China’s Republican revolution was a classic example of a revolution of rising expectations. The Qing dynasty was overthrown in 1911–12 not because it was resistant to reform, as it had been in 1898, but because it was not reforming fast enough. In its last decade, following the foreign intervention to suppress the Boxer Rebellion, the incumbent regime under the leadership of the chastened empress dowager Cixi finally recognized the need for change. Known as the New Policies (Xinzheng), the reforms she initiated were so radical and far-reaching that they have been described as revolutionary in and of themselves. Thus, Douglas Reynolds speaks of a “Xinzheng Revolution” that preceded the Xinhai (1911) Revolution in time and perhaps exceeded it in results.1 As he and others have described them, the New Policies comprised educational reforms and study abroad; reorganization of the army, navy, and police; and administrative and political changes, including plans to transform the millennia-old autocracy into a constitutional monarchy. While they were nationwide in scope, the overall reforms had the greatest impact among China’s urban minority. The five million or so Manchus, as a predominantly urban population, were one of the groups most affected by the New Policies. Thus, new schools were founded among the banner people in Beijing and in the provincial garrisons. Manchus were among the first Chinese to go abroad to study in Japan. Parts of the Metropolitan Banners were reorganized into the First Division of the new national army and into the new Beijing police. Manchus were also elected to the provincial assemblies and selected for the National Assembly.
One principal consequence of the post-Boxer reforms, unanticipated and unwelcome by the Qing court, was the politicization of China’s urban elite, who ever more boldly demanded a greater role for themselves in their government, whether in the form of a constitutional monarchy or a republic. While few, if any, Manchus were adherents of republicanism, they were not lacking among the reformist critics of the regime. Some took part in the first large-scale popular demonstration of Chinese nationalism, the anti-Russian agitation of 1903; some published periodicals, such as Beijing Women’s News and Great Harmony Journal, that supported various reformist causes; and some joined protopolitical parties, such as Liang Qichao’s Political Information Institute, and participated in the lobbying campaign of 1909–10 for the immediate summoning of a parliament. Thus, the post-Boxer reform movement, unlike the revolutionary movement, was never exclusively Han or inherently anti-Manchu.
It was the inability of the late Qing court to meet the rising expectations that its own reforms had created that estranged the elite from the incumbent regime and allowed the revolution to succeed. The court’s shortcomings in this regard were numerous. The most egregious was its steadfast opposition to the growing chorus of demands for a “responsible cabinet” and an elected legislature. It finally did accede to those demands, by proclaiming the Nineteen-Article Compact, but not until three weeks after the revolution had broken out and only after some of its own generals at nearby Luanzhou had threatened to march on the capital; by then it was clearly too little, too late. No less critical to the alienation of the reformers from the court were the attempt by the Qing court to reimperialize authority and its failure to eliminate Manchu-Han differences.
Reimperialization was related to but distinct from the court’s concurrent effort in the late Qing to recentralize, to reverse the process of political decentralization that had greatly accelerated in the mid-nineteenth century, when the court had had to rely upon such provincial officials as Zeng Guofan and Li Hongzhang and their armies to suppress the Taiping and other rebellions. During the post-Boxer decade, the Qing, with some success, reclaimed some of the political and military authority that it had previously given up. Most notably, it managed, first, to strip Li’s successor, Yuan Shikai, of his command over four of the six New Army divisions that he had personally organized and to turn them over to the Ministry of the Army under Tieliang; then, to remove Yuan and his equally powerful colleague, Zhang Zhidong, from their provincial bases in Tianjin and Wuchang by transferring them both to the capital; and finally, after Cixi’s death, to dismiss Yuan from all his posts. Each of these steps served to strengthen the central government at the expense of the provincial authorities and other local interests. The nationalization of the trunk railroad lines in May 1911 had the same centralizing effect, and, as is well known, it provoked such strong opposition among the merchant and gentry elite in the provinces, especially Sichuan, that it led indirectly to the outbreak of the revolution downriver at Wuchang five months later.
