UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The first known Chinese calendar : a reconstruction by the synchronic evidential approach Liu, Xueshun 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-105291.pdf [ 14.25MB ]
Metadata
JSON: 831-1.0099806.json
JSON-LD: 831-1.0099806-ld.json
RDF/XML (Pretty): 831-1.0099806-rdf.xml
RDF/JSON: 831-1.0099806-rdf.json
Turtle: 831-1.0099806-turtle.txt
N-Triples: 831-1.0099806-rdf-ntriples.txt
Original Record: 831-1.0099806-source.json
Full Text
831-1.0099806-fulltext.txt
Citation
831-1.0099806.ris

Full Text

THE FIRST K N O W N CHINESE C A L E N D A R : A RECONSTRUCTION B Y THE SYNCHRONIC EVIDENTIAL A P P R O A C H by Xueshun Liu M.A . , The University of Zhengzhou, 1993 A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE F A C U L T Y OF G R A D U A T E STUDIES (Asian Studies) THE UNIVERSITY OF BRITISH C O L U M B I A August 2005 © Xueshun Liu, 2005 ABSTRACT The first known Chinese calendar refers to the calendar embodied in the Yin oracle-bone inscriptions (OBI). Since 1925, the two-layered evidential method has been the standard approach to interpreting them. This method combines primary inscriptional evidence with secondary materials, resulting in inconclusiveness and inaccuracy in previous studies. By adopting the synchronic evidential approach, the present dissertation aims at accurately reconstructing the system by which the Yin divided time into fixed periods. Chapter 1 deals with background issues: it justifies the assertion that the Yin calendar was the first known Chinese calendar, presents inscriptional evidence indicating the existence of a prescriptive Yin calendar, proposes absolute dates for this calendar and justifies the adoption of the synchronic evidential approach. Chapter 2 focuses on time divisions in the Yin day. The two criteria for determining a time division in the OBI are defined as: 1) a word's usage as a time division in early Chinese texts, and 2) suitableness of this usage in inscriptional contexts. The order of the twelve time divisions shows that su WK is the first division of the Yin day. Su is thus the start of the Yin day. Chapter 3 analyzes the lunation in the Yin calendar. Inscriptional evidence confirms that the Yin month is either 3 0 or 2 9 days long. There is no proof of a long Yin month of 31 days or longer, or for a short one of 25 days. Long and short Yin months occur alternately. The Yin employed both year-end intercalation and in-year intercalation. By late periods, in-year intercalation replaced year-end intercalation. Chapter 4 addresses issues concerning the Yin year. A normal Yin year consists of 12 months, a leap year 13 months. The designation for the Yin year is TE. Reconstructions show the commencement of the Yin year is the second month before the month containing the winter solstice. Chapter 5 takes issue with a problematic attitude in the field. It is inappropriate to deny conclusions drawn from inscriptions. Rather, a researcher should give priority to inscriptional evidence over all other secondary materials. It is time to replace the two-layered evidential method with the synchronic evidential approach. ii T A B L E OF CONTENTS Abstract ii Table of contents i i i List of tables v Acknowledgements vi CHAPTER ONE INTRODUCTION 1 1.1 The Yin Calendar as the First Known Chinese Calendar 1 1.2 The Existence of a Prescriptive Yin Calendar 4 1.3 The Absolute Dates of the Yin Calendar 12 1.4 The Synchronic Evidential Approach 14 1.5 Aims of this Study 20 CHAPTER TWO "THE D A Y " IN THE YIN C A L E N D A R 23 2.1 Introduction 23 2.2 How to Determine Words of Time Divisions 23 2.2.1 Lexical Meaning 24 2.2.2 Inscriptional Context 25 2.2.3 Lexical Meaning and Inscriptional Context 27 2.3 Time Divisions of the Yin Day 29 2.3.1 Ri 29 2.3.1.1 Sub-divisions ofRi in Period III 30 2.3.1.2 Sub-divisions of Ri in Period 1 58 2.3.1.3 Sub-divisions of Ri in Period II 67 2.3.1.4 Sub-divisions of Ri in Period IV 70 2.3.1.5 Sub-divisions of Ri in Period V 71 2.3.2 Xi 73 2.3.2.1 Zhuo: a Time Division? 80 2.3.3 Order of Time Divisions of the Yin Day 82 2.3.3.1 Ri before Xi 82 2.3.3.2 Order of Time Divisions of Ri 84 2.3.4 Characteristics of Yin Time Divisions 88 2.4 The Start of the Yin Day 90 2.4.1 Evaluation of Previous Theories 91 2.4.1.1 The Midnight Theory 91 2.4.1.2 The Cockcrow Theory 93 2.4.1.3 The Dawn Theory 94 2.4.2 Su: the Start of the Yin Day 94 2.4.2.1 From Dacai to Xi: No Start of the Yin Day 94 2.4.2.2 Su before Dan 95 2.4.2.3 Su: the Beginning of the Yin Day 98 CHAPTER THREE "THE M O N T H " IN THE YIN C A L E N D A R 99 3.1 Introduction 99 3.2 The Number of the Month in a Yin Year 99 3.3 The Length of the Y i n Month 104 3.3.1 Yin Month Not Always 30 Days Long 104 3.3.2 No Yin Month of 40 or 50 Days 109 iii 3.3.3 No Yin Month of 31 Days '. 109 3.3.4 No Yin Month as Short as 25 Days 122 3.3.5 Yin Month is 29 Days or 30 Days Long 125 3.3.6 No Reference to Dayue or Xiaoyue 142 3.4 The Commencement of the Yin Month 145 3.5 The Arrangement of Y i n Months 149 3.5.1 Consecutive Long Yin Months 150 3.5.2 Consecutive Short Y in Months 151 3.5.3 Intercalation in the Yin Calendar 153 3.5.3.1 The Existence of Intercalation of the Yin Calendar 153 3.5.3.2 Year-end Intercalation 154 3.5.3.3 In-year Intercalation 154 3.5.3.4 Conclusion with Regard to In-year Intercalation 178 CHAPTER FOUR "THE Y E A R " IN THE Y I N C A L E N D A R 180 4.1 Introduction 180 4.2 The Designation for the Yin Year 180 4.2.1 Criteria 180 4.2.2 Si 181 4.2.3 Sui 190 4.2.4 Nian 193 4.2.5 Zai .' 197 4.3 The Commencement of the Y i n Year 199 4.3.1 Evaluation of Previous Studies 200 4.3.1.1 Evaluation of Inscriptions about Huo 202 4.3.1.2 Inscriptions about Lunar Eclipses 207 4.3.1.3 Inscriptions about Weather 214 4.3.1.4 Inscriptions about Agriculture 221 4.3.2 Reconstruction of the Start of the Yin Year 226 4.3.2.1 Absolute Dates for Lunar Eclipses in the OBI 226 4.3.2.2 Reconstruction of Calendars for Two Lunar Eclipses 232 4.3.3.3 Conclusion 238 CHAPTER FIVE C O N C L U D I N G R E M A R K S 240 Bibliography 253 iv LIST OF TABLES Table 1. The Order of Time Divisions of the Yin Day 89 2. Reconstruction of the Calendar for Heji 22404 116 3. Reconstruction of the Calendar for Heji 11546 127 4. Reconstruction of the Calendar for Heji 11485 131 5. Reconstruction of the Calendar for Heji 339 140 6. New Moons in 1228 BC and 1227 BC 148 7. Intercalary Months in the Yin OBI 174 8. Reconstruction of the Calendar and Five-Ritual Cycle for Three Sis 187 9. Opinions about the Start of the Yin Year 201 10. Winter of the Yin Year 217 11. Reconstruction of the Calendar for 1228 BC - 1227 BC 234 12. Reconstruction of the Calendar for 1227 BC 235 13a. Reconstruction of the Calendar for 1167 BC - 1166 BC 236 13b. Reconstruction of the Calendar for 1167 BC - 1166 BC 237 14a. Reconstruction of the Calendar for 1166 BC 237 14b. Reconstruction of the Calendar for 1166 BC 238 15. In-Year Intercalation in Late Spring and Autumn Period 247 V ACKNOWLEDGEMENTS I have finally managed to make this dissertation ready for defence after numerous revisions that include expanding both scope and depth, as well as adding major and fine details. The result of these revisions, which I consider as improvement over my previous versions, could not have been made without the help of many individuals. To each of them, I want to express my deepest gratitude. Allow me now to credit them. I naturally want to express my gratitude first and foremost to my academic advisor, Professor Ken-ichi Takashima, whose unsurpassed tutelage throughout my graduate work has not only taught me the knowledge of Chinese palaeography, but also trained me in how to conduct research by the distinctive and productive synchronic evidential approach. Moreover, his insightful and detailed criticisms, suggestions and corrections on numerous passages, all of which were given during his careful and tireless reading of drafts of this dissertation, have definitely made the content of this work more accurate and the flow of the main points more logical. To him, I express my sincere and everlasting gratitude. I also want to express my gratitude to my committee member Professor Edwin G. Pulleyblank. Even though he retired many years ago, he has been conducting graduate seminars at the University of British Columbia on a regular basis, in response to the interest of Professor Takashima who believes that knowledge of Chinese historical phonology and linguistics is essential in doing research on ancient Chinese civilization. Professor Takashima's graduate students also wanted to get first-hand tutelage from such an eminent scholar as Professor Pulleyblank. I consider myself lucky to have been one of those students who have taken a comprehensive examination for a PhD in the field of historical Chinese phonology from Professor Pulleyblank. As is well known, Professor Pulleyblank is a rare linguist who also does active work in the field of Chinese history. Although the period of his interest is later than the Yin Dynasty, Professor Pulleyblank has given me numerous valuable comments on the draft of this dissertation. I knew they came from the eyes of a keen historian. I am much indebted to him for all the help he has given me. The original members of my PhD committee included Professor David N . Keightley of the University of California at Berkeley, who has read draft versions of this dissertation and given me many valuable observations and comments. We have had very meaningful V I dialogues concerning questions on the Yin calendar. By the fall of 2004, however, his health unfortunately required his withdrawal, so that the reconstitution of my PhD committee became necessary. It was then decided that Professor David W. Pankenier, renowned authority in Chinese archaeoastronomy, be asked to take the place of Professor Keightley. He kindly agreed to serve as a committee member on a short notice, and amidst his heavy teaching and research schedules Professor Pankenier has read the entire draft dissertation and given me detailed comments and suggestions. The reader will find in the pages that follow my indebtedness to both Professor Keightley and Pankenier. Here I wish to record my profound gratitude to both scholars. I also wish to express my thanks to the following individuals: Dr. Liu Ciyuan for recommending me the Skymap software for calculating ancient Chinese astronomical phenomena; Professor Chang Yuzhi for sending me her new monograph about the Yin calendar; Dr. Shen Pei and X u Yihua for scanning new articles for me; Ms. Liu Jing and Mr. Wang Qingxiang, librarians in the Asian Library, UBC, for their enthusiastic help in locating rare publications; and Mr. David Dressier for his rich experience, serious attitude and commitment to the copyediting and proofreading of this dissertation. vii C H A P T E R O N E INTRODUCTION 1.1 The Yin Calendar as the First Known Chinese Calendar The word "calendar" has several meanings in modern English. Two of those meanings are particularly relevant to the subject of this study. The first one is "a chart showing the days, weeks, and months of a particular year." The second one is "a system by which time is divided into fixed periods, and of marking the beginning and end of a year" (Hornby 2002: 192). Since this dissertation aims to reconstruct the first known Chinese calendar, it certainly wi l l contain some charts that show days and months of some years of this calendar. However, its main purpose is to expound the system of this calendar that divided time into fixed periods. The traditional view is that the Chinese invented their calendar independently. According to extant early Chinese texts, Rong Cheng a subject of the legendary ruler Huang D i M'rfc, was the inventor of the first Chinese calendar.1 Huang D i ' s reign has been dated to a period in the 27 t h century B C . 2 It is also recorded in early Chinese texts that one of ' For instance, it is mentioned in the chapter "Wu gong ^J#j" of the Lushi chunqiu @ ft#$C that Rong Cheng made a calendar (^ fiEifr-jKi). It is recorded in the chapter " L i shulft'4$" of the Shiji $LML that Huang Di possibly observed and studied stars and established the calendar ('M.M^^'&S.IJJ). The present writer's translation of this sentence is based upon the usage of words kao 3% and ding 3? in classical Chinese. According to the Ciyuan WM, the word kao # may mean kaocha%^, 'to observe and study,' and the word ding 3! can mean zhiding%\\'$L, 'to make.' This writer's understanding of the expression kao ding xing li ^f'aiMDfj is that it means "to observe [the patterns of the motion of] stars and to make a calendar." That Huang Di and Rong Cheng both made calendars is not a contradiction because achievements by subjects could be attributed to their ruler in China. In any event, it is the traditional view that the Chinese calendar was invented during the reign of the legendary Huang Di. 2 According to the Shiji, Huang Di was the first Chinese ruler. Sima Qian WJ^/i3; did not give dates of the reign of Huang Di because there were no reliable sources about Chinese chronology before the Gonghe Period. After Sima Qian, many scholars have made efforts to reconstruct the Chinese chronology before that period. Unfortunately, none of them has produced an acceptable chronology for Huang Di. To date, there are no archaeological findings that can be regarded with much certainty as those of Huang Di. At present, it is not possible to determine the absolute dates of his reign, even if, indeed, he was an early ruler in China. A l l the alleged dates of the reign of Huang Di are therefore unreliable. One such example is the date 2697 B C given by J.A.G. Roberts (2003: xxi). 1 Chinese rulers' most important functions was to compile and issue calendars yearly. If these textual records concerning the invention of the Chinese calendar are accurate, then the Chinese calendar would have a history of about 4,700 years, and the one created by Rong Cheng should definitely be regarded as the first Chinese calendar. However, it must be acknowledged that the records mentioned above were compiled thousands years after the legendary reign of Huang Di. In the earliest extant contemporary writing in China, i.e., the oracle-bone inscriptions (abbreviated as OBI in this study) of the Yin IS dynasty (ca 1300 BC - ca 1046 BC), there are no references to Huang Di . In addition, the existence of Huang Di is not supported by archaeological findings; there is not a single archaeological site that is generally accepted as being related to Huang Di . These three facts certainly cast doubt upon the existence of both Huang Di and Rong Cheng. Accordingly, it is extremely difficult to accept the above records at face value. Needless to say, there is no reliable evidence whatsoever that supports the historicity of the invention of a calendar by Rong Cheng or by Huang Di . Therefore, the present writer does not consider the mythical calendar of Huang Di to be the first known Chinese calendar. It is recorded in the chapter "Wudi benji jEttf ^ £ E " of the Shiji $1 i E that the rulers who came after Huang Di in ancient China were Zhuan Xu M, Di K u f ^ # , Di Yao tfrfTti and Di Shun rfrff. At present," there are no contemporary materials, written or archaeological, that support the existence of these four legendary rulers of ancient China. Because of the uncertainty surrounding their existence in history, none of those alleged calendars of those 3 The pivotal role of calendars conceptualized by early Chinese can be clearly demonstrated by various early Chinese texts. For example, it is recorded in the chapter "Li shu" of the Shiji that, in early China, if the surname of a dynasty changed, the beginning of the calendar year must be changed accordingly. The followings are examples given by Sima Qian: the Xia calendar year began with the second month after the one that contains the winter solstice; after the Xia, the Shang changed the commencement of the year to the month immediately after the month of the winter solstice. After the overthrow of the Shang, the Zhou changed the start of the year to the month of the winter solstice (IE^l7&Slfr...&IE^...KlE(UiE.rj, SlEiiH"—MUEliH—M). This kind of record shows the importance attached to the compilation of calendars in early China. Further, there is the following passage in the chapter "Yao dian of the Shangshu ftii+i: "Thereupon Yao commanded Xi and He [the present writer has changed James Legge's romanization of Chinese characters to pinyin i f ] , in reverent accordance with their observation of the wide heavens, to calculate and delineate the movements and appearances of the sun, the moon, the stars, and the zodiacal spaces; and so to deliver respectfully the seasons to the people" (Legge 1872: 18). If Xi and He had to deliver seasons rather than years by observing the sun, the moon, the stars, and the zodiacal spaces, it is reasonable to assume it was done year by year. 2 rulers could truly be regarded as the first known Chinese calendar. After Di Shun, according to early Chinese texts, the Xia X. Dynasty (ca 2070 BC- ca 1600 BC) was founded. Also, in texts such as the Shijing i#^c, Zuozhuan Sift, Shiji and Honshu $£^r$, there are references to a Xia calendar. However, in the chapter "Luli zhi W JfiM" of the Songshu 5i5^, Zu Chongzhi TIffe (429-500), an astronomer of the Song 5i5 Dynasty (453-479 A.D.), pointed out that this so-called Xia calendar was actually created between the end of the Zhou Dynasty (ca 1046 BC ~ 256 BC) and the beginning of the Han Dynasty (206BC - 220 AD). Zu Chongzhi's view is still accepted by Chang Yuzhi (1998: 3) and Chen Meidong WMM (2003: 91). It thus appears that this so-called Xia calendar might in fact not be the calendar of the Xia Dynasty. In addition, there are no contemporary written records of the Xia Dynasty available to modern scholars. Sarah Allan (1984) has expressed doubts with regard to the existence of this dynasty. Because of these problems concerning the existence of the Xia Dynasty and the credibility of the so-called Xia calendar, there is no solid factual basis to consider the Xia calendar mentioned in early Chinese texts as the first known Chinese calendar. It is recorded in many early Chinese texts that the Xia Dynasty was followed by the Shang jgf Dynasty (ca 1600 - ca 1300 BC) and the Yin WL Dynasty (ca 1300 BC - ca 1046 BC). The existence of the Yin Dynasty has been proven by the discovery of the OBI at Anyang in 1899. Hu Houxuan WiWM. (1984) estimates that 154,604 pieces of Yin oracle bones have been unearthed.4 The followings are the two most important OBI collections: Jiaguwen heji ^ if^C^aM (abbreviated as Heji in this study), which published 41,956 pieces of the Y i n oracle bones, and Xiaotun nandi jiagu / J N 4 i l % J i i ¥ # (abbreviated as Tunnan below), which published 4,162 pieces of the Yin oracle bones. The OBI found at Anyang are the earliest contemporary written records in China. Among them are a large number of inscriptions that are clearly related to a calendar. These inscriptions have attracted the attention of many scholars since the earliest stage of the study of the OBI, and progress in the study of the Yin calendar has been continually made 4 Dong Zuobin i f f r ^ (1955: 184-185) and Chen Mengjia W&U (1956: 47-48) estimate that about 100,000 pieces of the Yin OBI have been found. Since Hu Houxuan (1984) has listed a specific number of oracle bones for each collection, his conclusion seems more credible than that of Dong Zuobin and of Chen Mengjia. 3 throughout the 20 t n century. To date, more than 400 articles and books relating to various aspects of the calendar seen in the Yin OBI have been published.5 Generally speaking, contemporary inscriptional evidence points to the existence of a Y i n calendar in the Yin OBI. The foregoing analysis leads to the following conclusion: up to now, the earliest contemporary written records are the Yin OBI, and the calendar attested to in the Yin OBI is the earliest Chinese calendar that can be reconstructed with contemporary records. Therefore, the Yin calendar may be considered to be the first known Chinese calendar. 1.2 The Existence of a Prescriptive Yin Calendar The previous section asserts that the calendar of the Yin Dynasty is the first known Chinese calendar. In order to reconstruct this calendar, the lost Y in calendar has to be recovered. Of course, this task rests on the assumption that there indeed existed a prescriptive calendar in the Yin Dynasty that guided the life of the Yin people throughout the whole year. If the existence of such a Yin calendar were in serious doubt, it would render this study meaningless. Therefore, it is absolutely vital that this issue be addressed. It appears that the ability of the Yin people to compile a prescriptive calendar has been accepted by most, specialists in the field, including eminent scholars such as Dong Zuobin If (1895-1963), Chen M e n g j i a l ^ ^ (1911-1966), and Chang Yuzhi However, there are a few scholars who doubt the existence of a prescriptive Yin calendar. For example, it is Shinjo Shinzo's ffiijftMxW, opinion (1936: 86) that the Chinese started compiling calendar in 600 BC. In other words, according to Shinjo Shinzo, before 600 BC, the Chinese people, including the Yin, were unable to compile a calendar. His view does not enjoy wide acceptance among scholars today, because his work, which was published at the very early stage of the study of the Yin OBI, was not based on comprehensive inscriptional evidence. The progress in the study of OBI that has been made subsequent to Shinjo Shinzo's work clearly demonstrates that his view contradicts the inscriptional evidence that will be introduced in the present study. 5 Song Zhenhao (1999: 439-463), the most recent bibliography of the study of the calendar of the Y in Dynasty, lists 398 relevant publications. 4 Yabuuchi Kiyoshi i £ r * ] ,#(1956: 72-74) holds the view that the commencement of each month of the Yin calendar was based on the actual observation of the new crescent moon. If this view is correct, it follows that the Yin people would have known the end of a month and the start of the next month only when they could actually observe the new crescent moon. In other words, the Yin did not have a calendar that had prescribed dates for each month. Without knowing dates of future months, it would have been absolutely impossible for the Yin to compile a prescriptive calendar for a whole year. This is a significant implication of Yabuuchi Kiyoshi's view. David N . Keightley's opinion (2000: 6, note 18) about the Yin calendar is similar to that of Yabuuchi Kiyoshi. He states that "I believe that the Shang did not use a prescriptive calendrical month of fixed length but employed an observational calendar, simply noting the number of each moon as it waxed and waned...." If it is a fact that the Yin simply noted the number of each month as it waxed and waned, it is reasonable for him to assume there was no prescriptive calendar in the Yin Dynasty. In the field of Yin calendar studies, scholars who hold the view that the Yin were unable to compile a prescriptive calendar spanning a whole year have been less influential than those who propose that the Yin were able to accomplish this feat. One reason for this is that scholars who hold the former view have yet to publish any detailed and systematic studies on the Yin calendar, whereas scholars who support the existence of a Yin calendar have already published two lengthy monographs. However, this does not constitute sufficient grounds for dismissing the view that the Yin could not compile a prescriptive calendar. The best and most practical way to judge the validity of these two different theories is to consult the inscriptional evidence. In the Yin OBI, there are a number of pieces of evidence that indicate, in one way or another, the existence of a prescriptive Yin calendar. First, David Pankenier has written the following comments to the draft of this dissertation on April 5, 2005: The fact that the count of moons could never exceed 13 proves beyond a doubt that they already had a concept of a year, otherwise there would be nothing to prevent their using the tiangan to enumerate 10 moons in the civil calendar, or even 60 moons, using the whole series of sexagenary designations. If they had a concept of a 5 year, almost certainly based on the cycle of seasons, and they made adjustments to the lunar count to synchronize with that tropical year, that means the calendar had a prescribed length. Also, in his comments to the draft of this dissertation, Edwin G. Pulleyblank points out that the Yin "obviously expected there to be 12 or 13 moons in a year." Indeed, it is a fact that a year in the Yin OBI generally consisted of 12 or 13 months, as will be demonstrated in Chapter Four. This fact suggests that the length of a Yin civil year was fixed at 12 or 13 months in length. This is the first evidence indicating a prescriptive calendar in the Yin OBI. Second, the Yin calendar had prescriptive months, as shown by the following six inscriptions on Heji 11485: [oi]H£h ^Jt:(=^)?gc, -Bo H0P h *0t:(=^)*8. f £ * K ^ ( = ^ 0 H B ^ M ^ j ! W I T , t 6 o / m . W h = ^[T_r=^C]?S0 « f ^ H » 11485 Crack-making on guihai (day 60), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the first month. Crack-making on guiwei (day 20), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the second month. On guimao (day 40), [Zheng] divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the second month. 6 The bone graph mm is often transcribed as wen H , 'to hear,' by other specialists. Ken-ichi Takashima (2004: 7) does not accept such an interpretation. He interprets "that the original graph 1^, structurally similar to the 'ancient form' (guwen rS"3t) given by the SW (12a) being complete with the phonophorica hun fr, written as hun H i stands for the word hun ff/| 'dark.' The Shang time keepers and others were watching this lunar eclipse at midnight on July 12, 1201 B.C., and the moonlit night turned dark momentarily." Since it is likely that the night turns dark momentarily when a lunar eclipse occurs, Ken-ichi Takashima's transcription is adopted here. 6 [Crack-making on gui]mao (day 40), [Zheng] divined: "[In the next 10-day week], there will be no [disasters." Day guimao was in] the fifth month. Crack-making on guiwei (day 20), Zheng divined: "[In the next] 10-day week, there will be no disasters." At the night of the third day yiyou (day 22), the moon was eclipsed and it became dark. [Day yiyou was in] the eighth month. On [gui]mao (day 40), divined: "[In the next] 10-day week, there will be [no] disasters." /fey/ 11485 In these six inscriptions, there are seven ganzhi dates and four month notations. According to Dong Zuobin (1952: 287-289), there is only one reconstruction that can accommodate all these dates and month notations. Below is his reconstruction. Moon First day Last day 1 s t moon yisi (day 42) guiyou (day 10) 2 n d moon jiaxu (day 11) guimao (day 40) 3 moon jiachen (day 41) renshen (day 9) 4 t h moon guiyou (day 10) renyin (day 39) 5 t h moon guimao (day 40) xinwei (day 8) 6 t h moon renshen (day 9) xinchou (day 38) 7 t h moon renyin (day 39) gengwu (day 7) 8 t h moon xinwei (day 8) The reconstruction above shows that the first, third, fifth and seventh months are 29 days long, and that the second, fourth and sixth months are 30 days long, whose durations are the same as months in later prescriptive Chinese calendars. The prescriptive lengths of these months are inscriptional evidence for the view that the Yin calendar had prescriptive months, indicating that the Yin calendar could be prescribed. Third, it can be inferred from some inscriptions that the Yin people were able to determine dates of future months in advance. This is another indicator of the existence of a prescriptive Yin calendar. As cited above, both Yabuuchi Kiyoshi and David N . Keightley believe that the Yin started their month with actual observation of the new crescent moon. According to their 7 theory, the Yin people must not have been able to determine a date of a future month or year in advance. The validity of their theory, therefore, can be judged by investigating whether or not the Yin actually knew dates of future months in advance. If the Yin people did know dates of future months, this fact would directly contradict Yabuuchi Kiyoshi's and David N . Keightley's view that the Yin did not employ a prescriptive calendar, making their view untenable. After examining all rubbings in the Heji, the present writer has found inscriptions that suggest that the Yin could know dates of future months. Below are two examples: [02] ^ B S , WM&^&o - h - j f . * B h T-n^mwM&o h o i^m 21221 Crack-making on xinchou (day 387), Xun [divined:] "On xinhai (day 48), [the king will] perform the T/ow-sacrifice and offer /iawg-millet." [Xun divined in] the twelfth month. On xinchou (day 38), [Xun] divined: "On xinyou (day 58) of the first month, [the king will] perform the j/ow-sacrifice and offer //awg-millet." In the twelfth month [Xun] divined. Heji 21221 Lo3],MPiK &mmft: ^ - M ^ ^ O 1^0= ^ £ A ^ 0 i^M)) 41756 On yimao (day 52), the king made cracks, and [he] was in the Geng encampment and divined: "I will attack X [an unknown character]; it should be in the tenth month, on wushen (day 45), that I destroy them." The king read the cracks and declared: "Auspicious." [Day yimao was] in the eighth month. Heji 41756 Example 2 is made up of two inscriptions from Period I 9. In these two inscriptions, there 7 The number after a certain ganzhi T'^c date indicates its position in the unending sixty-day ganzhi cycle. 8 Because of the gap of three thousand years between the OBI and the modern Chinese writing system, it is difficult to transcribe some bone graphs into simplified modern Chinese characters, even though there are detailed studies on those bone graphs. Whenever this occurs, this writer has simply copied the graph into his transcription. 9 There are two major theories with regard to the periodization of the Y in OBI. Dong Zuobin (1933) divides the Yin OBI into five periods. His theory has been modified later, based on progress in the study of the OBI and archaeological findings. For an English introduction of Dong Zuobin's theory, see David N. Keightley (1985: 92-94). The second theory has been systematically explained in L i Xueqin and Peng Yushang (1996). The major difference between these two theories is methodological. As for dating the vast majority of the Yin OBI, their results are almost the same except for the so-called L i -group inscriptions. L i Xueqin's conclusion about the Li-group inscriptions is not fully supported by archaeological data, 8 are three dates, xinchou, xinhai, and xinyou, and two month notations, the twelfth month and the first month. These inscriptions also show that the day xinchou was in the twelfth month and the day xinyou was in the first month. From xinchou (day 38) to xinyou (day 58), there were 21 days, which is apparently not enough for an intercalary month. Therefore, the first month immediately followed the twelfth month in this example. As will be demonstrated in Chapter Three, an intercalary month in Period I could be put at the end of a year, and such an intercalary month is referred to as 'the thirteenth month' in Period I inscriptions.10 Had the Yin not had a prescriptive calendar, as Yabuuchi Kiyoshi and David N . Keightley suggest, the diviner Xun would not have been certain as to whether there would be an intercalary month, 'the thirteenth month,' after the twelfth month when the divination of Example 2 was made. Example 2 proves diviner Xun actually knew there would not be an intercalary month after that twelfth month. Diviner Xun indeed knew day xinyou belonged to the first month. Thus, Example 2 is one piece of evidence that demonstrates that the Yin could know the date of a future month. Example 3 is an oracle bone from Period V . This inscription records two dates, yimao and wushen, and two month notations, the eighth month and the tenth month. Yimao was in the eighth month and wushen in the tenth month. The interval between yimao (day 52) and wushen (day 45) was fifty-three days, which is shorter than the length of two lunar months. Because of this, only one month can be inserted between yimao of the eighth month and wushen of the tenth month. This excludes the possibility of an intercalary month between the eighth and tenth month mentioned in this example. Chapter Three will present inscriptional evidence to show that the in-year intercalation was adopted during Period V . Since the two dates and two month notations in the third example exclude the possibility of an intercalary month between the eighth month and tenth month of that year, the diviner, who happened to be the king, was certain of the future date wushen. Heji 41756 is another piece of solid evidence that shows that the Yin knew the date of a future month. which the present writer analyzed in detail in a manuscript entitled "Can the Date of the L i Diviner Group Inscriptions be Settled?" In the present dissertation, therefore, this writer follows a refined Dong Zuobin' periodization of the Yin OBI, which is employed by the Heji. 1 0 According to Chang Yuzhi (1998: 302), the phrase "the thirteenth moon" occurs 121 times in the Bin-group inscriptions, which are one group of inscriptions from Period I. 9 Besides those two examples above, some time phrases in charges" indicate that the Yin knew when the current month would end and when the next month would start. The following are some examples of charges with different time phrases: [04 ]§ |EK f f - j S : 3E£H: mm. %E h - f r j g . l&M} 12487 Crack-making on guisi (day 30), Zheng divined: "It will rain in this first month." The king read the cracks and declared: "it will rain on a bing-day." Crack-making on guisi (day 30), Zheng divined: "It may not happen to rain in this first month." Heji 12487 In the two charges in this example, the time phrase is jin yiyue ^ ~B, 'this first month.' Generally speaking, divination was a means for early Chinese to resolve their doubts about a proposed plan of an action. It thus can be inferred that the time for an event in a charge should have been clear to the diviner. In other words, a time phrase in a charge refers to a definite time period. Since the phrase 'this first month' referred to a definite period, and since that period was clear to the diviner, it is reasonable to suggest that diviner Zheng and other Yin people knew when that first month would end and when the second month would begin. Besides jin yiyue, there are similar time phrases such as jin eryue -fyZLB, 'this second month,' jin sanyue ^~ELB, 'this third month,' jin siyue ^T~UMB, 'this fourth month,' jin wuyue -^JiB, 'this fifth month,' jin liuyue ^/\B, 'this sixth month,' jin qiyue ^"hB, 'this seventh month,' jin bayue 4*AB, 'this eighth month,' jin jiuyue -fytlB, 'this ninth month,' jin shiyue 4"f"H, 'this tenth month,' jin shiyiyue —B, 'this eleventh month,' jin shieryue -^"h—B, 'this twelfth month,' and jin shisanyue •^7~~JrELB 'this thirteenth 12 month' in other inscriptions . Like jin yiyue, 'this first month,' these time phrases are evidence showing that the Yin knew the end of the current month and the start of future months beforehand. 1 1 An inscription may consist of a preface, charge, crack number, crack notation, prognostication, verification and post-face. The term "charge" refers to the topic of the inscription. For more details, readers are referred to David N. Keightley (1978: 28-44). 1 2 A l l these examples are transcribed by Shima Kunio (1971: 260.2-4). 10 Examples 2-4 strongly indicate that the Yin people knew dates of future month in advance. In order for them to be able to do so, the Yin must have had prescriptive calendar tables to assist them. I believe that Heji 24440 can be considered a piece of direct evidence for the existence of that kind of calendar table. Those characters inscribed on this piece of oracle-bone are as follows. [05] R-JEBitM* ¥ ^ Z , f i , PHC> TFP> r£jg, B E , m*F> 3EEJ3> H WJ&* T S > E^P, mm. * E > Z*F, mm> ¥ ZJIh T ^ J f t T , B H > MW> 3 * 5 1 , f E . ¥ ^ F , z , * , p i ^ T I , B J ^ I^T> $ w , H £ P , f i , ZJEN m T * . E M , i f > ^ H > ¥IC> p l i l , T E , r j ^ , E * . ^Eja , H l&m 24440 Month/ one/ right/ called/ eating wheat: jiazi, yichou, bingyin, dingmao, wuchen, jisi, gengwu, xinwei, renshen, guiyou, jiaxu, yihai, bingzi, dingchou, wuyin, jimao, gengchen, xinsi, renwu, guifwei), jiashen, yiyou, bingxu, dinghai, wuzi, jichou, gengyin, xinmao, renchen, guisi. The second month/ father xuan: jiawu, yiwei, bingshen, dingyou, wuxu, jihai, gengzi, xinchou, renyin, guimao, jiachen, yisi, bingwu, dingwei, wushen, jiyou, gengxu, xinhai, renzi, guichou, jiayin, yimao, bingchen, dingsi, wuwu, jiwei, gengshen, xinyou, renxu, gui. Heji 24440 Heji 24440 is a piece of oracle bone from Period II. In 1933, Guo Moruo f l S ^ ^ f pointed out that this was a copy of an early calendar table.13 Since then, many scholars have accepted his conclusion. The most recent example is Chen MeidongP^H^ (2003: 22). Indeed, the content of Heji 24440 makes it easy for people to consider it to be a copy of a calendar table that recorded dates of the first two months of a certain year. It appears reasonable to assume that the Yin people used such tables to assist themselves in calculating future dates. More importantly, the existence of such a table strongly suggests the existence of a prescriptive calendar at that time. In this section, the writer has presented evidence in the Yin OBI that indicates that the 1 3 Guo Moruo made this comment in his annotation about Tongzuan MM 6, which was first published by Bunkyudo in Tokyo in 1933. It was reprinted by Kexue Chubanshe in Beijing in 1982 and is the version used in the present context. For his opinion on this particular oracle bone, see Guo Moruo (1982: 216-217). 11 length of a Yin civil year and that of a Yin month were prescribed, that the Yin people could know dates of future months with the assistance of a calendar table, and that there were calendar tables. This view is contrary to that which says the Yin did not use a prescriptive calendar. The conclusion drawn here is that the Yin people did compile a prescriptive calendar. As a matter of fact, Chen Zungui $kM$} (1980: 203-204), a specialist in the field of early Chinese calendar, points out that, with the knowledge of the length of both a civil year and a month, it was not at all difficult for early Chinese to compile their calendars. It is being asserted here that the inscriptional evidence listed above is sufficient for making the case for the existence of a prescriptive calendar in the Yin dynasty, which serves as a solid foundation for the thesis of the present dissertation. 1.3 The Absolute Dates of the Y i n Calendar If the Chinese calendar was indeed invented at the time of Huang Di , it would have a history of about 4,700 years. Since, based on the evidence, this writer does not agree with that belief and instead proposes that the Yin calendar is the first known Chinese calendar, it is necessary now to clarify the absolute dates of the Yin calendar. The Yin calendar is taken to mean the calendar reconstructed from the Yin OBI. It follows that the absolute dates of this Yin calendar are the same as that of the Yin OBI. Therefore, one can establish the dates of the Yin calendar by determining the absolute dates of the Yin OBI. There has been much research into determining absolute dates of the Yin Dynasty. The most recent study is the report by the Xia-Shang-Zhou Duandai Gongcheng j C j ^ J ^ l f f f ^ I M. published in 2000. In this report, absolute dates for the Yin Dynasty have been determined to be 1300 B C - 1046 BC. However, the absolute dates for the Yin Dynasty are not the same as those for the Yin OBI. During the Yin Dynasty, there were twelve kings: Pangeng Xiaoxin Xiaoyi ' h Z i , Wuding J F \ i T , Zugeng l&M, Zujia T l l ¥ , Linxin lM¥, Kangding MT, Wuyi J F ^ Z J , Wending ^ t T , Diyi T T ^ Z J and Dixin However, in the current corpus of the Yin OBI, no inscriptions can be identified as those of the first three, i.e., Y in kings Pangeng, Xiaoxin, 12 and Xiaoyi. In other words, current Yin OBI represent remains of only nine Yin kings. Therefore, the time period of the Yin OBI should be shorter than that of the Yin Dynasty. The report by the Xia-Shang-Zhou Duandai Gongcheng (2000) assigns 1250 BC - 1046 BC to the time period from Wuding to Dixin. But these dates cannot be accepted as absolute dates for the Yin OBI because they are not in agreement with records of lunar eclipses in Period I inscriptions. There are five lunar eclipse records in Period I inscriptions. Zhang Peiyu (1999) i^iflltj, a renowned specialist in early Chinese astronomy and a prominent member of the team of the Xia-Shang-Zhou Duandai Gongcheng, explains how the team made use of those records in determining absolute dates of the Yin Dynasty. However, not all of his conclusions are supported by relevant inscriptions, because of two reasons. First, Zhang Peiyu (1999) does not take month notations of those records into consideration at all. Among those five inscriptional lunar eclipse records, two have month notations. The eclipse on the night of yiyou Zllf (day 22) occurred in the eighth month and that on the night of jiwei (day 56) in the twelfth month. Zhang Peiyu (1999: 39) identifies them with the eclipses on November 25, 1181 BC and December 27, 1192 BC, respectively. His date for the yiyou eclipse indicates that the Yin calendar was three months ahead of the modern calendar, and his date for the jiwei eclipse suggests that the Yin month corresponded to the month of the modern calendar. If Zhang Peiyu's dates are correct, there would be a difference of three months when the Yin people made adjustments to their civil year to synchronize with the tropical year. Such a difference would be so large that the calendar would be useless to determine seasons. Moreover, other dates for these two eclipses, which will be presented shortly, can avoid this discrepancy that Zhang Peiyu has created. This is one reason that suggests that some of his absolute dates are unsupported by relevant inscriptions. Second, Zhang Peiyu selects December 27, 1192 BC as the date for the eclipse on jiwei because he thinks the word zhuo 5?jf, 'cut,' in that record indicates that the eclipse continued past midnight. This conclusion is not in agreement with his date for the eclipse on guiwei. He identifies the eclipse on guiwei as the one on July 11-12, 1201 BC which started at 22:24 and ended at 0:54. If his reason for selecting the date of the eclipse on jiwei is correct, the word zhuo should occur in the record of the eclipse on guiwei as well. The fact is that this word 13 does not appear in the record of the eclipse on guiwei. Zhang Peiyu's explanation could not, and still cannot, fit all relevant inscriptions. This certainly casts doubts on his conclusions. Given the two reasons just discussed, the present writer cannot accept Zhang Peiyu (1999) and Xia-Shang-Zhou Duandai Gongcheng's absolute dates for the Yin OBI. This writer will now present his view on the absolute dates of the Yin OBI. In 1998, this writer published an article on absolute dates for those five lunar eclipses in Period I inscriptions. At that time, both month notations of eclipses on yiyou and jiwei and the usage of the word zhuo were taken into consideration. The dates of those two eclipses were determined to be May 31, 1227 BC and August 14, 1166 BC, respectively, in that article. It was further proposed in that context that the reign of Yin King Wuding was around 1230 BC - 1162 BC (Liu Xueshun 1998: 23). In 2003, this writer published an article on the time span of the Yin OBI. According to the statistics presented, there are 2,179 pieces of oracle bone with inscriptions that divine on specific sexagenary dates whether the Yin king would have disasters in the next 10-day week. Because this kind of divination was done once in ten days, it is possible to calculate an approximate time period for the Yin OBI. The writer's conclusion is that the Yin OBI after Wuding was produced in about 116 years (Liu Xueshun 2003: 22). Based upon this writer's two studies just mentioned above, the writer's perspective on the temporal range of the Yin OBI will now be discussed. Since the Yin OBI are remains of Wuding to Dixin, the upper limit should not be earlier than the start of the reign of Wuding, i.e., ca 1230 BC. Because the inscriptions after Wuding were made in about 116 years, the lower limit should be 116 years later than the end of Wuding's reign, i.e., (1162-116=) ca 1046 BC. It is proposed that ca 1230 BC-1046 BC are absolute dates for the Yin OBI. Because the writer reconstructs the Yin calendar, or what is being termed the first known Chinese calendar, based on the Yin OBI, the absolute dates for this calendar should be ca 1230 BC-1046 BC as well. 1.4 The Synchronic Evidential Approach As mentioned in Section 1.1, there are more than 400 articles and books devoted to 14 studying the Yin calendar. Needless to say, significant progress has been made in the last century. For example, it is now a well-known fact that the Yin calendar was a lunar-solar calendar. On the other hand, specialists still cannot reach consensus on some major principles of this Yin calendar. Some disagreements have appeared because of shortcomings to the method employed in relevant studies. The view taken in the present context is that those controversies can be resolved by adopting a new, more reliable research approach. In the past century, the most dominant approach in the field of the Yin calendar has been the so-called erchong zhengju fa —fiiiEiS'V'i, 'two-layered evidential approach,' which was put forward by Wang Guowei iHift (1877-1927) in 1925 as follows: ^ H ' i T ^ B , mrm±mm, mmyztmo &&mm, %tmmnm mm*&±zw&, mmm-&vz^mfr±%%M, zmm^nmzt, # tt%&—mzm%o ik-mmte, m&^-ni&nftZo ^^z^mim #, ^mtiau&fe, m g e * # M # , ^m^Mwi-mfe, w§-tii. 14 In the time in which we presently live, besides documents on paper, it is fortunate that we have other kinds of materials [that have been dug from] underground. Based upon this kind of [excavated] material, we certainly are able to augment and rectify the paper documents. We also can prove that some portions of ancient books are entirely true records and that even those ungainly records of one-hundred schools of thought may contain one element of truth. It is only in the present time that this two-layered evidential approach can be applied [in the study of early history]. Even though some ancient books have not been proven [true records] yet, one cannot deny them; as for those books that have been proven [true records], one has to accept them. This is what I can say with certainty. The passage quoted above shows that Wang Guowei's two-layered evidential approach has two characteristics: 1) to augment and rectify paper documents with archaeological records, and 2) not to deny ancient paper records that have not been proven true. On its face value, there is nothing wrong with this approach in itself. However, there is a very serious This article is reprinted in Wang Guowei (1968). For this citation, see Wang Guowei (1968: 2078). 15 methodological flaw in its application, as will be shown shortly. At the time when Wang Guowei proposed his approach, there were two competing schools of thought concerning how to treat early Chinese texts. On the one hand, some scholars were very skeptical of the historicity of early Chinese texts. On the other hand, other scholars still believed those texts. By applying the two-layered evidential approach, Wang Guowei published some famous articles like "Buci zhong suo jian xiangong xianwang kao h S ^ ^ J E ^ ^ ^ f c - E E ^ , " "Buci zhong suo jian xiangong xianwang xu kao h S ^ ^ f jAL5fc£ 5fc3E§£#," and "Gushi xinzheng "rS^lfrilE." These articles proved some early text records about the Yin Dynasty, such as those in the Shiji, to be true, which was regarded as a clear demonstration of the historical values of early Chinese texts. These publications thus made him the most prominent scholar of that time in the field of early Chinese history. Because of this, his approach gained a warm reception. Gradually, it has become the most popular approach, and scholars researching the Yin OBI claim they employ it in their investigations. The fact is, however, that the so-called "Guowei" approach applied by most scholars is not actually the same one proposed by Wang Guowei himself. As already indicated, Wang Guowei's two-layered evidential approach has two characteristics. One is to augment and rectify records in early Chinese texts with archaeological records. The other is not to deny those texts that have not been proven true yet. It appears that he did not explicitly encourage people to treat those two kinds of materials equally in researches about early Chinese history. But it is true that he indeed asked others not to deny textual records even when they were not proven correct yet. His ambiguity about whether those two kinds of sources should be differentiated has encouraged other scholars to combine evidence of different natures rather than augment and rectify early Chinese texts with primary materials. This is a clear violation of the first characteristic of Wang Guowei's proposal. Consider Chang Yuzhi's ^ 3 1 2 Yin Shang Ufa yanjiuWLMfh¥&$\%, the second and most recent monograph on the Yin calendar, as an example. Chang Yuzhi (1998: 7) clearly describes her approach as follows. 