CHINESE HISTORY WRITING BETWEEN THE SACRED AND THE SECULAR
Yuri Pines
The nexus of religion and the writing of history in ancient China is a broad topic that may pertain to a large variety of texts and non-textual phenomena, depending on the definition of "religion" applied in the Chinese context. To focus the discussion more topically, in what follows I shall adopt a heuristically-convenient definition of "religious" and "secular" with regard to historical texts. I define a text, or an aspect thereof as "religious" insofar as it is related to communication with deities (particularly ancestral spirits), or insofar as it is supposed to have a certain sacral power of influencing the world through a proper choice of wording or proper arrangement of the material. Alternatively, I treat the text as "secular" if it lacks the above traits and is intended either for political education of the elite or for their entertainment.
My discussion will focus on those texts that had lasting impact on Chinese historiography, namely the canonical Chunqiu #fjlc (Annals) and its commentaries, and the first of the so-called "official histories," the Shiji j£|B (Records of the historian). In what follows, I shall outline, first, the cultic origins of the Chinese historiographical tradition and suggest that the Chunqiu should be understood primarily as a ritual rather than a historical text. Then, by analyzing two of the most important Chunqiu commentaries, the Zuozhuan ^rM (Zuo commentary) and the Gongyang zhuan ^-^^ (Gongyang commentary), I shall trace the bifurcation of history writing between the "secular" and the "religious" traditions. Finally, by briefly addressing the Shiji, I shall try to show that while this text belongs largely to what I define as a secular historiographical tradition, the invention of the genre of biographies by its author, Sima Qian WJHM (ca. 145-90 BC),1 added a new quasi-religious dimension to the Shiji and to the historical genre as a whole.
1 All dates are BC, unless otherwise indicated.
316
yuri pines
Cultic origins: from bronze inscriptions to the Chunqiu
The intrinsic connection between writing, historiography and cult is apparent in the two earliest examples of Chinese writings—oracle bones and bronze inscriptions associated respectively with the Shang M (ca. 1600-1046) and Western Zhou MM (1046-771) periods. The second of these genres is particularly germane to the present study because, as I hope to demonstrate, there are significant continuities between the bronze inscriptions and the annalistic genre exemplified by the Chunqiu of the state of Lu ||. As is well known, bronze inscriptions, especially those of the Western Zhou period, contain significant amounts of historical information, such as accounts of appointments, mentions of military campaigns and even—occasionally—references to land conflicts among the Zhou nobles. However, as insightfully pointed out by Lothar von Falkenhausen, it would be wrong to consider this information outside its cultic context. The very location of the inscriptions on the inside of the bronze vessels or on the verso side of the bells makes them inconvenient for reading. The purported audience of the inscriptions was thus in all likelihood not the living members of the kin who utilized the vessel in ancestral sacrifices, but primarily the ancestral spirits, who were supposed to "consume" the inscription during sacrificial rites.2
The bulk of historical information in the inscriptions is concentrated in what Falkenhausen aptly calls the "announcement of merit," in which the vessel's donor informs the ancestors of his recent achievements. These announcements of merit, particularly those performed in the so-called "documentary mode" are in turn based on earlier documentation created by court or lineage scribes (shi j£), the functionaries whose tasks comprised, among other things, the dual roles of keeping records of important events and communicating with the deities.3 The original
2 For inscriptions as historical sources, see Edward L. Shaughnessy, Sources of Western Zhou history (Berkeley, 1991); idem, "Western Zhou bronze inscriptions," New sources of early Chinese history: an introduction to the reading of inscriptions and manuscripts, ed., Shaughnessy (Berkeley, 1997), pp. 57-84; and Lothar von Falkenhausen, "Issues in Western Zhou studies: a review article," Early China 18 (1993), 139-95, on which much of the above discussion is based.
3 For the cultic origins of the scribes' function, see, e.g., Shirakawa Shizuka, "Shaku shi," Kbkotsu bungaku ronshu (Kyoto, 1955), vol. 1, pp. 3-68; Hu Dianxian, "Shi shi," Zhongguo gudai shi luncong (Fuzhou, 1981), vol. 1, pp. 210-24; Leon Vandermeersch, "Limaginaire divinatoire dans l'histoire en Chine," Bulletin de I'Ecole francaise d'Extreme Orient 79.1 (1992), 1-8.
chinese history writing
317
records, from which the inscriptions were extracted, were often lengthy and detailed. The mid-9th century inscription on the Jin Hou Su-bian-zhong WfllMifilit, for instance, tells of a military campaign that lasted for six months, supplying precise dating of major related events. Apparently, the scribes kept day-to-day records during the campaign, and at its successful conclusion they decided (perhaps in consultation with the lord of Jin, the vessels donor) what portion of their records was to be inscribed on the bells.4 It is important to note, then, that the inscription was not a primary historical document but a secondary record, selected and polished in accordance with the norms of this genre.
With this in mind, we may better understand the peculiar position of such inscriptions in Chinese historiography. While containing meticulous dating and an abundance of other details, such as place names, official titles and precise wording of investiture orders, the inscriptions were not meant to reproduce historical events as such, but rather those that would please the ancestors. Falkenhausen observes: "These documents do not convey the full range of human experience; the spirits were not supposed to know everything—there were to be no surprises, no irregularities in ritual exchange."5 Indeed, the inscriptions never record military defeats or other unpleasant events in the donors life. This selectivity results at times in euphemistic forms that completely distort the actual event. For instance, the famous narrative of early Zhou history on the Shi Qiang-pan j£Hif$£> states: "Great and excellent was King Zhao [r. ca. 977/5-957 BC]. He broadly overpowered Chu-Jing, opening the southern route."6
The author of the inscription, a scribe in the service of the Zhou royal house, was certainly well aware of the disastrous defeat of King Zhao's southern expedition, which marked the end of Zhou territorial expansion. Nonetheless, the inscription contains no reference to this; it was diplomatically refurbished to please Scribe Qiang's ancestors, who served at King Zhao's court. A similar formula appears in a recently
4 For the announcement of merits, see Falkenhausen, "Issues," 156-67; for the Jin Hou Su-bianzhong, see detailed discussion by Jaehoon Shim, "The 'Jinhou Su Bianzhong inscription and its significance," Early China 22 (1997), 43-76.
5 See Falkenhausen, "Issues," 152.
6 !&||B83£, JlctilJtjPJ* ftJfltifr- This example is discussed by Paul R. Goldin, in "Appeals to history in early Chinese philosophy and rhetoric," Journal of Chinese Philosophy, forthcoming. For the complete text of the Shi Qiang-pan, see Shaughnessy, Sources, 183-92.
318
yuri pines
discovered Qiu-pan MIS inscription from Yangjiacun ^^M, Shaanxi.7 Perhaps by presenting a revamped version of personal and general history, the donors sought to find favor with the spirits and increase their chances to receive the spirits' blessing.
All the features found in inscriptions, namely the meticulous dating, the abundance of technical details, formulaic language, selective recording and what seem like attempts to influence the future through properly recording the past, are evident in the genre of the state annals, of which the Chunqiu of the state of Lu is the only surviving representative. While many scholars routinely consider the Chunqiu as a purely "historical" text, I believe the evidence calls this assumption into question. Rather than being directed at posterity, the Chunqiu (and apparently other annals as well) were directed primarily at the ancestral spirits of the ruling house of each state, as I hope to demonstrate below.8
That the Chunqiu originated from the records of the ancestral temple is suggested by the Zuozhuan (hereafter the Zuo), the earliest and most detailed of its commentaries. Under the year 710, the Chunqiu record has: "Winter. The lord [i.e., Lord Huan of Lu, UWj> (r. 711-694)] returned from Tang." The Zuo comments:
"Winter. The lord returned from Tang jf|." It was reported (gao #f) at the temple. Whenever the lord departs, it is reported at the ancestral temple (zongmiao 5j5JH); whenever he returns, he drinks upon arrival, puts the cup down, and his achievements are written on the ce WL tablets: this is ritual jjUfe.9
7 The donor states that his ancestor "joined kings Zhao and Mu in appeasing and rectifying the Four Directions and clipping and attacking Chu Jing." See Li Xueqin, "Meixian Yangjiacun xinchu qingtongqi yanjiu," Wenwu 6 (2003), 66-73. For a detailed discussion of the Yangjiacun discovery, see Lothar von Falkenhausen, "The inscribed bronzes from Yangjiacun: new evidence on social structure and historical consciousness in Late Western Zhou China (ca. 800 BC)," Proceedings of the British Academy 139 (2006), 239-95.
