下载掌阅APP，畅读海量书库
立即打开
There are four separate texts among the Tsinghua manuscripts that correspond to chapters in the
Yi Zhou shu
: *
Ming xun
命訓 *
Instruction on Mandates
, which is the second chapter in the received text;
(1)
*
Cheng wu
程寤 *
Awakening at Cheng
, which is listed in the “Xu” 序 “Sequence” of the text as the thirteenth chapter and for which there are fragmentary quotations by title in several medieval sources, but the contents of which are entirely missing from the received text; *
Huang men
皇門 *
August Gate
, which is the forty-ninth chapter, and which is not titled in the manuscript (the characters in the text corresponding to the received title are
, which the Tsinghua editors transcribe as
門 and read as
ku men
庫門 “depot gate,” but which might better be read as
hu men
胡門 “great gate”); and the self-titled
Zhai Gong zhi gu ming
祭公之顧命
The Duke of Zhai’s Retrospective Command
, which corresponds to “Zhai Gong” 祭公 “The Duke of Zhai,” the sixtieth chapter of the received text.
(2)
With the exception of *
Cheng wu
, which as noted is missing from the received text, these texts match the received texts quite closely; careful comparisons of the manuscript texts and the received texts would leave little doubt that we are dealing with but slightly different versions of the same texts.
(3)
Nevertheless, despite the general similarity between the corresponding texts, it is also the case that all four manuscripts contain numerous variant readings vis-à-vis their received counterparts. In many cases, it is possible to show—or at least to suggest plausibly—how these variants came about, most of them being the same sort of variants seen throughout traditional Chinese textual criticism. Differences of single characters are often one of three different types: classifier variation, in which the same phonetic component appears in both characters, but the semantic component, usually referred to as the signific or radical (
bushou
部首 or
pianpang
偏旁) is different; phonetic loans, when two characters of similar pronunciation, often with similar meanings, appear in the same place; and variants caused by graphic similarity, usually with quite different meanings. In addition to these three types of variants, other typical variations in texts concern the addition or deletion of words, sometimes because of an eye-skip in copying from one text to another, and sometimes apparently for the purpose of making an intended reading clearer or more explicit. In addition, variations in texts are sometimes caused by different interpretations of punctuation, which is a feature more or less unique to manuscripts. Each of these types of variants can be seen in the four Tsinghua manuscripts vis-à-vis the corresponding chapters in the
Yi Zhou shu
. Although it would be mistaken to say that in all cases the readings of the manuscripts are preferable to those of the received texts, it is certainly true that the manuscripts are generally more sensible than the received texts. Indeed, comparing the manuscripts against the received texts, it is hard not to agree with the statement by Sun Yirang 孫詒讓 (1848-1908) quoted at the beginning of the first chapter, that “even the collations of Lu [Wenchao] 盧文弨 (1717-96) and Zhu [Youzeng] 朱右曾 (
jinshi
1838) are not without the failings of unfounded changes. Thus, it is not strange that this book has been filled with blind errors and broken passages!”
It is doubtless the case that some of the neglect of the Yi Zhou shu throughout Chinese history has been due to the corrupt state of the text, and it is probably also the case that the corrupt state of the text is due—at least in some part—to the neglect of the text throughout the same period. Although the four Tsinghua manuscripts provide early exemplars for well less than one-tenth of the original text of the Yi Zhou shu (if we can even speak meaningfully of an original text), still by showing how these four texts looked at least at one very early moment in time, not only do they provide the best evidence we are likely to have for these texts themselves, but they also suggest some of the problems that likely beset the other chapters for which early manuscripts have not yet been found. In the following discussion, I will illustrate these different types of variants found in these four manuscripts vis-à-vis their corresponding chapters in the Yi Zhou shu, and will provide some brief discussion of how they influence the reading of the two different versions of the texts. While scores of examples could potentially be cited, (4) I will restrict my discussion here to just one or two examples drawn from each of the four different manuscripts to be translated below. I have selected these both for their importance to understanding the contribution the manuscripts make to the understanding of the texts, and also to illustrate the different categories of variants to be seen across these manuscripts. I will treat the texts in the inverse order in which they are found in the received text of the Yi Zhou shu .
The first example comes from the
Zhai Gong zhi gu ming
祭公之顧命
The Duke of Zhai’s Retrospective Command
manuscript, which as noted above corresponds with the “Zhai Gong” 祭公 “Duke of Zhai” (Ch. 60) chapter of the received text. The text purports to record a conversation between an unnamed king, but one who is surely King Mu of Zhou 周穆王 (r. 956-918 BCE), and a senior—and dying—minister named Moufu, Duke of Zhai
公
父 (祭公謀父 in the received text). The text begins with the king recalling Zhou history, especially the receipt of the Mandate of Heaven and of the Zhou conquest of Shang, and also the role that the Duke of Zhai’s own family played in it. The king then asks the duke to overcome any reticence he might have and state frankly his advice for the king. The Duke exclaims the validity of this request (
yun zai
允哉) and then issues a summons (
nai zhao
乃召), according to the manuscript, or, in the received text “consents to your summons” (
yun nai zhao
允乃詔; or perhaps “consents to the summons”), followed by six characters which in the received text read as follows:
bi huan yu li min ban
畢桓于黎民般. These are explained by Kong Chao 孔晁 (fl. 266), the earliest commentator on the text, as:
般，樂也。言信如王告，盡治民樂政也。
Pan 般 means “happiness.” It says that truly as the king has reported, he has thoroughly ruled the people with a happy governance. (5)
In the manuscript, these six characters read rather differently (including punctuation embedded in the text):
利毛班
Reading backwards through these six characters, the last character
ban
班 is obviously phonetically related to
ban
般, the last character in the received text, and could certainly be an example of a phonetic loan, in which two different characters “spell” a single Chinese word differently, whether that word should mean “happiness,” as Kong Chao suggested, or not. However, the characters immediately preceding these two alternatives,
mao
毛 in the manuscript and
min
民 in the received text are not susceptible to the same sort of comparison. Instead, it is clear that their only relationship is the graphic similarity of the two characters. Although the shape of the two characters was only modestly similar during the Warring States period
(
mao
vs.
