The last chapter focused on “everyday resistance,” on the premise that most Uyghurs usually have been deterred from resisting openly by the threat of harsh punishments. Newspaper reports in the 1990s revealed to the outside world for the first time since the 1940s that Xinjiang was occasionally rocked by serious political violence or mass protests. These reports, coming on the heels of demonstrations in Tibet in the late 1980s and the collapse of the Soviet Union in 1991, fed speculation that China faced a looming crisis and might disintegrate along the lines of its former neighbor. Not until the end of the decade did new information emanating from China make it apparent that unrest in Xinjiang had occurred with some regularity over the previous fifty years.
This chapter turns to the evidence of open and organized resistance in the region and to the representation of that resistance. Rather than describing every reported major organized uprising or violent antistate attack (for that, see Dillon 2004), I focus on the frequency of protests and political violence and on the aims and strategies of the participants. There is no doubt that the events took place, with very few possible exceptions, but to stop at a straightforward accounting of the events and to ignore the ways they have been depicted by the Chinese government, Uyghur organizations, and other entities would be to miss a crucial dimension of the contention in Xinjiang.
The thousands of arrests each year since 2001 suggest that organizations have persisted and new ones are springing up, despite the state’s repressive efforts. But they also show that repression has deterred all but the most violent and fearless. Whereas large-scale protests had been largely peaceful from 1979 through the 1980s (which is not to say there were no violent events), by the 1990s small-scale riots or violent attacks had become more common. This shift began long before the “strike hard” campaigns, mentioned in chapter 2, officially emerged in Xinjiang in 1996.1 Forceful suppression of protests beginning in the late 1980s, the political atmosphere in China following the June 4 massacre in 1989, and the enactment of stricter regulations for demonstrations in 1990 combined to keep off the streets many people who might have joined protests in the more open climate of the 1980s. Repression increased even more dramatically after 2001, and episodes of protest fell further.
Curiously, official Chinese commentary depicted not a fall but an alarming rise in protest in the new millennium. In 2001, the chairman of the XUAR government, Ablät Abdurišit, announced that “the situation of Xinjiang is better than ever in history…. [T]here has been no room for national separatists and religious extremists. By no means is Xinjiang a place where violence and terrorist accidents take place very often” (quoted in Bao Lisheng 2001). Nonetheless, in 2005 the XUAR party secretary, Wang Lequan, warned ominously that “in Xinjiang the separatists, religious extremists and violent terrorists are all around us—they’re very active” (quoted in Sommerville 2005). Chinese academics suggested, too, that separatist threats and activity had exploded in that four-year period.2 In other words, there is no obvious relationship between official descriptions of the threat and the actual trends revealed by the independently compiled record of public protests. The depictions answer the exigencies of representational politics, rather than revealing the party-state’s perception of the threats.
This chapter makes four points. First, the sheer number of protests in Xinjiang since 1980 reinforces the contention of chapters 2 and 3 that the Uyghurs’ dissatisfaction with the region’s governance is deep and broad. The quantity of documented protest events also casts doubt on the Chinese government’s argument that major demonstrations were the work of a tiny minority of separatists and that the majority of participants took part out of naïveté or simple excitement. In addition, the political content of the demonstrations, as expressed in banners and shouted slogans, leaders’ programmatic statements, and handbills circulated secretly, strengthen the case that the specific criticisms raised by informants and artists were representative of widespread complaints and not the unhappiness of an isolated minority. At the same time, the evidence demonstrates that everyday resistance and the comparatively rare episodes of organized protest are part of a continuous political field.
Second, since protests have increased steadily throughout China proper since the 1990s, the dramatic decline in protests in Xinjiang (and Tibet) since 2001 is an anomaly. Whereas Xinjiang was once regarded as the wildest and most violent part of China, it appears to have ceded that reputation to the contentious factories and farmlands of China proper. Third, even though Uyghurs have expressed deep dissatisfaction with governance in Xinjiang and pointedly called for policy changes, Beijing and Ürümci have almost never responded by accommodating those demands or entertaining public discussions of the concerns. Instead, officials have strengthened unpopular policies and cracked down on both political speech and spaces for assembly outside party control. In other words, they have sought to limit as far as possible the further public articulation of discontent with those policies. Such unyielding responses have not resolved Uyghurs’ complaints and instead have often exacerbated them. Thus the anomalous drop in unrest in this famously contentious region cannot plausibly be attributed to the Uyghurs’ increased political contentment.
Nor is it easy to argue that rising material wealth has eased Uyghurs’ concerns with politics, given their high unemployment rate and the dire poverty of the Uyghur countryside in the south. Rather, the fall in protests reflects substantially increased political repression. Central Asian governments’ harsher suppression of Uyghur groups beginning in the late 1990s under strong pressure from Beijing also removed external sources of support for antistate activity in Xinjiang.
After the fall of communism in Eastern Europe and Moscow and the disintegration of the Soviet Union, Yugoslavia, and Czechoslovakia, many observers wondered whether China would follow suit. One motivation of various studies comparing the Soviet Union and China was surely a desire to avoid being caught flat-footed by sudden cataclysmic political changes, such as had shocked Sovietologists in 1991. Yet as the Chinese Communist Party continued to maintain its firm grip on power and the Chinese state seemed to remain strong year after year, talk of China’s being bound to follow the Soviet example grew progressively quieter. Then the heightened attention to conflict in China’s northwestern region of Xinjiang after September 11, and the publication in early 2006 of startling statistics on surging unrest in the Chinese heartland, seemed to make talk of Soviet-style regime and state disintegration plausible again.3 The figures on the rise and magnitude of protest events throughout China invited renewed comparisons with the revolutions of 1989 to 1991 that ended the reign of communist parties throughout the Soviet bloc. Discussion of violent unrest in Xinjiang again recalled the specter of the Soviet Union’s collapse, which began with protest and bloodshed in the Baltics, as well as the bloody departure of Bosnia from Yugoslavia. Reports of possible “Muslim terrorism” in Xinjiang, many of them hastily assembled by intelligence bureaus around the world, called to mind the brutal struggle in Chechnya as well as intrastate conflicts farther afield in Aceh and Mindanao.4
Chinese scholars have continued to think and publish about this topic. Academies of social science in Beijing and Ürümci sponsored research on the Soviet Union and Central Asia beginning in the late 1980s and with greater vigor after 1991. Numerous studies compared the Soviet Union and China on the dimensions of demography, minzu policies, governance, and economics. As recently as 2005 a book entitled China’s Borders and Minzu Problems began with a chapter on the “minzu problems and the lessons of the Soviet breakup” (Zhang Zhirong 2005:1–7). The second wave of political transitions in Georgia, Ukraine, and the Kyrgyz Republic similarly rattled the Beijing leadership, and now scholars all over China are researching the etiology of the “color revolutions” in hopes of helping the party stave them off. The sheer number of conferences and published articles focusing on the topic and Beijing’s decision to place severe legal restrictions on nongovernmental organizations (NGOs) in China belie officials’ public confidence that China will avert such an outcome (Xinhua 2006).5
Although there is wide agreement that the Chinese state is strong and maintains powerful authority over society, few honest observers deny that social groups have become much more restive in recent years.6 The transition from a socialist to a quasi-market economy has brought China one of the highest rates of growth ever sustained over two decades. It also has dislodged vast numbers of people from their jobs and farmlands. It has made many individuals rich and displaced an even greater number from the relative security of the socialist work unit into the stormy sea of market competition. Soon after Deng Xiaoping announced that class struggle was officially over in December 1978, real class struggle commenced in earnest. It is not idle to speak of the “unmaking” of the Chinese working class (Hurst 2004) or to conclude pessimistically that today’s Chinese laborers have lost everything that Marx promised the world’s proletarians had to gain from socialism (Blecher 2002, 2004). This process has provoked increasing waves of protest throughout the country (Tanner 2005).