8 For views of the Chunqiu as a purely historical text, see e.g., Timoteus Pokora, "Pre-Han literature," Essays on the sources for Chinese history, ed. Donald D. Leslie (Canberra, 1973), p. 20; Zhao Guangxian, "Chun qiu yu Zuozhuan" Zhongguo shixue mingzhu pingjia, ed., Cang Xiuliang (Jinan, 1990), pp. 26-29. See my earlier study of the Chunqiu, Yuri Pines, "Intellectual change in the Chunqiu period—the reliability of the speeches in the Zuozhuan as sources of Chunqiu intellectual history," Early China 22 (1997), 82-86.
9 Chunqiu Zuozhuan zhu, annotated by Yang Bojun (Beijing, 1981, hereinafter the
Zuo), Huan 2, p. 91. Cf. Xiang 13, p. 998.
chinese history writing
319
The Zuo identification of the Chunqiu with the records made on the ce tablets from the Lu ancestral temple10 helps us to understand several peculiarities of the Chunqiu text as ritual writing.11 For instance, the most important part of the Chunqiu are not records of political events but of seasons: on no less than 63 occasions the text records the season and its first month even when no event is reported for this period of time. In all likelihood, seasonal tablets were pre-arranged in the Lu ancestral temple, while records of events were added later to the pre-existent structure. Needless to say, for a lay reader these "blank" seasonal records are meaningless. Many other records convey information which might have been of significance—if any—only for the ruling family of the state of Lu, but not for an external audience. A reader of the Chunqiu accounts of the eventful year 632 may be puzzled by the entry, "Autumn. Bo Ji >fB{fi from Qi arrived,"12 which breaks the sequence of the records pertaining to the inter-state activities in which the lord of Lu took active part. The arrival of Bo Ji, a sister of the reigning Lord Xi of Lu H'dkfi- (r. 659-627), was a minor event in political terms; but for the ancestral spirits it was perhaps no less important than Lus newly-fostered alliance with the rising power of Jin. This peculiar selection of events diminishes the text's value as a historical chronicle, but serves its ritual purposes.
Ritual considerations obviously dominate the Chunqiu records. For instance, records of foreign events are made exclusively when the event was properly reported to the court of Lu and not in accordance with its overall political significance. Thus, while the Chunqiu routinely informs of deaths and funerals of allied rulers, including heads of tiny neighboring polities such as Zhu 9$ or Cao If, it fails to mention the deaths even of the Zhou kings if royal envoys did not report the matter.13 Similarly, while the Chunqiu usually tells of wars between other states, it fails to mention some fateful military encounters. In the winter
10 The identification of the ce records with the Chunqiu was analyzed by the most famous commentator of the Chunqiu and the Zuozhuan, Du Yu (tfcJS* 222-84 AD). See his comments and additional elaborations by Kong Yingda (574-648 AD) in Chunqiu Zuozhuan zhengyi, annotated by Du Yu and Kong Yingda (rpt. Shisanjing zhushu, compiled by Ruan Yuan [Beijing, 1991], "Xu" J% p. 1701).
11 Many insightful observations with regard to the ritualistic nature of the Chunqiu were made by Artemij M. Karapet'iants, "Chun tsiu i drevnekitajskij istoriograficheskij ritual," Etika i ritual v traditsionnom Kitae, eds L.S. Vasilev et al. (Moscow, 1988), pp. 85-154.
12 Zuo, Xi 28, p. 450.
13 For these omissions, see e.g., Zuo, Xi 7, p. 319; Wen 14, p. 602.
320
yuri pines
of 712 the state of Zheng defeated Song 5^, thereby solidifying its hegemonic status in the Central Plain area. The Chunqiu is silent about this event. The Zuo explains:
Winter, the tenth month. The earl of Zheng Hft{Q leading the Guo army invaded Song 5^. On the day renxu (14th), he greatly defeated the Song army, revenging its invasion of Zheng. Song did not report its [military] decree, hence [the Chunqiu] did not record it. Whenever the overlords issue a decree, if they report it, then it is recorded; otherwise it is not. The same is true about success and failure of military expeditions. Even if the state were annihilated, unless the defeat is reported, or the winners reported their victory, it would not be recorded on bamboo tablets.14
This passage clearly demonstrates that for the compilers of the Chunqiu—the Lu scribes—what truly mattered was not the political importance of the event but the ritually correct manner in which it was reported. Similar omissions occur with regard to domestic events that were not recorded in the Chunqiu if the lord of Lu was not in attendance; alternatively, the record could be modified and certain information omitted.15 Such omissions and modifications are notorious in relation to problematic events in the state of Lu. Like the bronze inscriptions, the annals concealed unpleasant news: thus the assassination of a Lu ruler or heir is invariably reported as the slain lord merely "passing away" (hongl&i), while the heir is said to have "died" (zu 2^).16 Similarly, it was taboo (hui M) for the annals to publish occasions when the lord of Lu was detained or otherwise humiliated by foreign powers.17 Also, when in 517 rebellious ministers expelled Lord Zhao H-BH^fi- (r. 541-510), the Chunqiu laconically recorded: "Ninth month; on [the day] jihai, the lord left for Qi."18 There are many similar instances.19
14 Zuo, Yin 11, p. 78. Indeed, even such an outstanding event as the defeat in 494 of the state of Yue H by its rivals from Wu ^ is not mentioned by the Chunqiu, since "Wu did not report victory, Yue did not report defeat" (Zuo, Ai 1, p. 1607).
15 See, e.g., Zuo Yin 1, p. 18; Huan 14, p. 138; Xi 2, p. 281; Wen 7, p. 562.
16 For the reports on the slain lords, see Zuo, Yin 11, p. 71; Huan 18, p. 151; Min 2, p. 261; for heirs, see Zuo, Zhuang 32, p. 251; and Wen 18, p. 629; in the latter case, the Zuo explains: "The book [Chunqiu] says 'the son died' because of the taboo" (Wen 18, p. 633).
17 See Zuo, Xi 17, p. 373; Wen 2, p. 522; Cheng 10, p. 851; Zhao 16, p. 375.
18 Zuo, Zhao 25, p. 1454; when other dignitaries or foreign rulers went into exile, the Chunqiu reports them as "fleeing" (ben ^).
19 The taboos were extended to the Zhou kings: in 632 when King Xiang Ml j|3£ (r. 651-619) was humiliatingly summoned to the inter-state meeting at Wen by powerful Lord Wen of Jin (r. 636-628), the Chunqiu laconically states: "Heavenly King
chinese history writing
321
The above examples suffice to indicate that the Chunqiu was definitely not composed as a factual record of historical events for the members of the educated elite, who would in any case find its terse language and skewed contents difficult to comprehend without proper commentaries.20 Rather, it was a means of communicating with the ancestors, predicated upon what Joachim Gentz aptly designates "ritual" rather than historical "reality."21 The omissions and skewed reports are only one aspect of a much more general tendency, that is, the general emphasis of the text on ritual precision. The Chunqiu is extraordinarily careful in its choice of words. Thus, it uniformly refers to foreign dignitaries according to their ranks within the Zhou original hierarchy, stubbornly refusing to recognize the ritual "upgrading" of powerful rulers of such states as Chu Jit, Wu Yue it and Qin |j| from their original bo {Q ("earl") or zi ^ ("viscount") to the gong 4a ("duke") and wang zE ("king") rank. Careful use of other terms, names and appellations convey the Chunqiu s judgment of political personalities of the age. Eventually, this formulaic language became the foundation of the subsequent exegetical tradition and the search for the "subtle meaning" of the text.
