 fl m 3 1161 * 3 1162 IP 1163 f # § * 214 ft ¥#* Wir;* * i r * $ i r * SLib #* &djft0 a f t a f t * Wir;* * i r * $ i r * ^0 l i f t * 1164 ffl 1165 fl * * 1166 « 1167 t 3 1168 ft ft 1169 I f 1170 * 1171 ft 1172 # 1173 1174 1175 H 1176 & 1177 * tf 1178 1179 i i i t 1180 1181 Kl V FI 1 1182 w 1183 1184 * 1185 & Si 1186 I* ^ ^ ^ ^ 1187 n E S 1188 215 fe ¥t* i r * Wir;* * i r * K i r * am ft* aft aft # Wir;* # i r * K i r * 5 f t * M £ * 1189 IK 1190 1191 fe HI 1192 # ft 1193 if 1194 m 1195 # » 1196 1197 1198 » 1199 4- | # 1200 if ft 1201 1202 # * 1203 n f 1204 ff a WW XI 1205 n 1206 1 i # 1207 n 1208 77T f tt 1209 1210 1211 If 1212 II 1213 216 ft ¥t* &% *&* $&* HB #• a f t a f t * * # *&* $&* a * * 1214 1215 A 1216 # 1217 1218 % oxo krtb I 1219 II 1220 # 1221 * 1222 1223 1224 ft m ft 1225 % ft 15 HI 1226 ft 1227 f £ * 1228 n Tr- <7N * 1229 1230 l i H gg 1231 HijA *W 1232 IP 1233 $ # 1234 IP 1235 IP 1236 |P 1237 ft IP 1238 fix i f 1239 H 217 ft WM #* ftOi^« aft aft # l i f t * 1240 "# # 1241 fi 1242 1243 ^ ^ • ^ 1244 f 1245 if ft 1246 ft 1247 JK 1248 H 1 1249 ff 1250 ft is */ 1251 ff If it 1252 * 1253 1254 1255 f if 1256 • A h 1257 t 1258 ft t 1259 ff $ 1260 $t 1261 h ff ft It If go oo . 1262 #f # 1263 1264 1265 1266 i i 218 it ¥ t * mM i r * # i r * K i r * & L b t t f l a f t 0* a f t m # sir;* * i r * K i r * 5 ikX l i s * 1267 i f 1268 # 1269 i $ ft 1270 1271 1272 1273 it H£ 1274 1275 It £ 1276 1277 1278 & If * 1279 1280 i t 1281 It 1282 i f 1283 II 1284 1285 ft * 1286 f t 1287 i f l f * 1288 M 1289 Ii 1290 1291 ft 219 fe ¥** WJ? i r * Wir3 * i r * K i r * ft* 0* a f t # Wir3 * i r * K i r * 5 ft* ffift* tSj&ilkttffi 1292 I f $ 1293 1294 1295 1296 1297 1298 1299 1300 f t 1301 1302 f 1303 &® 1304 rr 1305 * 1306 1307 SI ,g> | 1308 n * t f 1309 # IS 1310 P 1311 ? i fe IK $ I t ft ft i t 1312 1313 AM? 1314 m m • t j d ' U O i i <=& B B U i en B O O 9 EE b a a /IS O O P 220 ft •ft* &* Sir;* *&* $&* Midi ft* a f t a f t # *&* &&* *W l i f t * 1315 1316 w& 1317 tp If 1? 1318 # 11 $t 1319 If I M ,# 1320 #fe 1321 ft 1322 1323 « 1324 If 1325 ¥ 1326 f 1327 i 1328 & ft 1329 # 1330 1 * 1331 s ft 1332 1333 ik 1334 it 1335 1336 w f 1 It t 1337 & 1338 i f 1339 * $ 1340 1341 a 1342 ** a 221 fe ¥ t * WJ? i r * Wir;* * i r * K i r * ad/ 'S. ill # ffl a f t # Wir;* * i r * K i r * 5 f t * *n t&&* S^ftitffl 1343 II 1344 1 ft 1345 11 1346 fe $1 1347 ft 1348 ft 1349 Ik 1350 $ 1351 ti ft ft 1352 1353 ffl 1354 » 'ft 1355 ft 1356 f f % V ft 1357 s 1358 1359 1360 ft 1361 * 1362 1363 * tt 1364 1365 t ft « ft 1366 1367 1 9 222 ft ¥ f * WJ5 &* Sir;* * i r * $ i r * ft* a f t a f t * * # * i r * $ i r * t i t * 1368 « & 1369 n 1370 ft 1371 IS 1372 1373 1374 1375 1376 1377 & 1378 fit 1379 1? 