Flows of time in the centres and peripheries of Tang experience

Oliver Moore 本野英一

In this article I discuss asynchronous time flows in two stories translated from Lu Zhao’s 盧肇 History of Things Outside the Norm (Yishi 逸史, hereafter “Lu’s History”), a ninth-century collection of accounts of supernatural marvels. Like others in the same collection, both stories share their origins with verifiable biographies and historical events. The stories provide insights into the ritual procedures adopted during tomb thefts and into the magical lore surrounding tomb models (mingqi 冥器). The stories also describe the antagonism between agents of discipline in the city-based administration and disruptive forces in the extramural spaces outside. The tension in these accounts reveals a strong editorial fascination with the asynchronous flows of time, which I discuss in their relationship to late-Tang thinking about the qualities of temporal order and their insertion into the appropriate category of historical account. Beyond treating stories in Lu’s History as tales of the supernatural, then, my argument is first to read them as a category of historical writing that permitted their compiler to rehearse diverse time flows as forces contributing to subjects’ interpretations of experience. Second, I propose that this view of time gained currency and authenticity through conscious attempts to validate late-Tang compiling activities in relationship to both central and peripheral concerns of Tang historical writing.

Asynchronous Time and the Decentralised Tang World

In his last unfinished work, Siegfried Kracauer recorded several perceptions of time, which are extremely useful to historians practicing in the wake of a strong turn towards anthropology:

Since simultaneous events are more often than not intrinsically asynchronous, it makes no sense indeed to conceive of the historical process as a homogeneous flow. The image of that flow only veils the divergent times in which substantial sequences of historical events materialize. In referring to history, one should speak of the march of times rather than the “March of Time.” Far from marching, calendric time is an empty vessel. Much as the concept of it is indispensable for science, it does not apply to human affairs.[1]

Framed in these generic terms, Kracauer’s insights invite the possibility to theorise the nature of social formations whose space and time never comprised the primary object of his interests. Tang China is a case in point, since “divergent times” characterise some important recollections of the late Tang world. Divergence arises palpably in Tang perceptions of the passage of time in the period’s many social engagements. To consider with a modern historical imagination, then, how time may “apply to human affairs” offers a binary approach to specific recorded events, considered both as happenings that impacted forcefully within their social context and also as objects that exercised an intellectual fascination for the historians recollecting and preserving them.

The context of Tang events in the following discussion invariably comprises provincial spaces outside the central seats of government power. Attention to these areas has grown with interest in Tang people’s experiences not only in large urban centres, notably Chang’an 長安 and Luoyang 洛陽, but also at a dirstance from them. Denis Twitchett’s discussion of new social conditions and attitudes after the mid-eighth century pioneered Western scholars’ realisation of how the constituency of politics changed dramatically over the first and second halves of the Tang period, and he sensitised modern awareness to the greater breadth and variety of social participation in late-Tang affairs.[2] This analysis has gained extra definition with more recent research into Tang experiences of urban space and its changing uses.[3]

Although much less has been said about life outside cities, Tang reactions to life in the provinces are not entirely without trace, certainly not during the dynasty’s late period when writers commented on this dimension of existence much more readily. The apparent paradox is that these multiplying reactions to life outside Tang city walls accompanied ascending rates of urbanisation in many regions of the empire. Otagi Hajime has drawn from an impressive range of sources—including the new wealth of epigraphical discoveries—to show an acute rate of urban expansion in 30 case studies of the ninth and tenth centuries.[4] Late-Tang accounts reflect these conditions closely, since their authors are invariably members of a literate group who perceive their surroundings from the vantage point of town- or city-based commands. Even Tang accounts that detail events from the relatively prosperous decades of the early eighth century are more often than not stories retold by a later generation of writers who experienced an age when centralised imperial government was deteriorating, and when provincial sub-centres had experienced unprecedented commercial growth and assumed new military significance.

The intellectual mirror to these conditions is the deep fascination that late-Tang writers of minor historical accounts displayed towards human actors locked into struggles with time as an adverse agency. Their conception of time as a force that cannot be easily contested amounts also to an admission of their unease that one set of rules, one History Office, one broadly co-ordinated bureaucratic effort could account for the distance between centre and periphery—the divergence arising even in shared experiences, not to mention the strongly resistant temporalities of communities beyond Tang government’s regulative means of sight and control. Even when late-Tang writers preserved recollections of the early eighth century, they realised, sometimes painfully, that human action would never again be subject to familiar ambitions of political and cosmic unity, and that time was not so easily conceived as an agent of central control.

Late-Tang compilers may have reported on a more fragmented world with enthusiasm perhaps because they accepted what Robert Campany has described for an earlier period as a dialectical structure of centre and periphery. This description is still relevant to Tang conditions, insofar as Tang compilation was also the pursuit of a centrist cosmographic discourse—metaphorically a kind of data collection that helps to make sense of the world and to control it.[5] Yet, even if this centrist ideal endured throughout the Tang, the period’s most ambitious attempts to assert order from the centre outwards never assumed the creation of an absolute and shared continuum of time. The highest government control effected only what Marcel Granet famously called time’s “liturgical usage”,[6] which involved the categorisation of time into various shapes, or what Joseph Needham, following Granet, envisaged as time bundled into different forms of “packaging”.[7] The shapes spring into discernible forms—governed by space and ritual—as soon as we read the prescriptive ambitions of either the Kaiyuan Ritual Code (Kaiyuan li 開元禮), completed in 732, or surviving Tang manuals of Daoist liturgy, just two collective expressions from interlocking zones of the Tang intellectual outlook.[8] The men who so successfully established the final version of the Kaiyuan Ritual Code, for instance, did so after successive debates, during which they accepted or rejected what this monumental text should include. Given that this editorial experience was not abnormal, any assumption that Tang officials believed strongly that only one usage of time ordered the world is simply counterintuitive.

While codes, manuals and other philosophical expressions represented ideal packages that were not totally inclusive, whatever was excluded from these formations was not necessarily also suppressed. Its endurance, then, allowed the emergence of divergent temporalities, and against this background I analyse a group of late-Tang stories and their editor’s particular interest in time’s asynchronous flows. His records of time as psychological and historical experiences are drawn from both leading figures of the period and other informants of considerably lower social status. These form widespread contacts in a larger expanse of shared ideas that their recorder embedded in a category of primary source that remains today historiographically and ethnographically underemployed.

The first section of this article introduces Lu Zhao and his text. The next section translates and discusses two stories. Conflicting perceptions of time, whose apparent sovereign subjects are the senior- and middle-ranking officers of Tang government, can justify reading these stories as records of magical events and supernatural productions. The focus, however, is not primarily on magic and the supernatural, but rather on what their respective techniques and forces affected in the recorded experience of asynchronous time. The subsequent section sets descriptions of asynchronous time flows within the context of Lu Zhao’s editorial interests, and it proposes that this asynchronous quality comprised the primary motive to compile a text that contains so many historical examples of temporal divergence. In the final section discussing central and peripheral concerns in Tang historical writing, I look at how Lu Zhao’s editing reflects a general shift from the deep tradition of recording marvels, anomalies and mirabilia towards the creation of records that their compilers viewed in a growing relationship to the writing of history. Although the philosophical causes may be indirect, historiographical precepts first formulated in the early eighth century by Liu Zhiji 劉知幾 (661–721), Tang China’s most practical and eloquent philosopher on questions of historical writing, mark a founding moment in later generations’ changing evaluations of minor categories of record. Over the second half of the dynasty these changes enfranchised Tang compilers to collect data from both central and peripheral contexts, and to make sense of them in the corresponding antitheses of asynchronous time flows.

Lu Zhao and the History of Things Outside the Norm

The stories discussed below owe their earliest survival to inclusion in an encyclopedic compilation, The Taiping Reign Period’s Extensive Records (Taiping guangji 太平廣記) (hereafter the Extensive Records), named in honour of imperial time—by now a Song unity—during the progress of its commission in 977 and its completion one year later. The project filled 500 chapters (juan 卷]), and it was printed in 981.[9] The aim of this ambitious act of bibliography was to edit and preserve a large body of literature, which was seldom duplicated in extant official sources. Many of these texts were the content of private, semi- or unofficial historical record. Each text is usually cited from an original source, which often no longer exists independently and integrally. Some of the material dated to several centuries earlier; a few citations are from works compiled within living memory during the decades just before and after the Song reunification in 960. The fact that Song editors saw fit to draw from this material in order to create an editorial whole is a good indication of their own and previous generations’ comprehension of alternative constructions of time and its inscription in examples of non-canonical historiography.

The Extensive Records contains 78 stories attributed to Lu’s History. One instance of duplication reduces this total to 77. Another important Song source, the major Daoist anthology The Cloud Bookcase in its Seven Labels (Yunji qiqian 雲笈七籤) compiled by Zhang Junfang 張君房 (jinshi 進士 1004–1008), includes fourteen stories without any attribution. Long supposed to be items from Du Guangting’s 杜光庭 (850–933) Biographies Featuring Contacts and Encounters with Gods and Spirits (Shenxian ganyu zhuan 神仙感遇傳), from which Zhang also borrowed material, the true source of these stories has been recognised only recently.[10] The content and structure of thirteen of the stories match the same number in the Taiping Extensive Records. The fourteenth is not found in the Taiping Extensive Records, but its attribution to Lu’s History is documented in yet another Song collection of literature, Categorised Stories (Leishuo 類說).[11] 78 stories remains the total number of stories preserved in these three Song sources.