16 This book thoroughly analyzes the problems of how the Yin-Shang calendar designated day, month and year on the basis of tens of thousands of OBI and by combining them with relevant records in the Shang bronze inscriptions and texts. In her own words, she combines records in the Yin OBI, the primary and contemporary evidence, with records in texts, which are the secondary and much later records. Further, she even interprets inscriptional evidence based on later texts without fully analyzing other relevant inscriptional records. This is a clear example of how the so-called two-layered evidential approach has been applied by scholars in the field of the Yin calendar. Combining inscriptional evidence with textual records and using texts as the main basis for interpretation of the Y i n OBI result in very serious problems. That extant Chinese texts were written one thousand years or more after the Yin Dynasty is beyond dispute. This means that those records concerning the Yin are not necessarily historically accurate or reliable. Because of this, there is a risk to basing the interpretation of the Yin OBI mainly on questionable records of later Chinese texts. For instance, because the phrase dayue j\ $ refers to a long month in later Chinese texts, Chang Yuzhi (1998: 275) has regarded the expression da jin eryue Jz^ZL ^ as a reference to a long month of the Yin calendar. As will be shown in Chapter Three, the expression da jin eryue is not a constituent of that inscription and it was, and still is, incorrect to take it as a reference to a long month of the Yin calendar. In addition, there is contradictory information about the same issue of the Yin calendar in transmitted Chinese texts. If each specialist only accepts information he favors and then proceeds to interpret a given inscription, it follows that there will be disagreements among them. Unfortunately, that is the actual situation as it exists in the study of the Yin calendar. The following represents an example of this kind of biased approach. Chang Zhengguang %]EJt (1981), Wen Shaofeng ^mi>il# and Yuan Tingdong i l t l l ^ (1983), Zheng Huisheng 3$M?£ (1983), Zhang Peiyu ^ i f l f j and Meng Shikai iStftUl (1987), Feng Shi (1990), Wang Hui 3EB¥ (1994), and Chang Yuzhi (1998) have each conducted research concerning the start of the Yin year. Their approaches to this issue are the same, i.e., the so-called two-layered evidential method. The inscriptions they cited are generally the same as well. However, their conclusions are different and exclusive, primarily 17 because they have interpreted the same inscriptional materials according to different Chinese textual records, which will be evaluated in Chapter Four. The inconclusiveness of their research fully reflects the negative effect of their flawed approach to the study of the Yin calendar. In order to avoid yet another possible inconclusive study, it is absolutely vital that the present writer finds and adopts a new methodology for this dissertation. It is suggested that the synchronic evidential approach satisfies this requirement. Ken-ichi Takashima has long been an advocate of the synchronic evidential approach. More recently, he again defines this approach in the paper he presented at the Workshop on Early Chinese Civilization held at the University of British Columbia in Vancouver on March 9-13, 2005. Takashima (2005: 3) states: That is, we should try to interpret the data or issues at hand on the basis of as much intrinsic evidence as possible without drawing conclusions from the later transmitted texts and their commentaries. This seems idealistic, and sometimes interpretation is simply impossible due mainly to the paucity of relevant materials, but such a purist approach will uncover much that seems clouded by the application of the two-layered evidential approach. The differences between this synchronic evidential approach and the two-layered evidential method are significant. While the latter approach does not differentiate between inscriptional evidence and textual records, with regard to the relative importance of each, the former approach does encourage drawing conclusions from contemporary evidence as much as possible and discourages over-reliance on later evidence, i.e., records in late transmitted Chinese texts. As the above example shows, the combination of contemporary and late evidence contributes to the inconclusiveness of various studies of the Yin calendar. Theoretically speaking, by discouraging research based upon later evidence as much as possible, the amount of inconclusive research would be greatly reduced, making the synchronic evidential approach much superior to the two-layered evidential method. The superiority of this synchronic approach is clearly shown in Ken-ichi Takashima's paper presented at the International Symposium on the Historical Aspects of the Chinese 18 / Language in Commemorating the Centennial Birthday of the Late Professor L i Fang-Kuei (1902 - 1987) held at the University of Washington in Seattle on August 17, 2002. The paper concerns the rong ^> sacrifice of the Yin Dynasty, which was one of the most frequently performed sacrifices in the court of the Yin. In the Yin OBI, there are numerous inscriptions regarding this sacrifice. In addition, it is also mentioned in transmitted texts.15 Before Ken-ichi Takashima (2002), Kong Yingda's ?LMii sub-commentary on the rongri & Fj was repeated in various publications. Kong states that rong referred to a sacrifice conducted the day after another sacrifice, an interpretation that has subsequently been accepted by both Chinese and foreign scholars.16 It seems that Kong's commentary has become the standard interpretation. However, after examining the contexts of inscriptions where rong appears, Ken-ichi Takashima finds no evidence to support the traditional theory that rong is a following-day sacrifice. More specifically, the traditional interpretation requires that some sacrifices different from rong were conducted before it on the day previous to the rong sacrifice. But apart from a very limited number of possible examples, there were no sacrifices that were carried out before the rong sacrifice. Moreover, there are examples that are opposite to the interpretation of rong as "next day sacrifice" (Takashima 2002: 8-10). It thus appears that the interpretation of rong recorded in Chinese texts, though very popular among specialists, simply does not match the meaning of rong in the language of the Yin OBI. If so, another conclusion must be drawn: Kong's understanding of the rong sacrifice is inaccurate and it should be disregarded. Clearly, Ken-ichi Takashima's new approach has eliminated the misinterpretation of the word rong in the Yin OBI, which could not be achieved using the approach of the two-layered evidential method. Those scholars who follow the approach of the two-layered evidential method, be it consciously or subconsciously, have not made due effort to understand the actual meaning of the rong sacrifice in the language of the Yin OBI and have simply repeated an incorrect interpretation of the word. On the other hand, Ken-ichi Takashima (2002) is able to reach his decisive conclusion because of his synchronic evidential approach. While the researches mentioned in Note 16 are inconclusive, Ken-ichi Takashima (2002) seems conclusive insofar 1 5 See the chapter "Gaozong rongri ~\%7F.$> El" of the Shangshu, which James Legge (1872: 264) has translated into English. 1 6 To name just a few, Luo Zhenyu (1915.2: 16b), Rao Zongyi (1957: 11), Li Xiaoding (1965.3: 2762-2764), Xu Zhongshu (1988: 947-948, 995), James Legge (1872: 264) and Bernhard Karlgren (1950: 26). 19 as the rong sacrifice is concerned. The reason for the superiority of this new approach is that, to the greatest possible extent, it excludes records in later Chinese texts from the research. Compared with contemporary OBI, the later texts are far less reliable and less credible historically. Moreover, they may well have been further corrupted at even later times. A l l of these factors make these texts very unreliable, and cause them to be possible sources of errors in studies that are based on them. By excluding them from research whenever possible, the synchronic evidential approach can avoid their negative influence. This has been the rationale for adopting the synchronic evidential approach in this dissertation focusing on inscriptional evidence to reconstruct the Yin calendar. 1.5 Aims of This Study Since it involves reconstruction of a calendar, this study will produce calendar tables. It is firmly believed by this writer that accurate calendar tables for the whole Yin Dynasty would definitely be welcomed by scholars in the field of early Chinese history, since such tables would be a significant aid in determining the chronology of Chinese history. Unfortunately, the fact is that this worthwhile ideal is extremely difficult to attain. Dong Zuobin was unable to achieve this task in 1945. Half a century later, Chang Yuzhi (1998) also failed to provide such a reconstruction. It still is not a realistic aim for the following two major reasons. First, there are no generally accepted absolute dates for the Yin Dynasty yet. It is easy to understand that, in order to compile a calendar table for each year of the Yin Dynasty, absolute dates for each Yin king have to be conclusively determined. After nearly five years of work by a team of more than 200 Chinese scholars, the Dating Project of the Xia, Shang, and Zhou dynasties could not present final conclusions about dates of each Yin king, a fact that has already been admitted by the team. Because of the uncertainty regarding those absolute dates, there is no base for any reliable and meaningful reconstruction of a calendar table for each year of the Yin Dynasty. Second, the Yin calendar evolved over time. For example, as shown in Chapter Three, there is evidence to suggest that the year-end intercalation was replaced by the in-year intercalation. But it still is unclear as to how long that transitional period was. In addition, the 20 criteria for the Yin people to assign intercalary months have not been thoroughly studied yet. As a result, it will probably not be possible to reconstruct with certainty an accurate calendar table for each year, even i f the absolute dates for each Yin king were established. Given the two reasons above, this study will not aim at accurately reconstructing a calendar for every year of the Yin Dynasty, although a calendar for a few years will be constructed when necessary. Rather, this dissertation will focus on explaining the system of the Yin calendar. Chapter Two will thoroughly investigate the time divisions and the start of a Yin day. There are some time divisions that are generally accepted by specialists due to the fact that they were still used as time divisions in later periods. As for those alleged time divisions that are only seen in the OBI, this writer will utilize applicable criteria to determine whether or not they are true time divisions. For those time divisions in early morning, modern astronomy will be used to rationalize them. The start of the Yin day has long been a hot controversy in the field and will be dealt with in this chapter as well. Chapter Three will address several issues surrounding the Yin month. The writer will determine the number of months in a Yin year. Also, in order to ascertain the length of a Yin month, the plan will be to thoroughly investigate relevant inscriptions in order to determine i f Yin months could be longer than 30 days or shorter than 29 days. With regard to the arrangement of Yin months, inscriptional evidence will be cited to show the existence of both the year-end intercalation and the in-year intercalation. In addition, the beginning of the Yin month will be demonstrated by reconstructing several months surrounding a lunar eclipse recorded in the OBI. In Chapter Four, the discussion will turn to the designation and commencement of the Yin year, focusing on the beginning of the Yin year. First, this writer will utilize applicable criteria to determine the word si IE to be the designation of the Yin year. Then, not only will the writer evaluate different theories about the start of the Yin year, but he will also show that the Yin year started with the second month before the month of the winter solstice. Finally, in Chapter Five, the writer will further discuss the significance of the synchronic evidential approach in the research of the Yin calendar. This approach not only requires scholars to change their way of conducting researches about the Yin calendar, but also requires that they change their attitude toward results yielded by this new approach. It is this 21 writer's firm conviction that no one should reject these results on the basis of later Chinese calendars. From the point view of the history of the Chinese calendar, the development of early Chinese calendars is not linear. In early China, several calendars existed simultaneously. Moreover, their development was independent of one another. As a result, the evolution that occurred in one calendar might not have taken place in other calendars. Because of political reasons, when one calendar replaced another one, the same evolution might take place again. For example, the transition from the year-end intercalation to the in-year intercalation happened at least three times in early China. Therefore, one should judge the results by the synchronic evidential approach solely on the basis of contemporary evidence not on later calendars. 22 CHAPTER TWO "THE DAY" IN THE YIN CALENDAR 2.1 Introduction In any given day, there is a period of brightness and of darkness that may be called daytime and night-time, respectively. Generally speaking, the cycle of daytime and night-time, and that of people's working and resting, are synchronous. If the concept of "the day" occurred before the invention of the timepiece, which seems to have been the case, it must have been this synchronization that made people aware of the day as a period of time The scientific archaeological excavations at Anyang, where the residence of Yin kings was located, have been carried out since 1928. But no timepiece has ever been discovered.1 On the other hand, the Yin people indeed recorded dates with the cyclical 60 ganzhi in the Yin OBI. This is direct evidence showing that the Yin had the concept of a calendar day. Further, in the Yin OBI, there are numerous inscriptions about time divisions of a day, which will be discussed in this chapter. Specialists of the Yin calendar do not dispute the existence of time divisions of the day in the Yin Dynasty. However, they cannot reach consensus about the number of time divisions of the day there were during that dynasty. This absence of consensus is due to specialists lacking clear criteria forjudging a time division in the Yin OBI. Also, they cannot agree upon the time division with which a Yin day started. The present chapter has three objectives: first, to present this writer's criteria for determining time divisions of the Yin day; second, to analyze specific time divisions in the Yin OBI and characteristics of these time divisions; and third, to present this writer's concept of the start of the Yin day. 2.2 How to Determine Words of Time Divisions It is common sense that one has to have a criterion by which to pass judgment on any 1 For details of excavations at Anyang, readers are referred to the Yinxu de faxian yu yanjiu iSttM^iffi-^'fiif %, which was written by members of the archeologica! team at Anyang in 1994. 23 issue. However, with the exception of Song Zhenhao TJcfJlil:, scholars in this field usually do not clearly state their criteria, i f indeed they have any, for deciding time divisions in the Yin OBI. This apparent absence of criteria makes it difficult to evaluate the potential merit of these studies in an effective way. In order to avoid this difficulty, it is this writer's intention to make clear his criteria as to how he determines whether a word in the language of the OBI is used as a time division. 2.2.1 Lexical Meaning An authoritative dictionary is useful in deciding upon a word for a time division. If one meaning of a word denotes a time period of day, then this word can be regarded as a time division. Consider, for example, the word "dawn" in modern English. Hornby (2002: 362) lists the following two meanings under the entry of "dawn": "time of day when light first appears; daybreak" and "beginning; first signs of sth." Based on the first meaning, one can say that the word "dawn" is used as a division of time in modern English. It follows that a dictionary compiled by the Yin would be of great help to modern scholars in determining time divisions in the language of the Yin OBI. However, the unfortunate fact is that there was no dictionary during the Yin Dynasty. Because of this, lexical meanings of a word in the language of the OBI have to be deduced either from inscriptional contexts and/or from its usages in early transmitted Chinese texts. Below is an example. After citing five inscriptions where the word dan i=L occurs, Chang Yuzhi (1998: 136) offers the following explanation: i f : " 0 , TO" , ^n^mtMo The Shuowen says, "dan means 'bright.'" So dan refers to the time of daybreak. It is apparent that she has assigned the meaning "daybreak" to the word dan in the Yin OBI, based on its definition in the Shuowen. Then, she further classifies dan as a time division in the OBI. From the point of view of the synchronic evidential approach, it is unsatisfactory to assign meanings to a word in the Yin OBI based upon its usage in transmitted Chinese texts. 24 On the other hand, it would have been extremely difficult, i f not impossible, to achieve the progress that has been made in the study of the OBI in the last century i f the OBI were considered as completely separated from early Chinese texts. The present writer concurs that the synchronic evidential approach has limitations and that one should not take it to extremes. As long as records in Chinese texts do not contradict inscriptional records, they can be taken into consideration when one decides meanings of a word in the language of the Yin OBI. Therefore, it is this writer's opinion that lexical meanings of a word in early Chinese texts are helpful in determining time divisions seen in the Yin OBI. 2.2.2 Inscriptional Context It is a tradition that Chinese scholars deduce meanings of a word from contexts. If two words are found in similar environments, huwen S 3t in Chinese terminology, their meanings would be deemed similar or the same. This approach is criticized by Ken-ichiTakashima. Nevertheless, this approach is still adopted by many Chinese scholars. As for specialists in the field of the Yin calendar, Song Zhenhao (1985) has made efforts to establish a contextual criterion forjudging time divisions in the OBI. Chang Yuzhi (1998: 143-150) accepts Song's criterion. The following is Song's thesis: &z, ¥# 'm m' '\mn&~m®Am%±WBmi 1®, 'M.....M' ^#§Btn> Bmmnfto3 In sum, the sentence pattern of 'hui ...jitf in the bone inscriptions exactly shows a ritual practice of the Yin people attaching importance to the time and date of big events. The several words between 'hui ... jitf exclusively refer to time divisions, dates or month notations. In his comments to the draft of this dissertation, Ken-ichi Takashima points out that it is 2 When analyzing the words qi % and hui in the OBI, he makes the following comment about Han Yaolong (1972: 10b-1 la): "These two words provide another opportunity for Han Yaolong (1972: 1 Ob-1 la) to fall victim to the assumption that if two elements are found in similar environments, they must have a similar meaning." (Takashima 1996a: 470). 3 This criterion was first proposed by Song Zhenhao (1985: 305) and repeated by Song Zhenhao (1991: 38). 25 linguistically naive to take issue at length with this theory. On the other hand, a terse or cavalier rejection would not likely convince Song Zhenhao and Chang Yuzhi that Song Zhenhao's criterion does not work. It is necessary to demonstrate at some length why Song Zhenhao's proposal cannot be a valid criterion, by showing that his premise is indeed false. It is Song Zhenhao's opinion that all words between words hui and you4 refer exclusively to time divisions, dates, or month notations. If his assertion is correct, one can easily determine whether a word between hui and you is a time division in the language of the OBI, because time divisions are very different from dates and month notations. The fact is, however, that not every word between hui and you is a time division, date or month notation; and this can be clearly shown by the following inscriptions. [0i]*:#5feS^ . % £ 5 t m 27489 It should be a mother5 to whom [the king will] first perform the yow-cutting sacrifice. It should be a brother to whom [the king will] first perform the yow-cutting sacrifice. It should be a father to whom [the king will] first perform the you sacrifice. Heji 27489 [02]&±¥5fcl^o fei5feS^o «1=ril» 28272 It should be Shangjia to whom [the king will] first perform the yow-cutting sacrifice. It should be Zhuren to whom [the king will] first perform the j/ow-cutting sacrifice. Heji 28272 In those five inscriptions above, the words between hui and you are 'a mother,' 'a 4 Like other Chinese scholars, Song Zhenhao transcribes the bone graph F$ as jiu M, 'wine.' However, Ken-ichi Takashima points out that "the graph does not seem to stand for the noun 'wine.'" In addition, his paleographical study lets him understand the word "to have meant some way of cutting, the neat and beautiful execution of which was required as a preparatory sacrifice." (Takashima 1996a: 110-111). Given these reasons, it is appropriate to follow him by translating the word as 'theyou-cutting sacrifice.' 5 This is a Period I inscription. Inscriptional records show King Wuding had four fathers (Heji 2331) and six mothers (Heji 2536, 2537, 2575, 2576, 2581 and 2582). These inscriptions do not show the you-cutting sacrifice was performed to one person or persons. The present writer tentatively translates them as 'a mother' and 'a father.' 26 brother,' 'a father,' 'Shangjia,' 'Zhuren,' and xian, which means 'first.' In these inscriptions, a mother, a brother, a father, Shangjia and Zhuren are all recipients of the .yo w-cutting sacrifice. The other word xian ft, 'first,' is an adverb. Apparently, none of these six words refers to dates, month notations, or time divisions of a Yin day. Other similar inscriptions can be found on Heji 1351, 34221, 34223. It should be decapturi (to be used) in the ^ow-cutting sacrifice to Zuyi. Heji 190 In this inscription, the word between words hui and you is fa i% the victim of the vow-By now, it becomes obvious that not every word between hui ... you refers exclusively to a time division, date or month notation in the OBI. The premise of Song Zhenhao's theory is therefore incorrect. His criterion for judging time division in the language of the Yin OBI is untenable. 2.2.3 Lexical Meaning and Inscriptional Context Chapter One has made clear the fact that this study adopts the synchronic evidential approach. Every effort is made to draw conclusions from evidence actually found in the Yin OBI. It would be intellectually satisfying i f it were possible to cite clear-cut inscriptional evidence to prove or disprove whether a particular word in the language of the OBI is used as a time division. However, the fact is that the synchronic evidential approach has its limitations. Since the Y i n compiled no dictionary in which to look up meanings of words in 6 In his comments to the draft of this dissertation, David Pankenier raises the question: "Why are fa mdyou here not simply the names of two sacrificial methods?" The explanation is as follows: the word fa can be the name of a sacrificial method. Wang qi fa 3EKI35, 'the king will perform the dismembering sacrifice,' (Takashima 1996a: 252) is such an example. On the other hand, it can also be used as the victim of sacrifices. The expressions jiufa 'nine decaputuri,' and shifa -\~{%, 'ten decaputuri,' (Takashima 1996a: 220-221) are two examples. In the following inscription, fa is the victim of the yow-cutting sacrifice: ...B2TZ.-HftXE... ... ^ OM-cutting sacrifice fifteen decaputuri to Xiayi ... Heji 903 In addition, the word hui is a copula in the OBI whose function is to move the patient object to the front of the verb (Takashima 1996a: 453). This, too, indicates that fa in Example 3 is the victim of the _yo«-cutting sacrifice. Considering all these facts, it is suggested that one takes fa in this example as the victim of the yow-cutting sacrifice. cutting sacrifice6. It definitely is not a term for time division. 27 the Yin language, this writer joins other scholars in having to base his judgments in part upon the meanings found in transmitted Chinese texts. If there is evidence to show a word is used as a time division in classical Chinese, then this study will investigate whether it makes sense to carry such a temporal use back to the earlier OBI language. Only when such an interpretation fits inscriptional contexts will it be so interpreted in the OBI. Again, consider the word dan B. as an example. As stated in the previous section, Chang Yuzhi (1998: 136) already cites evidence for the word dan being used as a time division in early Chinese texts. Also, it makes sense to interpret dan as a division of time in the following inscription. Upon daybreak the next day, it will rain heavily. Heji 41308 Such an interpretation of the word dan in Inscription 4 is acceptable to all specialists. So the word dan in the OBI can indeed be taken as a time division. Since this word is a time division in classical Chinese, and since this usage fits inscriptional context such as Example 4, the present writer does consider the word dan as a time division in the OBI. The foregoing example shows that the first thing to do in determining whether a particular word is a time division in the OBI is to examine the early Chinese texts to determine whether the word in question is used as a time division there. Normally, a word will not be deemed a time division i f there is no evidence for the word being used as a time division in early Chinese texts. In such situation, even though an interpretation of a given word may seemingly fit some inscriptional contexts, this writer does not accept the particular word as a time division when the outcome of pursuing such a method could be the creation of many time divisions that are incorrect or false. The next section will pursue this discussion further. In sum, in this chapter lexical meaning and inscriptional context form the basis for determining a word as a time division of the Yin day in the language of the OBI. If a word occurs as a time division in the early Chinese language and such a meaning fits relevant inscriptional contexts, this word will be accepted as a time division in the OBI. If a word does not appear as a time division in early Chinese texts, confidence is lacking in considering it as a time division in the OBI. [04] f l B l T C S . 28 2.3 Time Divisions of the Yin Day It is common knowledge that a day consists of two parts: daytime, 'time of brightness,' and night-time, 'time of darkness.' The usage of the words ri B, 'day or daytime,' and xi , 'night-time' as a time division in early Chinese texts and various inscriptional contexts containing these two words show that the Yin day can be divided into ri B, 'daytime,' and xi & , 'night-time.' As for the time division ri, it can be further divided into different sub-divisions of time. 2.3.1 Ri In early Chinese texts, the word ri often refers to daytime. The following is an example: He stood leaning against the wall of the courtyard and cried. Day or night his voice was not silent; a spoonful of water did not enter his mouth. (Legge 1872b: 757) In James Legge's translation above, the word ri means "daytime," although he translates it as "day." It is one of many pieces of evidence for the usage of ri as a time division in early Chinese texts. Among the corpus of the Yin OBI, there are many thousands of inscriptions that contain the word ri. In the vast majority of these inscriptions, ri can be interpreted as a whole day. Determined effort has been exerted by this writer to find examples of the word ri as a time division in the OBI. Up to now, only a few inscriptions show that the word ri may actually refer to daytime rather than a whole day. Two such examples are cited below: 24358 Crack-making on guiyou (day 10), Chu divined: "[In the next] 10-day week, there will be no disasters." During this daytime, L i stopped; in the night-time, there were wild buffalos. [The king] was at Xiu . [Day guigyou was in] the eighth month. Heji 24358 In the prognostication of this inscription, the words ri and xi appear to be two time 29 divisions. Because two different events happened in these two time divisions, it is reasonable to consider them as different time periods. Based on their usage in early Chinese language, it makes sense to interpret the word ri, in this example, as daytime. [06 ]T£P h ^ F J M o ^ S . i^m 33871 Crack-making on dingmao (day 4), [X] divined: "During daylight today it will rain." "At night-time, it will rain." Heji 33871 The two inscriptions above divine the time of rain. Because of this, it is possible to interpret the word ri and xi as two different, possibly mutually exclusive, time divisions. If so, the word ri would be a time division that refers to daytime only. The word ri is accepted as a division of time in the Yin OBI by all scholars in previous publications, however, without providing any solid evidence. The reason for the lack of solid evidence might be that these scholars accept it as common knowledge that the word ri refers to daytime. In the present context, however, textual and inscriptional evidence is cited in order to demonstrate that, as a time division, the word ri can indeed refer to daytime in the language of the Yin OBI. In English, daytime includes time divisions such as morning, noon, afternoon, and evening. There is evidence showing that ri is divided into a number of sub-divisions in the Yin OBI as well. Before proceeding to discuss the time division xi, we turn to demonstrating, one by one, the sub-divisions of ri. As mentioned in Chapter One, the Yin OBI can be divided into five periods. The time divisions of ri change with periods. In order to reflect such an evolution, it would be appropriate to present these time divisions period by period, starting with Period III, because many inscriptions of this period contain more than one sub-division of the time division ri. 2.3.1.1 Sub-divisions of Ri in Period III 2.3.1.1.1 Su The word su M appears as a division of time in early Chinese texts. Consider the following example: 30 Mettle o ifov-mm Morning and night, you must be respectful. (Legge 1872:47) In this sentence, the word su means "morning." This is a piece of evidence for su as a time division. More examples can be found in Chapter "Gaoyao mo jp;pfi)iJ|" of the Shangshu; Chapter "Caifan 3£3I", "Xinglu "Xiaoxing /ML", "Ding zhi fang zhong fcZI:^", "Dongfang wei ming WjfrW, "Zhi gu g W , "Sheng min ^EK", "Zheng min U K " , "Min yu xiaozi K l M ^ T " , and "You bi WW of the Shying Wi&; Chapter "Shiguan l i ±M4L", "Shihun li dr if *L", and "Tesheng kuishi l i # TL" of the Yili ft 4L; and Chapter "Zhouyu J^il. ' ' and "Jinyu # iff' of the Guoyu Hip-.7 In addition, there is a set expression su xingye mei J U T S ^ ^ , 'to get up at the time of su and to go to bed at night.' A l l these show that the word su is often used as a time division in early Chinese texts. In the Yin OBI, the word su has such a usage as well. Su was inscribed asBSJ that consists of two components: yue B, 'moon,' and a kneeling figure. According to X u Shen's i^ ffJl Shuowen jiezi i&JcM^, the word su also has the same two components. Therefore, this bone graph is transcribed as su. For detailed studies about the graphic evolution and original meaning of this character, see the entry of su in Jiagu wenzi gulin. Song Zhenhao (1985: 307-309) and Chang Yuzhi (1998: 150-151) have already pointed out the usage of su as a time division in the OBI. Their view is fully supported by the following inscription. [07] gjsfcMtt^&A. 9 H ^ B ^ ^ ^ I A . 26897 On the gui-day, at the time of su, Guard will attack Zai because [that will] not 7 Ruan Yuan, a scholar of the Qing Dynasty, collected an enormous number of commentaries about the meaning of words in early Chinese texts. This book was reprinted in Taibei in 1967. For details of commentaries about the word su, see Ruan (1967: 886). 8 To be specific, readers are referred to Yu Xingwu (1996: 423). 9 Zai was an enemy of the Yin, which is shown by Yibicm 2503. For discussion of zhik%, see Yu Xingwu (1996: 1725-1730). Yu Xingwu points out that the word zhi is interchangeable with the word yi M, 'to exterminate,' in some texts. In these two inscriptions, zhi occurs after fai% 'to attack.' It seems to be the possible result of fa. If so, it makes sense to translate zhi here as 'exterminate.' 31 exterminate [his] people. On the gui-day, upon the time of dan, [Guard] will then attack Zai because [that will] not exterminate [his] people. Heji 26897 If one interprets the word su as morning, the word-by-word translation of the first inscription would be this: "On the gw'-day, Guard, in the morning, will attack Zai because [that will] not exterminate [his] people." It makes good sense. So it fits the above inscriptional context to take the word su as a time division. Since the word su is a time division in early Chinese texts, and since such an interpretation fits inscriptional context, this writer accepts the word su as a division of time in the language of the Yin OBI. However, it does not necessarily denote the whole morning in the OBI, as implied by James Legge's translation above. Both Song Zhenhao (1985: 308-309) and Chang Yuzhi (1998: 151) propose that the time division su refers to sometime in the night-time, an assertion with which this writer disagrees. Section 2.4, will make clear why the time division su does not refer to some period of the night-time. 2.3.1.1.2 Dan The word Dan K is explicitly defined as a time division in the following Chinese texts: Dan is the time just when the sun is rising. Dan is early morning. M., HJ-IHO micm^} Dan is daybreak. Strictly speaking, sunrise, early morning and daybreak do not denote the same time period. But the above records indeed show that dan is a time division in early Chinese texts. E l In the Yin OBI, there is a graph Its*, which has been transcribed as dan M.. The following inscription shows that the word dan is used as a time division in the Yin OBI. [08] h : S 0 £ 0 M 1 r B ( = H t ) u ^ [ M ] o 1 0 These two records are collected in Ruan Yuan (1967: 771). 32 £HS1TB ( = B t ) l S . ((^it)) 624 Crack-making on xinhai (day 48), [X] divined: "Next day ren(zi day 49), from dan to shishi, it will not rain." "On ren{zi), from dan to shishi, it perhaps12 will rain." Tunnan 624 [09] g f i l t H (=EWMo {T&m 42 "From dan to shishi, it will not rain." Tunnan 42 The topic of Examples 8 and 9 is when it will rain. It fits this context perfectly to interpret the word dan as a time division. Since the word dan is also used as a time division in early Chinese texts, it seems likely that it is a time division in the Yin OBI. However, specialists who accept the word dan as a time division in the language of the OBI cannot agree on the exact time to which it refers. Chen Mengjia's hypothesis (1956: 233) is that it is 6:00 a.m. Song Zhenhao's opinion (1985: 332) is that it corresponds to 3:00-5:00a.m. Cao Jinyan (1987: 197) assigns dan to the time of sunrise. Chang Yuzhi (1998: 136) follows the Shuowen's definition and understood the time division dan as tianming ^ B J S , sometime before sunrise. David N . Keightley's translation for dan is "dawn" (2000: 19). In light of the various opinions about the exact time represented by the time division dan, it seems problematic for Chen Mengjia (1956: 233) and Song Zhenhao (1985: 332) to fix dan to specific hours. According to Chen Mengjia (1956: 230), the bone graph dan depicts the sunrise (M.% EJ dj^M-tWIrl;^). Because of this, one can surmise that he thinks the word dan refers to the time of the sunrise. But it is common knowledge that the time of sunrise always changes with season and location. For example, in Anyang, the sun rose at 5:11 am on July 5, 1166 BC and at 7:39 a.m. on December 31, 1166 BC. The sun rose at 5:06 a.m. on June 21, 2002 and at 7:30 a.m. on December 22, 2002.1 3 Clearly, the sun does not always rise " The bone graph is often transcribed as ri El by other scholars. However, according to Ken-ichi Takashima 2003 and 2004, the word it represents is shi 'time,' an interpretation with which the present writer concurs. 1 2 Paul L-M Serruys (1974: 25 ff.) proposes the theory that word qi% often expresses a course of action or state that was undesirable to the Shang in complementary charges. It indeed seems true that qi always occurs in the undesirable charge of a set of complementary charges. But this might be the result of qi being a modal particle. As pointed out by Ken-ichi Takashima (1996a: 54), under certain specific conditions, qi adds a semantically variable element of 'unsureness' to the verb that it modifies. 1 3 July 5, 1166 and December 31, 1166 BC were the summer and winter solstice of that year, respectively. Similarly, June 21 and December 22 were the summer and winter solstice, respectively, of 2002^  The time of the sunrise was computed by 33 around 6:00 a.m., either during the Yin Dynasty or in modern times. It is a fact that the time of sunrise on July 5, 1166 BC actually was 2 hours 28 minutes earlier than that on December 31, 1166 BC. Therefore, it is inaccurate for Chen Mengjia to fix the sunrise in the Yin Dynasty around 6:00 a.m. Also, the sunrise on December 31, 1166 BC was 7:39 a.m., which was 2 hours 28 minutes later than 5:00 a.m. It is therefore equally inaccurate for Song Zhenhao to fix dan at 3:00-5:00 a.m. According to the Shuowen, dan refers to the time of ming EJE. Since ming denotes the three quarters before the sunrise ( 0 t i j F I U H j ^ B j ) 1 4 , it is understandable for Chang Yuzhi (1998: 136) to interpret the time division dan as a time period that does not include the actual rising of the sun, or sunrise. But the sunrise is a part of the time of dan in early Chinese texts, as cited at the beginning of this section. Moreover, this usage is in accordance with the structure of the character dan. It becomes questionable for Chang Yuzhi to exclude sunrise from the time division dan. If the word "dawn" can be used to refer to the time of actual sunrise and some time before sunrise in English, dawn is an appropriate translation of dan. But this is not a precise description for the time division dan. Normally, this is the point where other researchers have stopped. This author, however, would like to explore the possibility of defining the length of dan in terms of minutes. This does not in any way imply that the Yin already had astronomical knowledge as advanced as modern astronomy. Rather, this discussion merely represents this author's effort to explain the time division dan in a new way. Based upon the definition of dan in early Chinese texts, this time division generally corresponds to the time division tianming in modern Chinese, as proposed by Chang Yuzhi (1998: 136). Tianming is the time for people to do morning work. In other words, it is the time when there is no need for the assistance of any artificial illumination in order for people to conduct outdoor activities. In this sense, tianming might be the civil twilight in modern astronomy. The Astronomical Applications Department of the United States Naval Observatory the software SkyMap Pro v.9.0.9 provided by C. A. Marriot. The writer was kindly referred to this software by Mr: Liu Ciyuan of Shanxi Observatory, China. 14 Ke M is the Chinese term for a quarter hour. According to this record, which is collected by Ruan Yuan (1967: 332), the word ming Hj  refers to 45 minutes before sunrise. 34 provides the public with tools to access modern astronomical knowledge. On its website http://aa.usno.navy.mil civil twilight is defined as: ...the limit at which twilight illumination is sufficient, under good weather conditions, for terrestrial objects to be clearly distinguished.... In the morning before the beginning of civil twilight... artificial illumination is normally required to carry on ordinary outdoor activities. According to this definition, civil twilight in early morning is the time when people do morning work without the assistance of artificial illumination. It seems that the term civil twilight generally corresponds to the time period of tianming in modern Chinese or of dan in the Yin OBI. Since the start of civil twilight and sunrise can be accurately calculated in minutes, this provides the basis to explore the possible length of the dan in the language of the OBI. In Anyang, on July 5, 1166 BC, the summer solstice, civil twilight started at 4:28 a.m. and the sun rose at 4:59 a.m. On December 31, 1166 BC, the winter solstice, the start of civil twilight was 7:10 a.m. and the sunrise 7:39 a.m.15 In both cases, it is about half an hour from the start of civil twilight to sunrise. Accordingly, I would propose that the time division dan in the language of the Yin OBI is about half an hour long. 2.3.1.1.2.1 The Length of Su Before discussing the length of the time division su, it is instructive to recall Example 7: [07] ^ M i # « A o H ^ M 7 M # ^ t A o l&m 26897 On the gui-day, at the time of su, Guard will attack Zai because [that will] not exterminate [his] people. On the gui-day, upon the time of dan, [Guard] will then attack Zai because [that will] not exterminate [his] people. Heji 26897 As shown in previous sections, both su and dan are time divisions in the OBI. Moreover, 1 5 All these specific times were calculated with Skymap. 35 it fits the context of this example to interpret both su and dan as time divisions. So this example can be regarded as evidence for the usage for these two words as time divisions. But the most important reason for recalling this example is to discuss the sequence of the time divisions su and dan and the possible length of su. In the second inscription above, the word yu f occurs before the time division dan M.. Ken-ichi Takashima (1990: 36-37) points out that "whenyw is used, it always introduces a day more remote than hui does." The reason he suggests for this is that "in the bone inscriptions the word^w f had a clear 'futurity' meaning. This receives further support from its etymological doublet wang 'to go.'" He has already cited several inscriptions with clear dates to support his opinion. Since the word yu had a clear "futurity" meaning, it can be inferred that the word yu should always introduce a day that is more remote in a pair of inscriptions, even i f the word hui does not appear. This is certainly the case, which can be shown by the following inscriptions: [io]^sK m-. mmm^o h T-R^mnm&o - t - - ^ K mix Crack-making on xinchou (day 38), Xun [divined:] "On xinhai (day 48), [the king will] perform the jyow-sacrifice and offer //ang-millet." [Xun divined in] the twelfth month. On xinchou (day 38), [Xun] divined: "On xinyou (day 58) of the first month, [the king will] perform the j/ow-sacrifice and offer //ang-millet." In the twelfth month [Xun] divined. Heji 21221 [11] ^ 3 1 K j £ : ^ T / N ^ ^ o 4~-£Bm° i&m n 6 Crack-making on xinchou (day 38), [X] divined: "Perhaps, upon the coming of the sixth month, [she] will give birth." Divined: "In this fifth month, [she] will give birth." Heji 116 In Example 10, the date after the word yu is day xinyou (day 58) of the first month and the date not being introduced by yu is xinhai (day 48). The date introduced by yu is 10 days later than the other date. In Example 11, the month introduced by yu is the sixth month, and the month not being introduced by yu is the fifth month. The former is one month later than the latter. These passages amount to strong evidence that the word yu introduces a more 36 remote date in a pair of inscriptions, even when the word hui does not occur. In Example 7, yu introduces the time division dan but does not introduce the word su. Because the time introduced by yu is more remote than the time not being introduced by yu, it logically follows that the word dan should be a time division that comes after the time division su. This is, in actual fact, the order of time divisions dan and su in the language of the Yin OBI. As for the time division su, there must be some compelling reason for the Yin to have established this term. To date, however, that reason has not been found in the OBI. At this juncture, it is useful to consider su from the point of view of modern astronomical knowledge. In the natural development of the day, before civil twilight comes nautical twilight. At the beginning of nautical twilight, under good atmospheric conditions and in the absence of other illumination, general outlines of ground objects may be distinguishable. According to the United States Naval Observatory website, before nautical twilight, "sky illumination is so faint that it is practically imperceptible." Theoretically speaking, the change in the brightness of the sky is discernible, and it is reasonable to suppose that the Yin people would have noticed this change. This might be one possible reason for the Yin to have created a term to refer to the time period of nautical twilight. If this is the case, then by calculating the time of nautical twilight, one can establish the possible duration of the time division su. In Anyang, on December 31, 1166 BC, the start and end of nautical twilight was 6:38 a.m. and 7:10 a.m., respectively. That day's nautical twilight was about a half hour long. On July 5, 1166 B C , the nautical twilight started and ended at 3:50 a.m. and at 4:28 a.m., respectively. Its twilight was about 40 minutes long. These data imply that nautical twilight in the morning in the Yin dynasty was 30 to 40 minutes long. This author thus proposes that the time division su may refer to a time period of nautical twilight in the morning, which is about half an hour long. 2.3.1.1.3 Shishi and Dashi On the bamboo strips of Q i n found at Shuihudi @i^itk in 1975, the word shishi clearly is used as a time division: 37 Sunrise is the two-hour period mao, shishi is the two-hour period chen ... s # 0 T ± o m# • mm^n-m At the time of shishi, [he] handed it in. "The biography of the King of Huainan", Hanshu The second record above is full of interest for this author because it may throw light on which word the bone graph ri 0 represents in the Yin OBI. The phrase shishi lltBT is written as shiri #0 in the OBI, in which the word represented by the graph ri 0 is shi 0\J*, as pointed out by Ken-ichi Takashima (2003, 2004, and 2004-05). In the Han dynasty, shishi had become one word, and the character ri 0 can still represent the word shi Bf. To express the meaning "the time of shishi 1ft BT," the Han people created the phrase rishishi 0 litBT. In any event, those two records in early Chinese texts are evidence for the usage of the word shishi as a time division in early Chinese. This usage of shishi can be found in the Yin OBI as well. The following six inscriptions are examples in which shishi is used as a time division: [ 0 8 ] $ ^ h : S 0 £ B M l J r 0 (=fcf) T [ M ] . #0 (=&tyM4,B (=st) T I . ( = 0 T ) M * 0 (=BT) K M . i^m)) 624 Crack-making on xinhai (day 48), [X] divined: "Next day ren(zi day 49), from dan to shishi, it will not rain." "On renizi), from dan to shishi, it perhaps will rain." "From shishi to zhongshi, it will not rain." "From shishi to zhongshi, it perhaps will rain." Tunnan 624 [09] i I l t B (=Bt)TMo #0 (=BT)M'tJ 0 (=BT)TMo «t£pJO 42 "From cfcw to shishi, it will not rain." 1 6 Those Qin bamboo strips were published in the Yunmeng Shuihudi Qin mu :zj^8f§ife?SJ§ by the Yunmeng Shuihudi Qin mu bianxiezu in 1981. For the photo of this particular strip, see Yunmeng Shuihudi Qin mu (1981: fig. 156). 38 "From shishi to zhongshi, it will not rain." Tunnan 42 The topic of those inscriptions above is whether it would rain in a certain time period. It makes sense to take shishi as a time division in these inscriptions. To what time does shishi actually refer? Since shishi means the time of shi lit, 'meal,' as a time division, shishi should be related to meals,in the OBI. This conclusion is accepted by specialists in the field. 1 7 Dong Zuobin (1945.1.1: 5b) points out that, in early China, meals were commonly eaten only twice a day. Chen Mengjia (1956: 231) cites various early Chinese texts to support that statement. That people ate meals twice a day is in accordance with the Yin OBI. Up to now, there are records of dashi ^1ft, 'big meal,' and xiaoshi 'j>1ft, 'small meal,' in the OBI; there are no records of any other meal. It is thus highly likely that the Yin people ate only twice a day, once in the morning and once in the afternoon. If that is the case, which meal was served in the morning, the dashi or the xiaoshil Chao Fulin (1989: 162) proposes that da and xiao in such contexts do not refer to the size of a meal; they may refer to zao or zhao 'early,' and wan or xi , 'late,' respectively. According to Chao Fulin's proposal, dashi should be the meal served in the morning. However, Chao Fulin might be mistaken in saying that da and xiao in dashi and xiaoshi do not refer to the size of a meal, because the size of the meal in the morning was bigger than that in the afternoon in early China. In the Warring States Period, the meal in the morning was called yong d , and the meal in the afternoon was called sun In the chapter "Siyi "s] ijJC" of the Zhouli JR)?L, the small rite was called sun and the big rite was called yongxi ('ML S^t. ^LHlti'Ti). In the phrase xiaoli 'ML, 'small rite,' and dali ^IL, 'big rite,' da and xiao are related to the size. In addition, one ox, one sheep and one pig were called one lao $ in early China. According to the chapter "Zhangke of the Zhouli, five lao were served for sun, 'the meal in the afternoon,' and nine lao were served for yongxi, 'the meal in the morning' (^2L^» ^ ^ 7 1 ^ ) . If these two records in the Zhouli are deemed credible, they 1 7 To name a few, see Chen Mengjia (1956: 231-232), Wen Shaofeng and Yuan Tingdong (1983: 71), Chang Yuzhi (1998: 158) and David N. Keightley (2000: 19). It should be pointed out here that these specialists read the phrase as shin#0, which is different from Ken-ichi Takashima's interpretation of this phrase. 