SONG HISTORY NARRATIVES AS GRAND ALLEGORY

History imposes syntax on time.

Nancy Partner

INTRODUCTION

My interest in Song history emerged well after my initial training in pre-modern Chinese literature.* When I first began to read Song historical texts, I knew nothing about historiographical theory, so I blithely used the same techniques for close reading that I had used to decipher Tang-Song poetry. Using these strategies for reading literature—nothing much beyond old school European philology—I was able to document the different chronological layers in the Song History (Songshi 宋史) biography of Qin Gui 秦檜 (1090–1155).¹ When this article appeared in 1998, several colleagues more attuned to theoretical issues than I was told me its approach and findings were very much au courant. It seems, stumbling around in the dark, I had unwittingly taken the linguistic turn.

I was surprised and frightened, but also intrigued. My scholarship had never followed recognizable currents, and I knew nothing of postmodernism. I found its jargon mystifying and its complex psychological postures off-putting. Alun Munslow's Deconstructing History came to my immediate rescue, and then I understood what my colleagues had meant. The 1998 article on Qin Gui and my subsequent studies of Song historiography, while hardly postmodern, do share with deconstruction certain concerns about the relationship of history to historiography, about the role of language in the writing of history, and about the interplay of ideology and historical fact.²

The present article contends that the received version of Song history, as transmitted in the official Song History of 1345, can be understood as a grand allegory, or metanarrative, that grew from the rhetorical postures of Song dynasty political discourse. I did not pursue this research in response to any theory of history—deconstructionist or otherwise—nor do I intend this article to advance any specific theoretical perspective. On the contrary, what follows is the simple product of a once naïve philologist who for 25 years has continued to read historical texts as he once read literature.

As a student of literature, I did not look for facts; I looked for themes. I did not focus on which facts were true but on which fact the historian, as a writer, had chosen to emphasize, or foreground, over other facts that he also had at hand. Had he organized these foregrounded facts into coherent themes? Had he arranged these themes into consistent, larger structures of meaning? It soon became clear that many Song historical works—the biographies in the Song History for example—had indeed been composed in this way. They were not collections of disparate facts but collections of foregrounded facts chosen to illustrate specific themes, with these themes then organized into larger structured narratives.

Of course, most historians of Song do not read their texts in this way. Some sort through factual contradictions in order to determine which facts are more reliable than others. They then assemble these chosen facts into specialized histories of Song politics, institutions, or society. Others, more suspicious of received texts in general, use the stories told in historical texts to help reconstruct the stories the texts do not tell. Useful as these approaches are, their historiographical component usually remains a perfunctory bibliography of who, what, and (sometimes) when. Rarely do they ask the fundamental historiographical question: how and why were these texts composed, or secondarily, how have they survived into present times. However, if histories are collections of themes organized into larger structures, then historiography represents the first and most vital stage in the creation of history.

This interplay of history and historiography is central to my notion of Song history as a grand allegory that grew from the intersection of political conflict and historiographical writing. And so, I read historical texts not to determine reliable facts but to detect traces of the initial historiographical construction and subsequent reconstruction of facts. Often, the visible traces of this linguistic manipulation enable us to observe a story that is very different from and usually truer than the “facts” that that history purports to contain. As a scholar of medieval Western history has written: “History is meaning imposed on time by means of language: history imposes syntax on time.”³

History thus emerges from a historiographical process in which successive authors/historians use rhetoric to reorganize facts they cull, often for contemporaneous political purposes, from an already existing historical record. My perspective on the making of Song history thus shares much with theorists of history in the past century who emphasized the dependence of history upon language. Already in the nineteenth century, Fustel de Coulanges (1830–1889) declared “l'histoire se fait avec des textes,” implying that history results largely from the historian's manipulation of prior text.⁴ A century later Roland Barthes expanded this focus into the epistemological dictum that “le fait n'a jamais qu'une existence linguistique.”⁵

This article focuses on the Song History. But the Song History exists as the end product of a long series of official and private historiographical interventions that first began well after the dynasty began. As I have explained elsewhere, the official Song historiographical operation processed primary official documents through successive triages and redactions into a cascading series of daily calendars (rili 日歷), veritable records (shilu 實錄), and state histories (guoshi 國史). The Song History, for various reasons, is unique among the dynastic histories because its text often permits modern historians to detect traces of this earlier processing that began during the dynasty itself.⁶ Today, three major works survive as vestiges of these earlier processes.

The Essential Song Documents (Song huiyao 宋會要), a vast and confused cornucopia of once-original bureaucratic documents, provides our best view into the beginnings of the Song historiographical process. The Long Draft Continuation of the Comprehensive Mirror that Aids Governance (Xu zizhi tongjian changbian 續資治通鑑長編), completed by Li Tao 李燾 (1115–1184) in 1183, covers Northern Song history and is based largely upon the official veritable records. Li Xinchuan's 李心傳 (1167–1244) continuation of this work, known as the Chronological Record of Important Events since 1127 (Jianyan yilai xinian yaolu 建炎以來繫年要錄) and completed in 1208, relies upon daily calendars and veritable records to extend coverage through 1162. These works each derive from different periods in Song history, each reflects to some degree the political orientations of its compiler(s), and each has suffered a tortured transmission down to present times; nonetheless, each contains sizable chunks from the textual corpus of official Song historiography, and many of these texts eventually found their way into the Song History.

Yet the relationship between these Song works and the Song History is neither direct nor linear. The rise of the Learning of the Way movement (daoxue 道學) as a political and historiographical force in the mid-twelfth century profoundly shaped the eventual reception and understanding of these earlier works. Beginning with Zhu Xi 朱熹 (1130–1200) in the 1160s, politicians and historians, some closely, some loosely associated with daoxue, gradually formulated a metanarrative on Song history that validated their own political and intellectual convictions. By the middle of the next century, historians such as Chen Jun 陳均 (1174–1244) and Lü Zhong 呂中 (fl. 1250) had perfected this historical edifice—what I describe in this article as the grand allegory of Song history. This narrative then became the guiding historical vision which the Yuan compilers relied upon to organize the Song History. Once this final stage in the process was complete, the original textual corpus was no longer relevant. The Song state histories were lost, and the three works mentioned above fell into such oblivion that the Qing academicians could only partially reconstruct their texts.