1380 it TIE 1381 M 1382 1383 i IS 1384 # 1385 If 1386 1387 1388 1389 It 1390 » 1391 1392 1393 t i It t 1394 0,5 223 fe f t * mM i r * Wir3 * i r * K i r * aft 0* aft # m&% * i r * K i r * 5 ft* 7)cffl iaft* 1395 i t 1396 m 1397 ft 1398 ft 1399 « fti$ 1400 * f 1401 i t ft 1402 1403 m % ft 1404 n 1405 1406 X 1407 ft f l it 1408 ft ft 1409 i t ft ft 1410 $f If 1411 i t 1412 % » II ft 1413 IS 1414 ftft ft 1415 ft ft 1416 ft 1417 ft 1418 % 1419 1420 i$ 1421 U n S£ 224 it ¥** i r * Sir;* * i r * $ i r * ma; ft* lift «rt aft # Sir;* * i r * $ i r * 5 si* life* 1422 ft 1423 ft 1424 ft 1425 1 1426 H 1427 H IS 1428 fe 1429 ft 1430 II lit 1431 a a i B I f m II * s 1432 Is 1433 I f ft 1434 ft IS 1435 a 1436 ft ft 1437 f t « 1438 I f II 1439 t 1440 & t t 1441 i _ §!? a . 1442 ft 1443 » 1444 ft 1445 ft 1446 ft 1447 « 1448 i t 1449 ft$F i 225 ft &x a oi a f t a f t # &&X ZikX lift* 1450 1§ 1$ 1451 & 1452 it 1453 n 1454 tt 1455 aw Dt ft 1456 # 1457 If 1458 S It M 1459 * 1460 ft 1461 I 1462 IS 1463 # 1464 1465 8 B 1 1466 #^ ¥£ 1467 It ft 1468 ^ m 1469 If 1470 ft m 1471 $ fi 1472 fi 1473 m 226 Appendix II List o f Graphs from the Western Zhou Bronze Vessels This appendix provides an index o f the source materials from which the Western Zhou (T5 ffi&SO bronze-vessel graphical forms given in Appendix I have been taken. Appendix II gives between one and three Western Zhou forms for each entry, taken from the bronze vessels as numbered in Appendix III. The order of the bronze vessel numbers below matches the order of the Western Zhou variant forms. Entry Bronze vessels 1 78 2 78 3 79 5 24 6 78 9 24 61 10 78 12 158 13 62 14 79 15 155 16 135 17 78 18 62 19 2 20 143 10 21 24 99 22 78 5 23 61 24 143 25 61 27 155 28 158 Entry Bronze vessels 29 78 30 79 136 31 137 32 26 78 33 2 34 79 142 35 61 36 143 37 143 39 121 40 24 129 41 78 42 108 136 44 142 45 143 46 78 47 10 48 81 49 79 50 78 51 136 52 111 53 78 Entry Bronze vessels 54 79 55 78 136 144 56 8 57 2 58 5 59 24 2 60 9 61 61 62 126 63 78 64 110 61 65 45 66 28 67 78 68 107 69 61 70 46 71 78 72 78 73 46 158 74 172 75 27 76 2 227 77 138 78 79 80 79 111 82 122 83 5 84 82 86 100 87 6 118 177 89 37 90 6 91 38 93 145 94 76 158 96 143 97 2 99 78 101 158 104 155 105 1 106 158 107 143 108 84 109 78 110 101 111 78 112 39 113 20 114 61 115 143 10 116 64 117 27 118 78 119 43 120 85 121 73 122 79 123 26 124 47 125 11 126 49 127 12 100 128 78 61 131 2 132 78 134 78 135 78 136 143 140 127 148 136 149 79 150 50 151 78 154 150 155 45 160 143 161 23 162 102 51 163 101 165 78 166 87 167 46 169 35 100 170 52 171 53 172 136 173 100 174 65 175 54 176 60 181 128 145 182 87 183 111 184 78 155 185 149 186 78 187 66 2 188 136 191 5 192 88 198 6 116 143 199 89 200 67 202 143 204 136 205 136 208 69 209 55 210 5 211 119 212 66 215 115 216 116 217 5 226 90 228 