THE SCRIBAL WITNESS: NARRATIVE AUTHORITY IN ANCIENT CHINESE LITERATURE

Abstract

The scribe has been granted a special role in the creation of ancient Chinese narrative prose. Many texts seem to imply the presence of his person or written records, and scholars have often treated this feature as an indication of authorship. In this paper, I argue that another way of relating the scribe to ancient Chinese narrative prose is to see in him not an author but a witness of the events told. I will use several examples to demonstrate that the figure of the scribe stands out by its function of authenticating the narratives in which scribes takes part. Moreover, occasionally scribes appear to have been added to pre-existing “scribe-less” narratives. I will conclude my discussion by detailing how these findings shed light on the composition of individual pieces of literature and the nature of ancient Chinese narrative writing in general.

提要

史官在中國古代叙事散文作品中扮演著比較特殊的角色。許多文獻中經常明示或默示著史官本人或其書面記錄的存在，此一特點也往往被許多學者視為史官是文獻作者身份的標誌。然而，在本文中我將對史官與中國古代敘事散文的關係提出一個新的視角——我認為，史官未必文獻的作者，而是事件中的證人。文中我將舉數例為證，去說明史官這個角色在古代敘事文獻中的出場，是有著使其成為所參與事件的證人的特殊作用。此外，在某些例子中我們可以發現，史官似乎也被添加到了先前“無史”的敘事文獻中去增添可信度。最後，我將討論這些發現將如何有助於我們去理解個別敘事散文作品，以及整體中國古代敘事文學的形成與性質。

In this article, I argue that in early Chinese narrative literature of the first millennium b.c.e., scribes (shi 史 and zuo ce 作冊)¹ can function as witnesses to the events told. These scribes, whom I will refer to as “scribal witnesses,” figure in narratives not because the storyline requires their presence, but because their witnessing of certain events increases the authority (i.e. the perceived credibility and reliability) of the narrative. In this aspect, scribal witnesses seem comparable to the muses and other types of eyewitnesses in ancient Greek,² the ṛṣis, “seers,” in ancient Sanskrit,³ and (last but not least) scribes in ancient Mesopotamian literature,⁴ who have been argued to serve similar narratological functions.

Much has been written on the impact of scribes on early Chinese narrative literature. Scholars have often deemed them and their records responsible for the creation of early historiographical works such as the Chunqiu Zuo zhuan 春秋左傳.⁵ However, while there is no record of scribes composing narrative literature until Sima Qian 司馬遷 (c. 145–90 b.c.e.),⁶ there are plenty of passages in which scribes perform notary and other witness-related duties. The possibility that scribes might also fulfil a derived extradiegetic-rhetorical purpose by authenticating the narratives and texts they appear in has not yet been explored.

In what follows, I will attempt to substantiate my hypothesis with the help of several instances of scribal witnesses found in received and excavated early Chinese texts. I will start with examples in Western Zhou bronze inscriptions and subsequently move on to chapters of the Shang shu 尚書, the Guo yu's 國語 account of the Li Ji 驪姬 unrest, and the *Dan 丹⁷ manuscript excavated at Fangmatan 放馬灘, and end with a passage of the Chunqiu Zuo zhuan. This order generally follows the approximate historical age of the texts cited, but does not imply any definite judgement. It rather serves to arrange different narratological modes to utilize scribal witnesses in a clear and easily understood manner.

In some of these examples, the scribal witness and the relevant narrative may be based on historical facts. In others, however, this is improbable, if not impossible. Moreover, in a few cases, the scribe seems to have been inserted into an already-existing narrative. At this point at the latest, it transpires that attention to scribal witnesses bears the potential to shed light on the sources of early Chinese narrative literature and the general process of how narratives were written. At the same time, they also reflect epistemic considerations and concerns of those who shaped the respective texts.

That said, none of the following implies that scribal witnesses represented the only path to narrative authority in early China. As scholars have argued, accounts also may have become believable because they were attributed to certain sages,⁸ because they were transmitted in writing,⁹ or because they resonated with their readers’ moral and religious beliefs and described what “should have been true.”¹⁰ I do not wish to challenge these hypotheses but argue that they may not capture all that ancient Chinese texts offer.

