China, destined perhaps to be the world’s last Leninist state, was awash with rumors after the Tiananmen killings of June 4, 1989, that the aged leader Deng Xiaoping was about to die. In Beijing sullen survivors of the massacre symbolically broke small bottles (xiaoping) homonymic with the dictator, and residents of the capital spoke ominously of sorcerers’ creating tiny effigies of “old Deng” to be hexed and otherwise incised. Such death wishes made manifest in voodoo-like dolls (tishen, or “substitute bodies”) had ancient origins in China. The history of the Han dynasty (206 BC–AD 220) was punctuated with palace purges conducted by emperors who had discovered manikins designed to bring sickness and death.1 In later reigns Daoist magicians sometimes exerted a dark influence over monarchs who believed in the casting of spells. And the emperors of the last dynasty of all, the Ch’ing (1644–1911),2 while publicly espousing enlightened Confucian virtues of sagehood, practiced a particularly death-obsessed form of Tibetan Buddhism within the privacy of the imperial household.3

The most powerful of these Ch’ing rulers was Hungli, who reigned as the Ch’ien-lung emperor from 1736–1796. Under his rule, and especially after 1780 when an average of 1.6 million ounces of silver flowed into the central kingdom every year, the Chinese enjoyed one of the most prosperous eras of their entire history. At the height of his military might, Hungli cast his sway over a vast domain, extending China’s borders farther west than ever before. Surrounded by sycophantish courtiers, he allowed himself to be compared to the sage-kings of divine antiquity.4 Yet at the same time, we learn in Soulstealers, a masterful study by one of the West’s premier Chinese historians, Hungli was also prone to innate fears of sorcery and sedition—fears that drove him personally to persecute a motley succession of monks, medicine men, and mendicants believed to be “soulstealers” during the great sorcery scare of 1768.

The Chinese believed that a person’s soul consisted of dual entities: a sentient p’o that governed the senses, and a spiritual hun that ruled the mind. The hun could be separated from the body even while a person was alive, roaming while the subject slept and returning in wakefulness. This part of the soul was thus thought to be vulnerable to demons or vengeful ghosts, who might “call away” the hun and send the victim into a deathlike trance. Sorcerers also knew how to steal the soul by using a person’s proper name or stealing a lock of the victim’s hair. Once the hun was captured it could be used to infuse soul-life into paper or bamboo manikins, which thereby transmogrified into monstrous agents of the sorcerer.

In the spring of 1768 an epidemic of “soulstealing” (chiao-hun) incidents broke out in seven Yangtze River delta towns west and south of what is now Shanghai. The accused soulstealers were usually strangers, often Buddhist monks, who would be attacked by an angry mob for asking a child its name, or arrested on suspicion of clipping off the tip of another person’s queue. The latter act was regarded as seditious because the queue was part of the hair style that the ruling Manchus had forced the Chinese to adopt when they conquered the country a century earlier.

A typical case occurred in Hsiaoshan (Chekiang), where a constable heard talk of two monks with strange accents lodging in a local temple. According to his report to the district magistrate, the constable searched the monks’ baggage and found evidence of pieces of hair that may have pointed to queue cutting. After a crowd gathered and threatened to incinerate the suspects, the constable rushed the two monks to the district yamen. There the magistrate examined the evidence, which included two short braids, and ordered the monks to be interrogated with judicial torture. When their ankles were crushed beneath the pressing beam (chia-kun), the monks admitted to soulstealing. Together the two maimed clerics were remanded to higher courts, but while they were in custody the soulstealing failed to stop. Instead the epidemic of queue clipping spread northward. By the autumn of 1768 incidents of soulstealing had even been reported in the capital. Along with evidence forwarded by Governor Funihan from nearby Shantung, these incidents convinced the Ch’ien-lung emperor of the existence of an occult conspiracy, bent upon overthrowing the dynasty.