Although Lu’s History is not clearly attributed to a compiler in the New History of the Tang Dynasty monograph on literature, which was edited by eleventh-century bibliographers, it is listed directly after another work entitled Historical Records (Shilu 史錄) and attributed to Lu Zhao 盧肇.[12] This other work has vanished. Ye Mengde 葉夢得 (1077–1148), in notes on his own reading, is the earliest to confirm independently that Lu Zhao compiled the History of Things Outside the Norm.[13] Not many facts concerning Lu Zhao’s life survive. He was an outstanding jinshi degree-winner—first in his cohort in 843—having entered the competition in the prefectural selections at Yuanzhou 袁州 (modern Yichun, Jiangxi province).[14] During this period, or perhaps earlier, he attracted the patronage of Li Deyu 李德裕 (787–850), the most formidable figure of mid-century politics and an intellectually engaged Daoist.[15] Lu Zhao’s History shows that he shared many interests with his contemporary Duan Chengshi (d. 863), editor of the Youyang Miscellany (Youyang zazu 酉陽雜俎), one of the most famous late Tang collections of ancient and contemporary accounts of magic and science. Duan had worked under Li Deyu during his posting to Runzhou, and he was living in Chang’an when Lu Zhao arrived to take part in the state examinations in 842 or earlier. He was also prefect of Jizhou 吉州 (modern Ji’an, Jiangxi province) during 847–853, not long before Lu Zhao’s own tenure of the same office.

Lu Zhao never rose above a series of southeastern prefectural appointments, which sent him from his first and most distinguished posting in Shezhou 歙州 (modern Shexian, Anhui province) eventually to the relative southern obscurity of Jizhou where he probably died.[16] Like Li Deyu, he was famed for his rhapsody (fu 賦) compositions—several of his works in this genre are listed individually in the New History of the Tang Dynasty monograph on literature.[17] This body of work—besides letters and other writings not properly integrated again until 1160[18]—represented paradigmatic styles in the lyrical and discursive expectations of mid-ninth century examinations.[19]

Lu Zhao’s stories in the History mix official life, real events and strange happenings. His relationship with Li Deyu may have steered his work towards a common interest, since, despite complexities surrounding the attribution of story collections to Li Deyu, this category of literature certainly figured among the latter’s interests.[20] Lu Zhao and Duan Chengshi acknowledged no debt to each other, but they collected material in the same areas of China, particularly in and around Chang’an and Luoyang, the central Yangzi region, Yangzhou 揚州—where Duan Chengshi’s father Duan Wenchang 文昌 had once served as governor—and other southeastern command centres. For instance, both men had heard of the same celebrated seer called Wang 王 whom elite society in the Yangzhou area often consulted during the governorship of Li Deyu’s father Li Jifu 李吉甫 (758–814).[21] Both recorded stories of curing a sick crane encountered near Luoyang.[22] Both adopted some of the same idiosyncrasies of expression.[23]

Lu Zhao’s responsibility for the Historical Records, another integral compilation, reveals that he was engaged with historical editing on a broad front. Moreover, in his preface to the History, he linked his editorial aims prominently with his work on the Historical Records. Fortunately, the preface to Lu’s History survives, because it caught the attention of Tao Zongyi 陶宗儀, one the fourteenth century’s most expert collectors of old texts:

Once Master Lu had finished his Historical Records he collected the marvels of what he had seen and heard and titled them History of Things Outside the Norm. Whenever he touched on unions between gods and immortals, resonances from the other world, preordinations of [men’s] rise and fall, prophecies of disaster and fortune, he did no more than retrieve what was authentic and restore what was missing. He chronicled a total of 45 items, all of them affairs of our Tang rule. Written this eighth month in the first year of Dazhong [847].[24]

Forty-five items are considerably fewer than the 78 passages preserved in prestigious Song sources. The editors of the Taiping Extensive Records may have attributed material that did not belong with the original collection, but the linguistic homogeneity that the larger total of stories often share should discount errors of attribution. The discrepancy may also be due to an erroneous transmission of a statement giving “115”.[25] Finally, however, the work’s original size must remain a mystery.

The date 847 provides a credible term for Lu Zhao’s work on his collection, which no surviving story contradicts. His reference to an earlier literary commitment is highly significant for its assumption that in 847 readers already knew of his work on the Historical Records, and no less for its implication that he won some reputation from that work’s earlier circulation. These clues to his literary career comprise an unusually coherent claim to have been multiply productive across the various categories of minor individual historical writing. They suggest also how his production of Lu’s History was as much as anything an historiographical undertaking, a point to return to in the final section.

Surviving stories attributed to the History show that Lu Zhao was fascinated by the manifestations and lore of immortality, asynchronous time, parallel worlds and reports of hermits and other recluses. His accounts of these subjects are set invariably within real events of recent history. Only a few involve prominent figures, one or two of whom Lu Zhao may have met and known. The stories are often reports from religious cult centres in the mountainous areas of Hengshan 衡山, Huashan 華山, Kuaiji 會稽, Maoshan 茅山, Qingcheng 衡山, Qingxi 清溪 and Zhongnan 終南. One striking narrative feature is Lu Zhao’s attention to local surroundings. He reports events happening at particular gateways, streets, bridges, fords, waterways and mountain passes. Appropriately enough, these precise locations are often the settings for low-ranking officials, soldiers, physicians and commoners in some of Tang life’s more trivial engagements.

Stories in Lu’s History often develop the theme of asynchronous time in accounts of the world of spirits and its interactions with official life. Six stories recount respectively the success of men who renounced the opportunity to become immortal in return for a smooth ascent to high official rank, or else the failure of others to capitalise on an offer to achieve both aims together.[26] In one story a seaborne merchant is blown onto islands off the Zhejiang coast where he encounters a cult dedicated to Bai Juyi 白居易 (772–846) and maintained by immortal beings.[27] In another, the minister Zheng Juzhong 鄭居中 (fl. 820s), whose friends include a number of Daoist practitioners in the Hengshan area, proceeds to Songshan to become an immortal.[28] The larger significance concerning content of this nature is that it locates Lu’s History within a tradition already centuries’ old of recording individual quests to achieve immortality. At the same time, however, long-established immortal lore and its familiar narrative structures are combined with the credible facts of a senior politician’s social life and intellectual interests. More clear evidence of Lu Zhao’s debt to established literary tradition is his frequent use of tropes from Tao Qian’s 陶潛 (365–427), Tale of the Peach Blossom Spring (Taohua yuan ji 桃花源記), the classic asynchronous encounter between a fisherman of the Jin 晉 period and refugees from the Qin–Han transition. Four stories in Lu’s History feature discoveries of spirit worlds with separate chronicities. Respectively, these worlds are discovered by chance in remote uplands; once the discoverer has walked through a cave, before it becomes apparent that any re-entry is impossible; and, once their inhabitants have asked what age they are living in.[29]

Several stories just mentioned deride the worldly ambitions of scholars who nevertheless pretend to study Daoist arts, and others scorn sentimental attachments in the face of offers to begin an immortal existence. Some professed Daoists are simply frauds.[30] One story features a discreet and enlightened servant working for a charlatan physician on Maoshan.[31] Crassly naive are those who fail to grasp the deeper rewards of secret and patient contact with Daoist power—two stories feature women who berate their spouses for overindulging scruffy Daoist adepts.[32] Or, despite her son’s achievements in longevity techniques, a mother cajoles him to enter the state examinations.[33] Stories concerning experts who can predict future outcomes form the largest category of interest in the whole collection.[34] Particularly interesting are the stories in which asynchronous flows of time are confirmed by objects that magically transfer between different worlds and then sometimes back again.[35] Writing is another powerful medium confirming events from asynchronous vantage points.[36] So too are music and the board game weiqi 圍棋.[37] Stories of human efforts to stop the divergence of immortal time and the span of human existence feature popular ideas concerning eating and drinking, fasting, breathing and levitation.[38] In such contexts adepts are duped by family members into destroying their chances of becoming immortal after eating human food and thus breaking a cardinal taboo for immortal aspirants. One excellent example features a girl who abandons her strict diet and crashes earthwards from her long-practiced poise in midair.[39]

Tomb Robbing and Asynchronous Events

It is not necessary to detail every variation on the theme of asynchronous time that controls the narrative of each of the stories in Lu’s History. Rather, in this section, I present translations and readings of two stories that set wonders of temporal and ontological shifting within urgent concerns to guarantee the security of tombs. Tombs provide an excellent context in which to explore the theme of time, since tombs comprised not only spatial, symbolic and economic facts of Tang existence, but also represented the most elaborate alterity to everyday thought and action. Their enclosures provided space to enact and inhabit another world, and yet they belonged still to the world that constructed them. Lu Zhao is no exception among the many Tang people who speculated intensely on existence after death, and who sought explanations for spiritual communication between living and dead in the way that time allowed passage between their respective worlds.

Lu Zhao’s attention was drawn to qualities of time contested between the worlds of life and death, most especially when the intervention of robber gangs and law forces intensified the struggle. During his lifetime, the sanctity of tombs remained ritual and juridical, but by the 840s it had also formed a pressing political issue. The idea that robbing from the dead heralded the collapse of a known order had a long history,[40] and late-Tang statesmen were acutely aware that the stability of their state was manifest in its officers’ ability to prevent tomb looting. In fact, in 842 senior officials of the government—including now Lu Zhao’s patron Li Deyu—debated the recent increase in all categories of theft, and concluded that provincial officials needed to apply legal sanctions more assertively. Clearly, not only tomb robbery, but also collusion among officals was slipping out of control.[41] But, this official reflection of a broader social phenomenon, while historically relevant, was a secondary concern beside Lu Zhao’s perception of tomb robbers and law enforcers as the subjects of wholly asynchronous cycles.

In the two stories translated below, Tang administrators try to impose order in circumstances that feature ghosts of the dead and tomb robbers. The stories involve intense struggle, flowing, in one case, through ritual performance, and, in the other, through material agents. The progress of time in these stories is distributed between the central theatre of the sub-prefectural audience hall—effectively the courtroom—and the comparatively unruly spaces of town markets and open lands dedicated to burial.