39 indicate that the size of the meal in the morning was bigger than that in the afternoon in the Warring States Period. That the meal in the morning was bigger than that in the afternoon in the Warring States Period does not prove it was so in the Yin Dynasty. But it does suggest that the meal in the morning and the meal in the afternoon are possibly called dashi All", 'big meal,' and xiaoshi / MI' , 'small meal,' in the OBI, respectively, and these two designations may have something to do with the size of that meal. Between the dashi and the xiaoshi, to which one does shishi refer? It is Chen Mengjia's opinion (1956: 232) that shiri, which is read as shishi by Ken-ichi Takashima and the present author, could be abbreviations for either dashi or of xiaoshi ( h " A lit" " / J N & " t f ' f f l ^t-^-kZH, WL^M " ). In other words, according to Chen Mengjia, it could refer to either the time in the morning when dashi was served or the time in the afternoon when xiaoshi was eaten. However, all those inscriptions in Examples 8 and 9 show that shishi is the time period between dan, 'dawn to sunrise,' and zhongshi, 'time in the meridian.' As a time division in the OBI, shishi therefore refers to the time of dashi in the morning. Chen Mengjia is incorrect when he says shishi is the abbreviation of either dashi or xiaoshi. In the Shiji and the Huainanzi MW1?, zaoshi #ltf/ is a time division. David N . Keightley (2000: 20) notices that "The Former Han strips from Yinwan WW in northern Jiangsu give the 'early meal' (zao shi ilc[= - ¥ - ] l t t ) and 'late afternoon meal' (bu shi MH) as two of the five periods into which the day was divided." So zaoshi, 'the meal in the morning,' is a time division in early Chinese texts. In the OBI, dashi, 'the meal in the morning,' is used as a time division. The following is one such example: [12] [i]^Mo S ^ J f f l , Kilo A-fr^Mo «i=rlrl» 28618 "On [ren-d&y], it will not rain." "It will not be clear." "On ren-day, [the king] should not hunt because it might rain." 40 "On ren-day, [the king] will hunt then because it will not rain." "At the time of dashi, it will not rain." Heji 28618 The focus of this divination is whether it will rain on a ren-day. It makes good sense to take the last inscription as a charge that divines it will not rain at the time of the big meal in the morning. Because the phrase for the meal in the morning is a time division in early Chinese texts, and, because it also fits the above inscriptional context to interpret dashi as a time division, dashi can be accepted as a time division in the OBI. As shown above, the time division shishi means the time of dashi, and the phrase dashi itself is a time division. Dashi and shishi are two terms referring to the same time division of the Yin day. Is it possible to specify the time of this time division? It is Chen Mengjia's proposal (1656: 233) that the time of dashi was around 8:00 a.m. Wen Shaofeng and Yuan Tingdong (1983: 71) and Song Zhenhao (1985: 330) concur without criticism. However, Chen Mengjia's proposal is problematic. First, there is no evidence for Chen Mengjia's position in the Yin OBI or in early Chinese texts. His theory is merely speculation. Second, the exact duration of time divisions changes with seasons because the starting of daytime changes with seasons. The issue concerning the start of the Yin day will be addressed later in this chapter. For the moment, consider the example of sunrise as an illustration of the duration of a time division that naturally changes with the seasons in the Yin Dynasty. For example, in Anyang, on July 5, 1166 BC, the summer solstice, the sun rose at 4:59 a.m. On December 31, 1166 BC, the winter solstice, the sun rose at 7:39 a.m. The sun rose 2 hours 40 minutes earlier on July 5 than it did on December 31, 1166 BC. Because dashi or shishi is the time division after sunrise, it surely changed, though the meal on July 5 may not exactly be 2 hours 40 minutes earlier than the meal on December 31, 1166 BC. This author had a similar experience. When he lived in Xixiahan villageMJlCll:^, which is located about 15 km west of Anyang city, the time for breakfast and supper in summer indeed was different from that in winter. Because of the foregoing two reasons, especially the second one having to do with the change in the seasons, this author believes Chen Mengjia is incorrect when he seeks to fix dashi at 8:00 a.m. At present, there is no evidence to establish the exact time of the time division shishi or dashi. 41 2.3.1.1.4 Rizhong and Zhongshi Rizhong 0 ff- occurs as a time division in early Chinese texts, a fact that is proven by the following passages: From morning to midday, and from midday to sundown, he did not allow himself time to Jia did not arrive at midday. "The biography of Sima Rangju," Shiji In these two sentences, rizhong refers to the time of midday. The story in Chapter 64 of the Shiji shows rizhong means "exact noon." In his latest comments, Ken-ichi Takashima tells this author that he has written a paper called "Rizhong Fj "T4 and Zhongri 41 0 in Classical Chinese," which this author has not read yet. In this paper, he has "examined many classical texts and...[has] come to the conclusion that rizhong 0 ff-1 is a time division referring to noon, as well as a time duration referring to when the sun is still in its orbit in the sky with wu as its apical point." Those two records certainly support Ken-ichi Takashima's conclusion. In English, noon refers to 12:00 in the middle of the day. It also refers to time duration. In a similar way, as shown above, in Chinese the word rizhong can mean not only 12:00 in the middle of the day, but also a time duration. This author's understanding is that rizhong refers to some time around noon when it is pertaining to time duration. For example, elementary students go to school at 8:30 a.m. Chinese do not say those students go to school at rizhong. Students have a recess at 10:30 a.m. Chinese people also do not say they do that at rizhong. When rizhong is used as a word implying time duration, it may correspond to zhongwu f+p in modern Chinese, which means some time around noon. Whether rizhong refers to noon or some time around noon, it is a time division in early Chinese. 18 Sima Rangju and Zhuang Jia agreed to meet in Rangju's camp at rizhong next day. When Zhuang Jia did not come to meet him at rizhong, Rangju immediately destroyed the water clock. Zhuang Jia finally came, and Rangju killed him as a warning to others. eat. (Legge 1872: 469) 42 Rizhong also is used as a time division in the OBI. Below is an example: [13]*0*W*M<; KS. l&m 29789 It should be at rizhong that there will be a big rain. Perhaps it will rain. Heji 29789 In this example, it makes sense to take rizhong as a time division. As Ken-ichi Takashima points out to this author in his comments, rain is a durative verb referring to a certain length of time when the phenomenon lasts. It sounds very strange to say that it will rain exactly at noon and probably stop after noon is passed. His analysis is correct. In Example 13, rizhong should refer to some time around noon. This is a piece of good inscriptional evidence for the usage of rizhong as a time division. For rizhong as a word of time duration, Ken-ichi Takashima (2003: 5) offers more detailed analysis when he suggests that, in terms of linguistic structure, the relationship between 41 0 and 0 41 is very different: the former is an attributive, determiner noun phrase, and the latter consists of a " N + V " forming a noun phrase (literally "the-sun-being-in-the middle"), translatable to "during the day." He has noticed that while there are a few examples of the 41 0 phrase collocated with the M + (T) + X (where X is a time word such as H "dusk", "late afternoon", and B: "[the sun in decline:] afternoon"), no such collocation is found with the 0 "T4 phrase. It is his belief that the significance of the lack of that kind of collocation "is that the 0 "T4 phrase implies a certain duration of time (daytime, while the sun is still in the sky), whereas the other terms such as H. 'daybreak,' ... etc. refer to a time period shorter than the 0 ^ phrase." Again, this author accepts Ken-ichi Takashima's analysis above, with the exception of two minor points. First, it seems that there is an example of the collocation of the phrase rizhong with ... zhiyu.... The following is the example cited by Chang Yuzhi (1998: 159). [14]jg: 0 4>[M]T^ Mo i^m 13036 Divined: "From rizhong to ze, it will rain." Heji 13036 It is true, as already pointed out several times by Takashima, that this is not a perfect example of the collocation of rizhong with zhi yu, because zhi is supplied by Chang Yuzhi. On the other hand, her transcription is possibly correct. In addition, as cited at the beginning 43 of this section about rizhong, there is an example of the collocation of the phrase rizhong with the time division zhao and ze in the chapter "Wuyi" of the Shangshu. It is said that this chapter was created at the very beginning of the Western Zhou Dynasty, which was very close in time to the Yin Dynasty. If that is the case, it may shed light on the usage of that phrase in the Yin OBI. Given these two considerations, it is not certain that there is absolutely no collocation of the phrase rizhong with ... zhi yu.... Second, Ken-ichi Takashima mentions that other terms such as fi'daybreak,' ... etc. refer to a time period shorter than the 0 41 phrase. It would be ideal i f he were to present a system for comparing durations of time divisions in the Yin OBI. Now, the author would like to turn his attention to the word zhongshi. In the commentary to the Taiyuanzhou ^jLM, there is the following record: Exact noon is called zhong.19 In addition, Ken-ichi Takashima has found the following record in his Siku quanshu: As for the difference between the view before noon and that after noon, could it possibly not [be] affected by [different positions in] the left, right, high and low? Take the exact noon as zhong. Xinfa suan shu, vol. 69 Both records show that zhong means "noon." These records constitute evidence for zhong being used as a time division in Chinese texts. The same usage of zhong can be found in the Yin OBI as well. Below are two examples: [15] fpFJ (=trT)M^^Mo + 0 (=Ftf)M[if]^ [£M]o i^M)) 624 "From zhongshi to yongxi, it will not rain." "From zhongshi to (yong)xi, (it perhaps will rain)." Tunnan 624 [16] #0 (Htf)M*B (Htf)^M. 4>B (=Bt)M^Mo li&m 42 "From shishi to zhongshi, it will not rain." 1 9 Ken-ichi Takashima has made efforts to locate this record, cited by Ruan Yuan (1967: 2), in his Siku quanshu Wftfik+i. But he did not find it. It thus seems impossible to find the original text. 44 "From zhongshi to ze, it will not rain." Tunnan 42 The four inscriptions above divine i f it would rain in a certain time period. It fits such contexts to interpret zhongshi as a time division. Zhongshi is a newly found time division of the Yin. It is Ken-ichi Takashima (2003: 4) who puts forward the theory that the phrase zhongshi is a time division in the OBI. He points out that "the surface ' ^ 0' in the Shang language must have been Hf 'time in the meridian' in the same way as H" 0 and ^Ht 0 stands for 1ltfrT '(time of eating:) meal time' and '(time of big eating:)' breakfast time." Following the criteria for judging a time division in the Shang language, and in light of the fact that zhong is a time division in Chinese texts, and because it makes sense to take zhongshi as a time division of the Yin OBI, this author is compelled by logic and the facts to agree with Ken-ichi Takashima that zhongshi is a time division. Chen Mengjia (1956: 229), Wen Shaofeng and Yuan Tingdong (1983: 73), Song Zhenhao (1985: 307) and Chang Yuzhi (1998: 136-137) all transcribe the phrase zhongshi as zhongri 41 0. A l l of them claim that zhongri is another designation of rizhong 0 "T4. In spite of their consensus, this writer disagrees for the following reasons: First, they have provided no reasons as to why rizhong had to be written as zhongri. Without an explanation, their claims appear weak and arbitrary. Second, there is no occurrence of zhongri as a time division in early Chinese texts. This author's first criterion for determining a time division in the Yin language is the presence of the term in question in early Chinese texts. Because zhongri is absent from the early texts, this writer cannot accept zhongri as a time division of the Yin OBI. Third, both textual and inscriptional evidence supports Ken-ichi Takashima's reading, as shown above. For these three critical reasons, this writer concludes that the time division in the Yin OBI is, indeed, zhongshi rather than zhongri. As a time division, both rizhong and zhongshi can refer to the time of midday or noon. That is the reason this author considered both terms in the same section. 2.3.1.1.5 Ze 45 The word ze appears as a time division frequently in early Chinese texts. Two examples are cited below. From morning to midday, and from midday to sundown, he did not allow himself time to eat. (Legge 1872:469) Ze is the time when the sun is in the west. Shuowen jiezi These two records are clear evidence for the usage of ze as a time division in the early Chinese language. Examples 14 and 16 are evidence for its usage as a time division in the Yin OBI. The following is another example. [17] 4" S (=HT)£Mo OT. 29910 "At the time of the meridian, it will perhaps rain." "The king will make an inspection of the field. At the time of ze, it will not rain." "At the time of ze, it perhaps will rain." Heji 29910 The topic of these three inscriptions is whether it will rain or not at different times of that day. It fits the context to interpret ze as a time division. This is another piece of evidence for ze being a time division in the OBI. Simply based on the shape of the bone graph of ze, Dong Zuobin (1945.2.II: 41) proposes that the time division ze refers to the span of time between 2:00-3:00 p.m. but neglects to provide any rationale. Chen Mengjia (1956: 230) states that ze denotes the time around 2:00 p.m., and his assertion is repeated by Wen Shaofeng and Yuan Tirigdong (1983: 74), Song Zhenhao (1985: 330) and Chang Yuzhi (1998: 138). Chen Mengjia bases his statement on Kong Yingda's sub-commentary to the Chapter "Wuyi" of the Shangshu. However, as demonstrated by Ken-ichi Takashima (2002), Kong Yingda's interpretation of the rong ^ sacrifice lacks credibility because it is not supported by inscriptional evidence. And, Kong Yinda's sub-commentary on the word ze is not supported by contemporary evidence either. It is doubtful that Kong Yingda's comments can serve as valid evidence for the interpretation of materials seen in OBI. In other words, there is no hard evidence for the 46 view that the word ze refers to the time around 2:00 p.m. To what time does the time division ze refer in that case? If it refers to a time period, how long is it? Both textual materials, as cited in the beginning of this section, and inscriptional materials such as Examples 14 and 16, show only that ze is a time division occurring after noon. There is nothing to indicate its duration. 2.3.1.1.6 Yongxi The following inscriptions indicate that the word yongxi i$| ^ 2 0 is used as a time division in the Yin OBI: [08] * s (=fc f )M^^m. * 0 (=fcf)M[«]^[&M]o <i&m 624 "From noon to yongxi, it will not rain." "From noon to \yong\xi, [it perhaps will rain]." Tunnan 624 [ 1 8 ] H [ I ] ^ M „ M S M o l&m 29801 "From noon to \yong\xi, perhaps it will rain." "From yongxi to hun, it will not rain." "From yongxi to hun, perhaps, it will rain." Heji 29801 The two inscriptions in Example 8 divine whether it will rain during the time period from noon to yongxi. The three inscriptions in Example 18 divine i f it will rain from yongxi to hun, 'dusk.' A l l these inscriptions are about whether it will rain during a certain time period. In these five inscriptions, zhongshi, ze, hun are all time divisions. These contexts 2 0 This word was transcribed as guoxi in previous drafts of this dissertation. At the advice of Ken-ichi Takashima, the present writer has read philological studies collected by Yu Xingwu (1996: 1941-1949). Among them, Wang Guowei has pointed out that the graph in question is the same as that of guo % in the Shuowen. But it is also the early graph of yong ijf. Because of this, it seems that the graph can be transcribed either as guo or as yong. However, after comparing the word represented by this graph in the bronze inscriptions and relevant expressions in pre-Qin texts, Wang Guowei's analysis has shown that the pronunciation of the word represented by this graph is similar to the word yongM at that time. It is thus more accurate to transcribe it as, yong. 47 suggest that the word yongxi is a time division in the language of the Yin OBI. In early Chinese texts, however, the word yongxi never occurs, and it is impossible to find an example of yongxi as a time division in classical Chinese. According to the criteria this writer has adopted for determining a time division in the OBI, yongxi cannot be accepted as a time division of the Yin language. On the other hand, Examples 8 and 18 are compelling evidence for the usage of yongxi as a time division. After due consideration, this writer is making an exception to his criteria and is accepting yongxi as a time division in the Yin OBI. This is the only such exception. The word yongxi is sometimes abbreviated as yong.21 Guo Moruo (1965: 538),22 Dong Zuobin (1945.1.1: 6b-7a) and Chen Mengjia (1956: 231) argue that it is abbreviated as xi, too. However, as Chang Yuzhi's analysis (1998:138, note 1) shows, their transcriptions and interpretations of these three relevant inscriptions are incorrect. As a time division, to what time period does yongxi refer? It is the opinion of Guo Moruo (1965: 538) and Dong Zuobin (1945.1.1: 6b-7a) that it refers to early morning. However, these inscriptions in Examples 8 and 18 show that it is a time division between the time division ze and hun, i.e., sometime in the afternoon. Guo Moruo and Don Zuobin are shown to be mistaken, as is already suggested by Chen Mengjia (1956: 231). Chen Mengjia (1956: 231) points to yongxi as a time division for a time period in the afternoon. But he creates a problem when he specifies yongxi as the time period around 4:00 p.m., a view also taken by Wen Shaofeng and Yuan Tingdong (1983: 74). Song Zhenhao (1985: 330) fixes the time of yongxi at 2:00 p.m. As shown in previous sections, the exact time of time divisions of the Yin day changes with seasons. For this reason, the present writer does not accept Chen Mengjia's and Song Zhenhao's specifications for yongxi. 2.3.1.1.7 Mu 21 Heji 30203 is a good example: At the time yong of this^ '-day, it became clear and did not rain. In this inscription, it makes good sense to take the word yong as a time division. It seems reasonable to assume it is the abbreviation of the word yongxi. 2 2 Guo Moruo's Yinqi cui bian S ^ ^ t i was first published in Japan in 1937. In 1965, it was reprinted, which is the edition the present writer has been using. 48 There are many pieces of evidence to support the word mu M- as a time division in early Chinese texts. Below are two examples. Mu refers to the time when the day is about to darken. Shuowen jiezi ^ B#i£So w e • m^wnm At the time of dusk, I have a long way to go. "The biography of Wu Zixu," Shiji Similar usage of mu can be found in the Yin OBI, as shown by the following inscriptions. [ i 9 ] 5 £ B £ i j 2 3 , $~mto '%&Wto i^m 27401 As for the gui-cvX sacrifice to Father Ji, it should be at the time of mu that [the king will] perform the ^ow-cutting sacrifice. It should be at the time of xi that [the king will] perform the jyow-cutting sacrifice. Heji 27401 [20] n24, *mmo 30845 In praying, it should be at the time of mu that [the king will] perform the j>ow-cutting sacrifice. At the time of xi, [the king will] perform the ^ow-cutting sacrifice. Heji 30845 The topic of Examples 19 and 20 is to determine the appropriate time for performing the yow-cutting sacrifice between mu and xi ty, the time division for night-time in the Yin day. It fits these contexts to interpret mu as a time division. These two examples show that the Yin people needed to select a period between the time division mu and the time division xi. These two examples also show that the Yin people divined the suitability of mu before divining that of xi. These two facts indicate that these two time divisions should be mutually exclusive. As for the time period of mu, it is an issue that cannot be decided by inscriptional evidence yet. Since the Shuowen defines mu as the time when the day is about to darken, the tentative proposal is that it may refer to the time around sunset. 2 3 For detailed analysis of the word gui see Ken-ichi Takashima (1996a: 418-425). 2 4 For the rationale to transcribe the bone graph in question to dao, see Ji Xiaojun (1991). 49 2.3.1.1.8 Hun Hun If is a time division in classical Chinese. For instance, its definition in the Shuowen jiezi is ri ming ye BHiil (the day is dark). In addition, such usage of the word hun is supported by the following commentaries collected by Ruan Yuan (1967: 196): The time of 37 lA minutes 2 5 after sunset is called hun. B A B f t H j ^ t By water clock, three-quarters of an hour after sunset is called hun. According to these two records, the word hun refers to the time period of about three-quarters of an hour after sunset. These records represent evidence suggesting that hun is indeed a time division in early Chinese texts. There are inscriptions in which the word hun appears to be a time division. For example: [21]it^MtTWo J i ^ s t ^M- i^m 29801 "From yongxi to hun, it will not rain." "From yongxi to hun, perhaps, it will rain." Heji 29801 [ 2 2 ] J f ^ S t T M . «1=r*» 29794 "From yongxi to hun, perhaps, it will rain." Heji 29794 f-^-^mo 30838 It should be at this hun that [the king will] perform the .vow-cutting sacrifice. Upon this xi, [the king will] perform the.yow-cutting sacrifice. Heji 30838 These inscriptions in Examples 21 and 22 divine i f it will rain in the time period from yongxi to hun. The two inscriptions in Example 23 divine the time for performing the vow-cutting sacrifice. In these contexts, it does make sense to interpret hun as a time division. 2 5 Chen Meidong (2003: 130-135) has described in detail how the Chinese used ke M to measure time in the Qin and Han Dynasties. At that time, one day was divided into one hundred ke. The exact time of one ke is 14.4 minutes long, which is very close to the time of a quarter-hour in English. 50 As for the exact time of hun, Chang Yuzhi's opinion (1998: 139) is that hun refers to sunset only (zhi ri luoxia, ji riru zhi shi I H S ^ T . BP 5 A^LBT). Her opinion is not in agreement with those textual records cited at the beginning of this section. Chen Mengjia (1956: 230) and Wen Shaofeng and Yuan Tingdong (1983: 75) have reached the same conclusion that hun refers to a time period when it becomes dark, starting from sunset. Chen Mengjia (1956: 233) further specifies that hun denotes some time around 6:00 p.m. Again, such a specification is problematic because the time of sunset changes with seasons. For instance, in Anyang, the sun set at 7:38 p.m. on July 5, 1166 BC. According to Chen Menjia, hun started at 7:38 p.m. on that day. Clearly, this is about 1 hour 40 minutes later than 6:00 p.m. In the corpus of the Yin OBI, there is no inscriptional evidence to show the duration of this time division hun. Considering those textual records cited at the beginning of this section, the writer tentatively accepts Chen Mengjia's view (1956: 230), which is that it refers to a time period when it becomes dark, starting from sunset. 2.3.1.1.9 Zhou, Du, Zhi, and Zhu: Pseudo Time Divisions It is Song Zhenhao who first proposed zhou H , du H , zhi and zhu ft as time divisions of the Yin day. However, an examination of relevant inscriptions shows that these words are not time divisions in the OBI. 2.3.1.1.9.1 Zhou 13 The bone graph for the alleged time division zhou H is written as K>H. Song Zhenhao (1991: 40) suggests that zhou originally meant to determine time by erecting a wooden pole and measuring its shadow under the sun; and that at a later time zhou is used exclusively as a time word ( m ^ X ^ A L ^ M ^ B B U ^ IU, B^X^H W^BT). Chang Yuzhi (1998: 149) accepts Song Zhenhao's conclusion. Even so, there are some problems with Song's explanation of the graph zhou. First of all, the so-called original meaning of that word derives merely from Song 51 Zhenhao's speculation. He does not produce any textual or inscriptional evidence to justify his claim that the word zhou does, indeed, refer to a method of determining time. Second, Song Zhenhao interprets the hand-held object described in that graph as a straight wooden pole. But the object is actually more like a brush. This might be the reason why the authors of the Tunnan transcribe this bone graph as yuri ^ 0 rather than zhou. Third, Song Zhenhao mentions that the purpose of erecting a straight pole is to measure its shadow. However, there is nothing in the graph that can be considered as the shadow of that pole. Fourth, Song Zhenhao and Chang Yuzhi cite only one example of the word zhou being used as a time division in the OBI. The fact is that this example does not provide a compelling inscriptional context to show that this word is a time division. Below is their example. [24] h : H Mo i^M)) 2392 On jiawu (day 31), [X] divined: "Chang (aromatic wine) Two yous [a word referring to a unit of measure]. It should be oxen. Specially reared oxen. Today. Zhou. Tunnan 2392 On this piece of oracle bone, that bone graph occurs alone. This certainly is not good evidence for the usage of the alleged word zhou as a time division in the OBI. 2 6 Anne O. Yue mentions the occurrence of jinzhou 4" ft on Heji 22942. This is cited by Chang Yuzhi (1998: 169) as well. It is my mistake to say Chang Yuzhi has cited one example only. The rubbing of Heji 22942 shows that the inscription is not complete. More work needs to be done in order to understand the meaning of jinzhou on Heji 22942. 27 Chang B should be the liquid made from grains. The translation "aromatic wine" is that of Ken-ichi Takashima (1996a: 209). 52 Given these four reasons, this writer cannot accept the word zhou as a time division in the language of the Yin OBI. 2.3.1.1.9.2 Du The bone graph interpreted as du W by Song Zhenhao (1991: 34-35) is scribed as tS3M. According to him, like the graph of the so-called zhou, this graph also depicts the process of determining time by measuring the shadow of an erected wooden pole. He further speculates that this procedure is always done at noon, and that du is thus used as a time word to represent noon. As in the case of his interpretation of zhou, his explication of du is equally difficult to accept. First, his interpretation of the graph in question is mere speculation. There is no inscriptional or textual evidence to support his interpretation. Second, his interpretation of the graph cannot stand scrutiny. If this graph indeed depicts the process of determining time by measuring the shadow of an erected wooden pole, as Song Zhenhao claims, the pole must be as straight as possible so that its shadow can be measured accurately. But there is nothing that represents a straight wooden pole in the bone graph. In addition, in order to be able measure the shadow of a pole, the component sun must appear at the top rather than at the bottom of the graph. The so-called original meaning of this graph proposed by Song Zhenhao does not stand up to scrutiny. Third, since the purpose of measuring the length of a pole under sunshine is to determine time, it follows that people do not know the time when they do that. If so, there is no basis for Song Zhenhao to assert that this procedure is done at noon only. My analysis above shows that Song Zhenhao's interpretation of the graph H is incorrect, which casts serious doubt on its usage as a time division. More importantly, Song Zhenhao fails to provide valid evidence to support his opinion. In his article, he cites Heji 30893, 30894, 30599, and 30365 as his evidence. The reason he takes them as evidence is because this graph appears between words ^ and S .^ However, as shown in Section 2.2.2, the occurrence of a word between ^ and is not a criterion forjudging a time word in the Yin 53 language. Therefore, none of those four examples lends any support to his opinion. Chang Yuzhi (1998: 148) cites the following example to support the alleged word du as a time division in the OBI: [ 2 5 ] i : # [ @ ^ ] o tyWf° 30844 It should be at du that [the king will] perform [the.you-cutting sacrifice]. [It should be at] night that [the king will] perform the _yow-cutting sacrifice. Heji 30844 At first glance, the topic of these two inscriptions in this example is the appropriate time for performing the j/ow-cutting sacrifice. However, Chang Yuzhi's transcription is not accurate. The bone graph she transcribes as du is severely damaged. It is far from certain that it actually is the graph du. As a matter of fact, the authors of the Jiaguwen heji shiwen transcribe this graph as zhong l T J , 'middle.' Because of the uncertainty about the bone graph, it can hardly be considered a piece of evidence for the alleged time division du. Because Song Zhenhao's interpretation of the graph § is mistaken, because there is no textual or inscriptional evidence to show the word represented by this graph has ever been used as a time division, the so-called time division du is unacceptable. 2.3.1.1.9.3 Zhi I Tang Lan (1939: 46) interprets the bone graph Ifil as an early form of the character zhi tA, 'to seize.' As for the word represented by this graph, Tang Lan interprets it as re 'hot.' He further asserts that this word is used as a time division referring to the time of shangdeng Az'U, 'to light the lamp.' Tang Lan's opinion is followed by Song Zhenhao (1985: 307) and Chang Yuzhi (1998: 142-145); but they have not proven zhi is a time division yet. There are three reasons for the shortcoming in their reasoning. First, in early Chinese texts, there is no evidence to show the word zhi or re has ever been used as a time division. Second, there is no compelling inscriptional evidence to support the usage of zhi as a 54 time division. Tang Lan (1939: 46) bases his stance upon the pair of phrases zhiru #lA and xiru A , whose meaning will be explained shortly. But they do not conclusively show the word zhi is a division of time. Two inscriptions that contain the pair of phrases zhiru and xiru are cited below. [26] i £ # L A , T M * I ^ A f j J b TMc i&m 30094 + l&m 30113 2 8 The king might zhi ru, because he will not encounter rain. The king will xi ru at Zhi, because it will not rain. Heji 30094 + Heji 30113 In the first inscription, the word qi ^ occurs before the word zhi #U Ken-ichi Takashima devotes considerable time and effort to understanding the word qi. His conclusion is that it functions as a modal particle in the language of the O B I . 2 9 Qi can be construed as modifying a verb, and in this example, qi modifies zhi. So zhi should be interpreted as a verb rather than a noun. Since zhi is not a noun, it certainly cannot be a time division. Besides the first inscription in Example 26, there are other inscriptions where the word zhi occurs as a verb. The following two inscriptions are apt examples: [27] 20#U « ^ » 28939 [The king] will zhi. [The king] should not zhi. Heji 28939 These two inscriptions are a pair of duizhen M$l, 'a set of complementary charges.' In the first inscription, the word zhi is again modified by the modal particle qi, which shows that zhi is a verb. In the second inscription, the word wu ty) appears before zhi. Ken-ichi Takashima has done comprehensive studies on the negatives in the language of the OBI . 3 0 Wu is one of those *m-type negatives "that negate verbs whose salient feature is their 'controllability' - verbs which are thought of as being controllable by the will of living human beings" (1996a: 370). Clearly, zhi in the second inscription should be interpreted as a Chang Yuzhi (1998: 144) has rejoined these two pieces of oracle bone. 2 9 For examples of qi being treated as a modal particle, see Ken-ichi Takashima (1996a: 191-192). 3 0 For his detailed research on negatives in the OBI, readers are referred to Ken-ichi Takashima (1973, 1988 and 1996a: 364-383). 55 verb as well. As for the word represented by the graph that is transcribed as xi above, it can be used as a verb, which is shown in Example 28. In this example, that graph represents the word yue M, 'to amputate, to cut off a limb or other part of an animal body.' 3 1 [28] H W h S ¥ f X J ! i J + ^ o im 404 Crack-making on guiyou (day 10), Zheng divined: "On next day jiaxu (day 11), [the king] will amputate ten sheep." Yi 404 In this inscription, the word represented by the graph in question is the best candidate for a verb in this charge. Since sheep are often used in sacrifices, it makes the best sense to interpret the word represented by the graph as the verb .y we M. Now, let's return to Example 26. As preceding analysis here has shown, since the modal particle qi appears before the word zhi, the preference should be to construe zhi as a verb. In addition, xi could be an incorrect transcription; it should more judiciously be transcribed as yue M. According to such an understanding, there would not be a single time division in Example 26. This example does not support Tang Lan's opinion (1939: 46) that the word zhi is a time division in the Yin OBI. Song Zhenhao (1985: 307) and Chang Yuzhi (1998: 142-143) further cite the following inscriptions as evidence for zhi as a time division in the OBI. [29##lS^o 27052 *It should be at the time of zhi that [the king will] perform the j/ow-cutting sacrifice. Heji 27052 The translation above is based upon their understanding of the word zhi. The reason for them to regard this inscription as evidence for zhi being a time division is that the word occurs between ^ and As analyzed earlier in Section 2.2.2, such an argument is untenable. [30] i&m 30746 [The king] will zhi and perform the .vow-cutting sacrifice. Heji 30746 This is cited by Chang Yuzhi (1998: 142). Since the word zhi occurs after the modal 3 1 Ken-ichi Takashima (2002: 366) is the most recent study on the verb represented by the bone graph that has often been transcribed as xi. 56 particle qi, it should be a verb, as analyzed in Example 26. [ 3 1 ] i f i f f l , # l A , ^ f f i o i^m 28571 The king w i l l hunt. [When he returns, he will] enter from zhi [because] it w i l l not rain [there]. Heji 28571 [ 3 2 ] i ^ T f f l , ft A, ffio < ( ^ » 29003 The king w i l l inspect fields. [When he returns, he will] enter from zhi [because] it w i l l not rain [there]. Heji 29003 These examples are also cited by Chang Yuzhi (1998: 143) as evidence for the word zhi being used as a time division. But Zhang Bingquan (1959: 223) has already pointed out that zhi is a place name, which is supported by the following inscription. [33] T # l A , t r ^ o i^Ml 28984 [When the king] enters from zhi, there wi l l be no disasters. Heji 28984 Here, it makes sense to take zhi as the complement of yu and interpret it as a place name. If this is the case, Examples 31 and 32 are not good evidence for the usage of the word zhi as a time division. In short, there is no inscriptional evidence to support the word zhi as a time division. There is no occurrence o f zhi being used as a time word in early Chinese texts. Therefore, the word zhi cannot be accepted as a time division in the language o f Y i n O B I . 2.3.1.1.9.4 Zhu Song Zhenhao (1987:20) transcribes the bone graph HBH as zhu ft, 'to live, ' and interprets it as a time division referring to the time period from 21:00 to 23:00, which is accepted by Chang Yuzh i (1998: 145). They acknowledge that there is only one inscription in which zhu occurs as a time division. That inscription is as follows. [ 3 4 ] * f i : i ^ o <(1>&» 27522 *It should be at the time of zhu that [the king will] perform the yow-cutting sacrifice. Heji 27522 Again, the reason for them to interpret zhu as a time division in this inscription is 57 because it appears between ^ and M^. As shown earlier in Section 2.2.2, one cannot establish a time division simply based upon the occurrence of a word between ^ and Wl>. Strictly speaking, Example 34 is not evidence for the usage of the word zhu as a time division. Because the word zhu appears only once in the OBI, and because it is not certain whether this word is a time division in that inscription, it is safe to say that there is no example of the word zhu being used as a time division in the language of the Yin OBI. 2.3.1.2 Sub-divisions of Ri in Period I Some time divisions of ri in the other four periods, including Period I, are the same as those in Period III. In such cases, this writer will only analyze inscriptions of those periods in which those time divisions appear. For those inscriptions not found in Period III, the writer will cite both textual and inscriptional evidence to prove they are indeed time divisions in the Yin OBI. 2.3.1.2.1 Ming The following is an example of the word ming Bfj as a time division in early Chinese. At the time of ming, the king arrived at the Zhou temple. For its usage as a time division in the Yin OBI, the following is one of the most often cited examples: 32 Fa \% here refers to how sacrificial victims are handled when the yoM-cutting sacrifice is offered, which is clearer in the following inscription: M * h . S&fis teEirz,. 3 E £ H : um m^m. swafc Z.EH2, mm. &mm. «itr»» 11497 Crack-making on bingshen (day 33), Nan divined: "On the coming yisi (day 42), [the king will] perform the j/oH-cutting sacrifice to Xiayi." The king, having prognosticated [the omens], declared: "The performance of the yoH-cutting sacrifice means that there will be a curse [from Xiayi], and there will be thunder." On the yisi day, [the king] performed the you-cutting sacrifice, and in the morning it rained. [When he] dismembered victims, it was already raining. Heji 11497 The translation above is basically that of Ken-ichi Takashima (1996a: 195). In this example, fa clearly is a verb. 58 40341 3 4 On bingshen (day 33), [X] divined: "On the next day dingyou (day 34), [when the king] performs the ^ow-cutting sacrifice and dismembers victims, the weather will be clear." On ding\you], at the time of ming, it was foggy. At the time of the big meal, the day became clear. [Day bingshen was in] the first month. Heji 40341 In the verification of this inscription, both the word ming and the phrase dashishi occur. The translation above shows that it makes good sense to interpret both of them as time divisions. Below is another example of ming as a time division. [ 3 6 ] ^ K j£: fjjo ^B^Mo TMff i* $/h3£B (=Bt) M±, J*L S 0 J Bo • i^M)) 21016 Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, [there will be no disasters]." [Day guihai was in] the second month. At the night of yichou (day 2), it rained. At the time of ming on dingmao (day 4), it rained. At the time of xiaocai on wu[chen] the rain stopped; the wind [started]. At the time of ming on ji[si] (day 6), it became clear. Heji 21016 \ In this inscription, the word ming specifies the time when it rained on day dingmao and when it became clear on jisi. This is another piece of evidence for ming as a time division. In Period I, the word ming is a time division. In Period III, dan is a time division. It is accepted by specialists in the field that both dan and ming refer approximately to the same time period of the day, for two reasons. First, as cited in Section 2.3.1.1.2, the Shuowen defines dan as ming (i=L, fJj 4J3,). Second, it seems that both dan and ming are the time period immediately preceding dashi, 'the time of big meal.' In the verification of Example 36, ming and dashi occur as two time divisions of dingyou. Since ming appears before dashi, it should be earlier than dashi. Also, Examples 8 and 9 of Period III show that dan occurs before shishi. It is therefore reasonable to take ming and dan as two expressions of the same time period of the Yin day in the OBI. For studies on this bone graph, see Guo Moruo (1937: 560) and Yu Xingwu (1979: 107-111). 34 Heji 13450 and Heji 40341 are tongwen buci |H]3t hi§?', a set of inscriptions of the same event. For details on tongwen buci, see Hu Houxuan (1947). The difference between these two inscriptions is that the four characters riqiyiyue El/p —'Jf are completely missing on Heji 13450. Because these two inscriptions are records of the same event, the four lost characters riqiyiyue on Heji 13450 can be supplied based upon their appearance on Heji 40341. 59 2.3.1.2.2 Dashi, Dashishi and Shishi On Heji 20961, there is one example of dashi A l l " , 'big meal,' as a time division. [37]F>Mh = 0 1 = TJCA-frMo ( ( ^ ) ) 20961 On bingxu (day 23) divined: "In three days, it will rain." At the time of dashi on dinghai (day 24), it rained. Heji 20961 In the verification of this inscription, dashi, 'big meal,' is the time when it rained. In some cases, the word shi is added to dashi, as has already been seen in Example 35. In a few cases, dashi shi A'ftB'f is abbreviated to shishi llfftif. There is one such example on Heji 11506. The inscription reads as follows. (=Bt)AM 3 5o <(^H» 11506 Crack-making on jiayin (day 51), Nan divined: "The next day yimao (day 52) will be sunny." The king, having prognosticated the cracks, declared: "the next day yi, don't let it rain." On day yimao, at the time of ming, it was indeed foggy... at the time of [big] meal it became very sunny. Heji 11506 In this inscription, the time period referred to by dashi is written as shishi. What happened at the time of big meal was that it became very sunny. 2.3.1.2.3 Dacai and DacaiShi It was Dong Zuobin (1945.1.II: 5) who first proposed that the expression dacai i^M: in the chapter "Lu yu #ip-" of the Guoyu Hip- is a time division. £tt A T 0 . Therefore, the Son of Heaven at the time of dacai pays tribute to the sun. It seems that Dong Zuobin's interpretation of dacai in the Guoyu makes good sense. According to the chapter "Yao dian" of the Shangshu, Yao ordered X i Zhong H 3 5 For the usage of the word xingM. as becoming sunny in the OBI, readers are referred to Yang Shuda (1954.1: 20-21) and Li Xueqin(1981). 60 "respectfully to receive as a guest the rising sun" (Legge 1872: 18). It is reasonable to say that early Chinese had the thought that the Son of Heaven paid tribute to the sun at the time of dacai in the morning. Based upon this record, Dong Zuobin has suggested that the expression dacai in the OBI is a time division. The following are two examples. [39]Z^PK ^ B i i i f t o ZBA^m, i T f t . i&m 12814 Crack-making on yimao (day 52), Nan divined: "Today the king will go to Dun." This day it rained at the time of dacai, and the king did not go. Heji 12814 [40]MTh *5RMS:lfco 20960 On bingwu (day 43) [X] divined: "Today, perhaps, it will rain." At the time of dacai, it rained from the north. Heji 20960 The translation of verifications of Examples 39 and 40 shows that it fits those contexts to take dacai as a time division in the language of the Yin OBI. In some cases, the word shi is added to dacai. Two examples are cited below. [41]/a o KWftftB* {^m 20993 "[It will be] clear." At the time of dacai, it indeed became clear. Heji 20993 [42]H£h J£: tjo -j?o I l g l o 7 L 0 ^ * ^ # ^ H 4 b c -z fc^fcf&Sgzlb- <(^» 21021 Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, [there will be no disasters." Day guihai was in] the first month. At the time of ze, it rained from the east. At the time of dacai of the ninth day xinwei (day 9), clouds came from the north. ... at the time of dacai, clouds came from the north. Heji 21021 Chang Yuzhi (1998: 156) cites Heji 11727 as another example of the phrase dacai shi. Her transcription reads as follows. [43] T f f i T [ X M B ... divined: ". . . to ancestor.. .father." Not... the time of [da]cai... There are three mistakes in her transcription. First, the bone graph she transcribes as fu 5£ clearly is chou T i , as transcribed by the authors of the Jiaguwen heji shiwen. Second, the graph for cai 3^  itself is not complete. It is not certain that the word is indeed cai. Third, because the ,word immediately above the word cai is missing, it is impossible to know 61 whether that word is da A. Therefore, it is not certain that there is the phrase dacai on this oracle bone; and it is risky to cite Heji 11727 as an example of the occurrence of dacai shi in the OBI. As for the exact time referred to by dacai, Dong Zuobin (1945.1.II: 5) and Chang Yuzhi (1998: 156-157, 163-164) suggest that it is some time in the morning. Chen Mengjia (1956: 232), Wen Shaofeng and Yuan Tingdong (1983: 71), and Song Zhenhao (1985: 330) all agree that it refers to the time around 8:00 a.m. Judging from the record in the Guoyu, which is cited at the beginning of this section, it is reasonable to infer that the time division dacai refers to some time in the morning. If dacai is related to the activity of paying tribute to the rising sun, as suggested by the passage in the Guoyu, it is not accurate to fix dacai at 8:00 a.m., because the time of sunrise changes with seasons. For specific times of sunrise at Anyang, readers are referred to Section 2.3.1.1.2. 2.3.1.2.4 Zhongshi The following is an examples of zhongshi ^  as a time division in Period I. [ 4 4 ] P j £ h j£: t B ( = B f ) f f i . (£M)) 11775 Crack-making onX-xu day, [X] divined: "At the time of meridian, it will not rain." Heji 11775 2.3.1.2.5 Ze and Zeshi On Heji 20967, there is an example of the word ze as a time division in Period I. [45] Z , f i h i l l o J c M M , <b. i ^ m 20967 On yichou (day 2), [X] divined: "On yichou, it will rain." At the time of ze, it rained from the north, [and the rain was] light. Heji 20967 Chang Yuzhi (1998: 161) has provided an example of the word shi being added to the word ze. The phrase zeshi occurs on the back of Heji 11728. [46] ^ 0 (=Bt) ... at the time ofze ... 62 It is apparent that the context of this inscription is not complete. Because it is incomplete, there is some uncertainty as to how to understand this inscription. In any event, it is possible to read zeshi as 'at the time of ze.' 2.3.1.2.6 Xiaocai and Xiaocai Shi Like dacai, the textual evidence for xiaocai '\s9fc as a time division is also seen in the chapter "Luyu" of the Guoyu: M o [The Son of Heaven] at the time of shaocai pays tribute to the moon. In early Chinese, the word shao and xiao <h are interchangeable, which is common knowledge. Therefore, the phrase shaocai in early Chinese could appear as xiaocai. In the sentence above, it does make sense to take shaocai as a time division. It is an example of shaocai as a time division in early Chinese texts. There are several inscriptions in Period I in which the word xiaocai occurs as a time division. Below is an example: [ 4 7 ] § | E h 3H: tio H0p1$^Mi&, '«g5E36. 