Western Zhou Bronze Inscriptions

Texts written and compiled during the Warring States and the Han period tend to portray the scribe as the “perfect witness.”¹¹ We are told that scribes recorded all actions and words of their rulers,¹² that they only wrote down what they knew,¹³ and that they risked their lives to keep a record of the truth.¹⁴ We further read that scribes interpreted celestial omens, performed divination, and submitted accurate reports to the ancestral spirits.¹⁵ Like the muses and the ṛṣis, they were believed to have access to divine consciousness. The often-cited explanation of the composition of the Chunqiu Zuo zhuan recorded in the “Yi wen zhi” 藝文志 of the Han shu 漢書 finally puts a nametag on this role of the scribe. Citing Lunyu 論語 3:9, it has Confucius refer to scribal records as zheng 徵, “signs, witnesses” to his oral teachings from which the Chunqiu Zuo zhuan allegedly emerged.¹⁶

Given the uncertain dating of these texts, it is challenging to say when the scribe began to enjoy this reputation of being an exceptionally reliable and knowledgeable witness. However, already in Western Zhou bronze inscriptions, scribes figure as important legal witnesses in royal appointment ceremonies. By order of the king, they “issue” or “enact” (ce 冊) royal appointment mandates to lower nobles by announcing them orally. As the Song ding 頌鼎 (JC: 2829¹⁷) inscription illustrates, the scribe's actions turn a “mandate document” (ming shu 命書) into a legally binding “mandate act” (ming ce 命冊).¹⁸ This procedure seems to have been the privilege of the scribe. Except for the king, only scribes “enact” mandates.

The mention of the scribe and other participants of the appointment ceremony (such as the person “to the right” [you 右)] of the appointee) highlights witnesses of the event.¹⁹ Through his recitation, the scribe notarizes the written mandate. The presence of the scribe may have been necessary because he was one of the few who could read the mandate.²⁰ But this does not explain why he and the person to the right are mentioned in the inscription texts. After all, appointments were communicative acts between the king and the appointees, and the mention of the scribe in the inscription text was not essential to legitimize the mandate. However, he could confirm that the mandate orally announced was identical to the one in writing. As far as we can tell, all other attendees (including the king) only heard but did not read the text.

This role of the scribe is also evident in a few inscriptions that document legal affairs between non-royal nobility.²¹ The most eye-catching case is the inscription of the San shi pan 散氏盤 (JC: 10176).²² The inscription reproduces a contract between the houses of Ze 夨and San 散, in which Ze promises to transfer a part of its territory to San to compensate for an illegitimate attack. The course of the new border is described in geographic detail. Next, eyewitnesses of the demarcation procedure are listed by name. Then follows a dated record of the contract's official settlement, describing an oath-taking ceremony in which named representatives of Ze comply with the new agreement and penalties for violations are stipulated. We further read that a new map had been drawn in the presence of the king of Ze. Finally, the entire document ends with what Li Feng has referred to as a “signature line,”²³ an indented column that states in spatial separation from the rest of the inscription text: “his left-handed contract: chief scribe Zhong Nong” (氒左執要：史正中農。).

This line has led to different interpretations, but they all agree that a scribe named Zhong Nong vouches for the authenticity of the inscription text or, more precisely, the underlying contract.²⁴ Naming the eyewitnesses of the demarcation, the participants of the oath-taking ceremony, and the new map was not enough. A scribe had to certify that what we read corresponds to the “left” contract of debtor Ze.²⁵ We do not know Zhong Nong's precise relationship to the inscription text. He may have been involved in its composition or only counter-read the original contract. Even his affiliation is uncertain. It indeed seems most reasonable to link him to Ze, the debtor.²⁶ But San also remains possible or even a third, neutral party.²⁷ Regardless, as it is not made explicit, we have to conclude that the readers of the contract were either expected to know how to find him or that his reference alone was sufficient to authorize the text. If we follow Li Feng in assuming that the inscription served to eternalize the settlement of the border conflict for future generations, the latter scenario is most plausible.²⁸

The formulaic syntax, the indented position, and the unresolved identity of Zhong Nong indicate that the signature line is not an isolated phenomenon but a conventional feature of Western Zhou scribal repertoire.²⁹ Many more administrative documents were probably authorized with such references and laid the methodological foundations for the use of scribal witnesses in literary texts. That is not to say that the ancient Chinese distinguished sharply between what we nowadays call “literature” and administrative-legal writings reflected by the San shi pan. As is known, the transition between the two tends to be fluid. However, some texts are noticeably driven much more by aesthetic ambitions than others. This phenomenon is already evident in Western Zhou bronze inscriptions, as they rarely follow the model of the San shi pan, which reproduces the administrative documents completely, but instead contain summaries and collages thereof.