We must therefore distinguish carefully between daoxue historiography and daoxue history. The former refers to historians who write with a stated or implied daoxue intellectual or political orientation—Zhu Xi, Zhao Ruyu 趙汝愚 (1140–1196), Zhen Dexiu 真德秀 (1178–1235), Chen Jun, and the Yuan compilers of the Song History.⁷ In this sense, Li Tao and Li Xinchuan are not daoxue historians. Although Li Tao had close political ties to daoxue officials and sympathizers such as Lü Zuqian 呂祖謙 (1137–1181) and Zhou Bida 周必大 (1126–1204), his surviving writings do not manifest an intellectual commitment to the movement. Likewise, although Li Xinchuan's brothers were active promoters of Zhu Xi's legacy and early compilers of the Classified Conversations of Zhu Xi (Zhuzi yulei 朱子語類), Li Xinchuan's own stance on the movement was much more nuanced.⁸ The grand allegory is not the organizing vision or the ultimate goal of the Long Draft or the Chronological Record. It is, however, the impetus and overarching vision that animated the many thirteenth-century compressions and abridgements of these works, including those of Chen Jun.⁹ Thus, although the Long Draft and the Chronological Record contain most of the source texts, narratives, and motifs that ultimately made up the grand allegory, the vast bulk of these works presents a profusion of detail that does not contribute directly to this larger vision.

“Daoxue history” is then the metanarrative or grand allegory that came into being as the end product of daoxue historiography. To the degree that the Song History ratified this vision—and considering that scholars did not have practical access to the textual corpus necessary to deconstruct that vision until the twentieth century—this “daoxue history” was for many centuries the grand allegory that explained the entirety of Song history. One may compare the notion of such a daoxue history with Herbert Butterfield's well-known critique of Whig history.¹⁰ Butterfield charged that the needs of Whig politicians for historical justification had generated a history that led inexorably to their own political preferences for constitutional monarchy and British liberalism.

Seeking likenesses rather than unlikenesses between present and past, they created a narrative in which progress toward these political goals had flowed from the Reformation through Protestantism, the Glorious Revolution of 1688, the Whigs, and Parliament. In turn, they denigrated as retrograde and anti-progressive those countervailing forces that had run through Catholicism, the Tories, the king, and the absolute monarchy. Historical actors deemed to have fostered Whig goals received positive evaluations and became heroes; those who had opposed or thwarted them became villains. Whig historians employed radical abridgement of the complex and ambiguous primary historical record to foreground events that contributed to their preferred narrative and to omit those that did not. The result was a “giant optical illusion” so powerful that the research of most historians merely buttressed the illusion rather than challenged the basic story.¹¹

There are many similarities between Butterfield's conception of how Whig politicians created a British history that rendered inevitable their political ascendency over the Tories and how the daoxue politicians created a Song history that led inexorably to their own political triumph in 1241. Both are blatantly presentist: historical events are chosen because they contain positive or negative value as guidelines for present or proposed future action. Both rely extensively upon abridgment to foreground these examples and prune away the profusion of unrelated detail. Both generate clear heroes and villains, whose earlier struggles presage the political conflicts of the present. And finally, both create a teleological trajectory of moral rectitude that ensures the ultimate intellectual and political triumph of the writer's own beliefs.

RHETORICAL STRATEGIES OF CHINESE HISTORY WRITING

Western attention to the rhetoric of history, so central to postmodernism, finds a rough corollary in the Chinese adage that “literature and history cannot be separated” (文史不分家). Whatever this nebulous aphorism may mean, Chinese letters has always insisted upon a closer relationship than has been usual in the West between the creation of history and the creation of imaginative literature. Together, the canonical Book of Poetry (Shijing 詩經) and the Book of Documents (Shangshu 尚書) formed the twin staples of an advanced literati education; and the earliest commentators used allegory to align the meaning of the Poetry with the events of the Documents, thus unifying the central moral messages of the canon. In addition to the Book of Documents, the Spring and Autumn Annals (Chunqiu 春秋), a chronicle attributed to Confucius, and its commentaries constituted the canon's second historical work. Two of the five works in the classical canon were thus works of history. Later, the prose of the “three histories”—the Grand Scribe's Records (Shiji 史記) of Sima Qian 司馬遷 (ca. 145–86 BCE), the Han History (Hanshu 漢書) of Ban Gu 班固 (32–92 CE), and the Latter Han History (Hou Hanshu 後漢書) of Fan Ye 范曄 (398–445 CE)—also became standard repositories for the study of rhetorical technique and staples of the Tang and Song examination system.¹²

In Song, literati historians such as Ouyang Xiu 歐陽修 (1007–1072) and Sima Guang 司馬光 (1019–1086), major intellectual figures in the Confucian renaissance of the mid eleventh century, were also daring artistic innovators who brought their prowess as literary stylists to the crafting of historical narrative.¹³ At least three trends, the first new in Song and the second two rejuvenated from earlier models, contributed to this new infusion of rhetorical strategies into the writing of Song history.

First, mistrust of the Han–Tang commentarial tradition gradually freed the text of the classics to receive new interpretations. Mid-eleventh century exegesis of the Spring and Autumn Annals began to draw direct analogies between its Zhou dynasty history and contemporaneous Song dynasty events. The old text was thus made to speak directly to the present, and Spring and Autumn commentary grew to become an important forum for Song political discourse. This spate of Spring and Autumn commentary was unprecedented in Chinese history and continued unabated to the end of the dynasty.

A prominent feature of this commentary was the bifurcation of the text into “events” (shi 事) and “meaning” (yi 義), or more precisely “events” and “their morally incumbent meaning” (yili 義理).¹⁴ In this hermeneutical system, meaning arose from the proper juxtaposition of Zhou dynasty and Song dynasty events. More generally, this emphasis on the moral meaning of history grew along with the subsequent rise of the daoxue movement and found its fullest expression in the twelfth- and thirteenth-century proliferation of the “outline and details” (gangmu 綱目) format for abridging larger historical works.¹⁵

The second trend was the increased use of “precedents” (gushi 故事) among Northern Song literati as rhetorical props in political argumentation. The invocation of precedents for political purposes predated Song. But the Confucian literati revived this rhetorical strategy again in the 1030s in order to engage the monarchy and justify their political program for reform of the Song polity. “Precedents” catalogued actions supposedly undertaken by the first Song emperors, actions the literati claimed should be precedent-setting and thus binding upon their imperial successors. Although the first such collections were privately compiled, one of the earliest being the Records of Sagacious Governance from Three Courts (Sanchao shengzheng lu 三朝聖政錄) by Shi Jie 石介 (1005–1045), these collections soon evolved into a new genre of official historical writing known as “precious instructions” (baoxun 寶訓) or “sagacious governance” (shengzheng 聖政) that assembled precedents from each emperors’ reign as reference points for his successors.¹⁶

A third trend was a rejuvenation of the long-standing practice in Chinese history writing of “using the past to inform the present” (yingu yujin 引古喻今). Using this trope, the historian selects an earlier event as a parallel to inform a more recent circumstance. Variations of this trope are usually summarized with four characters, the second and fourth of which, as in 引古喻今, are usually “antiquity” (gu) and “the present” (jin). The first element is a generic character meaning “to take, cite, invoke” (yi 以, yin 引). The third element, however, crucially determines the tenor of the comparison. Most common are “to criticize” (fei 非), “to compare” (fang 方), and “to enlighten [by means of analogy]” (yu 喻). The earliest formulation occurs in the Grand Scribe's Records where the edict authorizing the burning of books in 213 BCE threatened “execution and exposure of his corpse in the marketplace for any who shall cite in conversation the Poetry or the Documents and extermination of his entire clan for any who shall invoke the past to criticize the present.”¹⁷

Formulations with fang and with yu also occur in contexts that define similar rhetorical strategies in literature.¹⁸ In this sense, the primary meaning of fang is “two boats lashed together” (fang 舫), and the formulation 以古方今 thus implies a relatively equal and neutral pairing of two events, with little hermeneutical or critical implication. Formulations with yu, however, imply not only that the two linked events possess an analogical relationship but that an understanding of that relationship will “inform, enlighten, or instruct” the reader.¹⁹ Yu became therefore the most widespread hermeneutical term for figuration.

