The political history of Xinjiang prior to 1949 was tumultuous and often violent. The substantial number of public protests and deadly clashes taking place there over the last quarter century have not been contemporary manifestations of an enduring culture of violence. Nor have they been the product of foreign intrigues. While there have been several contributory factors, Xinjiang’s very political structure is one root cause of the unrest.1 The full name of the territorial unit, the Xinjiang Uyghur Autonomous Region, suggests that Uyghurs largely govern themselves. But in fact, they do not. The system officially touted as providing autonomy in Xinjiang instead enacts heteronomy, or rule by others. Furthermore, the system of governance has not only denied Uyghurs the freedom to make some political decisions according to their own interests; it has also deprived them of the capacity to articulate and debate those interests in public or even to protest that deprivation. I argue in this chapter that this political system has exacerbated conflict and deepened Uyghur discontent.
The promise of autonomy arrangements around the world has lain in their capacity to protect both states’ territorial integrity and the fragile rights of minorities. Government advisers and scholars have proposed autonomy to avoid discord or, in some cases, to resolve it. Indeed, a major study of “minorities at risk” advocates autonomy as the sole solution acceptable to all sides in otherwise insoluble conflicts (Gurr 1993). Socialist states were among the first to adopt institutions of formal autonomy, and many claimed thereby to have “solved their national question” (Connor 1984:xvi).
Such abstract claims should not blind us to the fact that autonomy regimes emerged from a particular period in the history of nation-states. Nongovernmental organizations first mooted the establishment of formal legal and governmental institutions to protect minorities in discussions during World War I, and the League of Nations oversaw the writing of legal charters in the wake of the war’s resolution. Those states emerging victorious from the war saw the “protection of minorities [as] the means to achieve their end,” that end being not humanitarian but strategic: the preservation of international peace. It is important to recall that these charters were imposed on defeated states—or, rather, the successor states that replaced former imperial powers—by the victors in the war. As such, the charters were clear examples of victor’s justice. Italy, faced with the demand that it accept legal protection of its minorities, claimed to be “too great a Power to submit to such a derogation of [its] sovereignty.” This naturally made other states on whom such laws had been forced bridle at the imposition (Macartney 1934:252, 274). Decades later, many states still vehemently deny the right of outside actors to insist that they adopt autonomy regimes or to tinker with those regimes once installed.
All autonomy regimes privilege territorial integrity over absolute responsiveness to the demands of the autonomous group: they are a compromise between states that want unabridged sovereignty and homogeneous populations and peoples that want self-determination, generally meaning independence. Such compromises often leave neither side satisfied. Thus we should not be surprised to find both state actors and autonomous groups pressing for renegotiation. Yet there are dramatic differences in the degree to which states honor their formal commitments and the amount of pressure for change from the nominally self-ruling groups.
A foundational text on the law of autonomy regimes observes that “autonomy is understood to refer to independence of action on the internal or domestic level” (Hannum and Lillich 1980:860). Thus construed, the term ill fits the system of governance in Xinjiang. As I show, Beijing has allowed Uyghurs very little independence of action.2 The party-state has actively and with careful forethought thwarted the emergence of an autonomous political elite able to press for Uyghur collective interests, and it has similarly squelched ordinary Uyghurs’ attempts to respond to or influence policies in Xinjiang. For those silenced voices, it has substituted an official story that most Uyghurs are quite satisfied with the way Xinjiang is ruled. It has thus intervened in representational politics in both senses, actively selecting and managing the Uyghurs (and other non-Hans) charged with representing the Uyghur population, and carefully regulating the narrative representation of the region’s politics.
To be sure, the defects in the system are related to larger systemic problems in Chinese politics. The People’s Republic of China (PRC) remains a single-party state and has prevented various groups from articulating demands and from organizing to pursue their interests. Beijing employs propaganda, silences dissent, punishes transgressions with violence, and extirpates independent organizations throughout the country. All these practices are amplified in the political system in Xinjiang. Although desperate not to lose any territory, Chinese leaders attribute particular strategic significance to Xinjiang as both a source of energy and a conduit for still more from Central Asia, energy crucial to China’s continued economic growth. Concern about ethnic conflict and the fear that many Uyghurs aspire to independence has made party officials in Xinjiang much more sensitive, and often more brutal, than their counterparts elsewhere. The lodging of so much power in the hands of Han officials has only increased the region’s fractiousness.
To charge that China’s autonomous regions, prefectures, and counties do not enjoy the self-rule that Beijing claims is not particularly controversial. An early observer derided the system as “regional detention” (Moseley 1965:16, quoted in Shakya 1999:306). Many scholars have argued that the PRC’s system of “minzu regional autonomy” in fact provides little political autonomy (Dreyer 1976; Heberer 1989; Mackerras 1994). Zhao Suisheng noted, “Regional autonomy … by no means meant that the communist state would let minorities govern themselves” (Zhao 2004:183), and several scholars have made this argument with specific reference to Xinjiang (McMillen 1979; Moneyhon 2002; Stein 2003). Two scholars suggested that far from granting the locals self-rule, Xinjiang’s administration is colonial (Bachman 2004; Gladney 1998). In this chapter I move beyond formal critique to causal analysis, linking problems with the political system to consequences.
Some have argued that Xinjiang’s instability is the product of the system of autonomy and has little to do with Uyghurs’ grievances. On the basis of comparative study, Svante Cornell (2002) claims that systems of autonomy can themselves cause conflict rather than resolve it. In fact, in many cases they facilitate the emergence of secessionist movements. Cornell draws on the substantial literature (Brubaker 1996; Roeder 1991) to contend that the Soviet Union’s federal system of union republics contributed to the disintegration of the state. But he argues that the authors were wrong to ignore the institutions below the republic level. In subrepublican autonomous units, as in the republics, Cornell finds, the institutionalization and fostering of distinct identities increase autonomous groups’ “cohesion and willingness to act”; the institutions themselves enhance the “capacity of [groups] to act.” In general, a state establishing an autonomous region formally acknowledges the territorial boundaries of that region and the identity of the group or groups exercising autonomy. The state further allots resources to help protect the distinct identity of the titular group. In so doing, the state both strengthens group solidarity and sanctions a territorial frame for its political aspirations. Autonomy arrangements often allow for the establishment of distinct political institutions, the cultivation of leaders to run those institutions, and the local control of media and educational organs. Together, these institutions increase rather than reduce the likelihood of conflict and, therefore, of secession. Finally, a state’s formal recognition of an autonomous group may increase that group’s international visibility and thus raise the probability of international intervention in any resultant conflict (Cornell 2002:251–56, italics in original).
Three factors that Cornell identifies have been of undoubted importance in Xinjiang politics. First, formalizing the boundaries of Xinjiang and naming it the Uyghur Autonomous Region gave a convenient territorial shape to Uyghur political imaginings (Bovingdon 2002b). Second, institutionalizing minzu identities and assigning privileges on the basis of those identities strengthened them considerably.3 Finally, the formal assignment of one-sixth the territory of China to the Uyghurs has almost certainly given them greater international prominence than they would otherwise have enjoyed, although the violent protests of the 1990s played a role as well. Judging from the number of formal governmental inquiries about and international academic conferences on the Uyghurs and Xinjiang since 1997, they claim more attention than any others in China except Tibet and the Tibetans.
Although formally compelling, Cornell’s argument cannot explain the vastly different outcomes in China’s various autonomous regions. The PRC’s other provincial-level autonomous regions, Mongolia, Guangxi, Ningxia, and even Tibet in the 1990s, have been far quieter than Xinjiang.4 Nor does his model capture well the system actually operating in Xinjiang. Most of the features that Cornell considers as standard in systems of autonomy are absent or present only in truncated form in that region. The government institutions have been heavily colonized by Hans and have been subordinated at all levels to the heavily Han party structure. Uyghur (and other non-Han) leaders have been carefully chosen and forced to act as apologists, even boosters, for unpopular policies, to ensure they do not develop a popular constituency. The media and educational institutions have remained under tight state control in all periods except the 1980s. Indeed, it can be argued that political controls on these institutions now are tighter than they were at the beginning of the reform era. Thus while Chinese policies in Xinjiang might be credited with strengthening the Uyghurs’ sense of collective identity—in Cornell’s terms, their cohesiveness and inclination to act together—those policies have actually reduced their capacity to act collectively.
In sum, a purely systemic argument cannot fully account for unrest in Xinjiang. The institutions of autonomy have indeed contributed in various ways to the conflict, but we cannot fully understand how they have done so without considering the Uyghurs’ grievances. The manipulation of authority, violent policy swings, and state actions never given legal specification, all of which must be considered features of the system of minzu regional autonomy in Xinjiang, have provoked the Uyghurs’ discontent. Like other socialist regimes, the Chinese government has continued to boast that the system of autonomy solved China’s “minzu problems.” Instead, however, that system virtually guarantees there will be discontent and conflict for years hence.
The Xinjiang Uyghur Autonomous Region (XUAR) was established with great fanfare in October 1955. While in principle, the Uyghurs thereby received title to the property,5 what they confronted was in fact a condominium of nested autonomies, leaving them a patchwork of territories—principally Qumul and Turpan, Aqsu, Kashgar, and Khotän districts—divided and surrounded by the others (figure 2.1) and further overlaid by the overwhelmingly Han military–agricultural garrisons of the Production and Construction Corps (PCC) (McMillen 1981:70–71). The territory had been parceled out in a series of steps over the previous eighteen months and was presented to the Uyghurs as a fait accompli when the XUAR was established. It has recently come to light that in 1953, Xinjiang party officials proposed setting up autonomies “from the top down,” but the Central Committee in Beijing decreed that the order instead proceed from “small to large” (Wang Shuanqian 1999:249).