This has an impact on the scribal witness. Compare, for instance, the inscription of the Shi qi ding 師旂鼎 (JC: 2809):

唯三月丁卯，師旂眾僕不從王征于方雷。使氒友引以告于伯懋父，在艿。伯懋父廼罰得、顯、古三百鋝。今弗克氒罰。懋父令曰：“義殺，且氒不從氒右征。今毋殺，其又納于師旂。” 引以告中史，書。旂對氒䝳于尊彝。

It was the third month, day ding mao (four). The charioteers of Master Qi's retinue did not follow the king on his campaign to Fanglei. He sent his associate Yin to appeal to Elder Maofu at Nai. Thereupon, Elder Maofu fined De, Xian, and Gu three hundred lü. Until today they were unable to pay their fine. Maofu made an order saying: “By rights, they deserve to be killed and moreover they have not followed their superior on the campaign. But for now they shall not be killed. They shall rejoin Master Qi.” Yin reported this to the internal scribe, who wrote it down. Qi responded to his ruling by this esteemed vessel.

The inscription has some tricky passages,³⁰ but its purpose was undoubtedly to document Maofu's change of verdict. Despite their refusal to accompany the King on his campaign and their incapability to pay the subsequent fine, the three charioteers were spared their lives and allowed to return to Qi's retinue.³¹ We do not know what changed Maofu's mind. Maybe it was simply the realization that having them killed would not help the campaign either. Maybe Maofu did not want to deprive Master Qi of his servants, who probably had only appealed to his superior to escape punishment himself. Regardless, Maofu's ruling must have been quite significant to Master Qi (not to mention his charioteers) and was thus cast into bronze.

No less critical seems to have been that an “internal scribe” made a record of Maofu's ruling.³² Among all other witnesses, he is named last. This is not a matter of course. The scribe was informed after the lawsuit's settlement and his actions do not contribute to the understanding of the main event.³³ As in the case of Zhong Nong's signature line, we could remove him and the text would remain perfectly intelligible. However, we would then lose the insight that there existed a record of the verdict that confirmed the gist of this inscription. Thus, the message of this reference is essentially the same: it really happened, a scribe has a record of it. But unlike on the San shi pan, the scribal witness has become an organic part of the inscription text.

Scribal witnesses in later texts of the Eastern Zhou period and the early Chinese empire roughly fall into these two categories. Some follow the San shi pan and occur in structural separation to the main account. Others more closely resemble the inscription of the Shi qi ding by integrating the scribe into the text proper. In the following, I will first discuss examples of the last two categories and then return to the first group and their most significant representative, the Chunqiu Zuo zhuan.

The “Jin teng” 金騰 chapter

References to scribal witnesses similar to that of the Shi qi ding are found in the inscriptions of the Western Zhou vessels Ying yi 𠑇匜 (JC: 10285), Peng Sheng gui 倗生簋 (JC: 4262), and Wu Hu ding 吳虎鼎.³⁴ Like the Shi qi ding, the Ying yi inscription records a sentence reduction, which is said to have been reported to the scribes Xiong 𬆁 and Hu 曶. The Peng Sheng gui inscription describes a barter transaction between Elder Ge 格伯 and Peng Sheng, whose settlement was recorded by a scribe named name Zhi Wu 戠武.³⁵ The Wu Hu ding inscription portrays the allocation of land to Wu Hu by King Xuan 宣王 (r. 827–782), which is said to have been recorded by the scribes of You Shou 友守 and Fu Bin 甶賓. In all three inscriptions, the scribes are mentioned towards the end of the text and their role is peripheral to the described events.