Most demonstrations in China proper can loosely be called “economic protests.” Workers strike because of layoffs or unpaid wages. Farmers surround government offices to complain of land expropriation and exorbitant taxes. Both groups rise up against corruption by the cadres. In a pathbreaking article, William Hurst demonstrated that collective action by laid-off workers in China has varied dramatically by region. The precipitating factors, demands of protestors, and state responses differ among the northern “Stalinist rust belt,” booming central coast, and central inland regions, which he called “tentative[ly] transitional” because they have neither been crippled by layoffs by state-owned enterprises, as in the rust belt, nor enjoyed the same market-driven prosperity as the coast (Hurst 2004). Hurst’s analysis demonstrates the importance of multisited research in China, illustrating how regional differences in political economy crucially affect the aims and fates of workers’ protests. His study was spatially limited in significant ways, however. All three macro-regions in his analysis belong to “China proper.” We should broaden the study of protest to embrace China’s western periphery, and Xinjiang in particular, where we seem to encounter another realm entirely, if not several other realms.7
In the thirty years since reform began, only two demonstrations over purely economic issues in Xinjiang have been documented, one urban and one rural. In 2001 in the southern city of Khotän, around a hundred recently laid-off textile workers, mostly Uyghurs, demonstrated out of concern that their employer would not pay their severance. In that case, local government officials promised to make good any debts that the cash-strapped factory could not pay, and the protestors dispersed without any arrests (Dow Jones International News 2001). In 2004 Uyghur farmers and Qazaq pastoralists in Xinjiang’s northwestern Ili region protested what they saw as an unfair relocation package. They were angry about the construction of a hydropower plant requiring 18,000 people to relocate and about the gap between the promised compensation for the loss of land and an actual disbursement of only about 5 percent of the stated amount (RFA 2004).
In China’s peripheral regions such as Xinjiang and Tibet, most major protest episodes have not concerned economic matters (the pattern of protest in Inner Mongolia, with a more heterogeneous economy and an overwhelmingly Han population, more closely resembles that in China proper), and the political climate facing protestors is decidedly chillier. In fact, however, the two features are directly related. It is precisely because some protests in the peripheral regions do not target industrial firms or local officials but the very state itself that central and regional governments have been so much less tolerant of them. Mongols have protested Chinese attempts to thoroughly domesticate Chinggis Khan or officials’ suddenly canceling a performance by a popular band from Mongolia. Since the harsh crackdown in 1989, Tibetan monks have sporadically demonstrated against religious restrictions or the requirement that they openly condemn the Dalai Lama, and individuals have occasionally used bold gestures such as hoisting the Tibetan flag or crying out for Tibetan independence in the public square. Uyghurs have demonstrated against Xinjiang’s governors and policies and, in some cases, challenged the very incorporation of Xinjiang into China. What Uyghurs have not been able to do since Hu Yaobang’s fall from power is find high officials sympathetic to their claim that particular leaders have “failed to live up to some professed ideal or … not implemented some beneficial measure” (O’Brien 2003:53). Officials in Xinjiang and Beijing have taken the position that separatist aims lurked behind every protest—concerned, perhaps, that tolerating one kind of protest would be perceived as an opening for separatist agitation—and therefore have forbidden all.
Party officials in Beijing and Ürümci have worried that a large unchecked protest in one part of Xinjiang might mushroom into a broader anti-Chinese mobilization. If many Uyghurs are deeply dissatisfied (Bovingdon 2002a; Smith 2000, 2007; Yee 2003, 2005), and most of them refrain from expressing anger only because they fear retaliation, then it is quite plausible that protest would spread quickly were the party-state to stay its hand. The snowballing demonstrations in Eastern Europe and the Soviet Union showed that previously timid citizens could abandon a lifetime of quiescence in a very short time. Time and again in 1989, 1990, and 1991 as people flowed into the streets, more and more of their fellow citizens stripped off their public facade of support for the regimes until it became clear that the majority had joined the opposition and the leaders had no choice but to step down (Beissinger 1998; Kuran 1992). In March 1997, Rozi, a well-paid Uyghur professional with a steady job, told me that if conditions in Xinjiang continued to deteriorate, he might join the organized opposition. “I might decide that living is not worth more than dying,” he told me quite seriously.8 Without broad survey results, we naturally cannot say how broadly such a view was or is shared.9 But we should note the hedging in Rozi’s comment. If citizens living under repressive states remain in doubt about their neighbors’ true political views until moments of crisis, they surely also are uncertain of their own “tipping points”: that is, how far the situation must deteriorate before they act and what they might be willing to sacrifice for a collective goal in the heat of the moment.
In his article on workers’ protests, Hurst calculated that China faced “at least hundreds, and probably thousands,” of contentious events each year (Hurst 2004:95). This must have seemed quite a sensible estimate when the article was written, likely in late 2002. Hurst pointed out that the Chinese government had never publicized any figures on the quantity, frequency, or nature of mass protests. Many observers were consequently stunned when, beginning in early 2004, the State Security Ministry released a series of statistics indicating that Hurst’s estimate had been low by an order of magnitude.10 The statistics recorded 58,000 contentious episodes involving more than one hundred people in 2003, 74,000 in 2004, and 87,000 in 2005.11 The ministry also revealed at the time that there had been an almost monotonic increase in protests since 1997 (Tanner 2005).12 Because Chinese official statistics are notoriously unreliable, it is quite possible that even these large figures underestimate the number of disturbances. The numbers seem astonishing because of the party’s historical intolerance for organized protest and habit of repressing it harshly and because most Chinese citizens have consequently been loath to incur the wrath of the state by demonstrating. The government’s willingness to publish the figures was similarly startling.
Some scholars have leaped on these figures as the strongest indication yet that China may be on the verge of a new social revolution (Jiang 2006). But it is possible to read the publication of the figures differently.13 Beijing may have strategically released this information about unrest with both domestic and international audiences in mind. Domestically, it might have intended to convince the monied and middle classes to support the continued repression and oppose “premature” democratization, by implying that only the party and the thin line of security forces lay between the comfortable lives those people now enjoyed and a political-economic abyss that would make the Cultural Revolution look like a Mardi Gras celebration. The conventional wisdom has long predicted that a burgeoning middle class would press for political reforms in China as it has elsewhere in the world. But ironically, the increasing prosperity in China may have made economically successful citizens more skeptical of reforms and more sympathetic to hard-line party leaders (An Chen 2003; Tsai 2005). Some have argued that Public Security Bureau (PSB) officials publicized the numbers to wring more money for domestic security out of the national budget.14
Beijing also might have released the figures to gain sympathy from the international community. Such a move would not have been plausible ten years ago, since many international observers shared the opinion that popular unrest in China expressed deep and justified dissatisfaction with a brutal and unresponsive regime. Yet September 11 dramatically changed the international climate, giving many states the opportunity to recast domestic opposition as “terrorism” and their own efforts to squelch that opposition as contributions to the “global war on terror” (Dwyer 2001; Li Qi 2002; Millward 2004:10–11). Under these conditions, Beijing might well have considered it safe to acknowledge the sharp rise in protests to an international society increasingly intolerant of antistate violence.15 This has had especially poignant implications for Xinjiang and Uyghurs.
There is an irony to the role that Xinjiang has played in the story of post-1949 politics in the People’s Republic of China (PRC). Bloody clashes and bombings in China proper were frequently compared with those in the XUAR, the “really violent” part of the country, and many bombings were initially attributed to Uyghur separatists, only later to be revealed as the work of spurned lovers or laid-off workers.16 Transnational Uyghur organizations have asserted for years that Xinjiang was on the verge of a crisis and consequently broadcast news of every violent episode in the region. In some cases, they may have claimed responsibility for damage that actually was the result of natural disasters or industrial accidents. Possibly influenced by those claims, many foreign journalists and researchers have speculated that violence and unrest have increased in the region since the late 1990s.17
A careful and critical review of the evidence reveals that Xinjiang has been far quieter since 2001 than has any part of China proper. Despite the region’s reputation, no scholar has ever attempted to quantify the amount and frequency of violence there. Some have cited official Chinese statistics promulgated since 2001, but there are several problems with these statistics. Officials quickly and dramatically changed their strategy of representing unrest in Xinjiang. In the 1990s they generally suppressed evidence of protests or violence, sometimes even denying foreign reports of unrest (Agence France-Presse 1995, 1997). The rare revelations of episodic violence attributed it to “minzu splittists.” But the trend was not entirely systematic, and official numbers and attributions varied widely before being fixed by the State Council in 2002 (Guowuyuan xinwen bangongshi 2002). For instance, the XUAR government chairman, Ablät Abdurišit, claimed in a 1999 interview with reporters that “since the start of the 1990s, if you count explosions, assassinations, and other terrorist activities, it comes to a few thousand incidents” (Becquelin 2000:87).