41 234 143 236 100 239 177 147 240 135 228 241 66 155 242 2 243 153 244 78 143 245 116 254 81 138 257 78 258 43 259 62 264 5 90 267 143 268 92 269 78 276 78 277 73 278 28 279 89 78 283 173 286 99 78 170 288 61 289 136 290 3 292 143 293 43 10 296 143 301 119 303 116 305 71 306 93 309 78 310 140 328 68 329 110 330 70 332 149 334 46 95 341 89 349 139 354 152 356 66 359 78 361 78 362 36 82 364 78 365 101 371 10 375 78 378 143 380 14 381 151 382 90 383 78 384 143 385 143 387 6 2 388 66 389 86 391 2 392 57 399 128 401 2 403 118 407 117 409 103 412 104 421 56 422 26 423 155 424 5 425 105 430 5 431 78 448 145 451 143 453 115 454 78 458 13 459 155 463 143 469 78 142 471 149 473 5 477 158 478 87 493 43 495 87 501 136 506 17 507 78 510 24 157 511 44 513 81 515 100 516 16 526 4 535 135 2 536 101 538 15 541 177 229 563 120 569 143 570 124 571 2 575 2 580 148 584 164 590 5 29 591 77 595 7 597 119 136 601 116 605 143 606 80 612 118 619 34 623 136 625 88 627 78 628 142 637 179 638 142 639 78 640 119 641 27 642 97 647 136 650 130 662 35 663 81 670 91 2 671 125 673 155 691 116 693 73 696 78 709 18 10 717 85 720 2 727 61 730 135 732 24 733 131 739 5 743 111 749 61 756 40 759 166 761 101 762 5 775 137 777 167 778 78 141 781 71 98 169 786 29 788 171 143 793 2 795 78 798 5 90 806 2 808 138 817 78 819 34 820 132 821 32 830 134 27 832 143 845 111 846 78 851 117 854 61 860 72 878 26 881 2 882 112 21 885 106 892 5 142 894 2 135 921 100 923 74 927 61 932 5 933 133 934 2 938 143 73 943 146 945 61 953 2 980 83 1000 155 1001 110 1006 113 1014 42 1015 33 1021 143 1024 27 1025 161 1027 79 1041 89 164 230 1044 7 136 1051 94 1061 154 1073 63 5 1074 30 1083 142 61 143 1090 100 1092 26 1093 10 59 1094 135 2 1101 2 1115 136 1121 150 1124 158 1125 136 1130 143 1138 109 1146 19 1148 25 1151 78 1152 5 1163 61 1186 73 1206 6 1218 150 1221 123 1224 117 1228 2 1230 75 58 1238 87 1261 126 1278 89 1280 78 1307 5 1311 63 101 26 1314 61 1330 10 1336 87 1344 26 1351 142 1356 163 48 1371 149 1383 155 171 1393 5 1399 5 1406 176 1407 135 2 1431 61 114 1465 113 156 231 Appendix III List o f the Western Zhou Bronze Vessels 1 26 3 ftl 51 7^ "1 76 2 27 52 3ci 77 ^tei 3 A i 28 53 M l 78 i i 4 29 £ « 1 54 79 i t 5 A&i 30 55 ft* 80 6 31 A M 56 #*S1 81 7 * e&i 32 57 #JK£1 82 8 tbl 33 58 83 9 $1 34 59 # £ i 84 10 35 60 85 -frl 11 36 61 £ i 86 £fcte#l 12 37 *r&i 62 87 * £ 13 38 .^i 63 88 #!