Reimperialization, which was no less controversial, aimed to regain power not for the central government but for the Qing court itself. It seems to have been the main policy objective of Prince Regent Zaifeng, who evidently modeled himself after Kaiser Wilhelm of Germany. During the three years that he was in charge, Zaifeng ordered and oversaw the formation of an additional New Army division, the Palace Guard, which was under the direct command of the court rather than under the Ministry of the Army; he appointed his two brothers, Zaitao and Zaixun, to leading posts in the army and navy; he gave himself the title of generalissimo and took over as commander-in-chief of all the armed forces; he transferred control of the army’s general staff from the Ministry of the Army to the court; and when responding to the demands of the constitutionalists, he named a cabinet that was headed by Yikuang and included four other imperial princes as ministers. The political ascendency of these princes during Zaifeng’s regency was extremely unpopular, as the contemporary press and the debates at the newly convened National Assembly made amply clear.
While Zaifeng carried it to extreme lengths, the attempt to reimperialize authority did not begin with him; it had started half a century earlier, in 1861, when Yixin was named to the Grand Council. Yixin’s appointment is usually lauded as critical to the success of the Tongzhi Restoration, but at the same time it was a portentous departure from a dynastic tradition of more than a century’s standing—dating from the creation of the Grand Council in the 1720s and 1730s—that no imperial prince should be a council member or, by extension, a ministry head. It marked a return to the practices of the first century of the dynasty, when, under very different circumstances, imperial princes routinely participated in decision-making. Thus, beginning with Yixin in 1861 (aside from one brief break during 1901–3), one prince or another was to serve continuously as grand councilor until 1911. Furthermore, starting with the appointment of Yikuang as supervisor of the Ministry of Foreign Affairs in 1901 and that of his son, Zaizhen, as minister of commerce two years later, the mid-Qing ban on princes as ministry heads was also breached. These were the precedents, both of them set under Cixi, that paved the way for Zaifeng’s formation of the “imperial kinsmen’s cabinet,” with Yikuang as prime minister.
The outcome of the court’s efforts at recentralization and reimperialization was mixed. Reimperialization succeeded only in the short run and did not survive the revolution. Soon after the Wuchang uprising, the court was forced to beat a steady retreat. It removed Zaitao and Zaixun from their military and naval positions; it rescinded the appointment of Yikuang and the other princes as cabinet ministers, and, following a vote by the National Assembly, it named Yuan Shikai, recently recalled to office, to head up the new cabinet; it removed Zaifeng as regent; it conferred almost total decision-making authority upon Yuan and his cabinet; and last, at the time of the abdication, it transferred control of the Palace Guard from the court to the Ministry of the Army. On the other hand, though reimperialization obviously had failed, the cumulative effect of the parallel efforts at recentralization continued into the first few years of the new republic. They formed the institutional basis of Yuan Shikai’s presidency as a strong, central ruler. Only after Yuan died in 1916 did decentralization regain the upper hand.
Another shortcoming of the late-Qing court that contributed to the alienation of the elite and the success of the revolution was its failure to deliver on its oft-stated promise to “eliminate differences between Manchus and Han.” Such differences accounted for a substantial part of the republican revolutionaries’ indictment against the dynasty, but they figured in the reformers’ critique as well, such as Liang Qichao’s essay of 1898 in The China Discussion and the Manchus’ commentaries in Great Harmony Journal. They contradict the widespread belief, embraced by such disparate authorities as Mary Wright and Chiang Kai-shek, that the Manchus had become so assimilated to Han culture that by the late Qing they were no longer identifiably distinct from the Han; they also contradict the corollary of that belief, that the ethnic (or Manchu) issue was irrelevant to the 1911 Revolution. During the 268 years of Qing rule over China proper, the Manchus undeniably absorbed much of Han culture (even as they, to a lesser extent, had an indelible effect upon Han culture as well). After such a long period of sedentary idleness among the seductive pleasures of urban living surrounded by the multitude of Han Chinese, nearly all Manchus eventually turned their backs on the two core elements of the old Manchu way of life, Manchu speech and mounted archery. Yet the Manchus never entirely lost their identity as a separate people either in their own eyes or in the eyes of the Han. The institution of the Eight Banners remained intact down to the end of the dynasty and so did certain cultural markers, such as men’s names and women’s dress and natural feet, all of which continued to set Manchus apart from Han. It was upon the reality of these few but persistent differences that the revolutionaries based their propagandistic (and sometimes physical) assaults against the Manchus qua Manchus.