Modern historians first learned of the 1768 sorcery scare in 1931 when the Peking Palace Museum published a small collection of documents about the panic.5 The documents raised more questions than could be answered about the reasons for Hungli’s zealous reaction to what seemed to be a minor outbreak of public hysteria. However, historians could only speculate about the emperor’s motives because the full details of the sorcery scare were buried in documents stored away in the inaccessible palace archives.


In 1979 Philip Kuhn, who succeeded John Fairbank as the Higginson Professor of History at Harvard, was the first member of a delegation of US Ming and Ch’ing historians to enter the First Historical Archives in Beijing.6 The author of an influential study of nineteenth-century China,7 Kuhn was then conducting research on twentieth-century history. He quickly realized, however, what an extraordinary bounty of material the archives contained on eighteenth-century China.8 Making a conscious decision to abandon twentieth-century research in order to illuminate the little-understood 1700s, Kuhn spent the better part of the next decade mastering the bewildering “pathways of words” through the labyrinthine imperial bureaucracy of the High Ch’ing.9

Thanks to that mastery, we now have in Soulstealers a fascinating account of how the sorcery panic started to spread over the country in 1765, and why Hungli reacted as he did. In some ways the popular hysteria is easiest to explain, both by analyzing the “rich broth of local sorcery beliefs” prevalent in east-central China in the 1760s, and by demonstrating that the soul-stealing crisis occurred just before increasing imports of Mexican silver had begun to relieve the burden of population pressure, which had in turn produced internal migration. The homeless migrants themselves incited fear.

Suspicion of soulstealing focused on wanderers: strangers, people without roots, people of obscure origins and uncertain purpose, people lacking social connections, people out of control. The victims of lynch mobs and of torture chambers were mostly wandering monks and beggars; and if we consider monks a species of beggars, then the suspected soulstealers were all beggars.

A “clerical underclass” of wandering monks was thus the single most visible cause of popular hysteria. But what of the emperor’s alarm? Kuhn argues that the combination of sorcery (soulstealing) and sedition (queue cutting) created a “panic factor” in the mind of the emperor, who believed “that the credulous masses were ever on the brink of violent, panicky reactions to hints of political crisis or cosmic disorder,” and thus could be readily “deluded” by black magicians to rebel against the dynasty.

Sorcery, in this sense, can be seen as the “black” counterpart of the imperial cult. Just as the legitimate sacrifices conveyed to the public an image of firm and worthy control, so sorcery might convey an image of instability and imminent crisis. The representation and the reality were inseparable. No point in asking whether sorcery practices were “really” loosening the dynasty’s grip: the popular reaction to sorcery was what counted. Public disturbances, like astrological omens, were both signs and instruments of Heaven’s displeasure.

Precisely because Hungli was so obsessed by fears of sedition and sorcery, the emperor personally directed the campaign:

There can be no doubt that the chief prosecutor, from first to last, was the monarch himself. This is clear from his vermilion comments, both on memorials from the field and on court letters drafted by the grand councillors. The extra push, the sharper goad, the added injunction to speed and rigor, the acerbic abuse of laggard officials: all were his personal contributions.

The Ch’ien-lung emperor’s personal involvement then provides historian Kuhn with a remarkable opportunity now. By using Hungli’s own hand-brushed “vermilion rescripts,” Kuhn manages “to penetrate his ghostwriters and reach the man himself.”

The Hungli we finally do glimpse behind the Ch’ien-lung emperor’s different personas was far from the serene and self-confident sage-ruler presented to his subjects. There was a paradoxical contrast between the power and stability of his regime and the insecurity and lack of confidence of its monarch. Despite evidence that people who practiced soulstealing were marginal figures, Hungli was convinced that they were being manipulated by seditious conspirators of much greater importance. By October 1768 he had even managed to persuade himself that the conspirators were concocting the incidents precisely in order to attract government reprisals against the common folk, who would consequently rise in rebellion. In a court letter, Hungli warned his provincial governors against the indiscriminate use of torture lest the inquisition touch off uprisings. However, there was to be no slackening of the intense hunt for the main plotters. “If they are not treacherous monks, then they must be disheartened scholars,” Hungli wrote. “Their intentions are extremely dangerous, and their movements are extremely secretive.” The entire countryside—even “secluded villages and derelict temples”—must be searched to bring the conspirators to justice.