min
), the later
kaishu
楷書 forms,
mao
毛 vs.
min
民, are easy to confuse. Whereas the
min ban
民般 of the received text makes very little sense, the editor of the Tsinghua manuscript Shen Jianhua 沈建華 pointed out that the
mao ban
毛班 of the manuscript is the name of an important figure at the time of King Mu, known both from Western Zhou bronze inscriptions and from previously unearthed Warring States manuscripts, including both the
Mu tianzi zhuan
穆天子傳
Biography of the Son of Heaven Mu
and the
Zhushu jinian
竹書紀年
Bamboo Annals
.
(6)
Once this identification is made, it is possible also to identify the two preceding characters in the manuscript,
利, as another name. Li Xueqin李學勤, the general editor of the Tsinghua manuscripts, had already noted that it can also be identified with a figure from King Mu-period bronze inscriptions and also from the
Mu tianzi zhuan
.
(7)
The corresponding characters in the received text,
yu li
于黎, again make little or no sense in the context (despite the sense attributed to them by Kong Chao). Again, it is possible to see the process by which this corruption of the text may have come about, and again the process involves both graphic and phonetic confusion. The first character
or 汬 was doubtless written in some version of the text as
jing
井 (which is, indeed, the form of the character in bronze inscriptions for the figure’s name), which must have been miscopied at some point as 于 (essentially losing the right-hand vertical stroke). On the other hand, the character
li
利, which means “benefit, profit” and which was routinely written as
in Warring States manuscripts, could reasonably have been understood by a scribe as the protograph of
li
黎, in which it is the top component and provides the pronunciation.
Li
黎 is usually an adjective meaning “numerous,” and often forms a compound with
min
民, in the sense of the “numerous people” or “common people.” Once a scribe of the “Zhai Gong” text had miscopied 毛 as
min
民, it would have been a natural next step for him to understand
li
利 as
li
黎.
Having identified these two names, it is then an easy matter to identify the first two of these six characters,
in the manuscript (where they are set apart with a punctuation mark, presumably indicating that they are to be read separately) and
bi huan
畢桓 in the received text, as yet another name. The component
bi
畢 is clearly to be seen on the bottom right-hand side of the graph
, and surely serves as its phonetic element. Thus, the graph should be read as
bi
畢, which was the name of yet another important lineage of the Zhou royal family. The Bi lineage is very well represented among mid-Western Zhou bronze inscriptions (i.e., those more or less contemporary with the reign of King Mu), even if the individual name Bi Huan 畢桓 does not seem to appear.
(8)
Nevertheless, it is easy to see that
and
huan
桓 share the same phonetic, 亘, and thus can be interchangeable as ways of writing a name.
Corroboration that these three pairs of characters are indeed to be read as names is found in the text of the manuscript Zhai Gong zhi gu ming. These six characters are followed by a statement by the Duke of Zhai, addressed to the “Three dukes” ( san gong 三公). In the received text, the graph san 三 “three” is missing, and gong 公 “duke” comes before yue 曰 “to say,” such that this would indicate explicitly that it was the Duke of Zhai speaking, though this might be seen as redundant since the immediately preceding sentence already indicated that it was the duke speaking. Finally, the received text inserts the two characters tianzi 天子 “Son of Heaven,” indicating explicitly that these remarks are addressed to the king. While tianzi is sensible of understanding here, three subsequent passages even in the received text of the chapter refer to the “three dukes”; without introducing them as they are introduced in the manuscript, it is impossible to know who they are, and it is perhaps understandable that Kong Chao made no attempt to explain these references to them.
The passage reads in its entirety, in the two versions:
Zhai Gong zhi gu ming :
公懋拜手稽首，曰：“允哉。”乃召畢桓、井利、毛班，曰：“三公，謀父朕疾，惟不瘳，敢告天子。”
The duke earnestly saluted with his hands and touched his head to the ground, saying: “Truly, indeed,” and then summoned Bi Huan, Jing Li and Mao Ban, saying: “Three dukes, as for Moufu, my illness being incurable, I dare to report to the Son of Heaven.”
“Zhai Gong”:
祭公拜手稽首，曰：“允。乃詔，畢桓于黎民般。”公曰：“天子，謀父疾維不瘳，敢告天子。”
The Duke of Zhai saluted with his hands and touched his head to the ground, saying: “True is your summons, entirely to bring about wholeness in the common people’s service.” The duke said: “Son of Heaven, Moufu’s illness being incurable, I dare to report to the Son of Heaven.”