The division of Xinjiang into a number of smaller autonomies was a stroke of administrative genius. In parceling out “subautonomies,” the Chinese Communist Party (CCP) simultaneously satisfied the goals of embodying the idea that Xinjiang belonged to thirteen different minzu and of counterbalancing the Uyghurs’ overwhelming political and demographic weight.6 To a certain extent, the political and material interests of each of the other constituted groups were therefore aligned with the central government and against the Uyghurs. By the end of 1954, more than 50 percent of the area of the province had been allotted to autonomous townships, districts, counties, and prefectures.7 In fifteen out of twenty-seven units established, the titular minzu constituted less than half the population; in Tacheng and Emin county autonomous districts, the titular minzu (Daghuor and Mongol) made up, respectively, less than 17 percent and 12 percent of the population. Bayangol was made a Mongol autonomous prefecture, even though Mongols constituted only 35 percent of its population. Even more extraordinary, in 1960 the government annexed to it an “almost purely” Uyghur region, after which Bayangol Prefecture occupied fully one-third of Xinjiang’s area (Atwood 2004:39).8 As a result, in 2004 some 48,000 Mongols nominally exercised autonomy in a region with more than 370,000 Uyghurs (and, due to steady immigration, more than 660,000 Hans) (Xinjiang Weiwu’er zizhiqu tongjiju 2005:112–13).9 In a study on minzu relations in Xinjiang, government analysts Mao Yongfu10 and Li Ling observed that
FIGURE 2.1 Autonomous prefectures in Xinjiang. (Based on Yuan Qing-li 1990)
Qirghiz, Tajik, and other minzu, despite the small size of their populations, nevertheless have autonomous prefectures and counties belonging to their own minzu; in those areas, they belong to the self-governing minzu. By contrast, in [those places] Uyghurs have once again become non-self-governing minzu…. This is something we must look into diligently. (Yin Zhuguang and Mao Yongfu 1996:173)
The authors’ apparent surprise at this situation was rather disingenuous, as the order of establishment of the autonomies demonstrated that this had been precisely the intention forty years earlier.11 The intention endures today. According to 2000 figures, in none of the five autonomous districts in Xinjiang does the titular population exceed 25 percent of the total, and Bayangol Mongol District is only 4 percent Mongols. Nonetheless, a recent work cautioned that these figures “must not shake our determination to the exercise of minzu regional autonomy.” While administrative reform is all the rage in China proper, with localities becoming cities and some cities gaining provincial-level status, the boundaries and system of autonomous units will not change, the authors announced (Zhu Peimin, Chen Hong, and Yang Hong 2004:356–57).12 These odd circumstances then lead us to ask, in what sense are the various minzu “self-governing”?
Every official account of “minzu regional autonomy” (minzu quyu zizhi) makes liberal use of the expression “masters in their own house” (dang jia zuo zhu). The system of autonomies, advocated by theorists and enacted by the Chinese Communist Party (CCP) since 1949, is said to have transformed non-Hans from exploited others to rulers of their own domains. In Xinjiang, rather than making non-Hans heads of the house, the purpose of this system has been, above, all to keep them in the house. Granting to Uyghurs any influence over affairs in the Xinjiang Uyghur Autonomous Region (XUAR) has taken a back seat to consolidating the CCP’s control and crushing any movements advocating independence, or even the more modest goal of “real autonomy.”
Despite maintaining publicly that power stems from the people, the party leadership has always taken pains to extend authority from the top down and has therefore given no quarter to power organized locally. Second, party leaders have regarded the Uyghurs as a group to be politically untrustworthy and thus have allotted very little power to them. Instead, they have selected and promoted officials who exercise power only in a fashion consonant with the CCP’s goals and have reserved the decisive authority at virtually all levels for trusted Hans imported from posts in China proper.13 The system has failed to serve Uyghurs in a variety of ways. Vaunted as a means to lessen interethnic conflict and gain Uyghurs’ loyalties, it has instead contributed to their generalized dissatisfaction. To understand how, we must begin with the legal framework of autonomy.
Successive PRC constitutions and associated organic laws have codified a national plan for minzu regional autonomy in a set of institutions, cultural and linguistic rights that the institutions were intended to secure, and a carefully specified relationship between those institutions and the central government. Xinjiang’s distinctive features and political history notwithstanding, its “autonomous” government institutions emerged from that national plan and therefore strongly resemble those in China’s other autonomous regions.
Both the 1952 Program for Implementing Minzu Regional Autonomy and the 1954 constitution made clear the expectation that autonomous units at county, prefectural, and provincial levels would have governmental organs broadly similar to those at the corresponding levels in China proper. Stipulated elements included the standard branches of government, a people’s congress with elected representatives chosen from the titular group or groups as well as Hans, and obedience to higher-level government organs.14 Both documents provided that actual institutional forms could be determined by the wishes of the “great majority of the people and leading figures with links to the people” of the titular group or groups.15 Military organization was not left to local discretion. Although the program allowed for the establishment of local police cadres, all security personnel, including soldiers, fell under the “unified national military system.” Between 1950 and 1955, troops in the predominantly Uyghur and Qazaq Ili National Army (INA), formerly the military arm of the Eastern Turkestan Republic (ETR), were folded into the People’s Liberation Army (PLA) or demobilized and settled on paramilitary farms. In both cases, they fell under direct CCP control (McMillen 1979:53; Shichor 2004:127–29, 132).
The first constitution also committed the government to drawing up a national autonomy law. Even though the Government Administration Council had passed the program in 1952, a formal law was promulgated only in 1984 and then amended in 2001. In addition to these organizational principles, the 1984 autonomy law affirmed in general terms what had been practices of long duration in the autonomous regions: “affirmative action” in the recruitment of college students, the hiring of employees at state enterprises, and the training of base-level government cadres.
The articles of the autonomy law reveal the limitations built into the system.16 Article 15 indicates that all autonomous government organs are under the leadership of the State Council and all must “obey the state council.” Article 20 grants organs of autonomy the right to “alter or suspend” policies or orders promulgated by higher-level government units but makes such actions subject to approval by those superior units. And while it acknowledges the right of autonomous regional governments to draw up locally appropriate “statutes on autonomy and specific regulations,” article 19 also grants the National People’s Congress the authority to approve or reject such statutes.17
After studying the original law and its 2001 revision, Chinese legal scholar Yu Xingzhong concluded that the political system it specifies “certainly does not correspond to what is usually understood [by] the term ‘autonomy’” (Yu Xingzhong n.d.). Studies of the system with special reference to Xinjiang reached much the same conclusion (Moneyhon 2002; Sautman 1999; Stein 2003). The legal scholar Matthew Moneyhon (2002:137) argued that whatever rights the state constitution and autonomy law grant in principle to local decision-making bodies, by requiring that central government organs approve all local decisions, the national laws withdraw most of those rights in practice.
Indeed, the system of governance established in Xinjiang fails to satisfy a single one of five criteria that a pair of expert jurists have determined are fundamental to autonomy. In a widely cited article, Hannum and Lillich insist that legal autonomy requires, at minimum, an independent local legislature not subject to central veto power, a locally chosen executive, an independent judiciary, local decision making not compromised by the center’s “reservation … of general discretionary powers,” and binding power-sharing arrangements (Hannum and Lillich 1980:886–87). Beijing nonetheless retains veto power over the decisions of Xinjiang’s People’s Congress. The center has chosen each of Xinjiang’s successive executives, and the Supreme People’s Court retains supervisory power over Xinjiang’s courts. Furthermore, Beijing reserves broad discretionary power over Xinjiang’s affairs, including how autonomy is implemented; and the center allocates power over resource exploitation, policing, and other matters on its own initiative, rather than being bound by power-sharing arrangements (Moneyhon 2002:137, 142–44).18 Despite its dramatic departure from the Soviet model in many particulars, on this last point the PRC system clearly followed Moscow’s example. In relations with the union republics before 1991, Moscow “both define[d] the sphere of authority and exercise[d] it” (Gleason 1990:65).
As confining as the legal framework has proved, actual political practice in Beijing and Ürümci has made still greater incursions into the region’s hypothetical autonomy. In Yu Xingzhong’s estimation, rather than providing a new legal framework for the system, both the 1984 and 2001 versions of the autonomy law simply recorded in statutory form what already were features of established practice. Legal revisions followed policy changes, and not vice versa, and Beijing still governs by policy rather than by law (Yu Xingzhong n.d.).19 A full account of the workings of minzu regional autonomy in Xinjiang must therefore include not only structural limitations of the system but also a brief history of political developments and actual political practice.
Two features of the social climate of Republican-era Xinjiang figured prominently in Communist Party leaders’ calculations of strategy. As discussed in chapter 1, anti-Han sentiment was deep and widespread among the Uyghurs, partly as a consequence of decades of harsh, exploitative rule by Han warlords and local officials. At the same time, general antipathy to Hans, which might have drawn Uyghurs and other Turkic peoples together, was counterbalanced by religious, political, and cultural differences of long duration. The Turkic peoples had cooperated in founding the first (1933–1934) and second (1944–1949) Eastern Turkistan Republics. In the Ili National Army (INA), the military force raised by the latter republic, Uyghurs, Qazaqs, and members of other groups fought side by side. Yet those governments controlled only small parts of the vast territory of Xinjiang and fell apart as much from internal disagreements as from external attacks (Forbes 1986:229–34 and passim). Resentment of Hans posed a challenge to CCP strategists, but antagonisms within the Turkic population gave the party an opportunity to pit the groups against one another and thus to manage that challenge.
The political history of post-1949 Xinjiang reveals even more about governance in Xinjiang than does the study of autonomy law. If Xinjiang had truly been autonomous, it would have had its own political cycles and campaigns over the last fifty-odd years. Even a quick summary of major political events since the revolution shows that Xinjiang’s policies and campaigns closely followed those in China proper.
After leading the People’s Liberation Army into Xinjiang in 1949, General Wang Zhen demobilized thousands of soldiers and redeployed them on a network of paramilitary farms throughout the province, subsequently naming them the Production and Construction Corps (PCC).20 Having installed a tractable government leadership headed by the Tatar Burhan Shähidi and the Uyghur Säypidin (Saifudin) Äzizi, Wang and other party officials set about establishing strategies for managing non-Han groups.
In the early 1950s, those strategies were relatively tolerant. The CCP’s “united front” (tongyi zhanxian) policy counseled the establishment of links with “progressive members” of social and religious elites, which in turn required minimal interference in business, religious practice, and social norms. The party did, however, gradually take control of religious institutions through the China Islamic Association, as well as through the confiscation of mosque lands and the forcible replacement of religious courts with “People’s Courts” (McMillen 1979:113–14).21
By the mid-1950s, as Mao pressed regional leaders to make more sweeping economic changes throughout the country in the so-called socialist tide, the Xinjiang leadership faced resistance to such initiatives. Collectivization required antagonizing the “progressive elites” with which the party had previously cooperated. And again, the cadres’ attempts to mobilize the exploited classes (mostly peasants, in the overwhelmingly agrarian region) against the elites instead drove many Uyghurs and others together against the party. In China proper, Mao invited criticism of CCP policies from the masses in his 1956 Hundred Flowers campaign. The vehemence and volume of protest shocked the leadership, which then unleashed the antirightist movement in 1957 to silence the opposition. In Xinjiang, Uyghurs, Qazaqs, and others denounced the system of autonomy as a sham and demanded a far greater role in local governance. Party hard-liners swiftly redirected the antirightist movement into a movement against “local nationalism,” accusing such critics of seeking to “rule Xinjiang as an independent country” or of resisting CCP rule.22 Officials were especially irked by the charge that PCC soldier-farmers were “Han colonialists.” Faced with such challenges, the party redoubled its efforts to mobilize class against minzu interests (Dreyer 1976:150–57; McMillen 1979:116, 117).
