Dubious Facts

The Evidence of Early Chinese Historiography

Garret P. S. Olberding

• Publisher: SUNY Press
• Year: 2012
• ISBN:
  9781438443911
• Topic: Storia
• Subtopic: Storia cinese

What were the intentions of early China's historians? Modern readers must contend with the tension between the narrators' moralizing commentary and their description of events. Although these historians had notions of evidence, it is not clear to what extent they valued what contemporary scholars would deem "hard" facts. Offering an innovative approach to premodern historical documents, Garret P. S. Olberding argues that the speeches of court advisors reveal subtle strategies of information management in the early monarchic context. Olberding focuses on those addresses concerning military campaigns where evidence would be important in guiding immediate social and political policy. His analysis reveals the sophisticated conventions that governed the imperial advisor's logic and suasion in critical state discussions, which were specifically intended to counter anticipated doubts. Dubious Facts illuminates both the decision-making processes that informed early Chinese military campaigns and the historical records that represent them.

CHAPTER ONE

Introduction

THE EVIDENTIARY RELIABILITY OF MEMORIALIZED UTTERANCES AT COURT

Of the many problems surrounding the study of early Chinese historiography, one of the most troubling is the conspicuous tension between the narrator's moralizing commentary and the description of events. The earliest works of historiography—such as the Book of Documents (Shangshu 尚書), the Spring and Autumn Annals (Chunqiu 春秋), and the seminal commentary to the Annals, Zuozhuan 左傳—cannot be securely ascribed to a single author.¹ Perhaps because of this, scholars often attribute the tension to the complex accretion of commentarial intrusions and editorial changes. By contrast, with the Han histories, with the Records of the Grand Historian (Shiji 史記) and the History of the Former Han (Hanshu 漢書), there is the presumption of a dominant authorial voice, and with this authorial voice attends the assigning of a unitary authorial intent. Nevertheless, as in the earlier works, there often remain frustrating contradictions between moralizing commentaries and the specifics of the narrative. Consequently, as with the earlier works, the reader is often left to wonder what their authors' intentions actually were. Such concerns touch upon not only the ideological tensions between moral and narrative detail but also numerous tensions proceeding from various internal narrative contradictions, ironic voices, and complicated dramatic devices. Both the narrator and the historical characters at times appear to speak duplicitously, or at least disingenuously. Thus when interpreting the narratives, the discussions, the debates, and the references that comprise the material of the histories, we simply cannot take the speaker, whether narrator or character, at his word. This tendency toward vagueness and subtlety suffuses even some of the most outspokenly critical historical characters, particularly when their words are intended for submission to the court. Early Chinese historiography is a history frustratingly complicated by encoded speech.

Of course, encoded speech was vital to the monarchical circumstance. An anecdote about Emperor Yuan 元帝 misunderstanding the meaning of his own orders illustrates its ubiquity:

At this time the emperor had just come to the throne and he did not realize that the phrase “instruct the master of guests to summon them and turn them over to the commandant of justice” meant that they were to be taken to prison, and he therefore approved the recommendation. Later, when he asked to have Chou K'an and Liu Kengsheng called into his presence, he was told, “They are bound and in prison!” Astounded at this information, he said, “I thought they were only to be taken to the commandant of justice for questioning!” and he began to berate [the Chief of Palace Writers] Hung Kung and [his assistant] Shih Hsien. The two knocked their heads on the floor in apology, after which the emperor said, “See to it that they are released from prison and restored to their positions!” 時上初即位，不省「謁者 召致廷尉」為下獄也，可其奏．後上召堪﹑更生，曰繫獄．上大驚曰：「非但廷尉 問邪？」以責恭﹑顯，皆叩頭謝。上曰：「令出視事。」²