Although these three developments belong to the different genres of classical hermeneutics, political policy formation, and history writing, they share a similar rhetorical process: all create temporal analogies between events in the past and those in the present. In the first case, Confucius is held to have infused timeless political lessons (“meaning”) into the text of Spring and Autumn “events.” The commentator then purports to uncover and expound those lessons by forging links between those past Zhou events and analogous contemporaneous situations (“events”). Newly interpreted “meaning” thus arises from this juxtaposition of past and present. The Zhou event is aligned with a Song event; and Confucius's “meaning” is aligned with the commentator's interpretation of that meaning's relevance for the present. The direction or momentum of this analogical process is from past to present: the Zhou dynasty events are pre-selected and textually fixed in the classic, but the commentator selects the Song event with which to pair an earlier Zhou event. In this sense, hermeneutical analogies do not create history; they merely elucidate the moral relevance of that earlier history for Song.

In the second case of political precedents, the momentum of the analogy is bidirectional, and results in the creation of dynastic history. The policy maker selects both the modern event or issue and its earlier precedent. Collections of “precedents,” such as Precedents from [and for] the Age of Great Peace (Taiping gushi 太平故事), submitted by Fu Bi 富弼 (1004–1083) in 1044, thus straddle a murky middle ground between history and policy formation. Political policy formation drives the creation of a dynasty's history, since the resulting group of rhetorical precedents, foregrounding certain facts and not others, created an incipient “history” of the Song. As time moved on and dynastic experience stretched from decades into centuries, ongoing political developments generated the need for new precedents; and, as the number of precedents grew, the distinction between precedent and history blurred even further.

In the third case of using the past to inform the present, the writer combed the historical record and selected events that he posited as parallel to events in his own time. The direction or momentum of the analogy—from the present toward the past—is the reverse of classical hermeneutics: the writer has a fixed opinion on a present issue but is free to select prior events to pair with and thus illustrate his opinion. The result is a genre of historical commentary where the juxtaposition of present and past imposes value judgments on events in both times. The writer then uses the similarities and/or differences in these paired events to express a (usually) critical opinion on contemporary personalities and issues. Often the contemporary half of the parallel is left unstated, but verbal clues alert the careful reader to the unnamed target of the analogy. Usually employed to evade censorship or political persecution, this technique straddles a fine line between subtlety and ambiguity and remains to this day an active rhetorical strategy in Chinese historical and political discourse.²⁰

ALLEGORICAL EXEGESIS IN THE WEST

Because these three strategies create analogies between past and present events, they exhibit some features and concerns common to Western typology and allegory. Although an allegorical mode was already present in Graeco-Roman culture, the dominant strain of Western typology grew from early biblical exegesis. Its practitioners held that God, in creating the universe, controlled history and had guided the creation of the Bible. He had inserted into the Old Testament, and thus into history, prefigurations or “types” (from Greek τύροϛ, an impression made from a seal, a mold; thus a type, a figure) of events that were to later recur in the New Testament. Thus Moses who led the exodus out of Egypt is a type for Christ who leads the faithful to salvation; and Christ's actions fulfill the earlier type. The paradise of Adam is a type for the coming paradise of Christ. Christian typology insists that both the type and its chronologically later “antitype” were real historical events—Adam, Moses, and Christ are not myths. Yet type and antitype are not identical; the former is but an imperfect, unrealized intimation of the latter's fulfillment. Thus, for example, salvation is not a return back to its type. “The essence of typology … is to show how past events are a figure of events to come.”²¹ The movement of historical time is always forward, from creation towards salvation. Far from being a mechanical linking of two events, Christian typology became a fluid and creative technique to represent “historical manifestations of principles set forth in the New Law.”²²

Christian typology is thus primarily a mode of biblical exegesis, since both events are literally present in the same text. Christian allegory is a broader concept that relates to the perception of moral understanding in the text of scripture. Medieval commentators held that there were two levels of scriptural exegesis—the literal, which revealed the “letter,” and the allegorical, which revealed the “spirit.”²³ The first is history; the second is the moral meaning of that history. Or, as one scholar has well explained, for the Christian, history is “purposeful allegory.”²⁴ This Western sense of allegory held sway until well into the Renaissance, but the Reformation and then Romanticism dislodged it as the dominant mode of Western hermeneutics. Northrup Frye's bold declaration in 1957 that “all commentary is allegorical interpretation, an attaching of ideas to the structure of poetic imagery” presaged the return of a secularized allegory as an important component of postmodern theory.²⁵

A helpful point of reference to understand my conception of Song history as grand allegory is Hans Kellner's Language and Historical Representation.²⁶ Kellner accepts the premise that all history writing is ultimately allegorical. The historian's comparisons, interpretations, and generalizations all impose “values and structures” on sources. These resulting allegories are “the natural mirror of ideology.”²⁷ By juxtaposing his evidence and the supposed meaning presumed to be hidden in that evidence the historian creates a “counterdiscourse” with his sources that is dependent both on the sources and the “values and structures” that he proceeds to extract from those sources. “All forms of historical explanation as such thus make use of allegorical devices to mediate between the evidence and the history created from it.”²⁸ Kellner lists “the idea of the renaissance, notion of class, or la longue durée, or anxiety” as examples of “allegorical characters” through which the historian imparts his “values and structures.”²⁹

CYCLICALITY AND SONG TYPOLOGY

Before trying to determine how much of this theoretical posture may help to understand Song historiography, it may be useful to review some concrete examples of Song typology. As prelude however—and it cannot be insisted upon too forcefully—the ontological context differs in at least two fundamental ways. First, as opposed to a monolithic God who created the world and intervenes personally to chart its history, in China “Heaven” neither created the world nor does it direct human affairs. Rather an impersonal Heaven is the ultimate source and model of order, an order first replicated in history by the early sage-rulers whose governance the canon (especially the Documents and the Poetry) chronicles. Second, in contrast to the linear view of history imposed by the Christian narrative of creation and salvation, Heaven's order is cyclical, its ultimate model being the progression of the four seasons through regular cycles of change. Time moves in rhythmic cycles of alternating fluorescence and decay. Those who understand history can gage and anticipate these cycles, enabling them to maximize the periods of fluorescence and minimize those of decay. In this larger sense, the Book of Changes (Yijing 易經), the first of the canonical five classics, is also a book about history: having determined where the observer stands at any moment in a given cycle, history provides examples of earlier responses, both successful and unsuccessful, from those who once stood at the same point in the cycle.³⁰ The ultimate goal of much Song historiography, especially that written in the daoxue vein, is to enable the reader precisely to align political or personal action with the rhythmic fluctuations of history and thus to approximate the orderly governance of the ancient sage-rulers.