The adherence of the annalists to ritual reality explains the importance of the annals as a quasi-legal means of judging political personalities. The most celebrated case of such judgment was the condemnation of a Jin leader, Zhao Dun jUJlf, who orchestrated the assassination of his ruler, Lord Ling #®&- (r. 620-607), while pretending to flee the state. The court scribe, Dong Hu SM> nevertheless recorded for the annals: "Zhao Dun murdered his ruler." Zhao protested, but Dong Hu explained that as Zhao neither left the state at the time of the murder
hunted at Heyang" ^E^fMBII (Xi 28, p. 473). For further examples, see Chao Yuefeng, "Chunqiu shuo Ii," Guji zhengli yanjiu xuekan 1 (2000), 8-13, especially 10-11.
20 It is worth noting that from the very beginning of its circulation (presumably in the 5th century), the Chunqiu was accompanied by the commentaries (three to five, according to Ban Gu et al., Hanshu, annotated by Yan Shigu [Beijing, 1997], 30.1715). Interestingly, earlier Classics, written in archaic and less comprehensible language, like the Shijing (Book of songs) and the Shujing HIM (Book of documents), did not acquire systematic commentaries until the early Han. This shows both the demand to properly understand the Chunqiu and the difficulty of doing so. The Han scholar, Huan Tan flOPf (ca. 20 BC-56 AD) exclaimed: "If the [Chunqiu] classic lacked the [Zuo] commentary, the sage would close the door and ponder over it for ten years, and even then he would not understand it!" Cited by Zhu Yizun, Jingyi kao (rpt. Beijing, 1988), 169.875.
21 See Gentz, "The past as a messianic vision: historical thought and strategies of sacralization in the early Gongyang tradition," Historical truth, historical criticism and ideology: Chinese historiography and historical culture from a new comparative perspective, eds Helwig Schmidt-Glintzer, Achim Mittag and Jörn Rüsen (Leiden, 2005), p. 235.
322
yuri pines
nor punished the criminals thereafter, the legal responsibility was his.22 Courageous Dong Hu was praised as a model scribe by Confucius JL^P" (551-479) himself—precisely because he understood that the function of the annals is not to record events as such, but to present a ritually correct judgment of the rulers and their ministers. This judgment could have severe consequences for the culprit. Several Zuo anecdotes attest to the annals' legal importance.23 In 559, a Wei Hj potentate, Ning Zhi (d. 553), together with his accomplice Sun Linfu ££#5£ expelled Lord Xian (r. 576-559 and 546-544) from the state. On his deathbed, Ning Zhi reportedly repented:
Ning Huizi m~M~F of Wei (i.e., Ning Zhi) became sick. He summoned his son, Daozi [Ning Xi HHS, d. 546], saying: "I committed a crime toward the ruler, and although I repented, I was unable to mend it. My name is preserved on the ce -fflj tablets of the overlords, which say: 'Sun Linfu and Ning Zhi expelled their ruler.' When the ruler returns [to the state], you should conceal this [record]. If you are able to conceal it, you are my true son. If you fail, then, insofar as spirits and deities exist, I shall remain hungry and will not receive [your] offerings." Daozi made the promise, and then Huizi died.24
It is unclear whether Ning Zhi was afraid of condemnation on the ce tablets of the overlords because they would tarnish his memory or because of fear of divine retribution: his skepticism with regard to the deities' existence suggests that the former answer is more plausible.25 In any case, afraid of condemnation, Ning Zhi wanted his son to allow the ousted ruler to come back in exchange for a retroactive change of
22 Zuo, Xuan 2, p. 663. There are other cases in which the Chunqiu condemnation of a ruler's murderer hints at a legally responsible person, even if technically he was not the killer. See Zhao Shengqun, Chunqiu jing zhuan yanjiu (Shanghai, 2000), pp. 251-57.
23 In 612, the Song envoy, Hua Ou, had to decline a polite invitation from Lord Wen of Lu (r. 626-609), because a century earlier Hua's ancestor, Du, had participated in the assassination of his ruler and "was named in the bamboo tablets of the overlords" {Zuo, Wen 15, p. 609). Others were less compliant with the scribes' verdict: the Qi potentate, Cui Zhu IMff, had no less than two scribes killed in order to prevent them recording that he had assassinated Lord Zhuang HSBa (r. 553-548), but the scribes' persistence left him no option but to accept this damage to his name {Zuo, Xiang 25, p. 1099).
24 Zuo, Xiang 20, p. 1055.
25 The phrase "if spirits and deities exist," became widespread in the second half of the Chunqiu period, as suggested by the Zuozhuan {Zuo, Xuan 4, p. 680; Xiang 10, p. 977; Xiang 14, p. 1013; Xiang 20, p. 1055; Zhao 27, p. 1487). This conditional clause reflects growing doubts as to the deities' existence, reflected also in the later part of the Zuo (see Pines, Foundations of Confucian thought: intellectual life in the Chunqiu period, 722-453 B.C.E. [Honolulu, 2002], pp. 79-84).
chinese history writing
323
the record. Ning Zhis son managed to get this done, and the current text of the Chunqiu does not mention Ning Zhis role in ousting Lord Xian, stating instead: "The Marquis of Wei fled to Qi."261 shall return to the story of the retroactive manipulation of the records later; here it is important to assess the power of the annals to influence political reality. While not all Chunqiu statesmen were as concerned with their image "on the ce tablets of the overlords" as Ning Zhi, it is highly likely that Ning Zhi s reaction was precisely the one sought by the annals' compilers. Through their staunch preference for ritual reality over historical facts, and through their judgment of political actors, the court scribes and their employers hoped to preserve the deteriorating ritual order intact. Thus, the Chunqiu was not merely a means of communicating with the ancestors, but a creation—or re-creation—of reality as it should be, an alternative to the chaotic events of the real world. Perhaps this is why Confucius chose to publish the Chunqiu, turning thereby the Lu annals into one of the most revered canonical texts in Chinese history.
Profane or sacred? The Chunqiu and its commentaries
Throughout the Chunqiu period, the court annals were not designed to circulate widely; when a foreign dignitary was given the right to review them it was considered a gesture of extraordinary personal favor.27 Thus, when Confucius (or members of his entourage) published the Chunqiu, in effect re-addressing it from the spirits to living contemporaries, they radically altered the function of the text. This may have caused Confucius to doubt whether this action itself was in accord with ritual norms, as is reflected in his putative saying: "I will be understood only because of the Chunqiu; I will be condemned only because of the Chunqiu"28
Through its association with Confucius, the Chunqiu rapidly acquired the position of canonical text, and was identified as such since the
26 Zuo, Xiang 14, p. 1004.
27 See, e.g., Zuo, Zhao 2, pp. 1226-27.
28 Mengzi yizhu, annotated by Yang Bojun (Beijing, 1992), "Teng Wen Gong, xia" 6.9, p. 155. For the controversy over Confucius' authorship of the Chunqiu, see, e.g., Yang Bojun, "Qianyan," Chunqiu Zuozhuan zhu, pp. 7-18; Zhang Yiren, "Kongzi yu Chunqiu de guanxi," Chunqiu shi lunji (Taibei, 1990), pp. 1-59. The extant evidence does not allow a decisive answer as to whether or not Confucius (or his followers) modified the original text of the Lu annals, but it is highly likely that he was the text's publisher.
324
yuri pines
4th century BC.29 Almost immediately, the terse text of the Lu annals acquired commentaries, which have become indispensable to its readers ever since. Two of these commentaries, namely, the Zuo and the Gongyang zhuan, may be representative of the bifurcation of Chinese historiography between the bureaucratic account of events aimed at political education of the elite and the perpetuation of ritual reality at the expense of the facts.