Meanwhile the officials themselves were increasingly skeptical about the cases of soulstealing as such. In the Hsiao-shan case mentioned above, for example, the provincial judge grew suspicious of the constable who had produced the evidence that led to the maiming of the two monks. After an extended interrogation, the constable confessed that he had tried to extort money from one of the monks.

But the outraged monk insisted that he was going to file an official complaint. [The constable] started beating him, but without result. He realized that he was in serious trouble unless he could make the queue-clipping charge stick. Unfortunately, there was only one lock of hair in [the monk’s] box; furthermore, it was straight hair and did not really resemble a clipped queue-end. So [the constable] found an old lock of hair in his own house, went out in the alley where [the monk] could not see him, and carefully braided it. For a bit more evidence, he cut some strands of fiber from his own hat fringe and braided them up to resemble two little queues. This hastily concocted evidence he placed in the monk’s traveling box along with his own pair of scissors (making a total of four), and marched his prisoner off to the magistrate’s yamen.

Other soulstealing cases turned out to founded on equally shaky evidence, and almost always the charges were based upon confessions extracted under torture. As the “human debris” of lacerated prisoners were brought to Beijing from the provinces, Hungli’s grand councilors began to question the guilt of the accused soulstealers altogether. This skepticism, in the face of Hungli’s furious determination to find and punish the sorcerers, placed the grand councilors in a difficult position.


They may have shared his fears of sedition. Yet they also had to face the agonized prisoners, with their mangled bodies and muddled stories, who had been sent up from provincial courtrooms. When doubts began to multiply in their minds, they had a serious political problem on their hands.

Identifying cases of false accusation alone was not enough to change the emperor’s mind. After an alleged queue-clipping incident in September 1768 turned out to be another case of planted evidence, Hungli cautioned his prosecutors: “Do not, because of this case of false accusation, permit your will to be swayed, or show the slightest negligence in pursuing the queue-clippers, lest the true criminals slip through the net.”

However, it was another matter when false accusations and distorted confessions were linked to the disobedience of imperial orders to avoid indiscriminate torture. Governor Funihan of Shantung was the most reprehensible in this respect because he had insisted all along that his inquisitors had not used torture to extract confessions, which of course enhanced the credibility of the prisoners’ avowals of sorcery. Yet when the prisoners under his jurisdiction began arriving in the capital for reexamination, their suppurating wounds proved Funihan a liar and discredited his other cases—cases that had helped convince Hungli to call for a countrywide manhunt in the first place. Once Hungli learned of Funihan’s deceit and punished him, the councilors were emboldened to reexamine other cases. By the time the process of exoneration was over, even the very first case of sorcery that had provoked the initial spring panic of soulstealing in Chekiang turned out to be “a phantom conceived in ignorance and nourished in envy.”

On the face of it, Hungli’s reaction to the sorcery scare of 1768 was a fit of despotic panic that foreshadowed his decision twenty-five years later to order an analogous persecution of the followers of the Buddhist White Lotus sectarians.10 Since dealing with the White Lotus rebellion (1796–1803) that followed nearly bankrupted the imperial treasury, might we then regard Hungli’s fear of sorcery and sedition as politically excessive, and even self-defeating?

Kuhn’s response to that question would be negative. He would stress instead that in 1768 the persecution of political crime (sorcery and sedition) had a certain positive function within the bureaucratic-monarchic system: it may have provided an opportunity for Hungli “to whip his bureaucracy into line.” How does Kuhn reach this surprising conclusion?