By providing the names of these three figures from the time of King Mu of Zhou, and especially figures who are known only from paleographic sources, the Tsinghua manuscript Zhai Gong zhi gu ming provides important confirmation of the historicity of the “Zhai Gong” chapter. (9) That these names were unknown to Kong Chao in the third century CE, and doubtless were also unknown to the earlier copyist of whatever text he was using, not only suggests that the Tsinghua manuscript could not have been produced by a recent forger, but also that its original content doubtless goes back to close to the time of the events that the text purports to narrate.
The next examples of variants that I wish to introduce come from the manuscript that corresponds to the “Huang men” 皇門 “August Gate” chapter of the Yi Zhou shu. The closing passage of this text is a rousing denunciation of the last king of the Shang dynasty and his reliance on sycophantic ministers. It is entertaining enough—and still timely enough—that it deserves to be read in full, but I will content myself here with just the first two sentences. These two sentences alone include any number of variants vis-à-vis the received text, most of which will be discussed more fully in the notes to the translation of the manuscript in Chapter Five below. Here, I will consider only two examples that illustrate different sorts of variants. These sentences read as follows in the manuscript (in a more or less direct transcription with added punctuation to help with the interpretation); I highlight the two passages that contain examples of variants.
至于厥後嗣立王，乃弗肯用先王之明型， 乃惟汲汲胥驅胥教于非彝 ，以家相厥室，弗恤王邦王家。 惟媮德用，以問求于王臣 ，弗畏不祥，不肯惠聽亡罪之辭，乃惟不順是治。
Coming to their last successor and installed king, who then was unwilling to use the past kings’ bright model, then it was that he anxiously raced all over instructing about improprieties, with the families advising their houses and not caring for the royal country and royal family. It being false virtue that he used in order to question and seek out the royal ministers , not fearing what was not auspicious, and not willing generously to listen to the explanations of the innocent, then discordant was this governance.
The first accusation against the last king of Shang is that he engaged in improprieties, especially vis-à-vis his ministers and advisers:
乃惟汲汲胥驅胥教于非彝。
then it was that he anxiously raced all over instructing about improprieties.
In the received text, the corresponding sentence reads as follows:
維時及胥學於非夷。
it was this and the assistants studying about unevenness.
We can disregard for now the differences between jiao 教 “to instruct” and xue 學 “to study,” and fei yi 非彝 “improprieties” and fei yi 非夷 “unevenness”; jiao and xue are but two different readings of the same word, (10) and yi 彝 (*ləi) and yi 夷 (*ləi) are both exact homophones and synonyms and thus were routinely used for each other in ancient texts. (11) Here I propose only to discuss the difference between the 訯訯疋驅 of the manuscript and the ji xu 及胥 of the received text. Qing-dynasty commentators already identified the occurrence of ji 及, typically the conjunction “and,” as anomalous, suggesting different solutions. For instance, Lu Wenchao 盧文弨 (1717-1796), the author of the first critical edition of the text, noted that his contemporary Zhao Ximing 趙曦明 (1705-1787) suggested reading this as fan 反 “reverse” instead of ji 及. (12) Somewhat later, Chen Fengheng 陳逢衡 (1778-1855) emended the text to read nian 㞋. (13) Finally, Wang Niansun 王念孫 (1744-1832) noted that his son, Wang Yinzhi 王引之 (1766-1834), had suggested that ji 及 should originally have been written as nai 乃 “then.” (14) All three of these suggestions assume a simple graphic error, but none of them is based on any textual support or provides any apparent improvement in meaning. However, the Tsinghua manuscript provides yet a different variant, and one that makes the text immediately clear. The Tsinghua manuscript adds a “language” (言) signific to the base character 及, i.e., 訯, not necessarily to be read as sa 訯, which dictionaries define as “emphatic speech” ( qiangshi yanyu 強事言語), but probably indicative that it is to be read differently from ji 及. More important, the manuscript also adds a duplication mark after it, indicating that it is to be read twice. The Tsinghua editors suggest that it is a phonetic loan for ji 急, “anxious,” the reduplicative emphasizing the degree of anxiety, (15) which would seem to fit the context here. This is followed in the manuscript by shu qu 疋驅 and in the received text by just xu 胥. The Tsinghua editors suggest, reasonably, that shu 疋 should be read as xu 胥, the same graph as found in the received text. However, this word has different meanings when used as a noun, as it appears to be used in the received text, or as a verb (or an adverb), as it is in the manuscript. As a noun, it refers to a minor official, an “assistant.” As a verb it means “to await” or “to assist,” and as an adverb it means either “mutually” or “thoroughly.” It would seem that in the received text, the word is understood as a noun, following the conjunction ji 及 “and,” as it does. But this seems to deflect the focus from the improprieties of the last Shang king, and imputes them to his advisers. The manuscript, on the other hand, describes the last Shang king as very much personally engaged in these improprieties. Thus, the simple failure to copy a duplication mark seems to have caused a complete breakdown in the meaning of the received text. (16) The Qing-dynasty commentators identified this breakdown but, in the absence of early manuscript support, they could not adequately resolve it.
A second variant to note occurs in the second sentence, which reads as follows, first in the manuscript and then in the received text:
惟媮德用，以問求于王臣，弗畏不祥。
It being false virtue that he used in order to question and seek out the royal ministers, not fearing what was not auspicious.