Mao’s radical voluntarist Great Leap Forward, begun in 1958, led in Xinjiang to calls for rapid cultural homogenization to accompany and facilitate the Leap. This naturally meant much less tolerance of difference. Even ethnicity itself became an “obstacle to progress,” and party leaders stepped up their attacks on Islam and other “backward customs” (Dreyer 1976:157–63; McMillen 1979:118). As is now widely known, the combination of Great Leap policies, bad weather, and the central government’s ill-chosen decision to export grain to meet its debts to the Soviet Union brought on a terrible famine. Party leaders temporarily prevailed on Mao to restore a more moderate economic course in the early 1960s. Cultural policies in Xinjiang relaxed slightly during this period, as officials acknowledged that linguistic and cultural differences would persist over the long term. Muslims were again allowed to celebrate religious festivals, which had much subdued since 1957 (Benson and Svanberg 1998:139). Talk of speedy assimilation subsided.
Despite the policy retrenchment, widespread starvation in China’s interior intensified the Great Leap’s impact on Xinjiang. In addition to party-mandated population flows, vast numbers of hungry migrants poured into Xinjiang on their own initiative. This “surplus” flow drove Han immigration to more than 800,000, its highest level ever, in both 1959 and 1960 (Hannum and Xie 1998:324; Li Yuanqing 1990:52). PCC farms welcomed many of the refugees to settle and claim land, provoking increased resentment by Uyghurs and others. In 1962, more than 60,000 Uyghurs and Qazaqs fled across the border into the Soviet Union, exasperated with CCP policies and lured by ceaseless radio propaganda advertising far superior living conditions on the Soviet side of the border. The deepening Sino-Soviet split played a role in this exodus: Soviet consular officials had apparently connived in the mass migration by distributing previously prepared travel papers. For officials in Xinjiang and Beijing, the sheer numbers of emigrants raised the frightening prospect of hostile former citizens receiving military training abroad, then assisting in the cause of “Soviet social imperialism” by helping take Xinjiang by force. In response, the government sealed the border and forcibly relocated thousands of non-Han families away from the border zone, replacing them with Han PCC members (Dreyer 1976:169–70; McMillen 1979:120–23).23
Minzu policies changed course again in the mid-1960s, as renewed radicalization at the highest level of the party led to the Cultural Revolution (1966–1976). Officials appointed to the Cultural Revolutionary Small Group, which replaced the XUAR Party Committee for several years, and the initially mostly Han Red Guards raised demands for cultural conformity to a new extreme. While in China’s heartland, Red Guards collected and destroyed artifacts of the past, in Xinjiang (as in Tibet and other non-Han regions) they targeted non-Han culture. Difference once again became a sign of backwardness. Activists frightened the various Turkic peoples into shedding their habitual clothes, adornments, scarves, and hats for Mao suits. Activists also destroyed mosques and even forced many religious leaders and ordinary Muslims to raise pigs, apparently in an attempt to engineer rapid and thorough assimilation.24
The punishments that the Cultural Revolutionaries visited on intellectuals particularly targeted Uyghur culture. The famous linguist Ibrahim Muttä’i was tortured by having the huge volumes of a multilingual dictionary he had helped edit (with full CCP support at the time) dropped on his head (Clark forthcoming, 17). Ordinary citizens were not exempt. Residents who lived through the period describe witnessing men being subjected to shaving in the streets, for even beards were interpreted as signs of defiance. A Qazaq woman raising a towheaded Uyghur boy dyed his hair black and shaved his eyebrows to avoid persecution. Uyghurs meeting each other in the street learned to initiate every greeting with “Long live Chairman Mao” in Chinese.25 But in 1975, many Muslims finally lost patience and staged a protest when forced to work on Muslim holidays.26
In retrospect, it is clear that policies toward non-Hans were the most assimilationist and intolerant during the Cultural Revolution. After Mao’s death in 1976 and the fall of the Gang of Four, party leaders faced a crisis. The Cultural Revolution had alienated a large segment of the population throughout China. Resentment was particularly grave among Uyghurs and other non-Hans because for them, it had been not merely a political and social assault but an attack on their identities. The continuation of hard-line policies seemed certain to provoke increasing discontent and thus instability. But officials worried that more tolerant policies, allowing cultural exploration and freer religious practice, might similarly open the door to unrest. Nagged by these concerns, policymakers charted a zigzagging but narrowing course between openness and control.
In 1980 Hu Yaobang, one of the younger leaders in the Central Committee and soon to be promoted to CCP secretary general, traveled to Tibet to investigate local conditions. Appalled by the poverty and despair he found there, Hu urged that hard-line policies toward non-Hans be relaxed. To remedy Tibet’s situation, he advocated “genuine autonomy,” economic initiatives appropriate to local conditions, the renewal of cultural and scientific projects, and the gradual transfer of Han officials back to China proper. He made similar proposals for Xinjiang in July 1980.27 At the time, Hu felt that Xinjiang presented less of a separatist threat than Tibet because it lacked exiled religious or political leaders like the Dalai Lama and had no support abroad for independence (Dillon 2004:36). The archconservative Central Committee member Deng Liqun claimed to have heard that Hu had even proposed devolving nearly all political authority to Xinjiang, reserving power only over national defense and foreign relations and a veto over domestic politics to Beijing. Deng recalled asking Hu at the time, incredulously, “How can you propose such a resolution in a minzu autonomous region? How can it be OK … to let the autonomous region to exercise [all other powers] itself?” (Deng Liqun 2006:206–7). In the event, the XUAR Party Committee adopted Hu’s suggestions on cultural policies and economic reforms and announced in August 1980 that large numbers of Han cadres would be transferred out of Xinjiang, over the objections of such hard-liners.
Held responsible for the increasing number of student and popular demonstrations that convulsed Beijing and other major cities in 1986, Hu Yaobang was purged in 1987. Former Xinjiang Military Commander Wang Zhen, who had fought openly with Hu over his proposed changes, then ordered that the more accommodating policies be scrapped. A Han official siding with Wang is reported to have asserted, “You give [Uyghurs] autonomy and they will only turn around and create an East Turkistan.” The official was disgusted with the proposal to send Hans back to the interior and insisted that only tough-minded officials in the mold of Wang Zhen could keep Xinjiang stable.28 An official report produced for internal circulation a few years later concurred in this view. The dismissal of Han cadres had “made it impossible,” the report stated, “for Han cadres to work in Xinjiang with a sense of security, and it also sowed serious ideological chaos among minority minzu cadres. [It] was extremely harmful to minzu solidarity” (Zhang Yuxi 1993:358).
Over the previous decade, Deng Xiaoping’s economic reforms had brought unprecedented prosperity to China’s interior. The conservative leadership of Xinjiang initially sought to block the local implementation of reforms, fearing that they would destabilize the region. To many, the pace of the reforms that did come was frustratingly slow. In the mid-1990s, people continued to joke that while China proper had wholeheartedly embraced capitalism, socialism was still being pursued, if not realized, in Xinjiang. Nevertheless, by 1992 the XUAR leadership had come to an agreement that reform was inevitable. The popular Uyghur official Ismayil Ähmäd announced the same year that the central government would cede more autonomy to Xinjiang. He suggested that this would include local authority over foreign trade, control of the border, and administration. At least one analyst saw this move as an attempt to counter the appeal of separatists in the wake of the Soviet Union’s collapse (Cheung Po-ling 1992).
But the party complemented the loosening of economic policy with political tightening, a practice that has continued up to the present.29 There have been two central components of that tightening. The yearly “strike hard” (yan da) campaigns, like their counterparts in the interior, aim to capture large numbers of suspected criminals in massive dragnets and then prosecute them on an accelerated schedule. As in China proper, these campaigns target violent property crimes, drug rings, and the like; in Xinjiang, they also focus on political crimes.30 Second, public security personnel arrange periodic sweeps to shore up control in each locality, called “comprehensive management” (zonghe zhili).31 A number of my Uyghur informants received orders from their work units to take part in these sweeps in 1997 when the party was fretting about the return of Hong Kong. One ardently anticommunist man told me he was simply notified a week in advance that he would be traveling to the south to spend two months praising CCP policies. He was to go from house to house within “suspect” villages, chaperoned by two Hans, patiently correcting people’s misconceptions and erroneous political views. It was, he observed, like being forced to eat a steaming plateful of pork. Another informant told of being sent more regularly on short trips to areas around Ürümci, again without any choice in the matter. Interviews with a doctor revealed that the strategy did not involve surveillance and propaganda alone. With only a few days’ advance notice, she was dispatched for several months to several poor rural areas to treat patients and to pass on the party line while doing so.32
One of the party’s most effective tactics for counteracting political pressure from Uyghurs has been, in effect, to subcontract security work, by importing a Han constituency loyal to Beijing. Government-sponsored immigration of Hans into the region has been a central component of CCP policy in Xinjiang. It also has provided compelling evidence that officials in Beijing and Ürümci felt they could never persuade the majority of Uyghurs to be satisfied with their political status in China.33 Ma Dazheng finally expressed this premise in his internal-circulation report in 1999: “Hans are the most reliable force for stability in Xinjiang” (Ma Dazheng 2003:123). Local officials clearly understood this point, referring colloquially to the settlement of Han immigrants as “mixing sand” (Becquelin 2000:74).34
But this was never the government’s official rationale for encouraging Han immigration. The public story was that Hans were needed to spur development. In the mid-1950s the British journalist Basil Davidson, who had already published a paean to the first three postrevolutionary years in China, was invited on the strength of that work to travel to Xinjiang to see for himself the boons of the region’s “peaceful liberation” and then to broadcast them to the world.35 As the title of his popular travelogue, Turkestan Alive, suggests, it depicts a Xinjiang revivified by Chinese policies. Davidson nevertheless felt obliged to note that the great flow of Han immigration could be seen as in “flagrant contradiction to the fostering of home rule and minority development.”36 To compensate, he pointed out that the region desperately needed rapid economic development and lacked both the skilled workforce and able administrators vital to that task. Davidson thus reported sympathetically the comment of one of his Chinese hosts that “we need another ten million in the next fifteen years” (Davidson 1957:234–35). Soon after the PRC’s founding, the State Council prepared a plan to transfer some two million citizens from China’s interior. Though far below the figure proposed by Davidson’s interlocutor, it was extraordinarily ambitious given the difficulties of both mobilizing Han migrants and finding places for them in a largely arid region.37 Yet in the end, officials overfulfilled the plan. Between 1950 and 1978 the party induced roughly three million Hans to move to Xinjiang to help “build the borderlands” (Qiu Yuanyao 1994:223–40; Wang 1998:37–38).38 Many settled on PCC farms, and in the early years PCC cadres traveled to major population centers to recruit more volunteers directly (McMillen 1979:61). Massive and sustained immigration increased the Han proportion of the population from roughly 5 percent in 1949 to more than 40 percent in 1978. But while moving Hans into Xinjiang was relatively easy before 1978, making them stay after that time became increasingly difficult, and beginning in 1977 there was a net outflow of population (Li Yuanqing 1989:124).