$ 14 £ * 1 39 64 89 t l 15 40 *pl 65 90 *tel 16 41 66 91 t i l 17 42 $ i 67 te£i 92 t l 18 ^ i 43 68 fill 93 *1 19 44 69 te^3C^ 94 &#i 20 #A-£H 45 70 95 a t £ l 21 +;te£i 46 71 4 i 96 Ml 22 47 itki 72 97 23 48 73 S K I 98 24 41 49 74 99 25 50 75 #*A i 100 232 101 126 ft* 151 177 102 127 152 # s 103 t £ 128 153 # 1 104 ** 129 I f t H 154 105 130 155 106 131 156 107 132 157 108 l f £ # i 133 158 109 l f £ $ 134 #* 159 110 i f a * 135 # £ t 160 111 i f S . t 136 161 112 if M i 137 #3f 4 # 162 113 i f M 138 163 « 114 139 164 115 140 165 116 If W t 141 166 M S 117 142 *« 167 118 143 168 119 144 169 120 145 170 121 146 171 122 147 172 123 148 173 « $ 124 WS 149 174 125 150 ^ @ 176 233 Appendix I V List o f the Chu Bronze Vessels The Chu bronze-vessel graphs in the table have been collected from the following bronze vessels and artifacts. The bronze vessels and artifacts are listed in chronological order below, divided in province where they were found. The name o f the vessel or artifact, the place found, the dating, and the source are provided. List o f Abbreviations C W Z T Chu wenwu zhanlan tulu fe jti % % HI G M Q Guo Moruo quanji fp jfc £ H J H K G Jianghan kaogu i L ^ ^ " i " K I K S Kinshutsu In-Shu kinbun shusei $i tB WiM K G Kaogu L J D T K LiangZhou jinwenci daxi tulu kaoshi W M^^L^^^WM^ W W Wenwu W W C Z Wenwu cankao ziliao X % # # it 4^ S J W C Sandaijijin wencun 'feXfe SJT Shierjia jijin tulu - j - -Z. % fif & @ ^ S C C Y Shouxian Caihoumu chutuyiwu fH&H-'f^il ttl iift#7 S S S K Shunju sengoku seidoki no kenkyu ^^C^@#H^-(D9f^S x x c Xichuan Xiasi chunqiumu :M J'l ~F ^-^.S X Y C Xinyang Chumu It" |# fe H Y J J Yin-Zhou jinwen jicheng WiWi ZJT Zhensongtang jijin tu H ^ % if dkr m Z Y Zenghou Yimu f M £ S W Z the Western Zhou period; S A the Spring and Autumn period; W S the Warring States period; E early (first one-third); M middle (second one-third); L late (third one-third); E - M the first half; M - L the second half; pho. photograph o f rubbing 234 I. Bronze Vessels Excavated in Hunan Province (ffl $j) l.MHW W S / M W W 1960.8/9.P.81 W S / M K G 1963.9 P.467 II. Bronze Vessels Excavated in Hubei Province (ffl it) 1. # t e * * f i K M W Z / M W W 1973.5 P.22pho. 3 2. Mt# W Z / M W W 1973.5 P.25pho.8 K M W Z / M W W 1973.5 P.25pho.9 5. f W Z / L , S A / E W W 1972.1. pho.6 5. f 4 ^ ^ S W Z / L , S A / E W W 1972.2.P.53pho.l l 6. t?M + ^&Cli ^ l i / W Z / L S A / E WW1972.2 .P .53pho . l2 7.Sf-S W Z / L , S A / E W W 1972.2.P.53pho.l3 8. £ ^ H ^ F i S A / E J H K G 1984.1 P. 101 9. g f ^ + l t i S A / E K G 1975.4 P.223pho.3-l io. S A / E K G 1989.11 P. 1043 pho.6 S A / E , S A / M W W 1972.3 P.66pho.2 12. S A / E , S A / M W W 1972.3 P.68pho.7, 8 13. S A / E - M W W 1980.1 P.37 pho.6-2, 3; pho.7-2, 3 14. a: ft S A / M W W 1982.10P.17pho.2 QYT .P67.15 TCQ . P 2 0 15. g J B ^ f K M S A / M J H K G 1980.2 pho.l (below) 16. t * * # « K M S A / M J H K G 1980.1 P.77 pho.6 17. f 4 ^ * ^ : « S A / M J H K G 1980.1 P.73pho.3 18. K M S A / M K G 1982.2 P. 145 pho.6-1 K M S A / M K G 1982.2 P. 145 pho.6-3,4 20. f i l l S A / L WW1986 .4P .10pho .2 S A / M , M - L , W S / E - M S J W C 1.63-64; 22.JK.illl, £ K M S A / L J H K G 1983.1 P.75pho.2 23. S A / L W W 1976.11 pho.4-1 24. « S A / M , S A / L , S A / L — W S / E , W S / M J H K G 1992.1 P.73 2 5 . & j g - * £ K M S A / L - W S / E , W S / E J H K G 1985.1 P.59 235 27. f M ^ i ( t ) I M W S / E Z Y P . 