ON JULY 1, 1911, YUAN DAHUA, THE GOVERNOR OF XINJIANG, sent a telegram to the central government in Beijing. The topic was empires and their modes of governance. “In administering their interior and exterior territories,” Yuan wrote, “foreigners do not apply the same rubric of rule.” The newly minted constitution of his very own Great Qing Empire, he noted, emulated that of Japan. But “the system of rule that Japan applies to its interior is not the same as that which it applies to Korea or Taiwan, and the system Japan uses to rule its prefectures and counties is not the same as that applied to Hokkaido.” Why are there such differences? Yuan asked. “Because a system of rule must be suitable for that locality,” he said in answer to his own question. “It cannot be otherwise.” The point of Yuan’s memorial was that the Manchu court in Beijing should emulate the governing tactics of neighboring empires that Yuan considered to be more advanced. He concluded, “Xinjiang is to China as Hokkaido is to Japan.”1
This analogy resonated strongly with Han officials in Xinjiang. After all, they ruled over a vast Central Asian land of desert and steppe, home to a Mongol and Turkic demographic constituting more than 90 percent of the population. Just five years later, Yuan’s successor, Yang Zengxin, felt compelled to remind Beijing of the virtues of learning from its rivals. Xinjiang, he noted somewhat more insistently, “absolutely cannot be managed in a fashion similar to that of the inner provinces.” Yang emphasized that he, as governor, would be “reluctant to confine myself only to methods that are prescribed for China proper.” Once more, the example set by Japanese administrators in their expanding empire served as a model for frontier officials in China. “In light of its special circumstances,” Yang asked, “can we look into the possibility of allowing Xinjiang to be governed in the same way that Japan governs Hokkaido and Taiwan? Or perhaps in the accommodating manner that the Qing once stipulated for special administrations?” Yang’s choice of words, some drawn from precedents set as early as the Tang dynasty (618–907), traced a consistent logic. Because the non-Han frontier evinced cultural, ethnic, economic, and political norms strikingly different from those in the Han heartland, “special conditions” (tebie qingxing) mandated adherence to a form of “loose rein” (jimi) governance imbued with the “spirit of accommodation” or “flexible arrangements” (biantong).2
During the first half of the twentieth century, Chinese officials in Xinjiang had good reason to question the wisdom of exporting to the Muslim frontier policies originally designed for the Han heartland. Incorporated into the Manchu empire by military conquest in the mid-eighteenth century, Xinjiang nonetheless remained an attractive military and economic target for states and empires along its western and northern borders. It shared with the peoples of these regions extensive linguistic, economic, cultural, ethnic, and religious ties. By the late nineteenth century, all these neighboring lands and peoples had become territories and subjects of the Russian empire, which knitted together its far-flung domains via the Trans-Siberian Railway. In 1884, owing both to the Russian threat and to the brief rise and fall of an independent Islamic state in Xinjiang, the Qing court decided to consolidate its rule in the northwest by turning Xinjiang into a province (sheng), as distinct from its former designation as an outer dependency (shudi). In this new provincial bureaucracy, Han officials with civil service examination degrees began to replace the longstanding military aristocracy of Manchu and Mongol bannermen, who had once regarded all the non-Han peripheries of China as their unique political preserve.3
Regardless of whether the officials in Xinjiang were Manchu, Mongol, or Han, however, they all confronted more or less the same problems along the non-Han frontier, be it Mongolia, Tibet, or Xinjiang. By the late nineteenth century, rival empires across Eurasia had begun to leverage their superior economic and military weight to siphon wealth and resources away from Beijing and to reorient the political and economic networks of the Qing empire toward their own metropoles. In response, during the last decades of the Qing dynasty prior to the 1911 revolution, the Manchu court undertook a vigorous program of reform. Just three years after Xinjiang became a province, so too did Taiwan, and that was followed by the carving of the Manchu homeland into what would eventually become the three northeastern provinces of the early twentieth century. Many previously closed frontiers were opened to Han migration, non-Han elites found themselves grossly in debt to Han merchants, and large contingents of modernized New Armies, staffed almost entirely by Han, were stationed outside the inner provinces. In Xinjiang, new Chinese-language schools encroached on institutions of Muslim pedagogy, much to the dissatisfaction of native elites. Most upsetting, however, was the ubiquitous imposition of new taxes to pay for the government’s new policies, which for many felt like a form of sinification.4