20966 Crack-making on guisi (day 30), the king [divined]: "[In the next] 10-day week, [there will be no disasters.]" At the time of ze of the fourth day bingshen (day 33), it rained from the east; at the time oi xiaocai, the rain stopped. Heji 20966 In the verification of this inscription, xiaocai specifies the time when the rain stopped, which is a good reason to take it as a time division. In addition, this inscription shows that xiaocai occurs after ze. It can be inferred from this fact that xiaocai is a time division in the afternoon. Sometimes the word shi is added to xiaocai, as shown in the following inscription: [36] H ^ h , fjfo ZLBO ZaJi^M. TFpeJMo jft'HfcB (=Bt) Mih, M. B 3 6 The word jiIS has several meanings in classical Chinese. When it is used as a verb, it means 'to complete.' One example of this usage can be found in the chapter "Ying diwang jSiff 3Z" of the Zhuangzi 3 :7 , which reads as wu yu ru ji qi wen ?r -^j&SERi 'I and you have completed its pattern.' In this inscription, the word ji is related to rain. It fits the context here to interpret ji as 'to complete.' When the rain is completed, it does not rain anymore. That is the reason why the present writer translates ji as "stop." 63 mBo 2 1 0 1 6 Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, [there will be no disasters]." [Day guihai was in] the second month. At the night of yichou (day 2), it rained. At the time of ming on dingmao (day 4), it rained. At the time of xiaocai on wu[chen] the rain stopped; the wind [started]. At the time of ming on ji[si], (day 6), it became clear. Heji 21016 2.3.1.2.7 Xiaoshi In early Chinese texts, the phrase mushi §ltf, Tate afternoon meal,' is a time division. One piece of evidence is seen in the chapter "Tianguan shu A W ^ " of the Shiji. It reads as follows: [If Venus] appears at the time of late afternoon meal, [it] is small and dim. David N . Keightley (2000: 20) has cited the Former Han strips from Yin wan WtM to show "late afternoon meal" {hu shi MH) was one of the five time divisions. Bu shi is written as bu shi fif in the chapter "Changyi wang zhuan H S i f t " of the Hanshu. Thus: Bf&HTS£r%o At the time of late afternoon meal, it arrived at Dingtao. As mentioned in Section 2.3.1.1.3, only two meals were served in early China. Since both mushi and bushi refer to the late afternoon meal, they should be designations of the same time period. These records show that late afternoon meal is a time division in early Chinese. In the Yin OBI, xiaoshi A^^, 'small meal,' is the term for late afternoon meal. It also occurs as a time division, which is shown by the following inscription: [48] ^ J i h, j£: 10. ¥^A#SH4bc ZsW'httB* Mil^FJ (=fcf)*MS Pffc 21021 Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, [there will be no disasters]." At the time of big meal on jiayin (day 51), it rained from the north. At the time of small meal on yimao (day 52), it became very clear. At the time of meridian on bingchen (day 53), it rained heavily from the south. Heji 21021 64 In this inscription, xiaoshi should be taken as a time division. It specifies the time when it became clear on day yimao. Chen Mengjia (1956: 232), Wen Shaofeng and Yuan Tingdong (1983: 74), Song Zhenhao (1985: 330) and Chang Yuzhi (1998: 158-159) all agree that xiaoshi is the meal served in late afternoon. David N . Keightley (2000: 19) puts xiaoshi in the morning. But a personal communication of November 21, 2002 indicates that he has changed his view and accepts xiaoshi as a late afternoon meal. 2.3.1.2.8 Shuang and Xiang: Pseudo Time Divisions 2.3.1.2.8.1 Shuang Qiu Xigui (1992: 89) suggests that the word represented by the bone graph WM or ut**. has possibly a semantic element yue M, 'the moon,' and a phonetic element sang 'to lose.' He further proposes that this word can possibly be read as shuang M of meishuang 0jR His notion is followed by Chang Yuzhi (1998: 155-156) and Huang Tianshu (2001: 91-92). The difference between Qiu Xigui (1992), on the one hand, and Chang Yuzhi (1998) and Huang Tianshu (2001), on the other hand, is that Chang Yuzhi and Huang Tianshu are more assertive. For the present, shuang must be rejected as an alleged time division. There are three reasons for this refusal. First, it is clear in Qiu Xigui's own words that he is not sure i f that bone graph does indeed represent the word shuang. If so, it is mere speculation to say it is the time division meishuang in early Chinese texts. Second, even i f the bone graph can be read as shuang, it is a leap, as it were, to equate it with meishuang. These researchers have not provided any evidence to show how the word represented by this bone graph allegedly evolved to meishuang. Third, inscriptions on Heji 13751 and 13752 show that the word represented by this bone graph might actually be sang iJE, 'to lose.' Turning to the inscription on Heji 13752 here: 65 [ 49 ] j g : 3/f£W&. i r j j E i : im^m, mm-, 37-^w-s^*Effla. [X] Divined: "Jin might fall i l l . " The king read the cracks and declared, "Jin might fall ill. It will be on a bing-day; if not, [it will be on] a geng-day [when he falls ill]." On the twenty-first day gengshen (day 57), Jin lost his sight.38 Qiu Xigui, Chang Yuzhi and Huang Tianshu all take in this inscription as a time division. However, they do not explain how they understand the meaning of this inscription. If it is indeed a time division, the meaning of the verification would not be as certain as the translation above shows. Further, the relationship between the charge and the verification would not be as clear as the translation above indicates. It appears that it is problematic to interpret as a time division. Given the considerations above, up to this point, the so-called shuang cannot be accepted as a time division of the Yin OBI. 2.3.1.2.8.2 Xiang It is Chen Mengjia's opinion (1956: 232) that the word xiangM consists of two components: yang^-, 'sheep,' on the top and mu 'eye,' at the bottom. He suspects that xiang is a time division that is equivalent to noon. His opinion is followed by Wen Shaofeng and Yuan Tingdong (1983: 73). But Song Zhenhao (1985: 303) seriously doubts Chen Mengjia's conclusion. One example cited by Chen Mengjia is the following inscription on Heji 20397. [ 5 O ] ^ 0 / J N ^ A M , tU&B* The punctuation of this verification is proposed by Ken-ichi Takashima. He explains the rationale behind such punctuation, pointing out that "many specialists...take bu hui as a unit, with the bu negating hui. This, it is contended here, is a mistake. There should be a syntactic break after bu, and the main clause should begin with hui" (1996a: 482). As the translation above shows, his punctuation makes good sense. 3 8 This inscription clearly is a medical record. The king prognosticated that Jin might be sick. The result seems unfortunate: on gengshen (day 57), Jin lost his sight. Since one component of actually is sang ife, it makes perfect sense to interpret the word represented by this graph as sang fe, 'to lose.' It also fits the context of this inscription to read the graph as ming 'Uij, 'sight.' 66 At the time of xiaocai of today, it indeed rained heavily, *at the time of xiang, it should be clear. Heji 20397 There are two problems to considering the word xiang in this example as a time division. First, the position of the word xiang seems suspicious. This writer has never seen other time divisions appearing in sequence before the word wei. The occurrence of xiang before the word wei casts doubt on the proposition that xiang is used as a time division. Second, as noticed by Song Zhenhao (1985: 303), Example 50 shows xiangri occurs after xiaocai. Because of this, if xiang is a time division, it should be a time division of the afternoon. It is impossible for the word xiang to refer to some time period at noon. Chen Mengjia's interpretation of the relevant inscription appears incorrect. It follows that it is difficult to accept his opinion that xiang is a time division in the Yin OBI. 2.3.1.3 Sub-divisions of Ri in Period II 2.3.1.3.1 Chen Chang Zhengguang (1982: 141-46) interprets the bone graph HH as chen /jft, 'early morning,' an interpretation followed by Wen Shaofeng and Yuan Tingdong (1983: 68-69), Yao Xiaosui (Yu 1996: 1138) and Chang Yuzhi (1998: 155, 166-169). Because the component chen Jjf of this graph should be its phonetic element, just as with the element chen M of the character chen J R , the graph and the character chen J S likely represent the same word. In early Chinese texts, the word chen jS is a time division. Cited below are two examples: Towards day-break of Bing, Wei of the Dragon lies hid in the conjunction of the sun and moon. (Legge 1872b: 146) Chen means early, [referring to the same time period denoted by the time division] meishuang %M.. 67 These two records show the word chen is a time division about the time period around daybreak in early Chinese texts. This usage of the word chen fits inscriptional contexts as well. The inscription on Heji 23226 is a good example: [51][T]EK Mft: $LT$h f £ / « o Crack-making on [ding]si (day 54), Lu divined: "In offering the gwz'-cutting sacrifice to Father Ding, it should be at daybreak that [the king performs] the .yow-cutting sacrifice." Here, the word chen ft specifies the time when the >>0w-cutting will be performed. This is one piece of evidence for the word chen as a time division in Period II. As for the time period referred to by the time division chen, it can be inferred from the following passage in the Zuozhuan: ¥ ^ W f , i I J i f ¥ I P $ S M r K o ( ( £ ^ - M + A ^ ) ) Jiawu (day 31) was the last day of that month. At the time of chen, Chu came close to the army of Jin and deployed.... At the time of dan, [they] fought. It is apparent that chen is the time division before dan. It should refer to the same time period as su. 2.3.1.3.2 Zhao The word zhao %j\ is a time division in early Chinese texts, which is clearly shown by the following records. From morning to midday, and from midday to sundown,39 he did not allow himself time 3 9 In his comments to this dissertation, Ken-ichi Takashima writes: "I think his translation is wrong. Rizhong ze F4 *¥Bk, if quoted correctly, should be an NP which has the finer structure of NP=NP (El cr1) + N (iR), where the '+' sign indicates the WHOLE-PART relationship." Takashima's interpretation of the string rizhong ze is possibly correct. On the other hand, Legge's understanding is not necessarily incorrect. When a string of NP + N appears, the relationship between the NP and the N does not have to be WHOLE-PART. In the chapter "Kang gao Mia" of the Shangshu, for example, there is one sentence reads as follows: In examining the evidence in criminal cases, reflect upon it for five or six days, yea for ten days, or three months. (Legge 1872:390). In this example, the string xun shi lOHt has the structure of NP = N + N. The relationship between xun ill and shi ffl" is 68 to eat. (Legge 1872: 469) On that morning the king came to the open country of M u in the borders of Shang and addressed his army. (Legge 1872: 300) The next day was guisi when the king in the morning marched from Zhou to attack and punish Shang. (Legge 1872: 306) In all these examples, the word zhao is translated as "morning." Clearly, it is a time division in early Chinese texts. In the Yin OBI, zhao is used as a time division as well. Below is one such example. [ 5 2 ] ^ s h , £Tjg: ^•nmm.z>m, mw*. i^m 23ns Crack-making on guichou (day 50), Xing divined: "When the gui-cut sacrifice to Zu Y i is performed on the next day jiayin (day 51), at the time of zhao, [the king will] do the j/ow-cutting." Heji 23148 In this inscription, the word zhao specifies the time of the yow-cutting. In such context, it makes sense to interpret the word zhao as a time division. James Legge translates zhao as "morning." But it may refer to a much shorter time period in the Yin OBI. There are several pieces of evidence, though not strong, that indicate that the duration of the time division zhao may be the same as that of dan. First, Dong Zuobin (1945.1.1: 6b) has pointed out that the bone graph for the word zhao depicts the time when the sun rises above grass. In the Yin OBI, there are some pictographs such as the graphs for the word ri 0 , 'the sun,' mdyue H, 'moon.' The graph for the word zhao might be another one. If so, as a time division, the word zhao may refer to some time around sunrise. Second, Dong Zuobin (1945.1.1: 6b) and Chang Yuzhi (1998: 136) notice that the definition of zhao in the Shuowen is dan J=L (%$, H-tii). This suggests that X u ShenW'K, the author of the Shuowen, thinks that the two time divisions zhao and dan refer to the same time period of the day. Third, there is the phrase chongzhao in the chapter "Chuodong" of the Shijing. not that of WHOLE-PART because shi, 'three months,' is not part of xun, ' 10-day week.' 69 Legge (1893: 84) interprets it as zhongzhao referring to 'the whole morning,' i.e., "the space between dawn and breakfast." If so, the duration of zhao is rather brief. There is no inscriptional evidence to show the duration of the time division zhao. From those textual records above, it can be inferred that the time period of zhao is brief. If indeed zhao does refer to the time between dawn and breakfast, zhao and dan seem to be different terms for the same time period of a Yin day. 2.3.1.3.3 Mu In Section 2.3.1.1.7, it has been demonstrated that the word Mu # is a time division in Period-Ill inscriptions. Such usage of this word can also be found in Period-II inscriptions. Heji 23148 is a good example. 2.3.1.3.4 Zhou and Zhi: Pseudo-Time Divisions Chang Yuzhi (1998: 169-172) lists zhou I I and zhi #1 as time divisions in Period-II inscriptions. However, as analyzed in Section 2.3.1.1.9, neither zhou nor zhi are time divisions in the Yin OBI. 2.3.1.4 Sub-divisions of Ri in Period IV 2.3.1.4.1 5«? Chang Yuzhi (1998: 178) cites Heji 32485 as evidence for the word su JH as a time division in Period IV inscriptions. But the rubbing of Heji 32485 is so unclear that the authors of the Jiaguwen heji shiwen transcribe the graph as zhi #t There is no clear evidence for the usage of su as a time division in Period IV. [X] Divined: "At dusk, [the king will] perform the jyow-cutting sacrifice. Heji 23148 70 2.3.1.4.2 Z e There is possibly one example of the word ze ^ as a time division in Period IV. [54] K j£: J ^ T ^ M o «1=ril» 33918 ... crack-making [on day xx, X] divined: "At the time of ze, there will be enough rain." 4 0 'Heji 33918 On the rubbing of Heji 33918, the graph ze seems incomplete. The transcription above is based upon that in the Jiaguwen heji shiwen. If this transcription is accurate, it is an example of ze as a time division in Period IV. 2.3.1.4.3 M u ? Heji 32485 has been cited by Chang Yuzhi (1998: 178) as evidence for the word mu H being a time division in Period IV. Here, this inscription is being transcribed as follows:4 1 [55] Z , * f t l # „ « ^ | » 32485 Onyiwei (day 32), [the king will] X (word unknown) mu. In the language of the Yin OBI, a time division normally occurs before a verb. In this inscription, the word mu appears at the end of the sentence. This makes it uncertain as to whether or not the word mu is a time division in this context. Other than this inscription, there is no other example of mu being used as a time division in Period IV. 2.3.1.5 Sub-divisions of Ri in Period V 2.3.1.5.1 M e H 4 0 The meaning of zuyu AEM is not certain. Here it is tentatively translated as "enough rain." 4 1 Below is Chang Yuzhi (1998: 178)'s transcription: m- mmn =ftm. On bingwu (day 43), [X] divined: "Cha should be X ... son performs theyoM-cutting sacrifice at the time of mu." It appears that her transcription mixes up two inscriptions. Even if her transcription is correct, the end of a sentence is not a position in which a time division occurs. 71 It was Wang Xiang (1920.12: 54) who first made the proposal that the word mei 'sister,' was a loan word for mei ft in the Yin OBI. Based upon this, Chen Mengjia (1956: 232) asserts that the word mei is equivalent to meishuang ft$[, 'early morning,' a time division in early Chinese texts. Chen Mengjia's claim is repeated by Song Zhenhao (1985: 315). However, there are some problems with Chen Mengjia's interpretation. First, even i f mei ^ is a loan word for mei ft, mei ft is not necessarily a time division in Period V . It is apparent that mei is a word of one syllable; meishuang is a word of two syllables. The meanings of the word mei are related to darkness; meishuang is a time division. The word mei is apparently different from the word meishuang. It is not convincing to equate mei with meishuang without providing sound reasons, and these scholars have failed to do that. Second, the following inscriptions show that the word mei may be a negative in Period V . Crack-making on xinmao (day 28), [X] divined: "Today it will continue to be foggy." "It will not continue to be foggy." Heji 38191 On this bone, charges jinri yanwu and mei yanwu are inscribed in different places. Because of this, the string mei yanwu is not likely to be the verification of this inscription. Rather, jinri yanwu and mei yanwu should be a pair of complementary charges. In a pair of complementary charges, one is positive and the other is negative. In Example 56, the first charge is clearly positive. The second charge mei yanwu is thus negative. Among those three words of the second charge, the word mei is the only word that can be a negative. Such a pair of complementary charges also appears on Heji 38192, 38194 and 38197. These four examples show that it is not by accident that the charges jinri yanwu and mei yanwu are inscribed at different places. It is better to interpret the word mei as a negative. If the word mei is a negative, it cannot be a time division. Ken-ichi Takashima, on the other hand, does not accept mei as a negative. He states his reason as follows: [56] K j£: 4^011. mum 72 Meiffi has a *m- initial with the final *-dt/-*dd, so it is similar to the well-known *m- type negative *m0^J]. Now, wu M- is a stative verb, which [does] not agree at all with the *m- type negative. Also, *wu yan wu ty] MM- is, according to my theory, impossible (unless this is a prognostication, which does not appear to be [the case]). On these grounds, I think you and some other[s'] suggestion that it is a negative in these examples is incorrect. Now, it is difficult to determine whether the word mei is a time division in Period V. On the one hand, there is no solid basis to interpret it as meishuang BfcM, a time division in early Chinese texts. From the context shown in Example 56, it can be inferred that the word mei may be a negative. On the other hand, it is not in agreement with Ken-ichi Takashima's theory about negatives in the OBI. Further study is required in order to decide the usage of this word. 2.3.2 Xi Xi & is one of the most often seen time divisions in early Chinese texts. The followings are two examples. He charged him, saying, "Morning and evening present your instructions to aid my virtue." (Legge 1872: 252) Their servants, charioteers, attendants, and followers were all men of correctness, morning and evening waiting on their sovereign's wishes or supplying his deficiencies. (Legge 1872:585) It is also accepted by specialists in the field of the Yin calendar that the word xi is a time division in the Yin OBI. Below are two such examples. [57] ffifc h ^ 1 . i^m 2287 On bingxu (day 23), [X] divined: "At night it will rain." Tunnan 2287 [58] M h - ^ 1 . 34054 73 Crack-making on gengyin (day 27), at the temple [X divined]: "At night it will rain." Heji 34054 Although specialists in this field agree that xi is a time division of the Yin day, they cannot reach a consensus about the time period to which xi actually refers. There have been three opinions about the time period of xi. The first one is that xi refers to the time of dusk; the second opinion is that it refers to the whole night; and the third opinion is. that it refers to the period from midnight to dawn. In the Shuowen, xi is defined as mu, 'dusk.' (ty, Hfjl) Wang Xiang (1923.1: 1) cites this record, which indicates that he may think the word xi refers to the time of dusk in the Yin OBI. However, such an opinion is not supported by inscriptional evidence. There are inscriptions that show xi and mu are two different time divisions. One such inscription is cited below. 4 2 [59]5£B£ij, mmWo &tym° i^m 27401 In offering the gwz'-cutting sacrifice to Father Ji, it should be at dusk that [the king will] perform the ^ow-cutting sacrifice. It should be at the time of xi that [the king will] perform the yow-cutting sacrifice. Heji 27401 The topic of inscriptions in Example 59 is to select a more appropriate time for performing the .yow-cutting sacrifice. Such a context indicates that mu and xi are contrasted in the mind of Yin people, which shows that these two time divisions refer to different time periods. How is the duration of xi different from that of mu? Could it be that mu is part of xi? Judging from the order of the two inscriptions in Example 59, that is not the case. If mu were part of xi, the Yin people would divine performing the yow-cutting sacrifice at xi before they divined performing the sacrifice at mu. That is the logical order. For example, one may decide i f he will do something tomorrow before he decides whether he will do it tomorrow morning or tomorrow afternoon. If he has already made a decision not to do that thing tomorrow, there is no need for him to decide whether he will do it tomorrow morning For other four examples, readers are referred to Heji 27396, 30845, and 41409, and Tunnan 1443. 74 or tomorrow afternoon. Therefore, i f mu is indeed included by xi, the Yin should divine the appropriateness of xi before they divine that of mu. The two inscriptions in Example 59 show that the divination of performing the sacrifice at mu took place before that of performing the sacrifice at xi. Such an order suggests that mu cannot be part of xi. Because mu and xi refer to different time periods, and since mu cannot be part of xi, they must be two mutually exclusive different time divisions in the Yin OBI. Therefore, the opinion that the word xi refers to dusk in the Yin OBI is incorrect. It is Dong Zuobin's theory (1945.2.1: 4b) that a Yin night is generally called xi'($LM& I f c^ lS^ ' t i l ) . He has reached this conclusion because he has not found other time divisions of the night. Chang Yuzhi (1998: 116-134) agrees with Dong Zuobin's view that xi refers to the whole night. On the other hand, Chang Yuzhi (1998: 142-152) thinks xi can be further divided into different time divisions. Chen Mengjia holds two opinions about xi. In Chen Mengjia (1956), on page 229, he thinks xi refers to the whole night; but on page 239, he suggests that xi refers to the time period from midnight to dawn. His second opinion is based on his calculation about the time of lunar eclipses recorded in the OBI. Chen Mengjia has noticed that the eclipses on jiawu and renshen took place during the time of xi, while the eclipse on gengshen did not. It thus occurs to him that the times of these eclipses may shed light on the duration of xi. According to his calculations, the beginning of the partial phase of these three eclipses was 00:24 on December 17, 1229 B C , 4:00 on January 28,1183 BC, and 23:06 on November 15,1218 BC, respectively. It seems that the first two eclipses began after midnight but the third one started before midnight. Therefore, Chen Mengjia has reached the conclusion that xi refers to the time period from midnight to dawn. One major problem with Chen Mengjia's conclusion above is that his understanding of relevant inscriptions is incorrect. Recent studies such as Qiu Xigui (1993), Chang Yuzhi (1998), Liu Xueshun (1998), and Zhang Peiyu (1999) show that what Chen Mengjia calls the gengshen eclipse actually occurred on the night of jiwei. Because he got the date of that eclipse wrong, his calculation about the time of that eclipse is also wrong. It follows that the conclusion he has drawn from those times is equally incorrect. In other words, Chen Menjia's view that xi refers to the time period from midnight to dawn is mistaken. Although Chen Mengjia has reached an incorrect conclusion, his approach to this issue 75 appears reasonable and practical. Since it is clearly recorded in the OBI that these eclipses occurred at night, the actual time of the eclipses certainly can show the duration of the time division xi. As long as all information about these eclipses is correct, a credible conclusion can be reached. It is this writer's intent to proceed to draw a conclusion now. The first step in calculating the time of those eclipses is to establish a temporal range for them. Because of the fact that both the lunar eclipse and the Chinese ganzhi T - ^ date occur repeatedly, lunar eclipses can occur on the same ganzhi date in a long time period. The longer the time period is, the more lunar eclipses take place on a specific ganzhi date. In order to make calculations being developed here more accurate, it is absolutely necessary to determine the time range for those lunar eclipses. A l l those inscriptional records of lunar eclipses belong to Period I. It follows that the time period for their absolute dates should be within Period I, which includes the reign of Yin kings Wuding S^ T and Zugeng ISM. It is unfortunate that there are no credible absolute dates for those two Yin kings, because the Chinese chronology before 841 BC remains uncertain. It therefore becomes necessary for the present writer to estimate the time period for those lunar eclipses. The estimate will be based upon the number and average length of reigns of Shang and Zhou kings before 841 BC. The inscription on the L i gui MM has confirmed that the Yin Dynasty was overthrown by King Wu ^1 of Zhou. 4 3 From then to 841 BC, there were the following eight Zhou kings: King Wu, Cheng $ 3 E , K a n g J ^ I , Zhao B S I , M u f H , G o n g ^ l , Y i M l , Xiao Y i 3 j^£ and L i JTJBE . From Period I to Period V , there were also eight Yin kings in the Yin OBI: Wuding, Zugeng, Zujia | l¥, Kangding^T, Wuyif i^Zl , Wending JtT, Di Y i filrZj and Di Xin Lei Haizong (1931) has calculated the average length of royal reign throughout Chinese history. His conclusion is that the average length of each reign is twenty-five years. Based upon Lei's conclusion, the reigns of those eight Yin kings and eight Zhou kings should be about (16 x 25 =) 400 years long. The first year of King Wuding should be about (841 BC + 400 =) 1241 BC, which can be taken as the upper limit of the time range for those lunar eclipse records. 4 3 For a comprehensive study on this inscription, readers are referred to Ken-ichi Takashima (1996-97). 76 The lower limit of the temporal range should be the last year of the reign of Zugeng. If the average length of a royal reign is applicable here, the length of reigns of Wuding and Zugeng would be 50 years, and the lower limit would be (1241 BC - 50 =) 1191 BC. However, according to early Chinese texts such as the chapter "Wuyi" of the Shangshu, the reign of Wuding was exceptionally long, about 60 years. The validity of such records is supported by the enormous number of oracle bones for Period I. For instance, the Heji, the biggest collection of the Yin OBI, has thirteen volumes. Among them, seven volumes are rubbings of Period I inscriptions. If calculated proportionally, the number of inscriptions in Period I alone would account for 54% of the total number of inscriptions in all five periods! In Liu Xueshun (2003), this author's calculation for the length of those five periods is no less than 140 years. Accordingly, Period I is about 80 years long. The lower limit of the time period for those lunar eclipses should be around (1241 BC - 80 =) 1161 BC. Among those five lunar eclipse records, the ganzhi date of the alleged gengshen eclipse is complicated, and it will be analyzed in detail in Chapter Four. Here, it will be sufficient to determine absolute dates for the other four lunar eclipses, in order to explore the time duration of xi in the Yin OBI. These four lunar eclipses are: 1. Eclipse on the night of renshen (day 9): Heji 11482; 2. Eclipse on the night of guiwei (day 20): Heji 11483; 3. Eclipse on the night of yiyou (day 22) of the Yin eighth month: Heji 11486; 4. Eclipse on the night of jiawu (day 31): Heji 11484. According to Liu Baolin (1978), during the period from 1241 BC to 1161 BC, there were the following lunar eclipses on those four dates, all visible at Anyang: Eclipse on the night of renshen: A . October 25, 1189 BC, 19:27 to 21:53.4 4 Eclipse on the night of guiwei: A. August 23-24, 1232 BC, 22:53 to 1:35; B. November 25, 1227 B C , 4:11 to 6:36; C. July 11-12, 1201 B C , 22:24 to 0:54; D. February 18-19, 1185 BC, 22:31 to 1:15; E. May 22, 1180 B C , 17:22 to 20:54. The moon rose at 19:09 and the sun set at 4 4 All these specific times are local time at Anyang. 77 19:12/" Eclipse on the night of yiyou: A. May 31-32, 1227 BC, 22:29 to 1:52; B. April 11, 1206 BC, 4:01 to 7:55. The sun rose at 6:12 and moon set at 6:16; C. November 25, 1181 B C , 18:02 to 21:45; Eclipse on the night of jiawu: A. November 4, 1198 B C , 20:31 to 23:21 4 6 It is apparent that there are more candidates for the eclipses on the nights of guiwei and yiyou. If an absolute date for an eclipse is picked up at random, there can be fifteen clusters of absolute dates for those eclipses. In any one of these clusters, there will be at least two lunar eclipses, i.e., the eclipse on the night of renshen and another on the night of jiawu, both of which occurred before midnight. Therefore, it is inaccurate for Chen Menjia to say xi refers to the period from midnight to dawn only. It must be noted that the date above is based upon the assumption that the Yin day, like the modern day, started with midnight. If the Yin day started with the time around dawn (an opinion more popular in the field), the dates for those four lunar eclipses will certainly be different. The following are the dates for those four eclipses based upon the view that the Yin day started around dawn. Again, most dates below are from Liu Baolin (1978). Eclipse on the night of renshen: A. October 25, 1189 B C , 19:27 to 21:53; B. January 28, 1183 BC, 4:04 to 6:23. Eclipse on the night of guiwei: A. August 23-24, 1232 BC, 22:53 to 1:35; B. July 11-12, 1201 BC, 22:24 to 0:54; C. February 18-19, 1185 B C , 22:31 to 1:15; D. May 22, 1180 B C , 17.22 to 20.54. Eclipse on the night of yiyou: 4 5 The time was calculated by the software Skymap. 4 6 Liu Baolin lists the eclipse on September 14, 1177 BC (that happened from 5:23 to 9:13) as one visible at Anyang. However, according to the Skymap, the sun rose at 5:55 and the moon set 5:56 on that day. It appears that the time of the actual eclipse was too short for the Yin people to observe. As a matter of fact, Homer H. Dubs (1947: 172) lists this lunar eclipse as invisible at Anyang. 78 A . May 31-32, 1227 B C , 22:29 to 1:52; B. November 25, 1181 BC, 18:02 to 21:45; Eclipse on the night of jiawu: A . December 17, 1229, 0:00 to 3:00; B. November 4, 1198 BC, 20:31 to 23:21; There is more than one candidate for all four eclipses. If an absolute date for an eclipse is picked at random, there can be thirty-two clusters of absolute dates for those eclipses. In any one of the clusters, there will be at least two eclipses, i.e., on the nights of yiyou and guiwei, both of which occurred before midnight. Again, doubt is cast on the credibility of Chen Mengjia's view that the time division xi in the OBI only refers to the time period from midnight to dawn. After analyzing the first and the third opinions (the first opinion being that of Wang Xiang (1923.1: 1), i.e., that xi refers to late afternoon; and the third opinion being that of Chen Mengjia (1956: 239), i.e., that xi refers to the period from midnight to dawn), the present author would like to address the issue of whether xi can be divided into different time divisions. Dong Zuobin does not think xi can be further divided in the OBI. But at least three other scholars do think xi can be divided into several time divisions. It is Chang Yuzhi's view (1998: 152), for example, that xi can be divided into three time divisions: zhi $1, zhu ft, and su M- However, as shown in Section 2.3.1.1.9, both zhi and zhu are not time divisions in the OBI. As for su, as will be demonstrated in Section 2.4, that it does not belong to night-time. Rather, it denotes the start of the Yin day. Huang Tianshu (2001) is the most recent study on alleged time divisions of the night in the Yin OBI. Since he thinks sunset or dusk belongs to night, his definition of night seems to be the time between sunset and sunrise, which is different from the present writer's view and that taken in this dissertation. In the present context, the definition of night refers to the time of darkness that excludes sunset and dusk. According to this definition, the time divisions of a night proposed by Huang Tianshu (2001) include: yuechu R fct}, 'moonrise,' xiaoye / J S ^ , 'small night,' meiren^K, 'sleeping man(?),' zhiJS, 'wine vessel(?),' zhonglu^^, 'mid lu,' and unknown^, 'double night(?).' Huang Tianshu does not provide any examples of those words that he proposes as time divisions of night ever being used as time divisions in early Chinese texts, which is the starting point for determining a time division in the OBI. 79 Therefore, the present writer is compelled not to accept any of them as a time division in the Yin language. Moreover, it is preposterous to consider some of these proposed expressions as time divisions. Take moonrise as an example. It is a fact that the time of moonrise changes everyday. In Anyang, on July 5, 1166 BC, the moon rose at 8:38 a.m. How could that be a part of night? Also, Huang Tianshu cannot explain the meaning of the unknown 'double night (?).' Without knowing its meaning, how can he be sure it is a time division? He fails to prove any of these alleged time divisions of night. Song Zhenhao (1985: 314) proposes that there are three time divisions of night: zhi ft, xi ty, and su As shown in Section 2.3.1.1.9, zhi is not a time division. The time division su does not belong to night. Therefore, there still is only one time division for the Yin night, which is xi. This writer's foregoing analysis has discredited the first and the third view about the time period to which xi refers. As for the second view, there is no inscriptional evidence against it. It is actually common knowledge that xi can refer to the whole night. Moreover, it still is a valid conclusion that no study has proven xi can be further divided into sub-divisions of time. Therefore, this writer's conclusion is that Dong Zuobin's view about xi is correct: it refers to the whole night in the OBI. 2.3.2.1 Zhuo: a Time Division? Ken-ichi Takashima (1996a.2: 70) and David N . Keightley (2000: 20) read the bone graph ESI as zhuo 'to cut, to cleave,' replacing an old reading dou S , 'a vessel without cover,' by David N . Keightley (1978: 43) and Ken-ichi Takashima (1979-80: 54). When zhuo occurs between two ganzhi dates, they suggest that it denotes a time division. 4 7 David N . Keightley (1978: 43), for example, states that when this word appears between two ganzhi dates which are always consecutive, it "apparently referred to the night-time no-4 7 Homer H. Dubs (1951: 331) has discussed the usage of this bone graph. But he fails to specify which word it represents. Qiu Xigui (1993) reads it as xiang |n], 'toward.' Neither Homer H. Dubs nor Qiu Xigui has mentioned that it can be used as a time division in the OBI. It appears that there is no point in discussing their opinions here. 80 man's-land between two cyclical-day dates." Keightley (2000: 20) interprets zhuo as a term referring to "the period during the night (xi ty) when two consecutive ganzhi days 'cleaved' to each other." In a note about David N . Keightley (1978: 43), Ken-ichi Takashima (1979-80: 54) suggests that the basic meaning of the word zhuo is 'to cut'. He continues: Applied to 'sacrificial' activity, it means 'to cut up a victim, to dismember.' When used in a context of time-expression, the verb means '(when) ... cuts (i.e., separates, demarcates) ... [WJhile Keightley thinks that the word t'ou referred to the period which cleaved and divided two days, i.e., his 'no-man's land' time duration, my formulation would have to refer to a period of some duration in the day following the first kan-chih date. This is so because I am in effect suggesting the word t'ou is a verb and, semantically, a 'punctual' rather than a 'durative' one. David N . Keightley's (2000: 20) treatment is so brief that he does not provide any reason for zhuo to be considered as a term for the period during the night. On the other hand, when he defines the period as "when two consecutive ganzhi days 'cleaved' to each other," this suggests that he may consider zhuo as a verb, 'to cleave.' 4 8 As for Ken-ichi Takashima, he clearly states in the citation above that the word zhuo is a verb in a context of time-expression. Because it is common knowledge that a verb cannot be a time division, the word zhuo cannot be regarded as a time division in the Yin OBI. David N . Keightley (1978: 43) has noticed that the zhuo occurs between two ganzhi dates that are always consecutive. This is true. Below is an example. On the seventh day, when the night of the jiwei day "cut" the gengshen, the moon was In such a context, it is actually the phrase ganzhi + zhuo + ganzhi that refers to a time period. In Inscription 60, the expression jiwei zhuo gengshen specifies the time when the Edwin G. Pulleyblank points out that "The English word 'cleave' has two contradictory meanings, 'to cut apart' and 'to cling together.' When Keightley says 'cleaved to each other,' he seems to mean the second meaning." Since David N. Keightley interprets ganzhi + zhuo + ganzhi as "no-man's land," which implies that the time period does not belong to either ganzhi dates, it appears that he uses the word "cleave" to mean "to cut apart." 886 jk eclipsed. Yincang 886b 81 moon eclipsed. The word zhuo itself does not refer to that time duration. This analysis of the structure oi ganzhi + zhuo + ganzhi also suggests that the word zhuo is not a time division in the Yin OBI. 2.3.3 Order of Time Divisions of the Yin Day In previous sections, it has been demonstrated that the Yin day can be divided into ri 0 , 'daytime,' and xi , 'night-time.' As for the time division ri itself, it can be further divided into various distinct time divisions. In this section, this writer will first determine the order of ri and xi. Then, he will decide the order of those time divisions of ri. 2.3.3.1 Ri before Xi The order of ri and xi can be deduced from the following inscription. Wo $mm, JiM^Mo OTJ(=^iJ)4>l(=Hf). ZiE^WS£+Mo Crack-making on bingshen (day 33), Nan divined: "On the coming yisi (day 42), [the king will] perform the >>ow-cutting sacrifice to Xiayi ." His majesty, having prognosticated [the omens], declared: "The performance of the .yow-cutting sacrifice means that there will be a curse [from Xiayi], and there will be thunder."49 On the yisi day, we performed the j/ow-cutting sacrifice. At [the time of] ming it rained. [When the victim was being] dismembered, it was already raining. When the dismembering was done, it was still [likewise] raining. Having decapitated [the victims] and cut up a bird (?), 5 0 the weather became clear. On the night of yisi, there was thunder in the There are various transcriptions of the bone graphic. As pointed out by Ken-ichi Takashima, none of those transcriptions can be assigned either a reading or a meaning. Ken-ichi Takashima suggests that this bone graph represents a word for Wi 'thunder' on the basis of contextual and paleographical evidence (Takashima 1996a.2: 140-141). 5 0 In his comments, David Pankenier writes that he thinks that the word fa "meant 'to behead' as a verb, sometimes 'victims beheaded' as a noun." My translating fa here as "to dismember" is based upon Ken-ichi Takashima's research. Ken-ichi Takashima says that 82 west. Heji 11497 f & b In the verification, there are two time divisions: ming and xi. From the content of this verification, it can be inferred that the time division ming occurs before xi. Since ming is a sub-time-division of ri, it is safe to say ri is before xi in the OBI. The inscriptions on Heji 27396 also shed light on the order of ri and xi. These [The king] shall make an offer to Father Ji. It should be at dusk that [he will] perform the yow-cutting sacrifice because the king will receive assistance. [The king shall make an offer] to Father Geng. [He will] at night perform the you-cutting sacrifice because the king will receive assistance. Upon the coming day, [the king will] perform the .yow-cutting sacrifice. In these three inscriptions, there are three time words: mu, 'dusk,' xi, 'night,' and lairi, 'the coming day (=next day).' Moreover, the word hui occurs before mu andyw before lairi. the standard interpretation of the graph fa holds that it depicts a man's head being chopped off by an ax, and expresses the meaning 'decaputuri (human victims)' or 'prisoner.' However, both of these traditional meanings are deduced more from the shape of the graph than from consideration of the word it represented. In a sacrificial context, where a verbal interpretation is called for, we would suggest that it meant 'to dismember' or 'disintegrate' (destroy the unitary integrity of, reduce to remnants) rather than the usual 'to behead, decapitate; to attack (1996a: 253). Since an enemy state of Shang can occur as the object of the verb fa and it is impossible to physically behead a state, Takashima's new interpretation of the word fa makes better sense. David Pankenier further writes that "I don't think much of the translation of liu niao as 'a bird was cut up'; [it] doesn't make much sense to me. How do those who argue that xing 'star' here is actually qing deal with niaoT At first glance, the translation of liu niao as 'a bird was cut up,' which has been changed to 'cut up a bird' in this draft, may not make sense. On the other hand, the string tuo mao niao xing is an event in daytime when stars are invisible. It is difficult to interpret niao xing JSjii in this inscription as a star. In addition, Ken-ichi Takashima argues that the character mao represents the word liu, meaning "to split open/divide into two" (1996a.2: 88-89). He translates liu niao literally as "cut up a bird" and the present writer has followed him. Li Xueqin (1981), who has made the interpretation of xing as qing influential in the field, reads niao as the adverb shu $k, 'suddenly.' But that is not supported by philological evidence. Before this writer may find a satisfactory interpretation of the word niao in this inscription, he prefers Ken-ichi Takashima's direct translation of niao. inscriptions can be transcribed as follows: [ 6 2 ] ^ X 5 £ B , *.mm, BESfeo mxmxm, £ g # i . Heji 27396 83 Ken-ichi Takashima (1990: 36) states that "whenyw is used, it always introduces a day more remote than hui does. Conversely, of course, hui always introduces a day more immediate than yu does." He further suggests that the compelling reason for such usage of yu and hui is because "in the bone inscriptions the word yu had a clear 'futurity' meaning" (1990: 37). Therefore, the appearance of mu, xi, and lairi is a time order from recent to remote. Since mu is a time division of ri, it follows that the order of mu, xi and lairi can be expressed as ri, xi, and lairi. Again, in a Yin day, ri comes before xi. 2.3.3.2 Order of Time Divisions of Ri In previous sections, textual and inscriptional evidence has been cited to demonstrate that the followings are time divisions of ri: su M , chen jH, dan fi, ming BJ, zhao MB, dashi A i£, dashishi A ^ f f t , shishi dacai A3R, dacaishi A ^ & t , rizhong FJ rf3, zhongshi 41 Bf, ze IF:, zeshi ItkH, xiaocai <IN3 ,^ xiaocaishi ' h ^ B t , xiaoshi <hltr, yongxi iM^, mu U , and hun W- Among these twenty terms of time division, the following terms belong to different periods, but they refer to the same time period of the Yin day: su and chen refer to the same time period before dan; ming, zhao, and dan, all of which refer to the time of dan; dashi, dashishi, and shishi, all of which refer to the time of dashi; Dacai and dacaishi refer to the time of dacai; rizhong and zhongshi refer to midday; ze and zeshi refer to the time of ze; xiaocai and xiaocaishi refer to the time of xiaocai.5X In total, there are actually eleven time 5 1 In his comments to the last draft of this dissertation, David Pankenier raises very serious questions: "On p. 93 you mention the change from yi yue to zheng yue. Here you lump together all the time words without distinction as to period, which you do sort out in 2.3.3.2.6. Later you discuss the shift from inter-year to intra-year intercalation (e.g., p. 94, 108, 163-164). I kept waiting to find a discussion of what you think happened to cause these changes. Do you have any ideas about the cause of the calendar reform/change in terminology for which you provide such graphic evidence? This kind of terminology tends to be VERY conservative and resistant to Change. Something pretty radical had to happen for this change to occur. How does this relate to changes in the sacrifices, if at all? Do you really have nothing to say about this? Does anyone else have anything to say worth citing?" Dong Zuobin (1945.1.1: 2-4) notices that there are radical changes in many aspects of the Yin culture between Period I and Period II. In Period II, new sacrifices appear, new terms in the Yin calendar are adopted, new styles of calligraphy are created, and some events in Period I have disappeared in Period II. It appears that all of a sudden these big changes happened approximately at the same time. There must be something extremely radical behind them. Dong Zuobin attributes these changes to Zujia's ascending the throne. Dong Zuobin suggests that Zujia was a reformer and he was responsible for 84 divisions of ri: su, dan, dacai, dashi, zhongshi, ze, yongxi, xiaoshi, xiaocai, mu and hun. It remains now to determine the order of these time divisions. 2.3.3.2.1 The Order of Su and Dan The order of su and dan has been analyzed in Section 2.3.1.1.2.1. The inscriptions in Example 7 show that su is a time division before dan. For convenience, that inscription is cited again here. [07] % j & R m % ^ m A o H^B .73ft&*$tA. 26897 On the gui-day, at the time of su, Guard will attack Zai because [that will] not exterminate [his] people. On the gui-day, upon the time of dan, [Guard] will then attack Zai because [that will] not exterminate [his] people. Heji 26897 In this example, both su and dan are time divisions of that gui-day. In the second inscription, the word yu occurs before dan. According to Ken-ichi Takashima (1990: 36-37), the word yu has a clear meaning of "futurity" in the OBI and always introduces a more remote time. Su is thus closer to the time of the divination than dan. In other words, su comes before dan. 2.3.3.2.2 The Order of Dan, Dacai and Dashi Dan, dacai, and dashi all are time divisions of the morning. The inscription on Tunnan 42 shows that dan comes before dashi. The inscription reads as follows: [09] g BM-trFJ ( = 0 t ) ^ M . From dan to shishi, it will not rain. In the pattern "from ... to...," the complement after "from" is normally closer to the reference point than the complement after "to" is to the reference point. In this inscription, all those changes. If so, why and how Zujia could bring into being those changes remain unclear. In this study, the focus is on examining and explaining principles of the Yin calendar. Those issues mentioned by David Pankenier will be topics of future research. 85 dan appears after "from" and shishi after "to." Therefore, dan is earlier than shishi, another expression of dashi. It is unfortunate that there is no inscription that clearly shows the relationship between dacai and dan or between dacai and dashi. As a result, it becomes necessary here to determine their order based only upon circumstantial evidence. It is Dong Zuobin's idea (1945.1.1: 5) that dacai is equivalent to zhao and xiaocai to mu ib,). This opinion is followed by Chen Mengjia (1956: 232), Wen Shaofeng and Yuan Tingdong (1983: 71, 75), and Chang Yuzhi (1998: 163-164). Dong Zuobin bases his conclusion on the context in which dacai and xiaocai appear in the chapter "Lu yu" of the Guoyu. However, since dacai, xiaocai, zhao and xi all occur in the same passage of that chapter, it is this writer's opinion that there should be some subtle difference between dacai and zhao and between xiaocai and xi. According to the chapter "Lu yu" of the Guoyu, dacai was the time when the Son of Heaven paid tribute to the sun. It is easy to understand that it is proper to pay tribute to the sun shortly after sunrise. As a matter of fact, there are inscriptions showing that the Yin indeed offered sacrifice to the rising sun. Below is an example. [63] :^ #!i^Sfe^W.A.-H :^ 6572 Crack-making on xvuxu (day 35), Nei [divined]: "[The king should] call Que to direct the binding-sacrifice to the rising and setting sun of a specially reared ovine. 5 2 Heji 6572 It is also recorded in the chapter "Yao dian" of the Shangshu that Yao ordered He "respectfully to receive as a guest the rising sun." (Legge 1872: 18-19) It appears that early Chinese indeed paid tribute to the rising sun, which may have been performed at the time shortly after sunrise. These inscriptional and textual records indicate that dacai may refer to the time immediately after sunrise, which is the end of dan. This writer thus places the time division dacai between dan and dashi, rather than precisely at dan or zhao. The translation of this inscription is basically that of Ken-ichi Takashima (2003a). 86 2.3.3.2.3 The Order of Dashi, Zhongshi, Ze and Yongxi The following inscriptions clearly show the order of these four time divisions. [08] -fr 0 ( ^ I t 0 (=Bt)^M. #0 (=fft)M* 0 (=ttt)£M 624 "From shishi to zhongshi, it will not rain." "From shishi to zhongshi, perhaps, it will rain." Tunnan 624 [09] * B (=fcf)M^*M. i^m 42 "From zhongshi to ze, it will not rain." Tunnan 42 [64 ]^S[^]^Mo « ^ ) ) 29801 "From ze to [yo«g] xi, it perhaps will rain." Heji 29801 In inscriptions above, all these four time divisions occur in the pattern "from ... to...." Since the complement after "from" is closer to the reference point than the complement that after "to," the order of these four time divisions, from early to late, can be determined as follows: shishi, zhongshi, ze, and yongxi. 2.3.3.2.4 The Order of Yongxi and Xiaoshi Yongxi and xiaoshi are two different time divisions. But there are no inscriptions that, directly or indirectly, show the order in which they occur. Moreover, yongxi is the only exception, in that we cannot find an instance in which this word appears as a time division in early Chinese texts. Therefore, at the present time, there appears to be no way to determine their order with certainty. For the time being, the intent will be to determine the order of yongxi and xiaoshi based upon the order of dashi and dacai. As shown in Section 2.3.3.2.2, dashi is close to dacai, which in turn is very close to the sunrise. We may suspect that xiaoshi is close to xiaocai, which in turn is very close to the sunset. Tentatively, it is suggested that yongxi be placed before xiaoshi. 2.3.3.2.5 The Order of Xiaoshi, Xiaocai, Mu, and Hun 87 In early Chinese texts, there is evidence to determine the order of hun and mu. According to the Shuowen, hun refers to the time when the day is dark (S^ t f e ) and mu refers to the time when the day is going to be dark ( 0 j l ^ t i l ) . Therefore, mu is a little earlier than hun. Also, it is said that the time of 45 minutes after sunset is called hun (see Section 2.3.1.1.8.). So it can be inferred that mu should refer to some time after sunset. In Section 2.3.3.2.4, it was tentatively said that xiaocai is considerably before the time of sunset. So xiaocai should be a little earlier than mu. Further, the distribution of a pair of time divisions may be symmetric. Since dashi is after dacai, xiaoshi might be before xiaocai. Based upon the considerations above, the tentative order of these four time divisions, from early to late, can be established as xiaoshi, xiaocai, mu and hun. 2.3.3.3 The Order of Time Divisions of the Yin Day To summarize the analysis in Section 2.3.3, the order of time divisions of the Yin day in the language of the Yin OBI can be tabulated in Table 1 on next page. 2.3.4 Characteristics of Yin Time Divisions An examination of the twelve time divisions of the Yin day in Table 1 reveals the following characteristics of those time divisions. First, the duration of each of these time divisions is not equal, which has already been pointed out by Chang Yuzhi (1998: 180). For instance, on the one hand, the whole night-time is one division of the Yin day; on the other hand, between nautical twilight and the morning meal, there are four divisions. Also, there are three divisions in the short period around sunset. It goes without saying that each of these seven divisions is bound to be much shorter than xi, 'night.' The duration of each division is unequal in length. Second, those time divisions are not consecutive. Table 1 shows that the time division zhongshi, 'time of meridian, midday,' comes after dashi shi, 'the time of big meal in the morning.' The big meal in the morning certainly is not eaten in late morning. Between dashi shi and zhongshi, there must have been a period of time. This example shows that the time divisions of the Yin day do not follow each other consecutively. 