1 9 9 p h o . l 0 0 - l , 2 28. Bfc&gfc KM W S / E Z Y P . 2 2 7 p h o . 126-2 29. %&z,s. KM W S / E Z Y P . 2 1 3 p h o . H 2 - 3 30. f ^ ^ ^ M KM W S / E ZYP.561-579 31. KM W S / E Z Y P.580-582 32. KM W S / E Z Y P . 2 8 3 p h o . 174-3 33. f « $ ( N 1 0 5 ) KM W S / E ZYP.276pho .166 34. f M l M ( N 1 3 0 ) KM W S / E ZYP.273pho .163 35. t f M i i - $ KM W S / E ZYP.277pho .167 36. g £ * # # « f t e £ * H IM W S / E Y J J 8 5 3 7 . | p ^ f ^ i l f t W S / E K G 1980.5pho.3-ll 3 8 . ^ ^ ft W S / E K G 1980.5 pho.3-10 39. #39 3 W S / E W W 1962.11P.58.1 III. Bronze Vessels Excavated in Henan Province (53" $j) S A / E W W 1980.1 P.47pho.2 2 . # & £ ® /fi§ l# S A / E W W 1980.1 P.43 pho.3 3 . « M S A / E - M K G 1965.7 P.371 pho.1-2 4. # & 6 # i « S A W W 1980.1 P.43pho.4 5. S A W W 1980.1 P.43pho.5 6. S A W W 1980.1 P.44 pho.6 7. S A / E , S A / M K G 1984.4 P.311pho. 12-4 8. S A / E , S A / M K G 1984.4 P.311 pho. 12-5 9 . £ T # * A t S A / E , S A / M KG1984.4P.319pho.20-3 10. t ^ ^ ^ A i S A / E , S A / M K G 1984.4 P.320pho.21-5 n . - f c ^ f t ^ A f l S S A / E , S A / M K G 1984.4 P.311 pho.21-4 12. S A / M W W 1 9 8 1 . 1 P . 1 3 p h o . l l , 12 13. | * * A K S A / M WW1981 . lP .13pho .14 t r i p S A / M S J W C 18.24.1 15. m f£.fa S A / M - L , S A / L X X C P . 7 1 p h o . 6 0 - 2 16. W f i S A / M - L , S A / L X X C P.57 pho.46-1 17. tfc#mjt $fjn S A / M X X C P . 3 3 p h o . 2 5 - 1 , 2 18. J * * * ® S A / M X X C P.36 pho.29 19. S A / M X X C P.8 pho.5-2 236 20. %mM& 2 1 . i ^ ^ i 22. i ^ H i M fllfjll 23. 25. 26. S A / L S A / L S A / L S A / L S A / L - W S / E S A / L , S A / L - W S / E X X C P.69 pho.58 X X C P . l 17 pho.95, P . l 19pho.97, P. 121 pho.99, P.124pho.l01 X X C P. 143 pho. 118, P.144 pho.119, P. 145 pho. 120, P. 148 pho.122, P.149 pho.123, P.151 pho. 125 W W 1976.11 pho.4-4 W W 1980.1 pho.5-5 S S S K P.68No22 S A / L , W S / E , W S / E , W S / M X Y C pho. VIII , I X IV. Bronze Vessels Excavated in Anhui Province l . &£JtM S A / L W W 1982.5 P.59 2. S A / L W W 1986.2 P.64 3. S A / L S C C Y pho.39, 40 4 . # f c J £ S A / L S C C Y pho.32-2, 3 S A / L S C C Y pho.37 S A / L S C C Y pho.34-5 7 . & M M & S A / L S C C Y pho.34-6 S A / L S C C Y pho.38 9. S A / L S C C Y p h o . 5 0 , 51 S A / L S C C Y p h o . 5 2 , 53, 54, 55, 62, 65, 68, 69 S A / L K G 1963.4 P.205pho.l-1 12. f « * M " P £ W S / M S J W C 12.25.1 13. W S / M K G 1963.8 pho.8 14. g ^ i W S / L W W C Z 1957.7 P.83 15. W S / L W W 1959.4 P . l 16. W S / L W W 1978.5 P.96,pho.3 17. fei^i W S / L S J W C 3.25.1-4 18. fei^ll W S / L SJT Zun-21; S J W C 10.8, 3-5 19. XM%> W S / L W W 1980.8 P.28 237 20. $.5.