維德是用，以昏臣作威不祥。
It being virtue that was used, with ignorant ministers making awesome what was not auspicious.
I think it goes without saying that the received text here is nonsense, antithetical to the general meaning of the passage: what could it possibly mean that “ignorant ministers make awesome what was not auspicious”! The proximate origin of this confusion lies in the graph
found in the manuscript, made up of the signific
er
耳 “ear” and the phonetic
hun
昏 (*hmən) to write the word
wen
問 (*məns) “to ask” (or
wen
聞 [*mən] “to hear”), a form that was common during the Warring States period. However, by the Han period, the words came to be written with the phonetic
men
門(*mən), i.e., 問 and 聞.
(17)
However, in some version of the text “Huang men,” either the graph
or just the component 昏 survived, and was read as the word
hun
昏 “murky; dim.” Other problems must have subsequently compounded this misreading: the adjective
yu
媮 “false” dropped out of the text before
de
德 “virtue,” and
fu
弗 “not” was somehow substituted with
zuo
作 “to make,” though it is not so easy to see just how these changes may have come about.
(18)
However, their result was enough to render a text that was originally quite straight forward all but unintelligible.
Turning from the “Huang men” chapter to the third of the four chapters of the Yi Zhou shu under consideration in this book, the “Cheng wu” 程寤 “Awakening at Cheng,” we are faced with a very different set of problems concerning textual variants. This is because this chapter is one of the eleven chapters of the received text that was lost, presumably toward the end of the Northern Song dynasty. However, in the case of “Cheng wu,” because it concerned a dream, and dreams were a favorite topic for medieval Chinese encyclopedias, there are numerous quotations of the opening portion of the text. Indeed, it was these quotations, some of them quoting the text by name, that allowed the Tsinghua editors to identify their manuscript as the “Cheng wu” chapter. I have dealt with this manuscript in detail in a separate study, but the evidence that it presents both regarding textual variants and also regarding the historicity of the “Cheng Wu” text itself is compelling enough to warrant repeating here. (19)
The manuscript opens by recounting a dream that Tai Si 大姒, the wife of King Wen of Zhou 周文王 (r. 1099/1059-1050 BCE), had concerning bushes or trees at the court of the Shang king, the planting of another tree there by their son Fa 發, the future King Wu of Zhou 周武王 (r. 1049/1045-1043 BCE), and the transformation of some of this foliage into other types of trees. Tai Si is said to have awakened in alarm, and reported the dream to King Wen. King Wen too seems to have been startled by it, and enlisted clairvoyants at the Zhou court to interpret the dream. The clairvoyants exorcised the subjects of the dream and then performed various sacrifices and rituals in the Zhou court. King Wen and the Crown Prince Fa then pronounced the dream to be a “lucky dream” ( ji meng 吉夢) portending that they would receive from the “august Di on high” ( huang Shang Di 皇上帝) the Shang “mandate” ( ming 命) to rule. The Tsinghua manuscript then goes on to offer various interpretations and to draw lessons concerning government.
Most of the attention given to this text—both in medieval and also in modern times—has been devoted to this opening narrative of the dream. In the manuscript, this portion reads as follows:
惟王元祀正月既生魄，大姒夢見商廷惟棘，乃小子發取周廷梓樹于厥間，化爲松柏棫柞。寤驚，告王。王弗敢占，詔太子發，俾靈名總祓，祝祈祓王，巫率祓大姒，宗丁祓太子發，幣告宗祊、社稷，祈于六末、山川，攻于商神。望徵，占于明堂。王及太子發並拜吉夢，受商命于皇上帝。
It was the king’s first year, first month, after the growing brightness. Tai Si dreamed of seeing that in the Shang court were brambles, and then that the young son Fa took the Zhou court’s catalpa and planted it in their midst, transforming into pine and cypress, white oak and sawtooth oak. She awakened alarmed, and reported it to the king. The king did not dare to prognosticate it, but summoned Crown Prince Fa and had Clairvoyant Ming perform a general exorcism, Invocator Qi exorcise the king, Magician Shuai exorcise Tai Si, and Templar Ding exorcise Crown Prince Fa, reporting with silks at the ancestral temple gate and to the altars of state, praying to the six extremities and the mountains and rivers, and casting out the Shang spirits. They looked to the signs and prognosticated in the Bright Temple. The king and Crown Prince Fa together bowed to the auspicious dream and received the Shang mandate from the august Di on high.
Of the many medieval quotations of the text, for our purposes here it will suffice to quote just two, both of which appear in the Taiping yulan 太平御覽 Imperial Survey of the Taiping Era, completed in 984. Both of these quotations, which are found in widely different sections of the Tai-ping yulan, mention the text by title. As I will have occasion to note in the Appendix to this book discussing the textual history of the Yi Zhou shu , differences between these two quotations suggest that at least two different versions of the “Cheng Wu” text may have been extant as late as the early Northern Song dynasty.