Ritual Performance

Lu Zhao may have often enjoyed direct access to court proceedings, and obviously he exchanged experiences and gossip with those in stations of legal authority. Set in late eighth-century China, the following story provides valuable glimpses into the Tang tomb-robbing business and the military policing that tried to suppress it:

Fan Ze 樊澤 was the governor of Xiangyang 襄陽 [also Xiangzhou], and one of his security staff was a man named Zhang 張. Zhang’s father, who had been military commissioner of the Yong 邕 frontier [Yongzhou], was buried a good few li north of Dengzhou 鄧州.

The site of Xiangyang lies today beneath the city of Xiangfan in northern Hubei. Zhang’s father, who had served in a zone centred on Nanning in Guangxi province, was buried farther north in the area of modern Dengzhou in Henan province. At present nothing else is known about him. Fan Ze governed at Xiangyang from 784 to 787,[42] and once more from 792 until his death in post in 798.[43] Appointment as prefect here was usually combined with the governorship of the Eastern Shannan 山南 circuit, an area nearly twice the size of Belgium. According to his official biographers, Fan was a gifted commander, broadly read in military canons, and popular among his peers as a strong rider and huntsman. Already famous as the scourge of criminal gangs during his first governorship of Shannan, he was reappointed in 792 with a specific mission to reassert control over a total breakdown of military discipline that had led to widespread looting.[44] Perhaps this story of tomb robbing belongs to that period:

There were three Zhang brothers. Now, at precisely the same moment they had a dream in which their father said: “The tomb where I am buried is to be robbed this night, and the robbers will bring my clothes and effects into the city walls today. They will stop among the makers of mats and headgear. Your duty is to go fast and catch them. Once the sun is up you won’t succeed.”

Aside from introducing the personalities and their locations, the account so far recounts a prediction in the form of a dream. The dream is the device for crossing space and time, and eventually also the means to enable a forensic advantage to arrest and sentence lawbreakers. In another story attributed to Lu’s History exactly the same sequence occurs when a man, who has been robbed and murdered, appears in the dream of his relative, the military commander at Goushi 緱氏 sub-prefecture (near Luoyang), and predicts where to arrest the perpetrator of the crime.[45] The Zhang brothers’ dream and the events that follow are not easily categorised as fiction, for details of the account are not controlled with the kind of skill that would be commensurate with story plot. Since the prediction that the robbers will stop among the makers of mats and headgear is neither fulfilled nor further developed, it smacks of report rather than invention. A highly plausible report, in fact, since Xiangyang’s location in central China during this period places it at the heart of longstanding regional industries that produced its famous lacquer and bamboo products.[46] The members of the robber gang—men and women who are otherwise barely discernible—can be fitted into a verifiable social reality. If the robbers were also casual participants in the production of mats and headgear, robbing was, perhaps, an essential diversification when their craft production was insufficiently remunerative.

The Zhang brothers got up that night, and, moved to tears, told each other what had happened. Before dawn, having banged on the gates of the prefecture, they were admitted to Fan Ze’s presence where they explained the whole situation. Fan Ze immediately summoned the Chief of Police and ordered an arrest. Six members of the gang as well as the wife of the gang-chief were all captured. Fan Ze had them summoned and questioned them personally: “When you robbed this tomb, did anything unnatural occur?” [The leader of] the robbers said: “About this undertaking today I might as well conceal[47] nothing from you.[48] We have been ruined by the work of spirits. My wife and I have been in the tomb-robbing business[49] for over ten years now. Whenever we did a robbery, my wife and I would take along liquor and kindle a fire, while the rest of the gang opened the tomb. Once they reached the coffin cover, we two would proceed to pour and drink liquor with the deceased. I would drink a cup myself, and announce: ‘your guest drinks this cup’.[50] We would then pour the liquor into the mouth of the dead person, saying: ‘the host drinks a cup’. My wife would drink a cup in a second round. Then I would say: ‘From what source shall we pay for these drinks?’ My[51] wife would reply: ‘The host will pay the money for the drinks.’ Then we would take the clothes and any precious goods.”

Most of this long confession—it continues below—is an intriguing mixture of courtroom summary and the participants’ vernacular transmuted into a court record, hence the discrepancy between what the gang chief doubtless means and a lexicon and syntax that sometimes recall the sayings of Confucius. The extent to which these courtroom exchanges may have been remoulded in standard literary forms is not easy to gauge. Did the robbers really use the word “profession”, or did an erudite transcriber of their confession insert a term that locks their last words into the deeper lexical heritage of criminal law. The clerical record that inspired Lu Zhao’s story may have reflected only tangentially the words that the robbers actually spoke at their interrogation, but the history of tomb robbing does not suggest invariably that law officers and miscreants stood either side of the literacy divide. This was apparent as early as the Warring States,[52] and it would be so again when senior provincial officials plundered the imperial mausolea at the end of the Tang.[53] Thus, while the exact social status of the figures in this story remains elusive, the content of this confession has an authentic ring, quite plausibly because it was heard first during a trial.

The story concludes with the robbers’ report of unnatural conditions—notably a corpse that has not decomposed—and Fan Ze’s own realisation that his recent experience of time is synchronised with neither events in the tomb nor the Zhang brothers’ access to presentiments of an outrage. The robbers claim finally:

When we opened this tomb last night, we saw that the person in the coffin was wearing a purple robe and a jade belt. He appeared alive. I drank some liquor in the usual way, and had got to the point where we pour the liquor and pronounce ‘we honour the host with this cup’. As soon as the words were out, the dead man smiled. All of us were utterly shocked. Then we got him upright, and he was nothing but dry bones. Afterwards, when we unfastened the belt at his waist, the dead man cried out: ‘Make it loose. My back is sore.’ All of us panicked and then bolted outside. From that moment onwards my mind has been cut off from its senses,[54] and I knew that we would be defeated.” [Fan Ze] executed them all. It was not until some days later that Dengzhou reported the incident.[55]

Zhang’s purple robe, his jade belt, the paradox of his body’s “live appearance” and its “dry bones” present him as an aspirant Daoist immortal, already capable of delivering his spirit from its physical restraints. Borrowing the terminology of the age, Isabelle Robinet’s description of this metamorphosis could not be more to the point: “[the spirit] disencumbers itself ‘without loosening either the clothing or the belt.’ The adept is an ‘immortal who has slipped out of his belt in midday.’ He ‘moults like a cicada,’ or ‘like a snake’; he is even said to abandon his bones like the dragons whose remains are found in the mountains.”[56]

These hermeneutical extensions—as far back as Han literature—only enhance what remain immediate and tangible signs of religion’s most powerful techniques, manifested in a conjuncture that provokes the robbers’ sheer terror. Yet, apparently willing to stray beyond the single logic of one set of reactions, Lu Zhao also freights his criminal protagonists with the transcendental aspirations of classical lyricism, optimising their fate somewhat better than might be expected of an abject outcome in the justice system—the Tang legal code, which isolated tomb robbery as a specific category of theft, decreed strangulation for the most serious degrees of offence.[57] Not only does the robbers’ leader address his accusers as if they were disciples trying to match their intellects with his, he echoes a poet who has arrived in an empyrean region of detached enlightenment. Talking like Qu Yuan 屈原 (ca. 340–278 BCE), he uses words from an ubiquitous lyrical model of metamorphosis, whose positive transcendental vision was so popular that much of its terminology had long ago already made its way onto bronze-cast mirror backs.[58] The imaginative appeal of these metaphorical tropes is that they summon all the more strongly the sense of separated worlds. The conclusion of the story emphasises this more prosaically: the significant lapse of “some days” in the communications between Xiangyang and Dengzhou corresponds with the startling revelation that two realms of existence are not subject to the same progress of time, notwithstanding that a dream communication effected their momentary collision.

This story of robbery also provides valuable insights into deviant forms of ritual performance. Lu Zhao is not the only compiler to report how tomb robbers appeased the spirits of those whom they robbed. Duan Chengshi also collected a story of Tang looters who, having failed to outwit the lethal security devices of a Han tomb, forsook their exhausting efforts and instead offered a lavish sacrifice at the tomb entrance.[59] In another story in the History, Lu Zhao details the formal procedures for distributing stolen goods among the members of a robber gang.[60] But, the story of the robbers captured by Fan Ze is all the more interesting, because it links ritual performance to the passage of time.

The robbers follow prescribed ritual forms that imply their conscious management of time as well as their efforts to wrest control of events in their favour. All the while acting in criminal resistance, they adopt the language and exaggerative gestures of generic ritual performance to make time palpable during a drinking party that is intended as the medium to transform theft into consensual exchange. This mimetic set of actions features a banal conventionality of language at a micro level of social performance that appears only rarely on record. Ritual action gives a material quality to time that makes its different shapes all the more perceptible. Like the representatives of even the most ideal form of government, the robbers resort to prescriptive patterning to mark the flow of time. This is by no means exceptional, considering how much ritual performance affected every level of Tang theory and practice. Senior Tang officials’ involvement in ritual performance exceeded that in any other occupation, and this profile no doubt reflected an even more extensive involvement in Tang society at large. The robbers’ words and acts are the familiar morphological elements of all Tang ritual performance. Ironically, they render the official and criminal worlds indistinguishable, even though the outcome of the story—and the early portent of that outcome in the Zhang brothers’ dream—confirm that these worlds are asynchronous. The homologous appearance of representational forms and structures is common in Chinese attempts in art and literature to balance everyday reality with parallel, opposing and even antagonistic counterparts, more familiarly described as the underworld, hell or heaven. Religious acts in these conditions—both past and present—more often than not exemplify the paradigm of ritual technicians borrowing performance content from one set of circumstances and applying it to remold the shape of time in another.[61]

Material Agents

In another story of tomb robbery in the History, Lu Zhao reimagines time in its asynchronous shapes, only in this instance the agent of interaction between one sphere and another is a material thing:

At the beginning of the Tianbao 天寶 period [742–756] Yan Anzhi 嚴安之 was the Officer Charged with Thief Apprehension in Wannian 萬年 Sub-prefecture.

