The whole system of accountability revolved around control of information. An official could not be punished for failing to prosecute a crime, if the crime was not officially known to have been committed. This absurdly simple fact underlay the strained relations between Throne and bureaucracy in the soulstealing case. Although the Ch'ing regime lacked a professional secret service, the Throne had "eyes and ears" in the provinces. Through such a private source the Shantung cases had emerged. Through such a source Hungli discovered that provincial officials had covered up the sorcery cases of the past spring. It hardly matters whether the cover-up had resulted from official scorn of popular superstition (elite agnosticism about sorcery) or whether the tonsure offense hinted by the Soochow and Hsu-k'ou-chen affairs seemed, to provincial officials, too hot to handle. Hungli believed it happened because his officials feared awk wardness: court cases that might disrupt their comfortable routines, impeachments that might rend their patronage networks. We may forgive his suspicions: for a full two months, not a single provincial bureaucrat-whether Manchu or Han-had brought sorcery to the Throne's attention on his own initiative. Those suspicions were confirmed by the initial responses to his demands for information and action.
Hungli's Provincial Bureaucracy
How effective were Hungli's provincial prosecutors? China's civil provincial bureaucracy in 1 768 was a tiny elite corps of only sixtythree men. I use "provincial bureaucracy" in a special sense: men with duties on the provincial level or higher. Such duties included (i) comprehensive responsibility for the affairs of one province (the governors) or two provinces (the governors-general);' (2) specialized posts on the scale of one province (the provincial treasurers and provincial judges); and (3) specialized posts of a nonterritorial nature (such as the superintendents of the Grand Canal and Yellow River). Here, surely, was the world's most exclusive club. To enter it was to join the monarch's circle of special trust and personal communication. Its members (like officials of ministerial rank in the capital) were distinguished by being allowed, indeed required, to communicate directly and confidentially with the Throne. Although it lacked the old-boy co-optation and leather-chair coziness of a "club" properly so-called (that is, its internal cohesion was weak), membership gave a man a special self-image that transcended the world of the lower bureaucracy.2
The Men in Charge
To prosecute this case, Hungli depended on an experienced group of middle-aged and elderly men, each with a record of service in several provinces, and who had been (at their upper ranks) serving in the provincial bureaucracy for about a decade.' Manchus made up a disproportionately large fraction of the group as a whole (38 percent), and an even larger fraction (58 percent) at the top levels (governors-general and governors). For the provincial bureaucracy as a whole, this was a substantial ethnic shift from the beginning of Hungli's reign: since 1736, the Manchu component had increased by 84 percent; those cultural hybrids, the Han bannermen, by 33 per cent (though the number was so small as to be insignificant); and the ordinary Han had decreased by 51 percent. The composition of the governors-general remained almost unchanged, but there was a substantial increase in the proportion of Manchus in governor's rank (mainly at the expense of the Han bannermen), and another substantial increase of Manchus (at the expense of Han) in the post of provincial treasurer. More Manchus were working their way into provincial office by the standard route of promotion from provincial judgeships. This might be expected as Manchus became more sinicized and better at handling the affairs of the largely Han lower bureaucracy. But it also suggests a deliberate imperial policy to increase the representation of ethnic Manchus in local government.
By the standards of the eighteenth-century world, Hungli's province chiefs in i 768 ruled immense populations. In the three-province jurisdiction of the Liangkiang governor-general lived more than seventy million people, a population more than twice that of France at the time. The governor of Kiangsu, the largest province in this group and the most populous in the nation, ruled perhaps thirty million, at least triple the contemporary population of the United Kingdom. Even the smallest province involved in the soulstealing case, Shensi, contained some eight million people, which was roughly the size of Great Britain, minus Scotland.4 Clearly the tautness of such a ship had to be far short of what we (and the Chinese) are used to in modern bureaucracies. Chiefs of such immense societies had to be left substantial discretion. Yet the vast powers of these men were kept in check by numerous institutional safeguards. Moreover, collusion among them was balked by the confidential palace memorial system, in which a man never knew what was being reported to the Throne by his colleague next door. Finally, the emperor kept his province chiefs on the move by frequent transfers.5
Frequent rotation meant that province chiefs could hardly be deeply knowledgable about the special problems of their jurisdictions, and that much of the work of governance devolved upon the permanent staff of clerks.6 Each member of this highly mobile elite corps was kept busily circulating among provincial capitals, yet was bound to the imperial center by two iron cords: the personnel dossier in the Board of Civil Office that (as was the case with all officials) bore his cumulative record of promotions and demotions; and, more important, his personal tie of clientage and obligation to the monarch.