The first major disturbances in Xinjiang during the reform era involved not Uyghurs but Hans agitating to return to the interior. Many Han youths had been “rusticated” (xiafang) from Shanghai and other major urban centers in China proper during the Cultural Revolution. Some had already returned home clandestinely, discovering to their dismay that local authorities refused to find them jobs or housing. Their residence permits had been permanently transferred to Xinjiang. In February 1979, in Shanghai, Han youths unwilling to return to Xinjiang and other “remote” areas rioted, and in Aqsu, Han youths staged a series of actions petitioning in April and July for the right to return home. Seeing their petitions ignored, four thousand gathered to demonstrate in January 1980 (Zhu Peimin 2000:353–54).39 The government’s initial policy responses were rather tepid and clearly not directed to Aqsu. In July 1979 the government convened a meeting in the northern city of Shihezi to “make a plan to resolve the problem of educated youths.” In mid-November the government announced that it had taken out an interest-free loan of 2 million yuan to boost the economy and help find jobs for unemployed Ürümci youths. Later that month the government generously announced that workers in Xinjiang would henceforth be allowed to return home for a family visit once every five years (Chen Chao 1990:221, 27).
These steps clearly did not satisfy the Shanghai youth. In November 1980 they staged a demonstration in Aqsu twice as large as that in January. The party responded by dispatching former regional military leader Wang Zhen, aged and sick though he was, to stifle the disturbance, “requesting” that local units improve conditions for young Han settlers and stepping up propaganda stressing how important the youths were to countering Soviet designs on the region. Wang was dispatched to Xinjiang again in January 1981 after the Uyghurs staged a political protest, and in May after an armed uprising. In August he returned with Deng Xiaoping himself, an indication of the top leadership’s concern about recrudescent opposition in the region (Dreyer 1986:736). Despite their efforts, emigrants continued to outnumber immigrants. Increasingly desperate at the upsurge in violence and continued Han flight, Beijing recalled Wang Zhen’s old comrade Wang Enmao from retirement in October 1981 to serve once again as Xinjiang’s first party secretary. The latter proposed a compromise to keep the bulk of the former Shanghai youths in Xinjiang: a small number judged to have “qualifications” would be allowed to return to Shanghai, while the majority, though required to remain in Xinjiang, would be permitted to have one child resettled in Shanghai by the state (McMillen 1984:574–76; Zhu Peimin 2000:356). The most significant and far-reaching decision was made in Beijing. In June 1982 Deng Xiaoping announced the reinstatement of the Production and Construction Corps. This move meant renewed subsidies from Beijing, an expanded paramilitary force to respond to local uprisings, and an institutional framework for the recruitment and absorption of more Han immigrants (McMillen 1984:585–86; Zhu Peimin 2000:337–40).40
Despite the efforts of the two Wangs, not until 1990 was there again a net inflow of Hans, when a new wave of immigrants moved to Xinjiang by choice rather than compulsion. This inflow was the result of a combination of market forces and the declining significance of the national household registration system (hukou), which had been established in the 1950s precisely to stanch migration.41 Deng Xiaoping’s economic reforms had enabled farmers to lease land, individuals to strike out in private business, and underemployed rural and urban workers to seek jobs in new enterprises. Both geography and central policies initially favored the coastal regions, but by the early 1990s the government announced favorable land lease rates and tax abatements in China’s western hinterlands—the so-called west-leaning policies—to lure labor and capital to the west. Finally, burgeoning markets in food, land, and labor greatly diminished the power of the household registration system to fix people in one place. By the late 1990s, the combined effects were plain to see in the thousands of simply dressed, heavily burdened people pouring out of the Ürümci train station each day, drawn by rumors of land and jobs in the “great northwest.” Nicolas Becquelin noted that the 2000 census proved what Uyghurs had suspected but local officials had adamantly denied for a decade: Han immigration had risen enormously in the 1990s. The Han population in Xinjiang grew an astonishing 31.6 percent between 1990 and 2000, a pace almost exactly double that of non-Hans. And as Becquelin pointed out further, this figure would not have counted summer seasonal laborers because canvassing for the census took place in November (Becquelin 2002:61).42
Since 2000, while clearly aiming to maintain the inflow of large numbers of Hans, the party has given special attention to the scarcity of particularly desirable kinds of immigrants: educated youths, technical workers, and committed, politically reliable cadres.43 It has tried a variety of stratagems to remedy those deficiencies, including subsidies to college graduates willing to immigrate and temporary “swaps” of cadres from the interior with groups in Xinjiang. Strikingly, the party announced quietly in April 2000 that it was reassigning one hundred demobilized army officers from China proper to head local CCP branches responsible for “political, legal, military, and recruitment affairs” (Agence France-Presse 2000). In March 2001, the Xinjiang government acknowledged that of the 12,000 students from the region who traveled to the interior for schooling each year, only 20 percent returned. Over twenty years, the “brain drain” amounted to 210,000. The head of the Law and Regulations Office in the XUAR Personnel Department announced the division’s intention to put together a package of high salaries and subsidies to lure and keep talented people (“Xinjiang 20 nian liushi rencai 21 wan” 2001). A year later, an expanded version of the same article appeared on another government news site. This time the department’s party secretary announced a “human resources emergency” and called on the national government to set aside monies to support high salaries, subsidies, communications infrastructure, and other perks to tempt talented people to “happily put down roots in the borderlands.” He urged that long-term settlers receive even higher salaries and retain their subsidies even in retirement. Implicitly acknowledging that few talented people (understood to be Hans) were willing to stay in Xinjiang permanently, he suggested that individuals working there for more than twenty years should receive a “really eye-catching relocation allowance” to help them return to their home towns (“Rencai gaoji! Xinjiang 20 nian liushi rencai 21 wan” 2002). Xinjiang First Party Secretary Wang Lequan even made recruitment trips, observing pointedly during an inspection tour of Shanghai that Xinjiang was short on “qualified personnel” and its workers were of “poor overall quality.” One of China’s top newspapers interviewed him during the trip and broadcast his plea for “talented people of all types” to immigrate to Xinjiang to serve as party and government cadres, managers, or technicians (Xing Zhaoyuan and Wang Se 2003).44
Officials in Xinjiang have never been shy about calling openly for skilled and talented immigrants to come to the region. Such solicitations conformed to the public story that immigration was necessary to spur development. In 1999 the government began to speak openly about immigration as an end in itself. Ma Dazheng’s comment that Hans were the main force for stability in Xinjiang was originally written in 1993 and intended for only internal circulation. Nicolas Becquelin argued that by the end of the 1990s Beijing had taken Xinjiang fully in hand, so there was no need to “proceed with caution and coded double-speak any more” (Becquelin 2004b:374). This change was visible in an article in the very public venue of the Ta kung pao, a CCP-controlled newspaper in Hong Kong.45 Alarmed by NATO’s intervention in Kosovo and rumors that Tibet or Xinjiang might be a future target, the author argued that the government should “appropriately increase immigrants” to Xinjiang and settle them on PCC farms. The aim, the author continued, was “to adjust the proportions of the populations of different ethnic groups in this region, especially in its southern part” (Ai Yu 1999). In 2000, Li Dezhu, the leader of the State Minzu Affairs Commission (Minzu shiwu weiyuanhui) stated outright that the government expected a larger number of immigrants to flow to western China in response to its “open the west” (xibu da kaifa) campaign begun that year. He also made clear that Beijing intended this augmented influx of Hans to counter the threat of separatism by altering population ratios, although he acknowledged that in the short run, this alteration might actually “increase contradictions and friction” (Li Dezhu 2000:705–6, cited in Becquelin 2004b:373–74). This last admission directly contradicted officials’ earlier repeated assurances that Han immigration served the interest of all peoples in Xinjiang and would bring peace and prosperity to all (see, for example, Xinjiang jiushi ti bianji weiyuanhui 1993:16–8). As provocative as such announcements might have been decades earlier, by 2000 they merely acknowledged publicly what Uyghurs had been saying in private for years.
Uyghurs’ fears about the political implications of Han immigration were compounded when the party began to apply family-planning policies to non-Hans in Xinjiang. Knowing this action would be explosive, officials implemented them gradually.46 Even though Hans throughout China had been subject to birth limits since the early 1980s, Uyghurs and other non-Hans had been exempt. The government began to publish the new periodical Xinjiang Family Planning in Chinese in February 1985 and to post notices (presumably in Chinese), where Hans would be likely to see them, observing that all minzu had an “obligation to limit family size.” The family-planning periodical came out in its first Uyghur-language edition in December, and on December 12, students took to the streets in a demonstration that one observer called the “most serious political event” in Xinjiang since 1949 (Li Yuanqing 1990:71). The students shouted their firm opposition to birth limits for Uyghurs and raised other complaints as well; the eruption of the protest so soon after the periodical came out suggests that it was the triggering spark. No doubt dismayed by the violent response, party leaders maintained their measured pace, waiting two years before requiring Uyghur party officials to limit their fertility, another year before promulgating “temporary regulations” on minority minzu birth planning, and yet another year before enforcing limits of three children for rural Uyghurs and two for those in cities. In 1992 the State Council issued “strict ceilings” for Xinjiang’s population in 1995 and 2000 (Xu Xifa 1995:171, 173).47 As the decade wore on, the combination of a strict limit on Xinjiang’s total population growth and policies not just allowing but encouraging Han immigration lent considerable weight to the suspicion that Uyghurs were being made to limit their procreation to make room for more Hans.
My Uyghur informants condemned birth restrictions for at least three reasons. First, it was a clear external imposition by a government they regarded as alien. They argued that no other category of social life was more clearly part of “their own affairs,” to be carefully sheltered by the principle of autonomy. Second, many Uyghurs objected on religious grounds that the party-state was arrogating to itself an authority it had no right to wield. Third, an overwhelming number of Uyghurs felt that had millions of Hans not immigrated to Xinjiang, there would be no ecological case for birth planning.48 Uyghurs protested against Han immigration beginning in the 1950s (“Zizhiqu dangwei kuoda huiyi henhen de fandui difang minzu zhuyi” 1958). The influx of Hans and family-planning policies figured in student protests in 1985, 1988, and 1989 and continued to be a theme of demonstrations in the 1990s (see chapter 4 and appendix). When a Uyghur official in the Ürümci government told a BBC reporter in 2005, “There’s very little difference in the ethnic balance between now and the early 1950s,” his comment reflected the vast gulf—whether in perception or in honesty—between officialdom and the populace (Sommerville 2005).