十年正月，文王自商至程。大姒夢見商庭生棘，太子發取周庭之梓樹之于闕間。梓化爲松柏柞棫。覺而驚，以告文王。文王不敢占，召太子發，命祝以幣告于宗廟群神，然後占之于明堂。及發並拜吉夢，遂作《程寤》。
In the tenth year, first month, King Wen from Shang arrived at Cheng. Tai Si dreamed of seeing in the Shang court growing brambles, and the Crown Prince Fa taking the Zhou court’s catalpa and planting it between the gatehouses. The catalpa transformed into pine and cypress, sawtooth oak and white oak. Awakening and being alarmed, she reported it to King Wen. King Wen did not dare to prognosticate, but summoned the eldest son Fa and commanded the priests to use silk to report to the many spirits in the ancestral temple, and only afterwards prognosticated it in the Bright Hall. Together with Fa they both bowed to the auspicious dream, and subsequently made Awakening at Cheng . (20)
又《程寤》曰：文王在翟。大姒夢見商之庭産棘，小子發取周庭之梓樹於闕間，化爲松柏棫柞。驚，以告文王。文王曰：“召發于明堂，拜告夢，受商之大命。”
Again, Awakening at Cheng says: King Wen was at Di. Tai Si dreamed of seeing that the Shang’s court produced brambles, and that the young son Fa took the Zhou court’s catalpa and planted it between the gatehouses, transforming into pine and cypress, white oak and sawtooth oak. Being alarmed, she reported to King Wen, and King Wen said: “Summon Fa to the Bright Hall,” and bowed to report the dream, receiving Shang’s great mandate. (21)
It is easy to see that the opening of the Tsinghua manuscript * Cheng wu does indeed coincide to a very great extent with the information contained in these quotations, as it does with other earlier quotations as well. Notably, the quotations mention the same six trees seen in the manuscript text: “brambles” ( ji 棘), “catalpa” ( zi 梓), “pine” ( song 松), “cypress” ( bai 柏), “white oak” ( yu 棫) and “sawtooth oak” ( zuo 柞). However, while the quotations as well as the Tsinghua manuscript agree that there were brambles growing in the Shang court, and that Fa (i.e., the eventual King Wu), whether described as the “young son” ( xiaozi 小子) or “Crown Prince” ( taizi 太子), took a catalpa from the Zhou court and planted it in the midst of that court, there is one important point of difference between these two quotations (a difference seen also in other quotations): the first quotation states explicitly that it was this catalpa that “transformed into” ( hua wei 化爲) “pine and cypress, white oak and sawtooth oak,” while both the Tsinghua manuscript and the second Taiping yulan quotation leave unspecified what transformed into “pine and cypress, white oak and sawtooth oak.” This is a point of considerable importance in the interpretation of the dream, and one that I have examined in detail in that previous study. Here I forgo all of the interpretation of the dream and also much of the supporting evidence, focusing just on the immediate variants.
The Tsinghua manuscript contains one explicit variant vis-à-vis both quotations, writing that Fa planted the catalpa “in their midst” ( yu jue jian 于厥間), obviously referring to “the brambles,” whereas the Taiping yulan quotations (and indeed, all other quotations) both write that he planted it “between the gatehouses” ( yu que jian 于/於闕間). In addition to this one explicit variation, the manuscript and one of the quotations also reveal one implicit variation, as noted above: whereas the first Taiping yulan quotation writes the “catalpa” as the explicit subject of the following sentence, making explicit what transforms into “pine and cypress, white oak and sawtooth oak,” neither the Tsinghua manuscript nor the second Taiping yulan quotation provides a subject before the verb “to transform into” ( hua wei 化爲). Especially in the Tsinghua manuscript, it is my understanding that it is not the catalpa that is transforming, but rather the brambles, which is the immediate antecedent, even if these brambles are referred to only with the preposition jue 厥 “their.”
It is worth considering this explicit variant. It is easy to see the graphic relationship between the characters for que 闕 “gatehouse” and jue 厥 “their,” and also easy enough to understand why a copyist or copyists and/or an editor or editors of the “Cheng Wu” text might have chosen to add a “gate” ( men 門) signific to 欮 (certainly the original form of jue 厥), turning it into “gatehouse”: que 闕. This is perfectly consistent with traditional Chinese reading practices. However, it is also perfectly possible to read the phrase as written in the Tsinghua manuscript, with the pronoun jue pointing back to the brambles. Indeed, I have demonstrated that it is the only way to make sense of the dream.
However, once that copyist or editor transformed jue 厥 “their” into que 闕 “gatehouse,” the following sentence concerning the transformation into the other four trees was left without a subject, whether explicit or implicit. It is my contention that either this editor or some subsequent editor took it upon himself to correct the text, by adding such a subject, in this case the “catalpa.” This constituted a serious misunderstanding of the dream, but one that was to prove all too influential not only for medieval readers (who would go on to quote the text), but also for contemporary scholars, almost all of whom have accepted the reading of the quotations instead of the reading of the manuscript, and all of whom have badly misinterpreted the dream. (22) This goes to show just how one simple variant can lead to another variant, and so on until the text becomes unintelligible.
The fourth of the four chapters of the Yi Zhou shu represented among the Tsinghua manuscripts is entitled “Ming xun” 命訓 “Instruction on Mandates.” Unlike “Cheng wu,” “Ming xun” is represented in the received text, and unlike either the “Zhai Gong” or “Huang men” chapters, both of which are written in a quite archaic Chinese, it is written in a rather standard classical Chinese, reflective of the language of the early to mid-Warring States period. As such, it offers fewer problems of understanding than either “Zhai Gong” or “Huang men.” Nevertheless, the Tsinghua manuscript of the text also presents a number of variants vis-à-vis the received text. (23) Here I propose to examine just one of these variants that involves the addition of a signific in the received text. As in the case of the added “gate” signific for que 闕 in the received text of “Cheng wu,” the added signific in “Ming xun” is also consistent with Chinese reading practice. However, also as in the case of “Cheng wu,” it undercuts the argument of the text, and actually makes nonsense of the text.