Since September 11 the Xinjiang and national governments have had conflicting incentives in representing the scope and nature of unrest in Xinjiang. On one hand, officials at the regional and central levels seeking investment have habitually underplayed reports of unrest to avoid scaring away capital. On the other hand, regional governors seeking central grants for economic growth and policing, and officials in Beijing seeking global sympathy for China’s “plight,” have chosen to maximize the threat of separatists or “terrorists.”18 Most Chinese journalists and authors subscribe to the second strategy, touting large numbers of protests even in what had previously been depicted as placid periods and transforming splittists into terrorists and religious extremists. In other words, they first underrepresented and then exaggerated the number of episodes of political violence. I have attempted to replace speculations and distortions with systematic data collection, paying careful attention to content and sources. I have been able to document violent or organized protests or resistance in Xinjiang since 1949, including armed uprisings, peaceful demonstrations, and riots, as well as clearly political violence such as assassinations and bombings.
Figure 4.1 shows that between 1949 and 2005, there were at least 158 episodes of antistate violence or organized protest documented in printed sources, and of these, 142 had clear ethnonational content. The largest events involved 50,000 to 100,000 people, while most had only a few dozen participants. Only the armed resistance raised by various groups in the 1950s seriously challenged the party’s political-military control of the region. The period of greatest antistate or ethnonational protest since then was the mid-1990s, with a high point of twenty events in 1998.19 In addition, an event involving one thousand or more people took place in four of the five years from 1995 to 1999. Figure 4.2 plots events in the autonomous region against those throughout China from 1993 to 2005, showing that episodes in Xinjiang fell off just as they were increasing rapidly in China as a whole.20
The care with which it was assembled notwithstanding, my database of unrest in Xinjiang is unquestionably incomplete. One might infer its incompleteness from a comparison with the aggregate numbers cited in Chinese sources, even though the comparison would be misleading. For instance, Ma Dazheng, who presumably had access to internally circulated government statistics, cited “authoritative sources” to support the claim that there were 253 “violent terrorist episodes” in just the ten years between 1990 and 2000. Elsewhere he tallied 116 terrorist acts for 1998 alone, but these proved to include, in addition to bombings, assassinations, and arson, also livestock poisonings, kidnappings, and robberies (Ma Dazheng 2003:126–27, 153). These figures are an order of magnitude smaller than the “few thousand” invoked by the XUAR government chairman Ablät Abdurišit in 1999 but are equally out of step with Abdurišit’s 2001 comment that Xinjiang was “by no means … a place where violence and terrorist accidents take place very often” (Agence France-Presse 1999c; Bao Lisheng 2001).
FIGURE 4.1 Organized or violent events in Xinjiang, 1949–2005.
FIGURE 4.2 Protest events in China and Xinjiang, 1993–2005.
The elasticity of official numbers became obvious as the statistics promulgated in the State Council’s January 21, 2002, press release on “Eastern Turkestan terrorists” were applied without the slightest modifications to very different periods (Guowuyuan xinwen bangongshi, 2002). That document asserted that more than two hundred “terrorist events” left 162 dead and 440 wounded between 1992 and 2001. Less than two years later, officials assigned precisely the same figures to the period from 1990 to 2001, implying no one had died in violent attacks between 1990 and 1992, which hard to square with the well-attested evidence of casualties in Baren in 1990 and the bus bombings in Ürümci in 1992, among others. In 2005 the deputy director of Xinjiang’s Antiterrorism Bureau used the same figures for the “previous decade,” and officials continued to use the exact same numbers in 2006. Yitzhak Shichor put the case rather mildly when he observed that this use of an identical set of figures for very different time periods “casts a shadow over the rest of Beijing’s arguments” (Shichor 2006b:102).21
The numbers I use here are lower because my criteria are more restrictive than those adopted by officially sanctioned Chinese sources. I have not followed their examples in treating attacks on livestock or large robberies as instances of terrorism, or even as political violence, without supporting evidence, although I grant the possibility that some were perpetrated by individuals or organizations with political aims.22 Furthermore, my graphs and appendix include only those events described in narrative form in some source. Except in rare instances, I have insisted on at least two sources to confirm an event.
By contrast, whether in the State Council’s January 2002 document, in subsequent white papers, in the several graphic reports on “Eastern Turkestan terrorism,” or in reports for internal circulation, the numbers of violent events cited in statistics always far exceed those described in the narratives. The authors’ explicit mention of poisonings, crop burnings, and robberies in those narratives strongly suggests that the much larger figures on terrorism have been padded with figures from police blotters. Officially employed writers and spokespersons instructed to highlight “violent terrorism” in Xinjiang in order to garner international support seem to have elected only to count, and not to describe, episodes whose categorization as terrorist events might provoke skepticism abroad (Millward 2004:12). Since many individual cases claimed to underlie the aggregate numbers cannot be scrutinized, officials need not worry about drawing undue attention to the extraordinary breadth of the government’s definitions of terrorism and crimes threatening state security.23
Looking beyond the derogatory labels to the individual protest episodes themselves reveals much that is obscured by Chinese statistics. The events that triggered them, the organizations that spurred them, and the issues they raised are far indeed from the themes of global Islamism or transnational terror organizations. Not surprisingly, they are much more closely related to matters of governance and policy shifts in Xinjiang itself.
Careful scrutiny of the messages of public protests in Xinjiang reveals substantial overlap with the critiques discussed in the previous chapter. The willingness of large numbers of Uyghurs to march under particular banners or shout specific slogans strengthens our confidence that the criticisms raised by “everyday resisters” reflect broader views in Uyghur society.
In chapter 3 I argued that the myriad forms of everyday resistance in Xinjiang not only expressed dissent but carried out a kind of political work as well. That is, they communicated that dissent widely despite the powerful bans on public expression and organizing. Jokes traveled the breadth of the region in private conversations and via social gatherings of trusted friends. Writers and musicians made strategic use of the Xinhua distribution system itself, one of the party’s key tools for spreading propaganda, to broadcast well-hidden but subversive messages in tapes and books across the entire Uyghur-speaking community. Individual books passed through many people’s hands, and tapes could be duplicated at roadside stands (Dautcher 2000; Harris 2001). Turghun Almas’s historical writings gained a wide readership and an even wider “rumorship” (Bovingdon and Nebijan Tursun 2004). Letters and handwritten manuscripts circulated widely by hand, concealed in bags or clothes. Some textual and audiovisual materials from dissident groups in Central Asia and Turkey were smuggled into Xinjiang by traders or travelers and were passed around through social networks.24
Chinese sources provide some information about the Uyghurs’ spreading ideas through networks, although in their dogged emphasis on quantifying pieces of paper and documenting smashed organizations, such reports betray a studied uninterest in the messages being passed—or, perhaps more likely, a choice not to risk disseminating their contents any further. The aim has been to vilify separatists without attempting to understand them or make their objectives comprehensible to others.