Officials in Beijing and Ürümci deployed many Han immigrants in a manner that neatly complemented the division of Xinjiang into subregional autonomies. Redeployment of demobilized Guomindang (GMD), Ili National Army, and Red Army soldiers in the PCC, or bingtuan, at strategic points throughout the region enacted a subtler parcelization of the territory.49 PCC units set up along the margins of “troubled” regions, along key transport arteries, and around hubs provided the potential to control travel and isolate regions with very few enforcers. Heavy concentrations of PCC farms in Kashgar, Aqsu, and Qumul districts were intended to counterbalance the overwhelmingly Uyghur population in those areas. The bingtuan was initially billed as a force to protect Chinese sovereignty from external threat in sensitive border regions, but the pattern of its units’ deployment makes plain that defense against foreign invasions was never the planners’ primary worry. The concern of first importance was and would continue to be Uyghur agitation for an independent Xinjiang. The complexion of the PCC underscores this point: it has been overwhelmingly Han since its establishment. A recent publication praises the bingtuan not only for quelling the 1990 Baren uprising and similar disturbances but also for “preventing more than 8,600 people from leaving” during the mass exodus of more than 60,000 Uyghurs and Qazaqs in 1962 (Yang Zhenhua 1997:15). After the event, the government forcibly evicted farmers and pastoralists from a fifteen-mile-wide swath of territory along the entire border with the Soviet Union and set up new PCC camps there. Jiang Zemin’s explicit statement in 1990 that the PCC’s primary mission had changed from guarding against foreign threats to quelling domestic ones marked a change much less in function than in rhetoric (Becquelin 2002:63). Indeed, Deng Xiaoping had privately described the bingtuan as an “an important force in maintaining local stability” to Wang Zhen when authorizing him to reinstate it in 1981 (Seymour 2000:182).
It is a supreme irony that the PCC boasts more of the trappings of autonomy than any other entity in Xinjiang. According to a professor at the XUAR Party School, it “administers its own internal affairs,” claiming its own public security apparatus, courts, procuratorate, judiciary, and jails (Zhu Peimin 2000:423, 422). With regard to those powers it does not exercise itself, it is answerable not to the party organization or the government of Xinjiang but instead directly to the State Council in Beijing. A former Uyghur official commented that the PCC constitutes Xinjiang’s “second provincial authority,” and an expert on China’s military described it as “an independent government within a government and Beijing’s real and reliable power base” in the region. There is a strong case to be made that Beijing regards the organization as a better counter to the threat of Uyghur nationalism than the Xinjiang government itself (Becquelin 2004b:367–68; Rahman 2005:87–88; Seymour 2000:181; Shichor 2006a:144).
The organization’s demographic significance in Xinjiang is unmistakable. In 1974, the number of bingtuan members reached 2.26 million, or one-fifth of the total population and two-fifths of the Han population in Xinjiang. Despite being disbanded in 1975 and receiving no government support for seven years, the network of farms and construction units recovered and grew steadily after 1982, so that by 1994 the PCC again claimed 2.22 million souls, of which 88.3 percent, or 1.96 million, were Han, constituting 35 percent of Xinjiang’s total Han population in that year. By 2004 it had far exceeded that previous mark, with 2.56 million members.50
Officials and scholars have lately increased calls to expand the organization still further. A 1997 work on the history and function of the bingtuan insisted that Beijing “organize the transfer” of Hans from China proper to Xinjiang for the sake of national unity (Fang Yingkai and Li Fusheng 1997:1539, cited in Becquelin 2000:78). In 2001 Ma Dazheng wrote more bluntly still that “Hans are the most reliable force for stability in Xinjiang” and complemented this comment with the pithy statement that “the stronger the PCC, the more stable Xinjiang will be.” His text, intended for and read by policymakers, provided a detailed plan for further Han colonization. PCC sites near Qumul must be strengthened to enable full control of the passage from Xinjiang to the interior, and those near Kashgar should be beefed up to isolate Xinjiang from turbulence in Central Asia. Further deployments should “draw a circle” around southern Xinjiang by filling in “blank spaces,” one between Aqsu and Korla and another between Khotän and Ruoqiang. The expansion of PCC organizations in the latter region, Ma remarked brightly, “will prove an excellent conduit for changing the minzu population ratio in Khotän, which has gotten out of balance” (Ma Dazheng 2003:123, 124, 234–35). A Qing historian by training, Ma knew very well that Turkis had been the absolute majority of Khotän’s population for centuries; it is very hard to understand how he could have construed Khotän’s still overwhelmingly Uyghur population as “getting out of balance.” Yet there is absolutely no doubt that he meant the district needed more Hans.
The PLA and other security forces enjoy a decisive logistical and military advantage over any potentially violent challengers in Xinjiang. Both the system of nested autonomies and the network of strategically located PCC farms have augmented Beijing’s capacity to control the region by giving non-Uyghurs in Xinjiang a stake in the current distribution of power. The strategic deployment of immigrants and subprovincial units of autonomy has ensured that even if Uyghurs were to join together in a peaceful demand for independence tomorrow, they would be opposed by the nearly eight million Hans as well as by more than two million members of other Turkic, Xibo, and Mongol groups. To maintain the loyalties of numerically small groups, the party has incorporated members of each into the government through targeted recruitment, which mirrors and reinforces the effects of the territorial parcelization. By appointing members of those groups to offices in a higher proportion than their ratio in the wider population (Benson and Svanberg 1998:121), the system dilutes the already meager influence of Uyghurs and gives other groups disproportionate authority in the system.
The second major component of the CCP’s representational politics in Xinjiang is the careful selection of non-Han officials. Beijing has successfully co-opted many Uyghurs and members of the other, smaller groups. Through the careful selection, training, and promotion of loyal Uyghur cadres, the CCP has added substantial numbers to the government without compromising its policymaking autonomy. By law and tradition, the head of government of each autonomous unit from provincial down to township level is a member of the titular group; thus every head of Xinjiang’s government has been a Uyghursince 1955; each head of Ili Prefecture, a Qazaq; and so on.51 Uyghurs and other non-Hans in regional and local government are frequently called on to announce the party’s unpopular policies, all but guaranteeing that those officials will not develop a local power base and blunting the criticism that Hans alone rule the region.52 The recruitment has followed a consistent pattern.
Uyghurs and other non-Hans have been best represented at the lowest levels of government. By mid-1961, more than 85 percent of county magistrates and deputy magistrates were non-Han, and more than half the commissioners and deputies at the district, prefectural, and regional levels were non-Han. Yet according to McMillen, “the key departments and organs of Xinjiang administration … largely remained in the hands of Han CCP members.” In October 1965, non-Han cadres comprised 106,000 of 190,000, or 55.8 percent of the total. The continued (though reduced) preponderance of non-Hans was again tempered by a skewing toward the bottom: fewer than 10 percent of non-Han cadres were leaders at the county level or above (McMillen 1979:48, 75–76).53 Furthermore, the ratio dropped dramatically during the Cultural Revolution (1966–1976) when tens of thousands of non-Han cadres received damning political “labels” and were dismissed from their positions. The ratios fell most quickly late in the Cultural Revolution not because of purges but because of a massive influx of Hans into government positions. Between 1974 and 1975, the number of Hans in government more than doubled, from 133,000 to 269,000, while the figure for Uyghurs rose from 62,000 to 67,000 (Zhonggong Xinjiang weiwu’er zizhiqu weiyuanhui zuzhibu 1996:730). Even in 1979, three years after the fall of the Gang of Four, non-Hans held only 29 percent of cadre positions (Benson and Svanberg 1998:109).
Deng Xiaoping’s reforms, and in particular the policies advocated by Hu Yaobang, reversed this trend with moves both to reduce Han dominance and to restore non-Hans to their former positions. In July 1980 the Central Committee Secretariat, headed by Hu, heard a report on “Xinjiang work,” and in September it convened a symposium on the topic. Wang Zhen was then deputed to convey the sense of the September meeting to Xinjiang’s leadership. In November, the Xinjiang Party Committee explicitly stated for the first time that the proportions of minority minzu cadres in both party and government ought to reflect the proportions of those non-Han groups in the broader society. This objective was to be reached by a combination of recruiting more non-Han cadres and reducing the number of Han officials. To achieve the latter goal, some Han officials would be retired or furloughed; others would be given advisory roles; and many would be transferred to posts in China proper. Party scholars later described as “one-sided” the decision to make the complexion of officialdom numerically representative and condemned the initiative to remove the Han cadres as based on “an incorrect assessment of relations among minzu” (Dang Yulin and Zhang Yuxi 2003:277; Zhang Yuxi 1993:358).54
By the end of 1983, tens of thousands of non-Hans, now considered to have been “wrongly labeled,” had been reinstated. The combination of reinstatement of old cadres and the hastened training and selection of minority minzu stipulated by Hu brought the number of non-Han cadres up to 181,860, a substantial rise. On Hu’s orders, the ranks of Han cadres had shrunk dramatically over the same period. In 1982 alone, they were reduced by more than one-fourth, from 320,000 to 230,000. But despite the large increase in non-Han cadres and the departure of so many Hans, the percentage of non-Hans remained, at 43.1 percent, more than ten points lower than it had been in 1965. This reflected the fact that the purge of Han cadres in 1982 had been three-quarters less than the size of their prior influx into officialdom in 1975 (“XUAR gaikuang” bianxiezu 1985:52–54; Zhonggong Xinjiang weiwu’er zizhiqu weiyuanhui zuzhibu 1996:832).55
Chinese texts that showcase long series of figures on non-Hans at various levels of government—and all works that discuss the system of autonomy do so—delicately sidestep the key indicator of political authority: party first secretaries. In the Soviet Union, first secretaries in autonomous units were members of titular minority groups, while the top-ranking vice secretary was invariably a Russian. Beijing reversed the Soviet practice, with a Han in the top slot and a member of the titular group in the second position. One observer noted drily that there was “never … any suggestion” that top party officials in any of China’s autonomous units would “need to be members of the relevant nationality” (Mackerras 1994:156). Another averred that in Mao-era Xinjiang, the party secretary who was found in every government office and every work unit and who was the final authority on all decisions was “normally a Han” (McMillen 1979:48). Careful inspection of party rolls indicates that these observers were much too conservative, as the evidence suggests that there has been a tacit but rigid rule, with a vanishingly small number of exceptions, that the top party secretaries must be Han. This pattern has not changed appreciably (Yee 2003:449). Three decades into the reform period, it is exceedingly unusual to find a non-Han first party secretary in a government bureau, state-owned enterprise (SOE) work unit, or private enterprise.56 Non-Han first secretaries in local or regional government are never found. In 2000, all of the 124 secretaryships at prefectural, municipal, and county levels were occupied by Hans (Becquelin 2004b:363). In 2006, the official government Web site for Xinjiang party affairs displayed the names and photographs of all secretaries at the prefectural level and above, not one indicating an exception to this rule.57
Indeed, the percentage of non-Han party members has remained far below the proportions in the population. In 1987 only 38.4 percent of party members in Xinjiang were non-Han, even though non-Hans made up more than 60 percent of the population. By 1994, the percentage of non-Han party members had fallen further, to 36.7 percent, before rising to 40.3 percent in 2004.58 Despite the calls for “nativization” of government ranks in autonomous regions, there has never been a corresponding initiative in the party. The XUAR’s government chairman, Säypidin Äzizi, had warned in 1957 that proposing to increase the proportion of non-Hans in the party was “reactionary” (McMillen 1979:93). Decades later, officials continued to eschew targeted recruitment, claiming that party members had no particularist loyalties and were therefore capable of representing all groups without prejudice. With campaigns on the wane and the idea of “reactionary” political errors increasingly implausible, officials updated the language of justification, calling demands for minzu quotas in the party “unscientific” (Guo Zhengli 1992:92).