The final passage of the text, in both the manuscript and the received text, works forwards and backwards through twelve steps by which government should deal with the people. In the received text, this reads as follows, highlighting the relevant clauses:
撫之以惠，和之以均，斂之以哀，娱之以樂，慎之以禮，教之以藝，震之以政，動之以事，勸之以賞，畏之以罰， 臨之以忠 ，行之以權。權不法， 忠不忠 ，罰不服，賞不從勞，事不震，政不成，藝不淫，禮有時，樂不滿，哀不至，均不壹，惠不忍人。
Soothe them with generosity, harmonize them with equality, gather them in mourning, entertain them in joy, caution them with ritual, teach them with the arts, shake them with government, move them with work, encourage them with awards, terrify them with punishments, look upon them with loyalty, and mobilize them with balance. Balance is not legalistic, loyalty is not loyal , punishments are not submissive, awards are not loosely praised, work does not shake, governance is not complete, the arts are not excessive, the rites are timely, joy is not full, sorrow is not pervasive, equality is not unitary, and generosity is not tolerant of others.
Much of this seems straight-forward, and indeed is matched quite closely by the manuscript. However, there is one phrase that seems to be contradictory: zhong bu zhong 忠不忠 “loyalty is not loyal.” Although commentators such as Chen Fengheng 陳逢衡 and Zhu Youzeng 朱右曾 at least attempted to reconcile this phrase, suggesting in the first case that internal loyalty could manifest externally as disobedience or recalcitrance ( wu 忤), or, in the second case, that “minor loyalty” ( xiao zhong 小忠) is not what is meant by “loyalty,” this still seems awkward. It seems that only one critic, Ding Zongluo 丁宗洛 (1771-1841) saw that the text must be corrupt; he suggested that it should here read that “loyalty” is bu zhong 不中 “not centered,” which is to say that he noted an extraneous “heart” signific (心). (24) This is reasonable, and the Tsinghua manuscript shows that his instincts at least were right. However, the manuscript also shows a better reading. In the first list of government practices, the ruler is called to “look upon [the people] with centeredness” ( lin zhi yi zhong 臨之以中 ), zhong 中 “centered” here referring to notions of impartiality or equality, rather than loyalty. This is why in the second string of negative qualifications, it says that “centeredness is not loyal” ( zhong bu zhong 中不忠), which is to say that to be impartial is not necessarily to be loyal. On the other hand, the passage goes on in the manuscript to qualify this still further, with a sentence that is very hard to understand:
以中從忠則尚，尚不必中。
Using centeredness to accompany loyalty then there will be promotions, but the promotions will not necessarily be centered.
This seems to mean that if impartiality—middle-of-the-roadness—is equated with “loyalty,” the result itself may not be centered. In other words, impartiality is not the same thing as being wishy-washy and refusing to take a stand vis-à-vis loyalty. This is a very subtle distinction, and I admit I may be over-reading it. But the text is certainly striving to make an important distinction between two related, but different words: zhong 中 “centered,” and zhong 忠 “loyal.” It is perhaps not surprising that the received text, which as we have seen above had previously conflated these two words, would also conflate this sentence with another sentence to give a very different meaning.
以法從中則賞，賞不必中。
Using the law to accompany centeredness there will be awards, but the awards will not necessarily be centered.
This may or may not entail essentially the same meaning as the manuscript, but it certainly loses the wordplay between zhong 中 “centered” and zhong 忠 “loyal,” and thus makes for a much flatter presentation.
The variants between the Tsinghua manuscript versions of the Yi Zhou shu texts and the corresponding chapters in the received text examined here ought not be taken to exaggerate the difference between these texts. Except in the case of the * Cheng wu text, for which the corresponding chapter in the Yi Zhou shu has been lost since the early Northern Song dynasty, the other manuscripts are strikingly similar to their corresponding received text (of course, with due allowance for the Warring States Chu 楚 script in which they are written), such that there can be no question that in each case they are but two exemplars of a single text. As such, the manuscripts provide extraordinarily important evidence for the early history of the texts, not only some two thousand years earlier than the earliest extant printed text, but even five hundred or more years earlier than the earliest extant commentary. It is worth emphasizing that these manuscripts were copied in the age of Mencius 孟子 (c. 372-289 BCE) and Zhuangzi 莊子 (c. 369-286 BCE), not to mention more or less at the time of Plato (c. 429-347 BCE) and Aristotle (384-322 BCE). If they did nothing more than provide handwritten evidence of the intellectual thought of the time, the manuscripts would be well recognized as treasures of the world’s cultural heritage. But in the case of the Yi Zhou shu texts, the Tsinghua manuscripts do much more than just provide this handwritten evidence; they also allow us to see how at least these four texts looked and read in the hand of at least one or more scribes of that time, and how the “blind errors and broken passages” that later textual critics have detected in the received text of the Yi Zhou shu may have come about—and, perhaps more important, how they might be corrected. (25) In the case of these four chapters, I am confident that the manuscripts show that the texts—each in its own way—were originally very important historical and philosophical documents. Although they may have been “left over” from the compilation of the Shang shu 尚書 Exalted Scriptures, they very much deserve our attention today. I hope that the translations that follow will reflect at least some of the subtlety with which they were written.
