History of Xinjiang: Between China and the World (1990s-2000s)

Eurasian Crossroads: A History of Xinjiang. Editor: James A Millward. New York, NY: Columbia University Press, 2007.

In the summer of 2001, while walking through an underpass below one of Urumchi’s clamorous intersections, I came upon an old Han man kneeling on a scrap of plastic at one end of the tunnel, begging. He was shabbily dressed, as you would expect, in filthy cotton clothes. Frail, bony elbows poked through holes in his jacket. His few teeth were stumpy and yellow. I tossed a small bill into his bowl and walked on. Then, thinking he might have some interesting stories to tell about the old days in Xinjiang, I turned back and started talking with him. ‘Lao xiansheng’, I asked, ’How long have you been in Xinjiang?’

‘Since last month’, he answered. It turned out he came from Hunan province in central China. Recently his son had fallen seriously ill, and although the family had exhausted its meagre savings to buy medicine, the young man had died anyway. With no savings and no child to provide for him, the old man was reduced to begging. ‘But why did you come out here to Xinjiang?’ I asked. ‘Oh!’ he replied brightly, flashing me a gappy smile. ‘We’re developing the great North-west, aren’t we!’

The old beggar was not alone in scenting opportunity on the winds off the Tianshan. According to the most recent census, the population of Han Chinese in Xinjiang increased by 32 per cent, to 7.49 million, between 1990 and 2000. Much of this increase was due to migration, not natural increase (Han in Xinjiang have lower rates of natural increase than Uyghurs, but Han population in Xinjiang grew more rapidly over the decade between 1990 and 2000 than did the Uyghur population). Moreover, something like an additional three quarters of a million people officially registered in other Chinese provinces were residing temporarily in Xinjiang.

The changes of the 1990s and 2000s are dramatically illustrated by two snapshots I took from Hongshan, the rocky outcropping that overlooks Urumchi, in 1990 and 2004 respectively. To the far right side of the earlier photo, the Holiday Inn hotel is visible, still under construction in the spring of 1990. Then one of only a handful of tall buildings in the city, it is by the mid-2000s dwarfed by a forest of skyscrapers. New immigrants swelled the ranks of the construction workers erecting these high-rises, the occupants of the new apartments, employees of the new businesses, restaurateurs who feed them, truckers who supply them, shopkeepers who clothe them, and karaoke girls who entertain them. This same story has been repeated in Xinjiang’s other cities, even those of the south, on a smaller scale. As elsewhere in China, many old neighbourhoods and even the central bazaars of Tarim Basin cities have been swept aside by waves of urban renewal, to be replaced by new shopping centres and blocks of flats. Rural communities too, especially the Bingtuan settlements, have expanded since the 1990s, as labourers from the east come to harvest cotton and other cash crops. The Bingtuan farms now grow a phenomenal number of tomatoes, making the Xinjiang Production Construction Corps responsible for a quarter of the world’s trade in tomato paste.

Xinjiang’s international profile has risen along with the skyline. Once little known outside of China, Xinjiang and the Uyghurs now figure frequently in world press reports, as issues of human rights, religious freedom, terrorism, energy and China’s position in Central Asia bring the region if not into the headlines, then at least onto the internet and inside pages of print media.

A combination of international and domestic developments produced this accelerated change in the 1990s, causing PRC leaders to reassess Xinjiang’s situation and adopt policies to hasten its integration with the rest of China and with the world. The most significant of these developments was the disintegration of the Soviet Union in 1991, which eliminated the massive rival just over the Xinjiang border that had posed a threat to Chinese rule in the region since the nineteenth century. At the same time, continuing Sino-US tension has increased the strategic significance of Xinjiang’s and other Central Asian reserves of oil and gas. The explosive economic growth sparked by China’s experiments with market socialism since the 1980s, combined with the emergence of new Central Asian states, opened new niches for Uyghur entrepreneurs to trade Chinese goods in Xinjiang, exchange currency all across China, and extend operations across the borders to Turkic partners in the new Central Asian states.

Meanwhile, following the relative relaxation of state controls and rise of democracy activism throughout China in the 1980s, there were demonstrations, unrest and even armed resistance by Muslim ethnic groups in Xinjiang, especially Uyghurs. These continued in the 1990s and included large-scale demonstrations and violent incidents that gained international attention. Diasporic Uyghurs in Central Asia, Turkey, Europe and America organised, lobbied politicians, and employed the internet to publicise effectively their grievances to a global audience. As a result, in the twenty-first century the Xinjiang region, though still not the ‘pivot of Asia’ Owen Lattimore predicted, nonetheless looms larger in Chinese, regional and international affairs than it has for centuries.

The Soviet Union dissolved rapidly in August 1991 following the failed coup in Moscow in which Communist generals and politicians attempted to overthrow Mikhail Gorbachev and roll back the reforms he had ushered in. Some observers in the early 1990s predicted a re-emergence of pan-Turkism in Central Asia, or at least a resurgence of nationalism focused on Central Asian identities. Yet the Central Asian republics were the last of the territories to leave the Soviet Union, and did so only when no other option remained to their leaders.

Although the independence of former-Soviet Central Asia was not driven by centrifugal nationalism, the emergence of states named after other Central Asian peoples (Uzbekistan, Kazakstan, Kyrgyzstan, Tajikistan and Turkmenistan) did resound symbolically in Xinjiang, where by the early 1990s many Uyghurs were saying that there should be an independent ‘Uyghuristan’ to match. However, despite predictions of a reprise of pan-Turkic ideology or pan-Islamic ethno-nationalist movements across the region, it was the Chinese state that was best positioned to take advantage of the changed international environment, and it did so quickly. Xinjiang had been relegated to a status of strategic buffer zone and economic cul-de-sac since the rise of Sino-Soviet tensions in the late 1950s and 1960s. In the new international context Chinese leaders moved simultaneously both to open the region as a conduit to the rest of Eurasia and to integrate it more tightly with the rest of China.

Already during the 1980s Xinjiang’s direct foreign trade had increased fifteen times, from $31 million to $439 million, thanks to liberalised foreign trade regulations and the establishment of Xinjiang regional government and Bingtuan foreign trade entities. The upgraded Urumchi airport began to service international flights. Even in northern Xinjiang, where in 1990 the announcements of the few daily flights to Ghulja (Yining) could be readily accommodated on a chalkboard near the airstrip, a new spirit was evident: not far from that same blackboard, an enormous mural greeted arriving passengers with a depiction of people of all races in their native dress happily disembarking from an airliner at the foot of a range of snow-capped peaks.

Most goods, of course, went by land, accommodated by four land ports newly reopened in the late 1980s. The most important of these for international trade was at Horgos in the Yili valley, the truck route between northern Xinjiang and Almaty in Kazakstan (reopened 1983; projected to be a full-fledged free-trade zone by the mid-2000s).The other new border crossings were the Karakoram Highway between southwestern Xinjiang and Pakistan (1987); theTorugart Pass, linking Kashgar to Bishkek in Kyrgyzstan; and the rail line between Urumchi and Almaty via the Alatau Pass (Alashan kou), whose completion in 1990 was grandly hailed as the opening of the Eurasian Continental Rail Bridge. Most of these were old routes for which new road or rail construction had begun decades earlier but stalled for political reasons. Almaty itself began as a Russian outpost astride the China trade route.

After 1991 Xinjiang’s foreign trade with Central Asia expanded rapidly, as China began providing consumer goods no longer supplied by the Soviet economy, and in return importing steel, other metals, oil and other raw materials. The relationship that had characterised Xinjiang’s economic relationship with Russia and the USSR from the mid-nineteenth through the mid-twentieth century was thus now neatly reversed. Deng Xiaoping’s famous ‘southern tour’ early in 1992 announced to the world the Party’s intent to promote economic development beyond the already well-industrialised and commercialised parts of China. In keeping with this, his ‘three along’s policy’ (along rivers, along roads and along frontiers) likewise promised to extend benefits of economic reform to remote and non-Han areas. Accordingly, Xinjiang authorities declared Urumchi, Ghulja, Tacheng and Bole to be ‘frontier open cities’, and Prime Minister Li Peng personally inaugurated the first Urumchi Border and Local Trade Fair. The fair became an annual convention promoting Xinjiang’s foreign commerce.

By 1996 there were over three hundred companies engaged in this border trade, compared with only five in 1991. However, almost a third of these belonged to the massive Bingtuan conglomerate, whose branches now went by a number of names, including the Xinjiang New Construction Company (Xinjiang xinjian gongst). By the end of the decade, half of all Xinjiang’s imports and exports (reported at an overall value of $2.3 billion in 2000) were controlled either by the regional government under the Xinjiang Foreign Trade Group and another Bingtuan avatar, Bingtuan Chalkis. Most of the remainder was handled by various other state or Bingtuan-owned entities. Moreover, unlike in the eastern provinces, where foreign investment financed the rapid export-led growth, in the 1990s and 2000s there was little foreign investment in Xinjiang. Hence Xinjiang’s foreign trade, like its economy in general, has been highly centralised and state managed.

However, there was another side to economic liberalisation and the expansion of Xinjiang’s foreign trade. Not all the trade was in state hands: a great many ‘suitcase’ or shuttle traders also took advantage of the new border crossings and liberalised rules. Uyghurs went abroad or to eastern China to do business, and Pakistanis, Russians and Central Asians came to Xinjiang for the same purpose, as their compatriots had from the eighteenth until the mid-twentieth centuries.

Much of this private trade was small-scale and unreported, relying on personal networks to repatriate profits between inconvertible Central Asian and Chinese currencies. But it was no less lucrative for all that. From manufacturing centres in eastern China, Han and Uyghur merchants brought housewares, clothing items such as track suits, pirated global brands and virtually anything with English printed on it, and sold it in Xinjiang, especially Ghulja, which is close to the Kazakstan border. From there, some Uyghur merchants travelled to Almaty, where they were well represented in the large suburban Barakholka bazaar; others worked in Ghulja’s own wholesale bazaar selling to the Uzbek, Kazak and Russian merchants arriving on short term visas. Many of these Uyghur merchants had relatives who since the mass flight of 1962 or before had been living in Kazakstan, Kyrgyzstan or Uzbekistan; these contacts helped make sojourning in the Central Asian republics easier. This may also be a reason why many Uyghurs in Ghulja readily learned Russian, still the lingua franca of post-Soviet Central Asia. Multi-lingual Uyghurs in Ghulja filled the essential niche of translator and broker between Han sellers and foreign buyers, coming to be known as yangpungchi, a name constructed by adding a Turkic suffix to the Chinese word for the ’sample goods’ (yangpin)  they presented their Russian-speaking clients.

Yangpungchi could make far more money than wage labourers or small urban peddlers. Indeed, trading with the ‘Soviets’ (as Chinese still called all the Central Asian and Russian traders, much as ‘Bukharan’ or ‘Andijani’ were used as catch-all terms in the eighteenth and nineteenth centuries) could be profitable for all concerned. The anthropologist Jay Dautcher, based on fieldwork in Ghulja in the mid-1990s, reports that one Uyghur trader made enough money in a deal selling 4, 500 pairs of socks to support a household for three months. More highly capitalised merchants helped everyone bypass the problem of inconvertible currencies. These ‘bosses’ would bulk the Kazak currency (tengge) earned by many smaller Uyghur merchants selling clothes in Almaty, and use it to buy up large lots of steel and other commodities in Kazakstan. They would then officially import the steel to China, receive payment in Chinese yuan, and repay the small traders while taking substantial profits for themselves.

Though Ghulja was the most important conduit for Xinjiang’s foreign trade in the late 1980s and 1990s, other cities benefited as well. For example, Kashgar was the main focus for Pakistani merchants entering via the Kunjerab pass. Kashgar’s main bazaar, besides providing entertainment for tourists and locals with needed supplies, also became an international marketplace. Thus, besides the ethnic crafts, livestock pens and vegetable sellers at this colourful market, one could find stall after stall selling shoes, clothes, housewares and other consumer goods with a ready market in Pakistan.

By relaxing restrictions on domestic travel and temporary residence, the economic reforms also encouraged Uyghur traders to travel within China. Uyghurs from the countryside could stay in Urumchi or other cities legally if they purchased an inexpensive temporary permit. Others went further afield. Uyghurs sojourned in Beijing, Shanghai, Guangdong and other large cities where they bought goods to ship to markets in Xinjiang or Central Asia; some sold spiced mutton kebabs skewered on bicycle spokes and cooked over charcoal, a tasty street snack that spawned many Han imitations; others joined a large and shadowy network of Uyghur currency traders, allegedly linked to other underworld ventures. These koymochi (from Chinese kuai and mao, literally ‘dollar and dime-ists’) were a common sight in newly commercialised Chinese cityscapes during those years that the government still maintained separate domestic and foreign exchange currencies. Until 1994 average Chinese could not easily obtain dollars or foreign exchange certificates (FEC, waihuijuan); Uyghur koymochi traded in these instruments when the banks would not, and also bought dollars or FEC at a better rate than banks. Sometimes they cut their deals on the front steps of those same banks.