Some corresponding examples are also found in the Shang shu 尚書.³⁶ Thus, the “Luo gao” 洛誥 chapter consists of an instruction of King Cheng 成王 by the Duke of Zhou 周公 on how to rule the lands of Luo, which causes Cheng to leave Luo to the Duke and his descendants. The chapter ends with a dated note that Cheng ordered the famous “Document Maker” (or “Enactor”) Yi (zuo ce Yi 作冊逸) to “announce” (zhu 祝) the written record of his decision. Moreover, in the “Li zheng” 立政 chapter, the Duke of Zhou suddenly and without further elaboration addresses an unnamed Grand Scribe (tai shi 太史) in the last line of the text, revealing that the Grand Scribe (and presumably also his staff) were among the audience of the Duke's announcement. Finally, a slightly different and intriguing reference occurs in the “Jin teng” 金縢 chapter. Given the presumably late dating of the text, it serves as an example of how Eastern Zhou historical discourse reshaped the use of the scribal witness.³⁷

The chapter claims that soon after his victory over the Shang, King Wu 武王 fell ill and that his advisor and brother, the Duke of Zhou, promised himself to the ancestral spirits to save Wu's life. A scribe wrote down the promise, announced it to the spirits, and the record of the promise was then stored in a casket. A year later, King Wu died of another cause and his brothers spread the false rumor that the Duke of Zhou wanted to harm the heir to the throne, King Cheng. The Duke left the capital, killed his brothers, but failed to win back Cheng's confidence. Thereupon, severe storms and lightning threatened to destroy the harvest. Cheng opened the casket and discovered the record of the promise. His scribes and staff confirmed the record's authenticity and confessed that the Duke had told them not to speak about the promise. Cheng welcomed the Duke back to the capital, the weather changed and the harvest was saved.

The role of the scribe has recently received much attention, particularly since the publication of the Tsinghua manuscript version of the text. Rens Krijgsman, for instance, has argued that his verification of the recorded promise of the Duke of Zhou indicates that written texts needed oral confirmation to gain trustworthiness.³⁸ However, as I see it, the doubts run much deeper. The scribe's testimony verifies much more than just the recorded promise. It verifies the entire narrative of the “Jin teng” chapter.

What catches the eye is that the scribe and his fellow staff are told to remain silent. This is important, as had they been allowed to tell King Cheng about the promise, Cheng would have never doubted the Duke's intentions, and the crisis would not have escalated. As the manuscript and the Shi ji 史記 account show, ancient readers were acutely aware of this detail. Both versions differ from the Shang shu chapter in mentioning the Duke's order to remain silent immediately after the recording of the promise.³⁹ In these versions the reader is informed right after the Duke offers himself to the gods why no one attempts to protect him.⁴⁰ In the Shang shu, the reader has to wait until the end to understand this by then puzzling detail. However, neither the Shang shu nor the other two versions explain why the Duke told everyone to remain silent in the first place.

This did not go unnoticed by later readers of the text. The Kong Yingda 孔穎達 (574–648) commentary, for instance, claims that the Duke's behavior is testimony to his modesty and virtuous character.⁴¹ He did not want others to know about his selfless act because he had done it out of a pure sense of duty. However, this explanation will eventually leave us unpersuaded. Why then would the Duke not allow his staff to speak even after the rumors against him had begun to spread? Did he put his principles above the interests of the kingdom? And why would no one dare to betray the Duke's orders and secretly inform the King? The longer we dwell on these questions, the more evident it becomes that the reason rests on the extradiegetic level of the narrative. The scribe and his staff were made part of the narrative because someone had to accompany the Duke and bear witness to his promise. But since their testimony would have prevented the plot from progressing, the Duke was portrayed as ordering them to remain silent.

For the sake of clarity and argument, let us imagine that the Duke was alone during the ritual. It was he who wrote down the promise.⁴² Now his young and inexperienced nephew must realize on his own the falsehood of the accusations. But why would that happen? Thus, in a miraculous turn of events, Heaven steps in and releases ominous signs. This causes King Cheng to check if there are still unfulfilled promises to the ancestral spirits. He opens the casket with the written records of past promises inside and discovers a record carrying his uncle's name.⁴³ Cheng realizes his mistake and both are happily reunited. We note that, again, the scribe is not needed to sustain the storyline. Not only that, but the result is also less messy and more suspenseful than our extant versions of the “Jin teng” narrative. No scribes and other persons lurk in the narrative's background, distracting us from the much more meaningful interactions between the Duke, King Cheng, and Heaven. However, it also becomes vulnerable to doubts. On whose authority do we have that the Duke of Zhou acted as claimed? Only the Duke himself, the suspect of the allegations, would be left.