Although we are accustomed to calling provincial officials "bureaucrats," in an American political context these men would surely be called "political appointments." Although most had followed standard tracks into the provincial bureaucracy (either from a circuitintendancy, which oversaw several prefectures, or from a junior vicepresidency of one of the Six Boards), their elevation to provincial rank immediately signaled a special relationship to the Throne, a relationship marked by powerful rituals of loyalty and dependency. From such favored servants, the emperor expected not just reliability, but zeal: not merely to report accurately on local events, but to go the extra mile to further his royal objectives. Upon these qualities, more than upon the routine dossiers maintained by the Board of Civil Office, hinged the special trust and favor by which their future careers would he governed.
How did this special relationship affect the performance of a governor's job? The aspect of his job that most concerns us here is the insidious linkage at the heart of Chinese law enforcement: a territorial official was simultaneously policeman, prosecutor, and judge. Indeed, in every jurisdiction, the judicial power was simply an aspect of the executive. At the county level, the magistrate was in charge of arrests, prosecutions, and trials. Cases that carried penalties heavier than Hogging were referred upward to the governor's court, and sentences in all capital cases were automatically reviewed by the emperor. Because sorcery in many of its guises was classed by the Ch'ing Code as a capital crime, soulstealing cases might be expected to make their way to provincial courts, and ultimately to Peking. The governor had to exhort his subordinates to scour the counties and prefectures for sorcerers and then judge the cases of those they caught.
As Hungli turned up the pressure to crush the "evil arts," the governor-as-prosecutor loomed larger than the governor-as-judge. Most provincial officials in 1768 had some legal experience, generally obtained in the post of provincial judge through which they had entered the provincial bureaucracy. Yet few were legal scholars or administrators of wide reputation,' and I wonder whether many of the others had judicial self-images firm enough to withstand the political pressures inherent in their positions. Pleasing the monarch was a central part of administering the law.
The Communication System
Making government work effectively required carefully organizing the flow of information. From the point of view of the eighteenth century monarchy, this involved two problems: (i) separating routine from urgent business, so that problems could be handled at the appropriate levels and in the right order; and (2) ensuring that reports from field officials were timely and accurate. Neither problem was ever solved to Hungli's satisfaction."
To deal with the first problem, Hungli had in
herited from his predecessors a documentary system with a routine track and a confidential track. Ordinary matters were communicated by routine memorials (t'i-pen) through the Grand Secretariat, a committee of the highest ministers that oversaw the operation of the Six Boards (the ministries of the six traditional functional divisions of government). Through this channel flowed regular reports on tax receipts, criminal justice, public works, and routine personnel matters. The forms of these documents were rigidly prescribed, and irregularities were grounds for impeachment. Although today's social historian finds in these "routine" memorials the very pith of everyday Chinese life, the medium was ill suited to urgent matters that required speed and confidentiality, including such matters as sedition.