The Zuozhuan was composed between the 5th and 4th centuries BC. It incorporates various historical sources from major Chunqiu states, including materials that appear as auxiliary notes used by the scribes for the compilation of the annals, as well as more detailed narrative histories prepared by the scribes separately from the annals. Elsewhere I have suggested that these "scribal records" (shiji jfelE), being addressed to members of the educated elite rather than to the deities, reflect an alternative historiographic tradition to that of the court annals; and this alternative tradition is duly observable in the Zuozhuan.30
Even a cursory view of the Zuo shows its tremendous difference from the Chunqiu, in terms of structure, content and approach to history. First, far from being a dry ritualistic account, the Zuo presents history complexly, with explicit emphasis on cause-and-effect relations; not incidentally one of the most wide-spread terms in the text is gu tfc ("therefore"), which appears over 600 times. Second, the Zuo is markedly devoid of the formulaic language so characteristic of the Chunqiu. Thus, for example, there is no traceable correlation between the appellations it employs and its author's evaluation of the protagonists; nor does the text employ other hidden formulae to deliver "praise and blame"; this is done in a more straightforward way. Third, in terms of language, the Zuo does not steadfastly observe ritual conventions; hence, it frequently refers to the rulers of Chu, Wu and Yue as "kings"—much to the dismay of later Confucian purists. Fourth, there are no traces of taboos in the Zuo, and as far as I can tell, the text shows no clear political biases. It conceals neither defeats nor misdeeds of domestic and foreign rulers and dignitaries, as it mercilessly reveals their treachery, folly and cruelty.
29 The two earliest references to the Chunqiu are in the Mengzi ("Teng Wen Gong xia" 6.9 p. 155; "Li Lou xia" 8.21, p. 192), and in the "Yu cong fgjg; 1" slips (ca. 300), from Guodian where the Chunqiu is ranked together with other canonical texts, such as the Shijing and Shujing. See Guodian Chumu zhujian (Beijing, 1998), "Yu cong 1," p. 195, slips 40-41.
30 For a detailed discussion of the Zuozhuan sources, see Pines, "Intellectual change" and Foundations, pp. 14-26.
chinese history writing
325
The Zuo simply ignores "the rules of recording" that played such a great role in the compilation of the Chunqiu.
Having said this, I do not intend to claim that the Zuo is devoid of hidden or overt messages, or that it is a sort of Rankean history. Ideologically, the author obviously believes in the moral and ritually correct universe of the Chunqiu; but he delivers his ideological message neither through the Chunqiu-like subtleties nor through a tendentious arrangement of materials as in the later dynastic histories. Rather, ideological goals are served through a variety of interpretative techniques, such as moralizing speeches, long- and short-term predictions, commentaries by the narrator and by Confucius, and the like.31 Yet important as they are, these techniques do not turn the Zuo into a polemical treatise aimed "to validate Ru teachings... through writing them into a narrative of the past."32 Rather, its aim (like the aim of the "scribal records" incorporated into the Zuo) is to supply members of the educated elite with a working knowledge of past events—an indispensable asset in everyday political practice. This goal is yet another difference between the Zuo and the Chunqiu.
The Zuo is self-referential in its insistence on the importance of mastering the past to cope with current challenges. Its protagonists routinely invoke the past in a variety of court or inter-state debates, and their superior knowledge of former events becomes a useful polemical weapon. References to successes and failures of previous rulers and ministers, analyses of historical developments in a rival state or invocations of earlier precedents to justify a policy choice—are recurrent rhetorical strategies in the Zuo. The past is a tool in the statesmen's hands; and to make it more accessible, the author aims at providing the reader with as much useful information about important events in the life of major states and lineages as possible. This emphasis on detailed information is probably the single most important feature of the Zuo. Not a single pre-imperial text can even remotely match it in terms of precision of the historical data involved. Dates and names, office titles and locations, personal and lineage background of the protagonists—all these shape
31 The best discussion of these interpretative techniques is by David C. Schaberg, A patterned past: form and thought in early Chinese historiography (Cambridge and London, 2001). For the comments of the narrator and of Confucius, scattered throughout the Zuo, see Eric Henry, " 'Junzi yue' and 'Zhongni yue' in Zuozhuan" Harvard Journal of Asiatic Studies 59.1 (1999), 125-161.
32 Mark E. Lewis, Writing and authority in early China (Albany, 1999), p. 132.
326
yuri pines
the Zuo narrative. This abundant information is often presented in an
almost raw form, with minimal interpretation, which at times results in narratives that go against the moral messages enunciated in other parts, indicating that ruthless and immoral statesmen can attain political success. Some of the later readers of the Zuo were visibly annoyed by this occasional moral void in a canonical text. Zhu Xi (1130-1200 AD) bitterly complained: "The malady of Mr. Zuo is that he discusses right and wrong in terms of success and failure and does not root [himself] in the correctness of righteousness and principles."33
Zhu Xi and other Confucian purists were unhappy not only with regard to the insufficiently emphasized moral message of the Zuo but also with regard to its treatment of the Chunqiu text. Despite occasional praise of the Chunqius ability to use subtle words in order to "encourage the good and frighten the licentious,"34 the Zuo does not consider this text to be flawless or infallible, and its commentary, with its focus on the historical background of the Chunqiu entries, frequently leads the reader to question the correctness of some of the Chunqiu records. While the Zuo never criticizes the Chunqius adherence to formulaic language and its taboo-related omissions, it shows that some modifications or omissions were made not due to ritual considerations but out of political expediency. Thus, in the above mentioned case, Ning Zhi, who was originally condemned in the "annals of the overlords" as the ouster of his lord managed to get the record altered due to successful political maneuvering. The extant Chunqiu entry as it appears is therefore wrong not only factually but also morally, for it conceals the heinous crime of expelling the legitimate ruler. Elsewhere, the Zuo informs us that a certain record was falsified to appease the state of Chu, the erstwhile ally of the state of Lu, whom Lu betrayed.35 The Zuo also hints that the Chunqiu conceals cases of regicide in other states—apparently either to appease foreign powers or due to its uncritical acceptance of falsified
33 Zhuzi yulei, compiled by Li Jingde; collated by Wang Xingxian (Beijing, 1986) 93.2151. Six centuries later, Liu Fenglu MUSI (1776-1829 AD) condemned the Zuo: "The Chunqiu is not a historical text. Supporters of the Zuo treat the Chunqiu as a historical text, and inevitably lose its meaning" (Liu Fenglu, "Zuoshi chunqiu kaozheng," Guji kaobian congkan, ed., Gu Jiegang [Beijing, 1955], p. 599). For similar views, see Pi Xirui, Jingxue tonglun (Beijing, 1989), pp. 39-45.
34 Zuo, Zhao 31, pp. 1512-13; cf. Cheng 14, p. 870.
35 See Zuo, Xi 28, p. 452 (and the Chunqiu record on p. 448).
chinese history writing
327
reports from foreign courts.36 These instances, one of which is the case of regicide in the state of Zheng discussed below, strip the Chunqiu text of its supposed ritual and moral infallibility and effectively de-sacralize it. The resultant lack of an identifiable source of ultimate authority is yet another feature of "secular" historiography as exemplified by the Zuozhuan.
If the Zuozhuan approach represents a "secular" trend in traditional historiography, the Gongyang zhuan may be the clearest representative of a new quasi-religious approach, the adherents of which identified the Chunqiu itself as a sacred text that can magically influence the world. The Gongyang zhuan, which has been extensively studied by Joachim Gentz, to whom I am indebted for much of the following analysis, was composed in the second half of the Warring States period (Zhanguo, 453-221), approximately a century or more after the Zuo.37 Several important intellectual developments took place during these generations, and they are helpful in understanding the peculiarities of the Gongyang approach. First, the association between Confucius and the Chunqiu had been firmly established by then. Second, Confucius himself had been recast by some of his followers from an ordinary Master into a supreme sage that "nobody like him ever existed since the people were born."38 Third, the massive use and abuse of history by rival thinkers of the Warring States era began backfiring, generating negative reaction against invocations of the past as compelling arguments in political debates.39 These three developments resulted in the creation of the Gongyang zhuan—a text that insists on the sacral power of the writing of history when performed by the Sage, but simultaneously tries to disengage this sagely product from its historical setting.
36 See a list of these cases in Zhao Shengqun, Chunqiu jing zhuan, pp. 238-242; Zhao does not analyze these items in terms of concealment of regicides in foreign courts.