An internal-circulation report in 1993 suggested that in the latter half of 1988 in the four districts of southern Xinjiang, officials laid hands on 113 “reactionary posters,” handbills, or anonymous letters. All the 127 people they caught with these materials were under the age of twenty-five, and the youngest ones were only twelve; the vast majority were elementary and middle school students (Zhang Yuxi 1993:348). Ma Dazheng claims that between 1990 and 2000, the state destroyed 503 splittist or violent terrorist organizations or gangs. He includes under the subheading “violent terrorist incidents,” 953 cases of subversive propaganda or incitement, and of these, 458 cases involved “reactionary posters,” 107 cases leaflets, 157 letters, and 231 other types. Curiously, another highly placed author cited “incomplete” statistics showing that over the same period, the number of “reactionary” handbills and posters advocating minzu splittism averaged 5,000 a year, and reactionary audiotapes, another 1,000 cassettes. There was a marked uptick in 1996, with more than 8,000 handbills and more than 10,000 audiotapes (Yang Faren 2000:243).
This corpus of words and artifacts shaped and disseminated a critique of the political order in the autonomous region. Sociolinguists would recognize in this a kind of “framing”: the purposive selection of particular ways to represent social phenomena (Goffman 1974; G. Lakoff 1987; G. Lakoff and Johnson 1980; R. Lakoff 2001). Scholars of contentious politics use “framing” to denote two factors critical to social mobilization: the strategic representation of a sociopolitical situation as objectionable, and the proposal of action to remedy it (Benford and Snow 2000; Snow et al. 1986). Only if activists compose and propagate a frame with wide appeal will large numbers of people decide to join the movement, often a risky choice even in democratic political systems.25 In authoritarian polities, activists often rely on “mass frames,” which cannot be spread openly and are harder to shape, so they may rely much more heavily on already circulating ideas (Hurst 2004:102–5). It is not clear, however, that activists in democratic systems have all the advantages. Popular media are widely read and trusted, as in many liberal polities, whereas movements have only a limited capacity to reframe perceptions already shaped by those media (Tarrow 1994:23). Under authoritarian regimes, the official media—often the only kind—are treated with skepticism, and their influence on popular opinion is correspondingly weaker. Widely disseminated “hidden transcripts” may be more powerful, particularly when suddenly made public during demonstrations. And if suspicion of official media gives dissidents an advantage in China proper, the Uyghurs’ far stronger dubiety toward the Chinese media may give even greater power to the dissenters in Xinjiang.
The critiques and other forms of “everyday resistance” described in the previous chapter seldom called people to action. They were broad normative statements, not practical proposals: wealth needs to be distributed more fairly; Uyghurs need truly representative leaders; the government must not impose family-planning policies on non-Hans. As I suggested, in the spring of 1997, large numbers of Uyghurs seemed to trust, or at least hope, that others were taking care of organizing a resistance movement, since they themselves feared to do so.
The comparatively rare episodes of open political resistance in Xinjiang provoke fresh questions about framing. When the demonstrations were spontaneous, why did people join so quickly, and what did they hope to accomplish by doing so? When public protests or actions seemed to have been planned in advance, what messages did the planners propagate, and by what means? How did potential participants decide to take part, even after reflecting on the risks and the low chances of success? We also should ask what purposes or messages can be divined from acts of violence perpetrated by small groups. Were the targets of assassinations or bombings clear? Were the aims easy to understand? These questions are easy to ask but very hard to answer. The available sources of evidence pose particular difficulties for the study of framing in Xinjiang. These problems bear on the amount of information we are able to squeeze out of the available record, and thus on the soundness of interpretations.
For years, those interested in individual episodes of open resistance in Xinjiang—whether collective or violent or both—had no choice but to sift through scattered and sketchy foreign newspaper articles, accounts by human rights groups or transnational Uyghur organizations, and the very occasional Chinese news report. Because Xinjiang has been closed to foreign reporters for long periods, outside media reports have sometimes been hampered by relying on foreign travelers with little local knowledge. An Agence France-Presse report on the June 1988 protest in Ürümci, for instance, relied on the testimony of Western tourists who told the journalist that “the banner carried a lengthy inscription in Arabic script which they could not read” and admitted they had no idea of the point of the protest (Lescot 1988). Reports by human rights organizations frequently relied on the personal testimony of former prisoners (who would have had an incentive to play up their suffering for sympathy or to gain political asylum) or Uyghur organizations. Those groups, in turn, had every reason to maximize, even to embellish, the frequency and gravity of conflicts. As I discuss more fully in chapter 5, many leaders of organizations in Central Asia devoted most of their energies to media presentations as a strategy for keeping their movement alive. Few offered clear sources for their information, and some were serial fabulists. Uyghur news organizations in the diaspora such as the ETIC, the Uyghur Information Agency, and the Uyghur-language section of Radio Free Asia (RFA) have produced more plausible reports, but given their close association with political organizations, these cannot be considered disinterested or absolutely reliable.
After jealously guarding information about individual episodes of unrest (as with the protest numbers) in Xinjiang for decades, Chinese authorities began to release descriptions of that unrest in the late 1990s. Remember that these reports were compiled by officials whose job it is to present the party-state in the best possible light and, at the same time, to depict the protests as unsympathetically as possible. Like Uyghur news agents abroad, they have re-presented those episodes to suit their own purposes.26 Particular protests explicitly raised such matters as the dismissal of a Uyghur official without popular consultation, continued nuclear testing, perceived disrespect for Islam and Muslims, and the imposition of family planning. Yet in almost every case, the official representations of those events insisted that they openly challenged party rule, proposed the establishment of an Islamic republic, or aimed at secession. In other words, the state’s versions of events tarred them all with aspirations ruled unacceptable from the beginning. Next I describe a single example (briefly discussed in chapter 2) of a demonstration in Ürümci by two thousand students on December 12, 1985. An eyewitness reported that the students had protested the government’s plan to enforce birth limits on Uyghurs, announced only a short time before, and the continued shipment of criminals from China proper into Xinjiang (Li Yuanqing 1990:71). In a book on the party-state’s struggle against Uyghur separatism published nine years later, the “naïve and excitable students” prove to have been manipulated by splittists and so recede into the background. The reader learns only of splittists shouting “Hans get the hell out of Xinjiang” (Hanren gun chu Xinjiang), “Xinjiang must be independent, must be free, must have sovereignty,” and “Long live independent Xinjiang” (Xu Yuqi 1999:110).27
Chinese writers’ narratives are full of devious plotters, servants of foreign imperialism, and religious extremists, as well as innocent masses hoodwinked into marching or shouting along with these dangerous people. Reports in the early 1990s made elliptical references to protests by dates: “May 19” for violent protests in Ürümci in 1989, “April 5” for the 1990 Baren uprising, and so on. In offering only cryptic references, the writers intended to convey meaning to those in the know and remain mysterious to others. They had reason to fear that providing more information about major episodes of unrest or violence would backfire. Rather than making Uyghurs more supportive of the party-state and its policies, it would make them more hopeful about the possibility of widespread resistance.28 The concern was doubtless to avoid disseminating too widely the news of a considerable number of open protests since the mid-1980s. By the end of the 1990s, the events were given short descriptions and years, but finally they were chronicled in great detail.29 The stories of events have sometimes been subject to several revisions to suit changing political aims. For instance, the 1990 Baren uprising, the 1995 protest in Khotän, and the demonstration in Ghulja in 1997, all blamed for years on “splittists,” were transformed in a 2004 article into the work of “terrorists” (Zhu Jun 2004).
Two questions of particular interest in regard to protests in Xinjiang, whether they were organized and whether they had religious content, are also the two matters about which we must be most circumspect when reading the official accounts. Playing up the role of organization and religiosity in particular events, and possibly inventing those attributes where they do not exist, serves particular political aims. Official scholars describe episodes of unrest as planned and organized in order to challenge the idea that they were “natural” and “spontaneous” and expressed popular dissatisfaction. The same writers also may impute religious content (and attribute religious slogans) to uprisings to make them seem irrational, even radical, and the participants backward.