The Guo yu's 國語 account of the Li Ji unrest (Li Ji zhi luan 驪姬之亂)

Another example in which a scribal witness appears to have been added to a pre-existing narrative is the Guo yu version of the Li Ji unrest. In the lead-up to the events, the famous scribe Su 蘇 warns Lord Xian of Jin 晉獻公 (?－651 b.c.e.) against his planned campaign against the Li Rong 驪戎 because of a negative divination result. Xian ignores his warning, defeats the Li and returns with Li Ji. Su then continues to warn Xian and also Jin's grand masters. However, as none of them want to help him persuade Xian, the tragedy inexorably takes its course. Li Ji plots to make her son Xiqi 奚齐 successor to Lord Xian, drives the official heir to the throne, Shensheng 申生, to suicide and forces his remaining two brothers to flee the country. Her coup causes political unrest and punitive actions by other countries against Jin.

Su plays the wise but powerless advisor, a motive central to many ancient Chinese narratives. His warnings convey the moral lessons of the Li Ji unrest. However, examined more closely, Su's warnings are much more. He refers to historical precedents to substantiate his claims and also explains how the shape of the “tortoise crack” (zhao 兆) foretells future events.⁴⁴ Su's warnings edify the reader but also provide evidence for a remarkable turn of events. After all, Li Ji, a non-Xia woman, was to bring turmoil to one of the most powerful central states.

Verbatim repetitions in the received text suggest that Su's warnings form an independent narrative that was added to the main account.⁴⁵ Every time Su appears on stage, his words and the responses of his interlocutors are framed by proleptic elements that are sometimes repeated word for word only lines later.⁴⁶ The main account, on the other hand, remains unaffected by Su's speeches. There are two later, half-hearted references to his prophecies, but both are only superficially attached to their context.⁴⁷ They trigger no responses. Again the scribe can be removed from the text while the diegetic integrity of the main account remains intact. Moreover, Su's prophecies and their prolepses give away the ending.⁴⁸ Therefore, removing him would, as in the “Jin teng” chapter, add suspense to the narrative. The tragic ending would then slowly but steadily unfold before the reader's eyes.

The Resurrection of Dan 丹

Moving on to cases that separate scribal witnesses and narrative, an example illustrating the correlation between the scribal semblance and authority of an account is a manuscript text excavated in 1986 from a late Warring States or early Qin period tomb at Fangmatan 放馬灘. The text describes the resurrection of a man named Dan and his observations about the life and habits of the dead. His instructions on dealing with the spirits of the deceased seem to have been the central concern of the text. A contemporaneous manuscript published by Peking University in 2012 contains a closely related text that ends with a nearly verbatim list of instructions.⁴⁹ Apparently both were representatives of the same genre of paranormal literature. However, unlike the Peking University manuscript, the Fangmatan text purports to be an official report of an administrator called Chi 赤 to an unnamed Royal Scribe. Based on the mention of the Wei 魏 general Xi Wu 犀武 (d. 293 b.c.e.), we can conclude that the text portrays events of the late fourth or early third century b.c.e. However, the precise historical and geographical circumstances remain conspicuously uncertain.⁵⁰

The addressee suggests that the manuscript is a copy of Chi's report or a royal archive document. That this gives it authority needs hardly be mentioned. However, that is not all that the author of this text has done to substantiate its credibility. Meticulously he has Chi and Dan enumerate all the evidence that proves Dan's resurrection. He translated the narrative into the form of an administrative document to the extent that makes it difficult to distinguish it from a real one:⁵¹

八年八月己巳，邸丞赤敢謁御史：

大梁人王里徒髡曰丹□□。七年，丹矢傷人垣離里中，因自刎殹。□之於市三日，葬之垣離南門外。三年，丹而復生。丹所以得復生者，“吾犀武舍人。”犀武論其舍人尚（掌）命者，以丹未當死。因告司命史公孫強，因令白狐穴屈（掘）出。丹立墓上三日，因與司命史公孫強北，之趙氏之北地柏丘之上。盈四年，乃聞犬犻（吠）雞鳴而人食。

其狀：類（纇）益（嗌），少麋（眉），墨，四支（肢）不用。

丹言曰：“死者不欲多衣，死人以白茅為富，其鬼勝於它而富。”

丹言：“祠墓者毋敢嗀（哭）。嗀（哭），鬼去敬（驚）走。已，收腏（餟）而罄之，如此鬼終身不食殹。”

丹言：“祠者必謹騷（掃）除，毋以淘灑祠所，毋以羹沃腏（餟）上，鬼弗食殹。”

The eighth year, in the eighth month, day ji si (six), the Administrator of Di, Chi, dares to visit and report to the Royal Scribe:

A man from Daliang, Wang Village, shaved convict, named Dan [two undeciphered graphs]. In the seventh year, Dan shot and injured a man in Yuanli Village, and because of it slit his throat. [Dan] was [one undeciphered graph] at the market for three days and buried outside the south gate of Yuanli. Yet three years later, Dan was restored to life. Dan was allowed to be restored to life because [as he claimed] “I was Xi Wu's Caretaker.” Xi Wu appealed to those in charge of his Caretaker's life-mandate, because he thought that Dan did not deserve to die. He consequently made a declaration to the Scribe of the Director of the Life-mandate, Gongsun Qiang, who then had a white fox dig a tunnel. Dan stood on the tomb for three days, then departed northwards to Zhao in company with the Scribe of the Director of the Life-mandate, Gongsun Qiang, and went to Boqiu in the North Territory. Fully four years later, he heard dogs barking and roosters crowing and ate human food.

As for his appearance: he had a scar on his throat and sparse eyebrows. He was branded and could not use his four limbs.

Dan said the following: “The dead do not want many clothes. Dead people regard cogongrass as precious, to the ghosts it is more precious than anything else.”

Dan said: “Those who offer sacrifices at tombs should not dare to wail. If they wail, the ghosts depart and flee in fright. If they eat all offerings after the completion of the sacrifice, then the ghosts will never eat [their offerings] again.”

Dan said: “Those who offer sacrifices must carefully sweep and purify. Do not wash the place of sacrifice with muddy water. Do not pour the boiled dish over the sacrificial food, the ghosts will not eat it.”⁵²

In summary, the report relates that Dan committed suicide after he had shot and wounded another man. Dan was then brought back to life by his lord Xi Wu 犀武, who contacted the spirit of the renowned statesman Gongsun Qiang 公孫強 (d. 487 b.c.e.), the Scribe of the Director of the Life-mandate in the realm of the dead. Unfortunately, we do not learn how Gongsun Qiang was contacted, but it is possible that Xi Wu was already dead and reached out to him from the netherworld. Regardless, Xi Wu's request was granted and Dan reentered the world of the living, although in a permanently branded condition.

After the preamble, which states the alleged date, addressee, and author of the report, the text opens by identifying Dan based on his registered location and social status and continues to detail dates, places, and the names of the people involved. It further indicates that the administrator Chi and/or his staff saw and questioned Dan in person. This first becomes evident when Dan suddenly says that “I was Xi Wu's Caretaker” (wu Xi Wu sheren 吾犀武舍人), suggesting that Chi's report is based on Dan's own words.⁵³ Further proof of inquiry and even visual examination are found at the end of the report. Chi refers to the traces that his suicide and punishment left on Dan's body and lists several statements by Dan about the afterlife. To be sure, an important incentive to include them may have been to provide the living with helpful instructions. However, Dan's instructions also contain details that only a dead person could know. In the context of the report, they are presented as further proof of Dan's death and should be considered as such.

Similar to the inscription of the San shi pan, the reference to the Royal Scribe is just one item in a long list of things to convince us of the reality of Dan's resurrection. But that by no means makes him unimportant. For without him, this report and its underlying investigations would not have materialized. As in the San shi pan, it is the scribe and the official guidelines of administrative writings that determine the content and format of the text. It may be that the inscription of the San shi pan was based on authentic scribal records, while Chi's report drew inspiration from a folk tale, as scholars have argued. However, both shed light on the nature of scribal correspondence in their own ways.

The Chunqiu Zuo zhuan 春秋左傳

Scholarly consensus has it that the compilers of the Chunqiu Zuo zhuan undertook considerable efforts to cut formerly united and independent narratives into sections according to the chronology of the Chunqiu. In the case of parallels between the entries of the Chunqiu and the Zuo zhuan narratives, the compilers often added glosses explaining how the wording of the Chunqiu resonates with certain details of the Zuo zhuan accounts. Given the apparent farfetchedness of many of these comments, the compilers doubtlessly imposed their subjective readings on the Chunqiu to accommodate them with the Zuo zhuan, or, in some very intriguing cases, with their equally subjective moral interpretation of the true meaning of the Zuo zhuan narratives.⁵⁴