To interpret court speech successfully, the reader, like the monarch, is obliged to pay attention to the rhetorical undertones. Naturally, there can be unjustified and excessive interpretation of insinuation (as all too many traditional commentators of the Annals have engaged in), but a heightened awareness of insinuation is nevertheless important, particularly in the interpretation of the highly rhetoricized ministerial addresses. That ministers must rely upon muted insinuations is clearly reflected in the numerous recommendations by early Chinese thinkers about the need for extreme delicacy in the choice of language. Regarding the selection of illustrative examples, the Xunzi 荀子 proclaims: “Channel [examples] as if with canal ditches, force them as if with the press-frame, and accommodate them to the circumstances so that your audience will get hold of the idea under discussion, yet will not be given offense or be insulted.”³ In the selection of even small illustrative details, the minister must always be aware of the need to avoid risking offense.⁴ When directed toward political ends, court address had to be formulated carefully, both to avoid arousing the ire of the monarch and to address certain doubts that the speaker (or writer) presumes the monarch may have.

On a general level, this study is driven by a concern with the factual reliability of memorialized utterance, but its central focus is actually somewhat narrower—to understand the logic of ministerial address through an analysis of its evidentiary conventions. The material for my investigations are the “memorials” concerning the waging of military campaigns (which, for my argumentative purposes, I hereafter label as “hortatory addresses” or “addresses,” for short). Save for a few recent studies, most notably by David Schaberg and Yuri Pines, these texts are treated as contributing very little to the study of early Chinese historiography.⁵ Yet as reflections of the manner in which intellectual debates were actively drawn into contemporary court discussions, discussions of matters of immediate political and social import, these addresses are invaluable. They show how, or whether, various intellectual trends were imported into court debates, and thus can potentially reveal the extent to which these debates had direct and obvious influence on the handling of court affairs. They also can offer unique insight into the everyday developments of policy: what was discussed, what were the major concerns, and what considerations were possibly effective in shaping the course of state action, of the political realities as they were taking place. Naturally, much like the remonstrances of the Eastern Zhou period, these recorded addresses certainly cannot be treated as exact transcriptions of actually presented addresses but as either loose representations or reductions of such, whether originally presented in writing or orally. The accuracy of the representation of any original content does not fatally impact the quality of their general representation of the content of a particular address or of the rhetorical conventions of their time.

As my concern is with the role of evidence, I focus on those addresses that discuss the pursuit of military campaigns, for unlike numerous other topics of court deliberation, they regularly depend on evidentiary claims with bases in something other than pure hearsay or moralizing precept. While the military addresses do not completely avoid grandiose moralizing, in the main they are very considerate of specifics significant to any military decision, including details relating to geography, population, foreign relations, and previous military enterprises, of both recent and distant vintage. Unfortunately, as will be explained later, there are no obvious linguistic or genre characteristics by which the early histories signify a ministerial memorial or address. The criteria by which I denote recorded statements as “addresses” are two:

1. The statements are relatively thorough and deliberative; that is, they are not merely brief evaluative comments or reports. They present a certain vision of the undertaking, the problems that would be faced in the undertaking, predictions for its outcome, and explicit or implicit recommendations.
2. They are, in the main, not marked as part of a larger extemporaneous conversation.⁶ They employ certain rhetorical conventions common to those statements explicitly described as being presented to the monarch.

While these criteria do not delimit the number of statements to a quantity that allows for a comprehensive survey, they still permit the formation of a focused and coherent picture of the typical structures present in military address.

My ultimate concern in these analyses is to articulate how evidence was used to ground political deliberations, to determine in what way it functioned as warranting the knowledge claims and the consequent judgments and recommendations ministers submitted to influence executive action. Through these investigations, we can see how political persuasion, how rhetoricized speech appropriates externally produced information, and what role externally guided “proof” plays in the political process. In its broadest application, the study can assist in evaluating the extent to which ancient Chinese history is the handmaiden of rhetoric, how driven it is by the demands and outcomes of court politics, not just in terms of its narrative course (i.e., what policies the addresses appear to have been instrumental in effecting)⁷ but also in terms of its general evidentiary bases. But there are other narrower, yet equally significant implications, for example, the detailing of the limited effect of knowledge claims on political discussions.