It is fair to say, then, that following the implementation of China’s economic reforms and the opening of cross-border trade with post-Soviet Central Asia many Uyghurs became national and transnational entrepreneurs, albeit most on a relatively small-scale. Indeed, a common joke turned on the ubiquity of Uyghur traders in China and Central Asia: When Neil Armstrong lands on the moon he finds two young Uyghurs already there, eager to sell him a grilled kabob.

One result of the new economy and international context was that Uyghur and other non-Han families engaging in commerce began to join the small ranks of Uyghur professionals, political cadres and intellectuals who had comprised the Uyghur middle class in Xinjiang before the reforms. Centred on Urumchi, but likewise present in Ghulja and other cities, this growing middle class partook of the consumer society evident in any larger Chinese city, developing similar globalised tastes in entertainment and fashion—known here as yengi mädäbiyät, or new culture. They had fewer children, often educated in Chinese-language schools, and generally cultivated a more secular outlook than rural Uyghurs. They valued education as well as commerce and proved increasingly willing and able to pay tuition for higher education in universities or vocational schools. While thus converging in many ways with the growing Han middle class, they maintained a strong sense of distinctly Uyghur identity expressed through enjoyment of Uyghur music and dance, increasingly lavish celebrations of marriages, circumcisions and other life-cycle events, maintenance of Uyghur language, and fascination with Uyghur history.

Development

Xinjiang As a Late Developer

This, of course, is why there has been heightened official concern over development in frontier areas since HuYaobang’s 1980 visit to Tibet and discovery of the abysmal conditions there and in other minority areas (see Chapter 6). The initial state response to the non-Han expressions of dissatisfaction and separatistism in Xinjiang and Tibet in the 1980s and 1990s was, on the one hand, to crack down and tighten controls (discussed below); however, on the other hand the state promised increased development efforts, frequently and publicly predicting that with economic development, ethno-nationalist separatism would disappear. Thus in 1996, following a flurry of demonstrations and a wave of arrests of Uyghur separatists as well as bomb attacks and assassination attempts in Tibet, one top party official, writing in the influential journal Qiushi (Seeking Truth), proclaimed that ‘only a strong economy and improved material and cultural living standards can show the advantages of socialism … and promote the unification of all peoples towards the Communist Party.’ This line was both a pledge to actively develop Xinjiang and other non-Han majority areas and a tacit admission that minority areas remained mired in poverty after decades of PRC rule.

Several years later, in a 2003 white paper, the PRC government could indeed demonstrate its role in creating sustained and rapid growth in Xinjiang, at least on a gross statistical level. From 1950 to 2001 the state invested over 500 billion yuan in fixed assets in Xinjiang, transferred hundreds of thousands of educated personnel into the region, and furnished huge annual budget subsidies. Over the same period (1952-2001) Xinjiang’s GDP increased forty-three times. Per capita GDP in 2001 was almost 8, 000 yuan, compared with 166 yuan in 1952. However, these and similar figures mask the fact that most of this growth occurred after 1978, with a sharp acceleration in the late 1980s and 1990s. Why was Xinjiang, despite its strategic importance, developed so late?

The first Beijing ruler to promote Xinjiang development was in fact neither Jiang Zemin, Deng Xiaoping nor Mao Zedong but Hongli, the Qianlong emperor. As discussed in Chapter 3, in the mid-eighteenth century he used his dragon throne as bully pulpit to argue for developing and settling large parts of the ‘New Dominion’ to the west, rather than leaving it as a buffer under autonomous rule of local elites. At that time he was countering the sentiment in certain Chinese official and literati circles that Xinjiang was too remote, strategically vulnerable and environmentally and culturally unlike China proper to be worth major investment. As we have seen in earlier chapters, that debate re-emerged in various forms in the nineteenth and early twentieth centuries as authorities based in China proper pondered the dilemmas of controlling the region given unrest in Xinjiang, foreign aggression just over Xinjiang’s borders and pressing commitments closer to home.

In the PRC period, although the official position that Xinjiang is an ancient, inalienable part of China has not wavered, actual policy in the face of Xinjiang’s strategic challenges has reflected a less firm commitment. Yitzhak Shicor has written a detailed analysis of the PRC deployment of military forces in Xinjiang. His work reveals that the PRC deployment of troops and matériel in Xinjiang remained modest, despite the conflict with India in Aksai-Chin in 1962 and the still more serious on-going tensions with the Soviet Union. The Soviet Union doubled its forces along the 2,700 kilometre Xinjiang border in the 1960s, intruded many times into Chinese territory in Xinjiang and the north-east, clashed repeatedly with Chinese troops, and supported exile Uyghur forces. Nevertheless, the strength of Chinese force in Xinjiang remained well below what would have been necessary to repel a Soviet invasion, and Chinese leaders were well aware of this. Although the threat of Soviet attack did not wane until the advent of Gorbachev and the warming of Sino-Soviet relations in the 1980s, China’s north-western military command remained based far from the border, in Lanzhou, and planners viewed Xinjiang primarily as ‘strategic depth’ to slow a Soviet assault and stretch out its supply lines, rather than as a piece of the motherland to be held at all costs.

The Chinese planners’ view of Xinjiang as a tactically expendable, if strategically useful, buffer explains the delayed development of the region, and, conversely, its accelerated growth in the 1990s. The region’s late-blooming communications infrastructure is a case in point. The ports of entry, discussed above, remained closed until the late 1980s, despite their historic importance for trade. The rail link from eastern China reached as far as Urumchi by 1962, and a branch line extended to Korla by 1984. But not till 1990 was northern Xinjiang linked to the rail system (with the completion of the line to Alatau Pass), and not till 1999 was Kashgar and the western Tarim Basin linked to the national rail network. Perhaps more significant, the Lanzhou-Urumchi line itself was just a single track till as late as 1994, causing long delays and limiting freight capacity to one third of what is now carried along the double track. The network of improved roads likewise expanded greatly in the 1990s, financed in part by World Bank loans. The most notable new road was the transi-Taklamakan Desert Highway from Luntai to Minfeng which for the first time allowed direct travel by standard motor vehicles from the northern to the southern rim of the Tarim Basin, cutting out the lengthy circumambulation via either eastern or western rims of the desert. (Before the Desert Highway only camels and off-road vehicles could cross the centre of the desert, and that only late in the year along the dry bed of the Khotan River. A sand road, originally intended for retreat to Qinghai or Tibet in the event of a Soviet invasion, ran through the tamarisks and willows along the riverbank from Aqsu south to Khotan).

State Campaigns and the ‘One White, One Black’ Strategy

In the 1990s the Chinese state organised its efforts to develop Xinjiang under the rubric of high-profile campaigns: the first, the 1992 programme to ‘Open up the North-west’ was followed by the March 2000 inauguration of a broader ‘Great Development of the West’ (Xibu da kaifa), which targets Xinjiang and several other western provinces and regions that together comprise 60 per cent of China’s territory, inhabited by 25 per cent of its population. (Interestingly, the terminology and concept echo the ‘Come to the North-west’ [dao xibei lai] programme promoted by the Guomindang in the 1930s, and more generally, Qing efforts in the eighteenth and nineteenth centuries to encourage Han settlement and land reclamation in parts of Xinjiang). The ‘Great Development’ programme is the fulfilment of a promise China’s planners made explicitly at the beginning of the Deng reforms, namely, that the priority given to economic development of the coastal areas in the 1980s would be followed by stepped up investment in poorer inland areas. It also aims to mobilise the considerable natural resources of these regions to more efficient exploitation by the dynamic industries of eastern China. The campaign involves both direct state investments and incentives to encourage investment by Chinese and foreign firms in the underdeveloped regions. These material measures are accompanied by ample rations of boosterism.

Xinjiang’s share of ‘Great Development’ funds comes to over 900 billion yuan, most earmarked for major projects, including a rail-link from Kashgar to Kyrgyzstan, regional highway construction, telecommunications, water conservancy, environmental rehabilitation, agricultural expansion and especially the exploration, extraction, processing and delivery of energy resources.

One of the catchphrases of the Xinjiang development overall has been ‘one white and one black’, referring to cotton and oil, seen as the twin pillars of Xinjiang’s economy. As Xinjiang became self-sufficient in grain in the 1980s a growing proportion of the land was devoted to cotton. In 1978 Xinjiang produced 55, 000 metric tons of cotton; twenty years later the harvest reached 1.5 million tons. Between 1990 and 1997 alone acreage planted in cotton doubled. Many of the water conservancy projects under the development campaigns since the 1990s have related directly or indirectly to cotton production, either by supporting reclamation of land for cotton or, especially in the recent Great Development programme, by attempting to alleviate environmental damage caused by earlier reclamation efforts. Xinjiang became the single largest cotton producer in China in the early 1990s, and by 2001 was responsible for 25 per cent of national cotton cultivation. In 2005 Xinjiang’s 1.75-million ton cotton harvest accounted for 40 per cent of rural income, or as much as 70 per cent in some parts of the region. While cotton’s key role seems assured for the immediate future, the slogan ‘one white’ is no longer entirely accurate, for in 2004 Xinjiang also produced 16 per cent of the world’s total crop of so-called ‘naturally coloured’ cotton—cotton genetically modified to grow in hues other than white. According to unconfirmed rumours, in places where GM and normal cotton are grown too close together, the ‘natural colours’ are occasionally cross-pollinated, resulting in white cotton with red flecks and other unpredictable blends—an apt metaphor for the Chinese economy under market socialism.

Some 40 per cent of Xinjiang’s cotton is grown on Bingtuan farms, and not always profitably. At the end of the twentieth century, during an international cotton glut, Xinjiang cotton remained more expensive than that imported by eastern Chinese mills from abroad. Consequently large stocks piled up. Uyghur and Han farmers alike complained that they were forced to grow the crop at a loss. As during the Qing dynasty, state-requisitioned cotton is the lynchpin of today’s land reclamation and settlement policies in Xinjiang, and serves political as well as economic purposes.

Xinjiang’s oil too is a commodity with a history. Chinese records from the mid-seventh century report a river of crude oil emerging from a mountain north-west of Kucha (Qiuci). Hopeful Kucheans used the smelly ooze as a rejuvenating salve. By late Qing times many surface seeps were known and exploited for lamp-oil by locals. Russian explorers and Xinjiang-based consuls began to investigate these sites from the 1890s. Partly to keep foreigners at bay, and in an early example of the state-private partnership that still characterises Xinjiang (and much of Chinese) industry, the Qing government and private Chinese merchants formed a jointly-operated company to purchase Russian equipment and drill the first mechanised well in Dushanzi in northern Xinjiang in 1909. They struck oil 20 metres down, unleashing an impressive gusher, but the firm folded not long after when the Qing Governor Yuan Dahua pulled the funds and reallocated them to military use. In the late 1920s a Chinese geologist with the Sven Hedin’s Sino-Swedish expedition discovered reserves in the Tarim and Turfan Basins and through the 1930s the British, japanese, Americans (Sinclair Oil) and Germans (Siemens) all sent teams to survey Xinjiang oil possibilities. Extraction began in earnest during the Sheng Shicai era, when Soviet technicians opened the oil fields at Dushanzi, near Wusu, and worked them on-and-off through the 1940s. Soviet exploitation of north Xinjiang oil continued under the new guise of the Sino-Soviet joint stock companies after the PRC takeover, until the Soviets finally sold out in 1954.

China has controlled the continued development of Xinjiang’s oil industry, albeit with some continued investment by foreign oil companies. The main Xinjiang field at Karamay (’black oil’) was opened in 1955, and the city of that name has grown into Xinjiang’s petrochemical refining centre. Although most extraction of oil and natural gas has so far been in northern Xinjiang, Chinese (and, since 1994, foreign) firms have focused their exploration efforts on the Tarim Basin, despite the formidable obstacles to drilling in the Taklamakan Desert. Claims of new finds are published frequently and Chinese estimates of Xinjiang’s petrochemical reserves climb ever higher. In early 2005, for example, after twenty years of exploration, the state-owned oil company Sinopec proclaimed its discovery of a field containing over a billion tons of crude oil and 59 billion metres of natural gas in the Tarim Basin, which if proven true, would increase China’s overall known reserves by one third.

Xinjiang figures prominently in China’s quest for energy security, an increasingly important concern since China became a net energy importer in 1993. With its high rates of economic growth China’s energy imports have accelerated rapidly (oil imports increased by 31 per cent in the 2002—3 period alone). Xinjiang’s production of some 20 million tons annually makes it the fourth largest oil producing region in China, and some expect that it will become the foremost producer during the next century. Xinjiang also produces billions of cubic metres of natural gas per year. Xinjiang is, moreover, the conduit for Kazak oil. Construction began in September 2004 on a pipeline (‘a bridge of friendship between two peoples’) from Atasu in north-western Kazakstan to the Alatau pass (Alashankou) in Xinjiang; completed in May 2006, the pipeline began to replace rail shipments of Kazak oil, with plans for imports of up to 20 million tons annually by 2011. Some foreign analysts considered this pipeline ‘economically dubious’, but both Xinjiang domestic and Central Asian imports of energy are strategically important, given China’s high reliance on Middle Eastern oil. In 2004 some two thirds of China’s oil imports, those from the Middle East and Africa, had to pass through the Straits of Malacca, where they were vulnerable to piracy, terrorism or US blockade. Chinese leaders thus urgently want to develop more secure continental sources of energy.