This Chinese penchant for cyclicality provides the basis for a distinctive secular typology that pervades all levels of Song historical writing and interpretation. Consider the following passage from the formal congratulatory letter (qi 啟) written by Zhou Bida on the occasion of Wang Che's 汪澈 (1109–1171) appointment as assistant chief councilor (canzheng 參政) in 1162/5:

Rebels extinguished at Ganling,

[Wen] Yanbo enters [the court] to ascend the councilor's seat;

Come home as emissary from Xixia,

[Fan] Zhongyan returns to assist the military bureau.

Thus unplanned do former tidings from the Benevolent Ancestor

Appear again in our era of Continued Restoration.³¹

盜滅甘陵, 彦博入登於宰席.

使來西夏, 仲淹歸賛於樞庭.

不圖仁祖之前聞,

乃至紹興而復見.

Despite the contortions required to render such passages into English, the precise parallelism of the original conceals a process of secular typing that was routine in Song political discourse. Such parallels can also be viewed as a rudimentary form of precedent and were essential to the development of the grand allegory. The Qing historian Zhao Yi 趙翼 (1727–1814) included Zhou's analogy in his collection of parallel prose passages from Southern Song authors who had paired Southern Song and Northern Song events, a rhetorical device Zhao linked to “dynastic precedents” (benchao gushi 本朝故事).³² Such pairings foregrounded and thus assigned historical value to Northern Song history, all the while serving the political interests of their Southern Song authors. Wang Che's appointment followed as a result of his contribution to the successful military campaign to repel the Jin invasion in 1161. Zhou Bida therefore pairs this event with the appointments of Wen Yanbo 文彦博 (1006–1097) as chief councilor (zaixiang 宰相) in 1048 and of Fan Zhongyan 范仲淹 (989–1052) as assistant military affairs commissioner (Shumi fushi 樞密副使) in 1043, because both these appointments also resulted from successful participation in military campaigns.³³

Certainly, these pairings differ from Western typology. There is no sense that a higher force has predestined the two events. The Northern Song appointments do not foreshadow Wang Che's, nor does Wang's precisely “fulfill” the earlier events. Rather, in this example of a secular, cyclical type, Wang Che becomes and then is Wen Yanbo. Zhou thus claims to have noted a cyclical recurrence between his own time and Northern Song. Furthermore, Zhou Bida's pairing functions both politically and historically. Since Zhou and Wang Che were political allies, Zhou's equation of Wang and such illustrious forebears, boosts his political colleague's stature. At the same time, the analogy rehearses and reinforces the historical positions of Wen Yanbo and Fan Zhongyan. More importantly, Zhou proceeds to posit a much grander cyclical concurrence between the Qingli 慶曆 era (1041–1049) of Emperor Renzong 仁宗 (1010–1063; r. 1022–1063) and his present Shaoxing 紹興 era (1131–1162). This claim both ratifies the political values of the Qingli era and asserts that Zhou and Wang espouse a return of those values to their own time. The type also works to criticize the intervening periods in which, by implication, such values did not prevail. Such typing was a routine part of Song political discourse, especially during the Southern Song when daoxue rose to challenge the prevailing political structure.

Such rudimentary precedents could be extended into full-scale policy proposals and submitted to the emperor for consideration. Examples found in the collected works of their authors are often labeled “presented precedents” (jin gushi 進故事), preceded by the date of submission. The genre consisted of two parts: a direct quotation of the precedent text, then a commentary that applied the precedent to a contemporary issue.³⁴ The practice of presenting precedents at the emperor's Classic Mat session began in 1087, and early examples are confined largely to precedents from the Han and Tang dynasties.³⁵ The practice resumed in early Southern Song, and over the course of the twelfth century precedents were drawn increasingly from Northern Song sources, especially the Precious Instructions from the Three Courts (Sanchao baoxun 三朝寶訓), completed in 1032. By the end of Southern Song, authors drew their precedents largely from Song sources, even from the reigns of Emperors Gaozong 高宗 (1107–1187; r.1127–1162) and Xiaozong 孝宗 (1127–1194; r.1162–1189).

Many examples of fully developed, late Song precedents occur in the works of the daoxue official Hong Zikui 洪咨䕫 (1176–1236). For example, following the dissolution of the Jin dynasty in 1234, Song generals leading expeditionary forces in Shandong sent large amounts of plunder and confiscated cash south as personal gifts for Emperor Lizong 理宗 (1205–1264; r. 1224–1264). Hong urged the emperor to publically destroy the gifts as a signal both to the Shandong populace that the court was interested in their loyalty, not their property, and to the generals that imperial favor could not be purchased. In making this argument, he submitted a precedent from 1127/10 in which the young Emperor Gaozong had ordered two bags of pearls dumped into the Bian canal because the pearls had been looted from the former imperial treasury at Kaifeng and donated to him by eunuchs. In doing so, he cited his desire to emulate an ancient sovereign who had reduced thievery by publically destroying his own valuables.³⁶ This 1127 incident had already been included in the earliest layer of official historical documentation on the Gaozong reign, Gaozong's Sagacious Governance (Gaozong shengzheng 高宗聖政), completed in 1166. In that work, the historians had contrasted Gaozong's embrace of frugality, which they attributed to his desire to realize the Restoration, with Emperor Huizong's 徽宗 (1082–1135; r. 1100–1126) love of luxury, which they claimed had brought on the Jurchen invasions.³⁷ Hong's use of this precedent in the 1230s thus reinforced the motif of Gaozong as a founder on par with Taizu 太祖 (927–976; r. 960–976), and whose actions Lizong was therefore required to follow.