37 For Gentz's major study, see his Das Gongyang zhuan: Auslegung und Kanonisierung der Frühlings- und Herbstannalen (Chunqiu) (Wiesbaden, 2001); see also Gentz's articles cited in this chapter and his chapter in Volume Two.
38 Zi Gong's ^f-jf saying as cited in the Mengzi, "Gongsun Chou, shang" 3.2, p. 63. For the inflated image of sages, who were conceived of as divine figures by some of the late Zhanguo thinkers, see Michael J. Puett, To become a god: cosmology, sacrifice, and self-divinization in early China (Cambridge, Mass., 2002).
39 For the massive abuse of historical narratives and the diminishing appeal of historical argumentation in political debates, see Pines, "Speeches and the question of authenticity in ancient Chinese historical records," Historical truth, pp. 195-224.
328
yuri pines
For the authors of the Gongyang zhuan, the Chunqiu is not just a historical text to be commented upon, but a manifestation of the ultimate wisdom of the Sage. They clarify this point in the final lines of the text:
Why did the superior man [Confucius] make the Chunqiu'? To eradicate generations of disorder and return to the right there is nothing like the Chunqiu. Yet we cannot know anymore whether it was made for this purpose or because superior men liked to speak of the Way of Yao and Shun. And was it not in the end perhaps the delight that he would be recognized [in the same way] as Yao [recognized] Shun? To obtain the [hidden] meaning of the Chunqiu in order to await for later sages—it was also this in which the superior man was delighted when making [the Chunqiu].40
The authors are cautious with regard to their ability to comprehend Confucius's true intent, but they are unequivocal with regard to the potential of the Chunqiu "to eradicate generations of disorder and return to the right." While in the Zuozhuan the Chunqiu is presented as a means "to encourage the good and frighten the bad," a relatively modest goal, in the Gongyang zhuan this text becomes the single way out of the state of disorder, the repository of the ultimate blueprint of a correct world. This assertion sets the tone for the re-sacralization of the Chunqiu, as is reflected throughout the Gongyang zhuan narrative.
The Gongyang authors assume that the Chunqiu comprises two layers: the initial Lu court annals, and the modifications by Confucius. The first layer is a factual skeleton of the text to which Confucius added the "flesh," namely, specific wording through which the events should be properly understood and analyzed. The Gongyang authors attempt to discern the Master's hidden message from behind his "subtle words" fjlSnij". This is done through a particular exegetical method, according to which, first, the pattern of proper recording of the events in the Chunqiu is determined; second, all deviations from this putative pattern are identified; and third, each deviation is explained either as reflecting an abnormal historical situation, or representing Confucius' "praise and blame." At the next step one can establish the pattern of deviations from the regular pattern of recording, and then another set of yet more subtle deviations from that pattern, and the pattern of a new set of deviations, and so on. As Joachim Gentz summarizes, "the
I follow Gentzs translation ("The past," pp. 236-37), with slight modifications.
chinese history writing
329
Gongyang zhuan does not always succeed in offering a coherent exegesis and is time and again forced to invent new rules... and even rules that deviate from deviation rules."41
The Gongyang zhuan s method of exegesis is revealing. It is based on a somewhat cabbalistic assumption that the Chunqiu text is sacrosanct and infallible, not a single word of which is misplaced or miswritten due to authorial oversight, but rather hides the Sage's message. To preserve this assumption in light of obvious flaws in the Chunqiu records, the Gongyang masters perform remarkable intellectual acrobatics. This can be illustrated by a single comparison between their account and that of the Zuozhuan of the following Chunqiu entry from the year 566:
Winter, the twelfth month. [Our] lord assembled with the marquis of Jin, the duke of Song, the marquis of Chen, the marquis of Wei, the earl of Cao, the viscount of Ju, and the viscount of Zhu at Wei 0. Earl Kunwan 9%M of Zheng was going to the assembly. He did not meet the overlords; on the day bingxu he died at [the town of] Cao |;|I42
This record conceals an instance of regicide: Lord Xi of Zheng JtPfi|£-(r. 570-566) did not merely "die" but was assassinated. Why did the Chunqiu conceal this fact? The two commentaries offer entirely different explanations. The Zuo explains:
When Earl Xi -fl^fi- of Zheng was still heir-apparent, in the 16th year of Lord Cheng [of Lu, i.e., 575], he traveled together with Zi Han [a leading Zheng noble] to Jin, and mistreated him. Later he traveled with Zi Feng i21!! [another leading noble] to Chu and again mistreated him. In the first year of his rule, he attended the court of Jin; Zi Feng wanted to complain to Jin and to depose the lord, but Zi Han stopped him. On the eve of the meeting at Wei, Zi Si i^H [another leading Zheng noble] acted as chancellor, and [the lord] again mistreated him. [The lord's] servant remonstrated but was not heeded; when he remonstrated again, the lord killed him. When the lord arrived at Cao, Zi Si dispatched bandits who killed Lord Xi at night; while [Zi Si] informed the overlords that the lord died of high fever.43
The Gongyang zhuan presents an entirely different version:
Why was it written "died at Cao [Cao] is a Zheng town. When an overlord dies within his domain, the location is not recorded: why was the
41 See Gentz, "The ritual meaning," p. 133 (and pp. 128-34 for the detailed discussion of the Gongyang exegesis), and his article in Volume Two.
42 Zuo, Xiang 7, p. 949.
43 Zuo, Xiang 7, p. 953.
330
yuri pines
location recorded here?—To conceal the matter?—What to conceal?—[The lord] was murdered.—Who murdered him?—His nobles murdered him. Why is it not said that his nobles murdered him?—It is a taboo for the Central States.—Why is it a taboo for the Central States?—The earl of Zheng was en route to meet the overlords at Wei; his nobles remonstrated, saying: "It is not good to return to the Central States; it is better to follow Chu." The earl of Zheng said: "Unacceptable." His nobles said: "If you consider the Central States righteous, then what about them invading us during the mourning period? If you consider them powerful, then they are no match for Chu." Then they murdered [the lord].—Why then was the name of the earl of Zheng, Kunwan, recorded? He was wounded and was on the way back; he did not arrive at his lodging and died.—But he did not meet the overlords; why is it written that he went to the assembly?—To fulfill his will.44
Both texts agree about the basic fact: the lord of Zheng was murdered by his underlings en route to the inter-state assembly at Wei; thus the Chunqiu record is obviously misleading. The Zuozhuan explains that this record reflects a deliberately wrong report by the Zheng chancellor, a major culprit. The Lu scribes (and probably scribes of other northern states) went along with this lie in order not to alienate an important ally (and the Zheng leaders duly respected this by reconfirming their alliance with the northerners a few months later). That the lord of Zheng was an intolerable ruler whose cruelty and folly brought about this miserable end should not absolve the murderers of their responsibility: after all the aforementioned Zhao Dun was also justified in his plot against Lord Ling of Jin, but nonetheless the paragon of upright historians, Dong Hu, considered Zhao Dun guilty of regicide. The concealment of a similar event in the state of Zheng is therefore morally and ritually wrong and can be explained only by analyzing political circumstances.
This explanation is unacceptable to the Gongyang (as well as the parallel Guliang zhuan WMW) authors. To justify the false record they invent a fascinating explanation: the Chunqiu concealed the crime as a matter of taboo. While taboo regulations should not apply to the non-Lu rulers, in this case the record was modified out of respect to Lord Xis putative commitment to the cause of the "Central States" against the state of Chu. It is almost needless to state that this explanation is groundless. While Zheng frequently shifted its alliances, there are no indications that Lord Xi was a supporter of the northern states, or that
44 Chunqiu Gongyang zhuan zhushu, annot. by He Xiu and Xu Yan, rpt. Shisanjing zhushu, Xiang 7, 19, p. 2302 (hereafter Gongyang zhuan).
chinese history writing
331
his ministers leaned toward Chu; the fact that Zheng continued its alliance with Jin after the assassination of Lord Xi suggests that this explanation is wrong. More substantially, treating the state of Chu as the Other of the "Central States" is anachronistic: this conceptualization of Chu is a product of the middle Warring States period and does not reflect the conditions when the Chunqiu was compiled.45 The Gongyang authors go to great lengths to preserve the integrity and infallibility of the text upon which they comment.