Ma Dazheng’s study of protests of the previous decade from the vantage point of 1997 identified growing Islamic belief and practice as critical elements in recent events. Not a single protest lacked some religious content, the author found. Officials noted with alarm that more and more citizens were practicing Muslims, including students and party members. They blamed the influence of missionaries from Central Asia. Islamic missionary groups carrying out tabligh,30 or propagation of faith, reportedly operated throughout Xinjiang, using religious instruction as a cloak for spreading subversive political messages about independence and establishing an Islamic state in Xinjiang. In the first ten months of 1999, tabligh groups had reportedly spread from Ghulja in the north to the southern towns of Kashgar, Päyziwat, and Khotän, and officials had rooted out 91 sites and 1,600 practitioners.
Descriptions of protests in Chinese sources imply that if religious slogans were found in many protests in the 1980s, they were ubiquitous in the 1990s (Ma Dazheng 2003:92–105, 118). While some protests were clearly planned in religious settings and raised religious issues, they were not reducible to religious protests; much less can they be regarded as evidence of “Muslim extremism.” Uyghurs have often used religion as a vehicle to express wider grievances or have made the state’s repression of religiosity examples of broader repressions (Becquelin 2000; Dautcher 1999). Viewed from a distance, the Xinjiang government’s multipronged attack on religiosity was clearly intended to eliminate both an alternative source of meaning and a space for organization.
Students of contentious politics are not surprised to find organizations behind mass protests, even seemingly spontaneous ones. Indeed, Rogers Brubaker argues that the key actors in many ethnonational conflicts are “not ethnic groups as such but various kinds of organizations” (Brubaker 2002:172). If we want to find out whether and which organizations orchestrated protests in Xinjiang, we will face the challenge of extracting usable information from carefully constructed official or dissident narratives of those protests. Officials and academics describing many demonstrations or riots report in scandalized tones that they were planned and organized in advance, assertions sometimes seemingly strengthened when transnational Uyghur groups claim responsibility, though of course both have incentives to see efficacious organizations at work.31 One finds outraged accusations of “black hands” and “separatist organizations” behind mass events in both academic studies and reportage potboilers (Liu Hantai and Du Xingfu 2003; Ma Dazheng 1990; Xu Yuqi 1999).32 Such accusations are clearly intended to deny that particular protest episodes were authentic expressions of mass sentiment. For decades, officials broadcast to the citizenry the message that the party alone was allowed to organize people and orchestrate mass demonstrations and that the only legitimate way the masses might express grievances publicly was through purely spontaneous gatherings—which then had to be dispersed by officials and police in order not to disturb public order.33
In China proper, while demonstrations without prior official permission remain illegal and permission is nearly never granted,34 government officials have become somewhat more indulgent of local protests about economic matters. Peasants and workers have had some luck finding sympathetic officials who recognize their claims as “rightful,” thus reducing the chances of harsh repression of demonstrations (O’Brien 1996, 2003). Official treatment of such episodes, however, has varied dramatically by issue and region, as discussed earlier (Hurst 2004; Perry 2001).
By 1997, officials in Xinjiang were alarmed to find organizers drawing participants from across district and even county boundaries and to see demonstrations shifting from remote rural settings to Xinjiang’s major cities: Ghulja in early February 1997 and Ürümci at the end of the month.35 Observers also were disturbed to note that planned actions had grown in scope. Whereas they had previously seen only brief paroxysms of violence, they now faced “armed rebellions.” More and more police actions to round up suspects culminated in gun battles with well-armed holdouts. Politically motivated assassins now combined indiscriminate killing of Hans, intended to cause them to flee, with targeted killings of Uyghur officials loyal to the party-state, dubbed “bridge burning” (chaiqiao). The expression was a pointed barb directed at the official story that Uyghur, Qazaq, and other non-Han officials would serve as bridges between the party and the population. The most hardened partisans had received military training, at first in camps in rural southern Xinjiang and then, after the PSB closed those camps, in Pakistan or Afghanistan. Armed and battle trained, they now spoke openly of armed secession from China.36
The profile of individuals arrested in 1997 challenged a centerpiece of propagandists’ brief against separatists. Instead of the uneducated, unemployed, religious lumpen described in antiseparatist propaganda, the organization members turned out to be young and well educated—and growing more so over time. Suspects apprehended in connection with a spate of arson attacks in late May 1998, reportedly aimed at turning Ürümci into a “sea of fire” and causing Hans to flee, were found to include female students from two of Xinjiang’s top universities, Xinjiang University and the Medical College. Sweeps of suspected members of separatist organizations netted more than three hundred college students from ten postsecondary institutions, hailing from ten different districts.
The earliest reform-era protests appeared to be (even if they were not completely) spontaneous responses to inflammatory events, in much the way that the 1992 Los Angeles riots were touched off by the verdict in the Rodney King trial. Thus when a police officer killed a Uyghur man in PSB custody in April 1990, Uyghurs who caught wind of this stormed the jail, spirited his body away, and within hours staged a demonstration in which three thousand people marched through the streets demanding that Hans leave Xinjiang (Ma Dazheng 2003:47–48; McMillen 1984:575). Similarly, in October 1981 when a Han youth fatally shot a young Uyghur in Kashgar in a dispute over ditch digging, Uyghurs again marched the body through the streets until the crowd of protestors numbered more than six thousand. This time, the protestors reportedly shouted that they would kill Hans and called for a free “Uyghurstan.”37 In these and other cases, while a proximal cause can be identified, the speed, violence, and scope of the popular response point to pent-up anger that had grown over a long period. Widespread popular grievances at the nature of Chinese rule in Xinjiang and the myriad individual complaints of Uyghurs provided the background conditions. The sparking events seemed at once to capture features of the intolerable system in microcosm and to give the final push to tempers at their limits. In sum, while the precipitating events account for the timing of the protests, they cannot by themselves explain those protests.
In the latter half of the 1980s, students and other citizens in Ürümci organized three major demonstrations, each seizing on a recent happening that offended Uyghurs’ sensibilities—the replacement of a popular Uyghur leader, a slur found in a lavatory stall, or the publication of a salacious book—but all raising slogans that responded to matters far beyond the incidents’ provocations. Marchers protested the system of autonomy, nuclear weapons testing, Han migration, family-planning policies, and discrimination against Uyghurs or Muslims, among other matters. Officials worried that in each case, the protests lasted several days, and in the latter two instances they spread to (or had spread from) other cities in Xinjiang or elsewhere in China. There was evidence of coordination of both the content and the timing of demonstrations (for more information, see the appendix).
It was two major protests in the 1990s, however, that caused the most alarm in officialdom. Neither was among the largest protests in the reform era. But the two events’ organization, violence, and ideological challenge to the regime were without precedent in post-1949 Xinjiang. These were the Baren uprising in 1990 and the Ghulja uprising in 1997.
Before daylight on the morning of April 5, 1990, in the month of Ramadan, a group of several hundred men set out angrily from a mosque in southern Xinjiang where they had attended services and spoken publicly of their outrage at the Chinese Communist Party (CCP)’s policies on nuclear tests, the extension of family planning to Uyghurs, and the exploitation of Xinjiang’s resources for use in the interior.38 They marched on and surrounded the government offices in Baren, a rural township in Akto County, thirty miles southwest of Kashgar.39 They chanted the shahada in unison and some called for a jihad.40 Later in the day, a larger group of some three hundred returned to mount an armed assault on Baren party and PSB offices. When several carloads of police came to relieve the officials under siege, the insurgents stripped them of their weapons and killed a number of them, taking others hostage. The attack continued into the night, with the insurgents lobbing homemade bombs and firing on the government offices. The next day, much larger troop reinforcements entered the area and chased the remaining insurgents to the marshlands and mountains where they had fled, killing or capturing all of them by the third day. The official death toll was quite low, listing six police, one cadre, and fifteen or sixteen demonstrators or insurgents killed. International sources proposed a much higher figure of more than sixty killed.