Of the many Song works that use the past to criticize the present, the Tang Mirror (Tangjian 唐鑑) by Fan Zuyu 范祖禹 (1041–1098), set an early standard, was widely influential in Southern Song, and survives in several Song imprints. The work illustrates the propensity of the genre to move easily from simple typing toward larger allegorical structures. Fan had worked as Sima Guang's 司馬光 (1019–1086) assistant on the Tang sections of the Comprehensive Mirror that Aids Governance (Zizhi tongjian 資治通鑑). On his own initiative, he selected 332 events, arranged them chronologically, and provided his own comments. In his memorial that accompanied submission of the Tang Mirror to the court in 1086, Fan observed that his work “looks into the past that we may understand the present, reveals what has gone before that we may observe what is to come” 觀古所以知今, 章徃所以察來. The latter phrase paraphrases Confucius's definition of the purpose of the Book of Changes to “illumine the past and interpret the future” 夫易所以彰徃而察來.³⁸ Fan states that offering advice in the form of historical analogies is the most appropriate way for servitors to guide the sovereign toward correct policy choices. He uses Tang history to set forth standard principles of Confucian literati governance: the sovereign should control the palace's irregular exercise of imperial authority, delegate power only to carefully chosen civil officials, accept remonstrance, limit foreign entanglements, and generally exercise benevolent government. The work was clearly directed against the New Policies and was proscribed in 1103 along with other works of the Yuanyou 元祐 period (1086–1093) administrators. Although no serious historian today would consult the Tang Mirror as guide to Tang history, the work had a wide following in Song and was on the curriculum of historical works studied at the emperor's Classics Mat sessions.³⁹

Another example of the genre is Hu Yin's 胡寅 (1098–1156) My Humble Views Upon Reading History (Dushi guanjian 讀史管見), composed between 1150 and Hu's death in 1156. For most of this period, Hu was confined at Xinzhou 新州 for opposing the policies of Qin Gui. Hu held that Sima Guang's Comprehensive Mirror was “long on events but short on moral meaning” 事備而義少, so he spent his days adding his own “meaning” to events culled from the history. Later Song readers perceived immediately that Hu's primary purpose was to condemn an expansion of Song authoritarian government that he believed had begun under Wang Anshi 王安石 (1021–1086) and expanded under Cai Jing 蔡京 (1047–1126) and Qin Gui. Thus, although Hu was commenting on events that had occurred before the Song, he had used these events to “lodge his own ideas” (jiyi 寄意) about his own times.⁴⁰

For example, two late Song collections cite Hu's comment about the Five Dynasties Later Jin councilor Sang Weihan 桑維翰 (898–947). Sang had negotiated an alliance between his own dynasty and the Liao ruler Yelü Deguang 耶律德光 (902–947). Hu commented about this alliance that “although Sang became councilor because of Yelü Deguang, he desired only the prosperity of Later Jin. Never did he embrace the barbarians in order to enhance his own position, nor did he rob his lord in order to steal authority for himself. And so, one may still consider him a worthy.”⁴¹ Readers understood that Hu Yin's intent in this passage had been to criticize Qin Gui, that Hu had in fact created a type, albeit a negative type, between Sang Weihan and Qin Gui. Both had negotiated treaties with the northern states, whose rulers had then pressured the southern ruler to retain them as councilors; but Sang, unlike Qin Gui, had never arrogated authority to himself. By virtue of this comparison, Hu's praise for Sang also implies negative criticism of Qin Gui. Zhu Xi was much taken by Hu Yin's work and took over many of his comments into the Outline and Details from the Comprehensive Mirror that Aids Governance (Zizhi tongjian gangmu 資治通鑑綱目).⁴²

GRAND ALLEGORY AND MASTER NARRATIVE

Many scholars have noted the propensity for the moral categories of Song history writing to become increasingly well-defined and rigid as the dynasty progressed. This propensity closely tracks the growth of daoxue. As the reformulation of moral absolutes that began with the eleventh-century rise of the literati intensified into the twelfth-century daoxue movement, a drive also intensified for cleaner, thus more increasingly allegorized, interpretations of history.⁴³ Although modern scholars correctly recognize a sharp methodological divergence between Sima Guang and later daoxue historians, in truth Sima Guang had already planted the seeds of this later transformation. Not without justification did many Southern Song scholars count Sima Guang as among the progenitors of daoxue.⁴⁴ The Comprehensive Mirror implants a thin chronology of “fact” within a “sublimated moral universe” that originated in the classical canon, especially the Documents and the Spring and Autumn Annals.⁴⁵ Sima Guang expected his readers to apply “values and structures” from this moral universe to craft their own allegorical mediation between the world as it had been and the world as it could have been and might still be. The bulk of the Comprehensive Mirror—its lengthy quotations from memorials and earlier histories—forms a vast sourcebook of discourse on this moral universe to assist the reader in this interpretative task. Once the transition to the daoxue “outline and details” format was complete, the historian no longer asked the reader to craft his allegories for himself. Rather, they came pre-packaged along with requisite “facts”; and the transition to “pedagogical” history was complete.⁴⁶

The grand allegory of Song history is the end result of this process and finds its fullest articulation in the Song History of 1345. At first sight, the work's sprawling length and haphazard editing would seem to gainsay the presence of a controlling narrative. Yet closer reading reveals that the evaluations in the annals sections (zan 贊), the prefaces to the monographs, and the comments in the biographies (lun 論)—a total of 261 explanatory notes all penned by its Yuan compilers—provide the work an interpretive structure and overall vision. Li Huarui ascribes three functions to these comments. The prefaces to the monograph critique specific Song institutions; and those to the collective biographies define the criteria for inclusion in the group. Second, they comment on changes made to drafts of the Song state histories. And third, they present evaluative analyses of the successes and failures of Song governance as lessons for the Yuan court. Over two hundred of these comments remark, usually in moral terms, on specific incidents in individual biographies. The remainder treats larger issues of periodization, institutions, and political trends.⁴⁷ All comments reflect the Yuan compilers’ training and adherence to the Zhu Xi school of late Song daoxue.⁴⁸

Before the thirteenth century, the grand allegory has no definitive formulation, no touchstone passage that contained all its elements. It was rather an assembly of themes, motifs, “types,” and smaller allegories that late Song historians combined to express the “values and structures” that each perceived in their history. The process was well underway when Zhen Dexiu penned his preface to the Chronologically Arranged Complete Essentials in Outline and Details of the August Court (Huangchao biannian gangmu beiyao 皇朝編年綱目備要) by Chen Jun. This work was the first attempt at a comprehensive history of the dynasty from its beginning to the writer's own time, and Zhen Dexiu's preface constitutes the first synopsis of Song history to contain all three major motifs of the grand allegory to be delineated below.⁴⁹

As explained above, I prefer the term “grand allegory” because the Western notion of allegory indeed seems appropriate for how these late Song and Yuan historians accessed their own daoxue convictions to impose structure and meaning on the disparate data of Song history. In addition, Western allegories are often didactic, moral quests undertaken to explore grand existential struggles (think Dante, Spenser, Bunyan) similar to the absolute moral divide between the gentlemen (junzi 君子) and the petty men (xiaoren 小人) of the Song History, if not, of course, Song history itself. However, the distilled essence of the grand allegory might read as follows:

Motivated by concern for the welfare of the people and assisted by Heaven, Emperor Taizu, the Song founder, sought to restore the Benevolent governance of Antiquity. In order to dispel the pernicious effects of long Military rule upon the country, he and his successors fostered Civil governance. For this task, they recruited Gentlemen who supported their vision and propagated governance in the Public interest that considered the needs of all the realm's people. But Petty Men—unruly elements within the monarchy such as imperial affines and eunuchs, recalcitrant military, and apostates among the Gentlemen—motivated by Selfish interests thwarted and deformed the Founders’ model. Henceforth, the Upright and the Perverse contended, their struggles undermined the once strong dynasty, and half its territory was lost to the barbarians. Weakened by decades of suppression, the Gentlemen could only effect a partial Restoration. The Petty Men regrouped under a Nefarious Minister who once again betrayed the Founders’ commitment to Benevolence. Although the Gentlemen managed to resist and regroup, dynastic governance devolved into an unending cyclical struggle between the Upright and the Perverse, between advocates and opponents of the founders’ original conception of Benevolent rule.⁵⁰

Of course, all modern historians of Song will recognize in this morality tale a traditional, if somewhat banal, synopsis of the dynasty's history. Yet, even though many scholars long ago abandoned many of these motifs—for example the slogan that the Song “emphasized the civil and lightened the military” (chongwen jingwu 重文輕武)—if we translate this allegory into the jargon of the modern historian, we obtain a master narrative to which most scholars of Song history could ascribe:

Taizu desired to end of the militarism of the Five Dynasties and unify the country with a centralized, civil administration. He and his successors recruited such officials in large numbers, and they pursued a Confucian-based benevolent governance during the Renzong period. However, after 1067 Emperor Shenzong 神宗 (1048–1085 r. 1067–1085) and Wang Anshi, in order to raise revenue to finance the final conquest of North China, adopted New Policies, which undermined the governance of the founders. The resulting discord led to permanent factional division. The political ascendency of Wang's successors, especially Cai Jing, spurred renewed military expansionism, further weakened the state, and precipitated the Jurchen invasions of 1125. The Restoration under Emperor Gaozong, while rhetorically abjuring the New Policies, continued to employ many of their provisions. Opposition from the heirs of those who had opposed Wang Anshi, plus constant military threats from the north, fostered continued political division. Periods of authoritarian rule under Qin Gui, Han Tuozhou 韓侂冑 (1152–1207), and Shi Miyuan 史彌遠 (1164–1233) alternated with more benevolently inclined administrations favorable to daoxue. After Shi's death in 1233, Emperor Lizong kept the two sides locked in perpetual stalemate.

DECODING THE GRAND ALLEGORY

Although the Song History contains the grand allegory in its most extended form, the Yuan compilers conceived neither the narrative itself nor its components. Rather, they inherited a collection of themes, motifs, and maxims that Song historians had already generated a century earlier to impose structure and value on their own history. In other words, the grand allegory of Song history grew from the unfolding process of Song political history itself. As in all political cultures, Song politicians generated rhetorical strategies, stances, and themes to support their policy goals. Thus the “allegory” of which we speak is not actual history, nor fully identical to any master narrative that a modern historian might devise to explain Song history. Rather, the process and its narrative content resemble to some degree what we moderns might best understand as talking points from a political position paper, or more succinctly and negatively—Song political propaganda.

The allegory can be further analyzed into three large thematic structures or clusters. These are, in the order I will treat them below, “The Empire of Benevolence,” “The Deification of Taizu,” and “The Lineage of Evil.” Each cluster had a life of its own in Song political discourse, and each has numerous sub-divisions and attendant motifs. A brief examination of the historical origin of these clusters reveals that all three, and their ensuing narratives, were products of those who advocated, and to some degree practiced, Confucian literati governance. Each cluster arose at a specific moment, and for a specific purpose, in Song political history, and then maintained a life of its own until the three were combined into a unified narrative in late Song.

The origins of all three clusters can be traced to the early Renzong period when the politics of the Qingli reform movement required a redefinition of the dynasty's early history that would support the movement's political goals. For example, Fu Bi's Precedents for the Era of Great Peace stressed the Confucian virtue of “benevolence” as the basic nature of early Song governance.⁵¹ In 1038, Shi Jie had already equated the first three Song founders with the paramount Sage-emperors Yao and Shun.⁵² And in the 1040s Ouyang Xiu applied the Confucian moral dichotomy of gentlemen versus petty men to the contemporaneous world of Song politics.⁵³

For fifty years after the failure of the Qingli reforms, these three motifs remained random and unconnected fragments of political discourse. The first motif began to gain momentum when critics of the New Policies developed the rhetorical position that the reforms had undermined the “benevolent” character of Renzong's reign. In two memorials submitted in 1092, for example, Fan Zuyu, the author of the Tang Mirror, urged Emperor Zhezong 哲宗 (1077–1100; r.1085–1100) not to revert to the New Policies but to return to the benevolent governance of the Renzong era.⁵⁴ By 1100 this anti-New Policies rhetoric had generated a historical periodization that valorized the Qingli and Yuanyou administrations as kindred periods of benevolent governance.⁵⁵ Subsequently, the image of Renzong's governance rose and fell along with the historical legacy of the Yuanyou regime.

A crucial turning point came in 1125–1127 when the Jin invasions so thoroughly undermined the political viability of the remaining New Policies advocates that Emperors Qinzong 欽宗 (1100–1161 r. 1126–1127) and Gaozong were forced to embrace an anti-New Policies rhetoric as the only viable political base upon which to ground the Restoration. In the twelfth century, the authoritative Long Draft Continuation of the Comprehensive Mirror that Aids Governance, by Li Tao, and the rise of daoxue elevated the idea of the Renzong era as a golden period of literati governance into a dominant theme of Song history. By the end of the dynasty, development of this theme—that Wang Anshi had destroyed the spirit of benevolent literati governance under Renzong—would form a major structural principle of the grand allegory: cyclical revivals of the Renzong spirit had supposedly occurred in the Yuanyou period, in the 1130s under Zhao Ding 趙鼎 (1085–1147), again under Emperor Xiaozong, and finally during the Duanping 端平 administration in 1234–1236. The corollary to this chronology of political virtue was that the intervening periods had somehow failed to practice benevolent governance.