Such sensitive or high-priority information followed the confidential track: a direct, personal line of communication between every high provincial official and the Throne. Upward documents, carried either by the sender's personal servant or by military courier, reached the emperor's desk quickly and discreetly. His Majesty would normally brush his comments or instructions in vermilion ink directly on the memorial, then dispatch it back to its sender. These sealed missives we call "palace memorials" (tsou-che), because they were later preserved in the palace.9
The "palace memorial" was a personal document. Besides urgent local affairs, these documents dealt with matters growing from an official's personal relationship to the monarch.1) The form was simpler (for example, the elaborate expression of the sender's complete official rank was omitted in favor of a simple statement of the post he held). An exchange in the confidential track exuded reciprocity: the sender was expressing loyalty and gratitude by imparting confidential intelligence to his master; the monarch, in turn, replied with a stern (but occasionally also warm) paternalism. The routine memorial was shaped by bureaucratic form, the palace memorial by interpersonal ritual. The routine memorial communicated office to office, the palace memorial man to man.
The emperor's responses to official reports, as well as initiatives taken on his own, also followed both routine and confidential chan nels. The routine response might be no more than pro forma approval of a rescript drafted by the Grand Secretariat, perhaps instructing that one of the Six Boards take action, or perhaps simply that the information be filed. A more portentous matter or a normative pronouncement might be communicated by an "open edict" (ming-fa shang-yu), which was sent to every jurisdiction in the empire for ceremonious posting. The confidential response was nearly always, in the first instance, a "vermilion rescript" (chu-p'i), an instruction or comment brushed personally by His Majesty on the sender's memorial. The memorial, bearing the royal response, was then returned to its sender, generally through the emperor's powerful privy council, the Grand Council. Sometimes a returned memorial might be adorned with vermilion in many places, as the monarch responded to particular points by writing between the lines. A more formally organized response was drawn up as a "decree" (chih), which the Grand Council drafted at his instructions after having considered the original memorial, and then dispatched to the field as a "court letter" (t'ing-chi or tzu-chi). The "open edict" was a general message to the bureaucracy as a whole; the "vermilion rescript" and "court letter" were swift, confidential, and precise action-documents designed to instruct or admonish particular officials.
For understanding the emperor's own role in the soulstealing crisis, and indeed in Chinese politics in generall, the key is vermilion. Rescripts on memorials show us his instant, unmediated responses to reports from the field. And although court letters were drafted by those lofty ghostwriters, the grand councillors, the monarch always checked the final copies and often added his own remarks and editorial comments in vermilion. The amended version then went out to the field. The recipient was thus made aware of what points His Majesty considered particularly important, and the vermilion personal touch reminded him that the court letter as a whole faithfully reflected the imperial mind.
Cover-up in Kiangnan
Some Embarrassing Discoveries
The Liangkiang governor-generalship, which controlled the provinces of Kiangsu, Kiangsi, and Anhwei, was the empire's richest and most demanding provincial post. At its core was the Yangtze Delta region, which, along with part of neighboring Chekiang, was the Kiangnan spawning ground of soulstealing. In this sensitive post, Hungli was served by the eminent G'aojin, a master of river conservancy, who was sixty-two at the time of the soulstealing crisis. G'aojin was nothing if not well connected: a member of one of the upper three banners (His Majesty's own), his Han ancestors had served for generations as bondservants of the imperial household. This Man- chuized Chinese was nephew of a grand secretary and first cousin of an imperial concubine (which connection had led to the emancipation of his lineage by imperial decree). Rather than follow the usual banner-insider's route to high office, G'aojin had started his public career as a lowly county magistrate at age twenty-nine and only reached his first provincial-level post fourteen years later."