Within days, the government displayed on television the weapons and documents seized from the insurgents, including a booklet laying out the purposes and duties of jihad, among them killing “infidels,” and vaunting the imminent independence of “East Turkestan” (“‘Rebellion’ Quelling Detailed” 1990). Official sources later announced that years earlier, the leader, Zäydin Yusup, had begun recruiting forces for the uprising. He and his co-conspirators had traveled to several mosques stirring up a religious frenzy and secretly building an “Eastern Turkestan Islamic Party.” In each place they had broadcast the message that Islam would soon conquer socialism, that they would drive Hans out of Xinjiang, and that they would found an Eastern Turkestan republic. They also denounced the “colonial” exploitation of the region. Zäydin and others had made extensive preparations, including acquiring weapons and holding four planning meetings, but they were not completely ready to launch the resistance when they learned that the plot had been partially exposed in March 1990, at which point they chose to act in early April.
As soon as security forces had put down the Baren uprising, hard-liners in the government began to crack down on religion much more harshly. This included the questioning of imams, the dismissal of some and the training of the remainder; the closing of new mosques under construction and the halting of repair work on existing mosques; an official policy to find and destroy all private religious schools; and a much broader search for underground political and religious organizations. If the 1980s had provided a brief thaw after decades of anti-minzu policies, Baren ushered in a new era of repression and harsh policies.
A month after the Baren uprising, officials quietly promulgated new regulations governing protest in Xinjiang, superseding temporary ones from 1988. In May 1990 the XUAR People’s Congress passed the new administrative rule, officially termed a “method for implementing” the national law on protest. It stipulated that all marches or demonstrations must be cleared with the government in advance and must not “threaten the unification of the state, harm minzu solidarity, or compromise the interests of state, society, or collective.” The application for official approval must contain “the purpose, method, slogans or catchphrases, participant numbers, vehicles, and sound equipment of the assembly, march, or demonstration” and must identify a person responsible. Participants were forbidden to raise banners or shout slogans “incompatible with the aims” of the event. The rules even stated that security organs could set up security cordons protecting party and military offices, courts, jails, PCC offices, and broadcast stations—in other words, precisely those sites that the protests were likely to target (Xinjiang weiwu’er zizhiqu renda changwei 1990). The new rules would prove advantageous to the handling of several episodes of unrest nearly six hundred miles northeast of Baren, in Xinjiang’s northern city of Ghulja.
Most journalists’ accounts of the 1997 Ghulja protest begin only days before the event, with the sudden arrest of dozens of Uyghur youths in January or the police breaking up a circle of women praying in a private home on what proved to be the eve of the uprising. Chinese versions of the events begin a year earlier with a splittist organization, the “Eastern Turkestan Islamic Party of Allah.” The Ghulja uprising was clearly the product of a chain of events that began much earlier and was symptomatic of both the government’s repressive methods and the Uyghurs’ exasperated responses. As revealed only later in the work of foreign scholars and Amnesty International, this event was distantly connected with government efforts several years earlier to eliminate a popular form of Uyghur social organization (Dautcher 1999:328–29; 2000; Millward 2004:17; Roberts 1998a:686–87).
In 1994 a number of Uyghurs in Ghulja decided to revive a traditional social organization, the mäšräp, in order to combat endemic alcoholism and drug abuse in the region. The mäšräp met regularly, with memberships of several dozen, to share music and dance, learn more about Islam, and hold one another to account for their public behaviors. Leaders of the gatherings had both ritual and religious authority to punish participants in front of their peers for violating the group code. The groups were quite successful at reducing alcohol and drug use and also at giving Uyghurs a sense of collective capacity to help themselves. They multiplied quickly.41 In spring of 1995 the heads of all the mäšräp in Ili gathered and elected as the leader of all the groups one of the founders of the movement, Abdulhelil. He was detained for questioning soon after, and following this the government banned mäšräp, although the organizations continued to operate underground. An anthropologist living in Ghulja during spring 1995 concluded that what the party most feared about the groups was that they were organizations that “it did not initiate, supervise, [or] control” (Dautcher 1999:326).42
In July and August, Abdulhelil and other leaders organized a youth soccer league in Ghulja, and many youngsters joined. On August 12, several days before the tournament was to begin, military officials occupied the playing field, parked several tanks there, and announced that it would henceforth be needed for military exercises. Officials also reportedly removed the goalposts from the fields at all schools in the area to ensure that the tournament could not take place. On August 13 Abdulhelil was again taken in for questioning. The following day, hundreds of men marched peacefully through the streets and then dispersed, an event that officials later referred to as the “August 14 illegal march.” Remarkably, though there was no hint of violent intent in the march, by noon that day snipers stood conspicuously on the roofs of buildings in the center of town, and the People’s Armed Police (PAP) controlled the main intersections with barbed-wire barriers (Amnesty International 1999; Dautcher 1999:325–27; 2004:285–87; Roberts 1998a:686). Abdulhelil and others, angry at the government’s heavy-handed action to squelch a very successful social organization, went on to plan and lead the protest in 1997. Chinese sources claim that Abdulhelil and others joined the Eastern Turkestan Islamic Party of Allah (ETIPA) and that its leader, Päyzulla, had begun infiltrating Ghulja in early 1996, planning for the demonstration in January 1997 (Xu Yuqi 1999:177–78).43 No Chinese source I have seen explains the “August 14 illegal march,” and not one connects the Ghulja demonstration with the crackdown on mäšräp.
There were more proximal causes. A Uyghur organization in the United States asserted that demonstrators, mostly students, were marching to protest the arrests of thirty youths praying in a mosque on January 27, during the month of Ramadan. Yusupbäk (Yusupbek) Mukhlisi, the long-serving head of the Eastern Turkestan United National Revolutionary Front (in Kazakhstan) and, unfortunately, often not a reliable reporter, claimed that the thirty had been not only arrested but also executed (Hutzler 1997).44 Many sources agree that a series of raids on the night of February 4, picking up some two hundred worshippers at mosques and in private religious study groups, immediately preceded the peaceful demonstration beginning at around nine in the morning on February 5. There is little doubt that many marchers had religious motivations for taking part. An official Chinese account of the events has students carrying banners saying “It has begun” and “Use the Qur’an as a weapon” (Xu Yuqi 1999:178). A video shot by the Ghulja police shows the students marching under a white banner with the basmala and shahada handwritten in very large script.45 They marched speedily to the center of town, shouting “religious slogans” and picking up participants along the way until they numbered at least five hundred.46 Some sources suggest that demonstrators symbolized their rejection of the Chinese state’s authority by burning official documents such as identity cards and residency permits and even report implausibly that they “stripped off their ‘Han’ clothing” (another version has them removing all their clothing) as they marched, so as to disavow any connection with Hans (Becker 2001; Jiekai Xinjiang ‘Dong Tu’ fenzi de kongbu miansha 2001).47 About two hours into the demonstration, the police set upon the protestors in full riot gear and with dogs. Official reports asserted that many protestors were armed with bricks and knives and had begun to attack public security personnel and Han citizens as well as property. The police eventually fired live rounds into the crowd to put down the demonstration (Dillon 2004:96–97).