The backlash visited on the Qing court and its Republican successors as a result of these reform efforts conformed to a political mold unique to the twentieth century. As the discourse of the nation-state increasingly lent itself to opportunistic application in the realm of geopolitics, Chinese elites quickly became aware of the ways in which discontent on their own non-Han borderlands could breed results drastically different from those engendered by discontent in the heartland. The last time Xinjiang had been wrenched from Beijing’s control, for about a decade in the 1860s and 1870s, Beijing’s political rivals in Central Asia could only produce a single slogan around which to rally the Turkic natives against Qing rule: holy war.5 After the fall of the Manchus in 1911, the exclusionist discourse of the nation-state, premised on the novel ideal of ethnic self-rule, provided a new platform from which rival political elites could challenge Chinese rule. Yang Zengxin, governor of Xinjiang for seventeen years after the 1911 revolution, made clear what was liable to happen if China proved unable to keep the foreigners at bay. “Let us imagine,” he wrote to Beijing in 1917, “that trouble was to break out in the southwestern provinces today. Eventually they would be reunified. But if there is even the slightest slip in Xinjiang, so important for national defense, it will fall under the control of a foreign power. Even if we then had an eternity, we would not be able to recover it.”6
What Governor Yang was attempting to describe constitutes the narrative focus of this book. The creation of a putative “nation-state” by one empire on what was intended to become the former territory of its imperial rival was merely one permutation of a much larger phenomenon: the manipulation of ethnocultural difference. In the twentieth-century political history of Xinjiang, the only majority non-Han provincial-level unit of China never to have slipped from the hands of Han officials, we find the most complete case study for an interrogation of the legacy and inheritance of empire in the eastern Eurasian mainland after 1911. “Strategies of difference” consistently marked Xinjiang as a unique part of the Chinese state, a place where the normal rules of political discourse in the Han heartland ceased to apply. An extended analysis of what Han officials said and did in Xinjiang—as opposed to what Han intellectuals or politicians on the eastern seaboard merely said about Xinjiang—will thus bring us close to an understanding of the imperial foundation upon which modern Chinese political discourse actually operates.7
What are strategies of difference? The pronouncements of Governor Yang are instructive on this score. During his seventeen years (1912–28) in office, Yang made a habit of announcing surprisingly modest goals for himself. In 1925, he proclaimed as his highest ambition nothing more than the hope that “Xinjiang should not be delivered from our hands during my lifetime.” Elsewhere, Yang would gloss this sentiment in more dramatic terms, declaring that only his policies would prevent Xinjiang from becoming “the next Outer Mongolia” (wai Meng zhi xu), “the next Tibet” (Xizang zhi xu), or “the next Urga” (Kulun zhi xu), the capital of Outer Mongolia. After the creation of the Japanese proxy state of Manchukuo in the 1930s, “the next northeast” (dongbei zhi xu) also entered the political lexicon of Chinese elites who concerned themselves with the fate of Xinjiang. As for the rest of the country, Yang occasionally voiced his hope that “two or three great and powerful men will rouse their consciences, get rid of predatory armies, and implement policies beneficial to the people.” He did not, however, expect that any such savior would come from the non-Han borderlands. As governor of Xinjiang, Yang’s job was to sit tight and “wait for the day when the central plains have calmed down.” In the meantime, he wrote in 1919, “if I am able to maintain one portion [of the country], and the [central] government is able to maintain another, is this not for the best?”8
The novelty of Yang’s words becomes apparent only with the realization that the deployment of this type of political discourse was largely restricted to those areas of China evincing the requisite ethnic, cultural, and linguistic criteria capable of allowing rival imperial elites to engage in strategic manipulation of the “politics of difference.”9 Unless a foreign power (such as Japan) actually invaded the Han heartland, a warlord who claimed as his highest ambition the preservation of Chinese sovereignty in Hunan or Jiangxi would have invited ridicule and scorn. For men like Yan Xishan, Feng Yuxiang, and Wu Peifu, some of the most prominent warlords of the inner provinces, such an “achievement” merely marked the starting line. They were further obliged, in a way Han officials in Xinjiang were not, to strive for “national salvation” (jiuguo), a task made impossible by the geopolitical conditions of the day. In other words, far out in Xinjiang and any other non-Han borderland of which the Chinese state might find itself steward, local officials could employ the politics of difference to support their bid for wealth and power in a way that officials operating in the heartland could not.