The late Qing court itself acknowledged the reality of Manchu-Han differences and tried repeatedly to do something about them. On three widely separate occasions—in 1865,1898, and 1907—it issued decrees that either abolished the occupational and residential restrictions on the banner people or called for the disbandment of at least the provincial banner garrisons. However, none of these three edicts (nor that of 1902 lifting the ban on Manchu-Han marriage) was ever implemented. The banner people, for example, did not gain their freedom to choose where to live and what to do for a living until after the Qing abdication. Here, too, the court’s failure was not the fault of Zaifeng alone; Cixi shared in the blame as well.
Yet, when the revolution finally occurred, the Qing court, notwithstanding its unpopularity, fared surprisingly well. Though its overthrow was never in doubt, the terms of its surrender were by no means unconditional or abject. The abdication agreement was a product of genuine negotiations between the court and the Republicans, with Yuan Shikai as intermediary. Consequently, Puyi’s postrevolutionary fate contrasts dramatically with that of other deposed monarchs and autocrats of his historical era. He was not imprisoned and executed, as were King Louis XVI of France a century before and Czar Nicholas of Russia a few years later. He was not forever banished from his home country, as was Sultan Mehmed VI of the Ottoman dynasty following the Turkish revolution of 1922–24. Nor was he even obliged to give up his palace and take up residence elsewhere, as happened to the last Tokugawa shogun, Yoshinobu, whose castle in Edo (Tokyo) was turned over to the Meiji emperor while he himself retired to Shizuoka.2 The Articles of Favorable Treatment for Puyi and his court, which was signed in 1912, remained more or less in effect for over a decade afterward, despite the vicissitudes of warlord politics in Beijing. He was not expelled from the Forbidden City until 1924. Even then, the rest of the tattered agreement was not abrogated until 1928, when the Nationalists’ desecration of the Qing Eastern Tombs went unpunished.
It was also not until the mid-1920s that the Eight Banner system of the Qing was totally dismantled, thus hastening the completion of the process by which the Manchus evolved from an occupational caste into an ethnic group. As is evident throughout this study, the question of who the Manchus were is greatly complicated by the fact that the single English term “Manchu” embraces a number of different Chinese terms, ranging from Manzhou, Manzhouren, Manren, and Manzu to qiren, qijiren, and qizu. Although these words all refer to the Manchu people, their meaning could and did vary—sometimes slightly, sometimes greatly—depending on the social or historical context. As the concept of the Manchus changed over time, it helped to define who the Han were as well.
During the Qing, the question of Manchu identity was inextricably interwoven into the history and structure of the banner system, resulting initially in three concentric circles of meaning. First, in the narrowest as well as the earliest sense, the Manchus were the Manzhou, descendants of the Jurchen, whose scattered tribes in what was later called Manchuria (also Manzhou in Chinese) were unified by Nurhaci beginning in the 1590s and organized by him in 1615 into the Eight Banners. In 1635 Hong Taiji bestowed upon them the name Manju, from which is derived the Chinese transliteration Manzhou. To distinguish them from later additions to the Manchu banners, they were also known as the Old Manchus (Fo Manzhou, in Chinese). They and their descendants were the core element of the banner system. Throughout the Qing period, in the eyes of successive emperors they were the most highly esteemed, most trusted contingent of the system. The emperor himself was always one of them. It was their ancestral language and their equestrian lifestyle that constituted the essence of the “Manchu way.” Second, in a broader (and slightly later) sense, the Manchus were the Eight Banner Manchus (Baqi Manzhou). These included not only the Old Manchus but also the New Manchus (Yiche Manzhou), the nearby Tungusic and Tungusized Mongol peoples who were added to the Manchu banners after the Qing invasion of China proper in 1644.