Yan Anzhi served successively in disciplinary posts at Luoyang and Chang’an. In Wannian, one of the two sub-prefectures into which Chang’an was divided, the administrative offices were located near the southwest corner of the city’s East Market.[62] Wannian interests the compiler more than once: a different story records the Wannian law officers pursuing an offender as far as Gongxian 鞏縣, the other side of Luoyang,[63] and another story records a prediction concerning an official holding the same post as Yan Anzhi in the 790s.[64] Wherever Yan worked, he earned renown as a pathological sadist who thrilled in the administration of torture,[65] cynically exulting in his reputation through competitions with his rivals.[66] Like many brutes, however, he also enthralled Xuanzong 玄宗 (r. 712–756), characteristically obsessed to have a court ceremony performed without fault or interruption, was awed by Yan’s ability to make a rowdy mob stop heaving and shouting for the entire duration of five days.[67] Another late recollection of his service in the Luoyang area recounts how Yan, having traced a series of housebreakings to a culprit working in a palace factory, became locally celebrated as an uncompromising law enforcer.[68] That success parallels both his official function and his access to high circuits of power in this story:

One afternoon, a palace emissary dressed in yellow and riding a horse came rushing through the gate to announce a decree: “Ten li south of the city walls is a certain princess’s tomb. We have seen it being plundered by robbers. The orders are to send you there to arrest them. You must not let them escape.”

Although its magical significance will be revealed in the conclusion of the story, here once more is a time- and space-conquering device, formed now as a messenger in broad daylight. The story continues:

Anzhi then instructed those under his authority to assemble their weapons, and to go there and make a full arrest. He saw that six or seven people had dug a tunnel and just reached the tomb passage, so he arrested them that instant. Anzhi then gave orders to seek the palace official, but to no avail. Afterwards he thought: “If the robbers had only just opened the tomb, how did the emperor know of it?”

Yan Anzhi’s question concerning the emperor’s involvement raising the alarm is interesting, given what else is known of his career. The story of his campaign to solve a series of housebreakings near Luoyang is not clear in all its details. However, it is easy enough to grasp that Yan Anzhi resolved a case of repeated robberies at a country mansion only once he had adopted the expedient of submitting a report to a member of the government. Even that was not enough. The recipient of his report memorialised the throne, so that it took two steps up the bureaucratic ladder to uncover criminal links between a gang roaming in the Luoyang district and an official operating in the palace factories. Yan Anzhi’s access to an official sympathetic enough to his mission to take his information all the way to the top and to achieve a result was no doubt quite unusual and perhaps not entirely realistic. But it does reveal credibly that the system for reporting crimes in and around Chang’an and Luoyang comprised no firm structure, and this suggests that a deep division between central government and the life surrounding it was probably quite normal. The story concludes:

Upon arrival at the sub-prefecture he summoned all the robbers and questioned them on the matter. The robbers said: “We had just opened the tomb when we felt that there was something unnatural. We realized that we would be defeated, for, having reached the first doorway there were several tomb models of emissaries of decrees. They wore yellow and rode horses. One of them held a riding crop, and he was styled in the form of galloping. The ends of his cap ribbons seemed even to be blown straight by the wind. From his eyebrows downwards everything seemed to be in movement. We then knew that we would be defeated.” Anzhi now recalled the appearance of the earlier emissary of decrees. Obviously, he was the double of the tomb model of an emissary of decrees.[69]

Yan Anzhi’s story, like that of Fan Ze, is a reliable reflection of Tang material and economic realities, since immolating large clusters of tomb models reached the height of fashion in the Kaiyuan reign period and began to wane only slightly during the Tianbao reign period, the period of Yan Anzhi’s years in office. After the mid-century wars the industries that made these objects—for their biggest consumption in Chang’an, Luoyang and Yangzhou—seem to have declined rapidly.[70] However, senior officialdom never let go of its enthusiasm to create models that signified, as realistically as possible, individual figures and specific actions. Those who could afford this expense displayed tomb models at the graveside, and even mounted theatrical shows, using models mechanised to move and performs actions. Such efforts provided entertainment, multiplied social participation, and enhanced the status of funeral rites generally. David McMullen has suggested that competitive ostentation of this sort may even have been permitted during the entombment of emperors.[71]

The device for Yan Anzhi to solve this crime comprises beings who are ontologically interchangeable and simultaneously manifest in two different places. They function with the same effects as in the story of Fan Ze, since the narrative formulas of both stories hardly differ. A message transmitted from the world of the dead warns the living that a crime is happening; asynchronous slippage between two worlds allows the police an opportunity to do their work successfully. The single difference is that, while the voice of a dream addressing the Zhang brother arouses neither comment nor doubt, a talking and moving tomb model needs extra explanation, which is supplied via the robbers’ confession and Yan Anzhi’s own realisation of what has transpired.

Yan Anzhi’s most elementary deduction is that forms of material reality—horses and riders made of ceramic—are momentarily motivated with beliefs and attitudes in ways familiar only in human existence. What belongs usually in separate mental and material worlds has become fused into one set of conditions. This miraculous inter-substantiation—especially when it helps to solve crimes—appealed to Lu Zhao. He recorded another story in which a murder victim communicates through a dream to the sub-prefect of Wangwu 王屋 that his wooden effigy is hidden in his home. Once retrieved, the effigy is observed to be turning into flesh and, not long afterwards, the servants who murdered their master change into wood.[72] These various categories of ontological shifting are thoroughly consistent with Tang beliefs in the magical efficacy of tomb models.

Lu Zhao may have entertained a nostalgic fascination for funeral arrangements that were probably no longer current during his own day. Perhaps the events in the robbed tomb also satisfied an aspiration to deploy vicariously through Yan Anzhi the toys of funeral space and ritual as a spiritual recreation, perhaps even to the extent of compensating for the lost cosmic and political unity of the early Tang. Certainly significant is Lu Zhao’s acceptance of Yan Anzhi’s conviction that the experiential shift from object to person allows passage between what was otherwise two distinct dimensions of existence. This realisation, eventually condensed in the story’s forensic, judicial and magical closure, is linked to the powerful effects of duplication, representation and materiality—huge topics worth at least some brief gesture, since the material conditions of Tang burial are also integral to the perception of asynchronous time.

Yan Anzhi’s dramatic realisation is that an object perceived by the robbers as a tomb model—an everyday object in their line of business—is duplicated as a senior servant of government command and as an agent of action. The robbers must comprehend that action at the practical heart of politics is also manifest in an underground space outside the city walls. All participants in the same crisis are confronted with an efficacious pairing across time and space, in which objects project the aesthetics of symmetry whose sociological significance is a powerful link between periphery and centre.[73] Symmetry of this order is sudden and fugitive, because its reality is transient, but it subsists all the same in a permanent material pattern of human beings duplicated (liang 兩) as objects that express an hierarchical political order.

The art of representation was itself a source of power. Early elites in China had long regenerated the forms and functions of tomb models after an enormous re-evaluation of the tomb’s symbolism in the late Bronze Age. Theories and practices concerning tomb models matured when this ancient period’s ideas deepened the separation between the spheres of the living and the dead.[74] Ambitions to exert control among the dead were no longer expressed primarily though the violence of graveside sacrifices, but by means of objects whose visual accuracy translated into magical efficacy. Not simply bathed in the spot lit significance of modern museum displays, these things were intended for action. Recently, therefore, scholars have argued that models incorporated both a sign and its living referent, or that, in other words, models were not substitutes.[75] The robbers’ final utterance presents the same analytical work from direct experience: “everything seemed to be in movement”.

Literary history shows that the possibility of tomb robbers encountering ultra-realistic human models—or indeed models that could alternate their existence as human beings—was recorded long before the Tang. Although it survives in incomplete form, one of the Chinese tradition’s earliest repositories of collected historical facts and hearsay, the Random Records of the Western Capital (Xijing zaji 西京雜記), contains an account of the Han prince Liu Qubing 劉去病 robbing tombs in the vicinity of Chang’an. The work is attributed traditionally to Liu Xin 劉歆 (d.23 CE), and assumed to have been further edited by the Daoist polymath Ge Hong 葛洪 (283–343). It claimed an enthusiastic readership throughout the Tang, evidenced not least by its frequent citation in Tang encyclopedia, commentaries and style guides. Apparently, Liu Qubing regularly broke open tombs. When he and his followers entered the tomb of King You 幽, the last Western Zhou 周 king, they encountered the bodies of one hundred women and one man sitting, lying or standing in the tomb chamber. Their bodies had not decomposed, and their clothing and expressions made them seem fully alive.[76] If this eerie report refers to tomb models, it does not correspond with today’s archaeological records, and it seems more likely that Liu Qubing encountered the conditions of a later burial. Be that as it may, the particular interest of this account is the precedent that it provided for conceptualising tomb spaces as zones in which visitors confront asynchronous disjunctions, and it may be of added significance that the editor was no less a figure than Ge Hong, one of the giants of pre-Tang Daoist studies.

The agency of tomb models in the story of Yan Anzhi’s capture of a gang of robbers can benefit also from recent insights in the study of material culture to suggest that subjects do not invariably assume a sovereign dominance over their physical world. This latest turn towards the material, which has much to say on the idea of landscapes and objects occupying extra subject positions, suggests ways to understand Tang reactions to objects ranging from holy mountains to chemical elements. Instead of understanding objects as encompassed by subjects, Webb Keane suggests that objects are “forms of materiality that encompass their subjects”. Consequently, the house is a transitional locus between the materiality of a land mass that “encompasses” and the smaller things over which humans exercise more control.[77] Similarly, the Tang tomb, with its permeable skin between two worlds is a foyer suspended between the larger forces that govern it and the apparently servile agency of its contents, namely tomb models.