Traditionally the assumption has been that the compilers chose the Chunqiu because of its relationship to Confucius. However, the compilers do not attribute the entries of the Chunqiu to Confucius but to nameless scribal “officials” (guan 官).⁵⁵ Their anecdotes also mention a potential textual ancestor to the received Chunqiu in the possession of the Grand Scribe of Lu 魯.⁵⁶ While the Junzi 君子 commentary famously claims that only a sage could have “arranged” (xiu 修) an extraordinary text such as the Chunqiu, it is neither clear who is meant, nor can it be ruled out that the comment reflects a later perception of the text.⁵⁷ This certainly allows the thought that the initial compilers who put together most of the work regarded the Chunqiu primarily as a scribal work and were mainly interested in the Chunqiu because of its scribal nature. It was only at a later date that the work became associated with Confucius, which then led to the fact that the Chunqiu ends with his death.⁵⁸

The most instructive evidence for this hypothesis is that citations of scribal records appear to continue (now as part of the Zuo zhuan) after Confucius has died.⁵⁹ The compilers also reference entries of other (unnamed) annals elsewhere in the text.⁶⁰ It is crucial that these amendments generally tend to support but never challenge the material of the Zuo zhuan. We can only suppose that there existed annal entries that straightforwardly contradicted certain Zuo zhuan narratives. After all, the Zuo zhuan is full of moralistic, defamatory, and often simply impossible accounts.⁶¹ However, entries that contradict the Zuo zhuan are never cited. Minor deviations between the Chunqiu and the Zuo zhuan occur, but are explained away by the compilers.⁶² The compilers were not interested in drawing up a complete and transparent collection of annal entries. Instead, they wanted to strengthen the scribal foundations of the Zuo zhuan.⁶³ Take, for instance, the following passages attributed to the ninth year of Lord Xi 僖 (651 b.c.e.), which portray the final stage of the Li Ji unrest mentioned above:

Chunqiu:

冬，晉里克殺其君之子奚齊。

Winter, Li Ke of Jin murdered Xiqi, the son of his ruler.

Zuo zhuan:

冬，十月，里克殺奚齊于次。書曰:“殺其君之子，未葬也。”荀息將死之，人曰：“不如立卓子而輔之。”荀息立公子卓以葬。

十一月，里克殺公子卓于朝，荀息死之。

君子曰：“《詩》所謂‘白圭之玷，尚可磨也，斯言之玷，不可為也，’荀息有焉。”

齊侯以諸侯之師伐晉。及高梁而還。討晉亂也。令不及魯，故不書。

In winter, in the tenth month, Li Ke murdered Xiqi in the mourning hut. The record says “killed the son of his ruler” because Lord Xian had not yet been buried. Xun Xi wanted to die defending Xiqi when someone said, “Would it not be better to establish Zhuozi as ruler and assist him?” So Xun Xi established Zhuozi as ruler to complete the burial of Lord Xian.

In the eleventh month, Li Ke murdered the Lord [of Jin's] son Zhuo in the court, Xun Xi died for him.

The noble man says: “This is what the Odes refer to as ‘A flaw in a tablet of white jade still can be polished away. But a flaw in one's words—nothing can be done about that.’ Such is the case of Xun Xi.”

The ruler of Qi attacked Jin with troops of the feudal lords. They advanced as far as Gaoliang and then turned back. This was to punish Jin's unrest. The order did not reach Lu and therefore was not recorded.⁶⁴

Note that the Chunqiu only mentions the killing of Xiqi. The compilers argue that it states “son of his ruler” to indicate that Xiqi had not yet become the ruler of Jin, as his father, Lord Xian had not been buried. This detail is vital for the subsequently described event. For Xun Xi is said to have abandoned his original pledge to die for Xiqi and instead put Xiqi's step-brother Zhuozi on the throne to bury Xian.⁶⁵

The Chunqiu mentions none of the successive events. While Li Ke's killing of Zhuozi was at least as noteworthy as the killing of Xiqi, the Chunqiu does not mention it and the compilers do not explain why. However, they claim that the Chunqiu misses a record of Qi's campaign against Jin. We read that the campaign was to punish the disorder in Jin and that the scribes of Lu did not record the campaign because Lu did not participate in it.⁶⁶ This implies that those states that joined the campaign had records and what we read is a copy thereof. The argument is subtle but clear: these scribal records attest to the erupting turmoil in Jin and thereby also to the deaths of Zhuozi and Xun Xi.