The same political calculus evidently applied to the flagship project of the ‘Great Development of the West’ campaign. In September 2004 the 4,200 kilometre West-to-East Pipeline began delivering natural gas from Xinjiang to Shanghai. Much touted in official ‘Great Development’ literature, and at one time China’s single largest foreign-invested project, the 140 billion yuan pipeline was considered too expensive by the energy companies who initially planned to invest in it. Exxon-Mobil, Royal Dutch Shell, Russia’s Gazprom and even the Chinese firm Sinopec withdrew from the planned partnership at the last minute, leaving as sole owner PetroChina (majority owned by the Chinese government). The problem was again its economic rationale—or lack thereof. The natural gas shipped from Xinjiang, expected to reach 12 billion cubic metres per year by 2007, cost 45 per cent more than international imports, making natural-gas fuelled power plants in Shanghai lose money. One Chinese analyst criticised the pipeline as ‘a commercial project with characteristics of the command economy’, aimed primarily at stimulating growth in the far west, without consideration for cost.

This same critique could be applied more generally to the state-led efforts to develop Xinjiang. Several years into the ‘Great Development of the West’ campaign Xinjiang was only just beginning to attract sizeable investment from other Chinese provinces and regions: over 26 billion yuan in 2004—but this was a jump of 42 per cent over the previous year. Foreign direct investment in Xinjiang remained low into the 2000s (only 358 million yuan in 2002). In fact Xinjiang’s economy was much more tightly linked to the central government than was that of other provinces or regions of China. At the turn of the millennium a high percentage of its industry (approximately 80 per cent in 2004) remained state owned; a similarly high proportion of investment in capital construction (almost 60 per cent in 2000) came from the central government. Despite oil and gas revenues (which make up half of the regions’ fiscal revenues), Xinjiang runs huge annual deficits, its expenditures routinely exceeding its GDP by between 12 and 19 billion yuan. These deficits are covered by price subsidies, poverty alleviation plans, special allocations to Xinjiang’s state-owned enterprises and other mechanisms as well as by an annual fiscal subsidy from the central government to the Xinjiang region. Calla Wiemer has calculated the gap between GDP and expenditures—the net resource inflow into Xinjiang—for 1981-2000 as averaging over 20 billion yuan annually, or around 20 per cent of Xinjiang’s GDP. While the net inflow did appear to decline towards the end of the period, her calculation does not include the hundreds of billions of yuan recently invested in the region for the ‘Great Development’ campaign. Although Xinjiang is one of the richest of the inland provinces and regions, and its growth rate in recent years has been exceeded only by the coastal provinces, the region’s effective subsidy remains on a par with those of Yunnan and Guizhou, the poorest parts of China.

All of this demonstrates, once again, an uncanny similarity to the state of affairs in the eighteenth and nineteenth centuries, when the Qing imperial government shipped annual subsidies of millions of silver ounces westward to sustain its government, military and state farm operations in Xinjiang (see Chapter 3). Xinjiang’s energy industry and the prominent economic role of the highly unprofitable Bingtuan partly explain the centralised and subsidised nature of the region’s economy But why Xinjiang remains so commandist despite the phenomenal success of privatisation and liberalisation elsewhere in China ultimately comes down to political decisions by China’s leadership, which has consistently chosen centralised control and state-driven development over economic efficiency in this frontier region.

This section on development began with the question of how the party has improved the people’s lives in an ‘ethnic concentration region’, as Xinjiang is now commonly called. In fact the 2003 white paper on Xinjiang mobilises battalions of statistics to make the case that the PRC has measurably improved the lives of people in the region. One such measure involves consumption. In rural areas, according to official figures, people in 2001 devoted 10 per cent less of their income to food than they did in 1978. Although they still spent half of their money to feed themselves, on average they enjoyed enough discretionary income for every household to have a bicycle—alas, unlike Sayram Bek, not yet a motorcycle. Almost every rural home likewise had a television, and washing machines and tape recorders graced some village homes. Xinjiang’s city dwellers were on average nearly as blessed with the sine qua nons of modern life (colour TVs, refrigerators, hi-fis, VCD players, mobile phones) as their counterparts in Beijing, or for that matter in London or Washington. In all categories, rates of ownership of these consumer durables have increased many-fold since 1985.

However, other measures of development success are possible. In his talk with the Kazak herdsmen, General Secretary Jiang evoked ‘unity of the nationalities’ (minzu tuanjie), the common slogan and coded reference to separatist sentiments and ethnic unrest in the region. In his words, and in those of the official party line in the final decades of the twentieth century, prosperity in Xinjiang rested upon a foundation of ethnic harmony. But as noted above, the party in those same years also explicitly pegged its hopes for improved ethnic harmony in Xinjiang on its economic development, describing the latter as foundation for the former. By this measure it remains unclear to what extent Xinjiang development had succeeded by the end of the millennium.

According to an attitude survey conducted in 2000 by a scholar from Hong Kong University, some 47 per cent of Uyghurs sampled in Urumchi believed their standard of living under the reforms had risen as fast as that of the Han, and 38 per cent thought it had increased more slowly. (The perception among Han was almost a mirror image, with half thinking they were on a par with Uyghurs and 38 per cent believing Uyghurs had advanced more quickly). This study no doubt reflects the creation of a relatively well-off Uyghur urban middle class.

However, scrutiny of the official figures themselves reveals a different region-wide pattern. Although the PRC does not break down statistics on income in Xinjiang by ethnic group, Wiemer regressed county-level figures on GDP per capita and ethnic make-up to discover a strong correlation between ethnicity and income: the greater the Turkic population of a given area, the lower the GDP per capita for that area. Or, as Nicolas Becquelin puts it, southern Xinjiang, with 95 per cent non-Han population, has an average per capita income half that of Xinjiang as a whole.

Han Chinese are concentrated in cities of northern and north-eastern Xinjiang, including Urumchi; they work disproportionately in higher-paying professional and government jobs. Uyghurs in Xinjiang told foreign visitors that they felt shut out of new jobs created by large-scale projects, especially those in the energy sector. Uyghurs also commonly complain that they have no chance to work at the construction sites now ubiquitous even in southern Xinjiang cities. My own observations from the summers of 2001 and 2004 confirm that the labourers erecting new buildings around Id Kah mosque, paving the road through the Uyghur neighbourhoods in old Kashgar, demolishing the old bazaar in Khotan and working on a major renovation of Khotan’s central downtown intersection were all Han, as they were even in remote Tashkurghan (on the road to Pakistan). Those workers I talked to, moreover, said they came from provinces east of Xinjiang. Wang Lequan, Xinjiang First Party Secretary and Politburo member, expressed a commonly heard explanation for this when he told foreign journalists that ‘many Xinjiang minorities don’t want the “dirty, hard and tiring work”.’ The comments of a labourer from Sichuan to a Western reporter suggests a more nuanced reason:

‘We cannot communicate with them. We don’t know their language’, said Zhang Bizhong, a construction worker from Sichuan province. ‘We Sichuanese prefer to eat rice and pork, and dishes with strong flavours. The local minorities don’t eat pork. They don’t like rice.’

Most Uyghurs actually do like rice: polu, rice pilaf cooked with mutton, carrots and onions, is a favourite celebratory meal across Central Asia. But if Mr Zhang could not communicate with local Uyghurs enough to realise this, neither could the labour bosses who hired him. One suspects contractors have deemed it simpler to bring in migrant workers from their own provinces rather than work with local society. This is of course a common problem with development projects in many places: Washington, DC has passed laws forcing contractors from the neighbouring states of Virginia and Maryland to hire local labour if they wish to take on large municipal projects.

The Peacock Flies West’: Migration to Xinjiang

Han migration to Xinjiang is not new. Based on our best available figures and estimates, at the start of the nineteenth century, forty years after the Qing conquest, there were something like 155, 000 Han and Hui in Xinjiang, and something over twice that number of Uyghurs. Han population dropped precipitously in the mid-nineteenth-century wars, and climbed gradually thereafter until the mid-twentieth century. In 1947, on the eve of the PRC takeover, there were some 222, 000 Han in Xinjiang, comprising around 5 per cent of the population, compared to over three million Uyghurs, who made up 75 per cent. In 2000 XinJiang’s 7.49 million Han represented 40.6 per cent of Xinjiang’s total population of 18.5 million, or 43 per cent if the officially acknowledged ‘floating population’ of 790, 000 non-registered migrants is included in the Han total. Although the Uyghur population (7.19 million in 1990, about 8.35 million in 2000) has also increased markedly since 1949, Han population has grown much faster and by 2000 equalled that of Uyghurs region-wide when the floating population is taken into account, and exceeded it if military personnel stationed in Xinjiang are included. Figure 1 represents population changes in Xinjiang from 1947 to 2000.

Although the Urumchi environs—northern cities like Tacheng and even Korla and Aqsu—has had large Han populations for some time, one of the most obvious changes in Xinjiang in the 1990s-2000s was the increased numbers of Han Chinese in southern Xinjiang towns. This was evident not only in the faces of the people one encountered, but in the changing cityscapes themselves. For example, after the rail-line reached Kashgar in 1999, neighbourhoods of shiny, new, white-tiled buildings inhabited primarily by Han sprung up between the old town and the train station 11 kilometres away, much like settlements of Han merchants once filled the space between the old town and the new forts constructed by the Qing military in the eighteenth and nineteenth centuries. A night market on Youmulake Xiehaier (Yumulaq Shähär) Road near the intersection with Renmin Road, on the border of the old and newer parts of town, reflected the same reality noted by the construction worker Zhang Bizhong: in 2001 this market was ethnically segregated, with Han carts on one side selling liquor, pork dishes and such snacks as chicken feet; a few feet away on the other side Uyghur carts sold mutton kebabs, noodles and polu.

The urban changes proceeding at such a startling pace in southern Xinjiang in the 1990s and 2000s—the demolition of old housing and the growth of apartment blocks, the remodelling of city centres, the commercialisation and privatisation of former public spaces, the replacement of old-style covered bazaars with pedestrian malls and shopping centres—had as much to do with Chinese and global patterns of modernisation as with Han migration. Certainly, many criticised how these projects were conceived and implemented. As happens everywhere, developers often entered into cosy deals with city leaders and the desires of local residents and shopkeepers were ignored. But it is the nature of cities to change.

Still, by the early 2000s expatriate Uyghur dissidents, foreign observers and at least some Uyghurs in Xinjiang came to suspect that one goal of the Great Development of the West campaign was precisely to promote Han migration to Xinjiang. Nicolas Becquelin argues that by the late 1990s PRC leaders had determined that economic development alone could not eliminate ethnic dissent and separatist tensions in Xinjiang. Instead, it opted for a policy of accelerated integration of the region by means of Han migration in order to increase security. (As discussed in Chapter 3, the Qing dynasty revised its policy to allow Han settlement in southern Xinjiang for precisely the same reason in 1831). Becquelin points to a 2000 article by Li Dezhu, head of the State Ethnic Affairs Commission, which links Han migration to Xinjiang (‘The Peacock Flying West’; kongque xibufei) with the Great Development of the West campaign. In the article Li suggested that directing more policy attention and financial investment to the west would induce a westward flow of population to match the south-eastward flow to the coast that had occurred after the onset of Deng’s economic reforms. While population movement is a natural accompaniment to economic development, Li argued, it would result in a decline in the relative population of non-Han groups in western cities and thus could cause ethnic ‘contradictions’ and ‘friction’. Nonetheless, western development would help turn the entire country into ‘one big unified market’ and that, in turn, would increase the ‘centripetal force and cohesive force’ (xiangxinli he ningjuli) of each nationality towards the greater ‘Chinese’ nation (zhonghua minzu).

From the 1950s through the 1970s the PRC officially resettled millions of Han Chinese from eastern provinces to Xinjiang using the Bingtuan to accommodate and employ them. Although many of the young Shanghamese and others sent to Xinjiang during the Cultural Revolution eventually returned home, such forced population transfers were the main reason for the rapid growth of Han population in Xinjiang. In the 1990s—2000s the state no longer directly mandated the resettlement of large numbers of people to Xinjiang, nor could it effectively do so: In 1992 the China Daily reported a plan to relocate to Kashgar up to 470, 000 people destined to be flooded out of their villages and towns by the Yangzi Three Gorges Dam, then just beginning construction in Central China. After demonstrations by effected villagers, complaints by Bingtuan and Xinjiang officials and an international outcry Beijing officials dropped the idea. Nevertheless, official encouragement and recruitment—via, for example, public offerings of southern Xinjiang land to Han migrants willing to take out responsibility contracts—continued. During his 1990 tour of Xinjiang Jiang Zemin offered a wordplay on Mao Zedongs famous line about the Great Wall. Jiang’s new version of the epigram became ‘Bu dao Xinjiang, bu hao Hanf’ (if you haven’t been to Xinjiang, you’re not a good Han!). Hao Han has a general meaning akin to ‘a real mensch’ but in the ethnically politicised environment of Xinjiang, it takes on a more literal sense. As the optimism of our Urumchi beggar shows, even without direct state incentives Xinjiang’s boom-town atmosphere since 1991 and the call to ‘Develop the West’ proved a strong draw to hundreds of thousands of ‘hao Hans’ from the east.