Finally, in the broadest sense, the Manchus were the banner people (qiren), particularly when they were being contrasted to the Han. It is clear from the pronouncements of the Qing rulers themselves that when referring to Man-Han, they usually meant by “Man” not only the Old Manchus nor even the Eight Banner Manchus but the membership of the entire Eight Banner system, which Hong Taiji had enlarged by creating two new contingents in 1635 and 1642 respectively, the former made up of Mongols and the latter of captured or surrendered Han Chinese. Thus, in this sense, the Manchus embraced all three ethnic components of the banner system—the Manchu banners, the Mongol banners, and the Hanjun—as well as the bondservants. By this formulation, the Manchus were identical to the people who hereditarily belonged to one or another of the twenty-four banners and as such were officially classified as banner people, while the non-banner people, the vast majority of whom were Han, were generally classified as civilians (min). Since the Qing emperors and their officials used these two sets of terms—Man-Han and qi-min—more or less interchangeably, therefore just as the banner people corresponded to the Manchus, so the civilian population were, practically speaking, synonymous with the Han.
If, as this study contends, the Manchus are best viewed as equivalent to the banner people, then they were, during most of the Qing period, not so much an ethnic group as an occupational caste. In origin, the banner people were a multiethnic group. Not only were they divided among the three ethnic divisions, but the Eight Banner Manchus were divided between the Old and the New Manchus, and the New Manchus were divided among different groups themselves. As a whole, the banner people lacked the racial, linguistic, and cultural homogeneity of the original Old Manchus. While subsequent Qing rulers made an effort to Manchufy the entire banner population by requiring them all to learn and absorb the “national speech and mounted archery” tradition of the Old Manchus, in the end what differentiated the Manchus as banner people from the Han as civilians was not ethnicity or (as some have stated) political status, but occupation.3 From the very beginning, the primary function of the Eight Banner system was military. The core members of the system were the banner soldiers, who provided the main fighting force for the Qing conquest of China and who thereafter were required to protect and maintain the Qing dynasty from their more than ninety garrisons in Beijing, in their northeastern homeland, and in various strategic centers within China proper and across China’s northern frontier. Except for office-holding (which was rare) and in some cases farming, bannermen were prohibited from other employment, so as to better hold themselves in military preparedness. In return, the banner soldiers received a stipend of money and grain from the Qing state. The banner system, however, included not only the banner soldiers but also their families as well as numerous other relatives and dependents. By the end of the dynasty, when the banner population totaled five or six million, the ratio of banner soldier to banner people was roughly one to twenty. Every banner person (whether soldier or dependent) belonged to a banner company and was subject to the jurisdiction of his or her company captain and the banner’s lieutenant-general. Finally, membership in the banner system was hereditary and had been closed off to outsiders soon after the invasion of China proper. Intermarriage with non-banner people was prohibited and seldom occurred. In short, the Manchus as banner people can be characterized occupationally as a hereditary military caste, similar to the samurai of Tokugawa Japan.
The transformation of the Manchus from an occupational caste to an ethnic group began during the watershed decade of the late 1890s and early 1900s, when a growing number of Chinese scholars and officials were compelled by the threat of foreign imperialism to begin to think of their country no longer as a cultural sphere but rather in political and territorial terms as a nation-state. Heretofore, Chinese thinkers had viewed both themselves and others largely in cultural terms. They themselves were the “inner, civilized” people originating from the Central Plain of north China who called themselves Hua and/or Xia; the others were various “outer, barbarian” peoples, such as the Yi to the east and the Di to the north. It was, as is well known, Liang Qichao who was most responsible for redefining the notion of China from a civilization (indeed, the one and only civilization) to a territorial state (one among many). As part of that redefinition, Liang also reconceptualized the Manchus as a racial group. According to his late 1898 essay on Manchu-Han relations, just as the yellow race and the white race were locked in a life-or-death “racial conflict,” so too were the Manchus and Han, two subgroups of the yellow race within China. When referring to the Manchus and Han, Liang pointedly eschewed the traditional cultural terms (such as Yi-Di and Huaren) in favor of two essentially new terms, Manren and Hanren. (Though obviously derived from the established combination Man-Han, their usage as separate words had been infrequent in the past.) Liang, furthermore, equated the Manren with the banner people, the “five million people” who for two centuries had “eaten without farming and been clothed without weaving.”