While uninformed attitudes expect tomb models to be objects that their subjects’ control, Yan Anzhi grasps—as do the robbers—that this partitioning is no longer true. The objects that they encounter take decisions and act; they work as agents of government control; they take a hand in suppressing crime. Both asynchronous temporality and the encompassing power of material objects can work with equal effect to unseat humans from their assumed subject positions. The power attributed to objects seems to be fleeting—momentarily glimpsed and then gone—but it is nonetheless credible, given that the exacting and duplicative realism of tomb models offered enormous appeal to contemporary imagination, and owed its long history to religious axioms favoring the functional power of representation. Yan Anzhi accepts that his success in beating crime is due to powers greater than his resources for intervention. He sees too that the political authority invested in material objects can engage with and arrest asynchronous events and resistant temporalities.

Lu Zhao’s Perceptions of Time

What is the significance of Lu Zhao’s stories concerning tomb robbing and their reflections of asynchronous time? Firstly, Lu Zhao was not interested solely in divergent time, but also keen to describe the inner workings of the phenomenon. Significantly, several other compilers of his day shared this enthusiasm. Secondly, then, through this and other correspondences with more minor historical texts of the same period, Lu’s History forms part of a larger group of writings that comprised a kind of “report literature” of the ninth century’s middle decades. Since we cannot dismiss the content of this writing as purely fictional or as entirely supernatural, such literature needs to be characterised in its links to the categories of historical writing that it so much resembles. Lastly, therefore, a tradition of editing that collected the realities of life and religious belief usually located in the peripheral contexts of Tang existence is worth examining for what it reflects of a Tang undertaking in historiography. To be discerned here also is a parallel between asynchronous divergences operating back and forth between centre and periphery and a particular interest in the functions and values of historical writing.

Lu Zhao’s interest in time concentrated more than once on the literate functions of the historian, since the realities of historical composition and archival research presented powerful metaphors for historical time’s inner workings. In particular, the talismanic force of writing is the apotropaic medium that Anna Seidel analyses in her study of the huge Daoist intellectual debt to Han concepts of sovereign power over spirits.[78] In stories circulating during Lu Zhao’s lifetime, writing presented a medium of control that both humans and spirits contested. One more story in Lu’s History shows this with reference to time materialised in the forms of the historical archive and its associated operations of drafting and amending.

The story of Li Minqiu 李敏求 and his dream was extremely popular, giving rise to at least two versions in circulation before Lu Zhao completed his History. Zhong Lu 鍾輅 (jinshi 828) recorded one version in his Records of Predestination (Qianding lu 前定錄), completed between 830 and 834.[79] Another version is extant in Xue Yusi’s 薛漁思 Accounts from East of the River (Hedong ji 河東記), probably completed before 837.[80] Lu Zhao preserved a third and shorter version.[81] After spending fruitless years trying to win a state degree, Li Minqiu dreams one night of a journey to another world where all human events are preordained, and here he meets an erstwhile friend now working for the government. Through this connection, he is permitted to read the archives that determine future outcomes concerning his political success and financial benefits. In Lu Zhao’s story, he further avails himself of the chance to change the text of what he has been privileged to read, so that an amount of capital guaranteed to accrue to him is substantially increased. Inevitably, these changes dictate events to proceed entirely to his advantage. Aside from—yet also because of—the highly realistic forms and appearances of ninth-century writing and amending records, document handling, cataloguing, storage and retrieval in all three stories, Lu Zhao’s version reflects the writing process as critical to the power that Li Minqiu can bring to bear on his mortal destiny. Variously told by more than one writer, the interest of this story is the sense that it provides of historical records at the heart of contested control between those at the centre and those on the outside. Li Minqiu wins an advantage through an inner contact, but, even so, he is mindful not to disrupt the security regulations that are spelled out to him in each story. He has pitted himself against strong forces in a contest that is already apparent in the metaphor of a central archive recording outcomes at odds with the individual’s own preferences.

Whether it is through human dreams or the ontological shifts of objects, the asynchronous events reported in Lu Zhao’s stories provide both selfish rewards and judicial victory. In the more widely shared realities of Tang life, temporal separation—unified only by a dream communication from the dead in the night-time hours—is a powerful reminder that time is not universally shared. The government official, bound by the affairs of a provincial magistracy, performs his actions and rehearses his experiences more or less in tandem with other government agents of the day. Yet, quite clearly not embedded within that literate and calendrically even world are the acts and consciousness of other groups, most especially the disappointed, dispossessed, the poor, the insubordinate and the criminal. Moreover, even if they are not absolute, other opposites in many accounts reinforce this antagonism: day versus night; urban versus rural; living versus dead; legitimate versus criminal; mundane versus magical. In the two stories of tomb robbery, the forces of law and order leave urban centres to impose discipline in rural areas outside the city walls.

Kracauer’s analogy of times that march in divergent directions is particularly useful for imagining time in the Tang world as an indeterminate number of shapes whose relationship to each other is asynchronous. For Fan Ze, time flows one way; for the tomb robbers it flows another. In Fan Ze’s usual experience, time does flow through a calendric vessel, but the robbers live according to an entirely different schedule. Effectively, time for each group is not necessarily a commensurate dimension within a universal simultaneity. Or, as Lévi-Strauss also observes: “the dates belonging to any one of these classes [of history] are irrational in relation to all those belonging to the other classes”.[82] In Lu Zhao’s stories, supernatural powers account ostensibly for chronological slippage across space, but modern social insights on time are equally relevant. The supernatural forces in these stories would not have been convincing in Tang experience unless the period’s officers and criminals also accepted that their respective groups inhabited a realm profoundly isolated from each other.

Doubts that time dictates the same reality to every subject’s experience permeate a considerable amount of late-Tang writing. Despite their commitments to standards of measurement, the unity of the empire, and the single notion of the emperor, officials like Yan Anzhi and Fan Ze are forced to consider how time progresses in divergence, ultimately allowing a compiler such as Lu Zhao to note remarkable moments when hitherto isolated flows suddenly converge. Intuition directs that time is rated and qualified by those in power, and the historian Lu Zhao invariably loans this privilege to his stories’ protagonists of the same official background. But, Tang time, to which a far greater number of subordinate others are subject, is also an unfriendly or estranging notion that some in the larger and more amorphous whole of Tang society may elect to defy through different conceptions of temporality. Similarly, these recorded experiences reflect how Tang time, if measured at its heart of political and religious power and at its centre of economic activities, does not converge neatly with time endured at the periphery where other priorities—especially religious ones—reign. An asymmetry exists that is not easily overcome, and it is this nature of time that a large body of ninth-century historical writing was well adapted to report.

Central and Peripheral Concerns in Tang Historical Writings

Lu Zhao’s stories preserved in the Taiping Extensive Records and other collections once formed part of an edited collection that fits into three traditions of literary compilation at least: stories (xiaoshuo 小説), records of the strange (zhiguai 志怪), and history. How to define all three requires some qualification, since the status of at least the first two has been frequently distorted by the priorities of recent and current hindsight.

Despite a title that suggests predominantly historical content, Lu Zhao’s History of Things Outside the Norm, and many more works with similar titles and editorial aims, were soon classed as xiaoshuo, long a bibliographical category as much as a description of content. The earliest catalogue listing of Lu Zhao’s text in the New History of the Tang Dynasty places it under xiaoshuo, but the listing in the History of the Song Dynasty of two texts with similar titles, namely History of Things Outside the Norm and Mr Lu’s History of Things Outside the Norm (Lu shi yishi 盧氏逸史), places them respectively under the biographies (zhuanji 傳記) category of the history division and xiaoshuo.[83] Whether these two appearances represent two editions of Lu Zhao’s work is not certain. The elastic term xiaoshuo has sometimes been synonymous with “fiction”—an association that certainly thickened during early Chinese modernism’s adoption of new forms of literary creation and the enthusiastic hunt for antecedents to the novel (broadly defined). The result of these developments has been to minimise the historical value of many Tang works by validating them primarily as narrative exercises, the best of which writers in subsequent periods eventually adopted as source material for their own achievements in popular fiction and drama. Any attempt to make Lu Zhao’s work fit into this long development of literary history will lead off target. More useful, instead, is to examine the significance of Lu Zhao’s work to the two other aforementioned traditions and to revisit the exclusively historiographical significance of xiaoshuo for late-medieval and early-modern writers.

Lu Zhao’s History belongs also to the long zhiguai tradition of recording the strange. Numerous works, which appeared in the late-Han period and the following centuries, collect and systematise accounts of supernatural events and mirabilia linked in various degrees with human dealings. That these records of the strange, unlike xiaoshuo, with which they intersect, never gained recognition as a separate bibliographical category is out of all proportion to how this tradition and its contents saturated an enormous range of writing before and during the Tang period. Glen Dudbridge shows convincingly that Gu Kuang’s (jinshi 757) undated preface to Dai Fu’s Great Book of Marvels (Guangyi ji 廣異記) inscribes this large work closest to the zhiguai tradition, a primary object of the eighth century’s “middle-of-the-road” expectation for this kind of literature.[84] The extant stories of Lu’s History, are items that fit uncontroversially into the same broad tradition nearly one century later.

What the zhiguai tradition still comprised by Lu Zhao’s day requires another qualification. In common with a tendency shared by much writing of the eighth and ninth centuries, nearly every story in Lu Zhao’s History is attached to known individuals set initially in realistic and often verifiable social conditions. Even in Dai Fu’s day, the outrageously quirky was no longer the fashionable object that it had been one or two centuries’ earlier. Uchiyama Chinari notes that from the mid-eighth century onwards the status of many compilers equated increasingly to middle and low-ranked officials. Their attention focused more than hitherto on people and affairs of more modest origin and background. This late period also witnessed many compilers’ rising interest in Daoism. They collected information about this and other subjects rather in the manner of reporters, often trading their finds with each other—a reality that may have given rise to the aforementioned three accounts of Li Minqiu’s dream. Uchiyama notes too that one of the new literary developments throughout the Tang period in the hands of certain writers was the sophistication of imaginary plots and the use of more literary registers of language, especially true of a body of stories that later ages designated with the generic term “transmitted marvels” (chuanqi 傳奇). The description of this development is not apt for all Tang writing, and Lu’s History, which uses quite pedestrian constructions and vocabulary, fits closer with Uchiyama’s idea of a literature of report.[85]

The history category comprises another broad tradition—overlapping considerably with others—and its numerous categories of writing provided models for almost any of Lu Zhao’s topical interests. Engagements with history changed dramatically during the Tang. The expansion of historical interests, which accompanied the many deliberate resorts to private, small-scale historical compilations throughout the late Tang, was an important impetus to Lu Zhao’s compilation of his text. Closer scrutiny of Tang sources in recent historiographical research has retrieved many compilations from definitions that neglected their relevance to historical writing. Cheng Yizhong 程毅中has best described the most important theoretical implications of this shift in his discussion of Liu Zhiji’s attention to minor categories of historical compilation. In a section of his Generalities on History (Shitong 史通, completed in 710), Liu argued that professional historians should adopt a range of usually overlooked writings as worthy categories of historiography.[86] Cheng adopts these arguments to review the output of the following two centuries and to define an unprecedented number of its literary categories as genres of history.