‘A Dry and Semi-Dry Region’: Xinjiang’s Environment

Xinjiang has its share of baleful environmental stories. Radiation lingers from the nuclear testing grounds of the Lop Nor area, producing, some have claimed, clusters of birth defects. The Tarim Basin’s near monoculture of cotton renders Xinjiang’s agrarian economy vulnerable to pests. Extinction threatens the region’s indigenous species, including the wild Bactrian camel, the Yili pika, the Tibetan antelope and the snow leopard. Fires raging deep underground in over thirty coal fields have so far consumed over three billion tons of coal (the recently extinguished Liuhuanggou fire had been burning for 130 years). One third of China’s underground coal mine fires, which together reportedly emit a volume of carbon dioxide equivalent to that produced by all the automobiles in the United States in a year, are in Xinjiang. However, the crux of Xinjiang’s human and natural ecology, as in Loulan’s heyday, turns on the relationship between numbers of people and supplies of water.

If continued Han in-migration has become a prominent feature of Xinjiang’s post-1991 history, it is not the only source of population growth. Numbers of ’minority nationalities’ also increased by 1.5 million between 1990 and 2000. In fact, from a broader perspective, the story of Xinjiang’s demography throughout the modern period is one of faster population growth than for China as a whole. From the time of the Qing conquest until the mid-nineteenth century, Xinjiang’s population increased six fold—compared to an increase of four and a half times for China over the whole Qing dynastic period (1644-1911). Likewise, during 1990—2000 Xinjiang’s annual growth rate was 1.67 per cent versus 1.07 per cent for China overall, and in 1999 Xinjiang’s natural growth rate (crude birth rate minus crude death rate) was 1.28 per cent, compared with 0.88 per cent for China as a whole. A projection made in 2000 predicted that Xinjiang’s population would continue these high rates of growth, reaching 23.47 million by the year 2015. In the early twenty-first century, then, Xinjiang continued to look demographically like a frontier region, filling up with people due both to in-migration and high natural growth rates among Uyghurs and other non-Han groups.

Because of Xinjiang’s extreme aridity human life in the region has always been sensitive to the natural environment, and vice versa. The fact that the inhabitants of northern Xinjiang were until modern times primarily nomadic pastoralists is a reminder of how geography has shaped socio-economic patterns in Central Asia. The ancient cities now buried in the Taklamakan similarly testify to the key role of the environment and especially water in supporting human settlement in the Tarim Basin.

Against this background, the population growth and land reclamation that accompanied Xinjiang’s development in the second half of the twentieth century is, on the one hand, broadly understandable as a continuation of earlier patterns of settlement and environmental exploitation, particularly since the eighteenth century On the other hand, the past fifty years stand out for the sheer scale of human modifications to the environment.

Water, and its exploitation, is the key to modern human-environmental interaction in Xinjiang. Immediately after the conquest of Xinjiang, Qing officials eager to resolve the formidable logistical difficulties of the occupation set about improving water supplies for roads throughout the region and widely expanding irrigation to support increased agriculture, the harvest from which provided grain to the Qing troops. Till the early nineteenth century state farms and associated hydrological projects were concentrated in northern and eastern parts of the region, but from the 1830s imperial authorities turned more attention to the south as well. During his three years of exile in Xinjiang in the early 1840s, Lin Zexu alone charted out 800, 000 mu of new irrigated land, two thirds of it in the Tarim Basin. He supervised the digging of everything from karez wells to major canals. In fact, the impact of ex-Commissioner Lin’s hydrological works in Xinjiang are arguably as profound as his famous destruction of British opium, for he inaugurated the intensive agricultural exploitation of the Tarim Basin, and his surveys formed the template for later agricultural reclamation. Scholars estimate that from the Qing conquest to the late nineteenth century the area under cultivation in Xinjiang increased over ten-fold, to 11, 480, 190 mu (approximately 766, 000 hectares) in 1887.

In his plans to reconstruct the region after provincehood, Zuo Zong-tang was greatly influenced by the apparent success of this earlier reclamation programme, as he was by a famous essay written by the statecraft scholar Gong Zizhen, which stated confidently that investment in agriculture in Xinjiang and increased settlement by Han Chinese would make the region self-sufficient and stable (discussed in Chapter 4). Nonetheless, the lands abandoned in the mid-nineteenth-century rebellions proved hard to recover from the sands; moreover, neither Zuo nor successive Han governors in Xinjiang through the early twentieth century had the funds for a major investment. By the 1910s cultivated land was just over 700, 000 hectares, not yet back to its peak level during the Qing. Nevertheless, Gong Zizhen’s essay and the Qing record in Xinjiang suggested powerfully to later officials and farmers that popular effort guided by official planning could make even arid Xinjiang bloom and support large numbers of new settlers. By the 1930s cultivated area had reached some 995, 000 hectares and was at 1 to 1.2 million hectares in 1949.

After that, the Bingtuan, collective farms and more recently individuals contracting land followed the same philosophy, which easily dovetailed with the Maoist vision of nature as an enemy to be conquered by the party and the masses in the drive to build socialism—or, after Deng Xiaoping, the drive to create prosperity. Despite the excesses and environmental damage of the Great Leap Forward, the Bingtuan continued to settle new migrants and open new land. The reform policies since the late 1970s too encouraged expansion of arable area. The result was a total of 3, 404, 120 hectares under cultivation in 2001—an area the size of the Netherlands.

This rapid expansion of farmland occurred in a region with-minimal precipitation. Thus it depended almost exclusively upon channelling irrigation water from the streams flowing out of the mountains ringing Xinjiang’s basins. Rivers fed by run-off from snowpack and glaciers support settlement and grain fields in the fertile Yili valley and other parts of the north, as well as the Urumchi area, the Turfan Basin and the cotton-growing oases of southern Xinjiang.

However, the most acute environmental changes of the past half century have occurred in the Tarim Basin. After passing through and being tapped in the oases, run-off water in the Tarim system historically either emptied straight into the desert or joined the extensive river system consisting of the Khotan, Yarkand, Aqsu and other rivers whose confluence formed the 1,300 kilometre-long Tarim. Fed by these and smaller tributaries, the Tarim meanders from west to east along the northern rim of the basin. It has always been prone to course changes, but no water escaped the vast depression that is the Taklamakan, and the river always terminated in one or another of three lake basins in the south-east corner of the desert: Lop Nor, Kara-Koshun or Taitema. Most of the run-off water joining the Tarim system flows seasonally, with between 60 per cent and 95 per cent of the annual discharge flowing between June and September.

Despite vast infusions of new capital, labour and modern machinery, there was no way agriculture in the Tarim Basin in the second half of the twentieth century could transcend this basic hydrological framework. Agriculture enjoyed no new sources of water; even groundwater sources are linked to run-off discharges which annually replenish the oasis aquifers. The growth in population and expansion of cultivated land area, then, has simply drawn off more of the water flowing through the same Tarim system than had been done in the past. Reclamation engineers and farmers, many with the Bingtuan, were able to intensify exploitation of the Tarim system by constructing hundreds of dikes, hundreds of kilometres of canals and karez and hundreds of deep tube wells in oases and all along the reaches of the rivers. Ultimately this intensified agricultural and urban use, along with evaporation from uncovered irrigation channels and flooded fields, came to exhaust most of the water before it reached the lower reaches of the Tarim. Lop Nor, once a wetland of reed forests home to fish, birds, wild pigs and tigers, was reduced by 1964 to a thickly crusted salt flat. Lake Taitema, the terminal lake of the combined Tarim and Kongque rivers from 1952 to 1972, likewise dried up in 1972 when a dam was built on the lower Tarim to create the Daxihaizi reservoir.

Besides desiccation of the terminal lakes, intensified upstream use of the Tarim Basin’s rivers produced a range of unanticipated consequences. By the 1970s so much water was diverted upstream that little or none reached the lower 300 kilometres of the Tarim’s bed. This lowered the water table across this vast area from 3-5 m in the 1950s to 11-13 m below the surface in the 1980s. Not only did this undermine farming in the area’s state farms but it shrunk grass coverage by 75 per cent and reduced the poplar forest from 54, 000 to 13, 000 hectares. What was once called a ‘green corridor’ in the Taklamakan became largely desert as two thirds of its former farmland filled up with sand.

Population growth and increased land and water use had similar effects along each of the rivers and in each oasis of the Tarim Basin. Diversion and use of water upstream reduced water available to farmers downstream, lowered water tables and shrunk open bodies of water. Moreover, because of high evaporation rates and the common practice of flooding fields to irrigate them, mineral concentrations in downstream water increased markedly. This increased salinification of fields at the outskirts of oases, leading to degradation and abandonment of land. Abandoned land, with neither crops nor original natural vegetation to stabilise it, was easily eroded and covered by wind-blown desert sands. A study of the Cele oasis based on remote sensing data reveals that from the 1970s to the 1990s 8 to 9 per cent of good farmland along the outskirts of the oasis gave way to mixed bands of desert or moderately arable land. The area of bodies of water in the oasis shrunk by over half. This happened even as tree-plantings successfully maintained the overall area of the oasis against an advancing sand dune. At Cele and elsewhere desertification on the peripheries of the oases was exacerbated by the cutting of trees and scrub for use as lumber and fuel by oasis dwellers. Particularly damaging was the extirpation of tamarisk roots, which eliminated a tenacious stabiliser of the desert surface. The availability of tractors to pull out the massive root clusters and carry them into the oases made this practice possible and economical in a way it had never been before modernisation.

Human changes to the environment were not restricted to southern Xinjiang. To the east increasing numbers of Turfans ancient karez canals ran dry over the latter half of the twentieth century, with only 617 out of 1, 784 still functioning in 2004; these remaining karez were granted official protected status in 2005, but continued lowering of the water table caused by mechanised deep wells throughout the Turfan Basin was projected to dry out the last karez by 2030. In the north the drive to boost Xinjiang’s grain harvest led to a rapid conversion of pastures to farmland. In Bortala Prefecture, in the north-west corner of Xinjiang, from 1950 to 1977 the area of Ebinor (Lake Aibi) shrunk by half as the population in its drainage area shot up from 67, 800 to 550, 500 and farmed area and water use expanded accordingly. Ebinor’s area continues to shrink by 23 square kilometres annually Loss of vegetative cover around the lake led to wind erosion and dust storms: in the 1960s this area experienced high levels of airborne dust only one day in every couple of years. In the 1990s there were a month and half of such days annually

Encroachment of farmland throughout the region, especially in northern Xinjiang, led to a loss of some 240, 000 hectares of rangeland from the 1960s to the end of the century. However, from 1949 to 1998 the livestock population quadrupled from 10.4 million to 42.2 million head. Overgrazing, upstream water diversion and other factors have led an estimated 75 per cent of the available pasture to decline in productivity; the thinner cover of grass can now support fewer sheep per unit of land. Some 1, 000 square kilometres of grassland desertified.

Xinjiang’s forests likewise retreated, due both to excessive cutting and falling water tables. Besides the 84 per cent reduction of the poplar forest in the lower Tarim drainage, willow (hongliu) scrub on the fringes of the Taklamakan has been reduced by 65 to 90 per cent. The slopes of the eastern Tianshan have been virtually denuded, and treeline elsewhere in the range, at an altitude of between 1, 200-1, 400m in the 1950s, has receded to above 1,700 metres. Nearly 70 per cent of the sacsaoul trees of the Zungharian Basin were cut between 1958 and 1982, and the poplar forests of the southern rim of the basin are gone. Reduced tree cover contributes to flooding, erosion and lower atmospheric humidity.

In the war with the desert, the desert is winning of Xinjiang’s eighty-seven counties and municipal districts, fifty-three have suffered desertification. Annual economic losses from desertification were estimated in 2001 at 2.5 billion yuan. Of the 33, 317 square kilometres reclaimed through state and private efforts from the 1960s to the end of the century, one fifth (an area larger than the state of Delaware) has been abandoned again due to soil exhaustion, loss of water supply, salinification or encroachment by desert sands. Despite a massive programme of tree-planting intended to create greenbelts and windbreaks, in 2001 the Taklamakan desert was estimated to be expanding by 172 square kilometres per year; in 2004 the head of Xinjiang’s Forestry Ministry announced that the desert was growing by 400 square kilometres annually Over 47 per cent of Xinjiang is defined as ‘wasteland’.

Moreover, it appears that this human-assisted desiccation took place during a period when levels of water entering the Tarim system and other Xinjiang watersheds were actually on the rise. Average annual temperatures in this region, as elsewhere, have risen to unprecedented levels over the course of the twentieth century. Evidence of global warming is clear in oxygen isotope studies of Kunlun Mountain glacial ice cores. Higher average annual temperatures reduce overall precipitation in Xinjiang even while increasing it in the rest of China (Xinjiang lies outside the monsoonal system that determines eastern China’s weather). But unprecedented warming seems to have speeded the melting of the snowcap and glaciers in the Tianshan and Kunlun ranges over the past few decades, so that the average annual Tarim river discharge was some 10 per cent higher in 1976-86 than it was in the decade 1954-64. Xinjiang’s twentieth-century development has, therefore, drawn upon water reserves banked in the past: the most recent major deposits occurred during 1400—1900, when temperatures were markedly cooler than today, and especially during the global and local climatic minimums of that era, from 1420 to 1520, 1570 to 1690, and 1770 to 1890, when precipitation in the mountains peaked.