This, then, is evidence that some compilers not only used the Chunqiu to support the anecdotes of the Zuo zhuan but also entries of other annals. Confucius’ compilation of the Chunqiu cannot have been decsisive for the inclusions of these alternatives records. The implied belief that they were made by scribes, on the other hand, may have. While this single example cannot be taken as representative for a multifaceted and heterogenous work such as the Chunqiu Zuo zhuan, it demonstrates a link between the text and the scribal witness. The main difference being that the “witnesses” of the Chunqiu Zuo zhuan are not the scribes themselves but their written testimony.

As noted above, it may be worth considering this “non-Confucian” approach to the Chunqiu as the beginning of Chunqiu exegesis. Perhaps the initial compilers of the Chunqiu Zuo zhuan did not read deeper meaning into the entries of the Chunqiu because they believed that they had gone through the hand of Confucius or other sages. Although this idea is certainly present in parts of the Zuo zhuan, it fails to explain the presence of non-Chunqiu entries and therefore may well have emerged during the later transmission of the text. Perhaps Chunqiu exegesis started because compilers attempted to substantiate their stories with the help of the Chunqiu and/or other collections of annal entries, which then eventually became the text extant today. While this scenario naturally remains speculative, it would connect the textual genesis of the Chunqiu Zuo zhuan to the larger literary phenomenon of the scribal witness, whose general prevalence suggests that it did in one way or another affect the compilation of the Chunqiu Zuo zhuan.

Conclusion

With the Chunqiu and its formative influence on Chinese imperial historiography, we can turn to the Grand Scribe Sima Qian.⁶⁷ In the postface of his work, Sima Qian declares the Chunqiu an unattainable literary paragon, to which his writings are no match. This is an act of false modesty,⁶⁸ as he deliberately encouraged such a comparison by placing his attempt at reconstructing and imitating annalistic writings at the beginning of his work (the “elementary records,” ben ji 本紀). The sheer scope of his writings suggests that he wanted not only to emulate but to surpass the Chunqiu.

In this, Confucius’ authority was undoubtedly important. However, Sima Qian's comments about his travels and personal encounters suggest other factors were at play too.⁶⁹ In light of the examples above, these comments seem less exceptional and more like a continued development of an ancient tradition. From a larger historical perspective, Sima Qian acted as his own scribal witness: he shared records of his (a scribe's)⁷⁰ personal observations. The Fangmatan manuscript and the inscription of the San shi pan demonstrate that scribes and their staff had begun to conduct and record such “fieldwork” long before him. We only read about their endeavors sporadically because the resulting writings have survived the ages only in the rarest of cases. Initially, however, they probably comprised a large part of ancient Chinese literacy. Qin and early Han legal documents overwhelmingly confirm that on-site visits and the questioning of witnesses were the daily working routine of scribes and other administrative personnel.⁷¹ Sima Qian's remarks thus can be seen as an attempt to reconnect with the humble scribal roots of his ancestors. An unprecedented aspect of his approach was that he shared his findings in the context of composing a comprehensive work on world history. This, by all appearances, had not happened before.

As I have tried to show, new perspectives on individual texts and ancient Chinese narrative culture in general emerge when we approach the scribe as a witness. In particular the examples drawn from Eastern Zhou literature feed the impression that there existed a constant stream of fabulous narratives from which extant texts drew inspiration. The original audiences of these narratives might have been convinced of their credibility or simply did not bother. But the way those who composed, edited, or compiled our texts “scribalized” them suggests that they considered these stories of value while at the same time lacking authority. In some cases, the decisiveness (and sometimes also the awkwardness) of how the narratives were adapted seems to have left some indications on how they had originally been told. We encounter suspenseful plots, surprising twists, and stories about life after death. The origins of traditional Chinese fiction have been discussed extensively and controversially. It seems to me that a good share could be found behind the scribal surface of Zhou-period and Han-period texts.

At the same time, the scribal witness demonstrates that starting from the tenth and ninth centuries b.c.e., writers felt that the sages, textuality, or appeals to moral values were insufficient to lend their texts authority. These aspects may well have had utility in other contexts. But they were not compelling enough for the examples discussed in this article. Hence, writers found it necessary to provide epistemic proof of what was claimed to have happened. And although scribal witnesses were not always personally present at the scene of the event, their staff and records offered authoritative access to such evidence.

None of the above rules out in principle that these writers could have been scribes themselves. But given the aforementioned fact that we have no direct literary evidence of scribes composing narrative literature until Sima Qian, I find it important to end by noting that the rhetorical device of the scribal witness was naturally available to other writers as well.