Governor Yang, aware of this, took the unusual step of publishing, while still in office, several thousand telegrams culled from his own government archives. Records from the Studio of Rectification (Buguozhai wendu), issued in handsome thread-stitched bound sets and available for purchase throughout the country, effectively showcased the governor’s novel approach to self-legitimization on the non-Han frontier. The flip side of all this was that Yang’s rivals, usually but not always foreign, could just as easily turn the politics of difference against him, provided they deployed such discourse in support of their designs on Xinjiang. Han officials in the heartland, though obliged to strive for the impossible goal of national salvation—and then watch as the inevitable denouement sapped the morale of all but their most ardent enthusiasts—generally did not have to worry about hostile charges of Chinese “colonialism” or “imperialism.” Run-of-the-mill “autocrats” and “despots” they might be, but at least they were not characterized as chauvinist Han nationalists who “hated” their culturally alien subjects, as Russians officials once alleged of Yang. Viewed from this perspective, it should come as no surprise that Yang went to extraordinary lengths to select and publish for public consumption only those telegrams showing his manipulation of the politics of difference, while omitting entirely the hostile versions put forth by his many rivals.
This study is concerned not only with the abstract discourse of difference but also with the very real institutions of difference that continued to undergird Chinese rule in the former Qing imperium, long after the words empire and monarchy had succumbed to charges of illegitimacy. It is clear that Han officials in the new Republican state consciously revived selected rituals and modes of discourse from the Qing empire that had originally been designed for interaction with non-Han nobles, a tactic elsewhere described as “going imperial.”10 To probe the relevance of such a framework in Xinjiang during the first half of the twentieth century, it is necessary to shift our perspective from Beijing and Nanjing all the way to Urumchi,11 the capital of Xinjiang and home of its Chinese governors. The perspective of Han officials resident in Urumchi is indispensable for this task, for the simple reason that no central government of the Republican era (1912–49) ever succeeded in enforcing its writ throughout the province. Indeed, most of our evidence will come from the approximately four decades in which Han officials held the reins of government in Urumchi but, owing to the chronic disunity and fiscal poverty of the inner provinces, found themselves perennially threatened with political disenfranchisement. Almost always, this latter threat came at the hands of foreign powers and domestic warlords peddling their own versions of the politics of difference.12
Thus, the contest for wealth and power in twentieth-century Xinjiang—and by extension, the rest of the non-Han periphery—was mediated both in name and in deed through the strategic manipulation of difference. That there was still such difference to exploit underscores the growing realization among scholars that the twentieth century in no way marked the transition from “empire to nation-state,” in China or elsewhere. Yet if China did not become a Han nation-state, what did it become? Should we simply continue to refer to it as an “empire,” or is some sort of intermediary term (e.g., imperial formation) more appropriate?13 My analysis suggests that the phrase national empire—related to but distinct from the idea of the Soviet Union as an “empire of nations”—appears to be the most suitable label.14 Even some late Qing intellectuals coined a strikingly similar phrase—minzu diguo—in Chinese.15 The notion of difference is now regularly acknowledged as an indispensable feature of empires, and studies of the Ottoman, Russian, French, and other empires throughout history have all deemed the flexible institutionalization of ethnic and spatial difference as a defining characteristic of the largest and most enduring multiethnic states. In its most condensed form, the “politics of difference” is little more than the presumption “that different peoples within the polity will be governed differently” and that “distinction and hierarchy”—or the “non-equivalence of multiple populations”—will be maintained even as the state incorporates new people.16
Key to this interpretation is the distinction between empire as a type of state and empire as a type of power. As a type of power, empire is usually associated with an emperor or other absolutist monarch, who attempts to concentrate wealth and power in the hands of an exclusive elite whose ranks are often hereditary and usually closed to the mass of commoners who make up the taxpayers of the state. In its more lax and often popular usages, this interpretation of empire is almost always negative, concerned as it is with the use of brute military force, an elite dictatorship, and suppression of subaltern peoples. From this enduring paradigm, scholars and pundits have coined the phrase divide and rule, a concept usually imagined as the guiding ideology of a Machiavellian imperial elite. The idea of empire as a type of state, however, derives much of its explanatory power from scholarship conducted over the past several decades on the major Eurasian land-based empires: Ottoman, Russian, and Manchu.