Recent studies suggest that the benefit from increased glacial run-off cannot continue for long. An international team led by China’s preeminent glaciologist, Yao Tandong, concluded in 2004 that the glaciers of the high Tibetan massif, which includes the Kunlun range, are melting at a rate ‘equivalent to all the water in the Yellow River … every year.’ Those in the Tianshan are likewise imperilled. In fact China’s glaciers will be two-thirds gone by 2050, and may disappear entirely by the end of the twenty-first century. Since the major tributaries that form the Tarim derive between 41 per cent and 58 per cent of their volume from glacial melt-water, Xinjiang could face a major reduction of its water supplies in a matter of decades. On the other hand, in 2003 the Chinese Geological Survey Bureau announced the discovery of a massive underground reservoir beneath the Taklamakan Desert. Should their initial survey of its size bear out—they estimated its capacity at 36 billion cubic metres, the size of the lake behind the Yangzi Three Gorges Dam—and the obstacles to extracting this deep-lying water be overcome, this aquifer could help alleviate Xinjiang’s water crisis. But as fossil groundwater, its exploitation too is ultimately limited.

One sideeffect of the ecological changes in Xinjiang over the past half-century has been more qara buran, or sandstorms, blackening the sky, damaging property and crops and harming livestock. There were 105 major ‘wind disasters’ in the region in the 1980s, compared to only sixteen in the 1950s. Desiccation and desertification across north China has produced similar phenomena, and Beijing itself suffered debilitating sandstorms in the spring of 2000. The clouds of red dust enshrouding the capital spurred China’s leadership to greater efforts at environmental protection and rehabilitation. Indeed, one of the six publicly promulgated tenets of the Great Development of the West campaign was ‘ecological and environmental protection and building’. Although China’s environmental protection agency, SEPA, was not included among the twenty-two government agencies leading the campaign, the Great Development plan included measures to continue planting shelter-belt forests, to strengthen protections against logging, and to return farmland to forests and pastures. However, most salient among the environmental programmes in the Great Development and concurrent five-year plan was a major infrastructural project to restore the Tarim River watershed.

Chinese leaders have always displayed a predilection for monumental hydrological works. Throughout history rulers and ministers in China have concerned themselves with water control. One inspiration for this is Yu the Great, a legendary emperor who ‘regulated the waters’ and founded the Xia Dynasty (2205—1766 BCE), ushering in a period of luxuriant fertility and popular prosperity. Modern leaders, hoping to serve the people and leave their mark on Chinese history, still feel the pull of tradition; they have also been swayed by the high modernist impulses of Communism and grandiose tendencies of Soviet planners. SunYat-sen planned but never succeeded in damming the Yangzi. Mao Zedong, fascinated by the old saying that ‘when a great leader emerges, the Yellow River will run clear’, approved the construction of the enormous Sanmenxia dam on its main channel, designed—disastrously—to prevent the Yellow’s river’s copious sediment from flowing downstream. More recently Premier Li Peng pushed through plans for the enormous Three Gorges Dam on the Yangzi in the face of domestic opposition and international concerns over its economic feasibility and environmental and social impact.

it is not surprising, then, that as PRC leaders came to understand the severity of water scarcity in north-west China, the first solutions proposed in the 1990s were massive water transfer schemes. One plan envisioned channelling water from the Three Gorges Reservoir to the dry north via an elevated canal; another involved diverting it from the headwaters of several major rivers on the Tibetan plateau, including the Yangzi, Yellow and Yarlung Zangbo, towards Lanzhou or Xinjiang. One version of this scheme proposed an annual transfer of 40 billion cubic metres of water from a Yarlung Zangbo tributary to the upper Yellow River for use in the north-west. The project would have dug a 30 kilometre underground tunnel and built a hydropower dam three times the size of that at the Three Gorges. ‘Experts’ promised the project would ‘turn green’ the barren plateau of the upper Yellow River while at the same time preventing annual floods in India and Bangladesh.

But it was He Zuoxiu, a retired vice-chief of the Institute of Physics at the Chinese Academy of Sciences who had helped build China’s first atomic bomb, and Gong Yuzhi, Vice-President of the Central Party School, who shared the grandest vision. At a national session of the Chinese People’s Political Consultative Conference in 1996 they proposed using nuclear blasts to dig an 800 kilometre underground canal from the Yarlung Zangbo river, under the Tibetan plateau and through the Kunlun mountains to the Taklamakan Desert. (The Yarlung Zangbo or Bhramaputra River flows through southern Tibet, past the main Tibetan cities of Shigatse and Lhasa, before turning south into India). Gong and He reassured their audience that according to Russian scientists over 200 peaceful underground nuclear detonations in the Soviet Union had produced no pollution. Their project was thus safe, though it would take fifty years to complete.

China has in fact embarked upon a three stage north-south water diversion project, the final stage of which will move water from the Yangzi headwaters in Tibet to the upper Yellow River. So far neither nuclear nor non-nuclear options to transfer water directly from the southern Tibetan mountains to the Xinjiang desert are in the offing. Nevertheless, PRC planners have implemented other large water diversion projects in an attempt to ameliorate environmental damage and alleviate water shortage in Xinjiang. Between 2000 and 2004 engineers with the Ministry of Water Resources diverted to the Tarim six ‘infusions’ of water from the Kongque River and from Lake Baghrash (Bohu or Lake Bositeng). In 2004 they transferred some 300 million cubic metres from the lake.

In northern Xinjiang a canal completed in 2001 drew off some 10 per cent of the water from the Irtysh to the oil drilling and processing centre of Karamay A new canal under construction in 2004 will channel more water from the Irtysh to developing areas 300 kilometres to the south. According to the Chinese ambassador to Kazakstn, Zhou Xiaopei, this new canal will raise the amount of Irtysh water exploited by China to 40 per cent of the total flow. The issue is of great concern in Kazakstan, as the Irtysh fills Lake Zaysan, serves a major hydropower plant, factories and agriculture in central and northern parts of the country, and provides the drinking water for the capital, Astana. The irtysh also flows into Russia, where its waters are crucial to a large industrial centre. Similarly, the Yili River originates in Xinjiang and empties into Kazakstan’s massive Lake Balkash. Increased use of Yili water on the PRC side has started to shrink this lake, and both the Kazakstan parliament and the United Nations Development Programme have expressed concern that without changes in how Kazakstan and China use the Yili waters, Lake Balkash could go the way of the Aral Sea, which lost 60 per cent of its volume during the Soviet Era, causing one of the worst ecological disasters of the last century.

By the mid-2000s both officials and public in Xinjiang were much more aware of environmental issues than they had been fifteen years earlier. Billboards trumpeting such slogans as ‘Protect our Xinjiang’s Natural Environment’ stood in downtown Urumchi, not far from those advertising Xinjiang’s genetically modified cotton. Chinese geographers, hydrologists, ecologists and commentators were publishing alarming articles on the state of the Xinjiang environment, even delicately pointing out the tension between continued population growth and sustainable development in the region. Impassioned discussion of Xinjiang’s environment buzzed over internet bulletin-boards. Tree-planting campaigns and programmes to modernise irrigation with water-efficient (anti-evaporative) equipment have earned commendation from the United Nations. Yet the public pronouncements of the PRC’s news service-cum-propaganda organ, Xinhua , retained the old militaristic language (’attacks’ by the desert on oases; ’walling off the sand to build fields’ and allow ‘oases to advance upon the desert’; use forests as ‘green screens on the front lines’) and triumphalist accounts of renewed reclamation efforts. Such public pronouncements portray the deserts as autonomous, antagonistic invaders to be repelled by Great Walls of trees; they remain silent on the emerging connection between upstream water use (some of it for the new trees in shelter-belts) and downstream desertification. Although they acknowledge the contribution of irrational development policies of the past towards the current problems, these reports have little to say about the continuing—perhaps accelerating—damage done by state entities or corporations and individuals operating in the looser economic and political climate of recent years. In particular, Chinese official publications do not acknowledge the fundamental contradiction between the pro-migration thrust of the Great Development of the West campaign and the environmental constraints on Xinjiang’s carrying capacity that will only tighten as the glaciers melt away.

Dissent and Separatism

Violence

One result of political liberalisation in the 1980s was a modest revival of Islam, evident from the opening of new neighbourhood mosques and Quranic schools. According to an early-1990 survey, Xinjiang had 938 Islamic schools with over ten thousand students. Two hundred and fifty of these were in northern Xinjiang, the rest mainly in Kashgar (350 schools) with smaller numbers in Khotan, Aqsu, Yarkand and the Kizilsu Kyrgyz Autonomous Prefecture. Yarkand had only thirty-three schools, but two thirds of their 722 students came from elsewhere in the region—perhaps indicating that Yarkand had retained its old reputation as a centre for Islamic learning. The Communist Party worried about these schools, the unemployed or independently employed young men who studied there and the idea that Islam might be taught as an ideology competing with Marxism. As discussed in the last chapter, protestors had referred to Islam in their slogans during a few street demonstrations in the early 1980s. In particular the popularity of Yarkand religious institutions raised the possibility that graduates from these schools, returning home throughout the region, would form a region-wide network. (This is precisely the role played by the Kashgar new schools early in the twentieth century).

The party also grew concerned about participation by cadres and party members, especially those at the village level, in what the party defined as ‘religious activity’. Exactly what this term means is unclear: while it certainly includes prayer and mosque attendance, it may also include fasting during Ramadan and attendance at such ritual events as weddings, circumcision parties and wakes conducted a month after a death. These celebrations grew more elaborate as people gained wealth, and they remain deeply engrained in Uyghur culture, especially in the countryside. In any case, by 1990 the party was retreating from the more liberal and flexible position it had assumed a decade earlier with regard to religion in a ‘minority nationality autonomy area’ like Xinjiang.

In January of 1990 authorities closed down the privately-run Quranic schools in Yarkand (and presumably other Xinjiang cities as well) and ordered students from outside the city to return home. According to Chinese accounts, several hundred madrasa students (talip) in Yarkand demonstrated in response, shouting ’study and protect Islam’ and ‘down with the infidels’ as they marched. The event was dubbed ‘the Talip Incident’.

Much more serious—indeed the episode which did the most to sound Chinese alarms and perk up journalistic ears regarding Islam-inspired separatism in Xinjiang—was the Baren incident of April 1990, known in Chinese as the Baren County Counter-Revolutionary Armed Rebellion. Baren county, Akto township (about 10 km south of Kashgar), had been named an ‘Ethnic Unity Model Town’ in 1984. This designation proved ironic six years later, when evidence of an armed uprising against Chinese rule emerged. As told in ‘internal’ Chinese sources,  a group led by Zeydin Yusup and known by the name of East Turkestan Islamic Party planned a series of synchronised attacks on government buildings in the Akto and Kashgar area. Starting in March 1990 planners are said to have been raising money and issuing a call to arms through mosques and cassette tapes. According to one report, they were also buying up horses on local markets, causing prices to sky-rocket, and telling people in Kashgar to close their businesses on 5 April. Learning that something was afoot, Chinese authorities sent investigators to Baren, forcing the group to act. A Chinese source says that the group then dispatched men to Turfan to blow up the railway (they were apprehended before making any attempt to do so). On the morning of 5 April some two hundred demonstrators shouting Islamic slogans surrounded the compound containing government and security offices for Akto county and Baren township. Later that afternoon the confrontation grew more aggressive, with some three hundred men surrounding the compound. Five policemen who had been sent to arrest Zeydin Yusup were themselves seized and their guns and other equipment taken. At a nearby bridge men ambushed two cars from the Baren armed police, killed their six occupants with axes and knives, and seized a few guns before torching the vehicles.

In the middle of the night, according to Chinese accounts, Yusup’s group issued an ultimatum threatening to attack if those barricaded in the compound did not hand over any arms stored within. Shooting broke out as the rebels reportedly tossed homemade grenades into the compound. During this exchange Zeydin Yusup was shot and the besiegers fled. Four battalions of infantry and artillery troops from the southern Xinjiang military district cordoned off the entire area, scattered the men holding the bridge, and pursued sixteen rebels who fled into the mountains. An Englishman then living in Kashgar with his Uyghur wife (he was selling hashish to make ends meet) saw tanks, jet fighters and helicopter gunships in the Kashgar area for some time after these events.

One Chinese source reports that 16 rebels were killed in fighting, 508 detained for questioning, 124 arrested, 40 convicted, 3 executed and 378 released after ‘education’. These numbers may be questioned; news media subsequently reported a wave of arrests across southern Xinjiang, and Amnesty International has published claims from unofficial sources that up to fifty people were killed by security forces, including some killed by mortars and gunfire from helicopters.