88 Table 1: The Order of Time Divisions of the Y i n Day Period I Period II Period III Period IV Period V Chen Jt Ri 0 Ming HfJ ZhaoM DanS. Dacai X% Dacaishi ^ T ^ B t Dashi -X^k Dashishi X.'kH Shishi it H -Dashi X1k Shishi ft Wt Zhongshi 4 1 St Rizhong 0 f Zhongshi f Ht Rizhong B t 3 Zeshi 7o«gxz Xiaoshi Xiaocai 'h^K: Xiaocaishi /hjRBt Mu If Mu H Xity Xity Xity Third, the vast majority of those time divisions are closely related to the sun. For example, dan and dacai refer to the time around sunrise. Zhongshi means the sun is in the middle of the sky. A time period after the sun passes the meridian, it is called ze. The time around sunset is divided into three time periods: xiaocai, mu and hun. A t the time of xi, the sun is not visible. One way or another, eight of those twelve time divisions are related to the sun's motion in the sky. It is thus clear that the position of the sun in the sky plays a pivotal role in dividing the daytime by the Y i n . Fourth, because most time divisions are related to the sun's motion in the sky, the exact time and length of each time division vary with the seasons. In this connection, David N . 89 Keightley (2000: 20-21) reminds us that the exact time of dawn and sunset varies with the seasons. For instance, according to the calculation of the Skymap, on July 5, 1166 BC, the summer solstice of that year, the sun rose at 4:59 a.m. in Anyang. On December 31, 1166 BC, the winter solstice of that year, the sunrise was 7:39 a.m. The difference between these two sunrises is 2 hours 40 minutes. Accordingly, the exact times of dan and dacai on those two days are far apart. As for the length of the time period from noon to sunset in Anyang, it is 7 hours 50 minutes on July 5, 1166 BC; and it is 5 hours 31 minutes on December 31, 1166 BC. Between noon and sunset, there are five time divisions: ze, yongxi, xiaoshi, xiaocai and mu. It is likely that the length of ze and yongxi on July 5, 1166 B C would be much shorter than on December 31 of that year. Because Yin time divisions are not equal and consecutive, their length and exact time vary with the seasons, it is generally impossible and inaccurate to specify each time division of the Yin day using modern hours such as 8:00 a.m. or 7:00 p.m., etc. In her report about this dissertation, Anne O. Yue points out that Table 1 "showing that those designating divisions of the day time far exceed the one-expression for night. It may be pointed out that this is quite natural and to be expected, since divisions during the day are intimately related to human activities daily." Her observation is correct. 2.4 The Start of the Y i n Day Before the discovery of the Yin OBI at Anyang in 1899, there was only one view with regard to the start of the Yin day. According to the Shangshu dazhuan |RjT3Af#, the Yin day began with jimingj%$%, 'cockcrow.' Since the discovery of the Yin OBI, two more opinions have been proposed. Dong Zuobin's opinion (1945.2.II: 6b) is that the commencement of the Yin day is zhao or dan J i , a time division that includes actual sunrise and some time before sunrise. The other opinion, which was put forward by Homer H. Dubs (1951: 330), is that the Yin day started with midnight. Neither Dong Zuobin (1945) nor Homer H. Dubs (1951) has refuted the cockcrow theory. Below, this writer will first evaluate each of these three views. Thereafter, the present researcher's conclusion, drawn from inscriptional evidence, will be suggested: the start of the Yin day is su M. 90 2.4.1 Evaluation of Previous Theories 2.4.1.1 The Midnight Theory c-i The midnight theory was first proposed by Homer H . Dubs' (1951: 330). J His argument proceeds as follows: The Babylonians began the day at sunrise; the Jews and Greeks, at sunset; the Romans, at midnight, as did also China in Han times. The fact that most of these eclipses are reported as having occurred in the 'night', si, of a given day eliminates the Jewish and Greek 'day'. This fact also makes it possible that Shang China employed the Babylonian 'day'. Babylonia was the nearest area with a developed astronomy. Later Chinese practice however provides strong evidence that in the Shang period China used the Roman 'day'. Eclipse record IV confirms that probability. This view is reaffirmed in Homef'H. Dubs (1953: 102). The above citation shows that Dubs justifies his conclusion on two grounds. First, the Chinese day began with midnight in Han times. The other reason is that the absolute date on which he calculates an eclipse record, later published as Yingcang 886, coincides with his midnight theory. It appears that these two reasons are not strong enough to establish midnight as the start of the Yin day. Below the present author elaborates upon Dubs' two justifications for his conclusion. First, the fact that the Chinese day started with midnight in Han times is not conclusive evidence for determining the beginning of the Yin day. It is true that Chinese history is continuous since the Yin Dynasty. Because of this, it is possible that the beginning of the Chinese day in Han times is the same as that of the Yin day. However, whether the start of the Yin day is actually midnight has to be decided by contemporary evidence. The fact that Dubs is aware of this requirement is clearly shown in his comments about Dong Zuobin's This is an article about "The Date of the Shang Period" rather than the method of recording day in the Shang Dynasty ( «Kf^ WIB0>S-» ), as translated by Chang Yuzhi (1998: 181). 91 reconstruction of the Yin calendar. On this subject, Dubs says: The dating of this eclipse raises the problem concerning how much we know about the Shang calendar. Dr. Dung appears to go beyond the available evidence, fixing, by rules employed in Han times, the length of months, the interpolation of intercalary months, etc. for the whole later portion of the Shang period. It is however quite likely that astronomical calculations for the beginnings of months began to be used in China only about 600 B.C. Hence [it is] dangerous to employ a Han calendar for a period a thousand years earlier (Dubs 1951: 325). By the same token, it is dangerous for Dubs to apply the start of a Han day for a period a thousand years earlier. The beginning of the Han day is not "strong" evidence for the start of the Yin day. The second objection to Dubs' reasoning is that the absolute date he selects for the eclipse on the night of jiwei lends no support to his opinion about the start of the Yin day. According to Dubs (1951: 331), the date for that eclipse is December 27-28, 1192 BC: It began at 8:48 p.m. local time at Anyang and became total at 9:53 p.m. Totality lasted for an hour and three-quarters. The moon began to reappear at 11:37 p.m., and the eclipse ended at 42 minutes after midnight, on the morning of cyclical day 57. The debated unknown character in the record then probably means something like 'midnight' or 'continuing into', indicating that this eclipse endured into the morning of day 57. The reason Dubs selected this date is because he assumed that the Yin day began with midnight. If he had assumed sunrise to be the beginning of the Yin day, he would not have considered this eclipse on December 27-28, 1192 BC a candidate for this eclipse. The date he assigned to the eclipse does not confirm his opinion that the Yin day started at midnight. Third, the credibility of Dubs' argument is weakened by possible evidence he himself mentioned (1951: 332): 92 For record II, there is, within half a century, in addition to the eclipse of 1198, only the partial eclipse of 1229 B.C., Dec. 17, after midnight on the morning of day 32. If this latter dating is accepted, Shang China must have employed the Babylonian 'day'. Why does Dubs not accept that later dating? He offers no explanation. In the absence of providing any evidence to rule out this date, Dubs' opinion on the start time of the Yin day appears subjective and undermines the credibility of his assertion. Fourth, as analyzed in Section 2.3.2, there is inscriptional evidence to support the view that the whole Yin night-time is called xi. In other words, the Yin people did not have the time division for midnight. Without such a time division, it would be impossible for them to be able to select midnight as the beginning of a day. Because of these four reasons just discussed, Homer H. Dubs' view lacks solidity. Even so, his midnight theory has two followers: Chou Fa-kao (1964-1965) and Suetsugu Nobuyuki (1994). Both Chou Fa-kao (1964-1965) and Suetsugu Nobuyuki (1994) suggest that there were different beginnings of the day in Yin times. For example, it is Suetsugu Nobuyuki's theory (1994: 12) that the common Yin people reckoned that the day started with dawn, while day began at midnight according to diviners whose work took place at night. The fact is that neither Suetsugu Nobuyuki nor Chou Fa-kao has produced clear inscriptional evidence to substantiate their theory. Their theory therefore remains to be proven. 2.4.1.2 The Cockcrow Theory The cockcrow theory first appeared in the Shangshu dazhuan, whose author is Fu Sheng fA:J4. Fu Sheng lived in the beginning of the Western Han Dynasty, so it is a text of the Han times. There is no indication that this theory is based upon any contemporary sources. Its credibility is thus open to question. Moreover, in the corpus of extant Yin OBI, there still is no evidence that lends any support to such a theory. Therefore, judging by the synchronic evidential approach, it is difficult to accept that the Yin day started at the time of cockcrow. 93 2.4.1.3 The Dawn Theory Based upon his research about time divisions seen in the OBI, Dong Zuobin (1945.2.1: 6b) asserts the theory that the Yin day started with dawn. But he does not reject the view that the Yin day began with the time of cockcrow. Chang Yuzhi (1998: 193) makes a justifiable criticism of such an inconsistency. Dong Zuobin's argument is as follows: in the Yin Dynasty, a day has two parts: ri 0 , which refers to the whole daytime, and xi ty, which refers to the whole night-time; in a day, ri comes before xi. Dong Zuobin thus thinks that the earliest time division of ri would be the beginning of the Yin day. According to his research, the earliest time division of the daytime in the Yin OBI is ming BJ, 'dawn'. He then reaches the conclusion that the Yin day began with dawn. Dong Zuobin's argument seems logical. If a Yin day consists of ri and xi and the time division ming, 'dawn,' is indeed the earliest time division of ri, his conclusion would be accepted by many specialists in the field. For example, both Song Zhenhao (1985: 323) and Chang Yuzhi (1998: 188-193) agree with Dong Zuobin about dawn being the beginning of daytime in a Yin day, and they have followed his theory. However, as shown in Sections 2.3.1.1.2.1 and 2.3.3.2.1, there is the time division su M before dawn. It is incorrect for these researchers to regard dawn as the earliest time division of ri and, further, to take dawn as the beginning of the Yin day. 2.4.2 Su: the Start of the Yin Day As analyzed in Section 2.3, there are twelve time divisions in the Yin OBI: su, dan (or ming, or zhao), dacai (or dacaishi), dashi (or dashishi, shishi), zhongshi (or rizhong), ze (or zeshi), yongxi, xiaoshi, xiaocai (or xiaocaishi), hun, mu, and xi. It goes without saying that the commencement of the Yin day must be the earliest one among those twelve time divisions. By determining the earliest time division, one will get the start of the Yin day. 2.4.2.1 From Dacai to Xi: No Start of the Yin Day 94 In Example 61 cited in Section 2.3.3.1, there are two time divisions, ming and xi. Both ming and xi appear in the verification of this inscription. The content of this verification shows that these two time divisions belong to the same day yisi. More importantly, the order of the occurrence of ming and xi indicates that ming appears before xi. In other words, ming is the earliest time division between ming and xi. As shown in Table 1, from ming to xi, there are the following eleven time divisions: ming, dacai, dashi, zhongshi, ze, yongxi, xiaoshi, xiaocai, mu, hun, and xi. In can be deduced from Example 61 that ming is the earliest one of these eleven time divisions. As mentioned above, only the earliest time division can validly be considered the beginning of the day. If there are only those eleven time divisions in the Yin OBI, ming would be the start of the Yin day, as Dong Zuobin (1945.2.1: 6b) proposes. The fact is that there are twelve time divisions in the OBI. The other time division is su. It appears that one has to determine the beginning of the Yin day between ming (or dan) or su. If su is an earlier time division than ming (or dan), then it will be the start of the Yin day. 2.4.2.2 Su before Dan Both Song Zhenhao (1985: 307-308) and Chang Yuzhi (1998: 150-151) list su as a time division in the OBI. Also, both Song Zhenhao (1985: 312) and Chang Yuzhi (1998: 150-151) agree that su is the last time period of the Yin day, which would make dan the earliest time division of the Yin day. Therefore, they have reached the conclusion that the Yin day begins with dawn. They have cited three examples to show dan is the first time division of the present day and su is the last one of the previous day.5 4 However, their interpretations of those three examples are incorrect. Each interpretation will now be analyzed in turn. Their first example includes inscriptions on two oracle bones. [ 6 5 ] i ^ 0 * t l o The first two are cited by Song Zhenhao (1985: 323), and all three are cited by Chang Yuzhi (1998: 188-191). 95 i%m 1685 It should be a red cow. It should be a cow with different colors. It should be at this su that [the king will] perform the .yott-cutting sacrifice. [Upon] the dawn [of next day, the king will perform the >>ow-cutting sacrifice.] Anming 1685 [66] f S B I ^ I . mM)) 2336 Upon the dawn of next day, it will rain heavily. Yingcang 2336 Chang Yuzhi (1998: 189) acknowledges that the last inscription of Example 65 is damaged severely. Actually, only part of the graph dan survives. When Chang Yuzhi reconstructs that last inscription as "upon the dawn of next day, the king will perform the _yo«-cutting sacrifice," she creates two problems. First, Chang Yuzhi's transcriptions (1998: 150 and 189) for this last inscription are different. On page 150, her transcription is hui dan you, 'it should be at dawn that [the king will] perform the .yow-cutting sacrifice.' On page 189, it is transcribed as yu yi ri dan you, 'upon the dawn of next day, [the king will] perform the yow-cutting sacrifice.' She is not certain what the inscription really is. Second, the inscriptions in Example 65 are about the time to perform the .yow-cutting sacrifice. On the other hand, the inscription in Example 66 is about whether it will rain the next day. There is no direct link between these two examples. There is no reason for Song Zhenhao and Chang Yuzhi to reconstruct the last inscription in Example 65 as yu yi ri dan you solely based upon Example 66. If that is the case, there is no pair of jinsu and yiri dan in Example 65, and no definite conclusion about the start of the Yin day that can be drawn from these inscriptions. Their second example is two inscriptions on Heji 26897, which has been analyzed already in Section 2.3.1.1.1. To cite it again: On the gui-day, at the time of su, Guard will attack Zai because [that will] not exterminate [his] people. 26897 96 On the gui-day, upon the time of dan, [Guard] will then attack Zai because [that will] not exterminate [his] people. Heji 26897 Song Zhenhao (1985: 323) and Chang Yuzhi (1998: 189-190) have not offered any analysis of this example. They simply claim that this is a piece of evidence to support the view that the Y i n day starts with dan. As analyzed in Section 2.3.1.1.2.1, the two inscriptions in this example show su to be the time division before dan. Chen Mengjia (1956: 227) and Ken-ichi Takashima (1990: 36-37) have pointed out that the word yu T is used before a time word that is more distant. The reason is because yu has a clear "futurity" in the OBI. (Takashima 1990: 37) In the second inscription of this example, the word yu appears before dan. Therefore, dan is more remote than su. If so, dan is not the earliest time division of that gui-day and the start of that gui-day could not be dan, which contradicts Song Zhenhao's and Chang Yuzhi's conclusion. The third example cited by Chang Yuzhi (1998: 190) is an inscription on Heji 34601. [67]T£P K !%m&Ro £jgB, S® 0 55 « ^ ) ) 34601 On dingmao (day 4), [X] divined: "On wuchen (day 5), [the king will] repeat (?) dan. [The king] should not repeat (?) dan, [he] should continue." Heji 34601 Again, Chang Yuzhi does not explain how this example supports the view that the Yin day starts with the time division dan. On page 190, she first cites Y u Xingwu's interpretation that wuchen fu dan means that wuchen is the next day of dingmao(r%mMM., JJCJSBPT^ P .^IS 0 ill). Then, she pronounces that this inscription also shows that dan, i.e., dawn, is the beginning of the Yin day. It must be pointed out that it is extremely difficult to see any relation between Y u Xingwu's explanation of the phrase fudan and the start of the Yin day. It is a fact that wuchen is the next day of dingmao. Why would that make dan the beginning of the Yin day? There is no connection between Yu's interpretation of wuche fudan and Chang Yuzhi's conclusion. It also needs to be pointed out that the meaning of Example 67 is unclear. It appears that there is a pair of complementary charges in this example. In the second charge, negative wu appears before the phrase fudan. Since the word wu is an *m-type negative that negates verbs This writer believes what Chang Yuzhi transcribes here is a pair of complementary charges rather than one inscription. 97 "whose salient feature is their 'controllability' - verbs which are thought of as being controllable by the will of living human being," 5 6 fu must be a controllable verb. If so, Example 67 might have nothing to do with the beginning of a Y in day, even though the meaning offudan remains mysterious. The foregoing analysis demonstrates that those inscriptions cited by Song Zhenhao and Chang Yuzhi do not support their opinion that su is the last time division and dan is the first one of the Yin day. Rather, Example 7 advances evidence to show that su is the time division before dan. 2.4.2.3 Su: the Beginning of the Y i n Day There are the following twelve time divisions in a Yin day: su, dan (or ming, zhao), dacai (or dacaishi), dashi (or dashishi, shishi), zhongshi (or rizhong), ze (or zeshi), yongxi, xiaoshi, xiaocai (or xiaocaishi), hun, mu, and xi. It makes perfect sense to take the earliest time division as the beginning of the Yin day. The inscription in Example 61 indicates that none of these eleven time divisions from dacai to xi is the earliest time division of the Yin day, making dan (or ming) and su two candidates for the start of the Yin day. Moreover, it can be deduced from inscriptions in Example 7 that su is the time division before dan. The time division su is thus the beginning of the Yin day. For more details about *m-type negatives, see Ken-ichi Takashima (1996a: 370-375). 98 CHAPTER THREE "THE MONTH" IN THE YIN CALENDAR 3.1 Introduction It takes about 29 days 12 hours 44 minutes for the moon to revolve round the earth once. This period forms a unit for lunar-solar calendars that is called yue B, 'month,' in Chinese. In the corpus of the Yin OBI, there are an enormous number of month notations, a fact that testifies to the existence of the month in the Yin calendar. This chapter will deal with the following issues: the number of the month in the Yin year (3.2), the length of the Yin month (3.3), the start of the Yin month (3.4) and the arrangement of Yin months (3.5). 3.2 The Number of the Month in a Yin Year The months of a year are numbered, starting from number one to number twelve, or from number one to number thirteen, as is clearly shown by the following inscriptions. [oi]%M h], [£#!]. -Bo tjtrHWao ~Bo « ^ * » 16649 Crack-making on guihai (day 60), Bin [divined]: "[In the next] 10-day week, [there will be no disasters." Day guihai was in] the first month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiwei (day 20), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the second month. Heji 16649 Crack-making on guimao (day 40), Bin divined: "[In the next] 10-day week, there will [02]HifPh, g j j T : iotr(=3E)$o ~[B]o 16661 99 be no disasters." [Day guimao was in] the second [month]. Crack-making on guichou (day 50), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the third month. Heji 16661 [03]-£FPh f3Jfr(=^)^0 £H« H H h Ujt:(=5E)*S- 16657 Cracking on guimao (day 40), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the third month. Crack-making on guichou (day 50), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the fourth month. Heji 16657 StMh t / t r (=^n ) ^ o £J3„ « ^ ) ) 16668 Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fourth month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the fifth month. Heji 16668 M h , T 4 } t r ( = ^ 0 TNJ3O «iErDi» 16685 Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fifth month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the sixth month. Heji 16685 H B h ^t(=5E)^o - t iT 16689 Crack-making on guimao (day 40), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the sixth month. Crack-making on guichou (day 50), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the seventh month. Heji 16689 [07][HT£h I0t:(=^)*g. 4:J3o ^ I K ' 10£:(=3E)#io AJ3. 16716 Crack-making on [gui]hai (day 60), Bin divined: "[In the next] 10-day week, there will 100 be no disasters." [Day guihai was in] the seventh month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the eighth month. Heji 16716 [ O 8 ] H 0 J h j £ : t J t ( = ^ 0 ABo H H h j£: I0tr(=^)*g. Ai^o <(l=ril» 16733 Crack-making on guimao (day 40), divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eighth month. Crack-making on guichou (day 50), divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the ninth month. Heji 16733 [09]H£h tjt:(=^)^c J i M o H M K iOt:(=^)#S. i&m 16738 Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the ninth month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the tenth month. Heji 16738 [ l O j U B K fjfc: f3£:(=^)^o - M o . H ^ K fjS: *Ut:(=3c)*B. -+—-Ji . 16748 Crack-making on guichou (day 50), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the tenth month. Crack-making on guihai (day 60), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the eleventh month. Heji 16748 tjt:(=^)^o -r—Bo H E h ^ £ ( = ^ ) $ . +~^o i^M)) 16755 Crack-making on guiwei (day 20), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the eleventh month. Crack-making on guisi (day 30), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the twelfth month. Heji 16755 [ i 2 ] H E K JJ4: tltr(=^)?So -t--J!o h J£: tjt:(=^)^o - i f . 22404 Crack-making on guisi (day 30), divined: "[In the next] 10-day week, there will be no 101 disasters." [Day guisi was in] the twelfth month. Crack-making on guimao (day 40), divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the first month. Heji 22404 In Example 12, the first month occurs immediately after the twelfth month, which shows that from the first month to the twelfth month is a cycle. However, this is not the only cycle. There is inscriptional evidence showing that a thirteenth month sometimes occurs between a twelfth month and a first month. Below is one such example. Crack-making on guimao (day 40), Shi divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the twelfth month. Crack-making on guihai (day 60), Shi divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the thirteenth month. Heji 16670 Crack-making on guiwei (day 20), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the thirteenth month. Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in the first] month. Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the first month. Heji 26582 Examples 13 and 14 are proof for the occurrence of a thirteenth month between the twelfth and first month. It is apparent that in the Yin calendar from the first month to the thirteenth month is another cycle. There are two points that need to be made clear. The first point is that, from Period II to Period V , there is the expression zhengyue , 'right month.'1 Judging from the position of 1 David N. Keightley (2000: 43) translates zhengyue as "regulating moon." In his comments to the draft of this dissertation, Ken-ichi Takashima points out that such a translation is possibly incorrect. "Keightley took zheng IE to have meant 'to regulate,' which meaning is possible as a verb zheng IE, 'to rectify,' [is] wrongly written as in zhengfa tlEtfe; so zhengyue IE 16770 [ 1 4 ] H * h J5: 40t:(=^)*g o + = H E h j £ : f4jtr(= )^?So [-]Mo 102 the zhengyue, it can be deduced that it refers to the first month. The following inscriptions show that a zhengyue appears between a twelfth month and a second month: [ 1 5 ] H B h t f j g : tft:(=^)*g. K tr<£: TJJt(=A)^<. £ J E / 3 . «1=ril» 26517 Crack-making on guisi (day 30), Xing divined: "[In the next] 10-day week, there will be no disasters." [Day guisi] was in the twelfth month. Crack-making on guimao (day 40), Xing divined: "[In the next] 10-day week, there will be no disasters." [Day guimao] was in the zheng month. Heji 26517 [16]H£K triS: KJt:(=3E)#£. h frj£: fjjt:(=^c)^c £ - J3o «i=ril» 26513 Crack-making on guihai (day 60), Xing divined: "[In the next] 10-day week, there will be no disasters." [Day guihai] was in the zheng month. Crack-making on guiyou (day 10), Xing divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou] was in the second month. Heji 26513 Since inscriptions in Examples 12 and 1 show that the first month occurs between the twelfth month and the second month as well, it appears that both the first month and zhengyue occupy the same position. Moreover, i f zhengyue is not the first month of the year, i.e., the year to which the inscriptions in Examples 15 and 16 belong, the year would have to start with the second month, which would make that year incomplete. Given these two reasons, it can be deduced that zhengyue is the new designation for the first month from Period II to Period V . Since then, it has become the standard designation for the first month in Chinese calendars. FJ should accordingly be analyzed as a V + O structure, rather than Adj. + N structure. Yet, the latter analysis is most likely applicable to zhengyue l E f l , which probably should be analyzed meaning 'correct /right /just month,' not'regulating.'" 2 David Pankenier has written in his comments to this thesis that "This kind of terminology tends to be VERY conservative and resistant to change. Something pretty radical had to happen for this change to occur. How does this relate to changes in the sacrifices, if at all? Do you really have nothing to say about this? Does anyone else have anything to say worth citing?" He is right in suggesting that such a change of the designation for the first month is related to the change in sacrifices, and that there is probably something radical behind them. As Dong Zuobin (1945.1.1: 2b-4b) has made clear, this change of the terminology for the first month and changes in sacrifices, divination patterns and calligraphy of inscriptions all happened in Period II, which indicates something rather radical must have happened. Dong Zuobin suggests that it was Yin King Zujia who made these changes. Whether Dong Zuobin's opinion is correct or not is still an open question. 103 The second point worth noting is that the phrase shisan yue 'the thirteenth month,' only occurs in inscriptions from Period I and Period II. In inscriptions from Periods III to V , it disappears. The lack of this phrase in those inscriptions does not mean that there is only one cycle of twelve months, from the first to the twelfth month, in those periods. Rather, the reason for the disappearance of the phrase shisan yue is that a different method of intercalation is employed in late periods, which will be discussed in full detail in Section 3.5. The cycle of thirteen months still exists in those periods. As shown above, Yin months can form two cycles. One is twelve months long and the other one thirteen months long. The twelve-month cycle is a normal Yin year and the thirteen-month cycle an intercalary Yin year. 3.3 The Length of the Yin Month The average time for the moon to revolve round the earth is roughly twenty-nine-and-half days. Since the number of days in a month has to be a round figure, it is easy to understand that the length of a lunar month is either 29 days or 30 days. As for the length of the Yin month, however, there are differing opinions. Below, each opinion is evaluated against inscriptional evidence so as to determine which one is likely correct. 3.3.1 Yin Month Not Always 30 Days Long Liu Zhaoyang (1933: 151) and Sun Haibo (1935: 123) hold the opinion that the Yin month is normally 30 days long, starting with a jia ¥-day. If their opinion is correct, the Yin month would always start with a jia-day, as they state. It follows that their theory would be incorrect i f the Yin month does not always begin with a jia-day. A number of examples clearly show that the first day of the Yin month is not a jia-day. This kind of evidence is first presented by Hu Houxuan (1944a). He cites what is later published as Heji 26308, 26235 and 26293, as evidence that weighs against the opinion that the Yin month always starts with a jia-day. These same inscriptions are repeated by Chen Mengjia (1956: 219-220), Wen Shaofeng and Yuan Tingdong (1983: 109- 110) without giving due credit to Hu Houxuan. Below is one of those three examples first discovered by 104 Hu Houxuan. TW r. = 4^t:*g. M = i t h , <attj*: S T - J ^ o 3 m$ b , M M : * ? & t M o S i — M » i^M)) 26308 Crack-making on bingshen (day 33), Lu divined: "There will be no disasters tonight." [Day bingshen] was in the tenth month. Crack-making on dingyou (day 34), Lu divined: "There will be no disasters tonight." [Day dingyou] was in the tenth month. Crack-making on wuxu (day 35), Lu divined: "There will be no disasters tonight." [Day wuxu] was in the tenth month. Crack-making on jihai (day 36), Lu divined: "There will be no disasters tonight." [Day jihai] was in the tenth month. Crack-making on gengzi (day 37), Lu divined: "There will be no disasters tonight." [Day gengzi] was in the tenth month. Crack-making on xinchou (day 38), Lu divined: "There will be no disasters tonight." [Day xinchou] was in the tenth month. Crack-making on renyin (day 39), Lu divined: "There will be no disasters tonight."[Day renyin] was in the eleventh month. Crack-making on guimao (day 40), Lu divined: "There will be no disasters tonight." [Day guimao] was in the eleventh month. Heji 26308 There are eight inscriptions in this example. Day xinchou (38) of the sixth inscription and day renyin (39) of the seventh are consecutive. The post-faces of the sixth and seventh inscriptions clearly record that these two days belong to the tenth and eleventh months, 3 The authors of Jiaguwen heji shiwen transcribe it as the tenth month. Although the rubbing is not very clear, the present writer can discern that it is the eleventh month, and so does Chang Yuzhi (1998: 319). 105 respectively. This means that day xinchou is the last day of the tenth month, and that day renyin is the first day of the eleventh month. This is a piece of strong evidence against the view that the start of the Yin month is always a jia-day. Chang Yuzhi (1998: 320-321) believes that she has found two more examples. But only the following example is valid evidence against the opinion that the Yin month has to begin with a jia-day. [ 1 8 ] ¥ T K j&fejS: 4^-C*S. £ + - J i o £ * K MR-. <$-#tM. m$ h MM-. ^tyt:m° S T - J 3 , T M b , MR-. 4^fr?So 26306 Crack-making on jiawu (day 31), Lu divined: "There will be no disasters tonight." [Day jiawu] was in the eleventh month. Crack-making on yiwei (day 32), Lu divined: "There will be no disasters tonight." [Day yiwei] was in the twelfth month. Crack-making on bingshen (day 33), Lu divined: "There will be no disasters tonight." [Day bingshen] was in the twelfth month. Crack-making on dingyou (day 34), Lu divined: "There will be no disasters tonight." [Day dingyou] was in the twelfth month. Heji 26306 These are the four inscriptions on this piece of oracle bone. It is clearly recorded in the first two inscriptions that the divination date of the first inscription, jiawu, and the date of the second inscription, yiwei, belong to the eleventh and twelfth months, respectively. Since jiawu (day 31) and yiwei (day 32) are consecutive, it can be inferred that jiawu is the last day of the eleventh month and yiwei is the first day of the twelfth month. This is a good example illustrating that the first day of a Yin month is not a jia-day. The uncertain example cited by Chang Yuzhi (1998: 320-321) consists of the following two inscriptions scribed on Tunnan 4516. [ l 9 ] T W h ^ T ^ l l i f f i , f X . B £ h i U t i l , &-rM, ^ X o i^m 4516 On dingyou (day 34), [X] divined: "In this next tenth month, the king will attack Tong and [he will] receive assistance." 106 On jihai (day 36), [X] divined: "The king will attack Tong in this tenth month and [he will] receive assistance." Tunnan 4516 Chang Yuzhi's understanding of this example is based on Chen Mengjia's interpretation of the word sheng ^£ (1956: 117-118). Chen Mengjia has observed that the phrase jin 4* + month notation and sheng + month notation often form a pair. In such a context the word sheng always introduces a month notation that immediately follows a month notation introduced by jin, 'this.' From this observation, he infers that the word sheng in this context actually means "next." After mentioning such usage of the word sheng, Chang Yuzhi infers that the string jin sheng shiyue 4 * ^ E + ^ , 'this next tenth month,' indicates that the first divination was done on dingyou (day 34) of the ninth month; and that jin shiyue ^"tE, 'this tenth month,' indicates that the second divination was done on jihai (day 36) of the tenth month. She further deduces that the tenth month starts either with wuxu (day 35) or with jihai (day 36), neither of them is a jia-day. This is her rationale according to which Example 19 is evidence showing that the beginning of a Yin month may not be a jia-day. It is the present writer's opinion that there are uncertainties inherent in her argument. First, as cited above, Chen Mengjia has made it clear that the word sheng means "next" when it occurs in a context such as jin + month notation paired with sheng + month notation. Sheng may have a different meaning outside that context. For example, on Heji 10270, there is a string qi huo sheng lu%^k^E^, '[the king] may capture a living deer,' in which the word sheng means "living." Is there the context of a pair of of jin + month notation and sheng + month notation in Example 19? Apparently, there is not. This casts doubt on Chang Yuzhi's interpretation of sheng as "next." From this consideration, it appears that Chang Yuzhi's reading of jinsheng in Example 19 may not be accurate. Example 19 is therefore not a piece of strong evidence showing that the start of the Yin month may not be a jia-day. Beside those examples that have been discovered by other scholars, this writer has found two additional examples showing that the start of a Yin month can be a day other than a jia-day. The first example refers to the following inscriptions on Heji 26249. [ 2 0 ] f t f c - . * ? & t M o 107 [ $ ] f l h . T^JS: ^ t ^ . S t - H . « l^fc» 26249 Crack-making on wuxu (day 35), Xing divined: "There will be no disasters tonight." [Day wuxu] was in the tenth month. Crack-making on jihai (day 36), Xing divined: "There will be no disasters tonight." [Day jihai] was in the tenth month. Crack-making on gengzi (day 37), Xing divined: "There will be no disasters tonight." [Day gengzi] was in the eleventh month. Crack-making on [xin]chou (day 38), Xing divined: "There will be no disasters tonight." [Day xinchou] was in the eleventh month. Heji 26249 Day jihai (day 36) and gengzi (day 37) are two consecutive days. The post-faces of the second and the third inscription clearly show that these two days actually belong to two consecutive months: jihai in the tenth month and gengzi in the eleventh month. In other words, day jihai is the last day of the tenth month and gengzi the first day of the eleventh month. Without doubt, the beginning of the eleventh month is not a jia-day. [ 2 l ] ^ r £ h frjfc: QJEHo H f i h ftfc: Ar&tM, « f > * » 26264 Crack-making on gengxu (day 47), Xing divined: "There will be no disasters tonight." [Day gengxu] was in the first month. Crack-making on xinhai (day 48), Xing divined: "There will be no disasters tonight." [Day xinhai] was in the first month. Crack-making on guichou (day 50), Xing divined: "There will be no disasters tonight." [Day guichou] was in the second month. Heji 26264 In this example, xinhai (day 48) and guichou (day 50) belong to the first and second months, respectively. Unlike the previous examples, xinhai and guichou are not two consecutive days. There is but a single day, renzi (day 49), between them. It appears that renzi can belong either to the first or second month. If renzi belongs to the first month, it would be the last day of the first month and guichou would be the first day of the second month. If renzi belongs to the second month, xinhai would be the last day of the first month and renzi would be the first day of the second month. No matter which day the first day of 108 the second month is, renzi or guichou, it certainly is not a jia-day. The four examples cited above clearly show that renyin (day 39), yiwei (day 32), gengzi (day 37), and renzi (day 49) or guichou (day 50) are the first day of different Yin months. None of them is a jia-day, a fact that cannot be explained by the view that the Yin month is always 30 days long. It is thus incorrect to say that the Yin month must start with a jia-day and that the Yin month is always 30 days long. 3.3.2 No Y i n Month of 40 or 50 days Liu Zhaoyang (1933: 151) and Sun Haibo (1935: 123-124) reach the same conclusion, namely, that under some circumstances, 10 or 20 days might be added to a Yin month. According to such a conclusion, the length of a Yin month can be 40 or 50 days long. Edwin G. Pulleyblank suggests, in his comments to the draft of this dissertation, that there is no point in refuting this old theory that no one would take seriously nowadays. Notwithstanding, the present writer will briefly state his argument against this old theory. First, Liu Zhaoyang and Sun Haibo have produced no decisive inscriptional evidence to support their view. A l l inscriptions they have cited can be better explained by in-year intercalation. Second, they interpret adding 10 or 20 days to a month as intercalation of the Yin calendar. This is pure speculation. As will be shown in Section 3.5.3, the Yin calendar employs an intercalary month to adjust the difference between the length of the solar year and that of the Yin civil year. Third, the period of a rotation of the moon around the earth is about 29.5 days. This determines that the length of the month in the lunar-solar calendar must be approximately 29 days. As a matter of fact, it is common knowledge that the lunar month in Chinese calendars has been either 29 days or 30 days long ever since the Spring and Autumn Period. These factors make it impossible to accept the opinion that the Yin month can be 40 or 50 days long. 3.3.3 No Y i n Month of 31 Days Yabuuchi Kiyoshi's hypothesis (1956: 72-74) is that the Y i n month began upon the 109 actual observation of the crescent moon. Since the observation of the crescent moon is affected by many factors, weather being one of them, the interval between one new moon and another can be as long as 31 days. He thus suggests that it is natural for the Yin month to be 31 days long. In the section about the Yin calendar, Yabuuchi Kiyoshi (1974: 24-31) states the same opinion without offering analysis of any inscriptions. Among Chinese scholars, Wen Shaofeng and Yuan Tingdong (1983: 104-109) and Chang Yuzhi (1998: 276-281) hold a view that is identical to that of Yabuuchi Kiyoshi. A preliminary examination of inscriptions cited by Yabuuchi Kiyoshi (1956: 71-73) and Wen Shaofeng and Yuan Tingdong (1983: 104-109) has been made by Liu Xueshun (1992). Here, the present writer will make a more thorough investigation of all those examples cited by Yabuuchi Kiyoshi (1956), Wen Shaofeng and Yuan Tingdong (1983) and Chang Yuzhi (1998: 276-281). In total, they have listed nine examples to support their opinion. Each one will now be examined. Example I: [22] H f ih lOt:(=3E)#5. H E K 5uj£: *otr(=^ )#s. HE K t^:(= )^#|. H£P K TJjt:(=^)?So H5Ph fttKrftffio -f-Ro H^K t)tr(=^o H?P K ^ : t ) t (=^ 0 HE h J^JS: 10t:(=^)*S. + H J 3 „ « ^ » 26681 Crack-making on guichou (day 20), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters."[Day guisi was in] the tenth month. Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), Xiong divined: "[In the next] 10-day week, there will be no disasters." no Crack-making on guimao (day 40), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the twelfth month. Crack-making on guihai (day 60), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the thirteenth month. Heji 26681 The transcriptions above are made by Yabuuchi Kiyoshi (1956: 73). His interpretation of these inscriptions is as follows: since guisi (day 30) is in the tenth month, guichou (day 50) in the twelfth month, and guisi (day 30) again in the thirteenth month, there must be four gui-days in the tenth month: guisi (day 30), guimao (day 40), guichou (day 50), and guihai (day 60). This is accepted by Chang Yuzhi (1998: 279-280) as one of her two possible explanations of these inscriptions. However, there are some mistakes in Yabuuchi Kiyoshi's transcription. The correct divination dates of the first and fifth inscriptions are guisi and guichou, respectively. More seriously, the correct month notation of the second inscription is twelve rather than ten. It is unfortunate that Chang Yuzhi, a member of the team of the Jiaguwen heji project, has failed to check her transcription against the rubbing and correct these mistakes. Since the month notation in the second inscription is twelve, not ten, there are no direct inscriptions of the tenth month. Yabuuchi Kiyoshi's conclusion that the tenth month has four gwz-days thus becomes baseless. Wen Shaofeng and Yuan Tingdong (1983: 107) transcribe these inscriptions correctly. On the other hand, they argue that, since there are seven gwz'-days from the guisi of the twelfth month to the guisi of the thirteenth month, one of these two months must have four gui-days. Such an argument lacks certainty, not only because both Mo Feisi (1936: 303) and Chang Yuzhi (1998: 279-280) regard it as possible evidence for two consecutive intercalary months in a Yin year, but also because how to understand these inscriptions is a significant question. From the second to the last inscription, i f their order is from early to late, there would be nineteen xuns, '10-day week,' from the guisi of the twelfth month to the guisi of the thirteenth month. It is absolutely impossible for two or three months to have 190 days. i l l Without a reasonable explanation for this significant problem, any conclusion drawn from these inscriptions will remain uncertain. Example II: [23] H E j £ : fTJ -b ^ J t : ( = ^ ) ^ . AM. h m&: t J £ : ( = ^ ) # l . A ^ o l IMb, f5t:(=^)^0 AM. * h , ^ j g : t l A J 3 , 26667 [Crack-making on] gww7 (day 30), [Yi] divined: "[In the next] 10-day week, [there will be no disasters." Day guisi was in] the seventh [month]. Crack-making on guimao (day 40), Y i divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eighth month. Crack-making on guichou (day 50), Y i divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the eighth month. Crack-making on guihai (day 60), Y i divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the eighth month. Crack-making on guiyou (day 10), Y i divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the eighth month. [Crack-making on gui]wei (day 20), Y i divined: "In 10-day week, [there will be no disasters." Day guiwei was in] the ninth month. Heji 26667 Again, the transcriptions above are made by Yabuuchi Kiyoshi (1956: 73). There is one significant mistake in his transcription. In the first inscription, there is no trace for the number seven at all. The transcription provided in the Jiaguwen heji shiwen does not have the alleged number seven either. Without number seven, it is not certain that there are exactly four gwz'-days in the eighth month, because guisi could be in the eighth month as well. If that were the case, the eighth month would be at least 51 days long. Since no Yin month can be that long, and since an intercalary month would explain why so many days are present with the same month notation, this should be taken as evidence for in-year intercalation in the Yin calendar. Because of these considerations, this example cannot be considered as representing 112 evidence suggesting the existence a Yin month of 31 days. This example is also cited by Wen Shaofeng and Yuan Tingdong (1983: 104-106) and Chang Yuzhi (1998: 278-279). They have still not proven that the number seven appears on the rubbing. The present writer's analysis of this example in the previous paragraph remains valid. This is not an example of a Yin month of 31 days. Example III: [ 2 4 ] H * h , 5uj£i: lDt:(=^)*g. A J ! o H B h > Xfc-. I0tr(=^)$8. H ^ h > I0t:(=^)$. 7NJ3O H S h *0t:(=5e)*8. H E h, i u j ^ : ^ t r ( = ^ ) ^ „ H5P h, J^: t l t : ( = ^ ) ^ c H B h. tBiS: 1tJt:(=^)#S. - t J ! . H E h, J^rjT : ^ t r ( = 5 £ ) ^ 0 « ^ ) ) 26643 Crack-making on guiwei (day 20), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the sixth month. Crack-making on guichou (day 50), Da divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the sixth month. Crack-making on guihai (day 60), Da divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the sixth month. Crack-making on guiyou (day 10), Da divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the seventh month. Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Heji 26643 113 This is the third example cited by Yabuuchi Kiyoshi (1956: 71) as evidence for the Yin month of 31 days. However, he acknowledges that this example is not decisive but that it can be used as a piece of evidence for in-year intercalation. Indeed, Both Chang Yuzhi (1998: 314-315) and Liu Xueshun (1992: 5-6) offer this third example as evidence for two consecutive intercalary months in a single Yin year. The issue of the presence of two intercalary months in a single Yin year will be dealt with in Section 3.5. It thus appears that Heji 26643 is not strong evidence for the long Yin month of 31 days. Example IV: [ 2 5 ] H * h tft(=^)?So I i i i . M P h f j5 : tjtr(=^)^0 + - / 4 0 mmb, f jg : tit:(=^)?So + - ^ . 31* K fjfc: *0tr(=3E)*S. +-Jio H B h f jS: t j t ( = i ) i . -r-~Bo i^M)) 16751 Crack-making on guiwei (day 20), Zhong divined: "[In the next] 10-day week, there will be no disasters." Yong came out. Crack-making on guimao (day 40), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eleventh month. Crack-making on guiyou (day 10), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the eleventh month. Crack-making on guiwei (day 20), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the twelfth month. Crack-making on guisi (day 30), Zhong divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the twelfth month. Heji 16751 Both Wen Shaofeng and Yuan Tingdong (1983: 106) and Chang Yuzhi (1998: 277-278) point out that this bone clearly records that both guimao and guiyou belong to the eleventh month. Between that guimao and that guiyou, there are four gui-days: guimao, guichou, guihai and guiyou. Therefore, they argue that this eleventh month has four gui-days and it is 31 days long. It is true that the second and the third inscription show both guimao and guiyou belong to the eleventh month. This fact means that the eleventh month already has four gui-days. It does not mean this eleventh month has those four gw'-days only, because those inscriptions 114 do not specify that guimao and guiyou are the first and last day, respectively, of that month. So it is not certain that the eleventh month does not include any day before guimao or after guiyou. If this month does include days after guiyou and before guimao, a possibility that cannot be ruled out, it would be deemed as evidence for in-year intercalation. Because of this, one cannot be sure that the eleventh month has 31 days only. This example is not good evidence for the Yin month of 31 days. Example V: [26] H E h. ft-. *0t:(=^)*S. K ft-. ^Jt:(=?c)?So -Bo H B h ft: Tijt:(=^r:)^o H ^ h ft: *0t:(=^)#§o H S h ft-. ^Jtr(=^)^. H E h ft: t j t r (=^o EB° l&m 22404 Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the twelfth month. Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the first month. Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the second month. Crack-making on guiyou (day 10), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the fourth month. Heji 22404 The above transcriptions are those of Wen Shaofeng and Yuan Tingdong (1983: 106). Based upon these inscriptions, they have reconstructed a calendar follows. 115 Table 2: Reconstruction of the Calendar for Heji 22404 12 t h month Guisi 1s t month Guimao guichou 2 n d month Guihai guiyou [guiwei] 3 r d month [guisi] [guimao] [guichou] 4 t h month [guihai] [guiyou] [guiwei] guisi Wen Shaofeng and Yuan Tingdong argue that their reconstruction makes it clear that there are possibly four gw/-days in the fourth month. But they also acknowledge that there are other possibilities. If the second or third month has four g«/-days, then guisi could occur in the fourth month. In any event, it is their opinion that either the second, third, or fourth month has to have four gwz'-days. Thus, they regard these inscriptions as evidence of a Yin month of 31 days. However, their interpretation is not necessarily correct. There are other competing interpretations of these inscriptions. It is true that guisi is a day of the fourth month. But these inscriptions do not specify that day guisi is the last day of the fourth month. It is only one possible understanding to say guisi is the last day of the fourth month. There are other possibilities. Here is one scenario. If there is an intercalary month after the second or third month, and i f those several months from the first to the fourth month are long and short months alternately, guisi is the second day of the fourth month. Another scenario is as follows: if there is an intercalary fourth month, guisi is the second day of this intercalary month. More interpretations can be proposed. It appears that Wen Shaofeng and Yuan Tingdong's understanding at the very best is but one of several possible interpretations. This example is thus not decisive evidence for a Yin month of 31 days. Example VI: [ 2 7 ] H B K %ft-. t ( t : ( = ^ c ) ^ o -Ro HW h 1U£(=^)*S. H * K f j j t r ( = ^ o ~Eo 116 -Bo ELBo m n . « ^ # » 4 9 3 9 Crack-making on guichou (day 50), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the first month. Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the second month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiwei (day 20), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the second month. Crack-making on guisi (day 30), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the second month. Crack-making on guimao (day 40), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the third month. Crack-making on guihai (day 60), Yun divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the fourth month. Crack-making on guiwei (day 20), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the fourth month. Hebu 4939 This is a rejoined piece of Heji 16644, 16649, and 16660, done by Cai Zhemao (1999: 69, 365). The transcriptions above belong to Chang Yuzhi (1998: 276). Based upon her transcription, Chang Yuzhi argues that the second month has four gwz'-days, i.e., guihai, guiyou, guiwei, and guisi; and the second month is 31 days long. It must be pointed out that there is a serious mistake in her transcription. On the rubbings, both in Hebu and Heji, the month notation of guisi in the fifth inscription is not clear at all, and it is left blank both in the Jiaguwen heji shiwen and the Jiaguwen heji bubian shiwen. Apparently, the month notation of the guisi in the fifth inscription is uncertain. Chang Yuzhi If K l0t:(=5E)$. fJjt(=^)?S= t l t : (=5c)^o l0t:(=^)*B. 117 (1998: 276-277) has not explained why she is so sure that the month number is two not three. If that month notation is three, as Cai Zhemao (1999: 359) has transcribed, the second month would have only three gui-days: guihai, guiyou, and guiwei, and the month could not possibly be 31 days long. Therefore, Hebu 4939 does not support the view that the length of a Yin month can be 31 days. Example VII: [ 2 8 ] H M K fcHjfc: ^ - t ( = ^ ) # 5 « , EH* [H]0P h itift: TJj[t:?S]<. l ^ o i^M)> 26564 Crack-making on guiyou (day 10), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the fourth month. Crack-making on [gui\mao (day 40), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the fourth month. Heji 26564 It is clear in these two inscriptions that both guiyou and guimao are in the fourth month. Based upon this fact, Chang Yuzhi (1998: 278) suggests that between guiyou and guimao there are the following two gw/-days: guiwei and guisi. Then, she reaches a conclusion that there are four gw/-days in the fourth month and that this month is 31 days long. It is true that the inscriptions on Heji 26564 show the fourth month having four gui-days. What is not easy to see is how these inscriptions are supposed to show that the fourth month has those four gwz'-days only. It is a fact that those two inscriptions do not specify guiyou and guimao as the first and last day of that month. There is no way to exclude the possibility that this month includes days before guiyou and/or after guimao. If so, the fourth month certainly would have more than 31 days. Also, there is in-year intercalation in Period II, and Heji 26564 belongs to this period. If there is an intercalary fourth month, it is not unusual for the fourth month to have four gwi-days. This scenario cannot be ruled out. Given these two reasons, it appears rather subjective and arbitrary to assert that Heji 26564 represents evidence for the existence of a Yin month of 31 days. Example VIII: [ 2 9 ] H W i K ft: ^ t : ( = ^ ) ^ . m M E o 4 4 The inscriptions in this example are rather long. In order to save space, only relevant parts of these inscriptions are transcribed here. 118 n e i h , fc: l0t(=5E)^o SH^o H J K K J^: ^tr(=3E)^. £Biio J74: *o-t(=^ )*s. £HJ3<, 3 l ^ 3 E h JS: l0t:(=^)*B. On guiyou (day 10), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters."... [Day guiyou] was in the second month.... [It] was the king's eighth year. On guiwei (day 20), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters."... [Day guiwei] was in the third month.... On guisi (day 30), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters."... [Day guisi] was in the third month.... On guimao (day 40), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters."... [Day guimao] was in the third month.... On guichou (day 50), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." ... [Day guichou] was in the third month.... On guihai (day 60), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." ... [Day guihai] was in the fourth month.... Hebu 10958 The above transcriptions are provided by Chang Yuzhi (1998: 280-281). According to her transcriptions, guiwei, guisi, guimao, and guichou are all in the third month. Based upon this, she concludes that there are four gui-days in this third month and that this month is at least 31 days long. Her transcriptions, however, are uncertain because the month notation of the first inscription is open to different interpretations. Dong Zuobin (1948: 184) mentions that he asked Wu Jinding to check the month notation, and Wu Jinding told him it was two not three. This is followed by Chang Yuzhi, as her transcription above shows. On the other hand, both the drawing by Frank H. Chalfant and the rubbing in the Hebu show the number to be three, which is confirmed by the transcription in the Jiaguwen heji bubian shiwen. When Frank H. Chalfant made the drawing, he certainly examined the bone. In addition, it is difficult to understand how the graph "three" can appear on the rubbing i f the graph in question indeed is "two." It is probably accurate to transcribe that graph as "three." If so, the third month 119 would have five gwz'-days and its length can be 59 days. It is better to take this example as evidence for an intercalary third month; it is not an example of a Y i n month of 31 days. Example I X : [30] ¥ T K ft: &Bt: ffi ¥ f K ft-. &nt: (=^ c) m &Ht: ( = ^ n ) m Crack-making on jiawu (day 31), [X] divined: ". disasters...." Crack-making on yichou (day 2), [X] divined: " . disasters...." [Day jiawu was] in the ninth month. Crack-making on jiawu (day 31), [X] divined: ". disasters...." " . . . [In] this month, there w i l l be no disasters...." Qiu X i g u i (2002: 185) cites this example as decisive evidence for the existence of the Y i n month of 31 days. Before analyzing whether it is valid evidence for the Y i n month of 31 days, it is important to discuss the transcription of the key word of these inscriptions. The most important word in these transcriptions is yue, 'month.' In the Jiaguwen heji shiwen, what is transcribed above as zi yue wu huo $EBt! ( = ^ £ ) 'there wi l l be no disasters in this month,' is transcribed as zi xi wu huo, 'there w i l l be no disasters tonight.' If the bone graph represents the word xi, 'night,' there w i l l be no relation between these inscriptions and the length of the Y i n month. On the other hand, i f it indeed represents the word yue, it seems that they are related to the length of Y i n months. Therefore, the question is: should the bone graph be transcribed as xi or as yue! A s shown in the transcription above, the phrase zaijiuyue, ' i n the ninth month,' occurs in the second inscription. In this context, the interpretation or reading of the bone graph for < ( ¥ # ^ * » 256 . [In] this month, there wi l l be no . [In] this month, there w i l l be no . [In] this month, there wi l l be no Jiagu zhuiheji 256 yue is certain; it also is clear that this graph is scribed as B5fl. In the string zi yue/xi wu huo, the graph in question is scribed as urn. It appears that the structures o f these two graphs are the same. Therefore, the transcription zi yue wu huo is correct. 120 Dong Zuobin (1945.2.VII: 1-2) has noticed the inscriptions on what is later published as Heji 36542. He reads those inscriptions from top to bottom, which is opposite to the normal order of inscriptions on a scapula, which is from bottom to top. His reading of those inscriptions is thus inaccurate, as pointed out by Qiu Xigui (2002: 185). Qiu Xigui reads those inscriptions in the order from bottom to top. Since the first inscription is located below the second inscription, he infers that the first inscription is earlier than the second one. Further, in the post-face of the second inscription, there occurs the phrase zaijiuyue, ' in the ninth month.' He deduces that the first inscription should be in the eighth month. Moreover, these two inscriptions divine that there will be no disasters in a month. From such divination content, he infers that the divination date jiawu (day 31) and yichou (day 2) should be the first day of the eighth and ninth months, respectively, which indicates that jiazi (day 1) is the last day of the eighth month. From jiawu to jiazi, the duration of the eighth month, there are 31 days. Therefore, he considers the eighth month to be 31 days long. In Qiu Xigui's argument, there is a big assumption: the first inscription is only one month earlier than the second one. This assumption is not necessarily valid. First, among those 154,604 pieces of oracle bone discovered so far, an estimate made by Hu Houxuan (1984), there are only two pieces of oracle bone that bear inscriptions divining whether there would be no disasters in a month. It appears that such inscriptions are too few to establish a pattern that the divination about disasters in a month was done on a regular basis during the Yin Dynasty. Second, even i f the divination about disasters in a month had been done on a regular basis, it is not certain those inscriptions without month notations are records of consecutive divinations, which can be shown by inscriptions divining disasters during a 10-day week. Among those 41,956 pieces of oracle bone published in the Heji, 2,331 pieces bear inscriptions whose content is about disasters in a 10-day week (Liu Xueshun 2003: 20-21). This kind of inscription is so numerous that such divination must have been done once per 10-day week. Even so, there are many examples to show that these inscriptions on the same oracle bone are not complete records of consecutive divinations. Below is one such example: 121 H £ h 10t:(=^)#S. AJf, «1=ril» 16684 Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fifth month. Crack-making on guimao (day 40), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eighth month. Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the ninth month. Heji 16684 It is obvious that these three inscriptions belong to three different months: the fifth, eighth and ninth month. From guihai of the fifth month to guihai of the ninth month, there are 13 gw/-days, by inclusive counting method. Accordingly, there should be thirteen inscriptions on this oracle bone if the complete records of the divination about disasters in a 10-day week are scribed on this bone. But the fact is that there are only three inscriptions on Heji 16684. These three inscriptions do not comprise complete records of the divinations that were performed on a regular basis. For the same reason, it is not certain that those inscriptions in Example 30 are complete records of consecutive divinations about disasters in a month. If so, there is no inscriptional basis to consider that the first divination in Example 30 was done in the eighth month. It is therefore risky to say that the eighth month in Example 30 is 31 days long. Up to now, only nine oracle bones are alleged as evidence for the Yin month of 31 days. The analysis above shows that none of them is conclusive. This makes barely credible the view that the Yin month can be 31 days long. 3.3.4 No Yin Month as Short as 25 Days Chang Yuzhi (1998: 290-291) is the only specialist who proposes that a Yin month can be as short as 25 days. She bases her belief on the following inscription: [ 3 2 ] $ * K ^ A ^ - n f ^ ^ M o T i l . MT¥IC, T T J W A S . AJio 10976 Crack-making on xinwei (day 8), Zheng divined: "In the next eighth month, Di will order much rain." On dingyou (day 34), it rained. Till jiayin (day 51), [it rained for] 18 days. [Day xx was in] the ninth month. Heji 10976 122 After accepting Chen Mengjia's opinion (1956: 118) that the phrase shengXyue ^ . J h E often refers to the next month, Chang Yuzhi (1998: 290-292) interprets the eighth month as the next month of the divination date xinwei (day 8). In other word, xinwei is in the seventh month. In addition, she relates the phrase jiuyue, 'the ninth month,' to both dingyou (day 34) and jiayin (day 51). Then she argues that, i f xinwei is the last day of the seventh month, and if dingyou is the first day of the ninth month, then the eighth month would start with renshen (day 9) and end with bingshen (day 33); i.e., the month would be only 25 days long. There is a big problem with Chang Yuzhi's understanding of this inscription. In the verification, there are two dates dingyou and jiayin and one month notation jiuyue. As Qiu Xigui (2002: 187) suggests, one cannot say with assurance that both dingyou and jiayin are in the ninth month. From the charge of this inscription that divines much rain in the eighth month, Qiu Xigui infers that jiayin and many other days of the period from jiayin to dingyou must be in the eighth month. If so, the eighth would not be just 25 days long. The present writer agrees with Qiu Xigui's analysis of the verification of this inscription. When there is more than one date and a month notation in the verification, it is certain that the month notation is related to the latest date. Whether other dates are related to the month notation is a question that needs to be worked out case by case. This process can be demonstrated by reference to Example 33. [33]HFPh #<-j£: O^t:(=^ )*So ¥Jf ^ S f f Z a E ~£.n° M o HHK t jtr(=^o HBZ^P T E EBM^ ¥M £ 0 A i l * 5 «1 f^fc» 137 Crack-making on guimao (day 40), Zheng divined: "[In the next] 10-day week, there will be no disasters." On jiachen (day 41).... [When] that night cut into yisi (day 42)... the fifth month. [The king was] in Dun. Crack-making on guichou (day 50), Zheng divined: "[In the next] 10-day week, there will be no disasters." On the third day yimao (day 52) ... on dingsi (day 54) ... on the fourth day gengshen (day 57)... on jiachen (41).... On the fifth day wushen (day 45)..., the sixth month. [The king was] in Dun. Heji 137 5 This inscription is lengthy. In order to save space, this writer only transcribed those parts that are related to the relationship between dates and the month notation in the verification of an inscription. 123 In this example, these two inscriptions are scribed on the same piece of scapula, their divination dates only 10 days apart, and they have been made at the same place, as recorded in their verifications. It thus seems certain that the two inscriptions are records of two consecutive divinations about disasters in the next 10-day week. In the verification of the first inscription, dates jiachen (day 41) and yisi (day 42) and the month notation wuyue, 'the fifth month,' appear. Because yisi is the latest date in the verification, it is certainly in the fifth month. In the verification of the second inscription, there are the month notation liuyue, 'the sixth month,' and five dates: yimao (day 52), dingsi (day 54), gengshen (day 57), jiachen (day 41), and wushen (day 45). Among these dates, gengshen is the latest. One can say with certainty that gengshen is in the sixth month. Since yisi (day 42) and gengshen (day 57) are in the fifth and sixth month, respectively, it can be inferred that the sixth month starts with a day in the period from bingwu (43) to gengshen. But there is not enough information to determine which day is in fact the first day of the sixth month. Following Chang Yuzhi's interpretation of the verification of Example 31, one would think that all of those five dates in the verification of the second inscription of Example 32 are in the sixth month. However, because it is not certain which day among those fifteen days from bingwu (day 43) to gengshen (day 57) is the first day of the sixth month, whether wushen (day 45), yimao (day 52), and dingsi (day 54) are in the sixth month or not, remains uncertain. As for jiachen (day 41), it can be deduced from its appearance in the verification of the first inscription that jiachen is in the fifth month. Jiachen occurs in the second verification in which the month notation liuyue, 'the sixth month,' appears. If Chang Yuzhi's understanding of Example 31 is correct, jiachen would be a day in the sixth month. As analyzed in the previous paragraph, jiachen actually is a day in fifth month, which contradicts the conclusion that Chang Yuzhi would be expected to draw. By the same token, it is thus risky to say that all dates in the verification of Example 32 belong to the ninth month. The above analysis shows that, in verification, a month notation can only be related to the latest dates with certainty. Therefore, as Qiu Xigui (2002: 187) points out, not all 18 days from dingyou (day 34) to jiayin (day 51) are in the eighth month. It is a mistake for Chang Yuzhi (1998: 290-291) to interpret otherwise. There is no inscription to demonstrate the 124 existence of a Yin month of 25 days. 3.3.5 Y i n Month is 29 Days or 30 Days Long Dong Zuobin is the most eminent scholar in the field of the Yin calendar. Qne of his great contributions is that he has cited inscriptional evidence showing that the Yin month can be 29 or 30 days long. Below is an example cited by Dong Zuobin (1931: 503-504). [34]HMK #-J5: 10-t(=3E)#§. +J3o H E K l0t:(=5E)*B. + -mvb, lUt:(=3E)*S. H E K /iur:(=I:)^jo -f-HJ!. H M K f4jt:(=^)^o f E h , ijg: fjjt:(=^)^o BBo H B K M : l0t:(=3E)*8. 5H» ^ K J^t:(=^ )^ o 3£i^ o 11546 Crack-making on guiyou (day 10), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the tenth month. Crack-making on guisi (day 30), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the eleventh month. Crack-making on guimao (day 40), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eleventh month. Crack-making on guichou (day 50), Pin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the twelfth month. Crack-making on guiyou (day 10), Gong divined: "[In the next] 10-day week, there will 125 be no disasters." [Day guiyou was in] the twelfth month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the thirteenth month. Crack-making on guiyou (day 10), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the second month. Crack-making on guiwei (day 20), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the second month. Crack-making on guiyou (day 10), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the fourth month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the fourth month. Crack-making on guimao (day 40), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the fifth month. Crack-making on guichou (day 50), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the fifth month. Crack-making on guihai (day 60), Yun divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fifth month. Heji 11546 The inscriptions above are those that have month notations on Heji 11546. Based upon these inscriptions, Dong Zuobin has reconstructed a calendar for the period from the tenth to the fifth month, as shown by Table 3 on next page. Since guiyou (day 10) is both in the twelfth month and the second month, there could be only 59 days, from jiaxu (day 11) to renshen (day 9), for the two months in between, i.e., the thirteenth month and the first month. Based upon this inference, Dong Zuobin proposes that one of these two months must be 30 days long and the other one 29 days long. Dong Zuobin's analysis of these inscriptions is generally accepted in the field of the Yin calendar. This example is repeated by Chen Mengjia (1956: 219), Wen Shaofeng and Yuan Tingdong (1983: 103-104), and Chang Yuzhi (1998: 272-275). Among specialists of the Yin calendar, only Zheng Huisheng (1983: 110-111) raises questions about Dong Zuobin's interpretation of this example. He argues that those 59 days do not necessarily mean that one month is 30 days long and the other one 29 days long, because it is possible that one month is 28 days and the other 31 days. 126 Table 3: Reconstruction of the Calendar for Heji 11546 10th month Guiyou 11th month [Guiwei] Guisi Guimao 1201 month Guichou [Guihai] Guiyou 13th month [Guiwei] Guisi [Guimao]! 1st month [Guimao]! [Guichou] [Guihai] 2nd month Guiyou Guiwei [Guisi] 3rd month [Guimao] [Guichou] [Guihai] 4th month Guiyou [Guiwei] Guisi 5th month Guimao Guichou Guihai Theoretically speaking, 59 days can be divided into 31 days and 28 days; so Zheng Huisheng's argument seems possible. However, his argument totally ignores the fact that the average length of a lunar month is about 29.5 days, i.e., a lunar month is normally either 29 days or 30 days. Without this fact being the basis for research into the question of the length of the Yin month, it would be impossible to make that determination. If one follows Zheng Huisheng's logic, one will never be certain about the length of those two months. One may ask why the length of those two months is not 20 and 39 days, respectively? Why not 10 days and 49 days? Zheng Huisheng's argument is useless for making that determination; and it is for this reason that it is a bad argument. Based upon information on this oracle bone, Mark Halpern has calculated the range of the length of Yin months. Below are his calculations. Month 11 < 39 days; Month 13, 1 < 59 days, which means a month is < 29.5 days; Month 11, 12, 13, 1 < 119 days, which implies a month is < 29.75 days; Month 11, 12, 13, 1, 2, 3 < 179 days, which means a month is < 29.83 days; Month 11, 12, 13, 1, 2, 3, 4 < 209 days, which means a month is < 29.86 days. He has clearly shown that on average a Yin month is about 30 days long. If so, one should disregard Zheng Huisheng's argument and accept the above example as evidence for the view that the length of the Yin month must be either 29 or 30 days long. 127 Chang Yuzhi (1998: 283-286) claims that she has found another example of two months having 59 days. Below are relevant inscriptions. [ 3 5 ] H * K :t:j£ : tft(=^)?So Hjio H E h fjjt:(=^)^c ~Bo H l ^ K ^Jg: ^t(=5£)#i. H * K i>S: l0tr(=^)*S. EJio H E h ^tr(=5E)^. 3£i!o H?P h *0t:(=^)#J. H H h, ^ t r ( = ^ 0 Ajio H E K j^t:(=3E)^o A £ . H5P h iijS: t!t:(=^)?So A J i . H B K I J S : ^ t r ( = ^ . A ^ o H M h ^-t(=5c)^c A ^ o H * K f 4 j - t ( = ^ 0 A ^ o H E K ^t:(=^)^ 0 H£P K 1tjt:(=^)*g. H S h tltr(=^)^c +j?o HM h fjjt:(=5£)^o ( ( ^ ) ) 11546+18933+16721+16725+16752 Crack-making on guiwei (day 20), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the third month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the third month. Crack-making on guihai (day 60), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fourth month. Crack-making on guiwei (day 20), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the fifth month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be 128 no disasters." [Day guisi was in] the fifth month. Crack-making on guimao (day 40), Dun divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the sixth month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the eighth month. Crack-making on guimao (day 40), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the eighth month. Crack-making on guichou (day 50), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the eighth month. Crack-making on guiyou (day 10), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the ninth month. Crack-making on guiwei (day 20), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the ninth month. Crack-making on guisi (day 30), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the tenth month. Crack-making on guimao (day 40), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the tenth month. Crack-making on guichou (day 50), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the tenth month. Crack-making on guiyou (day 10), Dun divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the eleventh month. Heji 11546+18933+16721+16725+16752 According to the transcriptions above, guisi appears both in the fifth and eighth months. Between them, there are 59 days, which should be the length of the sixth plus the seventh month. Chang Yuzhi (1998: 286) proposes that this is an example of two months that, added together, total 59 days long, with one of them being 30 days and the other 29 days long. However, her conclusion is not on firm ground. Although Chang Yuzhi (1998: 284) still maintains that these five pieces of bone can be rejoined, the fact is that, Heji 16676 and 18933 on the one hand, and Heji 16721, 16725 and 16752 on the other hand, cannot be 129 joined together physically. Cai Zhemao (1999: 390), who suggested such a rejoining in 1984, acknowledges that those two groups of bone cannot be rejoined. If this is the situation, the fifth and eighth inscriptions may not belong to the same year, which would make it impossible to assert that the sixth month plus the seventh month total 59 days. Dong Zuobin (1931) not only produces inscriptional evidence to show that a long plus a short month total 59 days, but also presents evidence to show a Yin month of 29 days long. Heji 11485 is one such example. [36] 3 ^ K f 4 j t : ( = ^ ) ^ o - B o H * h , 10t(=3c)#i. H * b , I0t:(= )^#5. H 0 £ S ^ j l W 1 r , t o A £ . Crack-making on guihai (day 60), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the first month. Crack-making on guiwei (day 20), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the second month. Crack-making on guimao (day 40), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the second month. Crack-making on guimao (day 40), Zheng divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the fifth month. Crack-making on guiwei (day 20), Zheng divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiwei (day 20), Zheng divined: "[In the next] 10-day week, there will be no disasters." On the night of the third day yiyou (day 23), there was a lunar eclipse, and [the day became] dim. [Day guiwei was in] the eighth month. /fey/11485 Based upon dates and month notations above, Dong Zuobin (1952: 288) has reconstructed a calendar for those eight months as shown in Table 4 on next page. In this example, day guimao occurs in both the second and the fifth months. Between 130 these two guimaos, there are 59 days. Between the second and fifth months there are the third and fourth months. Chang Yuzhi (1998: 287) thus suggests that there must be a short and a long month in these two months. However, if the fact that day guiwei appears in the eighth month is taken into consideration, Dong Zuobin's reconstruction is the only one that can make guimao occur both in the second and fifth months. According to Dong Zuobin's reconstruction, there are only 29 days for the third month. Example 40 is a piece of evidence for a Yin month of 29 days. Table 4: Reconstruction of Calendar for Heji 11485 1st month [guichou] Guihai [Guiyou] 2 n a month Guiwei [Guisi] Guimao [3ra month] [Guichou] [Guihai] [4m month] [Guiyou] [Guiwei] [Guisi] 5 t h month Guimao [Guichou] [Guihai] [6m month] [Guiyou] [Guiwei] [Guisi] [7m month] [Guimao] [Guichou] [Guihai] 8 tn month [Guiyou] Guiwei Chang Yuzhi (1998: 288-290) mentions Heji 22404 as another example. [37]HEK ft-. Kjt:(=^ )#J. -r-JT %W K ft: T 3 t ( = ? G ) ^ o - B o HflK ft-. f3jt:(=^)^0 H^K ft-. lut:(=5£)*8. l&m 22404 Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the twelfth month. Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the first month. Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, there will be no disasters." 131 Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the second month. Heji 22404 It is clear from these inscriptions that day guisi belongs to the twelfth month while guihai belongs to the second month. Between these two dates, there are only 29 days. Between the twelfth and the second month is the first month. Therefore, this first month is 29 days long. David N . Keightley (personal communication dated May 20, 2002) notices that the eighth month on Heji 16706 is 29 days long. [38 ]HEh lUtr(=^)?Ko ±Bo 3S£K ^t(=?G)?8o KB* Crack-making on guisi (day 30), [Zheng] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the seventh month. Crack-making on guihai (day 60), [Zheng] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the ninth month. Heji 16706 Between the guisi of the seventh month and guihai of the ninth month, there are 29 days. Between the seventh and ninth month is the eighth month. The eighth month is 29 days long. Chang Yuzhi (1998: 287-288, 292-293) provides another two valid examples. The first example consists of the following inscriptions: [ 3 9 ] H B [ M , I10 t (=^)$o £+Jio H ^ h ItJt(=A)?So £ - h 8 „ rjT: itJt:(=5c)?S<= H E h J* : iiO£:(=^)#§. S + J i X - o 2627+ 37970+37974 [Crack-making] on guichou (day 50), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guichou] was in the tenth month. Crack-making on guihai (day 60), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guihai] was in the tenth month. 132 Crack-making on guiyou (day 10), [X] divined: " A s for the king, [in the next] 10-day week, there w i l l be no disasters." [Day guiyou] was in the eleventh month. Crack-making on guiwei (day 20), [X] divined: " A s for the king, [in the next] 10-day week, there w i l l be no disasters." [Day guiwei] was in the eleventh month. Crack-making on guisi (day 30), [X] divined: " A s for the king, [in the next] 10-day week, there w i l l be no disasters." [Day guisi] was in the twelfth month. Crack-making on guimao (day 40), [X] divined: " A s for the king, [in the next] 10-day week, there w i l l be no disasters." [Day guimao] was in the twelfth month. On guichou (day 50), the king made cracks and divined: " A s for the king, [in the next] 10-day week, there w i l l be no disasters." [Day guichou] was in the twelfth month. Yingcang 2627 + Heji 37970 + Heji 37974 Chang Y u z h i (1998: 292-293) has rejoined those three pieces of oracle bone. She points out that since day guihai (day 60) occurs in the tenth month, and guisi (day 30) occurs in the twelfth month, there are only 29 days between them. Between the tenth and twelfth months is the eleventh month. So the eleventh month must have 29 days only. Another example cited by Chang Yuzhi (1998: 287-288) is Heji 26682 that bears the following inscriptions. [40]Hfih. Djfc: 1 i J t : ( = ^ ) # § . UBo H B h M f t : tftr(=3E)^. f i B o h ^ t : ( = ^ o H W K tBj£: f5tr(=^)^0 h ttjg: f4jt(=?C)?So l±ij5: lot:(=^)*8. -Bo i ^ m 26682 Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, there w i l l be no disasters." [Day guichou was in the Y ] month. Crack-making on guimao (day 40), X i o n g divined: "[In the next] 10-day week, there w i l l be no disasters." [Day guimao was in] the ninth month. Crack-making on guichou (day 50), Zhu divined: "[In the next] 10-day week, there w i l l be no disasters." [Day guichou was in] the ninth month. 133 Crack-making on guihai (day 60), Zhu divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiyou (day 10), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the tenth month. Crack-making on guimao (day 40), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiwei (day 20), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the first month. Heji 26682 In this example, day guichou occurs in the ninth month and guiwei in the first month. Between these two months, there are the tenth, eleventh, and twelfth months. Since the length of the lunar month is about 29.5 days, the length of these three months would be approximately 88.5 days. On the other hand, between day guichou and guiwei, there are 89 days. Therefore, the length of these three months is 89 days and can be divided into three parts: 30 days, 30 days, and 29 days, which correspond to two months of 30 days and one month of 29 days. Heji 26682 is thus an example of the Yin month of 29 days. Chang Yuzhi (1998: 295) alleges that there are three more such examples. An examination of relevant inscriptions shows that they are not valid evidence for the existence of a Yin month of 29 days. Each of Chang Yuzhi's three examples will be analyzed, starting with Heji 37893. Below are Chang Yuzhi's transcriptions (1998: 293-295): Crack-making on guiwei (day 20), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guiwei] was in the second month. Crack-making on guimao (day 40), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guimao] was in the third month. Crack-making on [gui]hai (day 60), [X] divined: "[As for the king, in the next 10-day week], there will be no disasters." Crack-making on [guiwei] (day 20), [X] divined: "[As for the king, in the next 10-day week], there will be no disasters." [Day guiwei was in] the fifth month. Heji 37893 37893 134 Chang Yuzhi offers two interpretations for this example. First, she points out that day guiwei appears in both the second and fifth months. In order to provide the longest period between these two months, i.e., the third and fourth months, she assumes that guiwei is the last day of the second month and the first day of the fifth month, respectively. Even so, she finds that there are only 59 days between the guiwei of the second month and that of the fifth month, which means that the longest time period for the third and fourth month is 59 days, from which she deduces that one of the those months must be 29 days and the other 30 days long. At first glance, her interpretation seems logical. However, there is a serious mistake in her transcriptions. On the rubbing of Heji 37893, the divination date of the last inscription is almost entirely lost. The authors of the Jiaguwen heji shiwen do not transcribe that date. After examining the rubbing, the present writer thinks that date is more likely guimao than guiwei, because there are two dots that may be the ends of two vertical strokes. In any event, to identify the divination date of the inscription of the fifth month as guiwei is far from certain. There is no solid basis for Chang Yuzhi to say that these inscriptions show that there are only five gui-days for the two months in question. Therefore, Heji 37893 is not evidence for a Yin month of 29 days. Chang Yuzhi's second interpretation is that the guiwei in the second month is not the last day of that month. In that case, there would be only four gui-days in the third and fourth months. Both months would have only two gwz'-days, i.e., both are short months of 29 days each. The longest period of four gui-days is 49 days, which is pointed out by Qiu Xigui (2002: 184). That is nine days shorter than the time period for two lunar months. Her calculation is simply wrong. Therefore, Chang Yuzhi's second interpretation cannot possibly be correct. The second alleged example mentioned by Chang Yuzhi (1998: 282-283) is Heji 557, which bears the following relevant inscriptions. [42] H * b, ft: 'tomm-bo -Bo ¥ T K ft: IS,f?I§o ft: Wo -Bo T * h ft: TT£O ~BO H H K ft: T ^ f f f l B o 135 j£: T S # W * ) ( 6 O «f>*» 557 Crack-making on guiwei (day 20), [X] divined: "It should not be Niao whom [the king will] order." [Day guiwei was in] the first month. Crack-making on jiawu (day 31), [X] divined: "[The king will] perform the jyz'-sacrifice again; upon jiayin (day 51), [he will] perform the >>0w-cutting sacrifice." [X] divined: "[The king] should not [perform the ^/-sacrifice] again; [he will] perform the .vow-sacrifice." Crack-making on dingwei (day 44), [X] divined: "It is not [...]." [Day dingwei was in] the second month. Crack-making on guichou (day 50), [X] divined: "To small ancestor tablets, [the king will] offer Qiang." [X] divined: "Do not offer." [Day guichou was in] the second month. Crack-making on guiyou (day 10), [X] divined: "Perhaps, [there will be someone] bringing alarming news from Qin." [X] divined: "[There will] not [be someone] bringing alarming news from Qin." [Day guiyou was in] the eleventh month. Heji 557 Based upon her transcriptions above, Chang Yuzhi (1998: 283) lists that the first month has guiwei (day 20) and jiawu (day 31), the second month dingwei (day 44) and guichou (day 50), and the eleventh month guiyou (day 60). Then she asserts that there are three gwz'-days (guiwei, guisi and guimao) in the first month and that guichou is the first gwz'-day of the second month. She further calculates that, from the second to the eleventh month, if each month had three g«/-days, the eleventh month would not have guiyou; if one of the nine months from the second to the tenth month had only two gw/-days, guiyou could appear in the eleventh month. Therefore, she believes that there is one month of two gwz'-days among the nine months. Her argument is problematic. First, she misreads the crack notation yi, 'the first,' as a month notation. There is no inscription that indicates jiawu is in the first month. Second, even if both guiwei and jiawu are in the first month, those two dates do not imply the first gui-day of the second month is guichou, because it can possibly be guimao. If, from the 136 second to the eleventh month, each month has three gui-days, guiyou would have appeared in the eleventh month. There is no need to assume one of these months must have only two gui-Because Chang Yuzhi's transcriptions are mistaken and there is no need to assume one of those months has two gw/-days only, it is clear that Heji 557 is not evidence for the existence of a Yin month of 29 days. Heji 454 is the third invalid example cited by Chang Yuzhi (1998: 290). She cites the following three inscriptions from Heji 454. sacrifice]." [The king] indeed used [them in sacrifice]. "[The king] should not use Qiang [in sacrifice] upon the next day jiachen.'''' [It was in] the third month. Crack-making on xinwei (day 8), Nan divined: "Lady Nu's childbirth will be blessed." The king made prognostication and declared, "[If] it should be a geng-day [that she] gives birth, it will be blessed." [Day xinwei was in] the third month. Heji 454 Based upon her statistics that the word yi 3! often refers to the next day, Chang Yuzhi supplies guimao as the divination date for the first two inscriptions. Then, she infers from the postscript of the second inscription that guimao is in the third month. The month notation sanyue, 'the third month,' appears in the post-face of the third inscription as well, from which she infers that xinwei is in the third month. From guimao (day 40) to xinwei (day 8), there are 29 days. She thus reaches a conclusion that that third month has at least 29 days. She acknowledges that she is not sure if the third month may be longer than 29 days. Therefore, this is strong evidence to point to a short Yin month of 29 days. Dong Zuobin (1931: 504-505) tries to find inscriptional evidence for a long Yin month 6 In his comments on the last draft of this dissertation, Ken-ichi Takashima raises the following question: "How do you know it was the king?" The basis for my supplying "the king" is that the vast majority of the Yin OBI is remains of the Yin royal house. Generally speaking, divinations are made on behalf of Yin kings. Because of this, when the subject of a charge is missing, it is likely to be a Yin king, unless it can be deduced otherwise. days only. 137 of 30 days. The example he cites is what is published later as Heji 339. On this oracle bone, there are the following inscriptions relevant to the length of Yin month: [ 4 4 ] M H h gjfc: STWrfTTo J*: ^jfrlTTo £ J T T5P h. £ h*^ > tr^o h ^ J * : ^ H f r l T T o A J T i f h. ItltT^Eo J5: ^ E „ A i T T E h, ^ J S : f t T T - T o A J T h. ST£PffTT° $ * h. ^ J ^ : H B hi <£: ^ t t J a ^ f A f ^ . j£: A J T ¥Wh, JS: S Z ^ p i i f f , UTAo -S: ^Jfrl^, % c A J T Za^Ph, J £ : l U f ^ i f A « fflo A J T ¥ T h: S S f r l T l l ^ c ((^*)) 339 Crack-making on bingyin (day 3), Bin divined: "Next day dingmao (day 4), [the king] will make an offer to Ding." [X] divined: "[The king] should not make an offer to Ding:" [Day bingyin was in the] fifth month. Crack-making on dingmao (day 4), Bin divined: "In this season divination does not yield 138 results,7 there will be no harm." [Day dingmao was in the] fifth month. Crack-making on dingwei (day 44), Bin divined: "Today, [the king will] make an offer to Ding." [Day dingwei was in the] sixth month. Crack-making on renzi (day 49), Bin divined: "Dun and Bei will not die." [X] divined: "Perhaps, [they will] die." [Day renzi was in the] sixth month. Crack-making on dingsi (day 54), Bin divined: "[The king will] offer Ding one ox." [Day dingsi was in the] sixth month. Crack-making on bingyin (day 3), Zhong divined: "Next day dingmao (day 4), [the king will] make an offer to Ding." [X] divined: "[The king] should not offer." [Day bingyin was in the] seventh month. Crack-making on xinwei (day 8), Bin divined: "Next day, the king will chase." [X] divined: "[The king will] order [somebody] to chase." [Day xinwei was in the] seventh month. Crack-making on guichou (day 50), [X] divined: "[The king will] order Jian to fetch Qi and ten people from X (word unknown)." [X] divined: "[The king] should not order." [Day guichou was in the] eighth month. Crack-making on jiayin (day 51), [X] divined: "Next day yimao (day 52), [the king will] X (word unknown) ten oxen and ten Qiang people." [X] divined: "[The king] should not offer Qiang. It should be oxen [that he should offer]." [Day jiayin was in the] eighth month. Crack-making on yimao (day 52), [X] divined: "[The king will] X (word unknown) ten oxen and ten Qiang people." [This divination was] adopted. [Day yimao was in the] eighth month. [X] divined on jiazi (day 1): "Next day, [the king will] make an offer to Zu Y i . " Heji 339 Based upon the above dates and month notations, Dong Zuobin reconstructs the calendar for those four months as shown in Table 5. Dong Zuobin's reconstruction for those four months is based upon three crucial dates 7 For an example of the usage of the word suity as "season," see Heji 24225. The word xing¥< means to "get up" in classical Chinese. But it does not make much sense to say the divination "does not get up." What the string bu bu xing V ^ 7N means, it is surmised, is that the divination does not yield results. 139 and their month notations: bingyin of the fifth and seventh months and jiazi of the eighth month. It is unfortunate that there is a fatal mistake in his transcription of these inscriptions: no post-face states that jiazi is in the eighth month, and no inscription indicates jiazi is in the eighth month. The authors of the Jiaguwen heji shiwen place jiazi between dingsi in the sixth month and xinwei in the seventh month, which suggests that jiazi is in one or the other of these two months. If jiazi is not in the eighth month, there would be no crucial date to make his reconstruction the only one for those four months. Accordingly, Heji 339 is not a good example to show that a Yin month is exactly 30 days long. Table 5: Reconstruction of Calendar for Heji 339 5 t h month (long) bingyin (1 s t day) dingmao (2n d day) xinwei (6th day) [bingzi] (11m day) [bingxu] (21st day) 6 t h month (long) [bingshen] (1s t day) [bingwu] (11th day) dingwei (12th day) renzi (17th day) [bingchen] (21st day) dingsi (22nd day) 7 th moth (short) bingyin (1s t day) xinwei (6th day) [bingzi] (lTday) [bingxu] (21st day) 8 th month (long) [yzwez] (1s t day) [yisi] (11th day) , , guichou (19 day) rr jiayin (20 day) yimao (21st day) jiazi (30th day) The best inscriptional evidence to demonstrate that a Yin month is 30 days long is Heji 24440, a piece of oracle bone already cited in Section 1.2 of this dissertation. This oracle bone bears the following inscription. [45] £ - I E H # * o ¥ f > ZliEL jftJI, B B , i f , H M> M f , T f l , SFP> i ) I , $E> f f , H> £ m> j £ f > E H , M > $FP, f JI> HEo ¥ f ^ £ 5fc, T I > J f t j £ , EJu fW> H£P. ¥ J t , M f , 140 T * N EW> MR, i f , H S , £ £ P , S I . T E N fJc f> $ 1 , H « ^ » 24440 Month/one/right/ called eating wheat: y'/az/, yichou, bingyin, dingmao, wuchen, jisi, gengwu, xinwei, renshen, guiyou, jiaxu, yihai, bingzi, dingchou, wuyin, jimao, gengchen, . xinsi, renwu, gui, jiashen, yiyou, bingxu, dinghai, wuzi, jichou, gengyin, xinmao, renchen, guisi. The second month [called] father xuan: jiawu, yiwei, bingshen, dingyou, wuxu, jihai, gengzi, xinchou, renyin, guimao, jiachen, yisi, bingwu, dingwei, wushen, jiyou, gengxu, xinhai, renzi, guichou, jiayin, yimao, bingchen, dingsi, wuwu, jiwei, gengshen, xinyou, renxu, gui. Heji 24440 The above, inscription is not a record of divination. It appears to be a copy of the calendar table of the first and second month of a Yin year. What is relevant here is that this inscription indeed shows that the length of the first month is 30 days. As for the second month, since the ganzhi date of its last day is not complete, it is not certain whether the second month does have 30 days. In any event, this inscription at least records a Yin month that is exactly 30 days long. It is the strongest inscriptional evidence for the existence of a Yin month of 30 days. Qiu Xigui (2002: 181-183) suggests the followings as an example of Yin months of 30 days. [46] H B K ft [MB tl(=3c.)ffi] . £&b, ft-. &Bt:(=%)® ¥W K ft: MRtXrft)® ft: &Btl(r3Gffl *5P K ft: &£t:(=3c)#S • • K ft-. wikWi^mm 3 1 5 Crack-making on guisi (day 30), [X] divined: "... [In this month, there will be no disasters.] Crack-making on renxu (day 5 9 ) , [X] divined: "... [In] this month, there will be no disasters...." Crack-making on xinmao (day 28), [X] divined: "... [In] this month, there will be no disasters...." 