The “new Qing historians” have shown the ways in which the Manchus forged an expansive empire during the seventeenth, eighteenth, and nineteenth centuries, one that drew its wealth from the Han heartland but was rooted in institutions of ethnic and spatial difference throughout East and Central Asia. Historians of China, however, have generally stopped short of adapting such insights toward the twentieth-century successor states of the Qing. In tracing this Eurasian legacy of political difference into the twentieth century, Soviet historians have done the most to complicate the enduring paradigm of empire as a type of power. In demonstrating how the Soviet state was more a “maker of nations” than a “breaker of nations,” they have provided historians of modern China with the analytical tools necessary to reconsider the imperial legacy bequeathed to twentieth-century Republican and Communist administrators along the non-Han borderlands.17 This in turn enables us to interrogate the ways modern China—as a type of state—still resembled its imperial predecessors. The study of ethnic classification projects in the southwestern province of Yunnan during the 1950s, for example, has illuminated our understanding of the modern minzu (or minority nationality) regime as well as the imperial legacy in postrevolutionary China.18
The present study provides a similar analysis of Xinjiang and much of twentieth-century China. In the interests of analytical precision, the word empire here will refer to a type of state that creates, maintains, and ultimately institutionalizes ethnic and spatial difference over a vast swath of territorial holdings. The actual exercise of power by Han officials in an ethnoculturally alien land will be captured by reference to strategies or politics of difference. Every effort will be made to avoid deploying politically charged and highly subjective terms such as colonialism and imperialism, unless they form part of the discourse of historical actors. In evaluating the increasingly liberal deployment of provocative words such as imperialism in the scholarly literature, Bernard Porter, a historian of the British Empire, writes that “currently there is no general agreement over what the word means or covers,” and that “its negative connotations can make it difficult to use dispassionately or truly analytically.”19 In a very basic sense, the building blocks of empire are conscious performances of ethnocultural difference within a political arena. To describe something as “imperialist” or as the embodiment of “imperialism” is almost always to pass a negative moral judgment upon that performance.
In an attempt to drop an empirical anchor somewhere in this sea of partisan discourse, this study will confine its analysis of colonialism and imperialism to the contextual recognition of when and where it becomes rhetorically possible to affix one or both of these terms to the modern Chinese state. Less important than the ontological reality of such subjective and pliable terms is the recognition that only certain parts of the former Qing state were, in fact, suitable for their deployment by political elites. In practice, this meant only those places marked by a substantial “alien” ethnocultural demographic that lent substance to the definition of empire as a type of state built on the notion of difference. This study is not concerned with identifying or highlighting “victims” of imperialism, to the extent that such a phenomenon can actually exist outside the hypothetical abstractions of nation-states. Rather, it aims to identify and analyze the conditions under which rhetorical claims of ethnocultural victimhood can be made to sound plausible. For example, despite the fact that the vast majority of Han and Uighurs living in China today face similar restrictions in political, religious, and economic affairs, scholars do not generally refer to the “oppressed” or “restive” Han of Zhejiang or Anhui. Such “disenfranchised” Han are simply referred to as “exploited” but unmarked “people,” “peasants,” “prostitutes,” and so on. But once the ethnocultural difference of the Uighurs or Tibetans has been foregrounded, near identical forms of “oppression” are habitually described as “colonialist” or “imperialist” in nature.