Possibly more revealing than casualty figures is the list, published in the same Chinese source, of weapons confiscated by authorities from Baren participants: 16 guns (including 15 taken from police in the course of events on 5 April), 470 bullets of various types, 243 earthenware (homemade) hand grenades, 53 kilos of blasting powder, 512 blasting caps, 180 knives of all sizes, three motorcycles and five horses. While demonstrating signs of prior planning and organisation, and relatively well-outfitted with explosives (said to be easily accessible in China due to loose control over construction materials), the ‘East Turkestan Islamic Party’ apparently launched its attack on the PRC government with few if any firearms. This much is revealed in the official account of Baren events, which mentions little shooting. Perhaps they were hoping to spark spontaneous mass rebellion in Kashgar and elsewhere, though that did not occur. Another interpretation suggests the convergence around the government offices began as a demonstration, only turning violent as the standoff progressed.

Chinese sources are silent on grievances underlying the Baren unrest, implying that they arose solely from Islamic belief and East Turkestan separatism. Other reports, and my own conversations in Kashgar a few months after the Baren incident, suggest that one underlying complaint was the application (in 1988) of birth limits to minority families: Uyghurs living in cities were now legally limited to two, and those in the countryside to three children. Someone set fire to the state Planned Birth office in a county near Khotan at the end of March 1990, shortly before the Baren events.

In 1991 and 1992 another small group was responsible for three explosions and two attempted bombings of civilian targets. According to Chinese reports, Ablimit Talip and several others bombed a bus station video lounge in Kucha on 28 February 1991, killing one and wounding thirteen. A year later, on 5 February 1992, they exploded two bombs on public buses in Urumchi during Spring Festival. One bomb blew up at the terminus where the bus had emptied and no one was hurt, but another killed three and wounded fifteen. On the same day, authorities report, timed devices were planted in an Urumchi cinema and residential area but failed to explode. Ten men alleged to be conspirators in these attacks were apprehended later in 1992, and two were killed in the process. Five men convicted in the case were executed in June 1995. There were other, unrelated, bombings in Kashgar between February 1992 and September 1993, including an explosion in an agricultural equipment company on 17 June 1993 that killed two and injured six, and a bomb in a wing of the Seman Hotel (or a building next door) where no one was hurt.

From 1990 to 1995 security forces reportedly rounded up over one hundred ’separatist counter-revolutionary organisations, illegal organisations and reactionary gangs’, arresting 1,831 people. Despite this abundance of enemies, however, the next major outbreak of unrest was not until July 1995, and involved the townspeople of Khotan. Chinese authorities had recently arrested two imams of the Baytulla Mosque in that city for discussing current events in their sermons, and replaced them with a young and charismatic imam named Abdul Kayum. When Kayum began in his own lectures to advocate improving women’s rights, he too was arrested for raising proscribed topical issues. Some days later, on 7 July, a crowd converged on a party and government office compound near the mosque, demanding information about the imam’s whereabouts. The confrontation turned violent and the government called in large numbers of riot police who trapped the demonstrators in the compound, deployed tear-gas, and arrested and beat many of them. Official reports mention injuries to sixty-six officials and police, but supply no figures regarding demonstrator casualties. If the Amnesty International report on this event is correct, it is ironic that Abdul Kayum was detained for promoting women’s rights—especially in light of the common assumption that Uyghur unrest arises from Islamic fundamentalism.

Another large demonstration took place on 14 August 1995 in Ghulja in northern Xinjiang. Approximately 700 to 800 people marched along a main city street as police and paramilitary units responded with barricades, snipers and armed patrols; the standoff was ultimately defused without violence.

Underlying this demonstration, and the more serious Ghulja (Yining) Incident two years later (see below), was an important social movement among Ghulja Uyghurs and the states reaction to it. Starting in 1994 Uyghurs in Ghulja and surrounding villages began organising a type of social club known as mäshräp. Traditional mäshräp as once practised in many parts of Xinjiang refers to a social gathering involving musical performance, comedy or recitations of some kind. The revived form was a grass-roots response to the gathering crisis of unemployment, alcoholism and drug abuse besetting Uyghur youth, one fuelled by frequent all-night drinking parties that had become institutionalised as a primary form of male socialising. Adopting strict rules of personal conduct inspired by Islam, the new mäshräp clubs substituted other activities for drinking and meted out humorous ritual punishments (soakings with water and other mild humiliations) for young men who had gone off the wagon since the last assembly. As they grew in popularity, mäshräp in and around Ghulja staged a successful boycott of local liquor stores. This drew the attention of government authorities, who banned the mäshräp in July 1995: the Communist Party does not tolerate popular organisations outside its purview. Unfazed, the mäshräps went on to organise a city-wide soccer tournament, again in order to provide young men with a healthy outlet for their energies. Days before the opening games of the tournament that summer, the city cancelled it, commandeering the playing field for military exercises. To make the point perfectly clear, soldiers confiscated the goals and parked tanks on the pitch. This precipitated the August 1995 peaceful demonstration.

During the first half of 1996 there were three assassinations of, or assassination attempts on Uyghur cadres in the government or the Chinese Islamic Association. One man, who was killed with three family members in his home in Kucha county, had held various village and township party positions and had served as representative at the Sixth National People’s Consultative Conference. The most infamous assassination attempt occurred in Kashgar, when three men with cleavers attacked the senior imam, Aronghan Haji, and his son one evening while the imam was on his way to preside over prayers at the Id Kah Mosque. Aronghan, who survived the attack with cuts on his head, hands, back and legs, was Vice-Chairman of the Xinjiang Region Consultative Conference and head of the Kashgar District Islamic Association. Also during this period, police engaged in six shootouts with Uyghur suspects throughout the region.

Besides this violence, there was apparently a good deal of popular unrest in Xinjiang in 1996. Though the accounts are vague and unconfirmed, the foreign press indicated a spike in protests, as well as tightened official repression in the spring of that year. One exiled Uyghur leader in Almaty, Yusupbek Mukhlisi, issued press releases taking credit for bombings, claiming to run a network of underground cells in Xinjiang and alleging that as many as 18, 000 people had been arrested and hundreds injured in clashes in 1996. However, Mukhlisi was prone to exaggeration, and the PRC white paper on terrorism makes no mention of him or his group.

Possibly related to increased unrest in 1996 were three major political developments. The first was the release by the Standing Committee of the CCP Politburo on 19 March of a secret directive (CCP Central Committee Document No. 7) warning of illegal religious activities and foreign influence in Xinjiang. The second was the signing of a mutual tension-reducing and security treaty by China, Russia, Kazakstan, Kyrgyzstan and Tajikistan, the so-called ‘Shanghai Five’, on 26 April (later expanded to include Uzbekistan and known as the Shanghai Cooperation Organisation). The third was the announcement of the first ’Strike Hard’ anti-crime and anti-separatist campaign later that same month. (These are discussed in more detail below). The high numbers of ’suspected terrorists, separatists and criminals’ arrested, initially given by PRC sources as 1,700 and later raised to ‘several thousand’, may thus be the result not of any upsurge in separatist activity or ethnic unrest at this time, but rather of the ‘Strike Hard’ campaign itself, which placed a political premium on speed and quantity of arrests and convictions.

The second-largest protest in Xinjiang’s recent history (since the demonstration and hunger-strike of thousands of Han ‘educated youth’ in Aqsu in 1979-80), and the most severe street violence since the Cultural Revolution, occurred in Ghulja in early February 1997. Starkly different accounts of this event emanate from official Chinese sources on the one hand, and foreign press, rights organisations and Uyghur groups on the other. After denying that it had happened at all, initial PRC reports called the incident a case of’beating, smashing and looting’ by ‘drug addicts, looters and “social garbage”’. A source dated December 1997 describes organised rioting by several hundred to over a thousand ‘thugs’ (baotu) and disturbers of the peace (naoshi fenzi). In an internal speech in 1999 Xinjiang Regional Party Secretary Wang Lequan likewise referred to the ‘Yili February fifth beating, smashing, looting disturbance incident’. In 2002 the PRC white paper on terrorism went a step further and called the incident a ‘serious riot … perpetrated [by] the East Turkistan Islamic Party of Allah and some other terrorist organisations’, claiming that from 5 to 8 February terrorists shouted Islamic slogans, killed seven innocent people, injured 200, destroyed thirty vehicles and burned two houses.

Not surprisingly, accounts by Amnesty International and Uyghur dissident groups operating outside China differ from the PRC version in many respects. In these accounts, people in Ghulja took to the streets on 5 February out of accumulated frustration following the banning of mäshräp in 1995 (some of which continued to operate underground), and waves of arrests during the ‘Strike Hard’ Campaign, the principal targets of which included religious students, some imams and members of ’illegal organisations’, including mäshräp. According to testimony gathered by Amnesty International, the police presence in Ghulja had been heavy early in the year. Not long before 5 February, still during the month of Ramadan, police had converged on a mosque to arrest two Uyghur religious students, touching off a scuffle as other people at the mosque intervened. Many were arrested. Several hundred Uyghurs then marched on the morning of 5 February to protest these arrests and other examples of what they saw as repression. They carried banners and shouted religious slogans. Armed police confronted the demonstrators; some demonstrators were killed or injured, either by shots the police fired into the crowd or from ricocheting bullets fired into the ground in front of the crowd. Hundreds of demonstrators were arrested. That afternoon more protestors took to the streets, including family and friends of those detained earlier. Police broke up this latest wave with clubs and tear gas, and again detained many people. Amnesty International’s informants say many children were injured in this second demonstration. Protests and riots continued and spread to suburban districts over the next few days, with thousands of people in the streets, met by a strengthened police and military presence. Some of the rioters threw stones at police, attacked Han Chinese, ransacked shops, and destroyed vehicles. Videotape shown on Xinjiang television includes scenes both of orderly marches and of clashes between rock-throwing youths and police with shields and truncheons. Other Chinese police or soldiers shown in the video carry machine guns. Amnesty cites unofficial claims that security forces fired on protestors, but there are no firm estimates of how many protestors where killed. By 8 February soldiers had restored order and imposed a curfew on Ghulja; the city’s rail, air and road links to the rest of the region were sealed off for two weeks.

Only seventeen days after the conclusion of the worst recent street clashes came the single worst terrorist act in PRC Xinjiang. On 25 February 1997 three bombs exploded on Urumchi public buses, killing nine and seriously injuring twenty-eight others. Devices left on two other buses at the same time failed to explode. The date of the bombings was chosen with political intent: official memorial ceremonies for Deng Xiaoping were being held that day in Beijing. Press accounts around this time commonly associated the Urumchi bus bombings with a 7 March explosion on a bus in central Beijing. Although this allegation continues to be recycled by later Western press accounts and academic articles, XUAR Chairman Abdulahat Abdurishit denied in May 1997 that the Beijing bombing was connected to those in Urumchi or perpetrated by Uyghur separatists, and the 2002 PRC document on ‘East Turkistan terrorists’ does not mention the Beijing incident.

Amnesty International’s informants report thousands of arrests and incommunicado detentions over the weeks and months following the Ghulja riots and Urumchi bombings, enough to result in a severe overcrowding of facilities. There were many alleged cases of abuse or torture of prisoners, including beatings, hosings with water in outside detention areas in the February weather and other techniques. The region-wide crackdown following the events of February 1997 became almost a permanent feature of life throughout Xinjiang. Security certainly remained extremely high until after the 1 July return of Hong Kong to PRC governance—an event which, it was widely rumoured, would be accompanied by some sort of protest or violent act in Xinjiang. Human rights organisations claim that the sweeps resulted in hundreds of arbitrary arrests; ’illegal religious organisations’, religious schools and political dissenters were special targets. There were at least 190 executions by April of 1999, many after cursory legal proceedings: in the case of eight Uyghur men accused of the February bus bombings in Urumchi, Amnesty International calculates on the basis of published Xinhua reports that ‘the sentencing hearing by the court of first instance, the appeal and review process, and the executions all took place within 13 days, between 16 and 29 May 1997: Convictions were often followed by public sentencing rallies, and another clash between Ghulja residents and police occurred in April 1997 when a small crowd of bystanders, whom PRC sources say were attempting a rescue, approached prisoners on parade after such a rally. Guards shooting into the crowd killed more people.

State Response: International Aspects

The crackdown in Xinjiang continued into the new century. It was, from one point of view, successful: between 1997 and 2005 there were no further large-scale acts of political violence in Xinjiang—nor, for that matter, any large demonstrations. The white paper on terrorism does mention twelve further deaths of Uyghur cadres and police deaths between February 1997 and January 2002, as well as sabotage of economic targets. Chinese sources claim that security forces continued to smash terrorist cells and illegal organisations and uncover weapons factories during this period, but it also interdicted attempts to smuggle arms from Central Asia. The main focus of Chinese concern thus shifted beyond its borders.

There are several ways in which the Uyghur separatist issue became ‘internationalised’ in the 1990s. Chinese sources had long warned of’foreign forces’ lurking behind the separatists in Xinjiang. Specifically they pointed to Uyghur exile groups and especially the activities of Isa Yusuf Alptekin (former Xinjiang member of the Guomindang Legislative Yuan) and his son Erken, who were active in Turkey and Germany; the PRC also indirectly, and sometimes directly, accused the United States of supporting the separatists. For example, when Wang Fang, Minister of Public Security, toured Xinjiang in August 1989 he announced, ’the root cause of [separatist] instability [in Xinjiang] lies in the attempt by the USA and other countries to split and subvert our country’ Given their long historical memories, we may perhaps understand why Chinese leaders think the US is intent on detaching Qing-era acquisitions from modern China: they remember staunch US support for Taiwan, the CIA’s infiltration of Khampa guerrillas into Tibet, and CIA agent Douglas MacKiernan’s meetings with Osman Batur in 1949 just before Osman and his Kazaks launched their armed resistance to Communist Chinese takeover in Xinjiang (see Chapter 5).