$%% W S / L C W Z T 2 21. W S / L SJT Zun-24-25 22. W S / L C W Z T 4 W S / L SJT Bao-10 24. I # « W S / L S J W C 3.19.5 25. HF-^ -TII W S / L S J W C 3.19.6 26. i # « W S / L S J W C 3.26.1,2, 3 V . Bronze Vessels excavated in Jiangsu Province (%LM) l . ftgftft S A W W 1989.4 P.53 pho.2 2 . & B * # # * S A W W 1989.4 P.54pho.4 3. ft&^-t&lM S A / L W W 1984.5 P.17pho.5 4. ^ g x l . ^ ^ j S A / M - L W W 1988. 9 P96 pho.3 S A / L K G 1965.3 P.109pho.8, P . l 10 pho.9, P . l 12 pho. 11 6. W S / L K G 1995.1 P.76 pho.3 m% W S / L W W 1980.8 P .30pho . l left W S / L W W 1980.8 P.30 pho. l upper right, P.32 pho.2 right - W S / L W W 1980.8 P.32 pho.2 left V I . Bronze Vessels Excavated in Jiangxi Province (%L W) 1. ^ £ S A / L Z J T v o l . 2 , 13 I f l i f e l S A / L W W 1980.8 P.13 pho.l 3. M f t s S t S A / L W W 1980.8 P. 14pho.2 VII . Bronze Vessels Excavated in Shanxi Province (Jj W) S A M , L K G 1963.5 P.238 pho.9 VIII . Bronze Vessels Excavated in Sichuan Province ( 0 )\\) U P i l M fflP W S / E - M , W S / E W W 1981.6 P.7pho.4-2 I X . Bronze Vessels Excavated in Zhejiang Province (^fxH) 1. ftCflOHpf-f* $ H S A / L W W 1 9 8 4 . 1 P.12pho.4,P13 238 2. $H S A / L 3. K£$M 0 * S A / L pho.5 W W 1984.1 P .16pho. l4 K G 1983.4 P.372pho.3-l X . Bronze Vessels Excavated in Shanxi Province (Rfc W) 1. ^ i ^ H i i Ala S A / L W W 1981.1 pho.6-2 2. a # A ° 7 f f i ^-Jb W S K G 1984.8 P.761 pho.5 X I . Bronze Vessels of Unknown Excavated Location l . W Z / M - L , S A / E S J W C 1.5.2, 1.6.1, 1.6.2, 1.7.1 2. W Z / L L J D T K 177 3. &A£ S A / E S J W C 17.37.1 4. S A / E S J W C 17.10.3 5. £iJM+$Ni S A / E , S A / M L J D T K 179 6. ft^M S A / E , S A / L S J W C 10.15.2,3 7. t^Ht j | S A / E , W S / E S J W C 17.13.1 8. S A / M S J W C 1.9.2, 1.10.1 9.±^&i S A / M S J W C 4.14.1 l O . i ^ ^ ^ S S A / M , S A / L S J W C 18.12.5 11. S A / L S J W C 19.52.2 12. «ililt S A / L , S A - W S S S S K P . 4 8 Gui-25 13. W S / E L J D T K 179-180 14. £jg-j||| W S / E S J W C 3.11.4 15. W S / M J H K G 1992.2 P.68 239 Appendix V List o f the Qin Bronze Vessels The Q i n bronze-vessel graphs in the table have been collected from the following bronze vessels and artifacts. The bronze vessels and artifacts are listed in chronological order below. The name of the vessels and artifacts, the place found ( i f known), the dating and the source are provided. The original site o f many is unknown, but their inscriptions prove their Q in provenance, and on occasion supply their dating. Objects not scientifically excavated but located in museums or originally in private collections are marked with and their location is the earliest one for which a full chain of possession can be determined. List o f Abbreviations K G Kaogu K y W Kaogu yu wenwu K I K S Kinshutsu In-Sha. kinbun shusei $z & Jf£ M ^ tX^kfa L J D T K LiangZhou jinwenci daxi tulu kaoshi W M&XffrJZ&W Q M W X Qin mingke wenzixuan ^-^MX^M. Q T M B J Qin tongqi mingwen biannian jishi ^$$&$t%.