Because debate in the Chinese imperial court was most often conducted before a general, nonspecialist audience, the evidentiary bases had to be those that could garner the widest appeal, what Aristotle, still among the greatest analysts of political address, referred to as endoxa, or “reputable opinions,” that is, opinions that most or all could easily accept. Of course, not all “reputable opinions” were equally believable. Thus what we must explore is the difference between those statements that appear to be treated as beyond doubt—statements I formally label “common sense”—and those that are shown, through questions raised and critical comments made, as being somewhat open to doubt, statements I categorize as “popular knowledge.” Through an analysis of what was potentially available to doubt and what was not, we can begin to understand what was considered firm evidentiary ground, or “fact.” The analyses I will perform are naturally contingent upon discerning evidentiary distinctions. Given the indirect and sometimes cryptic nature of communication within the monarchic environment, the analyses will require careful attention to nuance and subtlety. If performed satisfactorily, they could help to reshape our current understanding of just exactly what was at issue in the deliberations, what was considered a fundamental truth or essential matter for the discussion of these military matters, and what was considered merely a suggestion or comment worthy of some brief consideration.

Making the above knowledge-related distinctions presumes that the addresses were composed to speak to substantive debates, rather than, for example, merely serving formulaic, ritualized ends. I argue they do, that they attend to the doubts specific to the circumstances in the most cogent manner possible. But a fair evaluation of the addresses' use of evidence, I argue, will demand that we establish a more appropriate notion of the standard of truth by which they were judged. Instead of accuracy, a standard that ideally admits of no competing claims to truth, I suggest that truth in this context must be defined as a standard of relevance to the overarching concerns of the executive, a formulation I will represent as the “truth of the matter.” Indeed, this formulation accords with what has been suggested to be a primary function of historical literature in classical Chinese civilization: to provide the reader with an education on the issues and details that would maximally impact his assessment of what action to take, or to reaffirm the conduct that is most becoming to a superior ruler. Such a function is most manifest in the ritual representations of historical occurrences but it is certainly not limited to such.

While my central concern is with the ministerial addresses of the early Han, I inaugurate my investigations with an analysis of two models of late Warring States military address presented by Su Qin 蘇秦 and Zhang Yi 張儀 to the King of Wei 魏, both included in the Zhanguoce 戰國策, a miscellaneous collection of persuasions and intrigues.⁸ My examination of these serves several ends. First, it offers support for my claim that the use of evidence in political discussions was restricted to neither an individual style nor a discrete time period. If the military addresses of the Intrigues are representative at all of the late Warring States, it appears that the evidentiary standards of the early Han shared significant resonance with those of the late Warring States. And if standards of evidence are diachronic, it follows that they were restricted to neither an individual style nor single official norm or guideline. Indeed, upon extensive survey of the early Han addresses, it appears the norms for the use of evidence were not significantly altered across station or person. The standards used by the highest levels of early Han officials were the same, or very similar, to those used by lower-level officials.

Assessments of the factual content of ancient historical literature find ample precedent in studies of classical Greece and Rome. Classicists, most notably Paul Veyne, have struggled to understand how history was understood, in what way history could or should include what we now might consider to be fictional accounts. My work takes inspiration from these studies and hopes to offer a greater insight into how ancient societies framed themselves to themselves. In previous studies of Han intellectual history, scholars have tended to represent the debates and arguments of the time ungenerously as intellectually deficient, as leaning too heavily on propaganda, stock formulae, and specious argument.⁹ Debates on military affairs, for instance, are usually characterized as being between two stereotyped positions—the dovish Confucian moralizer and the bellicose realist. This stereotype extends to debates on other issues as well, most notably the debates on economic policy recorded in The Debates on Salt and Iron (Yantielun 鹽鐵論). My analyses, I contend, will reveal that such stereotypes can often result from an incomplete understanding of what was generally accepted as legitimate evidence, augmenting past research through a careful evaluation of what theoretical notions, canonical references, and historical details were possibly accepted as general knowledge when justifying a course of action, in matters relating not only to military affairs but also perhaps to other topics.