This example will suffice to show how reverence for the Chunqiu required the authors of the Gongyang zhuan to invent or twist the facts to fit in with preconceived ideas. This reflects a general attitude of the authors toward history. The latter is important insofar as it serves as a foundation for Confucius' putative judgments in the Chunqiu, but in the final account it is subordinate to ritual considerations. What really happened matters very little; the true message of the Chunqiu is what ought to have happened. The Gongyang zhuan creates an imagined state of affairs in which there is a unified world ruled by a powerful Son of Heaven, who is served by the overlords and their nobles, with the latter being the ruler s minor executives and not powerful political actors. This picture, which the authors read into the Chunqiu, is completely at odds with the realities of the eponymous period (722-453), and depicts not the actual but the ideal state of affairs. The real world of Chunqiu events frequently threatens to break out from its interpretative bounds, reflecting the impossibility for the Gongyang authors to accommodate real history within their ritual framework.46 Thus, at times they display great reverence toward the Zhou kings, who are treated as if they were truly powerful sovereigns; at times, however, the text laments that during the Chunqiu period "there was no Son of Heaven above" ii^^f1, suggesting that the king was a nullity. The authors do not bother to
45 For the distinct identity of Chu being primarily a Zhanguo period phenomenon, see Yuri Pines, "Beasts or humans: pre-imperial origins of sino-barbarian dichotomy," Mongols, Turks and others, eds. Reuven Amitai and Michal Biran (Leiden, 2004), pp. 87-88.
46 It is significant I believe, that on a few occasions when the Chunqiu conceals the instance of a regicide in a foreign country, neither the Gongyang nor the Guliang zhuan comment upon this matter at all. See, e.g., Gongyang zhuan, Zhao 1, 22, p. 2317, Ai 10, 27, p. 2349; Chunqiu Guliang zhuan zhushu, annotated by Fan Ning and Yang Shixun, rpt. Shisanjing zhushu, Zhao 1, 17, p. 2433 and Ai 10, 20, p. 2450. Also noteworthy is that none of the major commentators of these texts addressed this concealment, although in the first case the culprit was the notorious usurper and tyrant, Prince Wei 3£-flH of Chu, who ascended the throne as King Ling 5§tlt3£ (r. 540-529).
332
yuri pines
resolve this apparent contradiction: they want to teach the reader the correct political principles bequeathed by Confucius, not to teach about the past.47
Disengaged from its historical setting, the Chunqiu in its Gongyang interpretation became adaptable to any imaginable political demand, as the history of the text under the Han dynasty (206 BC-220 AD) exemplifies. The so-called Chunqiu principles could be applied to domestic and foreign policy, and even to legal cases; the text became a kind of holy scripture. Its sacred status was further emphasized by those Han scholars who asserted that in the Chunqiu Confucius had predicted the Han ascendancy and preordained the political institutions to be used by the Han. Although this radical religious interpretation of the Chunqiu is not part of the original Gongyang zhuan design, it appears as a logical outcome both of its method of exegesis, which allowed reading into the Chunqiu almost any possible meaning, and of the elevation of the Lu text to supernatural realms.48
The Gongyang zhuan is certainly the most religious of the post-C/zwn-qiu historical texts, as well as the least "historical." Its a-historical view of Confucius and of his putative legacy places the Gongyang zhuan on a par with ritual texts such as the "Quli" tijjj!t or the "Wangzhi" zE0J chapters of the Liji jjjjljfE (Book of rites), or even with the Zhou li JUijH (Rites of Zhou), rather than with historical texts such as the Zuozhuan or the Guoyu Mlm (Discourses of the states). The Gongyang zhuan marks therefore both the apex and the dead end of religious historiography. Its advent in the early Han might therefore have contributed decisively toward what Li Wai-yee identifies as the "anti-historical tendency in Han thought."49 This a-historicism is what eventually diminished the importance of the Gongyang zhuan in the later historiographical tradition, as it was overshadowed by the rival commentary, the Zuozhuan. This does not mean however that history writing lost its spiritual elements. These were reintroduced at the beginning of the Han by the greatest early Chinese historian, Sima Qian.
47 For a brief but insightful analysis of the Gongyang zhuan s a-historical approach, see Lewis, Writing and authority, pp. 139-44.
48 For multiple applications of the Chunqiu under the Han, see Zhao Boxiong, Chunqiu xue shi (Ji'nan, 2004), pp. 102-27 and pp. 202-11. See also Joachim Gentz's chapter in Volume Two.
49 Li Wai-yee, "The idea of authority in the Shiji (Records of the historian)" Harvard Journal of Asiatic Studies 54.2 (1994), 353-58.
chinese history writing
333
Memory and immortality: the Shiji contribution
Insofar as I have identified correctly the "religious-secular" divide of ancient Chinese historiography, it is interesting to try to locate the Shiji, the fountainhead of Chinese official histories, within the sides of this divide. Recently, Michael Nylan has made several interesting observations with regard to possible religious motives in the Shiji. According to her interpretation, by compiling his universal history, Sima Qian tried to act out his filial obligation to his father, Tan §£, and to attain a kind of personal immortality in addition to addressing such a religiously significant topic as relations between Heaven and Man.50 My focus will be different from Nylans, though. In what follows, I shall address the relation of the Shiji to the Chunqiu traditions outlined above, and then explore possible religious aspects of the genre of biographies invented by Sima Qian.
The relation of the Shiji to the Chunqiu has been studied in several excellent works, and I shall therefore limit myself to a brief summary. In the autobiographic last chapter of the treatise, Sima Qian draws an explicit parallel between himself and Confucius, and, despite his polite demurral, it is clear that he considered his magnum opus as compatible with the Chunqiu.51 This desire to emulate the Sage did not mean, however, that Sima Qian had to adhere to the terse style of the Lu chronicle. First, by the early Han period, the term Chunqiu pertained not only to the original text attributed to Confucius but also to its commentaries and sub-commentaries, and, more broadly, to an entire list of historical works, numbering "myriad words."52 Second, living in the age of the unified empire, Sima Qian realized the need to modify the format of his work in order "to apply the model of the Chunqiu, the chronicle of a single state, to the entire known world."53 What mattered to Sima Qian was, then, not the style of the Chunqiu but the idea of a text that
50 See Nylan, "Sima Qian: a true historian?" Early China 23-24 (1998-99), 203-46.
51 See detailed analysis in Li Wai-yee, "The idea of authority," 358-61; Stephen W. Durrant, The cloudy mirror: tension and conflict in the writings of Sima Qian (Albany, 1995), pp. 9-12.
52 Sima Qian mentions that the text of the Chunqiu comprises "dozens of thousands" of words (Shiji 130, p. 3297)—a number that clearly refers to the Classic and the commentaries if not to the entire collection of historical texts. In the Shiji, the Chunqiu genre appears as particularly inclusive, as Sima Qian occasionally includes even those texts that we would define today as "philosophical," such as writings by Xunzi and Han Feizi Wt=f {Shiji 14, p. 510).
53 See Lewis, Writing and authority, p. 310.
334
yuri pines
is both comprehensive in its scope and authoritative in its judgments. As he explicitly stated, he hoped that his work would "exhaust the interchanges between Heaven and man, penetrate comprehensively the changes from ancient times to the present, and thus complete the words of a single school/family."54
In light of this conscious attempt to emulate the Chunqiu as a world-ordering text, it is remarkable that Sima Qian decisively opted to create a historical text in the Zuozhuan style and not a repository of "subtle words" in the manner of the Chunqiu in its Gongyang interpretation. The Shiji is indebted to the Zuo both as one of its most important sources, and as a model of informative historiography. Largely rejecting the a-temporal approach of the Gongyang tradition, Sima Qian depicts history in its complexity, avoiding a simplistic "praise and blame" mode and allowing his sources and protagonists to speak for themselves. His interpretative techniques, such as the protagonists' speeches and the narrator's remarks, clearly resemble those of the Zuo. And yet, the Shiji crucially departs from the Zuozhuan in two important aspects. First, unlike the Zuo, which remained limited temporally and geographically to the Chunqiu framework, the Shiji attempts to attain true comprehensiveness in time and space. Second, Sima Qian abandoned the rigid chronological form of the Zuozhuan, creating a hierarchically structured text, the form of which conveys a sense of order even more than the content does. The second feature in particular places Sima Qian's work on a par with the world-ordering function of the Chunqiu.