Liang Qichao’s social Darwinist redefinition of Manchus and Han carried the day. In the late Qing and early Republic, Manchus and Han were increasingly differentiated from each other along racial rather than—as in the past—either occupational or cultural lines, though the old concepts and the associated terminology lingered on. Thus, Wu Yue, the assassin of 1905, and Great Harmony Journal, the Manchu reformist publication of 1907, as well as the abdication agreement of 1912 all called the Manchus an ethnic group (zu or minzu). Yet the combined term “Manzu” (Manchu ethnic group), which clearly existed at the time, was rarely used as a reference for the Manchus. (It figured most prominently in the name of the Manchus’ main lobbying organization in the 1910s and 1920s, the Association for the Common Advancement of the Manchus.)4 Instead, the two most commonly used terms by which to refer to the Manchus in the late Qing and early Republic were “Manren” and qiren. The former, popularized by Liang Qichao, was favored by revolutionaries such as Zou Rong; the latter, by Qing administrators. Also, “Manren” seems to have been more popular before the revolution; qiren, or its variant qijiren (people of banner registry), more popular afterward. In addition to these terms, there were yet others in common use, particularly among the republican revolutionaries of the post-Boxer era. Chen Tianhua, for one, almost always used “Manzhouren,” which can be translated either as “Manchu people” (with “Manzhou” referring to the Eight Banner Manchus) or “people of Manchuria” (with “Manzhou” referring to China’s northeastern territory). Others drew upon China’s rich and ancient store of culturalist epithets. Thus, Zhang Binglin loved to vilify the Manchus as Donghu (Eastern Barbarians) and as lu (caitiffs), while others denigrated them as Dada (Tartars) and Dalu (Tartar caitiffs). In short, even as a consensus was emerging that the Manchus should be considered an ethnic group, there was as yet no agreement on what to call them.
As to who the Manchus were, the general view in the late Qing and early Republic was, following Liang Qichao, that they were synonymous with the total membership of the Eight Banners—in other words, the banner people. Thus, the revolutionaries’ seven-point indictment of the “Manchus” in the post-Boxer decade was, at heart, a condemnation of them as members of the banner system. This equation of the Manchus with the banner people lasted well into the early Republic, as did, of course, much of the banner system itself. Thus the Republican government, despite its recognition of the Manchus as an ethnic group, continued the Qing practice of classifying them as “banner people,” except in those cases where individual Manchus formally petitioned to have their status changed. (The change, though, was not from “Manchu” to “Han,” but from “banner person” to “civilian.”) This was why Manchus were so commonly referred to then as “people of banner registry.” Indeed, a group of politically active Manchus in 1914 even proposed qizu (banner ethnic group or “nationality”) as yet another name for themselves. Though it combined the old (Manchus as the banner people) and the new (Manchus as an ethnic group), the term did not catch on.
If there was no agreement about what to call the Manchus, there was no such problem with regard to the Han, who, following Liang Qichao, were almost universally known in the late Qing and early Republic as Hanren. For example, in contrast to their proclivity for labeling the Manchus as barbarians, the revolutionaries hardly ever utilized the correlatives Hua and/or Xia to refer to themselves; instead, they nearly always called themselves Hanren. (It is thus all the more surprising that when the Republic was established, they chose to use the traditional-sounding “Zhonghua” in its name.) As to who was meant by “Hanren,” here, too, Liang Qichao’s lead was followed. In his account of the racial competition within China, he focused on only two contestants—the Manchus and the Han. Since the Manchus were equated with the five million banner people, the Han were, by implication, the rest of China’s vast population, the four hundred million non-banner “civilians” of the qi-min dichotomy.