As for the origins of the second major theme of the allegory—the deification of the Song founder Taizu—the historical image of the “founders” (zuzong 祖宗), initially Taizu and his younger brother Taizong 太 宗 (937–997; r.976–997), evolved continually over the course of the dynasty and remains today a contentious subfield within Song studies.⁵⁶ There were, however, two contrasting lines of development. First, as the dynasty progressed, the definition of the “founders” expanded beyond the actual founders, Taizu and Taizong, to encompass, by the end of Northern Song, Emperor Zhenzong 真宗 (968–1022; r .997–1022), Renzong, and often Emperor Yingzong 英宗 (1032–1067; r. 1063–1067)—in short—all emperors before Shenzong.⁵⁷

A similar process occurred in Southern Song. Upon his abdication in 1162, Gaozong was positioned as the founder of the Restoration whose accomplishments equaled those of Taizu. By the thirteenth century, Xiaozong also counted as a Southern Song founder. These extensions of “founder” status provided literati policy makers a greater, more contemporary range of precedents to invoke as “policies of the ancestors” (zuzong zhi fa 祖宗之法) against their political adversaries. For example, Lü Zhong, writing in the 1240s, idolized Xiaozong so he could posit his reign as a golden age before the autocracy of Han Tuozhou. In doing so, he constructs a political trajectory for the Restoration that parallels that for Northern Song, equating Han Touzhou and Shi Miyuan with Wang Anshi as antagonistic to benevolence and to the policies of the founders.⁵⁸

At the same time, Taizu alone was gradually elevated to a unique status as the ultimate founder, the primal source from whom Song governance had flowed. Initially, attempts to camouflage the conflicts between the two brothers obfuscated the historical record of early Song. In addition, the founders, up through Zhenzong, had employed a largely Daoist rhetoric to legitimate their rule. The first formal history of the dynasty, the State History of the Three Courts (Sanchao guoshi 三朝國史) completed in 1030, was thus long on Daoist hagiography and vague on what Taizu and has actually done. Only in the late 1030s did the first wave of Confucian literati attempt to furnish Taizu with a “benevolent,” Confucian pedigree. Accordingly, they developed accounts that stressed his supposed preference for civil over military solutions, both on the battlefield and in general administration. For example, Sima Guang's narrative of the founding at Chen Bridge stresses Taizu's reluctance to become emperor: he agrees to do so only after extracting from his soldiers a pledge to refrain from looting and reprisals against the Zhou royal family.⁵⁹

This narrative of a “bloodless founding” formed the basis of a developing “type” that would link Taizu and the sage-emperor Yao: without coercion the people had flocked to both. Other narratives were developed to posit the evolving Renzong era tenets of literati governance back to Taizu: he had suppressed the military and reigned in excessive affines and eunuchs. Fan Zuyu, in the preface to his 1093 work, The Benevolent Emperor's Instructions and Statutes (Renhuang xundian 仁皇訓典), linked together Heaven, Yao, Taizu, and benevolent governance. Fan stressed that Taizu, like Yao, had bypassed his own son to transmit his authority to a Sage, in Yao's case, Shun, and in Taizu's case, Taizong.⁶⁰ The repression of the Yuanyou party, however, curtailed the development of these associations for the remainder of Northern Song.

Once again, the early 1130s witnessed a remarkable reworking of the dynasty's historical narrative. As is well known, Emperor Gaozong, perhaps because of his own inability after 1129 to sire children, decided in 1131 to return the throne to the Taizu line. The move not only accorded with his earlier dynastic embrace of Qingli-Yuanyou rhetoric but also enabled Gaozong to posit himself, like Taizu, as Yao, who would then pass the realm to a Sage rather than to his own son, thus enhancing his own stature as founder of the Restoration. Major motifs in the Taizu cluster—that he had won the empire by acclamation and without military struggle at Chen Bridge and that he had “taken back military power over a cup of wine”—all assume their final textual form during the mid-1130s at the hands of Gaozong and Fan Zuyu's son, Fan Chong. Such narratives were then used to justify the 1142 peace treaty with the Jin. When Gaozong visited the Imperial University in 1143, the students proclaimed that “Your Majesty has now hushed the military to promote our culture (yanwu xiuwen偃武修文) and matched the age when Taizu first settled the world.”⁶¹ Although the daoxue historians beginning with Zhu Xi worked quietly to remove Gaozong from this equation, the link between Taizu and Yao was strengthened, and Taizu became invested with all the qualities of a daoxue Sage.⁶²

Reflecting this paramount status of Taizu, the Song History editors have inserted a collection of fifteen anecdotes about him between the end of his formal annals (太祖本紀) and their evaluation of his reign (zan), a unique feature they accord no other emperor.⁶³ Among them is the story of how Taizu ordered the doors and gates of his palace realigned to permit an unobstructed view into his audience chamber, “for just so is my [open] mind/heart in which all may see even the slightest deviation” (此如我心, 少有邪曲, 人皆見之). A textual analysis of these anecdotes reveals that, although several probably derive from the first state history of 1030, most first appear only later in the Renzong era baoxun and shengzheng collections. Thus, they reflect literati attempts to cloak the founder in a Confucian aura, and Sima Guang included many of these stories in his Su River Records (Sushui jiwen 涑水記聞).⁶⁴ In the Long Draft, Li Tao treats this material with great caution, often pointing out inconsistencies that call specific details and events into question.⁶⁵ Taken together, these fifteen anecdotes reflect a thirteenth century image of Taizu as a daoxue sage-ruler. Lü Zhong, for example, cites the palace renovation anecdote to demonstrate that Taizu inherited the mind of Yao and Shun. His commentary, which cites Zhu Xi, repeats the daoxue precept that all governance emanates from the mind of the ruler and good governance ensues when the ruler emulates the mind of the sage-kings of Antiquity.⁶⁶

The political concept known as the “policies of the ancestors” (祖宗之法) also comprises a vital component of the second cluster. Deng Xiaonan's detailed study has revealed their centrality to the development and practice of Confucian literati governance.⁶⁷ Early formulations, which date from the 1030s and mature during the Qingli period, attribute to the founders (Taizu, Taizong, and sometimes Zhenzong) a set of generalized principles, as manifested through their specific actions, that the writers frame as fundamental policies of dynastic governance. Politicians who reference the policies often emphasize that because these policies originated from the founders, succeeding emperors are required to follow them. The writers thus attempt to create a body of precedent whose validity supersedes any contemporaneous authority, even that of the reigning emperor. Loyalty becomes loyalty to a set of universally valid dynastic principles for good governance rather than personal loyalty to a monarch.

Yet ongoing political imperatives, rather than any historical sense of what the early emperors had actually done, drove the formulation of this canon of precedents. For example, the initial iteration of the policies created a source of higher authority to which the Qingli reformers could appeal in their attempt to undermine the existing political power structure that had solidified under Empress Liu 劉皇后 (969–1033). Thus, the actual content and interpretation of the canon of “ancestor” precedents evolved with every new political struggle. What did not change was the authority of the founders as dynastic exemplars. Wang Anshi's supposed attempt to diminish their authority was counted among his most heinous crimes. After the 1127 Restoration, the stature of the founders, especially Taizu, increased to such a degree that to diminish his authority was to remove the legitimacy, even the possibility, of effective advocacy for political change.