Sima Qian's arrangement of the Shiji materials discloses his attempt not just to depict the world but also to regulate it. Of the five sections into which he divided the Shiji, three, namely, "Chronicles" (ji IE), "Hereditary houses" (shijia titHO and "Arrayed biographies" (liezhuan are arranged in a clearly hierarchic pattern. The twelve "Chronicles" (the number of which might have been patterned after the twelve lords of Lu surveyed in the Chunqiu) deal with the rulers who supposedly governed All under Heaven; the "Hereditary houses" deal with the high nobility; while the "Arrayed biographies" focus on particularly noteworthy individuals of lesser rank.55 This structure, emulated henceforth in
54 From Sima Qian's letter to Ren An        Han shu 62, p. 2735.
55 For detailed analyses of the structure of the Shiji, see Grant Hardy, Worlds of bronze and bamboo: Sima Qian's conquest of history (New York, 1999), pp. 27-60; Lewis, Writing and authority, pp. 309-13.
chinese history writing
335
the dynastic histories, allowed the historian to buttress the hierarchic order of the universe, placing every protagonist in a bracket that befitted his or her social position. What is remarkable, however, is the degree of looseness in Sima Qians arrangement of narratives. Much to the dismay of later ideologues of dynastic legitimacy, he placed accounts of the pre-dynastic state of Qin and of the Han competitor, Xiang Yu (d. 202) in the "imperial" section, while the story of the peasant rebel, Chen She Wffi (d. 208), appears among the "Hereditary houses."56 This obvious violation of the ritually sanctioned hierarchical norms discloses the major difference between the Shiji and the Chunqiu. While in the latter, ritual order mattered much more than historical setting, for Sima Qian the actual power of certain individuals or groups was compelling enough to make him deviate from the norms of social hierarchy. In the final account, the "secular" historian in Sima Qian overwhelms the "religious" one.
If in terms of its relation to the Chunqiu lore, the Shiji belongs squarely to the "secular" tradition of Chinese historiography, the text nonetheless does not lack religious dimensions. Aside from those outlined by Michael Nylan, I would like to focus on the genre of biographies, which is one of Sima Qians important innovations. Prior to the Shiji there is no evidence for the existence of a biographic genre at all. Although its seeds can be traced to the inscription of the donors achievements or pedigree on bronze vessels, or to the collection of anecdotes about and sayings of important historical personalities that circulated in the Warring States period, prior to Sima Qian no attempt was made to systematically present biographies. Sima Qians invention of this genre—if he really did it—became one of his lasting contributions to Chinese historiography.57
Sima Qians decision to commemorate outstanding individuals may reflect his personal experience, which made him particularly sensitive to the fate of a personality in history; it may also reflect his general
56 For further examples of Sima Qians violations of the ritually correct hierarchy, see Zheng Zhihong, Shiji zhi wenxian yanjiu (Chengdu, 1997), pp. 202-08.
57 Li Wai-yee raises the possibility that the biographic genre first appeared in the now lost late Warring States compilation, Shi ben fji$; see "The idea of authority," pp. 378-79, n. 55. Cf. Denis Twitchett, "Chinese biographical writings," Historians of China and Japan, eds. William G. Beasley and Edwin G. Pulleyblank (London, 1962), p. 96. For the novelty of the biographic genre, see also Nylan, "Sima Qian," 214-15, especially n. 40.
336
yuri pines
emphasis on the individual as a prime mover in historical events.58 It would be incongruous, then, to argue that he created biographies for explicitly religious purposes; but from the inception, the genre had a religious dimension. This dimension is explicit in the first of the "Arrayed biographies" in the Shiji, where Sima Qian briefly narrates the story of the two legendary righteous hermits, Boyi {fi^ and Shuqi /jfiflOt, and then begins an unusually long dialogue with the readers, in which he reflects upon the historians tasks and upon the importance of commemoration of outstanding personalities.59
Boyi and Shuqi, two morally impeccable persons who died of starvation, enable Sima Qian to question more generally Heavens justice. By providing examples of righteous men, like Confuciuss disciple Yan Yuan l|t#l, who died prematurely, and of arch-villains, like Robber Zhi iSSK, who enjoyed a good life, Sima Qian asks in despair: "So, what is called 'the Way of Heaven ^M.'- is it right or is it wrong?"60 Questioning Heavens justice and lamenting ones fate was a common topos in Warring States discourse, the prevalent answer being that the superior man will cultivate his virtue whatever the external circumstances.61 Sima Qian mentions this solution, proposed by Confucius, but then provides an additional way of coping with injustice:
[Confucius said:] "The superior man detests that after he passes away his name will not be mentioned." Master Jia [Jia Yi jftl> ca. 200-168] said: "The covetous man seeks wealth; the zealous shi dr seeks name; one who boasts dies out of expediency; the masses cling to life." [The Yijing HH? says:] "Those of identical light illuminate each other; those of identical
58 For Sima Qians emphasis on personality in history, see, e.g., Li Wai-yee, "The
idea of authority"; for his focus on the individual as a prime mover in history, see Vitalij Rubin, "Kak Syma TSian' izobrazhal period Chuntsiu," Národy Azii i Afriki 2 (1966), 66-76.
59 For a brilliant discussion of this chapter, see Durrant, The cloudy mirror, pp. 20-27; see also Shan Shaojie, " 'Boyi liezhuan' zhong de gongzheng linian he yongheng linian," Zhongguo renmin daxue xuebao 4 (2005), 129-37.
60 Shiji 61, p. 2125.
61 This issue is discussed in numerous texts, the paradigmatic one being the recently unearthed Qiong da yi shi HiUíífSf ("Failure and success depend on the times") from the Guodian site. For details, see Li Rui, "Guodian Chu jian Qiong da yi shi zai kao," Xin chutu wenxian yu gudai wenming yanjiu, eds Xie Weiyang and Zhu Yuanqing (Shanghai, 2004), pp. 268-78. A recently published short text Guishen zhi ming
from the Shanghai Museum collection, which may have been written by Mozi's (ca. 460-390) followers, explores the issue of the good fate of villains and the bad fate of sages to question the numinosity of the deities. See "Guishen zhi ming," annotated by Cao Jinyan, Shanghai bowuguan cang Zhanguo Chu zhushu, ed., Ma Chengyuan (Shanghai, 2006), vol. 5, pp. 307-20.
chinese history writing
337
kind seek each other." [It also says:] "Clouds follow the dragon; wind follows the tiger; the sage rises and myriad creatures eye him." Although Boyi and Shuqi were worthy, it was because they attained the Master [Confucius] that they became more illustrious. Although Yan Yuan fg$}[ studied industriously, it was because he attached himself to the tail of thoroughbred [Confucius] that his conduct became more renowned. There are shi of rocky caves who approach and reject [a position] in a timely [manner]; how sad it is that their name will perish like smoke and not be mentioned! The men from village gates and lanes who want to behave steadfastly to establish their names—how will they carry it to the later generations unless attached to the lofty shi?62
The detailed analysis of this extraordinarily rich passage would lead me too far afield; what is important for our discussion is Sima Qian's assertion that establishing ones name may serve as a sort of compensation for Heavens injustice. In Durrant's words, "the historian thereby becomes the savior, those attached to him are saved, living on through the power of his writing brush."63 Indeed, by preserving one's name for posterity, the historian corrects Heavens wrongdoing, providing a sort of immortality for those who failed to fulfill their aspirations in life. An after-life in a historical text becomes a compensation for under-appreciation or failure in life.