The governor-general had ample reason to feel secure in offering a bland response to his master's inquiry about sorcery in Liangkiang. In early August, he wrote that indeed he had heard rumors of soulstealing in Chekiang while temporarily in Soochow as acting governor the preceding spring. Local officials told him that the rumors came from the Hangchow area, and that Kiangsu itself had experienced no queue-clipping outrages. Once the rumor-spreaders were arrested and the spreading of rumors prohibited, the problem vanished. But Hungli's vermilion rejoinder showed that he believed not a word: other provinces had reported queue-clipping, so "how can Kiangnan alone have none?" The Kiangnan bureaucracy was "substandard," and its "practice of making something appear to be nothing is really hateful."12
The monarch now turned his ire on G'aojin's subordinate, Governor Jangboo of Kiangsu, where the Shantung confessions had revealed that several important sorcerers were hiding. An experienced Manchu official and wily bureaucratic infighter, Jangboo had risen steadily through the provincial ranks, gaining Hungli's trust for his effective prosecution of a Shansi corruption case of 1766.13 Another corruption scandal confronted him when he moved into the governor's yamen at Soochow in the spring of 1768. This time the trouble was in the Yangchow salt administration. The most prominent official culprit, G'aoheng, was embarrassingly well connected: his first cousin was none other than Jangboo's superior, G'aojin, and his sister was the imperial concubine whose charms had won freedom for Gaojin's lineage." It was just as Jangboo was prosecuting this awkward case that the hunt for soulstealers began in earnest. Governor Jangboo quickly found himself the target of heated court letters from Peking.
The latest Shantung intelligence (confessions of beggars Ts'ai and Chin, complete with names of ringleaders) had been distributed to all high provincial officials in the east. By mid-August, however, it was apparent to Hungli that his province chiefs were not following up leads to the chief sorcerers, monks Yu-shih (Kiangsu) and Wuyuan (Chekiang). Though numerous arrests had been made in Chihli and Shantung (Funihan had now caught five more queue-clippers), the hotbed of soulstealing in the Yangtze provinces had yielded no culprits.
To his embarrassment, Jangboo now had to admit that certain "rumors" about queue-clipping sorcery had seeped across the Chekiang border the previous spring. He had seen no need to report them, because his "investigations" turned up no evidence that anyone had actually been clipped. In early August, however, the plot thickened, as reports came in from northern Kiangsu districts near the Grand Canal. Back in the late spring, in the county of An-tung, a certain Liu Wu had clipped the queue of a man named Tsou and was now in official custody. In a P'ei-hsien market crowd, a Shantung man named Yao was reported to have "bumped" the mother of one Yang, causing the lady to feel "dizzy." And in P'i-chou, a man named Wang had hidden in some bushes, accosted the wife of Ch'iu Ta- feng and clipped a piece of cloth from her lapel. Of the last two criminals, the first had been beaten to death by the crowd, the other hounded to suicide. The surviving culprit was a wily rasca
l: Liu Wu had convinced county authorities that he had clipped the queue only so he could cut purses during the ensuing uproar. Governor Jangboo assured the emperor that he would interrogate him "personally." He was also rushing agents to Hai-chou to intercept the master-sorcerer Ming-yuan, who was scheduled to turn up there August 26, according to the confession of his apprentice, Harr P'ei-hsien. He had also instructed local officials to be on the watch for the master-sorcerer Yu-shih, who was hiding in Su-chou across the Anhwei border, according to Chin Kuan-tzu's confession, lest he try to enter Kiangsu.'5
Hungli snapped back that Jangboo had done "extremely improper" work: how could county officials have relied on cutpurse Liu's "slippery confession" and put such an important case on ice for several months? (Vermilion: "How could you have failed to impeach such a refractory subordinate?") And if the spring rumors had been followed up rigorously, the "little people" would not have had to lynch the culprits but would have reported them to officials, as in Shantung. The gap in quality between the administrations of Shantung and Kiangsu was all too plain. jangboo's failure to cross the Anhwei border in pursuit of Yu-shih was further evidence of bureaucratic laxity: although in ordinary criminal cases hot pursuit across provincial boundaries might be thought excessive, how could it be so in a case like this?'s
On an encouraging note, Jangboo cheerfully informed his master that the corruption case in the Yangchow salt administration had yielded clear evidence and would soon be solved. Not the least mollified, Hungli hectoredJangboo on priorities: salt administration is only "one of the normal affairs of local government. Moreover, once it is dealt with, there's an end to it, and it should not be unduly troublesome. But if criminals conceal themselves and carry out their evil plots, troubling the villages, the damage to people's lives will be great." Jangboo seemed to have "reversed the serious and the trivial." 17
Soulstealers: The Chinese Sorcery Scare of 1768 Page 14