As with non-Russian groups in the Soviet Union that were underrepresented in the Communist Party of the Soviet Union (CPSU), several explanations were plausible: either the CCP considered non-Hans politically suspect, or they were averse to joining the party, or both (on the Soviet Union, see Armstrong 1992:243). Not only have non-Hans been reluctant to take responsible positions in government or enter the CCP, but many have become acutely frustrated with the political system after doing so. An official Chinese study provides a tantalizing piece of evidence of Uyghur cadres’ deep and increasing frustration. Whereas in the 1950s, Uyghur separatist organizations were composed mainly of Guomindang diehards, former officials and soldiers from the Eastern Turkestan Republic, and angry ordinary citizens, by the 1990s the organizations contained a substantial and increasing number of cadres and party members in their ranks (Ma Dazheng 2003:80).59
Thus Uyghurs are quite underrepresented in the government apparatus, particularly at the higher levels, and are even less well represented in the party. In a sense, recruitment policies have dovetailed neatly with the policy on immigration. Over time, the increasing numbers of Hans have made it easier to justify the Hans’ predominance in government. In interviews, Hans invariably approved of recruitment patterns, whereas those Uyghurs willing to discuss the matter strongly objected.60 Even those who themselves were not implicated in the Han–Uyghur rivalry felt that the system was skewed. A Xibo man who had once announced proudly to me that he had not suffered any discrimination because “no one knows I’m not Han” admitted months later that he had lost patience with official propaganda contrasting equal opportunity in China with discrimination in the United States: “Our media say the U.S. discriminates, but even so, a black man can become head of the Defense Department. Of Commerce. Of the Supreme Court. In China, does the Supreme Court have a single minzu? Does the Military Affairs Commission in Beijing? Damn!”61
Such comments cannot be found in newspapers or other media. Indeed, the same system that placed Han party officials in charge of non-Hans at all levels of government has, with the exception of a brief period in the 1980s, prevented the open expression of grievances about that fact.
The rarity of Uyghurs and other Muslims in positions of ultimate authority in Xinjiang has had a clear impact on the handling of religion there. While the government’s control of religion was noticeably relaxed during the reform era in many areas of the PRC, as among the Dai in Yunnan, the Hui in the north and southwest, and even to some extent in Tibet, the same cannot be said for Xinjiang.62 To be sure, the party did loosen control of religion in the early 1980s as part of its overture to Uyghurs and others after the antireligious excesses of the Cultural Revolution. One obvious sign was the great number of mosques that were rebuilt. Before the Cultural Revolution began in 1966, there around 5,500 mosques in Kashgar District in southern Xinjiang, of which 107 were in Kashgar City. The prefectural figure was far below that in 1949, as a “religious investigation team” dispatched from the Xinjiang Academy of Social Science noted with satisfaction.63 This had been the consequence of the “greatly elevated consciousness of the masses,” although the team’s report also admitted that quite a few mosques had been taken over by communes during the collectivization movement in 1958. One indication of the Cultural Revolution’s devastation can be seen in the further reduction in these figures. Of the original number for the Kashgar District, only 392 mosques remained usable by the 1970s, and only two stayed open in Kashgar City.
Muslims responded vigorously to Deng’s broadening of cultural policies at the beginning of the reform period. In 1980 and 1981 alone, communities reclaimed or rebuilt more than two-thirds of the original number, so that by the end of the latter year, there were 4,700 mosques in Kashgar District and 93 in Kashgar City (Zongjiao yanjiusuo Yisilanjiao diaocha zu 1983:21–23). Mosques continued to be built through the 1980s. Many villages had more resources as a consequence of the agricultural reforms of the early Deng era, and they had decided to build mosques with their new wealth.
A violent uprising in the southwestern town of Baren in 1990 caught officialdom by complete surprise. In addition to the uprising’s violence and its evident popular support, what most stunned party leaders was the news that the participants had announced openly that Islam would conquer Marxism–Leninism (Liu Zhongkang 1990:50). In the aftermath of Baren and in response to the shocking collapse of the Soviet Union, the government reversed its previous policy of tolerance. Officials prosecuted “illegal religious activities,” defrocking suspect clerics, breaking up unauthorized scripture schools (mädräsä), and halting the construction of mosques. In 1991, 10 percent of roughly 25,000 clerics examined by officials were stripped of their positions (Harris 1993:120–21). The party instituted new regular political examinations for imams, decreeing that only those judged patriotic and politically sound could continue to serve. Those restrictions, along with the stipulation that all new clerics be trained at Xinjiang’s sole religious institute in Ürümci, have continued to the present. After a decade of turning a blind eye to mosque building, officials felt that their construction had exceeded acceptable limits. In 1990, cadres in Akto County (where Baren is located) closed fifty mosques judged to be “superfluous” and canceled the construction of one hundred more out of fear that religion was getting out of control (Dillon 2004:73).
Despite the crackdown, propagandists in Kashgar continued to complain of “indiscriminate (lan) construction of mosques” five years later: by 1995, the Kashgar District had some 9,600 mosques, more than twice the number in 1981 and, in the judgment of atheist party members, “already plenty enough to satisfy the needs of normal religious practice by the believing masses.”64 Certain religious personnel, officials warned, engaged in wanton building, with the excuse that the number of religious sites were inadequate.65 Officials then sought to halt as well as to reverse the trend of mosque construction. Between 1995 and 1999, officials in Ili Prefecture demolished seventy mosques or religious sites, and their counterparts in Ürümci razed twenty-one sites in 1998 alone. Mosques near schools were specially targeted for destruction, lest they continue to exercise a “negative influence” on pupils (Agence France-Presse 2001a; Human Rights Watch 2005:56). In 1996, I observed the flattening of several mosques in the vicinity of Xinjiang University.66
Officials were even more concerned about clerics opening private schools and Qur’an study classes without official permission, or taking on religious pupils (talip).67 There were reportedly more than four thousand talip in Kashgar District in 1995 (Kashi diwei xuanchuan bu 1995:37). Concern about renewed private religious study and growing popular religiosity led to further repression in the following year. Michael Dillon noted that both national and XUAR governments directed more attention to the influence of religion on regional unrest beginning in the latter half of 1996 than they had since the founding of the PRC (Dillon 2004:90). Officials closed all underground schools and cracked down on both pupils and teachers, labeling all private scriptural study “illegal religious activity.” Even so, such schools continued to spring up (Ma Dazheng 2003:99). Alarmed at the emergence of religious networks that were neither organized nor sanctioned by the state, Wang Lequan imposed new restrictions on religious education in December 1999:
In general, patriotic clerics [should] have two to three students or, in big mosques, three to four; more is unacceptable. The practice of appointing mollas from outside districts is strictly forbidden. We cannot give them the opportunity to link up by forgetting this. Originally, there were no contacts between village and village; once mollas were laterally appointed, it produced a problem.
Wang derived this “lesson” from the activities of Abduqadir Ayum, damolla of Qaghiliq (Yecheng), who had been the teacher of members of an alleged separatist gang in Khotän that officials were trying to “mop up” in December 1999. Equally, if not more, significant, Abduqadir had managed to train some eight hundred students and dispatch them to various parts of Xinjiang, building a network that Wang regarded as deeply threatening. Wang’s new rules were plain: “Wherever an ahong [a Muslim cleric] is trained, that’s the mosque where he serves; there will be no lateral appointments” (Wang Lequan 1999:6). Wang’s fiat was later transformed into law. The 2001 revision of the 1994 law on religion, much stricter than its predecessor, included an “on-the-spot” rule, which meant that imams from one town could not preach in another, nor could a mosque receive people from different parts of the XUAR (Human Rights Watch 2005:39 and n. 80).68
The treatment of Islam in China’s interior has been much more relaxed. Religious schools for young students continue to thrive and enjoy state protection in China proper. It seems clear that Hui Muslims have been granted so much more religious latitude because almost none have challenged secular authorities (Lipman 1997; Savadove 2005; Yardley 2006). The Hui religious leader of a Sufi sect in Ningxia was reportedly allowed to establish a “virtual religious state” with one and a half million followers and a wide network of mosques and religious schools. Remarkably, this leader was able to do so despite openly acknowledging having heard Osama bin Laden speak and meeting many fundamentalist clerics while studying for five years in Pakistan. The price of this freedom was his profession of absolute loyalty to the Chinese state and service in the Ningxia People’s Congress (Savadove 2005).69
Not only Islamic institutions but Muslims themselves have been treated quite differently across regions. Every version of the PRC’s constitution has maintained, and textbooks on religious policy repeat, that all citizens have two freedoms with respect to religious belief: the freedom to believe and the freedom not to believe. Officials have worried for years that continued religious belief threatens party authority and provides a breeding ground for Uyghurs’ political mobilization. The Xinjiang government’s chosen strategy has been to protect the freedom of people not to believe and to “dilute religious consciousness” in the population (Yin Zhuguang and Mao Yongfu 1996:233). The principal aim of official policy in securing freedom of belief is to make religion a “personal matter” (sishi). Officials have clearly grasped that attempting to force people to stop believing in or practicing religion does not work and may in fact provoke a strong backlash. They have come up with a clever second-best strategy that, even if not fully effective, will have a profound impact on Xinjiang’s Muslim populations. Over the long term, the party aims at no less than ending the transmission of religiosity from one generation to the next. Interrupting that transmission will in turn serve the larger project of reducing the distinct (and oppositional) identities of Uyghurs and other Turkic Muslims.
The party has placed special emphasis on eliminating the pull of religion on two groups: party members and students, which has been a matter of growing urgency. Since 1978, substantial numbers of Uyghur and other Muslim party officials have become religiously observant. Surveys in 1983 determined that 20 to 30 percent of rural party members believed and practiced and that in some districts the rate was a high as 70 percent (Zhu Peimin, Chen Hong, and Yang Hong 2004:222). The proportions appeared to have held steady or risen further by the end of the decade. In 1989, a survey in Lop County found that more than 58 percent of party members took part in religious activities, and another survey in Qaraqash (Moyu) found that 83 percent took part. To higher officials’ horror, even party branch committee members and secretaries prayed daily, and some secretaries even became imams. These outcomes were characterized as a “loss of party and political control” in the bottom rungs of the party (Ma Dazheng 2003:15).
Even though the constitution guarantees these two freedoms, party cadres and students are now openly denied the right to believe. As one text explains,
Ordinary citizens are permitted two freedoms. Though party members are also citizens, they are first of all members of the party of the proletariat, and therefore enjoy only one freedom—the freedom not to believe—and absolutely do not enjoy the freedom to believe. They cannot have a foot in two boats. (XUAR Party Committee Propaganda Bureau 1997:52, italics added)
This novel explication of freedom has not persuaded many people. Many party members continue to believe and pray, even if they rarely dare to attend mosques. In at least one documented instance, one group responded to this difficult situation by simply removing their foot from the party boat, and in a very public setting. In 1999 in a village outside Kashgar, twenty-eight party members, including some village leaders, mounted a stage in front of more than a thousand locals and announced that they were quitting the party, whereupon they were embraced as Muslims by local religious authorities (Ma Dazheng 2003:117–18). In June of the following year in another village not far from Kashgar, eleven party cadres openly entered a mosque together, and fully one-half the party members in the village were religious, according to an official report. A month later, authorities in Qaghiliq, the site of frequent unrest since the beginning of the reform era, punished seven party members for various offenses related to religion (Kashi diqu shizhi ban 2001:318). There is little doubt that these scattered reports reflect a broader pattern of defiance by Muslim party members. The mere fact that in 2006, officials placed signs over the gates of mosques forbidding party members and government officials to enter suggests that the prohibition had not been entirely successful (RFA 2006a).