141 Crack-making on xinyou (day 58), [X] divined: "... [In] this month, there will be no disasters...." Crack-making on xinmao (day 28), [X] divined: "... [In] this month, there will be no disasters...." Crack-making on [XX], [Y] divined: "... [In] this month, there will be no disasters Jiagu zhuihe xinbian 315 i n Qiu Xigui's opinion (2002: 183), these inscriptions are records of consecutive divination about the auspiciousness of a whole month. Based upon this assumption, he proposes the following lengths for those months on this bone: Month I: from guisi (day 30) to xinyou (day 58), 29 days, short month. Month II: from renxu (day 59) to gengyin (day 27), 29 days, short month. Month III: from xinmao (day 28) to gengshen (day 57), 30 days, long month. Month IV: from xinyou (day 58) to gengyin (day 27), 30 days, long month. While Qiu Xigui's reconstruction above seems tidy, there are some problems with his interpretations. Example 30 in Section 3.3.3, i.e., Jiagu zhuihe ji 256, is very similar to this example. That section has already analyzed why such inscriptions cannot be taken as complete records of consecutive divinations about each month's auspiciousness. That analysis applies here. It still is risky to take this example as evidence for the length of the Yin month. In this section, inscriptional evidence for the length of the Yin month has been examined. Among them, Heji 11485, 16706, 22404, 26682 and the rejoined piece of Heji 37970 + 37974 +'Yingcang 2627 are clear evidence for the existence of a Yin month of 29 days. Heji 24440 is strong evidence for the existence of a Yin month of 30 days. Heji 11546 is evidence demonstrating that two Yin months are 59 days long. Al l these pieces of evidence lead to the conclusion: the Yin month can be either 30 or 29 days long. 3.3.6 No Reference to Dayue or Xiaoyue In Chinese calendars since the Qin Dynasty, the month of 30 days has been called dayue ~XB, 'long month,' and the month of 29 days has been called xiaoyue 4sB, 'short month.' In these two expressions, both da and xiao are adjectives that modify yue. As evidence in 142 Section 3.3.5 shows, the Yin month is either 30 days (long month) or 29 days (short month). There are long and short months, then, but are they called dayue and xiaoyue in the language of the Yin OBI? It is Chang Yuzhi's opinion (1998: 275, 282) that there are indeed references to dayue and xiaoyue in the OBI. She cites the following inscriptions as direct references to xiaoyue. The word-by-word translations below do not make complete sense, which indicates that there are difficulties with Chang Yuzhi's interpretation of the references to 'short month.' [47] I / H - h ^ [ A ] f i o 7790 * [X] divined: "The king [will enter] Shang in the short next seventh month." Heji 7790 [48] j£: i ^ M A f l o « ^ * » 7791 * [X] divined: "The king will enter Shang in the short next seventh month." Heji 7791 [49]£)W h JS: '.hSJiSfcW*. «1^fc» 21637 * Crack-making on renyin (day 39), [X] divined: "In the short fifth month, we will hold a [religious] service." Heji 21637 Chang Yuzhi interprets xiao wuyue and xiao sheng qiyue as "the short, fifth month" and "the short, next, seventh month," respectively. That is the reason for her to regard them as direct references of the xiaoyue, 'short month,' in the OBI. However, there are several problems with her interpretations. First, if the string xiao sheng qiyue / J N ^E~b^ l indeed means 'the next, short, seventh month,' it should be written as *sheng xiao qiyue ^ E / J N ' h ^ . The difference between these two strings is shown below: Xiao sheng qiyue Sheng xiao qiyue Short next seventh month next short seventh month 143 If the word xiao means 'short' and is a modifier for qiyue, the structure of the expression for 'the next, short, seventh month' should be like t^hat one on the right. But what appears on the surface level in the examples above is that on the left. Such structure of the string xiao sheng qiyue indicates that xiao does not modify qiyue and that it does not mean "short." Second, the complete context for the string xiao wuyue, as cited in Example 50, implies that the word xiao does not mean short. [50] i W K T,£i!fSW*o Crack-making on renyin (day 39), [X] divined: "Xiao, in the fifth month, we will hold a [religious] service." "Not xiao." Heji 21637 The word xiao often means 'small.' Such a meaning, however, does not fit the context of Example 50. In these two inscriptions, xiao and fu xiao, form a pair of complementary charges. Between the two, the second inscription is the negative charge with fu being its negative. According to Ken-ichi Takashima (1996a: 365), fu is a *p-type negative that "negates verbs whose salient feature is their 'uncontrollability' - that is, verbs expressing actions which are beyond the control of living persons." It appears that the word xiao in this context should be taken as an uncontrollable verb. If the word xiao is an uncontrollable verb, what does that imply? There is no answer to this question yet, which is the reason why the present writer cannot translate xiao. The fact is that the occurrence offu before xiao in the second inscription of Example 50 shows that xiao is a verb. It is not the adjective xiao. Because xiao is not an adjective, it is impossible that xiao means "short." If so, it is not related to the length of the fifth month. The foregoing analysis shows that it is problematic to interpret xiao wuyue and xiao sheng qiyue as 'the short, fifth month' and 'the next, short, seventh month,' respectively. There is no evidence to support Chang Yuzhi's view. As for references to dayue in the OBI, the following inscriptions are cited by Chang Yuzhi (1998: 275) as her evidence. The word-by-word translation below appears ungrammatical, which indicates a possible problem with taking them as references to dayue in the Yin OBI. 144 [51] JS: M H ^ I O «1=ril» 12528 * [X] divined: "In the big this third month, it will rain." Heji 12528 [ 5 2 ] * ^ - 8 J 3 T K M o « ^ * » 12529 * "In the big this second month, it perhaps will not rain." Heji 12529 Both "big this second month" and "big this third month" are literal translations. Professor Takashima's comment is that they are impossible English, i.e., they cannot be rendered into grammatical English. To some degree, the absurdity of the translation shows the awkwardness of interpreting those expressions as references to dayue. In this writer's opinion, the grammatical analysis of the string xiao sheng qiyue is applicable to both da jin sanyue and da jin eryue. From the point view of grammar, da jin sanyue and da jin eryue cannot mean 'this big third month' and 'this big second month.' They are not references to dayue in the OBI. Chang Yuzhi is the only scholar who suggests that there are direct references to the dayue and xiaoyue in the Yin OBI. The analysis above shows that her interpretations of all those relevant inscriptions are problematic. To date, no direct references to either dayue or xiaoyue have been found in the OBI. 3.4 The Commencement of the Yin Month With regard to the commencement of the Yin month, there are three theories: the Yin month starts with a jia-day ¥ B,fei IUJ, or shuo The first theory is proposed by Liu Zhaoyang (1933: 151) and followed by Sun Haibo (1935: 123). Section 3.3.1 cited inscriptional evidence to show that the Yin month does not always start with a jia-day. This theory turns out to be incorrect. There also are problems with the second theory that was proposed by Yabuuchi Kiyoshi (1956: 72). It is his opinion that the Yin had not reached the stage of compiling a prescriptive calendar, and that instead the Yin month started with actual observation of the new moon. His opinion is followed by Zhang Peiyu, Lu Yang and Xu Zhentao (1984: 70), Chang Yuzhi (1998: 324-340) and David N . Keightley (2000: 43). Since the Chinese term for the 8 The authors of the Jiaguwen heji shiwen transcribe this graph as san, 'the third,' rather than er, 'the second.' Since the graph in question represents the word er on the rubbing of Heji 12529, Chang Yuzhi's transcription is correct. 145 observable new month is fei, 'the first waxing crescent,' Yabuuchi Kiyoshi's theory can be paraphrased as follows: the start of the Yin month is fei. Those who believe in the second theory assert two reasons in support of their position. First, Yabuuchi Kiyoshi and other scholars assume that the Yin people were not able to calculate the time of new moon, a position clearly articulated by Chang Yuzhi (1998: 324). The Yin people, whose astronomy was not developed enough, could not possibly calculate the time of the new moon. These scholars, however, do not specify how developed astronomy must be for early people to be able to calculate the time of the new moon. This position thus appears rather subjective. For example, to Chang Yuzhi (1998: 324), Chinese people were still not able to calculate the new moon in the Western Zhou Dynasty. But the fact is that the following passage in the chapter "Shiyue zhi jiao B Z7^" of the Shijing demonstrates that the new moon was already the start of a month in Western Zhou: ^ 0 « o m^z, miZB.o At the conjunction [of the sun and month] in the tenth month, On the first day of the month, which was xinmao, The sun was eclipsed, A thing that is a very evil omen. (Legge 1872a: 320-321) It is the standard view that this chapter of the Shijing was written in the Western Zhou, as James Legge (1872a: 321) states clearly: "Ll-3 give us a certain date for the composition of this ode, and determine it as belonging to the reign of king You [of the Western Zhou]...." In this piece of contemporary record, new moon is already been used as the start of that tenth month. It is the proof that the people in Western Zhou were already able to calculate the new moon, even though Chang Yuzhi does not think that the astronomy at that time was advanced 146 enough for them to do so. Clearly, Chang Yuzhi has drawn an inappropriate conclusion. By the same token, Chang Yuzhi's and others' assumption that the Yin astronomy was not developed enough for people to calculate the new moon does not mean that the Yin calendar did not start with the new moon. This assumption is not sufficient justification for their theory ihatfei is the start of the Yin month. Their second reason derives from the fact that in some ancient calendars, such as the Babylonian, Hebrew and Greek calendars, the month started with actual observation of the new crescent moon. This historical fact is not sufficient justification for their position either. The start of the lunar month may differ in ancient calendars. For example, consider the beginning of a day: "For 'days' have not always begun at the same time. The Babylonians began the day at sunrise; the Jews and Greeks, at sunset; the Romans, at midnight" (Dubs 1951: 330). If the start time for the day in those calendars was so different one from the other, how can one be sure that all ancient calendars had the same commencement of the month? Moreover, none of those scholars who hold the second theory has produced clear-cut inscriptional evidence to show the Yin month actually starts with fei. To date, there is no solid basis to say the Yin month begins with actual observation of the new crescent moon. Specialists such Dong Zuobin, Chen Mengjia and Feng Shi hold the view that the Yin month begins with shuo, 'new moon.' Among these scholars, Feng Shi (1990) produces important research about the commencement of the Yin month. He distinguishes the astronomical shuo from shuo as the beginning of the Yin month. The astronomical shuo can be calculated to seconds, while shuo marking the beginning of the Yin month refers to one of the two days when the moon cannot be observed. The astronomical new moon is thus much more precise than the new moon defined as the start of a lunar-solar month. It is Feng Shi's opinion (1990: 155) that, based on the observation of the moon, early people could choose one of the two days when moon was not visible as the start of a month. More importantly, Feng Shi (1990: 149-154) reconstructs the commencement of Yin months based upon the absolute date of the lunar eclipse on yiyou. His approach is in agreement with the synchronic evidential approach. Feng Shi's discussion about those inscriptions of the yiyou eclipse indicates that he was not aware of the work of Dong Zuobin (1952) that was previously evaluated in Section 3.3.5. Dong Zuobin's reconstruction is the only one that can accommodate all dates and month 147 notations recorded in the inscriptions of this eclipse. We should make best use of it by citing Dong Zuobin's reconstruction below: 1st month: from yisi (day 42) to guiyou (day 10), short month. 2 n d month: from jiaxu (day 11) to guimao (day 40), long month. 3 r d month: from jiachen (day 41) to renshen (day 9), short month. 4 t h month: from guiyou (day 10) to renyin (day 39), long month. 5 t h month: from guimao (day 40) to xinwei (day 8), short month. 6 t h month: from renshen (day 9) to xinchou (day 38), long month. 7 month: from renyin (day 39) to gengwu (day 7), short month. 8 th month: from xinwei (day 8) to... The astronomical new moons that correspond to the above Yin months are as follows:9 Table 6: New Moons in 1228 BC and 1227 BC Yin month Gregorian date Ganzhi date Astronomical new month 1st month Oct. 22, 1228 BC Jiachen (day 41) 13:36 2 n d month Nov. 21, 1228 BC Guiyou (day 10) 0:35 3 r d month Dec. 20, 1228 BC Guimao (day 40) 12:21 4 t h month Jan. 19,1227 BC Renshen (day 9) 0:43 5 t h month Feb. 17, 1227 BC Renyin (day 39) 13:41 6 t h month Mar. 19, 1227 BC Renshen (day 9) 3:24 7 t h month Apr. 17, 1227 BC Xinchou (day 38) 17:53 8 th month May 17, 1227 BC Xinwei (day 8) 9:8 A comparison of ganzhi dates in this table and those ganzhi dates of Yin months reconstructed by Dong Zuobin shows that six Yin months begin one day after the astronomical new moon and two Yin months begin with the astronomical new moon. It appears that the start of the Yin month is consistently closely related to the astronomical new 9 These specific times for new moons are cited from Zhang Peiyu (1990: 475). Their ganzhi dates are based on the view that the Yin day started with su, which corresponds to nautical twilight in the morning. 148 moon. On the other hand, it is common knowledge in the field that the observable new crescent moon is two days after the astronomical new moon when a lunar month is a long one, or three days after the astronomical new moon when a lunar month is a short one. In Dong Zuobin's reconstruction, none of the eight Yin months starts two days after the astronomical new moon. This contradicts the theory that the Yin month starts with fei, 'the first crescent new moon.' In short, the reconstruction of first days of those eight Yin months related to the lunar eclipse on yiyou demonstrates that the start of the Yin month is consistently closely related to the astronomical new moon. It is strong evidence for the view that the Yin month starts with shuo, 'new moon.' 3.5 The Arrangement of Y i n Months A Yin month is either 30 days or 29 days long. In a lunar calendar, the order most often seen is that long and short months occurring alternately. In Section 3.3, there are some examples where a long and a short Yin month appear alternately10. Are there other orders such as consecutive long Yin months or consecutive short months? Also, the Yin calendar is a lunar-solar calendar. It has to employ intercalation to adjust the difference between the 1 0 Chang Yuzhi (1998: 295-297) cites Heji 6 as new evidence for the alternation of short and long Yin months. However, she has made a mistake in calculation. The following are relevant dates on that oracle bone: The 4th month jiyou (day 46) The 5,h month jiaxu (day 11) yihai (day 12) dingchou (day 14) The 6th month xinmao (day 28) gwi'si(day30) jiawu (day 31) yiwei (day 32) dingyou (day 34) guimao (day 40) The 7th month guihai (day 60) yichou (day 2) xinwei (day 8) wuyin (day 15) guiwei (day 20)? The 8,h month gengyin (day 27) xinmao (day 28) It is true that xinmao appears in both the sixth and eighth month and that there are 59 days between them. This does not mean that the total number of days in the sixth and seventh months, is 59, as Chang Yuzhi suggests. Only after establishing that both the sixth and eighth months indeed start with xinmao is it certain that the sixth and seventh months together total 59 days. However, there are no inscriptions on Heji 6 that show the first day of the sixth and eighth month is xinmao. There is no basis to say that the number of days in the sixth and seventh months totals 59 and that one of these months is a long month and the other a short one. Therefore, Heji 6 is not evidence for alternation of a long Yin month with a short Yin month. 149 length of a Yin civil year and that of a solar year. Where does one put the intercalary month in an intercalary year? These are issues to be addressed in this section. 3.5.1 Consecutive Long Y i n months The appearance of consecutive long months is not unusual in lunar-solar calendars. For instance, the first, second, fourth, fifth month in the Chinese year 2003 are all 30 days long. They are two examples of consecutive long months in one year. If consecutive long months occur in the Yin calendar, it should not be a surprise. Dong Zuobin (1931: 505) has touched on this issue when he demonstrates the length of a long Yin month, suggesting that there are consecutive long Yin months in the Yin calendar. Relevant inscriptions are those of Example 44 in Section 3.3.5. As analyzed in that section, Dong Zuobin transcribes one crucial date incorrectly. Accurate transcriptions of Example 44 neither confirm nor deny the existence of consecutive long Yin months. Chang Yuzhi (1998: 297-299) also cites Example 44 as evidence for the occurrence of consecutive long Yin months. Her argument is as follows: because bingyin (day 3) appears in both the fifth and seventh month, the number of days in the fifth and sixth month, totaled together, must be 60 days. Therefore, she reaches the conclusion that the fifth and sixth months are both 30 days long and offers her conclusion as an example of two long months appearing consecutively. It has already been pointed out by Qiu Xigui (2002: 183)11 that Chang Yuzhi made a mistake in her calculations. It is true that bingyin appears in the fifth and seventh months, but that does not mean that the number of days in the fifth and sixth months total 60. Only when bingyin is the first day of the fifth and of the seventh months can one be certain that these two months add up to 60 days. Without this condition, it is baseless to claim that they are 60 days long and that both of them are long Yin months. It follows that Example 44 cannot be regarded as evidence for consecutive long Yin months in the Yin OBI. Both Chen Mengjia (1956: 219) and Chang Yuzhi (1998: 295) cite Example 45 as evidence for possible consecutive long months. Since the last character hai is not inscribed, it is not certain whether the second month indeed ends with guihai. In addition, Qiu Xigui 1 1 Qiu Xigui mistakes Heji 339 as Heji 389. 150 (2002: 183) points out that the nature of Example 45 needs to be studied further. It seems that the uncertainties about this inscription are far too strong. Qiu Xigui (2002: 181-183) proposes Example 46 as evidence for the existence of consecutive long Yin months in the Yin calendar. As already analyzed in Section 3.3.5, Qiu Xigui has not established that those inscriptions are complete records of consecutive divination about the auspiciousness in a month. Because of this, one cannot draw a conclusion about lengths of consecutive Yin months. Example 46 is therefore not evidence for consecutive long Yin months. Consecutive long months are not unusual in lunar-solar calendars. Because the Yin calendar is a lunar-solar calendar, it may have consecutive long months. However, so far, there is no clear-cut evidence for this phenomenon in the Yin calendar. 3.5.2 Consecutive Short Y i n Months In his introduction of Chinese calendar, Ran Xuezhen (1984: 1) makes it clear that it is common for a lunar-solar calendar to have consecutive short months. This is true. Take Chinese year 2003 as an example. In this year, the eighth month is a short month. So is the ninth month. This is an example of consecutive short months in the present Chinese calendar. Consecutive short months may appear in the Yin calendar as well. Dong Zuobin (1931: 505) mentions the existence of consecutive short months in the Yin calendar. After Dong Zuobin's publication, other researchers such as Xu Jinxiong, Chang Yuzhi and Qiu Xigui published their studies on consecutive short months of the Yin calendar. Xu Jinxiong (1985: 177-178) has rejoined eight pieces of oracle bones: Heji 37840+35529, Heji 37846+35422, Heji 37838+35756, Heji 35424+35534, Heji 35585+35649+35700, Heji 35653+35752, Heji 35409+35416 and Heji 35892+38274. According to Xu Jinxiong, inscriptions on these eight bones show guiyou (day 10) is the last gui-day of the eleventh month of the third year of Diyi, and guiwei (day 20) is the first gui-day of the fifth month of the seventh year of Diyi. After assuming that jiaxu (day 11) is the first day of the twelfth month of the third year, and that long and short Yin months occur alternately, he finds that the first day of the fifth month of the seventh year would be jiashen (day 21), i.e., the next day of guiwei (day 20). In order to make guiwei the first day of the 151 fifth month of the seventh year of Diyi, he suggests that there must be two short Yin months occurring consecutively during the period from the twelfth month of the third year to the fifth month of the seventh year. It is necessary to immediately point out that Xu Jinxiongs' reconstruction is only one of a number of possible interpretations of these inscriptions. It is true that guiyou (day 10) is the last gui-day of the eleventh month, and guiwei (day 20) is the first gui-day of the twelfth month of the third year. But these two gwz'-days do not necessarily make jiaxu (day 11) the first day of the twelfth month. Rather, any day from jiaxu (day 11) to guiwei (day 20) could be the first day. In addition, the five-ritual cycle has two periods, which certainly affect the reconstruction of those inscriptions. Moreover, as will be demonstrated shortly, in-year intercalation is adopted in Period V. This is a factor Xu Jinxiong neglects to consider. Taking all these factors into consideration, the possibility of Xu Jinxiong's construction being correct is very slight. Accordingly, Xu Jinxiong's reconstruction is not proof for consecutive short months in the Yin OBI. Chang Yuzhi (1998: 287-288) cites Heji 26682, Example 40 of this chapter, as evidence for consecutive short Yin months. She has argued that if guichou is the second gui-day of the ninth month, or that guiwei is the second gwz'-day of the first month, then two months out of the tenth, eleventh and twelfth months would be short months. Chang Yuzhi's argument is invalid because she makes a simple but very serious mistake in calculation that has been noticed by Zhang Peiyu and Qiu Xigui (2002). If guichou is the second gui-day of the ninth month, as Chang Yuzhi has claimed, these four months can only be reconstructed as follows: The 9 t h month: guimao (day 40) The 10'" month: guiyou (day 10) guimao (day 40) [guiyou] (day 10) The 1st month: guiwei (day 20) Apparently, there are seven gwz-days for the period between the tenth, eleventh, and twelfth months. The longest time period for seven gui-days is 79 days, which is 8 days shorter than three consecutive short Yin months. Chang Yuzhi's interpretation of Heji 26682 1 2 According to Qiu Xigui (2002: 184), Zhang Peiyu informed him about Chang Yuzhi's miscalculation. guichou (day 50) guihai (day 60) [guiwei] (day 20) [guisi] (day30) [guichou] (day 50) [guihai] (day 60) 152 does not make sense. Chang Yuzhi (1998: 293-295) also cites Heji 37983, Example 41 in this chapter, as possible evidence for consecutive short Yin months in the OBI. But she makes yet another mistake in calculating which has been discussed in Section 3.3.5. Again, there is no evidence here for consecutive short Yin months of the Yin calendar. Qiu Xigui (2002: 181-183) regards Example 46 as evidence for consecutive short months in the Yin calendar. However, since the nature of the inscriptions in Example 50 is very uncertain, Qiu Xigui has not established that they are complete records of consecutive divination about auspiciousness in a whole month. As analyzed in Section 3.3.5, no credible conclusion can be drawn from Example 46 with regard to the length of the Yin month. Needless to say, those inscriptions are not solid evidence for consecutive short months in the Yin Dynasty. It is Xu Jinxiong's opinion (1985: 181) that consecutive short months are something irregular ( T I E ^ ) . Qiu Xigui (2002: 184-185) also thinks this occurrence cannot take place in a calendar whose month starts with shuo, 'new moon.' However, as presented in the first paragraph of this section, the fact is that even the present Chinese calendar has consecutive short months. Xu Jinxiong and Qiu Xigui's opinions about consecutive short months do not correspond with the facts. Consecutive short months in a lunar-solar calendar, is not a rare phenomenon, and it is possible that it occurs in the Yin calendar as well. On the other hand, there is so far no good evidence to prove the existence of consecutive short months in the Yin calendar. 3.5.3 Intercalation in the Yin Calendar One lunar year is about 354 days long and a solar year is 365 days long. There is a difference of 11 days between a lunar year and a solar year. In order to adjust this difference, lunar-solar calendars have employed intercalations. The Yin calendar is not an exception. 3.5.3.1 The Existence of Intercalation of the Yin Calendar In Section 3.2 it has been demonstrated that a Yin year can be twelve or thirteen months 153 long. When it has twelve months, it is a normal year; when it has thirteen months, it is a leap year with one intercalary month. The Yin calendar does employ intercalation. 3.5.3.2 Year-end Intercalation In inscriptions of Period I and Period II, the phrase shisan yue ~f"ZE,J=j, 'the thirteenth month,' occurs in 147 inscriptions (Chang Yuzhi 1998: 302). It is Luo Zhenyu (1914.2.VII: 14) who first points to the fact that this is the designation for the intercalary month in the Yin OBI. All scholars except Liu Zhaoyang (1933: 143-145) and Sun Haibo (1935: 101-114) accept this view. Liu Zhaoyang and Sun Haibo take the phrase shisan yue to be another expression for yiyue —B, 'the first month.' However, such an opinion runs against the grain, as it were, of inscriptional evidence. [53] jg: #&&4*+ = £.^ ©. f f r g f ^ - J ^ W o i^M)) 14127 [X] divined: "Di may happen to [be able to] order Thunder in this thirteenth month." "Di may happen to [be able to] order Thunder upon the next first month." Heji 14127 In these two inscriptions, both the phrase jin shisan yue, 'this thirteenth month,' and the phrase sheng yiyue, 'the next first month,' appear. Moreover, they are used contrastively. It is thus clear that phrase shisan yue is not another expression of yiyue. Liu Zhaoyang and Sun Haibo are mistaken. As shown in Section 3.2, a Yin year can end with either the twelfth or thirteenth month. It appears that the thirteenth month is an "extra" month, which should be regarded as the intercalary month of a leap year. The phrase shisan yue in the OBI is direct evidence for the year-end intercalation in the Yin calendar. 3.5.3.3 In-year Intercalation Dong Zuobin (1945.2.V: 1-24), Chen Mengjia (1956: 220-222), Chang Yuzhi (1998: 307-318) and other scholars have offered inscriptional evidence to show the existence of in-year intercalation of the Yin calendar. However, Yabuuchi Kiyoshi (1956: 68-74), Chang 154 Zhengguang (1981: 105-106), Zheng Huisheng (1983: 111-114), and Zhang Peiyu (1984: 70-71) do not accept this view. Yabuuchi Kiyoshi, Chang Zhengguang, Zheng Huisheng and Zhang Peiyu hold the opinion that sophisticated astronomical knowledge is necessary for people to actually use the in-year intercalation. It is their assumption that the Yin people did not have such advanced astronomical knowledge, and therefore these scholars refuse to accept evidence for the in-year intercalation in the Yin calendar. The view of Yabuuchi Kiyoshi, Chang Zhengguang, Zheng Huisheng and Zhang Peiyu appears subjective. It remains a fact that none of them has established a criterion for how advanced astronomical knowledge would have to have been to enable people to employ the in-year intercalation. Without such a criterion, there is no way to pass judgment about the relationship between Yin astronomy and the in-year intercalation in the Yin calendar. It is merely an unsupported assertion for researchers to say that Yin astronomy was not advanced enough for the Yin to adopt the in-year intercalation. In this writer's opinion, their assertion does not lead in any productive direction. Moreover, these researchers reject the inscriptional evidence for the in-year intercalation presented by Dong Zuobin (1945), Chen Mengjia (1956), Wen Shaofeng and Yuan Tingdong (1983), Yang Shengnan (1986), Liu Xueshun (1992), and Chang Yuzhi (1998) by claiming that there must have been errors in engraving or that these inscriptions are interpreted incorrectly. If this claim is correct, it is a real challenge to the view that the Yin people adopted the in-year intercalation. It would be very difficult for any serious scholar to accept in-year intercalation in the Yin calendar if such a view is based on evidence that is flawed or interpreted inaccurately. On the other hand, if the evidence is valid, it would be unreasonable not to accept the in-year intercalation in the Yin calendar. Therefore, it is essential to make a thorough evaluation of all inscriptional evidence discovered so far, in order to determine if there is indeed the in-year intercalation in the Yin calendar. As discussed briefly in Chapter One, the Yin OBI can be divided into five periods. Below, inscriptional evidence for the in-year intercalation is presented and analyzed one period at a time. 3.5.3.3.1 Evidence in Period I 155 There are three examples of the in-year intercalation in Period I. The first example is Heji 10111. It bears the following relevant inscriptions. [ 5 4 1 T W K X^-£*° AB* H I C K *JS: ^ S T E P M T ^ J S c JL^c B E K ^ [ # ] ^ f ± ? o A ^ o Crack-making on dingyou (day 34), Dun divined: "To big temple tablets, [the king will offer] five oxen." [Day dingyou was in] the ninth month. Crack-making on guihai (day 60), Dun divined: "[The king will] pray for harvest to many kings, starting with Shangjia." [Day guihai was in] the ninth month. Crack-making on jiazi (day 1), Dun divined: "[The king will] pray for harvest, starting with Shangjia." [Day jiazi was in] the ninth month. Crack-making on jisi (day 6), Dun divined: "[The king] may happen to [be able to pray for] harvest to Shangjia." [Day jisi was in] the ninth month. Heji 10111 This evidence is presented by Yang Shengnan (1986). He points out that all month notations in these inscriptions refer to the ninth month. Then he proposes that all the divination dates are in the ninth month. For those divination dates, he suggests two possible orders. One order is from jiazi (day 1) to guihai (day 60), whose time span is 60 days; the other one is from dingyou (day 34) to jisi (day 6), whose time span is 33 days. The time period of those two orders, whichever may have been the case, exceeds the length of a Yin month, because a Yin month is either 30 or 29 days long. In order to explain such a long period for the ninth month, Yang Shengnan infers that there is an intercalary ninth month. He thus takes these inscriptions as evidence for the in-year intercalation of the Yin calendar. David N . Keightley (personal communication dated May 20, 2002) notices that Akatsuka Kiyoshi (1977: 548-49) lists these inscriptions, starting with the divination on guihai. Keightley notes that this reading gives "ninth month" dates spanning 35 days, i.e., from guihai to dingyou. He also mentions that Yao Xiaosui (1988: 240-241) starts the series with the divination on jiazi, which gives "ninth month" dates spanning 54 days, i.e., from jiazi to dingyou. These two more orders for the divination dates of Heji 10111 do not negate Yang Shengnan's conclusion. Because the time span of both orders still exceeds the length of a Yin 156 month, it is still necessary to use an intercalary ninth month to accommodate these long time periods. Among those four possible orders above, only the one from dingyou to jisi would place these inscriptions in a sequence neatly ordered from left to right across this scapula. The other three would place these inscriptions randomly. For this reason, the order from dingyou to jisi is likely the correct one. David N . Keightley, however, does not think this is strong evidence for the in-year intercalation in the Yin calendar. It is his opinion that the divination date dingyou ought possibly to be read as jiyou (day 46); the rubbing at least permits this possibility. He further argues that if the date were jiyou, then all the "ninth month" inscriptions would fit into a 21-day period from jiyou (day 46, not dingyou, day 34) to the putative jisi (day 6), thus removing any need for an in-year intercalary ninth month. Is the date dingyou possibly jiyou? An examination of the rubbing of Heji 10111 shows M that it is impossible. On Heji 10111, the graph in question is scribed asmBM, and the graph for ji is scribed asE3. In the first graph, there is no trace of a third horizontal stroke. The first graph should, therefore, still be transcribed as ding. For the four divination dates - jiazi, jisi, dingyou, and guihai — of Heji 10111, there are four possible orders. The time span of each of the four dates exceeds the length of a Yin month. Because these four dates all belong to the ninth month, it is correct for Yang Shengnan to propose an intercalary ninth month to accommodate such a long time period. Heji 10111 is a piece of strong evidence for the in-year intercalation in the Yin calendar. The second example is Heji 22404, on which there are the following inscriptions. [ 5 5 ]HEh j£: T4j-£(=5£)#3o + B ^ A I . H£P h t j j t : ( = ^ ) ^ o - / § < , Hf l h >5: 1 i J t : ( = ^ ) $ o H £ h, jji: ^ t r ( = ^ o ~Bo HH h ^ t r ( = ^ . (H)*h 157 %Wft: %tl(r3G)ffio %B.ft: ^}tl(=3c)mo K ft-, aitxrjtiffio m s * ( H ) * h ft: ^tr(=^)?Sc £J?c 22404 Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the twelfth month. On jihai (day 36), it rained heavily. Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the first month. Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the second month. Crack-making on guiyou (day 10), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiwei (day 20), [X] divined... Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi1 was in] the third month. [The king did] not catch [any game]. Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fourth month. Crack-making on guiwei (day 20), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the fifth month. Heji 22404 Heji 22404 is a rejoined fragment of Jiabian 3625, 3633, and 3635 by Yan Yiping (1951: 3-4). But he does not present these inscriptions as evidence for the in-year intercalation. In his reconstruction, he has assigned wuyue, 'the fifth month,' to guisi. This is a mistake because there is no guisi near wuyue on that fragment. Rather, wuyue is located right next to guiwei (day 20), and wuyue must be part of the inscription of guiwei (day 20). By assigning 158 wuyue to guiwei (day 20), it is possible to reconstruct the calendar for the period from the twelfth month to the fifth month, as follows: Guisi (day 30) 12th month Guimao (day 40) 1st month Guichou (day 50) (1 s t month) Guihai (day 60) 2 n d month Guiyou (day 10) [2nd month] Guiwei (day 20) [2nd month] Guisi (day 30) 3 r d month Guimao (day 40) [3rd month] Guichou (day 50) [3rd month] Guihai (day 60) 4 t h month [Guiyou day 10 4 t h month] [Guiwei day 20 4 t h month] [Guisi day 30 *4th month or *5 t h [Guimao day 40 *4th month or *5 t h [Guichou day 50 *4th month or *5 t h [Guihai day 60 5 t h month] [Guiyou day 10 5 t h month] Guiwei (day 20) 5 t h month Since guisi (day 30) is in the twelfth month and guihai (day 60) is in the second month, there are only two gui-days in the first month, as shown above. If long and short Yin months occur alternately, as is often seen in lunar-solar calendars, it can be calculated that at least seventeen months following this first month would have three gui-days. Based upon this construction, if the last gui-day of the fourth month is reconstructed as guiwei (day 20), then between guiwei (day 20) of the fourth month and guiwei (day 20) of the fifth month, there have to be 60 days. Apparently, there are six gwz'-days, the length of two Yin months, for the fifth month. An extra month is needed. This extra month is an in-year intercalary month. Therefore, Heji 22404 is a piece of good evidence for the in-year intercalation in the Yin calendar. 159 The third example is Hebu 4931, which is a rejoined piece of Heji 11545 and 16685. On this scapula, there are nine complete inscriptions. [56]HlCh tft(=5E)?S<= - J f . H * h = 1ut:(=^)#io H5P K ^ J S : Hj-£(=^)*g0 H H h, ^ J S : tlt:(=^)^o £J3° H M h ^ J S : tltr(=5E)?So AJi° H * h ^Jt:(=5E)^o H E h ^Jg: fjjt:(=^)?So « ^ # ) > 4931 Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the second month. Crack-making on guiyou (day 10), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the third month. Crack-making on guiwei (day 20), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guimao was in] the fifth month. Crack-making on guichou (day 50), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the fifth month. Crack-making on guihai (day 60), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the fifth month. Crack-making on guiyou (day 10), Bin divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the sixth month. Crack-making on guiwei (day 20), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Bin divined: "[In the next] 10-day week, there will be no disasters." Hebu 4931 160 It is Chen Mengjia (1956: 220-221) who first presented Heji 11545, part of this rejoined bone, as evidence for the in-year intercalation during the reign of Wuding. Based upon the dates and their month notations, he reconstructs the following calendar: Guihai, 2na month Guiyou, 3 r d month [Guiwei, 3 r d month] [Guisi, 3 r d month] Guimao [leap month] [Guichou, leap month] [guihai, leap month] Guiyou [4th month] Guiwei [4th month] [guisi, 4 t h month] Guimao, 5 t h month [Guichou, 5 t h month] Guihai 5 t h month Chen Mengjia's reconstruction is adopted by Chang Yuzhi (1998: 308-310). The present writer agrees with Chen Mengjia's arrangement of those dates, except that the leap-month does not necessarily have to be situated between the third and fourth months. It can as appropriately be inserted between the fourth and the fifth months as well. Heji 11545 was first published as Zhu 199. In order to avoid interpreting this piece of oracle bone as evidence for the in-year intercalation, Jin Zutong (1939: 16-17) suggests that either the phrase eryue, 'the second month' in the post-face of the first inscription, or, alternatively, sanyue, 'the third month' in the post-face of the second inscription, is mistaken. Such a handling of these inscriptions is highly problematic. As Chang Yuzhi (1998: 309) asserts, without enough evidence, one should not rashly doubt the correctness of original material; rather, one should draw conclusions from those materials— not change them to support one's own conclusion (&&fi%frffii£ffii£WZW,^M&Mfoft&&Pfctt:&&} Therefore, Jin Zutong's speculations warrant no serious consideration. They represent no real challenge to Chen Mengjia's reconstruction. Dong Zuobin (1945.2.V: 6) notices this piece of oracle bone as well. By assuming that the first two inscriptions and the remaining three belonged to two years, he inserts a year-end leap month between them. However, he fails to provide any evidence for his assumption. Therefore, his interpretation of this example lacks credibility. In short, Chen Mengjia's interpretation of this example is very straight and simple. More importantly, his interpretation does fit the context of the inscriptions. This is the third strong case in Period I for the in-year intercalation in the Yin calendar. 161 3.5.3.3.2 Evidence in Period I I In Period II, there are three examples for the in-year intercalation as well. The first one is ^Ut:(=3E)*8. tlt:(=^ )^ o - M . *Qt:(r3c)ffio i^M)) 26569 Crack-making on guiwei (day 20), Chu divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Chu divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), Chu divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the tenth month. Crack-making on guisi (day 30), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the tenth month. Crack-making on guimao (day 40), Chu divined: "[In the next] 10-day week, there will be no disasters." Heji 26569 According to the fourth and fifth inscriptions, both guichou (day 50) and guisi (day 30) are in the tenth month. From guichou to guisi, there are 41 days, which is 11 days longer than a long lunar month. In order to accommodate these extra 11 days, it is necessary to propose an intercalary month. That is the reason it is being presented here as a piece of inscriptional evidence for the in-year intercalation. David Keightley (personal communication dated May 20, 2002) says that "one could argue that the two charges are out of order on the bone, so that the 10 month ran from day - - j -[57] H * h tfcjS: H E K tb£: H5P h ttiiS: H A h H E h H2P h £ B J £ : 162 30 to day 50." Such an argument does not fit the inscriptional context on this bone. On a scapula, the inscriptions are normally read from the bottom to the top. If one reads these inscriptions on Heji 26569 in the usual way, i.e., from the first inscription to the sixth inscription, their order is from earlier to later inscription. The placement of these inscriptions is very clearly sequentially ordered; this does not support the argument that the two charges of the tenth month are out of order. David Keightley has not made his case that Heji 26569 is not a piece of evidence for the in-year intercalation of the Yin calendar. The second example is Heji 26643: [ 5 8 ] H * K R f t : I0t:(=^)*go A I T H l h - ±ft: T J C t ( = ^ ) ^ c TsBo H ^ K Xft: TJJ-£(=?E)?80 jsEo HW h t J t r ( = ^ 0 H E h 1jjt:(=^)#S. HFP h ft: iOt:(=^)*Bo H B h ttift: fjjtr(=5>c)#30 - t M o H E h R f t : T4jtr(=^c)^0 {^m 26643 Crack-making on guiwei (day 20), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guiwei was in] the sixth month. Crack-making on guichou (day 50), Da divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the sixth month. Crack-making on guihai (day 60), Da divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the sixth month. Crack-making on guiyou (day 10), Da divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), [X] divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Chu divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the seventh month. 163 Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Heji 26643 This inscription has been cited as an example of two intercalary months being inserted in the middle of one year, according to Liu Xueshun (1992: 4-6). Based upon the first seven inscriptions, the calendar for the sixth and seventh month has to be reconstructed as follows: Guiwei (day 20) 6 t h month [Guisi day 30 6 t h month] [Guimao day 40) 6 t h month] Guichou (day 50) 6 t h month Guihai (day 60) 6 t h month Guiyou (day 10) [*6th month] [Guiwei day 20 *6 th month] Guisi (day 30) [•6 th month] Guimao (day 40) [*6th month] Guichou (day 50) 7 t h month This writer's reconstruction demonstrates that, from the guiwei of the sixth month to the guichou of the seventh month, there are 10 gw/-days. The shortest time period for 10 gui-days is 91 days, which is longer than the length of three months. In order to account for such a long period, two months have to be inserted between the sixth and seventh months. This interpretation is also seen in Chang Yuzhi (1998: 312-315). Dong Zuobin (1934: 346-347, 1945.2.V: llb-12b) has cited this piece of bone as evidence for the in-year intercalation. In his transcriptions, he moves the first inscription to the place of the fourth inscription. This changes the interval between the guichou of the sixth month to the guichou of the seventh month. According to his transcriptions, from the guichou of the sixth month to the guichou of the seventh month, there are 61 days, which is one month shorter than the interval indicated by the present writer's transcriptions. The result is that he proposes only one intercalary month in his reconstruction. Dong Zuobin's transcriptions are followed by Chen Mengjia (1956: 221), Edward Shaughnessy (1985-87: 58-59) and Zheng Huisheng (1983: 111). David N . Keightley 164 (personal communication dated May 20, 2002) is interested in them as well. However, Dong Zuobin provides no reason for his change, and there is no compelling reason for him to change the position of those two inscriptions. The change Dong Zuobin makes appears subjective, and it undermines the credibility of his reconstruction. Comparatively speaking, the present writer's interpretation requires fewer assumptions than does Dong Zuobin's interpretation. In addition, those inscriptions can be adequately explained by the present writer's interpretation. According to the rule of Ockham's Razor (that assumptions must not be needlessly multiplied), the present writer's interpretation is preferable to that of Dong Zuobin's. David N. Keightley (personal communication May 20, 202) finds that two consecutive in-year intercalary months are "highly unlikely. Why would the months be so far out of synch with the solar year, especially in mid-summer? This makes little sense." It may appear that two consecutive in-year intercalary months make little sense. But the fact is that people in early China actually failed to put intercalary months in their calendars. It is recorded that, in the Zuozhuan, until the 27 th year of Duke Xiang i.e., 545 BC, two intercalary months had been omitted (Legge 1872b: 531). It is understandable that those two missing intercalary months had to be restored. When they were put back, there would be two consecutive in-year intercalary months within a single year. If this happened in the Spring and Autumn Period, it certainly could have happened in the Yin Dynasty. The seeming improbability of two consecutive intercalary months in one year is not a serious challenge to the present writer's interpretation of Heji 26643. Heji 26643 is an example of two consecutive in-year intercalary months within a single year. Needless to say, this is strong evidence for the in-year intercalation in the Yin calendar. The third example for in-year intercalation in Period II is Hebu 8197. [ 5 9 ] H S h-ftjfc: 10t:(=3E)*8. %B° H E h GitXrftfflo H H K R$X: iQtKrftffio HICK JSTU/J: f j jtr(=^ 0 - t - B o H E h tijtKrftffio H£P K tjtr(=^c)?So 165 Crack-making on guiyou (day 10), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the ninth month. Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guichou was in] the tenth month. Crack-making on guihai (day 60), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the eleventh month. Crack-making on guisi (day 30), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guimao (day 40), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guichou (day 50), Xiong divined: "[In the next] 10-day week, there will be no disasters." Crack-making on guiyou (day 10), Xiong divined: "[In the next] 10-day week, there will be no disasters." [Day guiyou was in] the twelfth month. Hebu 8197 Hebu 8197 is a rejoined fragment of Heji 26628 and 26630. This writer's transcriptions above basically follow those provided by the Jiaguwen heji bubian shiwen, except the order of the inscriptions. Based upon the dates and month notations of those inscriptions, the calendar table for those months has been reconstructed by the present writer as follows: Guiyou (day 10) 9 t h month [Guiwei day 20 9 t h month] Guisi (day 30) [*9th month or * 10th month] [Guimao day 40 10th month] 1 3 In the transcriptions provided by the Hebu shiwen, the inscription of the eleventh month is put before that of the twelfth month; that of the twelfth month appears before that of the ninth month; and that of the ninth month is put before that of the tenth month. It seems that the person who prepared these transcriptions did not follow the convention adopted by specialists in the field of OBI. 166 Guichou (day 50) 10m month 11th month 11th month] 11th month] [11th month] [11th month] [11th month] 12th month] 12th month Guihai (day 60) [Guiyou day 10 [Guiwei day 20 Guisi (day 30) Guimao (day 40) Guichou (day 50) [Guihai day 60 Guiyou (day 10) The third and fourth inscriptions show that guichou (day 50) is the last gui-day of the tenth month, and that guihai (day 60) is the first g«/-day of the eleventh month. From the guihai (day 60) of the eleventh month to the guiyou (day 10) of the twelfth month, there are eight gui-days. The longest time period of eight gui-days is 89 days, which correspond to the length of three months. In order to accommodate these 89 days, one intercalary month has to be inserted between the eleventh and twelfth month. Therefore, this is a piece of evidence for in-year intercalation of the Yin calendar. David N. Keightley (personal communication dated May 20, 2002) questions the present writer's interpretation. He points out that looking at the rubbing, I find that '12 t h month' record [is] uncertain. If it were an '11 t h month,' then no in-year intercalation would be needed, I think. I do grant that the '2' of the putative '2' may be present on HJ 26628, but the bottom line of the '2' is particularly faint. It is unfortunate that the month number was split in two when the scapula fractured. The result is that any conclusions drawn from this particular set of inscriptions do not strike me as fully reliable. It is certainly an unfortunate fact that the bottom line of the "2" is particularly faint. It is appropriate for David Keightley to be cautious about drawing a conclusion with regard to an in-year calendar from these inscriptions. But one can still tell which word the bone graph represents. Here are three graphs on that bone:ISS, ESP , and ES3. The first and third 167 graphs are crack numbers. Although the bottom line of the first graph is not as clear as that of the third graph, it still is er, 'the second.' The second graph is what David Keightley refers to. Like the first graph, its bottom line is not as clear as that in the third graph. But it can be discerned that it is er. So the guiyou of the eighth inscription indeed belongs to the twelfth month, and thus an in-year intercalary month must be inserted in that year. 3.5.3.3.3 Evidence in Period V In Period V, there are two such examples. The first one is Heji 35745: [ 6 0 ] H £ h ft: il0t:(=^)#5o £ + 4 . ¥ ^ S P H ¥ . ft: 3 E ^ t ( = ^ ) ^ 0 M S B M * HFP h> ft: 3EtJt:(=^)^0 H £ K ft: ii0t:(=^)*8o ? m ¥ o H * K ft: ifjjt:(=^)^o [H0P]h l£: [i10]£:(=3c)#i. S = ^ o « ^ » 35745 Crack-making on guihai (day 60), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guihai] was in the tenth month.14 On jiazi (day 1), [the king will perform] the y/'-sacrifice to Yangjia. Crack-making on guiwei (day 20), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." On jiashen (day 21), [the king will] perform the yi-sacrifice to Zujia. Crack-making on guimao (day 40), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." Crack-making on guihai (day 60), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guihai] was in the first month. On jiazi (day 1), [the king will perform] the .//-sacrifice to Dajia. Chang Yuzhi (1987: 12) suggests that the scope of the month notation in these inscriptions covers both the gui-days and the jia-days. But in the fifth and sixth inscriptions, there are month notations and gui-days only. The month notation is thus related to the gui-days of these inscriptions only, which has already been pointed out by Xu Jinxiong (1985: 178). 