During the twentieth century, the presence of ethnocultural difference provided fertile ground for the introduction of a novel and unprecedented threat to extant paradigms of difference in East Asia: national determination. How did Han officials in Xinjiang respond to what they perceived as the nationalist threat? Despite a growing interest in Uighur unrest among mainstream intellectual and media circles, there is at present no historically informed interrogation of this important question from the perspective of those who exercised paramount power in modern Xinjiang: the Han ruling class. In part, this is due to a long-standing methodological bias within modern Chinese history. Because modern China is often portrayed first and foremost as a victim of Western and Japanese imperialism (both in the narratives of its own political elites and in those of foreigners), historians have been slow to recognize that not every significant region of ethnocultural difference managed to slip from Chinese control after 1911. Despite the permanent loss of Outer Mongolia, the lengthy estrangement of Tibet, and the brief emergence of the Japanese puppet state of Manchukuo, fully one-sixth of the presentday People’s Republic of China remained subject to a handful of Han rulers who were grossly outnumbered by their non-Han subjects. That their writ continued to be observed in Xinjiang without break throughout the twentieth century could not be said of Han officials in Manchuria, Taiwan, Shanghai, or Beijing. And yet these latter locales are often called upon within the scholarly literature to serve as representative microcosms of the “Chinese nation-state” in toto.
Scholars of all persuasions will be better served by the recognition that twentieth-century China was much more than simply a nation beset by imperialists. Certainly, from the perspective of a great many residents of the Han heartland, this was an apt characterization, and scholars who study such communities exclusively can likely rest content with this framework. But when we shift our perspective from Shanghai to Urumchi, the historian of modern China encounters a case study substantial and continuous enough to illustrate the ways Chinese political elites continued to engage the rhetoric and realities of empire on the ground throughout the twentieth century. Owen Lattimore recognized as much during his travels along the northern borderlands in the 1920s, noting that “while the Chinese could rightly claim that they suffered from foreign imperialism, there was also a second level of Chinese imperialism against the Mongols.”20 The words and deeds of Han officials resident in Xinjiang constitute the missing link necessary to portray twentieth-century China first and foremost as an empire among empires, and later as a nationalizing empire among nationalizing empires. In this study, the empirical basis for such a portrayal is a detailed analysis of how Han officials in Xinjiang attempted to meet, counter, or otherwise defuse the nationalist threat over a period of five decades. In so doing, they drew from what we might describe as a comparative cabinet of imperial “best practices,” borrowed and adapted from rival empires with which they were in direct geopolitical competition.
The link between the nationalist threat in Xinjiang and the idea of modern China as an “empire of difference” lies in the realization that the ideal of national determination has been manipulated for the purposes of geopolitical competition among rival empires since its very inception. The specter of the nation-state and the notion of ethnic self-rule first make their appearance in the documents of Han officials in Xinjiang in the form of offensive tactics deployed by rival Russian imperial elites. In response, successive Chinese governors formulated defensive tactical responses designed first to combat, and later to co-opt, nationalist politics among their non-Han subjects. Over time, the larger transformation in evidence here is a new form of the politics of difference: from an ethno-elitist alliance of culturally and politically conservative Manchu and Mongol bannermen, former imperial Han officials, and Turkic and Tibetan nobles, to an ethnopopulist alliance overseen by “progressive” and “revolutionary” cadres committed to the evolutionary development of ethnocultural and economic levels among a multitude of “backward” subjects. For Han officials in Xinjiang during the late Republican and early Communist eras, the appeal of the Soviet version of ethnopopulism was to be found in the realization that a preemptive sponsorship of national development within one’s own state could do much to defuse the appeal of separatist platforms underwritten from abroad. From this awareness arose the impulse to nationalize the various territorial and human components of one’s own empire before rival imperial elites succeeded in one’s stead. Whoever managed to do so first could then orient that difference toward their own metropole.