Finally, what was the proper place for the Manchus? Where in late Qing and early Republican China did they fit in? The concept of China among the republican revolutionaries in the post-Boxer decade was that of a unitary nation-state. Geographically, it encompassed China proper, where nearly all the residents were Han and where the Han should rule. As Chen Tianhua asserted, “China belongs to the Han people.” In a China so defined, the Manchus, simply because they were non-Han, had no rightful place; they should be, according to the manifesto of the revolutionary Alliance, “expelled.” As Zhang Binglin more than once intimated, the appropriate refuge for the Manchus was Manchuria, their ancestral homeland, where they could create a nation-state of their own, as the Japanese later on tried to do in their name. The Qing, on the other hand, regarded China as a multiethnic empire, of which Han-populated China proper was but one region. The Qing empire, in this view, also included Manchuria, Mongolia, Turkestan, and Tibet, all of which, along with China proper, owed equal but separate obeisance to the emperor. The peoples of these different regions were administered separately. Whereas the Han were governed by the regular civil bureaucracy and the Manchus by the Eight Banners, the Mongols, Muslims, and Tibetans were subject to the Court of Colonial Affairs. The Qianlong emperor, who was instrumental in subjugating Turkestan and bringing it firmly within the Qing empire, took particular delight in his self-image as ruler and unifier of the five peoples. During the post-Boxer decade, the Manchu reformers associated with Great Harmony Journal reiterated the Qianlong ideal when they sought to “unify the Manchus, Han, Mongols, Muslims, and Tibetans as one citizenry.”
By the time Qing rule was overthrown in 1912, however, the republicans had abandoned their earlier, narrowly defined view of “China proper for the Han Chinese.” They did not, for example, drive the Manchus back to Manchuria but generally allowed them to remain where they were. Instead, they came around to a modified version of the Qing’s expansive concept of China as a multiethnic state. Sanctioned by Empress Dowager Longyu’s abdication decree, the new regime claimed to be the legitimate successor to the Qing and hence to all of its far-flung territories and variegated populations. (Not everyone heeded its claim, as the Mongols in Outer Mongolia refused to transfer their allegiance from the Qing empire to the Republic.) The revolutionaries’ previous view of China as caught in a struggle between two groups, Han and Manchus, was replaced by the concept of the Republic of China (no longer the Qing empire) as composed of five peoples. This idea, recycled from the Qing, was symbolized by the five-bar flag of the early Republic. Moreover, it was explicitly spelled out in the abdication agreements, which referred to the four major non-Han peoples—Manchus, Mongols, Muslims, and Tibetans—as ethnic groups and promised that they would be treated on an equal basis with the Han people (Hanren). Such, at least, was the stated aim of the early Republic, even if the reality for the Manchus was quite otherwise.
Temporarily rejected by Chiang Kai-shek and the Nationalists, the early Republic’s ideal of China as a multiethnic state was resurrected by the Communists. Chiang and the Nationalists, successors to the republican revolutionaries, insisted that China was a unitary nation-state, populated by one undifferentiated “Chinese ethnic group” (Zhonghua minzu) that was essentially synonymous with the Han. Though they referred to the Manchus as Manzu, as if they were a separate entity, they viewed them as having been completely assimilated with “us Han.” The People’s Republic, on the other hand, proclaimed itself a multiethnic state, and the number of officially recognized ethnic groups ballooned from five to fifty-six as ethnographers differentiated the Chinese population more carefully, more precisely than before. For example, during the early Republic the sizable Muslim population, most of whom are concentrated in the northwest, were considered one ethnic group, the Hui. Under the Communists, they were divided on the basis of linguistic and cultural differences into ten separate groups, and the term “Hui” came to be reserved for the Chinese-speaking Muslims, who were now called “Huizu.” The Han, however, did not undergo a similar subdivision, as the regime refrained from recognizing as official ethnic groups such distinct groups as the Hakka and the Subei people. Instead, much like Liang Qichao, the Communists consider the Han, with 93 percent of China’s total population, one huge, undifferentiated mass of people.5
Like the Hui and unlike the Han, the Manchus too multiplied amoebalike under the Communists. What in the early Republic had been one people, based on common membership in the Eight Banner system, became six: the Xibe, Daur, Evenki, Oroqen, Hezhe, and, of course, the Manchus. As in the late Qing, ancestral membership in the Eight Banners is again the defining characteristic of the Manchu ethnic group, but it is now shared with five other groups. Though qiren is still used informally in some localities to refer to the Manchus, it has by and large been supplanted by the single term “Manzu,” thus ending the terminological confusion of the late Qing and early Republic. In short, the banner people (qiren) have completed their transformation from the hereditary military caste that they were during the Qing, to the broad but vaguely construed ethnic group whom the republican revolutionaries and others in the late Qing and early Republic called “Manren,” and finally to the narrowly defined, officially recognized Manchu ethnic group (Manzu) that they are under the Communists.