Liu Zhiji described an entire group of such writings as “corollary accounts and minor persuasions” (pianji xiaoshuo 偏記小説), and he identifies ten categories of clerical and historical record in precise terms:

“Corollary accounts” (pianji), “minor records” (xiaolu 小錄), “extra facts” (yishi 逸事), “light statements” (suoyan 瑣言), “commandery writings” (junshu 郡書), “family histories” (jiashi 家史), “individual biographies” (biezhuan 別傳), “miscellaneous accounts” (zaji 雑記), “geographical writings” (dili shu 地理書), “manuals on capital cities” (duyi bu 都邑簿).[87]

It is useful to recall this list in its entirety, because it demonstrates the breadth of vision with which Liu Zhiji recommended sources that “can participate in the progress of official history” (neng yu zhengshi can xing 能與正史參行). Moreover, Liu Zhiji suggests that many, if not all, of these categories owe their existence to urges of random spontaneity, natural forces that outdo human plans and creative ingenuity. Expanding on the characteristics of his fourth category, “light statements”, he acknowledges the earliest bibliographical definitions of xiaoshuo set down by Ban Gu 班固 (32–92) in the History of the Han Dynasty (Hanshu 漢書) and reiterated by Zhangsun Wuji 長孫無忌 (d. 569) in the History of the Sui Dynasty (Suishu 隋書), but he also projects beyond this tradition to a Daoist notion of disciplined chaos: “talk on the streets and opinions in the lanes[88] are sometimes worth considering; minor persuasions and spontaneous talk[89] are even wiser than our selves” (jietan xiangyi shi you keguan xiaoshuo zhiyan you xian yu ji 街談巷議時有可觀小説巵言猶賢於己).

Did Liu Zhiji’s rigorous framework of working methods provide relevant conditions for compiling a collection such as Lu’s History more than one century later? No single answer suffices. Lu Zhao’s attention to his past and present surroundings fell short of a major historical critic’s ambitions; on the other hand, it sometimes exceeded them.

The ten categories in Liu Zhiji’s list represent an elite bias for the top end of human affairs. Liu Zhiji’s interest in capital cities, for instance, which he discussed again as one of several topics deserving research for new monographs in the official histories,[90] emerged from the common antiquarian interest in the organisation of southern and northern capitals, such as Jiankang 建康, Luoyang and Ye 鄴 (capital of the Northern Qi 北齊 government under which one generation of Liu’s ancestors lived). The category is tenth on his list, which may signify an unusually low priority for political centres. Even so, any historical interest in places of such elevated status is bound to be qualitatively different from late-Tang perceptions of the growing importance of provincial centres, such as Xiangyang, the city where Lu Zhao recorded Fan Ze’s operations. Similarly, an antiquarian-topographical interest in Luoyang was a more elitist concern than Lu Zhao’s fascination with the experiences of officers on active service in the field, leastwise with that of criminals in the countryside.

Conversely, Liu Zhiji’s ten categories suggest a historian’s outlook based on catholic interests and a new desire to investigate neglected areas of research. In this respect he exerted a major influence on later generations of historians, and it is reasonable to suppose that he licensed Lu Zhao and contemporaries to turn their attention to a new and unprecedented status for minor historical records. No absolute correspondence between Lu Zhao’s work and Liu Zhiji’s precepts is discernible. But, beyond doubt, Liu Zhiji confirmed a new departure in the functions of an historian. The range of interests displayed in his list of ten categories confirms his disenchantment with the central establishment’s management of historical writing along a correspondingly narrow front. His positive regard for zhiguai literature marked rather a step forward from a generation earlier when Zhangsun Wuji oversaw considerably less committed views recorded in the literary monograph of the History of the Sui Dynasty. In short, the first truly buoyant evaluation of multiple categories of xiaoshuo writing as a category of historical record is an attitude that Liu Zhiji first documented.[91] Zhang Xuecheng 章學誠 (1738–1801), another brilliant innovator, got the wrong end of the stick entirely when he denigrated Liu Zhiji for upholding only the ambitions of the History Office.[92]

Fully in step with the outlook of his own generation, Lu Zhao had assimilated the lessons of a highly critical historian writing more than one century earlier. The title of his collection, History of Things Outside the Norm, which recalls Liu Zhiji’s category “extra facts” (yishi 逸史 and yishi 逸事 respectively), suggests too a ninth-century debt to the clearest surviving Tang theorisation of minor historical compositions. Even more striking is a debt to Liu Zhiji’s dual formulation of sources that “can participate in the progress of official history” and official history itself. Recall that in the preface to his work, cited above, Lu Zhao asserts that he first finished his Historical Records and only then set to work on the History. The mention of both works in one statement reveals an intention to define his work as a dual achievement, one in which the significance of each part sustains the other. It seems too that a priority is implied. The compiler expends effort on a text whose title suggests its centrality to the expectations of official history before he is ready to turn to another text concerned with what others had overlooked in order to “retrieve what was authentic and restore what was missing”.

Lu Zhao seems even to have welcomed quite cynical views of the value of official history. These are clearly presented in one last story: an early seventh-century Chan monk residing on Hengshan 衡山 meets Yao Hong 姚泓 (388–417), the last ruler of the Later Qin 後秦 regime, which collapsed in 417.[93] Captured, Yao Hong was transported to the Eastern Jin capital at Jiankang and apparently executed in the market place. Two centuries later Yao now explains to the monk that he soon escaped captivity, so that, in order to maintain their performance of punishment, the Jin authorities were forced to execute another prisoner who closely resembled him. Significantly, this exchange between monk and immortal is presented primarily as a discussion of historical sources. The monk claims to know what happened to Yao Hong because he has read the history of the Jin (晉史)—meaning most likely The History of the Jin Dynasty 晉書 by Fang Xuanling 房玄齡 (578–648). He could also have read a similar account in Xu Song’s 許嵩 (fl. 750s) The Veritable Record of Jiankang (Jiankang shilu 建康實錄).[94] To the monk’s familiarity with his life, Yao Hong retorts “This is proof that historians talk nonsense”. He gives other examples of official history’s redundancy and further elaborates on what has happened in the last two centuries in order to convince his listener that “there was even more that the historians had omitted and not written down”. Although they critique objects at a safe enough distance from the present, these are, nevertheless, quite bold remarks. They reveal a sceptical attitude to history that can only have encouraged Lu Zhao to collect stories that oppose the standard versions of events and to establish asynchronous patterns for their explanation.

Whatever reception Lu Zhao’s text gained during his own day is not known. It did attract the notice of Zhang Junfang, who included at least fourteen stories from Lu’s History in his Daoist anthology The Cloud Bookcase in its Seven Labels , presented to the throne after 1025. The two stories of tomb robbing are not among Zhang Junfang’s selection, but this does not reduce the significance of Lu Zhao’s work as a prime source of Tang biographical facts and hagiological arguments in the hands of a major interpreter of Daoist doctrine.[95]


The larger point of reference in this article is to the History of Things Outside the Norm as a text of history. Lu Zhao’s claims on behalf of his text suggest that a late-Tang historian could approach history twice: once in the formal setting of compiling the dynastic record organised by the central bureaucracy, and once more in the less constantly patrolled limits of service in the field. Bureaucratic methods tend to prescribe their aims closely, and to achieve them through vigorous editorial reduction and deletion, whereas private compilation allows less constrained intellectual inquiry and more equal exchanges of knowledge. It seems unsurprising, therefore, that Lu Zhao’s stories attracted the notice of a later generation of editors engaged in their more sizeable effort to document the evidence of natural, scientific and cosmic conditions so important to the practical (and secret) workings of the world manifested according to shared religious precepts. Lu Zhao initiated a contribution that later editors completed.

The decentralised structures of late-Tang life, accompanied by an unprecedented growth of urban centres throughout many provincial circuits of the empire stood in no causal relationship to the editing of any branch of history, but the social experience of urban dwelling once magnified can only have deepened the distinctions between diverse inhabitations of space, so that hierarchical levels of administration, different degrees of political control, and unequal possibilities of encountering danger became all the more perceptible. These spatial asymmetries threw up views of temporal disjunction that accentuated understanding historical processes not simply within one homogeneous flow.

No less important were the intellectual changes that took place in Tang historiographical thinking. The title History of Things Outside the Norm that Lu Zhao selected for his text suggests his historical imagination at work for accounts that were literally extraneous to the established canonical models and not subject to the standards of writing official history (zhengshi 正史). The contents of his collection also perpetuate an alternative way of thinking about time—one that was embedded in perceptions of time’s asynchronous flows far away or even isolated from the formally calibrated mainstream of government time. Time flows at the centre and the periphery are asynchronous, if only to be perceived by special effort or merely by chance during rare moments of spillover. Lu Zhao satisfied his interest in this phenomenon not through fanciful inventions alone, but according to inquiries into and reports on real conditions in his immediate surroundings. His stories often lead his readers back to the same place more than once, revealing his and his acquaintances’ regular priority to work from verifiable experiences.