Toward the end of his enthralling work, Kuhn asks a key question: “Why were not the bureaucrats as concerned about [political crime] as the monarch?” The answer rests upon Max Weber’s fundamental distinction between “routine” and “arbitrary” power in imperial China. Instead of accepting Weber’s notion that autocrats in the end succumb to bureau-crats, and arbitrary authority to routinization, Kuhn argues that the two forms of authority coexisted in the Chinese system of bureaucratic monarchy. In other words, rather than succumbing to bureaucratic authority and becoming a “top functionary” like the Prussian monarchs described by Hans Rosenberg, 11 Kuhn’s Ch’ien-lung emperor was able to resist being trapped by his own bureaucracy in red tape and routine. Of course Hungli used the bureaucratic machine’s routine methods both of surveillance and constant evaluation of the performance of all officials to “tighten the screws” on them with varying degrees of success; but—more importantly to Kuhn—the emperor also managed to succeed in “finding ways to inject his own arbitrary power into [the bureaucratic machine].”

Arbitrary power was necessary because the political culture of the bureaucracy, which stressed the “magnanimity” of patron-client ties, readily accommodated the formation of cliques consisting of both senior and junior officials that kept the bureaucracy from routinely policing itself. (This characteristic of the system is what Chinese today describe when they speak of “feudal” vestiges in the Communist state.) Of course, there was the system of bureaucratic evaluation through the emperor’s secret palace “memorial” system, and this worked outside the usual bureaucratic routine; but Kuhn’s own research shows that officials frequently got together and compared so-called secret reports, smoothing out the differences between them and frustrating the emperor even more because of their uniformity.

The emperor also used his court audiences in order to evaluate officials, and Kuhn analyzes Hungli’s own comments to show how the throne valued “gumption” on the part of the small number of officials who belonged to the “club” of the higher provincial bureaucracy. With these men the emperor formed a special relationship indeed, Kuhn likens these officials to the emperor’s own personal “political appointees,” and he examines their “gratitude memorials” to show how a kind of verbal kowtow created a ritual of symbolic dependency on the part of such officials with respect to their emperor.

This analysis reveals much about the court as a social system, but Kuhn goes a considerable step beyond conventional reflections on the Confucian ruler-minister relationship by arguing as follows. Because routine controls could be evaded, the emperor sought opportunities to exercise arbitrary authority. What might be seen as a disabling obsession with political crimes on Hungli’s part (as shown by his overwhelming conviction that crimes of sorcery were being committed, and by the suspicious spirit behind the literary inquisitions a decade later) is taken by Kuhn to have served a useful function. So far as Hungli’s interests were concerned,

certain classes of events—preeminently “political crime,” as I have defined it—provided the best medium for nourishing the personalistic discipline that bound the upper layers of China’s bureaucratic monarchy. It was the sort of occasion Hungli could use to keep his top officials from slipping away from his personal control and into the rhythms of routine and cronyism.

In this, turning what others might feel to be common sense on its head, Kuhn boldly states that

the soulstealing crisis was a particularly suitable context for personalistic discipline because it was so ill founded a case. The imperial spleen could be vented upon provincial officials for failing to turn up master-sorcerers—a failure that was inevitable because no master-sorcerers existed.

Kuhn cannot argue that Hungli actually intended to use the sorcery cases to shake up the bureaucracy. There is no evidence of such purposiveness. Rather, Kuhn believes that the impersonal “long-term structural features of the bureaucratic monarchy” creating an “overall impetus” that would “shake bureaucrats out of patterns of routine behavior….”

Here, I think, the argument grows weakest. For, in order to give the structure of bureaucratic monarchy a soul-life of its own, Kuhn has to create his own manikins: instances of willful refusal on the part of provincial officials to carry out the emperor’s wishes during the sorcery scare. The evidence of this official resistance is very slight, and Kuhn himself offers little more than his own impression that it took place:

Every one of the measures I am about to describe can be explained on other grounds. Taken together, I am persuaded, they indicate a cautious, pervasive resistance to autocratic pressure. That they were concerted is unlikely, that they were deliberate cannot be proved.

Altogether Kuhn is only able to find three examples of supposed bureaucratic resistance: the failure of the Kiangsi governor’s secret agents to turn up any queue clippers in that province, the discovery by the Kiangsu provincial judge of several long-proscribed Buddhist sects that had somehow escaped official discovery earlier, and the attendance of the Hukuang governor-general at the provincial trial of a queue clipper whose innocence was proved.