Stressing the importance of leaving one's name to posterity as a means to post-mortem justice if not "commemorative immortality" is not Sima Qian's invention. Its roots are traceable to the following Zuozhuan dialogue:
Twenty-fourth year [of Lord Xiang, 549], spring; Mu Shu MM. [Shusun Bao fflMWl] arrived at Jin. Fan Xuanzi ffisrf [Shi Gai dr1^] greeted him [at the capital's outskirts]. [Fan Xuanzi] asked him: "The men of old had a saying: 'dead and not rotten: what did they mean?" Mu Shu did not answer. Xuanzi said: "My, Gai's, ancestors at the time of Yu [Shun] and earlier were the Taotang lineage; during the Xia they became the Yulong lineage, during the Shang the Shiwei lineage, during the Zhou the Tangdu lineage, and, when Jin began presiding over the Xia ["Chinese" states] alliance, we became the Fan lineage. It is said about this!"
Mu Shu said: "According to what I, Bao, heard, these are the so-called hereditary emoluments and not 'being not rotten.' Lu had a former noble Zang Wenzhong %$J>Ci$; he is dead already, but his words are still established: it is told about that! I, Bao, heard: 'The best is to establish virtue; second to it is to establish merits; next is to establish words' If even after
Shiji 61, p. 2127.
Durrant, The cloudy mirror, p. 25.
338
yuri pines
a lengthy period they do not fade—this means being not rotten. Preserving ones clan and establishing a lineage to guard the ancestral temple so that sacrifices do not fade for generations—this exists in every state. It is named 'great emolument,' but it cannot be named 'not being rotten.' "64
This fascinating exchange, recorded also in the Guoyu, epitomizes a change of mentality from the lineage-oriented to an individual-oriented notion of continuity and immortality. Shi Gais views reflected lineage consciousness, according to which the survival of the lineage and perpetuity of ancestral sacrifices were the true measure of one's success.65 Shusun Bao, a leading noble himself, did not reject this idea, but outlined a higher degree of personal success: to attain personal fame either through political achievements or at least through establishing one's "words" to be transmitted to posterity.66 One's posthumous fame was one's true immortality.
The idea of preserving one's name for posterity as a true measure of personal success gained further popularity in the Warring States period and thereafter, as suggested among others by the sayings of Confucius and Jia Yi, cited in the Shiji passage above. It resonated well both with the increasingly popular quest for name/fame (ming ^) that became one of the prime movers of elite behavior and with the religious idea that through remembering the deceased one could "enliven" them. This latter topic is presented in the "Ji yi" HH ("Meaning of the sacrifice") chapter of the Liji, which was perhaps compiled in the early Han period, but which may well reflect earlier religious views:
64 Zuo, Xiang 24, pp. 1087-88.
65 This "lineage immortality" is traceable in turn already to the "auspicious words" (gu ci MM) of the bronze inscriptions in which the donors asked the ancestors to allow future generations to "forever treasure [this vessel]" iS^jfcif. Significantly, the requests of the "auspicious words" are directed either to lineage or to personal well-being, but never to remembrance by posterity. See more in Xu Zhongshu, "Jinwen guci shili," Xu Zhongshu lishi lunwen xuanji (Beijing, 1998), pp. 502-64.
66 The difference between establishing "virtue" (de Hf) and "merit" (gong ~Sfo), may be in the degree of political achievement; establishing de could pertain in certain early Zhou contexts to establishing a new dynasty or at least a new regional polity (for these early usages of de, see Kominami Ichiro, "Tenmei to toku," Toho gakuhd 64 [1992], 1-59). Merit could pertain to a more modest achievement, such as those for which meritorious ancestors of the noble lineages received their ranks. As for "establishing words," this may well refer to the tradition of preserving ideologically important speeches of leading statesmen. For the preservation of such speeches of Zang Wenzhong, see Guoyu jijie, comp. by Xu Yuangao (Beijing, 2002), "Lu yu 1" 4.9, p. 170.
chinese history writing
339
Thus, the filiality [taught by] the former kings [requires] that the [son's] eyes should not forget [the parents'] looks, his ears should not forget their voices, his heart should not forget their hearts, will, inclinations, and desires. When he delivers them his love, they seem to live again; when he delivers his sincerity, they reappear. When those who reappear and live again are not forgotten, how can [the son] be irreverent?67
I whole-heartedly concur with Michael Nylan that this idea of the transformative power of mortuary rituals could have been a source of inspiration for Sima Qian.68 Indeed, the notion of "commemorative immortality" appears in several texts. The Laozis view, that "to die and not be forgotten means longevity" ^EM-^F^^S^69 reflects well this quest for commemoration in the world of the Warring States. Some thinkers hoped to do so through the written media. Several Zhanguo texts speak of the importance of one's merits being recorded—preferably inscribed or incised on bronze or on stone to become truly indestructible—to perpetuate one's glory.70 All these ideas might have contributed toward Sima Qian's idea to immortalize the heroes of the past through preserving their individual deeds and statements in his magnum opus.
The idea of "commemorative immortality" might have had even deeper religious importance in Sima Qian's times. It could have been particularly appealing to educated persons, in light of the notorious lack of clarity with regard to the afterlife. In the course of the Warring States and early Han period, the previously dominant idea of the ancestors dwelling above and communicating with their descendants during sacrificial rites gave place to a variety of competing approaches. The coexisting ideas of a bureaucratically organized netherworld, of the tomb as a dwelling for the soul and of the "paradise" of the Spirit
67 Liji jijie, compiled by Sun Xidan (Beijing, 1995), "Ji yi" 24, p. 1209.
68 Nylan, "Sima Qian," 220-222.
69 Boshu Laozi jiaozhu, compiled and annotated by Gao Ming (Beijing, 1996), 33, pp. 403-05.1 follow both Mawangdui versions; Gao Ming suggests following Wang Bis 1£M recension, to emend "forget" (wang to "die" (wang £), because in his eyes the idea of commemoration does not resonate well with the Laozis thought.
70 For the idea that one's deeds should be inscribed or incised to preserve the name for posterity, see e.g., Xunzi jijie, compiled by Wang Xianqian (Beijing, 1992), "Li lun" 19, p. 367; Lushi chunqiu jiaoshi, compiled and annotated by Chen Qiyou (Shanghai, 1990), "Shen shi" 17.6, p. 1108; Liji, "Ji tong" 47, p. 1250. The earliest extant example of such commemoration is a late 4th-century BC stele of a Zhongshan cf [if official, who recorded his appointment "to announce it respectfully to the later exalted worthies" (cited from Martin Kern, The stele inscriptions of Ch'in Shih-huang: text and ritual in early Chinese imperial representation [New Haven, 2000], p. 56).
340
yuri pines
Mother of the West (Xiwangmu HrES) might have been insufficiently compelling to satisfy the elites' need of positive retribution in the afterlife.71 It is thus that the historian moved in to propose a way toward lasting endurance.
It is impossible to ascertain the full range of Sima Qians motives when he undertook the compilation of the "Arrayed biographies"; nor will we ever be able to clarify entirely the role of religious considerations in his project, although religious implications of the biographic genre are undeniable. To paraphrase Grant Hardy, Sima Qians magnum opus became eventually as imperishable as bronze and stone, which were singled out as proper media for commemoration by the Warring States thinkers. By the 1st century AD, with the increasing circulation of the Shiji, a new commemorative genre ensued, that of stone steles erected on the tombs of the elite. The similarity in the structure of the epitaphs and the Shiji biographies may not be incidental.72 It is not impossible, therefore, that while the Chunqiu reflected the ancestral cult of its age, the Shiji eventually contributed toward the development of a new strand in the ancestral cult of future generations. If this observation is correct, it adds yet another dimension to the discussion of the complex relations between religion and historiography during the formative age of Chinas intellectual tradition.
71 For changing views on the afterlife in the Warring States period, see Lothar von Falkenhausen, Chinese society in the age of Confucius (1050-250 BC): the archeological evidence (Los Angeles, 2006), pp. 293-325. For co-existence of divergent approaches to the afterlife in the early Han period, see Wu Hung, "Art in ritual context: rethinking Mawangdui," Early China 17 (1992), 111-44.
72 This similarity is noticed by Twitchett, "Chinese biographical writing," pp. 96-7; see also K.E. Brashier's chapter in Volume Two.