While in theory, students also are citizens, they now also are limited to a single religious freedom. Official explanations stress both the vital role of education in ensuring the future prosperity of the nation and the importance of allowing youths, once old enough to choose, to make a free, informed choice to believe or not to believe. For example,
youths and children are in the growing-up stage; their worldviews have not yet formed. They lack scientific knowledge and life experience. They cannot yet make responsible and scientific choices appropriate to their goals. To irrigate the minds of immature youths with religious thought is to allow someone to impose belief in a particular religion on them.(Luo Yingfu 1992:171)
The author pointedly ignores the fact that to prevent youths from practicing religion and others from teaching them about it is to allow another agent to impose unbelief. But such intervention has proven ineffectual. Prosecuting religious activity and expressly forbidding the teaching of religion on school campuses have not eliminated belief. A scholar noted with alarm in the mid-1990s that when asked which was greater, the “strength of Khuda [Allah] or the strength of science,” 71 percent of high school students in Changji Prefecture sided with Allah (Gong Yong 1997:4). Party strategists also have placed hopes on mandatory classes in atheism.
An Amnesty International report on Xinjiang claimed that the CCP began the “education in atheism” campaign in 1997 (Amnesty International 1999:n. 52), but a textbook published in 1992 that I purchased in Kashgar shows that the campaign began earlier. The text of the book makes clear that students no longer enjoyed “freedom of religious belief,” and the fifth lesson is devoted entirely to the bald assertion that “teenagers must become atheists” (Li Ailing et al. 1992:144 ff.). A reporter’s visit to a middle school in the central city of Korla in 1999 shows that the message did not remain inside books. An educational poster on the wall of the school listed “two musts,” the first being the obligation to fight “splittism” and the second, the obligation to “believe in Marxist atheism and not attend religious activities” (Marquand 2003). By 2000, academics in China were writing openly that atheism education was the way to “gradually reduce religious consciousness among the masses.” They urged that the government “try every conceivable stratagem to lighten the influence of religion and shrink the terrain held by religion.” The authors’ hostility to religious belief and practice was quite plain in the language they chose: the term “terrain” (zhendi) connotes territory held by the enemy in a military conflict (Long Qun and Guo Ning 2000:78). During interviews in 1996–1997 and again in 2002, secular Uyghur intellectuals, even actively antireligious ones who bemoaned the conservatizing influence of Islam on Uyghurs, criticized the party’s putting so much pressure on believers.70
The party-state has taken the more aggressive step of forbidding anyone under the age of eighteen to enter a mosque. Although signs excluding minors are not visible outside the major city mosques in the capital city of Ürümci, they can be seen elsewhere in the region (Fuller and Lipman 2004:335; Mackerras 2004b). The placard warning party members and government officials not to enter also includes youths under eighteen. Questioned in 2006 about who was permitted to pray in mosques, the imam of the Idkah mosque in Kashgar, the largest in Xinjiang, initially responded that every Muslim could do so. Then, after being reminded about the rules in other mosques, he admitted that the same restrictions applied in Idkah (RFA 2006a).71
In the mid-1990s, students at Xinjiang University were fully aware of the increasingly stringent official policy. They observed privately that many classmates continued to pray five times a day and to participate secretly in study groups. But the costs of doing so were readily apparent. For instance, in the spring of 1997, at the entrance to the campus computer building, a series of posters with gaudy vermillion stamps indicated that six students from Khotän had been expelled and arrested for attending religious study groups and that they had received substantial prison sentences. In 1999, parents of students at another postsecondary school received letters warning them their children had been caught praying or fasting and threatening expulsion if they were caught again. Two years later, a female student in Kashgar was in fact expelled for performing namaz (Human Rights Watch 2005:62, 110). In 2005 Xinjiang University distributed a 142-page student handbook with specific rules against students’ religious observance, although they are somewhat inconsistent. Some articles forbid students to participate in “illegal religious activities,” while others prohibit participation in religious activities of any kind. One even bars students from activities with a “religious coloring.” One regulation (dated 2001) specifically forbids fasting at Ramadan and mandates penalties for students who do.72 Another states that students will be given a disciplinary warning the first time they are caught taking part in religious activities, a demerit in their dossiers for a second occurrence, and expulsion if caught a third time (Xinjiang daxue xuesheng shouce 2005:57, 88, 124–25).
If students have been subject to particularly severe religious restrictions since the 1990s, the wider religious community felt a palpable and immediate effect in the aftermath of the September 11 attacks. In October 2001, authorities in Ürümci announced that they would increase the intensity of the “strike hard” campaign against “separatists and terrorists,” with particular emphasis on punishing “illegal religious activities.” In November the police closed down thirteen “illegal religious centers” in Kashgar and arrested more than fifty people they found praying there. In December the police arrested nine people in Bayangol for translating the Qur’an into local languages and “illegally preaching” to locals; they were accused of using religion to incite separatism. In the same month, police in Khotän detained a man and a woman, each accused of teaching young pupils privately about the Qur’an (Amnesty International 2002).73 Since that time, the restrictions have only increased. In 2001 the national government promulgated a revised form of the 1994 law on religion, with a subtle change in wording that indicated a major change in policy. Whereas the 1994 law had listed among its purposes the “protection of normal religious activities,” the 2001 law spoke instead of “regulating religious activities according to law” and “inducing religion to conform to socialist society.” As researchers at Human Rights Watch pointed out, even though the party-state had always been the sole arbiter of the distinction between “normal” and “illegal” religious activities, the 2001 law left the pious in even greater doubt about where the distinction lay. The law also enacted another subtle but significant change in applying to all believers the loyalty requirement originally applied only to clerics. This made the practice of religion by private citizens “conditional on support for government and Party leaders” (Human Rights Watch 2005:33–34, 37).
Further restrictions on religion came into force in March 2005, and again the effects could easily be seen. In July 2005, three Uyghurs waiting at a bus stop had their bags searched and were detained for holding “unauthorized” religious texts. In August of the same year, police burst unannounced into the home of a religious teacher and arrested her along with thirty-seven pupils for studying the Qur’an privately (Hoo 2005; Human Rights Watch 2006). Over time, the official rubric of “freedom of religious belief” has worn quite thin indeed.
The Uyghurs’ dissatisfaction with the political system in Xinjiang has been clear since the founding of the People’s Republic of China. Demands for “real” or expanded autonomy figured in most of the organized protests in the 1980s, and Uyghur informants privately expressed their wishes for greater collective autonomy to many researchers in the 1980s and 1990s (Beller-Hann 1997, 2002; Bovingdon 2002a; Smith 2000, 2002). Pressures to change the legal framework or the policies on autonomy have also come from other sources: Soviet-bloc writers and officials, more liberal Xinjiang officials urging the expansion of autonomy, and hardliners demanding its reduction. In the last decade, the United States, some European Union members, and international organizations have also begun to criticize the system.
From very early on, Marxists inside and outside China condemned the framework of autonomy. Following the Sino-Soviet split, Soviet authors scornfully denounced China’s system of autonomous regions as a betrayal of socialist principles. One argued that Beijing was denying non-Hans their “legal right to self-determination,” while another complained that in its place, the government was offering “only a truncated territorial autonomy” (Connor 1984:236).
Deng Xiaoping showed he was aware that this was still a live issue in the 1980s, as a comment that came to light much later revealed: “Xinjiang’s fundamental question is [whether it should be] a republic or an autonomous region.” Echoing Zhou Enlai’s 1957 remarks, Deng asserted that “we’re different from the Soviet Union. We can’t have a republic; we are an autonomous region.”74 Also like Zhou, he defended the Chinese system against Soviet-bloc criticisms. In 1987, Deng admitted to Hungary’s János Kádár that the Chinese government had consciously chosen not to have a Soviet-style federation of republics; yet he boasted that the system of autonomous regions was one of the best points of China’s socialist system and must not be abandoned (Zhu Peimin 2000:334, 337).
Party theorists attempted to lay the matter to rest in lengthy treatises. For instance, in a work billed as the first account of China’s system of regional autonomy integrating Marxist–Leninist theory and Chinese practice, the author acknowledged the CCP’s gradual abandonment of the principle of self-determination but claimed that it was entirely justified. The CCP did indeed “emphasize many times and over a long period the principle of the right to minzu self-determination,” though it ceased to do so after the victory over the Japanese in 1945. The CCP’s supporting the principle of self-determination in the 1930s and 1940s was “understandable” given its membership in the Communist Third International; so was its “use of the slogan of minzu self-determination” to motivate non-Hans to fight against imperialism. So, finally, was its decision to abandon the principle entirely in September 1949 as the Chinese People’s Political Consultative Conference deliberated the shape of New China. At that point, the repudiation of self-determination was “correct” and “a matter of course” because acknowledging it any longer would simply have served imperialist plans to split up China (Zhang Erju 1988:33–36).75 Officially supported scholars adopted a still firmer line in 1999, declaring that in contemporary Xinjiang, “to pursue ‘self-determination’ is to engage in splittism” (Pan Zhiping 1999:178–81).
These official comments and theoretical works have clearly been intended to answer Marxist critiques of the political-legal framework of autonomy. While neither silenced those critiques, Chinese officials can observe with bittersweet satisfaction that many of the critics lost their jobs and, in several cases, their countries. CCP officials often point to the disintegration of the Soviet Union and Yugoslavia as proof that Mao and his cohort made prudent choices in 1949 (Zhu Peimin, Chen Hong, and Yang Hong 2004:357). In the meantime, Uyghurs have continued to clamor for more autonomy (Becquelin 2000; Bovingdon 2002b; Gladney 2002, 2004b; Rudelson 1997; Smith 2000).
Already at the beginning of the reform era, a few officials and scholars had acknowledged the limitations, even the failures, of the system’s implementation. This is revealed, for example, in the recently published report of a group of specialists on minzu problems (which would have included a large number of non-Hans), trading ideas at a symposium on that topic in Beijing in 1980. Despite the party’s proclivity for “emphasizing the positive,” an occasional critical note slipped through. It was the consensus of the conferees that
the principal problem at present is that the powers of self-government and decision-making powers pertaining to the policy of minzu regional autonomy have not yet really been implemented (wei neng zhizheng de luoshi)…. Some comrades noted that minzu equality has not yet been successfully established in people’s minds, compromising the implementation of decision-making authority. There are those who see minzu regions purely as “materials supply sites” and seldom concern themselves with the economic development of [those regions]; they infringe on the legitimate interests of minzu regions. The problem of nativization of organs of self-government has for quite a few years now, and especially during the time of the “gang of four,” become a virtual taboo. The proportion of minzu cadres in organs of self-government has fallen drastically, producing a situation in which “minority minzu manage household affairs (dangjia), Hans make the decisions” (zuozhu).