The following chapters narrate this momentous shift in precise chronological detail, showing when and where twentieth-century China emerges as an empire within the world rather than a nation set against it. In constructing this narrative, I have primarily drawn upon Chinese archival sources in Taiwan and China, both published and unpublished. At key points, I also incorporate relevant insights from Soviet, British, and American archives, in addition to new and exciting research based upon Uighur-language source material. Some scholars may wonder about the utility of relying so heavily upon published Chinese archival sources, particularly those produced in and about a non-Han borderland. Yet for the historian who wishes to uncover the perspective of the Han governor and his ranking officials in Xinjiang, there are few alternatives. Of course, there are key moments when unpublished archival sources play a decisive role in filling in the gaps. But the idea that five decades of Han rule and contentious ethnopolitics in twentieth-century Xinjiang can be narrated primarily from source material made freely available to foreign researchers working in mainland Chinese archives is little more than wishful thinking. Fortunately, it has often been possible to corroborate sources published on the mainland with those made accessible in the less ideologically restricted political climate of Taiwan. In every available instance, published Chinese archival sources—even those of Governor Yang Zengxin, who personally edited his telegrams for publication while still in office—have aligned with unpublished and unedited counterparts in all but the most trivial of details.
Though we can be reasonably certain regarding the integrity of those Chinese archival sources that are published, we can only speculate about those that were not. The suspicion, of course, is that we are seeing only what Beijing wants us to see, and that these materials must somehow replicate or otherwise reinforce the official view of the Chinese government. If that was the goal, then we can only conclude that the censors somehow fell asleep on the job. How else to explain the inclusion of some of the most scathing and damning indictments of Communist ethnic policies ever seen, communicated by Uighur, Kazak, and Hui workers on the factory floors of Urumchi in the 1950s? Also on full and uncensored display is Governor Yang’s unrelenting persecution of Uighur expatriates in the Soviet Union during the 1920s, along with rebel proclamations characterizing the Han people as “yellow filth” one decade later. Rare published archival documents from the Sheng Shicai era freely and openly admit the near total failure of the sort of ethnic policies the Communists themselves would one day try to implement. The facsimile edition of Governor Wu Zhongxin’s handwritten diary reveals his chronic frustration and bitter disagreements with top Nationalist leaders, and includes every entry that he later attempted to cross out to protect both his reputation and that of his colleagues. Of course, it is true that these sources portray the most contentious instance of indigenous resistance to Han rule—the Ili rebellion of the 1940s—as entirely the result of Soviet manipulation and instigation. Fortunately, newly available Soviet archives and Russian-language scholarship on this topic yield the reluctant admission from behind closed doors that Han officials in Urumchi were right to be paranoid about Moscow’s role in this uprising.
Last but not least, it is important to point out that the uses to which this source base have been put are about as far as possible from the “official line” of the Chinese government. To argue that postrevolutionary China continued to act as an empire while calling itself a democratic republic is to risk perennial ostracism and vitriolic denunciation in scholarly and political circles on the mainland. Many prominent foreign scholars of Xinjiang have been denied visas for even suggesting that modern China has been anything less than a harmonious multiethnic state for the past five thousand years.21 Most recently, in April 2015, the Chinese Academy of Social Sciences described prominent American proponents of the “new Qing history” framework—which focuses on the eighteenth and nineteenth centuries—as “arrogant,” “overbearing,” and “imperialist.” Their scholarship, from which much of the inspiration for this book was derived, was dismissed as “academically absurd.” Their crime? Having the temerity to suggest that the Qing dynasty, which flourished more than two hundred years ago and fell from power in 1912, was an “empire” that had “conquered” non-Han peoples and the lands they occupied.22
It remains to be seen what the Chinese government will think of a scholar who argues for the extension of the empire paradigm into the twentieth century and is foolish enough to suppose that an imperial legacy can be identified and studied through Chinese archival sources published under its own watch.