Chinese officials initially denied there were any casualties from the police action. In fact, a police spokesman in Ghulja refused to acknowledge that the protest and crackdown had even occurred, saying, “Nothing happened here last week.”48 This fit poorly with the autonomous regional government’s announcement on the same day that 10 had died and 130 had been arrested.49 Non-Chinese sources reported up to 130 killed that day and up to 500 arrested. Later reports by human rights organizations indicated that the protestors had been hosed down with cold water and then held outdoors in subzero temperatures for hours, with the result that many developed frostbite and had to have their feet or hands amputated. Some protestors returned to the streets on the following two days, again facing riot police and the PAP. There were further arrests, and some Uyghurs reportedly assaulted Hans they found in the street and destroyed cars. The government enacted a curfew and closed the city to outsiders for two weeks. Unconfirmed reports state that independence activists, some of them from as far away as Kashgar, had planned a major demonstration for February 9, the final day of Ramadan. They were betrayed to the police and arrested, and according to one source, they were among the first group to be executed after the demonstrations. Abdulhelil was reportedly tortured and executed secretly months later (Amnesty International 1999; Campion 1997; Hutzler 1997; Tyler 1997).50
Although they were very different events, the Baren and Ghulja uprisings shared certain important features. Both apparently had been planned in advance. In the case of Baren, Zäydin Yusuf, head of the Eastern Turkestan Islamic Party (ETIP), is supposed to have spent the three years from 1987 to 1990 building his organization by inducting members in trips to various mosques. The uprising took place in Baren, but the ETIP reportedly had members in Ürümci, Kashgar, Turpan, and at least ten other major cities in Xinjiang (Zhang Yuxi 1993:349). Investigators reportedly found that the Ghulja uprising had been plotted by the Eastern Turkestan Islamic Party of Allah, an organization founded three years earlier in October 1993.51 Like Zäydin’s ETIP, it had branches and members throughout Xinjiang. Its leaders decided to set the protest in motion at a “Xinjiang-wide congress” of that party on November 27, 1996 (Ma Dazheng 2003:95). Both the Baren and Ghulja uprisings were religiously motivated, and both emphasized public repudiation of the official policy on religion and the party’s claim to be the highest authority. The biggest difference is that the Ghulja protest began peacefully, and according to most reports, it became violent only when police began to crack down.
A number of gun battles might be interpreted as armed rebellions that did not come off; this is how Chinese sources generally represent them. Many started when police tried to apprehend individuals suspected of seeking independence through violent means. According to one source, between 1990 and 2000 Chinese forces reportedly fought 57 gun battles, with 26 police or soldiers killed and 74 wounded, with 140 civilians dying and 371 injured. All told, security personnel fatally shot 106 “rebels.” Much of the bloodletting took place in the latter half of the decade. In a two-month period in 1996, PSB officials engaged in gunfights six times, with one officer killed, while eighteen suspects were killed and another thirteen injured. In the first half of 1999, the PSB had seven more gun battles. In that period, PSB forces lost one, with sixteen injured. Seven “terrorists” were shot dead and eight injured (Ma Dazheng 2003:73, 126–27, 153).
Between 1980 and 1997, the governments in Beijing and Ürümci made concessions in only four instances to matters raised during demonstrations. After a series of protests by Han former “educated youths” desiring to return to their home cities in the interior in the late 1970s and early 1980s, officials granted them the right to periodic home visits, agreed to resettle some individuals, and allowed for all the individuals who remained in Xinjiang to send one child back to China proper. In response to the Muslims’ protests in spring 1989, Beijing halted the publication of the book Sexual Customs, which contained offensive (and wildly inaccurate) descriptions of Muslims’ sexual behavior, and punished both the authors and publisher of the book. In this case, the government responded before the protests spread to Xinjiang, and although the authorities treated demonstrators in China proper quite leniently, they were much less generous with their counterparts in Ürümci (Gladney 1991:3–4; 1992). When 130 uranium mine workers,52 whose radiation sickness had been ignored by authorities for years, traveled to Ürümci and staged a sit-in on May 13, 1989, officials agreed to address their concerns but then scolded them for the form of their protest, saying that a sit-in was “inappropriate” (Zhang Liang, Nathan, and Link 2001:170). It seems evident that had they not protested, their problems would have continued to be ignored. Finally, in 1996 Beijing ended the testing of nuclear weapons at Lop Nur, although this surely was prompted by the hope of wringing arms control concessions from other countries rather than the many Uyghur protests against the practice (Johnston 1996).
In all other documented cases, the government responded to protestors’ demands with either stony silence or even more restrictive policies. When protestors called for greater religious freedom, Ürümci stepped up the repression of religious belief among students and officials, zero tolerance for private religious instruction, and arrests of religious pupils deemed underage or unsuitable (as, for instance, with all children and youths in high school or college or technical schools at equivalent levels). When demonstrators called for increased representation by Uyghur, Qazaq, and other non-Han officials, officials and their advisers pushed for more Han cadres to preserve stability. When Uyghurs repeatedly insisted that Han immigration stop, the government reinstated the PCC and then enacted a series of policies that dramatically increased the inflow of Hans. Officials expressly targeted those regions of Xinjiang where Hans were the scarcest, lavishing great state largesse on the completion of the Kashgar rail link with this aim in mind.53 When students asked for greater respect for Uyghur culture, the government chose to phase out bilingual education and has made a bid to eliminate the use of Uyghur (and Qazaq) as a high-prestige language (Dwyer 2005). And when Uyghurs sought local indigenous remedies to social ills such as alcoholism and drug abuse, the government cracked down on these autonomous social organizations (Congressional-Executive Commission on China 2009; Dautcher 1999, 2004:286–92).
The party-state has relied heavily on a particular strategy for breaking up existing organizations and thwarting the emergence of new ones. Security officials make a point of targeting the leaders of protests for prosecution and heavy sentences as a cautionary example to others. This practice broadcasts the message that potential movement leaders have nothing more to gain than do rank-and file participants and they also have more to lose (Cai 2002:333; see also Tanner 1999:11). Deterring would-be leaders from taking the initiative has so far been widely effective. One researcher found in interviews with disgruntled workers that many were waiting for someone else to organize a protest, with the excuse that once that happened, “I would definitely participate” (Cai 2002:333). This echoes the comments of the many Uyghurs expecting others to take the initiative in 1997, as described in chapter 3.
Officials in China’s inland regions have admitted that they are seeking to convey a “strong signal” to the wider population that “there is nothing to be gained from causing trouble” (Hurst 2004:108). In Xinjiang as in the interior, PSB and other officials have similarly gone after the leaders of movements, ostentatiously singling them out for arrest and harsh punishment while treating most participants in demonstrations leniently. Unlike in the interior, movement leaders in Xinjiang have, on numerous occasions, been publicly executed for the crime of “splittism.”
Chinese scholars have attempted to carry out in their descriptions of protest events what police have done on the ground: isolate the leaders from the putatively guileless and therefore blameless masses. The strategy on paper has been to condemn “a few bad people” (Chen Chao 1990:234; XUAR Local Gazetteer Editorial Committee 1997:77) or people with “ulterior motives” (J K P Š U A R komiteti täšwiqat bölümi, 2000?:49; Xu Yuqi 1999:110–12) for fomenting uprisings. In the case of the October 1981 riot after the shooting of a Uyghur youth, official sources identified the “Central Asian Uyghurstan Youth Sparks Party,” formed only the month before, as the instigator. Three members of the organization supposedly rushed to the scene within half an hour of the shooting and whipped bystanders into a riotous fury (Zhu Peimin, Chen Hong, and Yang Hong 2004:209). The sources’ authors do not try to explain, but instead explain away, the participation of large numbers, asserting that the masses “did not know the true situation” or noting that college students, because of their “ignorance and susceptibility to incitement,” could be induced to march in the streets and shout anti-Han and pro-independence slogans (Xu Yuqi 1999:110–12). In other words, they worked hard to find an explanation for large protests safely distant from the far simpler and more straightforward political diagnosis that only because substantial numbers of Uyghurs are deeply disgruntled are they therefore available for, and willing to participate in, protests at the drop of a piece of fruit, the display of an offending slur, or the description of a scurrilous book. But the study of social movements around the world makes it clear that people participate in them for a great variety of reasons, and that variety does not vitiate their participation or the significance of the movement. Quite clearly, this rhetorical gesture by officials and scholars is a panicked attempt to avoid acknowledging the obvious and pervasive problem of Uyghurs’ anger at the government.54 In fact, if we discount the argument that Uyghurs are somehow more excitable and therefore prone to participate in “troublemaking” without inquiring into its purpose or likely outcome, we are led more strongly to the conclusion that ordinary Uyghurs’ availability for impromptu protests and organized ones alike is a clear index of that anger.