Public acknowledgment of the problem did not bring about much change over the ensuing decade, in spite of Hu Yaobang’s reforms. In 1988 the author of an officially supported national study admitted with concern that most non-Han groups continued to be underrepresented in government in the autonomous regions, as reflected in the gap between cadre ratios and population ratios. He argued that this was a critical problem for the system’s implementation. Without a substantial body of non-Han cadres, he wrote, “the true realization of …autonomy is impossible.” The only solution was to train non-Han cadres in large numbers (Zhang Erju 1988:287). Yet after another twenty years, the gap had not changed appreciably.
A dominant group of high officials viewed with suspicion the renewed calls for nativization. As previously noted, many believed that rather than satisfying the demands of disgruntled Uyghurs, the looser and more tolerant policies enacted under Hu Yaobang had only whetted their appetite for more and thus caused the demonstrations and violence of the 1980s and 1990s. Many contended that the criticism of the Cultural Revolution’s leftist excesses had provided cover for the emergence of ethnonationalist extremism. Ma Dazheng cited as evidence a letter submitted to the Xinjiang government in 1980. The letter’s author, understood to be Uyghur, complained that “in Xinjiang, the leadership of the party has always meant leadership by Han cadres.” To remedy the situation, the Uyghur author proposed setting ratios for cadres:
Xinjiang is the Uyghur autonomous region. This point absolutely must not be forgotten or overlooked. Therefore … Uyghur cadres should constitute no less than 45 percent of the total (not less than 60 percent in southern Xinjiang), and cadres from the autonomous region’s other minority minzu should make up not less than 15 percent.
Reading the letter almost two decades later, in 1998, Ma was disturbed by two aspects. First, its call for “full autonomy” smacked of “serious minzu attitude” (qianglie de minzu qingxu) and appeared to him to conceal a wish for independence. This concerned him particularly because Kosovar Albanians had recently raised the standard of full autonomy to cloak their quest to secede from Serbia. Second, in the two decades since the letter was written, its spirit had not only developed into an “erroneous intellectual current” but also spread widely among the Uyghur masses, making them susceptible to incitement and organization by separatists (Ma Dazheng 2003:185–86).
Facing pressures for more autonomy, hard-liners instead have pushed measures to diminish its scope. These include stimulating increased Han immigration and reducing the proportion of non-Han cadres by recruiting more Hans: “Hanization” instead of nativization. Policy advisers on minzu affairs to the Xinjiang government argued in the mid-1990s that Han flight and more relaxed family-planning controls on non-Hans had damaged the balance of populations in the region. They urged the government to adopt targeted policies to remedy the “out-of-balance population ratio” by luring Han cadres and members of the masses to Xinjiang, although they noted pessimistically that no policies could swiftly bring about radical change in population ratios (Mao Yongfu and Li Ling 1996:180).
Ma Dazheng made a more searching critique. It was not just the ratios themselves that were troubling, but non-Hans’ focus on them. Directly contradicting the policy analysts of a decade earlier, he opposed equating cadre ratios with autonomy, hammering away at a fundamental premise of representational politics. To do so carried a grave risk: “The minute one raises ‘autonomy’ in Xinjiang, people, and especially comrades from minority minzu, think immediately of the question of ratios of cadres and civil servants. The two have practically become synonyms. This has become a ‘big and intractable problem’ difficult for local governments to resolve.” In fact, it has become fertile “soil for ‘ethnocentrism.’” Ma repeatedly proposed recruiting still more Han cadres, though he urged that local Hans be chosen in preference to those from China proper.76 He also put into words the principle that had governed cadre recruitment over decades but had never previously been stated openly: The government preferred Hans over non-Hans as cadres in Xinjiang. “We should,” he announced, “see the special usefulness of Han cadres from the lofty strategic vantage of Xinjiang’s stability” (Ma Dazheng 2003:21, 124, 188–89). In his policy proposals for counteracting the growing minzu consciousness and assertiveness of Uyghurs, Ma candidly urged that the government “weaken the ‘minzu ratio system’ for appointing cadres, changing over to a ‘minzu inheritance system’ for cadre positions” (Ma Dazheng 2003:201).77 To eliminate the “ratio system” would mean decisively repudiating the principle of political representation. There is no discernible doctrinal justification for switching to “inheritance” of cadre positions, but such a change would have two practical consequences. First, assuming that the Han population will continue to outstrip that of non-Hans, it would not just perpetuate but increase the underrepresentation of Uyghurs. Second, it would justify the permanent adoption of the now fifty-year-old practice of assigning Hans to all party first secretary positions at all levels.
Although Ma’s study was supported by the highest levels in Beijing and Ürümci, it is not always clear whether his policy suggestions were true proposals or scholarly justifications for existing practices. Given that two of the top officials in Xinjiang acknowledged having read the chapters of his book in the prefaces they wrote for it, we can be certain that had his policy suggestions not been officially approved, they would not have been published. In either case, we can detect high officials’ inclination to move in this direction in their retreat from the language of definite ratios. By the late 1990s, in calling for the training of “politically aware minority minzu cadres” who could implement party policies correctly, Jiang Zemin announced simply that China must have “a group” (yi pi) of them (Huang Guangxue 2001).
Scholar-officials have attempted to endow these changes, which might otherwise be seen as manifestations of hypocrisy or blatant chauvinism, with the sanctity of Marxist principle. In a work earning the imprimatur of the XUAR Party Committee, a professor at Xinjiang’s Party School wrote that “it is no longer permissible, as in 1957 or 1980, to raise demands for ‘nativization.’” He went on to argue that nativization and ensuring that various minzu were masters of their own house were fundamentally different matters. Korenizatsiia, once slavishly copied from the Soviet Union, was inappropriate for China’s national conditions, and “judging from the present vantage, one cannot even say it was appropriate for the Soviet Union” (Zhu Peimin, Chen Hong, and Yang Hong 2004:357).
Some scholars have even obliquely proposed abandoning the commitment to autonomy altogether, almost certainly with official support. In the reform era, academics and low-level functionaries frequently publish theoretical essays that float ideas being contemplated in the government. This provides what Americans now refer to wryly as “plausible deniability” for potentially unpopular actions. Thus the young scholar Zhu Lun, a researcher in the Minzu Institute of the Chinese Academy of Social Sciences, asserted in 2002 that the idea of autonomy was obsolete and no longer suited to global realities. Collective rule (gongzhi) is now the “basic tenet of minzu politics in contemporary multi-minzu states.” He therefore proposed that policymakers begin to think about “postautonomy” (hou zizhi) (Zhu Lun 2002:4, 6).
The appearance of a similar article later in the same year strengthened the supposition that the idea had official support. Commenting several months later on Zhu’s article, legal scholar Du Wenzhong argued that “China’s praxis for resolving problems of minzu relations should be understood as … ‘collective rule’” (gongzhi). Chinese jurisprudential scholars’ continued use of the term “autonomy” (zizhi) to describe minzu policies was both unwise and “inaccurate.” In reality, Du maintained, the newly amended law on regional autonomy stipulated “collective rule” by the state and local autonomous organs (Du Wenzhong 2002:8). This idea was echoed in policy proposals made specifically for Xinjiang, once again by Ma Dazheng. Prior work on the theory of minzu policies, Ma contended, had “put too great an emphasis” on minzu autonomy. This ought to be balanced by a greater stress on “collective rule” (Ma Dazheng 2003:188–89).
This chapter began by suggesting that Chinese policies in Xinjiang have been a key source of discontent and conflict, thereby implying that different policies might have produced less contentious outcomes. Historical counterfactuals are notoriously shaky ground for comparison. Furthermore, the frequency of ethnic conflict and violence in Xinjiang during the Republican era might have been expected to continue after 1949. The argument here is not that some set of ideal policies would have eliminated discontent and friction entirely. Instead, it is, first, that had the CCP hewed more closely to what Hannum and Lillich described as the minimal principles of autonomy, there would have been less friction. Second, the specific departures of Chinese practice from that minimal model exacerbated discontent in particular ways.
The large discrepancy between the promise of autonomy and the reality of Xinjiang’s governance did not originate at the regional level. As Pei Minxin (2002:319) pointed out, two features of China’s national political system have “in reality allowed almost no genuine local autonomy.” First, Beijing has kept power strongly centralized and resisted the devolution of much authority to regions, and second, party leaders have adamantly opposed substantial democratization, which might have allowed the articulation of, and bargaining over, local demands. Both features remain obstacles to beneficial changes, and they demonstrate the perspicacity of Hannum and Lillich’s argument that democracy and federalism are crucial guarantors of robust autonomy (Hannum and Lillich 1980:885, 887).78
Pei suggested that China’s system of minzu regional autonomy faced a crisis in the mid-1990s. The country’s market transition had dramatically reduced Beijing’s capacity to redistribute resources and therefore to subsidize the less-developed border regions. Second, both international criticism and domestic opposition placed greater restraints on Beijing’s capacity to use Han migration as a tool of control. Finally, in recent years, repression has proved to be less effective and more explosive in regions like Xinjiang and Tibet. Pei argued that the crisis would have to be resolved by strengthening institutions to protect the “rights and political representation” of non-Hans (Pei Minxin 2002:319–21, 323, 327).
Since Pei made this argument, changes in the international climate and China’s economic situation have reduced pressures to enhance autonomy in Xinjiang and other peripheral regions. In the 1990s, international scrutiny and the specter of humanitarian intervention gave repressive states pause, particularly in the wake of NATO’s intervention in Kosovo (Finnemore 2003). For a time, Beijing was gravely concerned that the United States or international organizations might use human rights abuses as a pretext for intervening in Xinjiang or Tibet (Dreyer 2000; Lawrence 2000; Ma Dazheng 2003:106–24). Then the attacks on September 11 gave Beijing the cover it needed to further reduce the scope of autonomy and step up political pressure in Xinjiang from 2001 on (Amnesty International 2002; Human Rights Watch 2005).
This chapter has argued that the system officially billed as providing Uyghurs “minzu regional autonomy” has in fact enforced heteronomy, as illustrated by the Hans’ monopoly on the most powerful party offices. The system of heteronomy has in turn enforced a series of policies that Uyghurs would not have chosen had they been able to govern themselves. The response of party officials to Uyghurs’ complaints has been to further concentrate power in the hands of Hans and to enact even less popular policies. The new proposals to do away with the very rubric of autonomy are unlikely to reduce the discontent, even if they help silence its expression. I have suggested that the configuration of power and the unpopular policies have been key sources of discontent and therefore conflict in the region.