(1) Li Xueqin 李學勤 ed.-in-chief, Qinghua daxue Chutu wenxian yanjiu yu baohu zhongxin 清華大學出土文獻研究與保護中心 ed., Qinghua daxue cang Zhanguo zhujian (wu ) 清華大學藏戰國竹簡 (伍) (Shanghai: Zhong Xi shuju, 2015), Shang 上 6-9 (full-size photos), Shang 45-57 (double-size photos), Xia 下 124-133 (transcription and notes, edited by Liu Guozhong 劉國忠).
(2) Li Xueqin 李學勤 ed.-in-chief, Qinghua daxue Chutu wenxian yanjiu yu baohu zhongxin 清華大學出土文獻研究與保護中心 ed., Qinghua daxue cang Zhanguo zhujian ( yi ) 清華大學藏戰國竹簡 (壹) (Shanghai: Zhong Xi shuju, 2010): * Cheng wu 程寤: Shang 上 6-7 (full-size photos), Shang 47-54 (double-size photos), Xia 下 135-141 (transcription and notes, edited by Liu Guozhong 劉國忠); * Huang men 皇門: Shang 上 18-21 (full-size photos), Shang 45-57 (double-size photos), Xia 下 163-172 (transcription and notes, edited by Li Junming 李均明); Zhai Gong zhi gu ming 祭公之顧命: Shang 上 22-25 (full-size photos), Shang 99-116 (double-size photos), Xia 下 173-179 (transcription and notes, edited by Shen Jianhua 沈建華).
(3) Although the text of “Cheng wu” is missing from the received text, there are more or less extensive quotations of the text in medieval sources and the manuscript matches these rather closely. Thus, it too can be identified as but a different version of the same text.
(4) Concerning just the one text * Huang men 皇門, one tabulation gives 153 variants between the Tsinghua manuscript and the received version of the text. According to this tabulation, of these 85 are graphic errors, 35 involve characters missing in the received text, 26 where there are extraneous characters, and 7 cases where characters are inverted; see Zhao Yasi 趙雅思, “Cong jianben yu chuanshiben Huang men kan gushu liuchuan yu jiaokan wenti” 從簡本與傳世本《皇門》看古書流傳與校勘問題, Wenxue jie ( Lilun ban ) 文學界(理論版), 2012.5: 202-208. The examples in all four manuscripts will all be discussed in the notes to the four separate translations.
(5) Yi Zhou shu 逸周書 (Sibu beiyao ed.), 8.2a.
(6) Qinghua daxue cang Zhanguo zhujian ( yi ), 173. For a more detailed discussion of this identification in English, see Edward L. Shaughnessy, “The Mu tianzi zhuan and King Mu Bronzes,” Bulletin of the Jao Tsung-I Academy of Sinology 1 (2014): 58-60.
(7) Li Xueqin 李學勤, “Mu gong gui gai zai qingtong qi fenqi shang de yiyi” 穆公簋蓋在青銅器分期上的意義, Wenbo 文博 1984.2: 7. A figure named Li 利 and identified as an “intendant” ( zai 宰) is mentioned in the Shi Ju fangyi 師遽方彝; Zhongguo Shehui kexueyuan kaogu yanjiusuo 中國社會科學院考古研究所 ed., Yin Zhou jinwen jicheng 殷周金文集成 (Beijing: Zhonghua shuju, 1984-1994), #09897. Li Xueqin argued persuasively that he was from the Jing 井 (or Xing 邢) lineage of the Zhou royal family, and thus would have been referred to as “Jing Li.”
(8) A case can be made that the name Bi Huan can be identified with a figure in the Mu tianzi zhuan named Bi Ju 畢矩:
己巳，至于文山，西膜之所謂
，觴天子于文山。西膜之人乃獻食馬三百、牛羊二千、穄米千車，天子使畢矩受之。
On jisi (day 6), arriving at Wenshan, which in the western regions is called .., they feasted the Son of Heaven on Wenshan. The men of the western region then presented three hundred feed horses, two thousand cows and sheep, and a thousand cartloads of grain. The Son of Heaven sent Bi Ju to accept them.
It is possible that
,
huan
桓 and
ju
矩 are but different transcriptions of the same original graph. Although the “bird” signific (i.e., 鳥) of
and the “arrow” signific (i.e., 矢) of 矩 would seem to be quite different, the “arrow” signific is reasonably similar in shape to the “wood” signific (i.e., 木) with which the name is written in the received “Zhai gong.” It is easy to see the graphic similarity between 亘 and 巨; if the two forms were not different transcriptions of the same archaic graph, they may well have been mistaken one for the other in the course of the two texts’ transmission in later ages. For this suggestion, see Shaughnessy, “The
Mu tianzi zhuan
and King Mu Bronzes,” 65-67.