The third cluster arose from motifs that relate to the political dynamic between gentlemen and petty men. This is the most complex of the three clusters, but also perhaps the most historically robust. Unfortunately, this cluster presents the greatest problem for modern historians, who tend either to embrace its terms uncritically, and often unconsciously, or to ignore them totally. However, the ubiquity of this dichotomy in Song political discourse after the Qingli period became a fundamental dynamic of Song political culture. Later, in the hands of the daoxue historians, it became the fundamental driver of the grand allegory. A complicating factor in decoding the role of this language in Song political discourse is that while junzi always refers univocally and positively to the writer, his associates, and their policies, xiaoren can refer to a broader range of political actors whom the writer opposes. First, within the context of Northern Song factional politics, xiaoren refers to one's political opponents within the literati class. For example, as the debates over the New Policies unfolded, both sides used this language to distinguish themselves from their opponents. Thus Sima Guang defined junzi as motivated by “public good” (gong 公) and xiaoren by “private interests” (si 私), with the emperor being responsible for discerning one from the other.⁶⁸ But xiaoren also occurs often in literati discourse to refer generically to non-literati actors such as eunuchs, clerks, imperial affines, and female bureaucrats within the monarchy. The assumption that these latter non-literati groups will always act to bolster their own “private” interest links the literati and non-literati frames of reference for xiaoren.

During the Southern Song, a variation of the junzi/xiaoren dichotomy developed as a major motif within this third cluster, and I term this motif “the lineage of evil.” This motif contends that after (and because of) Wang Anshi, a series of powerful, literati xiaoren, termed “nefarious ministers” (jianchen姦臣), succeeded in dominating Song governance for extended periods. Four Song History chapters gather together their biographies. Its preface frames Song history as a cosmic struggle between junzi and coalitions of xiaoren led by these nefarious ministers. The first chapter contains biographies of the post-Wang Anshi purveyors of the New Policies: Cai Que 蔡確 (1037–1093), Lü Huiqing 呂惠卿 (1032–1111), Zhang Dun 章惇 (1035–1105), and Zeng Bu 曾布 (1036–1107); the second is devoted to Cai Jing; the third to Qin Gui, and the fourth to Han Tuozhou and Jia Sidao 賈似道 (1213–1275).⁶⁹ Since all of these save Han and Jia possessed sterling credentials as literati officials, their biographies describe them as apostates to a concept of governance based on Confucian moral values. In this view, the nefarious ministers pursued their own “private interests” and so forged political coalitions with the non-literati xiaoren.⁷⁰

The lineage of evil was the last motif in the three major clusters to evolve from Song political discourse. It required not only a negative historical assessment of Qin Gui, which Zhu Xi had begun to formulate as early as the 1160s, but also of Han Tuozhou. The latter's assassination in 1207 sparked immediate comparisons that linked him to Qin Gui. But, since Qin and Han had pursued opposite policies towards the Jurchen, the link was based upon the contention that both had pursued analogous styles of “private” governance. The analogy also sharpened the negative character of Qin Gui and his administration: in the face of strong opposition, both Qin and Han had relied extensively on coalitions of non-literati actors to form their governing coalitions. This vilification of Han served as political rhetoric to prepare the way for an anticipated return to a better governance model under Shi Miyuan. When Shi failed to deliver on that promise, his opposition cast him as the most recent addition to the newly conceived lineage of nefarious ministers. After Shi Miyuan deposed the rightful heir and imposed Emperor Lizong on the throne in 1224, this conception of a lineage of evil ministers quickly morphed into a formidable political and historiographical weapon. In politics, late Song daoxue adherents employed the motif to devastating political effect, toppling the chief councilors Shi Songzhi 史嵩之 (1189–1257) in 1245 and Ding Daquan 丁大全 (d. 1263) in 1259.⁷¹

CONCLUSION

Hayden White has reportedly “denied the charge of relativism, averring that the reality of events in the past is not contradicted by literary portrayals of those events.”⁷² Indeed, the existence and decoding of a grand allegory of Song history does not negate the reality of individual events in the Song historical record, merely the arrangement of those events into subjective narratives. As we have seen above, certain periods in Song history spurred the creation of these narratives, namely the Qingli and Yuanyou periods in Northern Song, and the early (pre-Qin Gui) Shaoxing and the Jiading 嘉定 (1208–1224) periods in Southern Song. All of these eras experienced either defensive or offensive wars that sparked domestic political upheavals, and these political conflicts then generated historiographical revisionism. The result was the emergence of a narrative of positive political value that moved from Taizu through Qingli and Yuanyou. This narrative of supposed Northern Song values and structures took shape under the descendants of the Yuanyou administrators in the mid-1130s, was then held in check during the Qin Gui years from 1138 through 1155, but took definitive form in Li Tao's Long Draft, composed over a forty year period between the 1140s and 1183. In essence, our received narrative of Northern Song history grew from the political struggles of the Restoration. Later Southern Song historians then wrote the Restoration's own history by plotting its events back onto the earlier model their predecessors had created for the Northern Song.

Within a temporal worldview rooted in cyclicality, the most distinctive Song element affecting the structure of the grand allegory was the expression of Confucian political values through precedents. Although there is nothing uniquely Confucian about precedents, the emergence of a robust Confucian renewal in the Qingli period profoundly shaped subsequent Song historiography. As we have seen, precedents are bidirectional (the writer chooses both type and antitype). Thus, as Southern Song politicians increasingly cited dynastic rather than pre-dynastic events as precedents, the narrative structures of Northern and Southern Song became increasingly aligned. And, as the dynasty progressed and the influence of daoxue grew, the grand allegory of Song history came increasingly to reflect Confucian values—benevolent governance, the primacy of Yao/Taizu/Gaozong as founders of that governance, and the political supremacy of gentlemen (junzi) as paramount administrators.

One may ask how this allegorical structure for Song history might apply to other dynasties. Certainly, even for Song, the model could be refined: other clusters could be developed and additional sub-motifs identified. As for other dynasties, the cluster relating to the primacy of the founders is probably universal, as research on the image of the Ming founder suggests.⁷³ However, the decidedly Confucian tenor of the first and third clusters reflects a Confucian intensity that marks the eleventh century as a turning point in Chinese historiography. If Ouyang Xiu's revisions to the existing histories of the Tang and Five Dynasties signaled his frustration with the moral ambiguity of earlier history, then his fellow Song historians abided no similar frustrations when they wrote the history of their own dynasty.