Minor historical works, such as Lu’s History and other texts of the late-Tang period, are not of a higher order than the productions of official government historians; nor does Lu Zhao deserve promotion to a special pantheon of historical talents. Instead, differentiating between historians who did and did not accommodate asymmetrical time flows within their reports is a valuable key to discern not only their outlook on human events but also to individuate the kind of work that they did. One motive for compiling the History of Things Outside the Norm was that compilation represented the metaphorical act of testing different, yet nonetheless valid, worldviews in alternative spaces. Not only does this metaphor align with others concerning the asynchronous flows of time, it enhances the spatial duality of Tang experiences of social reality at either centre or periphery. Such a proposal may help to shift our modern thinking a fraction closer to the philosophy of a Tang historian when he retold events through the creation of his text, and it may redirect recognition for the considerable achievements in Tang historical observation to sources where it is less usually confirmed.

Oliver Moore
University of Leiden
The Netherlands



I owe profound thanks to the organisers of the conference T’ang Studies: The Next Twenty-Five Years for the opportunity to present a draft of this article at SUNY, Albany, New York on 9 May, 2009, and to the Leids Universitair Fonds for its grant to fund my participation.

1 Siegfried Kracauer, History: The Last Things Before the Last (New York: Oxford University Press, 1969), p.149.

2 Denis Twitchett, “Merchant, Trade and Government in Late T’ang,” Asia Major 14:1 (1968) 63–95.

3 Seo Tatsuhiko,Chōan no toshi keikaku [Metropolitan City Planning in Chang’an] (Tokyo: Kodansha, 2001).

4 Otagi Hajime, Tōdai chi–iki shakaishi kenkyū [Investigations of Regional Social Histories in the Tang Period] (Kyoto: Dōhōsha, 1997), pp.415–50.

5 Robert Campany, Strange Writing: Anomaly Accounts in Early Modern China (Albany: University of New York, 1996), pp.1–17.

6 Marcel Grant, La pensée chinoise (Paris: La Renaissance du livre, 1934), p.97.

7 Joseph Needham, “Time and Eastern Man,” in The Grand Titration: Science and Society in East and West, ed. Joseph Needham (London: George Allen & Unwin, 1969), pp.228–29.

8 For a survey of the contents of the Code, see David McMullen, “Bureaucrats and Cosmology: The Ritual Code of T’ang China,” in Rituals of Royalty: Power and Ceremonial in Traditional Societies, eds David Cannadine and Simon Price (Cambridge: Cambridge University Press, 1987), pp.181–236. On Daoist ritual, see Kristofer Schipper and Wang Hsiu–huei, “Progressive and Regressive Time Cycles in Taoist Ritual,” in The Study of Time, V: Time, Science and Society in China and the West, ed. J.T.Fraser (Amherst: University of Massachusetts, 1986), pp.185–205. The authors remark (p.186) that liturgical manuals comprise half of the Daoist canon (after its last recomposition in 1445).

9 Glen Dudbridge, Lost Books of Medieval China (London: The British Library, 2000).

10 Yunji qiqian (DZ 1032, HY1026), 113A. This chapter contains fourteen stories from Yishi, and the preceding chapter contains thirty stories from Shenxian ganyu zhuan. On the transmission of these texts in Yunji qiqian, see Kristofer Schipper and Franciscus Verellen, eds, The Taoist Canon: A Historical Companion to the Daozang (Chicago: Chicago University Press, 2004), Vol.1, pp. 430–31; Vol.2, pp.943–45.

11 Zeng Zao, Leishuo [Categorised Stories] (Beijing: Wenxue guji kanxingshe, 1955 preface dated 1136), 27:4a–5a.

12 Xin Tangshu, [New Tang History] (Beijing: Zhonghua, 1975), 59:1542.

13 Shilin bishu luhua [Conversations recorded during summer retreats in the Stone Forest] (Shanghai: Shanghai shudian, 1990), 1:7a–b.

14 Xu Song, Dengke jikao [Evidence for the Records of Examination Degrees], ed. Zhao Shouyan (Beijing: Zhonghua, 1984), 22:791–92.

15 On Lu Zhao’s participation in the examinations, see Oliver Moore, Rituals of Recruitment in Tang China: Reading an Annual Programme in the Collected Statements by Wang Dingbao (870–940) (Leiden: Brill, 2004), pp.120; 224–25. For a full discussion of all the documents concerning Lu Zhao’s career and movements, see Li Jianguo, Tang Wudai zhiguai chuanqi xulu [Overview of Tang and Five Dynasties’ Tales of Marvels and Records of Anomalies] (Tianjin: Nankai daxue, 1993), pp.670–79.

16 For Lu Zhao’s office at Jizhou and earlier at Shezhou, see Wang Dingbao, Tang zhiyan [Collected Statements from the Tang] (Shanghai: Gudian wenxue, 1957), pp.2.21; 10.110; see also Moore, Rituals of Recruitment, Appendix, nos 41, 260b. See also documents collected by eighteenth-century editors and their remarks at Yu Xianhao, ed., Tang cishi kao [Evidence for the Tang Prefects] (Nanjing: Jiangsu guji, 1987), Vol.4, pp.1871, 2074.

17 Xin Tangshu, 60:1615.

18 Wan Man, Tangji xulu [Overview of Tang Literary Collections] (Beijing: Zhonghua, 1980), pp.292–93.

19 For this implicit recognition, see Wang Dingbao, Tang zhiyan, 2:21, 3:40, and 12:137.

20 Several lost works are attributed to Li Deyu. For a tentative attribution of the collection Youguai lu [Records of Hidden Anomalies] (listed in Songshi [Song History] 206:5220), see Li Jianguo, Tang Wudai zhiguai chuanqi xulu [Overview of Tang and Five Dynasties’ Tales of Marvels and Records of Anomalies, pp.626–28; alternatively, for the suggestion that the record of this work in Songshi confuses it with the Xuanguai lu [Records of Dark Anomalies] by Niu Sengru (779–847), see Cheng Yizhong, Gu xiaoshuo jianmu [Brief Catalogue of Ancient Stories] (Beijing: Zhonghua, 1981), p.62.

21 Taiping guangji, 48:297; see also You-yang zazu, ed. Fang Nansheng (Beijing: Zhonghua, 1981), A8:85 (No.326).

22 Both stories at Taiping guangji, 460.3767–68; see also each account at respectively Yunji qiqian, 113A:14a–16a, and Youyang zazu, A2:19–20 (no. 77).

23 The most remarkable is their use of the postpositional particle ci 次 in place of the more common zhong 中 to express the sense of “in” or “during”. The locus classicus of this usage is found in the ‘Tianzi Fang’ [chapter in Zhuangzi jishi [Collected Glosses for Zhuangzi], ed. Guo Qingfan (Beijing: Zhonghua, 1961): 7B:714.

24 Shuofu (ed. Commercial Press, 1927), 24:21b; also cited at Cheng Yizhong, “Tangdai xiaoshuo suoji ” [Brief Notes on Tang Stories] Wenxue yichan [Literary Heritage], 1980, 2:56.

25 百十五 transcribed as 四十五. See Li Jianguo Tang Wudai zhiguai chuanqi xulu, 692.

26 Taiping guangji, 19:129–31, 35:222–23, 42:267, 48:297, 48:301, 64:400–401.

27 Taiping guangji, 48:299; also at Yunji qiqian, 113A:4a–b.

28 Taiping guangji, 55:341–42.

29 Taiping guangji, 23:154–56, 29:189–190, 42:267, 424:3451; first story also at Yunji qiqian, 113A:7b–9a.

30 Taiping guangji, 42:265.

31 Taiping guangji, 74:463.

32 Taiping guangji, 39:245, 42:265.

33 Taiping guangji, 23:157.

34 Taiping guangji, 23:154–56, 84:546–47, 96:639–40, 122:862 & 864–65, 127:902–03, 148:1069, 149:1072 & 1073, 151:1085–86, 153:1096–1101 (four stories), 156:1125, 157:1126–29 & 1129–30, 159:1147, 172:1263–64, 180:1339, 224:1726, 277:2198–99, 278:2204–05 & 2206–07, 357:2824–25.

35 Taiping guangji, 36:229, 69:428, 72:451–52, 80:506, 125:885–86, 127:902–03, 128:905, 146:1048, 151:1086, 405:3268, 424:3451, 460:3767–68.

36 Taiping guangji, 17:118–19, 49:304, 70:433, 157:1129–30, 378:3012.

37 Taiping guangji, 204:1552–53, 305:2417–18, 420:3421 (featuring music), 42:265, 45:279 & 281–82, 460:3767–68 (featuring weiqi).

38 17:118–19, 29:189–90, 80:506 & 506–07, 153:1100–01.

39 Taiping guangji, 23:157, 65:407.

40 See especially Lüshi chunqiu jishi [Collected Glosses for Mr Lü’s Spring and Autumn Annals], ed. Xu Weiyu (Beijing: Wenxue guji kanshe, 1955): 10:9a.

41 Tang huiyao [Collected Essentials of the Tang] (1955) 39:714.

42 Jiu Tangshu [Old Tang History] (Beijing: Zhonghua, 1975), 12:340 & 357:

43 Jiu Tangshu, 13:373, 388; Fan’s two appointments to Xiangzhou are also noted in a section on the appointment of military generals in Wang Qinruo, ed., Cefu yuangui [Primary Indicator in the Store of Documents] (Beijing: Zhonghua, 1960), 119:23b.

44 Biographies in Jiu Tangshi, 122:3505–06; Xin Tangshu 159:4952–53.

45 Taiping guangji, 127:902–903.

46 Xin Tangshu, 40:1030.

47 yin 隱. See Lu Zhao’s use of this word in another robber’s confession at Taiping guangji, 127:903.