(9) For studies making strong arguments for the historicity of the “Zhai Gong” chapter of the Yi Zhou shu, see Li Xueqin 李學勤, “Zhai Gong Moufu ji qi delun” 祭公謀父及其德論, in Li Xueqin, Li Xueqin ji 李學勤集 (Harbin: Heilongjiang Jiaoyu chubanshe, 1989), 186-192; Li Xueqin, “Shi Xun gui yu ‘ Zhai Gong ’” 師詢簋與《祭公》, Guwenzi yanjiu 古文字研究 22 (2000): 70-71. See, too, Shaughnessy, “Texts Lost in Texts.”
(10) Many ancient Chinese verbs involving communication can be read in either direction, the cases commonly cited being mai 買 “to buy” and mai 賣 “to sell,” and shou 受 “to receive” and shou 授 “to give,” now written with different graphs but originally written identically. Examples of this phenomenon could be multiplied many times over, and certainly include jiao “to instruct” and xue “to study,” which were also originally written with the same graph (though I should hasten to add that the graph in the manuscript version of the text does include the signific pu 攴 [a hand holding a stick], which came to differentiate jiao from xue ).
(11) For examples of the two characters being used as phonetic loans for each other, see Bai Yulan 白於藍, Jianbo gushu tongjiazi daxi 簡帛古書通假字大系 (Fuzhou: Fujian Renmin chubanshe, 2017), 503, 504.
(12) Lu Wenchao, Yi Zhou shu , 5.10b.
(13) Chen Fengheng 陳逢衡, Yi Zhou shu buzhu 逸周書補注 (Jiangdu 江都: Xiumei shan guan 修梅山館, 1825), 12.31a.
(14) Wang Niansun 王念孫, Du Yi Zhou shu za zhi 讀逸周書雜志, in Huang Qing jingjie xubian 皇清經解續編 (Jiangyin 江陰: Nanjing shuyuan 南菁書院, 1888), 2.9b ( juan 210.9b).
(15) Qinghua daxue cang Zhanguo zhujian ( yi ), 168 n. 39.
(16) It goes without saying that there is no direct connection between the Tsinghua manuscript of * Huang men, which was buried already in the fourth century BCE, and the received text of the Yi Zhou shu, the transmission of which must have involved numerous copyists and texts over the course of many centuries.
(17) For some indication of this distinction, see Gao Ming 高明 and Tu Baikui 涂白奎 ed., Guwenzi leibian ( Zengdingben ) 古文字類編 (增訂本) (Shanghai: Shanghai Guji chubanshe, 2008), 866. See too Ji Xusheng 季旭昇, Shuo wen xin zheng 説文新證 (Taibei: Yiwen yinshuguan, 2014), 842.
(18)
I suspect that the latter difference may have been the result of graphic similarity, though indirectly. It is possible that some version of the text had the negative
wang
亡 instead of the negative
fu
弗 for not “fearing what was not auspicious” (
wei bu xiang
畏不祥).
Wang
was written as
. On the other hand, the
zuo
作 “to do” of the received text was commonly written in early sources as
zha
乍, the early form of which was
. The two characters are quite similar, and easily mistaken.
(19) For this study, see Edward L. Shaughnessy, “Of Trees, a Son, and Kingship: Recovering an Ancient Chinese Dream,” Journal of Asian Studies 77.3 (2018): 593-609.
(20) Li Fang 李昉 et al. ed., Taiping yulan 太平御覽 (Beijing: Zhonghua shuju, 1960), 84 (“Ji meng shang” 吉夢上), 1836.
(21) Li Fang et al. ed., Taiping yulan, 533 (“Ming tang” 明堂), 2418.
(22) I am delighted to say that the Tsinghua editors have now accepted this interpretation; see Cheng Hao 程浩, “Qinghua jian Cheng Wu jiaoshi yijie” 清華簡《程寤》校釋譯介, manuscript dated 10 March 2020.
(23) For a study of some of these variants, see Edward L. Shaughnessy, “Varieties of Textual Variants: Evidence from the Tsinghua Bamboo-Slip * Ming Xun Manuscript,” Early China 39 (2016): 111-144.
(24) All three of these comments, originally published as Chen Fengheng 陳逢衡, Yi Zhou shu buzhu 逸周書補注 (Jiangdu 江都: Xiumei shan guan 修梅山館, 1825), Zhu Youzeng 朱右曾, Zhou shu jixun jiaoshi 周書集訓校釋 (Jiading 嘉定: Guiyanzhai 歸硯齋, 1846), and Ding Zongluo 丁宗洛, Yi Zhou shu guanjian 逸周書管箋 (Haikang 海康: Yuyuan 迂園, 1830), are quoted at Huang Huaixin 黄懷信, Zhang Maorong 張懋鎔 and Tian Xudong 田旭東 ed., Yi Zhou shu huijiao jizhu 逸周書彙校集注 (Shanghai: Shanghai Guji chubanshe, 2007), 36.
(25) I am well aware that recent studies of philology, now often termed “New Philology,” emphasize the distinct nature of each textual exemplar and seek to understand how it was produced and understood in its own individual historical and intellectual context, such that there is no “correct” text; for the best known of these, see Bernard Cerquiglini, Éloge de la variante: Histoire critique de la philology (Paris: Éditions du Seuil, 1989). I accept the insights that studies such as this have made to our understanding of different textual exemplars; nevertheless, they do not supplant the contributions that traditional textual criticism—both in the West and in China—has made to our understanding of textual transmission.