168 Crack-making on guiwei (day 20), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guiwei] was in the second month. Crack-making on [guimao] (day 40), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guimao] was in the third month. Heji 35745 Based on the transcriptions above, this writer reconstructs the calendar for these months as follows: Guihai (day 60) 10th month yi Yangjia [Guiyou day 10 11th month] Guiwei (day 20) [11th month] yi Zujia [Guisi day 30 11th month] Guimao (day 40) [12th month ji Shangjia] [Guichou day 50 12th month] Guihai (day 60) 1st month ji Dajia [Guiyou day 10 1st month ji Xiaojia] [Guiwei day 20 1st month] [Guisi day 30 *2 n d month ji Jianjia] [Guimao day 40 *2 n d month ji Qiangjia] [Guichou day 50 *2 n d month ji Yangjia] [Guihai day 60 2 n d month] [Guiyou day 10 2 n d month ji Zujia] Guiwei (day 20) 2 n d month [Guisi day 30 3 r d month] [Guimao] (day 40) 3 r d month Guihai (day 60) occurs in both the tenth and first months, and there are only 59 days left for the eleventh and twelfth months; i.e., one of them must have two gui-days only. Calculations show that the other month with two gui-days would be about 20 months away if long and short Yin months appear alternately, which is common in lunar-solar calendars. Therefore, the first, second, and third months should have three gwz-days. If there is no 169 intercalary month, guiwei and guimao would not appear in the second and third month, respectively. By inserting an intercalary month, be it an intercalary first or second month,15 the reconstruction can accommodate all of those dates and month notations. This example serves as a demonstration of the in-year intercalation in the Yin calendar. However, there is some uncertainty about this example. The month number in the fourth inscription is scribed as KiSi£. It is clearly different from B l in the fifth inscription. The latter graph is without doubt transcribed as er, 'two,' or 'the second.' The white impression in the upper portion in the former graph does not look like a stroke, however. Thus, this writer transcribes the graph as yi, 'one,' or 'the first.' On the other hand, Chen Mengjia (1956: 395) and Chang Yuzhi (1987: 175) transcribe it as er. After transcribing the graph in question as er, Chen Mengjia and Chang Yuzhi propose different interpretations of these inscriptions. It is Chen Mengjia's idea (1956: 395) that the order of these inscriptions is from top to bottom. Such a reading makes the second month run from guiwei (day 20) to guihai (60). This second month would have at least five gui-days, the longest time period of which is 59 days. The period of 59 days corresponds to the length of two months. In order to account for such a long time period, one has to propose an intercalary month. Chang Yuzhi (1987: 179) does not agree with Chen Mengjia's reading. The present writer also disagrees with Chen Mengjia's interpretation, even though it does make these inscriptions into an example of the in-year intercalation. The main reason for objecting to Chen Mengjia's interpretation is that the ji H-sacrifice is immediately after the yi 31-sacrifice in the five-ritual cycle whose characteristics are analyzed fully by Chang Yuzhi (1987). The order from bottom to top is in agreement with the order of the yi and ji sacrifices, 1 5 David N. Keightley (personal communication dated May 20, 2002) expresses his reservations about the intercalary first month. "Why would the 'regulating month' (zheng yue) have had to be made intercalary? The Shang would presumably have added a 13th month at the end of the previous year, to 'slow' their calendar down. It seems unlikely that they would have wanted to duplicate their 'regulating month' - indeed, to do so would seem to be almost a contradiction in terms." The questions he raises appear very reasonable. However, it seems that such questions did not bother early Chinese people. I have looked through the calendar table for the period from 100 BC to 1 BC and found intercalary first month or zhengyue in the following four years: 88 BC, 69 BC, 31 BC and 12 BC. For more detail, see Zhang Peiyu (1990: 79-95). It seems that Keightley's doubt about an intercalary first month does not affect my proposal of an intercalary first month for this example. 170 as shown in the reconstruction above. On the other hand, Chen Mengjia's reading goes against such an order. Chen Mengjia has misinterpreted this example. Chang Yuzhi (1987: 177) suggests that the month number in the first inscription is a mistake. She changes the month notation to the twelfth month so that the records of the five-ritual cycle do not violate their patterns. It appears, however, that such a change is not necessary if the month number in the fourth inscription is transcribed as yi ~ , as the graph indicates, rather than as er H , as she has done, which is clearly demonstrated by the present writer's own reconstruction above. Similarly, David N . Keightley (personal communication dated May 20, 2002) does not think there is any reason to change the month number from 10 to 12. He points out that the month number in the fourth inscription on the rubbing is unclear, which is true. By his calculation, if the month in question is the second month, no intercalation is needed. David N. Keightley has not given details about his calculations. In any event, if the month number 10 is not changed to 12, there are at least 13 gui-days from the guihai of the tenth month to the guihai of the second month. On the other hand, from the time of performing the ^/'-sacrifice to Zujia to the time of performing the y'z'-sacrifice to Dajia, only 7 gui-days are required. The numbers of gui-days clearly do not match. Keightley's calculations are thus open to question. Admittedly, the month number in the fourth inscription is visually unclear, although it looks like yi, 'one,' or 'the first.' However, the month number in the first inscription is clearly shi, 'ten' or 'the tenth.' When there are two graphs, one clear and the other unclear, a researcher should decide how to transcribe the unclear one based upon the clear one, not change the clear one. If, in the present context, the unclear graph is transcribed in light of the clear one, as the present writer suggests is the appropriate procedure, his transcription for the month number in the fourth month will be confirmed. In addition, the present writer's transcription is seen to be in agreement with the pattern of the five-ritual cycle in Period V. By transcribing the month notation in the fourth inscription as the first month, the reconstruction for the period from the tenth month to the third month shows that an in-year intercalary month must be inserted. Heji 35745 is an example illustrating the in-year intercalation in the OBI. The second example is Hebu 10962 that bears the following inscriptions. 171 [6i ]HMK J£: i^jtr(=^)#i0 4 0 1 . ¥j£1ftB*B¥. HM h. J^ : £ 1 0 (t:) (J!) . 10962 Crack-making on guiyou (day 10), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guiyou] was in the fourth month. On jiaxu (day 11), [the king will] perform the xz'eTz'-sacrifice to Zujia. Crack-making on guichou (day 50), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guichou] was in the fifth month. On jiayin (day 51), [the king will] perform the rongrz'-sacrifice to Dajia. Crack-making on guiyou (day 10), [X] divined: "As for the king, [in the next] 10-day week, there will [be no] disasters." [Day guiyou] was in the fifth month. Hebu 10962 It is clear that guiyou occur both in the fourth month and in the fifth month. Between these two dates, there are seven gwz'-days whose length is at least 61 days. It exceeds the length of two Yin months. This writer proposes an intercalary month to account for those extra days. If the intercalary month is a fifth month, this writer offers one possible reconstruction of the calendar table for these two months: Guiyou (day 10) 4 t h month xieri Zujia [Guiwei day 20 *4th month] [Guisi day 30 *4th month rong Shangjia] [Guimao day 40 *5 t h month] Guichou (day 50) 5 t h month rongri Dajia [Guihai day 60 5 t h month] Guiyou (day 10) 5 t h month [Guiwei day 20 5 t h month] [Guisi day 30 5 t h month] If the intercalary month is the fourth month, the reconstruction would be slightly different: 172 [Guichou day 50 4 t h month] [Guihai day 60 4 t h month] Guiyou (day 10) 4 t h month xieri Zujia [Guiwei day 20 *4th month] [Guisi day 30 *4th month rong Shangjia] [Guimao day 40 *4th month] Guichou (day 50) 5 t h month rongri Dajia [Guihai day 60 5 t h month] Guiyou (day 10) 5 t h month As far as the present writer is concerned, it does not really matter whether the intercalary month is the fourth or fifth month. Whichever may have been the case, Hebu 10962 is a piece of evidence for the in-year intercalation in the Yin calendar, and this is what matters. There is one possible example of the in-year intercalation in Period V. Those inscriptions are scribed on Hebu 10958: [ 6 2 ] H M i K ft: - SHHo m^Afco 16 fS*3EK ft: tlt:(=^)^o ? ± z ^ H E T K ft: Hjtr(=^)*8. £HJ3o mmb, ft-. 1iJt:(=3E)$. 4 H J !o H l l h , ft: *0-£(=^)*B. S = ^ o M l h . ft-. trfr(= )^*B. £ = H o On guiyou (day 10), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters."... [Day guiyou] was in the third month.... [It] was the king's eighth year. On guiwei (day 20), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters."... [Day guiwei] was in the third month.... On guisi (day 30), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters."... [Day guisi] was in the third month.... 38 The inscriptions in this example are lengthy. In order to save space, the present writer only transcribes relevant parts of the inscriptions on this oracle bone. 173 On guimao (day 40), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters."... [Day guimao] was in the third month.... On guichou (day 50), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters." ... [Day guichou] was in the third month On guihai (day 60), the king made this crack and divined: "[In the next] 10-day week, there will be no disasters." ... [Day guihai] was in the third month.... Hebu 10958 The transcriptions above are provided by authors of the Jiaguwen heji bubian shiwen. It is clear that the third month has six gui-days, the length of two Yin months. There should be an intercalary third month in that year. David N . Keightley (personal communication dated May 20, 2002) points out that the month number in the post-face of the first inscription is read as the second by Dong Zuobin (1945.2.II: 6b), Qu Wanli (1961: 86-88) and Yan Yiping (1975: 329) and Chang Yuzhi (1998: 280-281). Moreover, the month number in the fifth inscription is read as the fourth, rather than the third, month by Yan Yiping (1975: 329) and Chang Yuzhi (1998: 280-281). If the transcription of the Jiaguwen heji bubian shiwen is correct, Hebu 10958 is an example of the in-year intercalation in Period V. If other readings are correct, there is no need for an intercalary third month. Since it is difficult to determine which transcription is accurate, one cannot be sure about whether there is an intercalary third month recorded on Hebu 10958. It is merely potential evidence, not proof, for the in-year intercalation. Table 7: Intercalary Months in the Yin OBI Oracle Bone Intercalary Month Heji 10111 9 t h month Heji 22404 4 t h or 5 t h month Hebu 4931 3 r d, 4 t h or 5 t h month Heji 26569 10tn month Heji 26643 6 t n month Hebu 8197 11th month Heji 35745 1st or 2 n d month Hebu 10962 5 t h month *Hebu 10958 *3 r d month 174 The results of these nine examples, eight certain and one possible, for the in-year intercalation in the Yin calendar, can be tabulated as Table 7 on the previous page. 3.5.3.3.4 False Evidence for the In-year Intercalation Whether the Yin calendar adopts the in-year intercalation is an important issue. In the past, a number of scholars have made efforts to address this issue. They have discovered an increasing body of inscriptional evidence. However, it is also the case that not every example presented to date constitutes valid evidence. The following examples indicate some problems that arise when they are used as evidence for the in-year intercalation in the Yin calendar. Example I: [ 6 3 ] - ^ o H f i h IS, leu J F U M . 0 ^ ° .. .the second month. Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, [there will be no disasters]." On ren[sheri\ (day 9), there was a storm and wind. Crack-making on guichou (day 50), [X] divined: "[In the next] 10-day week, [there will be no disasters]." On jiayin (day 51), it rained. [Day jiayin was in] the fourth month. Crack-making on guiyou (day 10), [X] divined: "[In the next] 10-day week, [there will be no disasters]." On gengchen (day 17), it rained. [Day gengchen was in] the fourth month. Heji 13361 On this piece of oracle bone, the inscriptions are scribed in three rows. The phrase eryue, 'the second month,' appears in the top row; one siyue, 'the fourth month,' in the middle row; and the other siyue in the bottom row. Because the second month is earlier than the fourth month, those month notations suggest that these inscriptions should be read from top to bottom. Reading in this order, the fourth month runs from guichou (day 50) to guiyou (day 10). There are three gw/-days in that month, and there is no need for an intercalary fourth month. 175 Chang Yuzhi (1998: 310-311) reads these inscriptions in an opposite order. By doing so, the fourth month runs from guiyou (day 10) to guichou (day 50). If so, there are at least five gui-days in that month, the longest time period of which is 59 days. In order to account for such a long time period, she proposes that there is an intercalary fourth month in that year. As mentioned above, the placement of those month notations indicates that these three inscriptions should be read from top to bottom. Chang Yuzhi reads these inscriptions in the wrong direction. Because of this, her proposal of an intercalary fourth month is not credible. Heji 13361 is not good evidence for the in-year intercalation. H £ K tlt:(=^)?So 7 L ^ o Crack-making on guisi (day 30), [Zheng] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the seventh month. Crack-making on guihai (day 60), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guihai was in] the ninth month. Heji 16706 Chang Yuzhi (1998: 311) has presented this example as evidence for the in-year intercalation and the present writer followed her in the draft of Liu Xueshun (2002). However, David N . Keightley (personal communication dated May 20, 2002) points out that, if one interprets guisi as the last day of the seventh month, and guihai as the first day of the ninth month, then the eighth month between them has 29 days, the length of a short month. There is no need for intercalation. David N . Keightley's interpretation seems simple and direct. Thus, this is not a good example of the in-year intercalation in the Yin calendar. Example III: Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." Onyiwei (day 32).... [Day yiwei was in] the seventh month. Crack-making on guisi (day 30), [X] divined: "[In the next] 10-day week, there will be no disasters." [Day guisi was in] the eighth month. Example II: 34991 176 Crack-making on guiyou (day 10), [X] divined: "[In the next] 10-day week, there will be Liu Xueshun (2002: 12) presented these three inscriptions on Heji 34991 as evidence for the in-year intercalation in the Yin calendar. At that time, the present writer did not notice that the phrase qiyue, 'the seventh month,' occurs after yiwei, which indicates it is the month notation of yiwei rather than of guisi. In addition, as David N . Keightley (personal communication dated May 20, 2002) points out, the link is not definite between the phrase jiuyue, 'the ninth month,' and guiyou. These changes make it unnecessary to reconstruct an intercalary month. Example IV: Crack-making on..., [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [XX was in] the first month. Crack-making on [guimao] day 40), [X] divined: "[As for the king, in the next] 10-day week, there will be no disasters." On jiachen (day 41), [the king will] perform the ji-sacrifice to Shangjia. Crack-making on guichou (day 50), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guichou was in] the first month. On jiayin (day 51), [the king will] perform the B9-sacrifice to Shangjia. Crack-making on guiyou (day 10), [X] divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guiyou was in] the first month. On jiaxu (day 11), [the king will] perform the EE-sacrifice to Dajia. Hebu 10949 The above transcription is provided by authors of the Jiaguwen heji bubian shiwen. Based upon this transcription, it was possible to reconstruct an intercalary first month. However, the month number of the last inscription was not transcribed accurately. On the no disasters." [Day guiyou was in] the ninth month. Heji 34991 177 rubbing, the month number appears as 13, which can be transcribed as yue er, 'month two.' This change makes it unnecessary to insert an intercalary first month. David N. Keightley (personal communication dated May 20, 2002) points out that this month number was transcribed as shier yue, 'the twelfth month,' by Hu Houxuan's drawing of Xucun 2.965 and Shima Kunio (1971: 520.2). He inclines to read it that way, too. But the present writer does not think that the month number can be transcribed as the twelfth month, because of the following considerations: First, the bone graph shi, 'ten,' cannot be found in the last inscription. Second, by transcribing the month number as the twelfth month, one has to put the last inscriptions before the other three inscriptions. However, according to the pattern of the five-ritual cycle, the y/-sacrifice to Shangjia is performed before the E^ Lsacrifice to Shangjia, which in turn is performed before the ^ 8 sacrifice to Dajia. This pattern determines that the month notation of the fourth inscription is later than that of the other three inscriptions, i.e., the first month. The month following the first month is the second month rather than the twelfth month. In light of these two reasons, the month notation in the post-face to the fourth inscription should be transcribed as the second month, not the twelfth month. 3.5.3.4 Conclusion with Regard to In-year Intercalation Section 3.5.3.3 presented an evaluation of inscriptional evidence for the in-year intercalation in the Yin calendar. An analysis of various issues concerning dates and month notations of each piece of evidence led to the following conclusion: there are eight correct examples, one possible example, and four wrong examples of the in-year intercalation in the current corpus of the Yin OBI. If anyone treats those inscriptions objectively, i.e., interpreting them without changing dates and month notations recorded on the bones, one would reach the same conclusion. It is this writer's firm view that those eight correct examples are decisive proof for the in-year intercalation in the Yin calendar. Chang Yuzhi's statistics (1998: 302) show the phrase shisan yue, 'the thirteenth month,' occurs 142 times in inscriptions of Periods I and II. This phrase never appears in inscriptions 178 of Periods III, IV and V. As analyzed in Section 3.5.3.2, the phrase shisan yue is the designation for an intercalary month in Periods I and II. If there are intercalary months in inscriptions of early periods, there certainly have to be intercalary months in inscriptions in late periods, too. The lack of the thirteenth month in late periods indicates that the intercalary month is not put at the end of the Yin year anymore in late periods. Although the phrase shisan yue, 'the thirteenth month,' occurs frequently in early inscriptions, there are six certain examples of the in-year intercalation in inscriptions in Periods I and II. One should not deny these examples simply because of the existence of the thirteenth month in these two periods. It should not be difficult to understand that there could be a transitional period when the year-end intercalation and the in-year intercalation coexist. The coexistence of the thirteenth month and examples of the in-year intercalation in early inscriptions indicate that the transition from the year-end intercalation to the in-year intercalation took place in Periods I and II. The absence of the thirteenth month in late periods indicates that the year-end intercalation was completely replaced by the in-year intercalation by the time of the late Yin Dynasty. 179 CHAPTER FOUR "THE YEAR" IN THE YIN CALENDAR 4.1 Introduction The month notations in the Yin OBI show that Yin months are numbered. Numbering starts with 1 and ends with 12 or 13 before starting with number 1 again. This indicates that twelve or thirteen Yin months form a unit in the Yin calendar. It is generally accepted that such a unit in the Yin calendar is a Yin "civil year." A Yin year of twelve months is a "normal year" and that of thirteen months is an "intercalary year." With regard to the Yin year, specialists have not reached a consensus on the designation for the year in the Yin language, nor have they agreed on the commencement of the Yin year. They are issues that will be addressed in this chapter. 4.2 The Designation for the Yin Year The designation for the Yin year has been a rigorously debated issue in the study of the Yin calendar. To date, si IE, zai tic, nian and sui have been proposed as terms for the Yin year. In order to judge whether these words are valid designations for the Yin year, it is necessary to establish a set of clearly defined criteria. Below, this writer first makes clear his criteria for how to determine the appropriate designation for the Yin year. Then, the discussion proceeds to analyze which word is an appropriate designation or term for the Yin year in the Yin OBI. 4.2.1 Criteria If a word is a designation for the year, it must refer to the period of a calendar year, which means one that lasts from the first to the last month of a given year. A word that refers to a time period of one year is not necessarily a designation for the year, a fact that can be shown by the usage of the words zai 1±% and sui £. 180 In phrases such as san nian wu zai zi^E-iLWu, 'three or five years,' and sui sui ping an # ^ T ^ , 'be safe every year,' words sui nian and zai tic all refer to a time period of one year. However, not all of them are terms for a calendar year. For instance, there are no expressions such as *2004 sui & or *2004 zaz'fc, 'year of 2004.' On the other hand, 2004 nian 'year of 2004,' is perfect Chinese, where the word nian refers to the time period from the first day of 2004 to the last day of 2004. This demonstrates that among these three words, sui, zai, nian, although all of them can refer to the time period of one year, only nian is the designation for the year in modern Chinese language. It is to be suspected that such distinctions exist in the language of the Yin OBI as well. In other words, even though a word in the Yin OBI may be used to refer to the time period of one year, if it is not used to refer to the period of a calendar year, this word can still not be considered as a designation for the Yin year. The present writer now proposes two criteria for judging whether a word in the Yin OBI is a term for the Yin year. First, there must be inscriptional evidence to show that a word is used to refer to the time period of one year. It is easy to understand that a word cannot possibly be a designation for the Yin year if it is not used to refer to the time period of one year in the inscriptions themselves. Second, there must be inscriptional evidence to show that a word is used to refer to the period of a calendar year, i.e., the time period from the first month of a year to the last month of that year. Without such evidence, it is not certain whether it is a designation for the Yin year in the Yin OBI. Having stated criteria for determining a designation for the year in the OBI, this writer will now analyze which of the words, si &, zai ^c, nian and sui $r, is a designation for the Yin year. 4.2.2 Si It is a standard practice to transcribe the bone graph HH2 as si IE, 'to sacrifice.' The following inscription is an example of such usage of the word si: [ O T J ^ h ^tEffio 14851 Crack-making on gengzi (day 37), Zheng divined: "[The king] will perform a sacrifice to He." Heji 14851 181 This example is a piece of inscription from Period I. From Periods I to IV, the word si means "to sacrifice," as shown by Example 1. By Period V , this word gains another usage: a designation for the Yin year. The Erya ^K?t is a dictionary compiled in the Han Dynasty. In this dictionary, there is a record that states that Shang yue si i ^ S l R , ' in the Shang Dynasty [the year] is called si.' Based on this record, Luo Zhenyu (1914.7: 53b) proposes that the word si is a designation for the year in the OBI. His opinion is followed by Dong Zuobin (1931: 518-519) and Hu Houxuan (1944. 1: 3a-6a) in the early stage of the study of the Yin calendar. On the basis of his research about the five-ritual cycle, Dong Zuobin (1945.1.Ill: 2) explains why the word si is used as a term for the Yin year: -mzm, t & v i = + A t i , ^-^ZRWL, B t f £ & Z^, gpi^A 'IffTUB' In other words, among sacrifices during the reigns of Di Y i and Di Xin , rong,yi,ji, and xie are five main ones. The period for performing these five sacrifices to all ancestors and ancestresses is exactly 36 xuns, '10-day week,' which are close to the number of days in a year. Therefore, [the Yin] called a year a si, and the current year of the contemporary king is recorded as wei wang ji si, 'it was X' th si of the king.' Dong Zuobin's explanation is followed by Chinese specialists in the field of the Yin calendar. However, his explanation is not accepted by non-Chinese scholars. For example, it is Shima Kunio's interpretation (1958: 128, 502) that the word si in the phrase wei wangji si still means "to sacrifice." He interprets the phrase wei wangji si as a notation of the number of sacrifices offered by the king, rather than a notation of the year of the Yin king's reign. Shima Kunio's view has been followed by Ito Michiharu (1996: 99) and David N . Keightley (2000:50). ^ It appears that there are still disputes over whether the word si is a designation for the 182 Yin year. Liu Xueshun (2003) has cited some inscriptions showing that si is a term for the Yin year in Period V inscriptions. Now, we investigate this issue more thoroughly. According to criteria stated in Section 4.2.1, the first step in deciding whether a word is a designation for the Yin year is to examine whether it refers to the time period of one year. Inscriptional examination proves that the word si does. [ 0 2 ] H J i h ft-. W c THo M'^ZM1, f£iA?Eo i^M)) 37855 Crack-making on guichou (day 50), [X] divined: "Lady ... has ...." It will not be blessed. [Day guichou] was in the first month and coincided with [the day when] sacrificial meat and what was amputated were offered to Xiaojia. It was the ninth si of the king. Heji 37855 [03] Z , 2 H K ft: F34-#SfS, ACTAliio i £ H : ilffiZ^o ^AlEo l&m 37852 On yihai (day 12), the king made cracks and divined: "From this spring to next [spring], the Renfang will not take the field on a massive scale." The king read cracks and declared, "Auspicious." [Day yihai] was in the second month and coincided with [the day when] sacrificial meat was offered to Zuyi. It was the ninth si. Heji 37852 [ 0 4 ] [ H ] £ £ K ft-. m^U, r H [ ± ¥ M ] f ^ J 5 ^ , t : W 3 [ S ^ c i £ ] B : i i f . £ HJ3o ^ittPfE],, 4 37864 On [gui]hai (day 60), the king made cracks and divined: "[I will] perform the you-cutting sacrifice on the ro«g-sacrificial day, starting [from Shangjia] to many kings, there will be no [disasters." The king read cracks and] declared: "Auspicious." [Day guihai] was in the third month. It was the twentieth si of the king. Heji 37864 [ 0 5 ] H * £ h ft: mZU, g ± ¥ M T £ J 5 * » t f f t lSo i r J j S : M o ^ 3 E - * E . «1=ril» 37865 On guiwei (day 20), the king made cracks and divined: "[I will] perform the yow-cutting ' For the study on the expression rongyue j&M, readers are referred to Ken-ichi Takashima (2002). 2 The supplement of character yi 7L is based on the pattern of sacrifice inscriptions in Period V. In this period, a sacrifice to an ancestor is normally performed on the date whose tiangan is the same as that of the ancestor in question. 3 For the interpretation of the word hai #, see Qiu Xigui (1992: 11-16). 4 Based on Example 5, the missing characters are identified because the content of both inscriptions is the same. 183 sacrifice on the rorcg-sacrificial day, starting from Shangjia to many kings, there will be no disasters." The king read cracks and declared: "Auspicious." [Day guiwei] was in the fourth month. It was the second si of the king. Heji 37865 [06]H*f K j£: [tjtr(=^)^o jfrJjS: ^ . ££Hf£ j^ [}E ]o i & M } 37846 On guiwei (day 20), the king made cracks and divined: "[In the next 10-day week, there will be no disasters.]" The king read cracks and declared: "Auspicious." [Day guiwei] was in the fifth month... It was the seventh [si] of the king. Heji 37846 [07]HSK itlt:(=^)^o £ T N J 3 O TOI^±¥<. itJ-teo 37867 Crack-making on guichou (day 50), Yong divined: "As for the king, [in the next] 10-day week, there will be no disasters." [Day guichou] was in the sixth month. On jiayin (day 51), the yow-cutting sacrifice and the yi sacrifice were offered to Shangjia. [It was] the twentieth si of the king. Heji 37867 [08] [Hfl]f h , j£= 10t:(=^)#§. [ i £ ] H : B ° £^J3o EPW P^H¥. £ i H lEc i^M)) 37839 [On guichou] (day 50), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." [The king read cracks] and declared: "Auspicious." [Day guichou] was in the seventh month. On jiayin (day 51), the sacrificial meat was offered to Yangjia. It was the third si of the king. Heji 37839 [09]HBK J*: 4#g7fc. ftJKHo £IAfEo «f>il» 37849 Crack-making on guichou (day 50), [X] divined: "This season [we will] reap a harvest." It was greatly auspicious. [Day guichou] was in the eighth month. It was the eighth si of the king. Heji 37849 37844 On guimao (day 40), the king made cracks and divined: "[I will] perform the yow-cutting sacrifice on,the ^ /-sacrificial day, starting from Shangjia to many kings, there will be no disasters." [Day guimao] was in the ninth month. It was the fifth si of the king. Heji 37844 184 [11]PIK [Jg]: I^[^]t(=^)io S+^o i^M)) 37842 Crack-making on x-yow (day ??), [X] divined: "As for the king, there will be no disasters tonight." [Day x-you] was in the tenth month. It was the fourth si of the king. Heji 37842 [12]HM3Eh- J*: t J t (=^)?S° i £ H : t o ^ i H f E o « ^ | | » 37840 On guiyou (day 10), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." The king read cracks and declared: "Auspicious." [Day guiyou] was in the eleventh month. It was the third si of the king. Heji 37840 [13]HWiK j£: t l t r (=^)^ 0 i r ^ S : ^ . S-r-JfX-o ^ i A I E o 37845 On guiyou (day 10), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." The king read cracks and declared: "Auspicious." [Day guiyou] was in the twelfth month. It was the sixth si of the king. Heji 37845 It is clear from Example 2 to Example 13 that the word si can refer to any month, from the first to the twelfth, of a Yin year in Period V. In other words, si is used to refer to a time period of one year in the Yin OBI. The word si does meet the first criterion for a word being a designation for the Yin year. Can the word si refer to the time period of a calendar year, i.e., from the first to the twelfth month of the same year? This can be decided by investigating whether the commencement of a si coincides with the first month of the Yin year. If that is the case, one can safely conclude that the word si is used to refer to a calendar year. If so, the word si will meet the second criterion for judging a word as a designation for the Yin year, and it can be determined as a term for the Yin year. If the start of si does not coincide with the first month of the Yin year, it means the word si does not refer to the time period of a calendar year; and in this case, the second criterion is not met. Then, one is driven to the conclusion that si cannot be regarded as a designation for the Yin year. Whether or not the beginning of a si coincides with the first month of the Yin year, this matter can be decided upon the basis of inscriptional evidence. The following inscriptions show that the start of a si does coincide with the first month of the Yin year. 185 l&M)) 37844 On guiwei (day 40), the king made cracks and divined: "[I will] perform the yow-cutting sacrifice on the y/-sacrificial day, starting from Shangjia to many kings, there will be no disasters." [Day guimao] was in the ninth month. It was the fifth si of the king. Heji 37844 [15]HSihj£: fQtXrftffio f ^ B : t o ^iAP?E]o [H*]ih[j£: tf]-t(=^o [3Er^]B: t o [S+]J!X-, [¥]¥Sffi¥o 5 « ^ ^ » 37845 On guiyou (day 10), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." The king read cracks and declared: "Auspicious." [Day guiyou] was in the twelfth month. It was the sixth [si] of the king. On [guiwei] (day 20), the king made cracks [and divined: "In the next 10-day week,] there will be no disasters." [The king read cracks and] declared: "Auspicious." [Day guiyou] was in the twelfth month. On [jia]shen (day 21), the yz-sacrifice was offered to Zujia. Heji 37845 [16] H * 3 E hi£: [tjjt:(=^)?S]o i ^ 0 : t o &IL[R> ¥$]H*&¥, ^ i - fc [IE], H E i h l * : f5tr(=^o i r ^ B : l*f- ¥ T ^ l l ¥ o [H5P3E MJT;: [TJjtr^epfgo [ i £ H : Hf. [ ? J I I ] l l [ I I ^ o 6 37846 On guiwei (day 20), the king made cracks and divined: ['in the next 10-day week, there will be no disasters."] The king read cracks and declared: "Auspicious." [Day guiwei] was in the fifth [month. On jiashen (day 21), the king] performed the llH sacrifice to 5 Chang Yuzhi (1987: 240) cites these two inscriptions in order to reconstruct the five-ritual cycle. It is her opinion that the divination and sacrifice dates in the second inscription are missing. However, the damaged graph mtm (scale is 125%) of the sacrifice date indicates the date is jiashen (day 21), as she infers from the pattern of the five-ritual cycle. 6 The divination date guimao, sacrifice date jiachen, and the month notation five, are supplied by Chang Yuzhi (1987: 241) on the basis of the pattern of the five-ritual cycle. 186 Zujia. It was the seventh si of the king. On guisi (day 30), the king made cracks and divined: "[In the next] 10-day week, there will be no disasters." The king read cracks and declared: "Auspicious." [Day guisi] was in the fifth month. On jiawu (day 31), [the king] performed the wez'-sacrifice to Zujia. [On guimao, day 40, the king made cracks and] divined: "[In the next 10-day week, there will be no] disasters." [The king read cracks and declared: "Auspicious." Day guimao was in the fifth] month. [On jiachen, day 41, the king] presented tablets, [performed the yow-cutting and] the Rong sacrifices. Heji 37846 Based upon the inscriptions above, the present writer reconstructs the following calendar and five-ritual cycles. Table 8: Reconstruction of the Calendar and Five-Ritual Cycle for Three Sis Month Si Date Ritual Inscription 9 5 Guimao Yiri Shangjia 14 [9] [5] [Guichou] [9] [5] [Guihai] [Yiri Dajia] [10] [5] [Guiyou] [Yiri Xiaojia] [10] [5] [Guiwei] [10] [5] [Guisi] [Yiri Jianjia] [11] [5] [Guimao] [Yiri Qiangjia] [11] [5] [Guichou] [Yiri Yangjia] [11] [5] [Guihai] [12] [5] [Guiyou] [Yiri Zujia] [12] [*5] [Guiwei] [Ji gongdian] [12] [*5] [Guisi] [Ji Shangjia] [1] [6] [Guimao] [1] [6] [Guichou] [Ji Dajia] [2] [6] [Guihai] [Ji Xiaojia] [2] [6] [Guiyou] 187 Table 8: Reconstruction of the Calendar and Five-Ritual Cycle for Three Si [2] [6] [Guiwei] [Ji Jianjia] [3] [6] [Guisi] [Ji Qiangjia] [3] [6] [Guimao] [Ji Yangjia] [3] [6] [Guichou] [4] [6] [Guihai] [Ji Zujia] [4] [6] [Guiyou] [4] [6] [Guiwei] [5] [6] [Guisi] [Rong gongdian] [5] [6] [Guimao] [5] [6] [Guichou] [Rong Shangjia] [6] [6] [Guihai] [6] [6] [Guiyou] [Rong Dajia] [6] [6] [Guiwei] [Rong Xiaojia] [7] [6] [Guisi] [7] [6] [Guimao] [Rong Jianjia] [7] [6] [Guichou] [Rong Qiangjia] [8] [6] [Guihai] [Rong Yangjia] [8] [6] [Guiyou] [8] [6] [Guiwei] [Rong Zujia] [9] [6] [Guisi] [9] [6] [Guimao] [Yiri gongdian] [9] [6] [Guichou] [Yiri Shangjia] [10] [6] [Guihai] [10] [6] [Guiyou] [Yiri Dajia] [10] [6] [Guiwei] [Yiri Xiaojia] [11] [6] [Guisi] [11] [6] [Guimao] [Yiri Jianjia] [11] [6] [Guichou] [Yiri Qiangjia] 188 Table 8: Reconstruction of the Calendar and Five-Ritual Cycle for Three Si [12] [6] [Guihai] [Yiri Yangjia] 12 6 Guiyou 15 12 6 Guiwei Yiri Zujia 15 [1] [7] [Guisi] [Ji gongdian] [1] [7] [Guimao] [Ji Shangjia] [1] [7] [Guichou] [2] [7] [Guihai] [Ji Dajia] [2] [7] [Guiyou] [Ji Xiaojia] [2] [7] [Guiwei] [3] [7] [Guisi] [Ji Jianjia] [3] [7] [Guimao] [Ji Qiangjia] [3] [7] [Guichou] [Ji Yangjia] [4] [7] [Guihai] [4] [7] [Guiyou] [Ji Zujia] 5 7 Guiwei Zai Zujia 16 5 [7] Guisi Wei Zujia 16 5 [7] Guimao Rong gongdian 16 According to Example 14, there is guimao in the ninth month of the fifth si. Example 15 records that guiyou and guiwei are in the twelfth month of the sixth si. Based upon the pattern of the five-ritual cycle, guimao must be taken as the first gui-day of the ninth month of the fifth si, and guiwei the last gui-day of the twelfth month of the sixth si, so that all dates and month notations in these two examples can be accommodated. In addition, in Example 16, guiwei occurs in the fifth month of the seventh si. According to the pattern of the five-ritual cycle, it has to be interpreted as the first gui-day of that month. Such a reconstruction fits the pattern of the five-ritual cycle and accommodates all those dates and month notations in all three examples. It is thus a credible reconstruction. In the reconstruction above, the sixth si ends with the twelfth month and the seventh si begins with the first month. It appears that si refers to a time period from the first month to 189 the twelfth month, i.e., a calendar year. That si refers to a calendar year is supported by the materials in the bronze inscriptions as well. Chen Mengjia (1956: 237) and Chang Yuzhi (1998: 365) cite a number of examples of the phrase wei wang ji si in the Western Zhou bronze inscriptions. The Zhou people did not have the five rituals. There is no basis to take the word si in the phrase wei wang ji si, in the Western Zhou bronze inscriptions, to mean "sacrifice." Therefore, Chen Mengjia and Chang Yuzhi interpret this phrase as a designation for the year of the Zhou king. Since the Western Zhou Dynasty was founded immediately after the Yin Dynasty, it is reasonable to assume that the phrase wei wang ji si was used in the same way in the Yin OBI and Western Zhou bronze inscriptions. The forgoing analysis shows that the word si can refer to both the time period of one year and the time period of a calendar year in the Yin OBI. These two requirements for a word to be a designation for the Yin year have been fulfilled. Therefore, si is a designation for the Yin year. Following Dong Zuobin's explanation (1945.1.Ill: 2) on why si is a designation for the Yin year, Chen Mengjia (1956: 236-237), Xu Jinxiong (1968: 77) and Chang Yuzhi (1998: 354) all propose that as a designation for the year, si should refer to a solar year and not a lunar year. In this writer's opinion, their proposal is mistaken. Like other Chinese calendars, a Yin civil year consists of twelve lunar months in a normal year and thirteen lunar months in an intercalary year, as discussed in Chapter Two. A normal Yin civil year is 354 days long and an intercalary Yin civil year 383 days long. Whether it is a normal year or an intercalary year, a Yin civil year is not 365 days long. Because a Yin year cannot be as long as a solar year, it is impossible for the word si to refer to a solar year. 4.2.3 Sui The word sui & is inscribed as K2I in the OBI. From Period I to V, sui has two usages. First, it can be used as a verb, meaning "to cut." In this case, it should be read as gui #'J. Second, it can also be used as a time word, which is the usage related to this study. The following inscription is one example of its usage as a time word. 190 [17 ]Hf ih JS: 4 * ^ ^ . £3EA*E. 37849 Crack-making on guichou (day 50), [X] divined: "This season [we will] reap a harvest." It was greatly auspicious. [Day guichou] was in the eighth month. It was the eighth si of the king. Heji 37849 In this inscription, the phrase jinsui specifies the time scope of the charge. There are no disputes over the word sui being a time word in such a context. However, there are different opinions about the length of the time period to which sui actually refers. David N. Keightley (2000) offers two translations for sui. On page 52, he translates sui as "season," without specifying the length of a season. On page 61, he translates it as "year." In the strict sense of the word, a "season" is not a "year." Hu Houxuan (1944: 2b-3a) proposes that sui is used as a designation for a year. This opinion is shared by Dong Zuobin (1945.1 .III: 2b) and Chang Yuzhi (1998: 344-351). On the other hand, Chen Mengjia (1956: 225-226) suggests that the Yin year has two suis. If so, sui cannot be a designation for a Yin year. This view is repeated by Wen Shaofeng and Yuan Tingdong(1983: 95-96). Given these two viewpoints on the time word sui, which is correct? The answer has to be decided by inscriptional evidence. The following inscriptions are relevant to the time period referred to by the word sui. [18]H0Ph 4-^^^o -B. i^M)) 24427 Crack-making on guimao (day 40), Da divined: "In this sui, the Shang will reap a harvest." [Day guimao was in] the first month. Heji 24427 [ 1 9 ] ^ f i h . &&%%-o ~Bo 24429 Crack-making on xinchou (day 38), Da divined: "In this sui, [the Shang] will reap a harvest." [Day xinchou was in] the second month. Heji 24427 [20] P P h j£: 4*£g[^o 3E ]£H: tfo &-£B° «f>H» 36977 Crack-making on X X , [diviner Y] divined: "In this sui, [the Shang] will reap [a harvest." The king] read cracks and declared: "Auspicious." [Day X X was] in the fifth month. Heji 36977 [ 1 7 ] H E h 4-£%%o £AJ!o 4£iA*E. l&m 37849 Crack-making on guichou (day 50), [X] divined: "This season [we will] reap a harvest." 191 It was greatly auspicious. [Day guichou] was in the eighth month. It was the eighth si of the king. Heji 37849 [21]H0Ph Aj£: ^ § i ¥ o - f B o l&m 24431 Crack-making on guimao (day 40), Da divined: "In this sui, [the Shang] will reap a harvest of broomcorn millet." [Day guimao was in] the tenth month. Heji 24431 [22]£Tih, 4-+-JT l&m 9650 Crack-making on yichou (day 2), the king divined: "In this sui, [the Shang] will reap a harvest." [Day yichou was in] the twelfth month. Heji 9650 These six examples show that the time period of sui includes the first, second, fifth, eighth, tenth and twelfth months, which indicates that sui refers to a period that is longer than a month. Currently, there are no inscriptions to show that the third, fourth, sixth, seventh, ninth and eleventh months are included in sui. On the other hand, as far as the present writer knows, there are no time words that refer to several separate months. It seems reasonable to draw the conclusion that, as a time word, sui can refer to any month of the Yin year. In other words, sui may refer to a time period of one year, which is the present writer's first criterion for a word being a designation for the Yin year. Does the word sui meet the second criterion? Can it refer to the time period of a calendar year, i.e., from the first to the twelfth month of the same Yin year? There are some inscriptions that may shed light on this question. Example 22 is already cited above. It divines whether the Shang would reap a harvest in this sui and its divination date is in the twelfth month, the last month of the Yin year. If sui ends with the twelfth month, it makes more sense for the Shang to divine whether they would reap a harvest in the next sui, rather than in this sui. Example 22 therefore suggests that the time period of sui may not end with the twelfth month. The Yin people did divine if they would reap a harvest in the next sui. Below are several such examples. [23]rJJHjg: ^ ^ g ^ . S A ^ ho 33241 On wuyin (day 15), [X] divined: "In the coming sui, [the people of] big settlements will reap a harvest." In the sixth month, [X] divined. Heji 33241 [24] ¥ T h: AJ!o i&m 9659 Crack-making on jiazi (day 1), [X divined]: "In the coming sui, [the Shang] will reap a harvest." [Day jiazi was in] the eighth month. Heji 9659 192 ID) 4.134.27 Crack-making on xinmao (day 28), the king divined: "In the coming sui, [the people of] For the phrase laisui , 'the coming sw/',' in these three examples, it is Chang Yuzhi's conclusion (1998: 250-255) that the word lai means 'next.' Also, it makes sense to understand "the coming sui" as "the next sui.'" If so, the topic of these three divinations is whether those people will reap a harvest in the next sui. According to the post-faces of Examples 23, 24, and 25, these three divinations are conducted in the sixth, eighth and ninth months, respectively, which indicates that the next sui may start in the period between the seventh and tenth months. Since Examples 23 - 25 suggest that the time period of sui possibly starts between the seventh and tenth months and Example 22 indicates that it may not end in the twelfth month, it appears that the word sui cannot refer to the time period of a calendar year. The second requirement for judging a word as a term for the Yin year is not fulfilled. Therefore, sui cannot be accepted as a designation for the Yin year. The bone graph for word nian in the Yin OBI is scribed as H i . It consists of two components: he 7R, 'millet,' and ren A, 'man.' Its meaning is "harvest," as is shown by Examples 18, 19, 21, 22, 24 and 25 in Section 4.2.3. This is the meaning of the word nian most frequently found throughout five periods of the Yin OBI. In Period I, however, there are several inscriptions in which a numeral occurs before the word nian. Without analyzing these inscriptions in detail, Hu Houxuan (1944: 2b) takes them as examples of nian being used as a designation for the Yin year. His opinion is repeated by Wen Shaofeng and Yuan Tingdong (1983: 94) and Chang Yuzhi (1998: 341-344). Dong Zuobin (1945.1 .III: 1) is more cautious. He points out that it is difficult to prove from these examples that nian is a term for the Yin year. Chen Mengjia's understanding (1956: 224) of the word nian in these examples is different from Hu Houxuan. He suggests that nian means "harvest season." Zheng Huisheng Y i will reap a harvest." [Day xinmao was in] the ninth month. Sude 4.134.27 4.2.4 Nian 193 (1984: 14) agrees with Chen Mengjia's rendering. It is apparent that there is no consensus yet as to whether the word nian is a designation for the Yin year. Below, the present writer will analyze whether nian is a term for the Yin year according to criteria stated in Section 4.2.1. It is very difficult to determine whether nian is a term for the Yin year, mainly because there is no inscription that indicates what the relationship is between this word and the month of the Yin year. Wen Shaofeng and Yuan Tingdong (1983: 94) have made effort to show a relationship; however, it appears that their interpretation of the following relevant inscriptions is problematic. S T ^ ¥ o «i=rii» 7049 On x-yin, [diviner Y] divined: "Que, [Z], andU [will reap a] harvest." [Day x-yin was in] the twelfth month. " Y i will not reap a harvest." Heji 7049 In their transcriptions, Wen Shaofeng and Yuan Tingdong do not put a period between the word nian and the phrase shier yue -f"Zl Jf. But the second inscription clearly shows that the topic of that divination is whether those persons would reap a harvest. The word nian in the first inscription of this example thus means "harvest." It is not a time word and should be separated from shier yue. [27] —Ho 9908 "... harvest." [Day xx was in] the first month. Heji 9908 [28] -tBo 9910 "... harvest." [Day xx was in] the seventh month. Heji 9910 [29] EBo i£M)) 11571 "... nian." [Day xx was in] the fourth month. Heji 11571 The authors of the Jiaguwen heji shiwen put a period between nian and these three month notations, which implies that they should be separated. If so, these examples cannot be taken as evidence for the word nian as a time word. A comparison between these three examples and Example 26 suggests that it is reasonable to separate the word nian and other month notations, a practice that Wen Shaofeng and Yuan Tingdong do not follow. Moreover, 194 even if Wen Shaofeng and Yuan Tingdong's punctuation is correct, it is not in agreement with the method of recording dates in Period V inscriptions. In Period V, a date occurs before a month notation, which in turn appears before a year notation, as is clearly shown in Examples 2 to 17 in this chapter. But in Examples 26 to 29, a month notation occurs after the alleged year notation. Such a disagreement certainly casts doubt on interpreting nian as a year notation in the three examples just considered. [ 3 0 ] H * K Jg: $ffi+<N2£0p-HF-, [ W o +F4 C ffi„ « ^ » 14770 Crack-making on guiwei (day 20), [X] divined: "[The king will] offer from the holocaust ten small specially reared sheep and cut open ten oxen for 111 in praying for nian." [Day guiwei was in] the tenth month. [This divination was] adopted. Heji 14770 This example is different from Examples 26 in that it is unlikely to be correct to supply the word shou 'to receive,' before nian because its context is different. In Example 26, there is no mention of any sacrifice. In this example, there are sacrificial victims. It seems a good choice to supply the word dao 'to pray for,' before nian because of the similar context in the following inscription. [31]EfJf h I T N ^ O i&m 10067 Crack-making on jiazi (day 1), Zheng divined: "[When the king] prays for harvest from Kui, [he will] offer six oxen from the holocaust." Heji 10067 In this example, the king offers from the holocaust six