48 亦無所隱: Compare Confucius’ words to his disciples: 二三子以我為隱乎; 吾無隱乎爾. See Lunyu [Analects], in Shisan jing zhushu [Commentaries to the Thirteen Canons], Song edition, repr. 1816 (Kyoto: Chūbun, 1971), 7:7b. According to Bao Xian’s 包咸 (6 BCE – 65 CE) gloss, the sage’s disciples, having despaired of catching up with his learning, assumed that he was concealing as much as he revealed. James Legge translates Confucius’s staement: “Do you think, my disciples, that I have any concealments? I conceal nothing from you”. Bao Xian, Confucian Analects, The Great Learning and The Doctrine of The Mean (Oxford, 1893, repr. New York: Dover Publications, 1971), p.202.

49 ye 業. For this term, current in literature since the Han to describe grave robbing as the “digging business” (jueye 掘業), see Shiji [Historical Records] (Beijing: Zhonghua, 1959), 129:3282.

50 The text 客 欲一盞, which might contain a transcription error, could be translated as: “your guest desires a cup”. Consistent with what the robbers say next, I propose 客飲一盞 and its translation above.

51 其 should be translated as: “his”. However, this single use of a third person pronoun conflicts with the dominant use of first person 某, and is perhaps an error caused by the close resemblance of 某 to 其.

52 For a recollection of tomb robbing as a sort of sport pursued in combination with country outings, concerts and other entertainments, see Hanshu [The History of the Han Dynasty] (Beijing: Zhonghua, 1962), 28B:1655.

53 See biographies of Wen Tao 溫韜 in Jiu Wudai shi [Old History of the Five Dynasties] (Beijing: Zhonghua, 1976), 73:961; Xin Wudai shi [New History of the Five Dynasties] (Beijing: Zhonghua, 1974), 40:441.

54 Changhuang 惝怳: this expression, which is difficult to translate, describes a subject who can barely trust his sensory perceptions. It occurs more than once in Qu Yuan’s “Far-off Journey” (Yuanyou 遠遊) and in several Han texts, but the locus classicus in Yuanyou is in one of the final verses. See Tang Bingzheng, et al., eds, Chuci jinzhu [Modern Annotations to the Chu Songs] (Shanghai, Guji, 1996), p.192. David Hawkes translates: “When I looked, my startled eyes saw nothing. When I listened, no sound met my amazed ear [changhuang]”.The Songs of the South (Harmondsworth: Penguin Books, 1985), p.199.

55 Taiping guangji, 390:3117.

56 Isabelle Robinet, “Metamorphosis and Deliverance from the Corpse in Taoism,” History of Religions, 19 (1979): 37–70, esp. p.58.

57 Liu Junwen, ed., Tanglü shuyi jianjie [Annotated Edition of Commentaries and Meanings of Tang Law ] (Beijing: Zhonghua, 1996), 19:1364 (no. 277).

58 Hawkes, Songs of the South, p.193.

59 Youyang zazu, A13:124–25 (no. 521).

60 Taiping guangji, 357:2824–25.

61 For insights on this point of ritual performance in modern engagements, see ch.5 “Ghostly Epiphanies: Recalling the Dead on Mount Osore” in Marilyn Ivy, Discourses of the Vanishing: Modernity, Phantasm, Japan (Chicago: University of Chicago Press, 1995).

62 See the reconstructed map of Chang’an in Shi Nianhai, et al. eds, Xi’an lishi ditu ji [Collection of Xi’an Historical Maps] (Xi’an: Xi’an ditu, 1996), pp.80–81.

63 Taiping guangji, 130:920–23.

64 Taiping guangji, 153:1096.

65 See the recollections of Yan Anzhi at Luoyang during the early 730s in the chapters devoted to “Stern officers” in Jiu Tangshu, 186B:4857,and Xin Tangshu, 209:5913.

66 Feng Yan (fl. 795), Feng shi wenjian ji jaozhu [Annotated Edition of Mr Feng’s Records of His Experiences] (Beijing: Zhonghua, 1958), 9:83; also at Leishuo, 6:27a.

67 Zheng Qi 鄭綮 (d. 899), Kaitian chuan-xin ji [Authentic Records Transmitted from the Kaiyuan and Tianbao Eras], in Ding Ruming, ed., Kaiyuan Tianbao yishi shi zhong [Ten Categories of ‘Ignored Events’ from the Kaiyuan and Tianbao Eras] (Shanghai: Guji, 1985), p.52; also in Taiping guangji, 164:1192–93; Qian Yi 錢易, Nanbu xinshu 南部新書, compiled during 1008–1016, ed. Huang Shoucheng 黃壽成 (Beijing: Zhonghua, 2002), 甲/5; Wang Dang 王讜 (d. ca. 1107), Tang yulin jiaozheng [Critical Edition of the Tang Forest of Talk], ed. Zhou Xunchu (Beijing: Zhonghua, 1987), 1:54 (no. 93), Leishuo, 6:30b.

68 Nanbu xinshu, 戊/65.

69 Taiping guangji, 390:3114.

70 Feng Xianming, An Zhimin, et al., Zhongguo taoci shi [China’s Ceramic History] (Beijing: Wenwu, 1982), p.215.

71 David McMullen, “The Death Rites of Tang Daizong,” in State and Court Ritual in China, ed. Joseph McDermott (Cambridge: Cambridge University, 1999), p.187–88.

72 Taiping guangji, 128:905.

73 On the importance of symmetry in both material and social patterning, see “Sociological Aesthetics” (1896), reprinted in K.P.Etzkorn, ed., Georg Simmel: The Conflict in Modern Culture and Other Essays (New York: Teachers College Press, 1968), pp.71–76.

74 Ladislav Kesner, “Likeness of No One: (Re)presenting the First Emperor’s Army,” The Art Bulletin 77:1 (March 1995): 115–32, esp. p.116.

75 Jessica Rawson, “Thinking in Pictures: Tomb Figures in the Chinese View of the Afterlife,” Transactions of the Oriental Ceramic Society, 61 (1996–1997): 19–37.

76 Xijing zaji (Ming printed edition in Sibu congkan ), 6:1b–3a; also in Taiping guangji, 389:3100–3101.

77 Webb Keane, Introduction to Part III “Subjects and Objects,” in Handbook of Material Culture, eds Chris Tilley et al. (London: Sage Publications, 2006), p.200.

78 Anna Seidel, “Imperial Treasures and Taoist Sacraments: Taoist Roots in the Apocrypha,” in Michel Strickmann, ed., Tantric and Taoist Studies (Brussels: Institut Belge des Hautes Études Chinoises), pp.291–371, esp.pp. 322–23.

79 Qianding lu [Records of Predestination] in Zuo Gui, comp. 1273, Baichuan xuehai [The Hundred Rivers’ Sea of Learning] (Beijing: Zhongguo shudian, 1990), one juan, 8a–9a.

80 Taiping guangji, 157:1126–28.

81 Taiping guangji, 157:1128–29.

82 Claude Lévi–Strauss, La pensée sauvage (Paris: Plon, 1962), p.344.

83 Xin Tangshu, 59:1542; Songshi, 203:5118, and 206:5222.

84 Glen Dudbridge, Religious Experience and Lay Society in T’ang China: A Reading of Tai Fu’s Kuang–i chi (Cambridge: Cambridge University Press, 1995), p.18.

85 Uchiyama Chinari, Sui Tang xiaoshuo yanjiu [Research on Sui and Tang Stories] [first published as Zui Tōshōsetsu kenkyū (Tokyo: Mokujisha, 1977)] (Shanghai: Fudan University, 2010), pp.3–26.

86 Cheng Yizhong, “Lun Tangdai xiaoshuo di yanjin zhi ji,” [Discussion of the Evidence for the Development of Stories in the Tang Period] Wenxue yichan [Literary Heritage] 5 (1987): 44–52.

87 Shitong [Generalities on History] (1577 edn, repr. Beijing: Zhonghua, 1961), 10.1a–b.

88 Hanshu (Beijing: Zhonghua, 1962), 30:1745; compare 街說巷語 in Suishu (Beijing: Zhonghua, 1973), 34:1012.

89 Three Ming imprints of Shitong (1535, 1577 and 1602) give: xiaoshuo weiyan 小説為言. See Shitong, 10:2a. Other editions, including notably the Shitong tongshi 通釋, prepared by Pu Qilong 浦起龍
(b.1679) on the basis of a Song imprint, give the expression cited above. See Shitong tongshi (Shanghai: Guji, 1978), 10:273–74. The striking image of literally “tumbling beaker remarks” occurs in the Yuyan 寓言 (“Parables”) chapter of Zhuang Zi: “Daily we utter spontaneous remarks” 巵言日出. The Tang editor Cheng Xuanying 成玄英 (fl. 650) glosses this as unplanned, spontaneous talk during drinking games featuring a footless zhi–type beaker that repeatedly tumbles between its users. See sub-commentary in Nanhua zhenjing zhushu [Commentaries to the Nanhua True Canon] (DZ 745), 29:1b–2a.

90 Shitong, 3:9a–12a. See also Edwin G. Pulleyblank, “Chinese Historical Criticism: Liu Chih-chi and Ssu-ma Kuang,” in Historians of China and Japan, eds Edwin G. Pulleyblank and W.G. Beasley (London: Oxford University, 1961), p.145.

91 Cheng Yizhong, “Lun Tangdai xiao-shuo di yanjin zhi ji,” p.45.

92 See Cheng Qiangfan’s citation of Zhang Xuecheng in Shitong jianji [Annotations to the Generalities on History] (Beijing: Zhonghua, 1980), p.318.

93 Taiping guangji, 29:189–90.

94 For Yao Hong’s death, see Jinshu (Beijing: Zhonghua, 1974), 119:3017; Zhang Chenshi, ed. Jiankang shilu [The Historical Record of Jiankang], (Beijing: Zhonghua, 1986), 11:384.

95 For an example of how this material can be used to research Tang religious history, see Franciscus Verellen, “Luo Gongyuan 羅公遠: Légende et culte d’un saint taoïste,” Journal Asiatique 275 (1987): 283–332, on Lu Zhao’s Yishi, esp. 291–92.