However, of more immediate official Chinese concern was the increased international attention to human rights in Xinjiang that accompanied the greater openness of the region. Amnesty International, Human Rights Watch and the US State Department’s Human Rights Report all began tracking the Xinjiang crackdown. From the 1990s the issue of Xinjiang separatism along with that of Uyghur rights took on an international dimension.

Besides these political issues, real security threats were brewing outside Xinjiang. Though the actual details and scale remain unclear, Uyghur groups were engaged in political organisation and some military training abroad. Chinese sources claim that till 1994 there had been secret training camps in southern Xinjiang, which later, under Chinese pressure, moved abroad. It seems likely that at one time or another in the 1990s some sort of weapons training went on in Kazakstan, Kyrgyzstan and Afghanistan. China was also concerned about activities of Uyghurs in Pakistan.

The Chinese response to these threats was a new diplomatic initiative, which, combined with China’s rising economic clout, vastly increased its influence in Central Asia. In 1996 officials representing China, Russia, Kazakstan, Kyrgyzstan and Tajikistan met in Shanghai and announced the formation of the ‘Shanghai Five’, a loose alliance meant to resolve border disputes lingering from the Soviet era, address mutual security questions, and facilitate economic cooperation. The border issues were quickly resolved, and with concerns mounting over Islamic extremism in the region and possible influence from the Taliban regime in Afghanistan, security rose to the top of the agenda. In 1999 a series of bombs in Tashkent targeted Uzbek President Islam Karimov, and in 2001 Uzbekistan, though not bordering directly on China, joined the organisation, which was then renamed the Shanghai Cooperation Organisation (SCO). The SCO thereafter devoted itself primarily to military coordination and anti-terrorism efforts, while providing a symbolic political counter-weight to the influence of the United States, then actively promoting NATO expansion to include Baltic countries once under Soviet control. President Jiang Zemin pointedly referred to the SCO as the ‘Shanghai Pact’ in what was most likely a conscious echo of ’Warsaw Pact’, the Cold War era Soviet counterpart to NATO in Europe.

Under the auspices of the SCO, China established a joint anti-terrorism centre in Bishkek, provided military aid to Kyrgyzstan, and engaged in joint military exercises with SCO partners. As a multi-lateral security organisation the SCO is a departure from earlier Chinese foreign policy approaches, but it has channelled and amplified Chinese influence in Central Asia, at the expense even of Russia, and is now seemingly taking on broader regional, even global significance. Mongolia received observer status in the organisation and in 2005 India, Pakistan and Iran were each pursuing membership. The United States by 2005 still would not deal directly with the organisation, but had taken notice. The SCO, together with the commercial ventures of mostly Chinese state firms, thus represented the greatest extension of Chinese power into Central Asia beyond the Pamirs since the Tang period. However, China’s first goal for the SCO had not been to project power into Central Asia, but to deal with Uyghur dissidents and potential militants abroad.

From Beijing’s point of view, the SCO succeeded brilliantly at this task. SCO membership presented the rulers of Central Asia’s new states with a choice; some commentators called this ‘playing the Uyghur card’. In return for good relations with the PRC, security cooperation, aid and, for Kazakstan, lucrative oil deals Central Asian governments sharply narrowed the scope of Uyghur activities in their countries. Such Uyghur militant training as had gone on in Kazakstan and Kyrgyzstan apparently ended by the mid-1990s; Central Asian governments also curtailed the rights of political assembly and fair legal process for both their long-term Uyghur minority citizens and recent immigrants and sojourners. From the mid-1990s Kazakstan and Kyrgyzstan began extraditing Uyghur suspects to China on Beijing’s behest. By the early 2000s only Uyghur ‘cultural organisations’ could legally operate in Central Asian countries, and then only under close government scrutiny. Uyghur political organisations were banned. In short, whereas during Soviet times Moscow determined that a certain level of support and official sympathy for Uyghurs in the Soviet Union was politically in its interest, in the post-Soviet era Almaty, Bishkek and Tashkent responded to a new calculus.

Though Pakistan is not a member of the SCO, China was likewise able to pressure Islamabad in the 1990s to close a market where Uyghur businessmen operated and to drive Uyghur students out of religious schools; unfortunately the students followed other graduates of Pakistan madrasas to Afghanistan, where some (hundreds in some accounts, 2, 000—3, 000 according to Chinese claims) wound up as soldiers for the Taliban. Chinese ambassador to Pakistan, Lu Shulin, met with Taliban leader Mullah Muhammad Omar in Kandahar in December 2000, seeking guarantees that Uyghurs would not receive military training. Muhammad Omar’s price for such an assurance was Chinese opposition to UN sanctions on the Taliban government in Afghanistan, and because China was not willing to block these sanctions, which stemmed from the 1998 al Qaeda bombings of US embassies in Africa, the Sino-Afghan talks yielded no results. Some Uyghurs fought with the Taliban against the Afghan Northern Alliance; others reportedly joined the Islamic Movement of Uzbekistan, a group connected with al Qaeda.

The 9/11 al Qaeda attacks on Washington and New York changed the ground underlying both anti-terrorism policies and China’s new diplomatic position in the region. The US response presented China with a dilemma, and in the first weeks after 9/11 China considered the new situation carefully. While on the one hand the US-led war in Afghanistan would eliminate the Taliban and presumably shut down any camps training Uyghurs in Afghanistan, it would also place the US military squarely in China’s backyard. In fact, ignoring the SCO entirely, in preparing for its invasion of Afghanistan the United States concluded separate agreements with Central Asian countries for bases and landing rights extending indefinitely into the future—a strategic disaster from China’s point of view. (By 2006 US forces had left the Uzbekistan base on Tashkent’s request, and faced pressure from Bishkek to quit Kyrgyzstan as well).

In the end, in responding to 9/11, Chinese leaders played their own Uyghur card, taking the opportunity to turn the official, public PRC position on Xinjiang separatism on its head. The earliest PRC approach to separatist tension in Xinjiang had been to minimise it by keeping most information under wraps and blaming only small core groups of vicious separatists backed by shadowy ‘foreign forces’, while at the same time turning out stories and images of beaming minorities thankful for the improvements in their lives made possible by the party. In the late 1990s Xinjiang authorities began discussing the violence in Xinjiang more openly. In March of 1999, for example, Xinjiang Uyghur Autonomous Region (XUAR) Governor Abdulahat Abdurishit claimed that there had been ‘thousands’ of explosions, assassinations and other incidents in the 1990s. However, officials moderated their statements in the early 2000s, perhaps concerned about the impact of such publicity on Xinjiang’s development under the Great Development campaign. Thus in welcoming Chinese and international business representatives to the Urumchi trade fair on 2 September 2001, Xinjiang Party Secretary Wang Lequan together with Abdulahat Abdurishit proclaimed that the situation in Xinjiang was ‘better than ever in history’. While mentioning separatism, they stressed that ‘society is stable and people are living and working in peace and contentment.’ Xinjiang’s nightlife, Wang enthused, continues until 2 or 3 a.m.! (Because Xinjiang officials set their watches to Beijing time, this is really the equivalent of 12 or 1 a.m.)

Nine days later the September 11 attacks and US response required a revised rhetorical approach to Xinjiang separatism. The new paradigm became apparent by the end of 2001, and was codified in a document released by the PRC State Council Information Office in late January 2002, entitled’ “East Turkistan” Terrorist Forces Cannot Get Away with Impunity’ (’”Dongtu” kongbu shili nantuo zuize’). The paper outlines a history of resistance to Chinese rule in Xinjiang, tracing the ideological origins of the ‘East Turkistan’ idea to the machinations of the ‘old colonialists with the aim of dismembering China’. It then catalogues violent acts of the 1990s and lists several groups operating inside and outside Xinjiang. The document’s wording and argument represent two shifts in the PRC international position with regard to Xinjiang separatism. First, it explicitly links what it now called ‘“East Turkistan” terrorists’ to international Islamist terrorism; the document also refers to Osama bin Laden and al Qaeda several times. However, neither ‘East Turkistan’ nor ‘terrorism’ were terms commonly used in earlier writing about separatism in Xinjiang. In internal party documents from 1996 and 1999, for example, the terms ‘national separatist’ (minzu fenliezhuyi fenzi) and ‘enemy’ were standard; but ‘terrorist’ or ‘terrorism’ did not appear. The name ‘East Turkistan’ had been taboo except in carefully vetted historical contexts; only after 2002 did official sources begin using it in scare quotes. Second, through frequent use of generic terminology (’”East Turkistan” forces’) the document suggests that all of the 200 terrorist acts and 162 deaths of the 1990s it lists were the work of a single, unified terrorist organisation, though a close reading of the document reveals this not to be the case.

Despite a measure of scepticism in the international press, the documents account of Xinjiang’s terrorist problem made for a simple and dramatic story: a single Uyghur terrorist organisation, linked to al Qaeda, opposing both China and the United States. The US government inadvertently amplified this impression when in August 2002 a State Department spokesman at the US embassy in Beijing announced the US designation of the East Turkestan Islamic Movement (ETIM) as an international terrorist organisation associated with al Qaeda and threatening US interests. ETIM is one of the groups listed in the 2002 document, but is not blamed there for any specific violent acts. Unnamed US government sources claimed that the US designation of ETIM was based on interrogations at Guantanamo of Uyghurs taken prisoner by US forces in Afghanistan. According to this intelligence, ETIM was plotting to blow up the US embassy in Kyrgyzstan. However, despite this independent corroboration of Chinese claims, in the public announcement the US spokesman adopted the language of the 2002 PRC document and specifically accused ETIM of 200 acts of terrorism, 162 deaths and 440 injuries in the 1990s. The PRC document itself stresses ETIM’s al Qaeda ties but does not mention it in connection with any specific acts in Xinjiang. Although US officials privately acknowledged the spokesman’s mistake, the State Department chose not to correct it publicly, and PRC press reports subsequent to the US announcement capitalised on the US error by proclaiming that the United States had designated the ‘East Turkestan movement’ in toto to be a terrorist organisation.

A more nuanced assessment of the record of political violence in Xinjiang in the 1990s would not describe a unified movement, let alone blame a single organisation. Rather, as the above survey shows, there were a series of incidents and attacks through the decade, a few of them clearly terroristic in the sense that they hit random people, irrespective of status, job or ethnicity; others not random, but rather carefully targeted political murders; and still others more in the nature of protests that turned violent rather than planned attacks. In the 2000s no Uyghur organisation publicly acknowledged militant or violent acts in Xinjiang or Central Asia. Finally, although it is difficult to judge on the basis of limited information, violence in Xinjiang seems to have tapered off from 1997 to 2005.

State Response: Domestic Policy Shifts

The other side of the PRC response to both the new international environment and outbreaks of unrest in Xinjiang from 1990 was a broad shift in domestic policy and popular attitudes regarding non-Han ethnic groups, particularly the Uyghurs. This shift was comparable in some ways to earlier pendulum swings in the 1940s, late 1950s-1970s and early 1980s, and involved reforms or reinterpretations of official policies towards religion, education, cultural expression and the ethnic make-up of Xinjiang’s government and party cadres, in addition to the encouragement of Han migration to Xinjiang discussed above. It thus entailed a rethinking of the foundations of the nationalities system in place since the 1950s, accompanied by subtle but portentous redefinitions of such key terms as zizhi (self-rule, autonomy) and minzu (nationality, ethnicity) upon which that system hung. While they were far less extreme than the chauvinistic attacks on Uyghur culture of the Cultural Revolution era, the overall thrust of these changes was nonetheless integrationist and even assimilationist, comprising a reversal of the relative liberality and tolerance of diversity of the early 1980s.

Hints of this new direction are evident in a document issued by the CCP Central Committee in 1996, a month before the first ‘Strike Hard’ campaign began. Known as ‘Document No. 7’, it comprises a set of recommendations on Xinjiang security from the Politburo Standing Committee in the face of threats from what it called ‘national separatism’ and ‘illegal religious activity’ aided by ‘international counter-revolutionary forces led by the United States of America’. While blaming the troubles on the standard ‘very small number of national separatists and criminals’, the document nevertheless focuses more generally on religion in Xinjiang as an urgent problem. It advocates strict controls on mosque construction and religious students and, in particular, on the religious belief and practices of basic branch level cadres, party members and students. The document refers to ‘village level organisations which have fallen into the hands of religious powers’ and warns that ‘Communist Party members and cadres are Marxist materialists and, therefore, should not be allowed to believe in and practise religion.’ In the peculiar parlance of the Chinese constitution, citizens have two religious freedoms—the right to believe and the right not to believe in religion. In the 1990s party propaganda made the Orwellian point that party members and students enjoyed only one freedom: the freedom not to believe in religion. This reversed the 1982 policy allowing a flexible interpretation of rules on party atheism in parts of China where religion was an important part of social life (see Chapter 6).