Because the force of the endoxic claims forwarded at court in overcoming prospective knowledge-related, or epistemic, doubts cannot be fully assessed in terms of their accuracy or justifiability, I introduce two approximate evaluative concepts: “epistemic quality” and “epistemic weight.” The term “epistemic quality” derives from H. P. Grice's seminal paper, “Logic and Conversation,” wherein Grice defines four categories of maxims necessary for meaningful contribution to dialogue. The most important category relates to the “quality,” or general truthfulness, of the contribution: “Under the category of quality falls a supermaxim—‘Try to make your contribution one that is true.’—and two more specific maxims: 1. Do not say what you believe to be false. 2. Do not say that for which you lack adequate evidence.”¹⁰ For a conversational contribution to really be any kind of contribution at all, Grice asserts, it must be truthful. If not, the contribution is not an inferior kind of information, “it just is not information.”¹¹ The primary question is what can be considered truth and under what auspices it can be considered such. I use the term both in the singular, “epistemic quality,” when speaking of the overall quality, or “truth-content,” of a statement, and in the plural, “epistemic qualities,” when speaking of the factors that impact whether it can be seen as truthful. I describe the epistemic qualities of a statement as relating to its derived inferences, its place in the context of the argument, the appearance of its generally being undoubted by proponents and opponents, its pertinence to the argument, and so forth.

The second evaluative concept, “epistemic weight,” bears close kinship with the standard notion of justification. As with the justificatory value of a statement, in its ideal expression epistemic weight depends solely upon the epistemic qualities of a statement or class of statements. The epistemic qualities of a statement are those that give a statement weight. Yet in contrast to the notion of justification, the epistemic weight of a statement is constrained by the rhetorical demands of the politicized court environment; thus it is not solely limited to the objectively testable effect the evidence bears upon the course of the argument. Simply put, the epistemic weight of rhetoricized evidence has both subjective and objective components. Its objective components derive from features that can be tested against a generally accepted standard—distances to be traveled or the comparative size of troops or provisions relative to the enemy's or what is needed for a campaign, for instance. Its subjective components are related to the “popularity” a belief has acquired, without any obvious relation to an objective standard.¹² Both would be of significance in the weighing of the credibility of the evidence presented in the politicized court environment.

In the typical scenario in which an address is presented, the monarch, faced with a problem of state, requests his ministers present their thoughts. Among the discursive challenges faced by ministers, two impinge heavily on the problem of doubt: the monarch neither takes well to challenges to his authority nor does he often have patience for lengthy instruction.¹³ The former speaks to the general risk of advising the monarch, for advice naturally involves some measure of critique, and even oblique assertions of error may result in gruesome punishment. The latter, a demand for concision, bears on avoiding raising not only the monarch's ire and personal self-doubts but also his epistemic doubts. Such a concern is aptly exemplified in Grice's maxim relating to the “quantity” of information one should provide when one is following what he calls the “cooperative principle”: “Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.”¹⁴ When cooperating in discursive interaction, according to Grice, one should limit the quantity of information to the amount that is needed and of relevance, no more, no less. Overinformativeness is as detrimental to cooperative conversation as underinformativeness because, in Grice's words, it can raise “side issues” and because the hearer “may be misled as a result of thinking that there is some particular point in the provision of the excess of information.”¹⁵ This caution about being overinformative bears upon my concern with raising doubts in the hearer, for not only can the hearer be led to think that there is a particular point in the excess information, he can also be led to reassess the issues involved and be more prone to objections, of his own devising or others'. I will show that ...