$iia%-M.% \ S J W C Sandaijijin wencun jE. ft llf % fc W W Wenwu X% uk. The place where the object was W Z the Western Zhou period; W S the Warring States period; M middle (second one-third); E - M the first half; found is unknown. S A the Spring and Autumn period; E early (first one-third); L late (third one-third); M - L the second half. 1. uk. 3. £ uk. W Z / M - L Q T M B J pho. P.2. 3 W Z / M - L W W 1981.9 P.26 pho.3 S A / E S J W C 19.53.2 240 4. S A / E Q T M B J p h o . P . 7 . 10-1, P.8. 10-2 5. S A / E Q T M B J pho. P.18. 12-4, P.19. 12-5, P.20. 12-6 6. S A / L L J D T K v o l . 5.288 7. uk. S A / L L J D T K v o l . 5.289, 290, 291 8. + uk. W S / M Q M W X P.47 9. W S / M Q T M B J pho. P.32. 17 10 .* t , f t# W S / M K I K S vol.2 P. 154 (reprinted from W W 1979.9 pho.8-1) i i . m^-mnw^ uk. W S / M S J W C 20.26.2; S J W C 20.27.1 W S / M Q T M B J pho. P.40. 27 1 3 . + = uk. W S / M W W 1964.2 P.49 pho.l 14. W S / M W W 1964.9 P.43 pho.2; P.44 pho.3 15. - fa^_r-«F*1£ W S / M W W 1987.8 P.61 pho.3, 4 16. + - # ± I P ^ # ^ W S / M W W 1977.5 pho.3-1, 5; P.35 pho.9; Q T M B J pho. P.49. 35 (handcopy) 17. J H * - h 3 ^ W S / M W W 1977.5 pho.3-3; Q T M B J pho. P.51.37 (handcopy) is. -Ym%-m§&% uk. W S / M Q T M B J pho. P.52. 38 19. uk. W S / M W W 1964.2 P.50 pho.2 20. - t " f a # « M £ W S / M W W 1986.3 P.43 pho.2, 3, 4 2i.-y-s#_hip^^^ W S / M Q T M B J p h o . P . 5 5 . 4 3 W S / M W W 1980.9 P.94 pho.2 2 3 . ^ ^ ^ - H W S / M W W 1980.9 P.94 pho.3 24. W S / L W W 1975.6 P.75pho.l3-2 241 25. 3 + 7 N # ^ W S / L W W 1986.4 P.22pho.4 26. uk. W S / L S J W C 14.11.3 27. 0 + #Ji iP^^ W S / L W W 1992.8 P757 28. fcXftfc* W S / L Q T M B J pho. P.63. 53 29. ^ # ± 1 1 5 ^ ^ ^ W S / L W W 1982.11 P.75; Q T M B J pho. P.65. 55 (handcopy) 30. - # # X # W S / L K y W 1983.6 P.5 pho.5; Q T M B J pho. P.61. 52 31. # X ^ W S / L W W 1989.6 P.73 pho.2 3 2 . S # - * ^ g ^ * ^ ; - uk. W S / L S J W C 20.28.2, 20.29.1 3 3 . £ 4 M B # g : F * £ - uk. W S / L Q T M B J pho. P.77. 68-1 34. S ^ ^ ^ g ^ t ^ = uk. W S / L Q T M B J pho. P.79. 69 35. X*B?ft% IS;II#;II W S / L K G 1991.1 P17pho.3 ,4 36. + 0 ^ M # P ^ W S / L Q T M B J pho. P.83. 74 3 7 . + - b ^ # X ^ [ ws/L W W 1982.3 P.12 38. ff#P uk. W S / L L J D T K v o l . 5.292 39. mmm uk. Beginning of Qin dynasty Q M W X P.48 uk. Beginning o f Qin dynasty Q M W X P . 4 1 242