Government officials in Beijing and Ürümci have, with very few exceptions, shown no tolerance for open protests by Uyghurs, whatever the motivation (Hastings 2005). In other words, no matter what the issue, Uyghurs do not have a right to express their discontent openly. A document promulgated by the XUAR party secretary in February 2002, shortly after the arrest of the poet Tursunjan Ämät, showed that officials in Xinjiang “equate any expression of dissatisfaction … even metaphorical or ironical, with separatist thought” (Becquelin 2004a:44).55 In July 2002 Liu Yaohua, vice director of the Xinjiang PSB, told a foreign reporter that “any Uighur who advocated independence for Xinjiang was probably a terrorist” (Pan 2002). In December 2008, administrators squelching a planned protest against the sale of alcohol and cigarettes in shops told the Ürümci college students involved that their demonstration would have been “an act of beating, smashing, and looting … forbidden by our country’s laws.” The event, they said, would have broadcast “reactionary speech” and undermined “stability and unity” (Congressional-Executive Commission on China 2009).
Government regulations and governors’ comments demonstrate how much more restrictive the political climate is in Xinjiang than in China’s interior. The atmosphere in the XUAR has always been more tense precisely because so many Uyghurs resent both the fact and the nature of Chinese control. Despite denials by Ürümci and Beijing, restrictions actually increased over the last decade.
The government has emphasized the message that protest is unacceptable and that any form of public dissent will be regarded as “splittism” and punished severely. There was at least one major political campaign in Xinjiang each year between 1996 and 2004, and every campaign “involved the arrest of hundreds,” often followed by expedited convictions under drastically reduced evidentiary standards. The governing principle of the courtroom proceedings, underscored by Wang Lequan in a 2001 speech, has been the so-called two basics: “As long as the basic truth is clear and …basic evidence is verified,” the legal apparatus is obligated to approve arrest, carry out speedy prosecution, and deliver a sentence (Becquelin 2004a:41; Human Rights Watch 2005:57).
The “strike hard” campaigns begun in Xinjiang in 1996 and repeated every year since have substantially raised the level of repression. At the outset, officials in Beijing worried that this move might trigger international disapproval, but they later found that this was not so. Ma Dazheng noted with pleasure that between 1996 and 1998, the “forcefulness of our ‘strike hard’ [campaign] was massively increased [and yet] there was not a peep from the United States government” and that Western media paid little attention to the matter. Then in 1998, articles “sympathetic to splittist activities” began to appear in the International Herald Tribune.56 Worse, the U.S. State Department began to cover police action in Xinjiang in its annual report on human rights, and Western countries began to use this as a pretext to make trouble for China (Ma Dazheng 2003:208). In the end, Beijing was able not only to repeat the campaigns every year but even to increase their intensity. September 11 provided an excellent opportunity to ratchet up the force of repression yet again. Shielded by international concern about global terrorism, Beijing launched a “high-pressure strike hard” in 2002, a special “100 days’ strike hard” in 2003, and a “high-pressure strike hard” in 2004 with no time limit (Human Rights Watch 2005:67). In interviews with a reporter in 2002, Uyghurs admitted that they feared the police much more than they did terrorists (Pan 2002).57
There is abundant evidence of continuing Uyghur discontent, or the party-state’s fear of it, since 2001. Han Zhubin, once the top prosecutor in China, revealed in mid-2003 that between 1998 and the end of 2002, the government had arrested 3,400 individuals throughout the country for threatening “state security.” Han indicated that there had been a sharp increase in prosecutions since September 11, with 1,600 of those individuals prosecuted after that date. One knowledgeable source calculated that roughly one-quarter of individuals known to have been prosecuted were non-Hans, even though Hans then made up 92 percent of the national population (“A Grim Reminder for the Central Government’s Opponents” 2003). Depending on how comprehensive the former prosecutor’s figures were, the proportion might have been much higher. A paper released by the Ministry of Justice reflected that 9.2 percent of all Uyghurs convicted in 2001 had received sentences for “state security crimes” (Human Rights Watch 2005:72). Statistics culled from various editions of the Xinjiang Yearbook reflect that 2,353 individuals were arrested in Xinjiang alone during the period that Han Zhubin cited, 1998 to 2002 (Xinjiang yilnamisi, 1998 through 2002). In the first eight months of 2004, the government had, according to its own reports, exposed and destroyed twenty-two groups carrying out “separatist and terrorist activities” and handed down fifty death sentences to people convicted of separatist activities (Ruwitch 2004). In August 2004, according to Agence France-Presse, “ethnic and religious tensions [were] flaring up again,” and an official in Khotän told AFP that eight people had been indicted in the last week of July for “endangering state security.” A Uyghur dissident organization reported that seventy-five people, twenty-seven of them children, had been taken into custody in Khotän for “illegal religious activities” (Agence France-Presse 2004a). More recent official statistics count 1,300 people arrested for threatening state security in Xinjiang in the first eleven months of 2008, as against 742 in all of China in 2007, of which roughly half were in Xinjiang. It seems clear that Beijing greatly broadened the definition of a crime threatening state security in the months before the 2008 Summer Olympics. At the same time, the numbers can be read as an index of continuing concern in Beijing and Ürümci about Uyghur discontent.
Officials’ deep fears of unrest can be read as well from moves to shore up the region’s political stability. In March 2005 the Ürümci Evening News reported that police in the region’s capital city had been issued heavier weaponry, including submachine guns, and given training in counterterrorism (Congressional-Executive Commission on China 2005b). A month later, officials announced that of seven hundred new government jobs opening in southern Xinjiang, where Uyghurs are the overwhelming majority of the population, five hundred would be open only to Han Chinese (U.S. Department of State 2006). The government-run Xinjiang Daily newspaper reported in September 2005 that 947 Hans had been dispatched from China proper to take up various government posts (“947 ming yuan jiang ganbu fen fu Tianshan nan bei” 2005). And in November 2008, the Central Military Commission in Beijing promoted the Xinjiang contingent of the People’s Armed Police from deputy to full corps command in order to “safeguard national security and social stability” (Xinhua 2008).
The relative rarity of protest on the periphery since 2001 should not be mistaken for evidence of increasing satisfaction among Uyghurs, Tibetans, and other non-Hans; not even resignation.58 If the hegemony of market and state partly account for the relative quiescence of labor in China proper (Blecher 2002, 2004), we cannot attribute the rarity of protest by Uyghurs, Tibetans, and Mongols to hegemony of the “Chinese nation,” or the state.59 There is too much evidence of everyday resistance, even in periods with little open protest. Xinjiang Party Secretary Wang Lequan’s bitter comment in a private meeting that “our cadres have no place in the hearts of the people” makes this point eloquently (Wang Lequan 1999:17).60 The small and decreasing number of public protests and acts of violent resistance in Xinjiang since 2001 should not be interpreted as a sign that steady economic growth has made Uyghurs as a whole more materially contented and less concerned with politics and thus less inclined to engage in public resistance. Instead, viewed against the backdrop of increasing protests and violence in China proper and evidence of a pervasive wealth gap between Hans and Uyghurs in Xinjiang, the falling protest numbers indicate the success of the party-state’s actions to root out organizations and deter would-be protestors into quiescence—in short, not to resolve Uyghurs’ grievances but to deprive them of the resources and opportunities to articulate them publicly. In fact, instead of addressing Uyghurs’ dissatisfactions, many of the policy instruments used to quell protests actually exacerbated them.
But it would be a mistake to stop with the consideration of domestic effects. Paralleling the domestic crackdown was a regional clampdown on Uyghur individuals and organizations in Central Asia. This reduced or eradicated organizations, sources of weapons, the spread of propaganda, and other sources of support for activities in Xinjiang. More influential still was the dramatic reversal of an international trend toward more frequent humanitarian intervention, indeed, of a seeming revision in the status of state sovereignty, developing in the 1980s and 1990s. Antistate actors who might have won international sympathy and even logistical support, only a year or two earlier now found themselves recast as terrorists. States of all stripes from the most democratic to the brutally authoritarian could now repackage their efforts to squelch challengers as part of the “global war on terror.”