Readers will have to judge for themselves whether or not these three instances provide plausible evidence of a “pervasive resistance to autocratic pressure”: a supposed capacity to resist an emperor who, because they failed to discover the existence of the proscribed Buddhist sects, retroactively disciplined 205 officials from district magistrates all the way up to governor-general. I myself was not persuaded by the argument, and wondered as well if Kuhn’s search for signs of bureaucratic resistance might not reflect his will to believe that Confucian literati were somehow able to defend themselves quite readily against aroused imperial authority. Attributing a degree of autonomy to Ch’ing bureaucrats may be a mark of the functionalist “school” of eighteenth-century studies which Philip Kuhn has come to lead in American Sinology, and which stands opposed to an earlier historiography that stresses the overwhelmingly despotic nature of High Ch’ing rule.12

If heroes are to be found, Kuhn implies that they are among the bureaucrats themselves who emerge as principled opponents of Hungli’s fanatic campaign. Kuhn is not suggesting that the emperor’s councilors were a group that acted legally to check the arbitrary power of the monarch. But he does believe that “in certain extraordinary cases” officials could invoke a set of higher principles to restrain the “lash of state power.” However, only men “who believed themselves to be certified carriers of a cultural tradition” had the confidence to invoke that “superior code under which all human governments might be judged.” When the Confucian tradition foundered, the self-confidence of the elite sank; and—so we read by implication—the Chinese people lost their higher political guardians and face their present dictators defenselessly on their own.

Was that the contrast we were meant to see when Kuhn wrote his final paragraph, sometime after June 4, 1989?

Nobody mourns the old Chinese bureaucracy. The social harm it did, even by the standards of its day, went well beyond the crushed ankles of helpless vagrants. Yet its nature impeded zealotry of any sort, whether for good or for ill. Without that great sheet-anchor, China yaws wildly in the storm. Without a workable alternative, leaders can manipulate mass fears and turn them with terrible force against the deviants and scapegoats of our own day—anyone vulnerable to labeling, either for his social origins or his exotic beliefs—with none to stand between.

Though we may question Kuhn’s conservative alternative to the present state of affairs in China, and prefer the prospect of a democratic future to nostalgia for a bureaucratic past, Kuhn draws a telling contrast with contemporary tyranny.

Modern China has its sorcerers, too, and since the economic reforms began in 1979 there has been a visible recrudescence of religious sectarianism. The most acceptable form of this belief in soul-power is qigong, which superficially appears to be a set of breathing exercises to keep one fit. Actually, as taught by some qigong masters, the exercises draw in powerful popular sources of religious inspiration for the cultivation of physical invulnerability, trancelike possession, and cosmic correspondence. These sources are associated with the Chinese martial arts, certain biodynamic forms of traditional medicine, syncretic shamanism, tantric Daoism, and chiliastic sectarian Buddhism.

Yet because it is ostensibly just a form of physical conditioning, the nourishing of one’s gi (breath, pneuma, material force, energy) can be treated by the Communist authorities as a somewhat mysterious but properly materialistic way of increasing one’s life span. Indeed, qigong is frequently presented in official publications as a paranormal pseudo-science that cures cancer and helps people live longer. Fang Lizhi, the dissident physicist now in exile, once joked to me while he was still in China that it was quite appropriate for the regime to permit qigong to flourish because China was ruled by a handful of gerontocrats obsessed with their own longevity.

And flourish it did. More than fifty million people belong to the national qigong association, and visitors to China in the last few years have been struck by the large crowds of practitioners who gather around the masters and copy their movements, sometimes in the classically alienated way of soulpossession. These association meetings hardly represent a form of civil society that threatens to restrain the dictatorship. But it seems to me telling that these assemblies have now been prohibited in many parts of the country, and that the national qigong association has been declared a form of political party that will not be allowed to continue to exist. This is certainly not because Hungli’s soulstealers have somehow come back to haunt the land. Rather it is just one more sign that China’s present leaders have no tolerance at all for even the faintest specter of an autonomous sphere of public life.

This Issue

May 16, 1991