That the party suspected village cadres is further confirmed by another recommendation in the document: in addition to training more minority cadres dedicated to unity of the motherland, it orders ‘party members and soldiers’ from the PLA and the Bingtuan to be assigned to the county and township-level to ‘improve the structure of cadres’—in other words, to increase the percentage of Han cadres at the lower levels in Xinjiang. Likewise, Document No. 7 recommends training ‘a large number of Han cadres who love Xinjiang’ and ‘then relocating them to Xinjiang’. Indeed, it established a policy to ‘continuously import talented people’ to Xinjiang. This was a response to a brain-drain problem among educated Han cadres and technical personnel in Xinjiang already acknowledged internally by the 1980s: from 1980 to 1990 nearly 10, 000 cadres, 93 per cent of them Han, had left Xinjiang for the east where they expected easier living, more security, better pay and better education for their children.

Document No. 7 also expressed the Party’s apprehension about education in Xinjiang. In order to eliminate illiteracy, since the 1950s the national minority policies had permitted, even encouraged, education in autonomous regions in major non-Chinese languages. Uyghur-language education was available from primary through university levels in Xinjiang. However, directives in Document No. 7 to ‘investigate and organise schools’ and warnings about teachers and textbooks ‘which inspire national separatism and publicise religious ideas’ suggest that party leaders felt they had insufficient control over the Uyghur-language education system. The document also dictated tight limits on foreign cultural exchanges and instruction by visiting foreign teachers, and made attitude and political background the prime considerations in permitting Uyghur students to study abroad, even with their own money. It is of course ironic that even while promoting Xinjiang as the hub of the Silk Road and of the new Eurasian Land Bridge, the state would restrict foreign contacts and educational exchange for Xinjiang’s youth.

In addition to intensified efforts to apprehend criminals and separatist groups under ‘Strike Hard’, then, the 1996 Central Committee document codified a general tightening in the cultural arena among Xinjiang’s minorities in the 1990s. This tightening was especially evident in religious and educational matters. Religious personnel and teachers underwent courses of ’patriotic re-education’ and inspections by party work teams, and the state restricted the number of permits granted people applying to go on hajj. There were reports of new rules governing prayer: loudspeakers could not be used in giving the call to prayer; prayer before 9 a.m. was prohibited; and praying was restricted to those people who could fit within the mosque structures themselves—thus prohibiting the traditional practice of worshipers spilling over into squares outside mosques at prayer time on festival days. Regulations forbidding minors from participating in religious activities were strictly enforced in Xinjiang (though not in other parts of China); signs declaring no admittance to anyone under eighteen appeared above the doorways of Xinjiang mosques. University students were told that prayer, fasting for Ramadan, or, for women, wearing a head scarf, were inconsistent with communism and that they would be expelled if they continued the practice.

The state also began more vigorous policing of Xinjiang history. The decade began with the banning of historical works by the writer Turghun Almas, who was criticised in a public campaign and put under house arrest in 1991. Almas’ works, published in the more relaxed 1980s, famously propounded a version of the past that placed Uyghurs in Xinjiang in ancient times, before the first Han dynasty colonies, and thus ran counter to official historical interpretations that embedded Xinjiang and the Uyghurs as a part of Chinese history and denied Uyghurs’ claims to be Xinjiang indigenes. While the works and ideas of Almas and other Uyghur scholars and historical novelists continued to circulate privately, major state-sponsored historiographical projects through the decade turned out a stream of new pedagogical, popular and scholarly historical works publicising the official line of Xinjiang’s primordial Chineseness. The state symbolically demonstrated its intention to further ratchet up its enforcement of history in the spring of 2002, when the Kashgar Daily newspaper published a list of 330 banned books. These included Turghun Almas’ works, but also many other titles, including one on Uyghur craftsmanship that had been published years earlier with government approval. Thousands of volumes of these condemned works were collected and publicly burned in Kashgar.

Even tourist sites reflected the intensification of the historical message: a monument was built in Kashgar commemorating the Han Dynasty general Ban Chao—who in the first century CE had conquered the city by killing its king after inviting him to a banquet (see Chapter 1). InTurfan, the Su Gong Ta (Sulaiman Mosque) received a makeover. The Qianlong Emperor had built this graceful Samanid style minaret and mosque in the eighteenth century to commemorate the collaboration of the local ruler Amin Khoja with the Qing conquest of Xinjiang. The mosque once stood alone on a small rise amid grape arbours and mud-brick houses; by 2004 a large parking lot and garden had been added, and tourists approached the mosque by walking past a large modern statue of Amin Khoja. His hands and eyes are raised in an attitude of supplication as he receives the Chinese emperor’s edict. In the 1990s and 2000s Uyghur students and travel guides frequently commented on the pressure they felt to conform to the official version of history, even while doubting it themselves.

The most potentially significant change in Xinjiang cultural policy in the early 2000s was a reform of the multi-lingual education system to increase levels of Chinese fluency and literacy for all students. Like affirmative action for minorities (also implemented in Xinjiang), education in languages other than Chinese (in Xinjiang, including Uyghur and on a smaller scale Kazak, Mongolian, Sibo, Kyrgyz and Russian) posed dilemmas even as it afforded benefits to non-Chinese speaking students. While arguably helping non-Han students in their primary school years and promoting the culture of the nationality, lack of early immersion in Chinese could handicap students who later wished to go on to higher education and seek professional careers in Xinjiang or elsewhere in China. Xinjiang University, the top university in the region, had long maintained two faculties, one for Chinese-language and one for Uyghur-language instruction. In 2002 the government ordered that from the following year all classes except Uyghur literature would be taught in Chinese, and professors accustomed to lecturing in Uyghur were given a year to brush up their Mandarin. Around the same time it was reported that Uyghur schools in Hami (Qumul) were to be merged with Chinese-language schools. Uyghur groups abroad have presented these reforms as an effort to eradicate the Uyghur language, and certainly Xinjiang Party Secretary Wang Lequan’s own comments at the time displayed a dismissive attitude toward languages other than Chinese: the shift to Chinese-language education was necessary, he said, to ‘improve the quality of ethnic minorities’ because ‘the languages of the minority nationalities have very small capacities and do not contain many of the expressions in modern science and technology which makes education in these concepts impossible. This is out of step with the twenty-first century’ In fact the introduction into Uyghur and other Turkic languages of scientific terms from Russian from the late nineteenth through the middle of the twentieth century has afforded them a Latinate technical vocabulary very close to English—and thus to international scientific terminology

Nevertheless, solid knowledge of Chinese is important for all PRC citizens, and official reports from China suggest that the goal of the Xinjiang education reform was not to eliminate Uyghur, as dissidents claim, but to turn the system of Uyghur-language primary school instruction into a bilingual education system through increased use of Chinese from earlier grades. A Uyghur professional with a prestigious government job described to me the educational reforms as they affected her own young child. In her characterisation, the new system resembled bilingual schools in the United States, amounting to a replacement of mono-language Uyghur instruction with bilingual Chinese-Uyghur instruction, and to her this was a positive development. Towards this end, in 2004 Xinjiang authorities announced plans to train 55, 000 bilingual teachers. Whatever the motivation and extent of these educational reforms, they betoken a change from the second half of the twentieth century when large numbers of non-Han in Xinjiang had little grasp of Chinese.

As with development projects, then, the thrust of both official historiography and these educational reforms was to integrate Xinjiang more closely to the rest of China. Indeed, there is reason to believe that by the early twenty-first century Chinese leaders and ideologues had grown uneasy with the policy towards non-Chinese peoples and regions implemented half a century earlier. As discussed in Chapter 6, when inaugurated in the 1950s this system had identified fifty-five official ‘minority nationalities’ (minzu) and created ‘autonomous’ units at township, district, prefectural and regional levels in areas where these peoples were concentrated. State cultural organs reinforced the definitions of each nationality through linguistic, literary and historical publications in its language, codification and promotion of its particular music and dance forms in arts schools and similar means.

The system had served a propagandistic purpose in the 1950s by casting the Chinese Communist Party as more sympathetic to minority concerns than the assimilationist Guomindang. Gerrymandering of nationality enclaves and maintenance of real power in the hands of party officials guaranteed that the Xinjiang Uyghur Autonomous Region did not deviate far from central directives. Nevertheless, the fruitful ambiguity of’national autonomy’ had defined Xinjiang as a non-Han cultural space and at times—the early 1950s or early 1980s in particular—even justified flexible implementation of central policies and official support for non-Han peoples against ‘Han chauvinism’.

Although it was not nationalism per se that led to the secession of the Central Asian republics from the Soviet Union, the latter did break up along the same national lines defined by Stalin. The Chinese constitution, unlike the Soviet, never granted the right to secession to Chinese ‘nationality autonomous regions’; nevertheless, PRC minzu policy over fifty years had highlighted, rather than erased, the seams in China’s patchwork national fabric. The US-led NATO intervention in Kosovo in 1999 in the name of human rights and self-determination for a Muslim minority sent a chill through Chinese leaders concerned about the strategic implications of Xinjiang instability. In 2000 Ma Dazheng, head of a research institute dedicated to frontier matters and a prolific writer on Xinjiang, responded to internal debate over minzu policy with an interesting reinterpretation of the notion of zizhi (the term translated as ‘autonomy’). Whenever discussion turned to the implementation of minzu policy or Xinjiang autonomy, Ma wrote, many people, especially minority comrades, immediately raised the issue of the representation: the fact that non-Hans hold fewer government positions proportional to their numbers in the Xinjiang population than do Han cadres. Debate over these percentages in Xinjiang was an old problem—indeed, although Ma does not mention this, it goes back to the era of Guomindang rule in the 1940s.’Nativisation’ of Xinjiang leadership was promised in the 1950s; critics raised the issue again in the early 1980s, in the reform era following the Cultural Revolution. However, Ma argued, the phrase ‘nationality regional autonomy’ (minzu quyu zizhi) should not be understood to mean ruling power in the hands of a single nationality, but rather collective rule by all nationalities in the region {quyu nei ge minzu gongzhi). Autonomy by a single nationality is only a stage on the way to such collective nationality rule. Exclusive focus on one’s own group, in a diverse region like Xinjiang, is excessive and divisive. Thus while sounding very reasonable, Ma upturns the premises of fifty years of minzu theory.

Ma presents his argument as a personal opinion, and it is unclear what influence it had on the leadership. However, in the early 2000s party theorists were already revising basic concepts of the minzu system in a similar direction. One indication of this was a change in the official English translation of the term minzu: In place of the old translation, ’nationality’, which reflected Stalinist nationalities theory and policy, official PRC materials adopted the term’ethnic’. For example, the Nationalities Affairs Commission became the Ethnic Affairs Commission, and a white paper released in February 2005 bore the title ‘Regional Autonomy for Ethnic Minorities in China’. This departure from the defunct Stalinist nationalities policy turned China at a stroke from an ideologically cumbersome ‘multi-national state’ into a ‘multi-ethnic country’ like the United States. Moreover, although the original Chinese terminology remained unchanged, the new English translation effectively downgraded Uyghurs, Tibetans and others from the status of’nationality’ (which in English might imply rights to self-determination and perhaps a nation-state) to that of an ethnic (and hence sub-national) group. In the aftermath of the violent incidents of the 1990s reinterpretations of zizhiand minzu demonstrated the PRC’s new integrationist approach to Xinjiang as clearly as the extension of the railway line to Kashgar.

Inter-ethnic Relations and Subtle Dissent

If the sporadic violent episodes of the 1990s seemed to have tapered off by the 2000s, personal relations between ethnic groups, particularly between Uyghur and Han, were if anything more tense than at the start of the decade. This tension emerged more often in mundane ways than in open acts of protest or oppression, but it remained palpable nonetheless.

Indications of less than harmonious relations between minzu were readily apparent in the 1990s-2000s. Many Han Chinese in Xinjiang, particularly recent arrivals, often harboured fearful, derogatory or stereotyped images of Uyghurs and other minorities, and there were few taboos about openly expressing these views. For example, visitors from abroad were cautioned that Uyghurs carry knives and rob the unwary. At the same time, they learned that Uyghurs ‘excel at song and dance’ (nengge shanwu). Qnce in 1990 on a visit to Heaven Lake (Tianchi, Tengri köl) I astounded a group of Han tourists by photographing some Kazak children who were mugging for my camera. The tourists asked, ‘Why are you taking their picture? Those are Kazak kids!’ (Tarnen shi Hazu wawa). On another excursion, I rode a bus where the Han bus driver had retained the services of a Kazak teenager, evidently in return for free passage to Urumchi. When the Kazak youth made a small mistake while helping the driver, one Han college student rolled his eyes and loudly groaned ‘Kazaks!’ (Ha-zu!) for the benefit of his friends and other riders.

Because of the language difference inter-ethnic friction is often apparent in the educational arena: One Han first grade teacher at Xinjiang University’s Chinese-language track primary school expressed dismay when she discovered that half of her students for the new year would be minorities. ‘They have their own class’, she said. ‘Why can’t they go there?’ Slurs may be unconscious: I have heard well-intentioned Han Chinese praise professional Uyghurs for ‘speaking Chinese very well, for a Uyghur’. Nevertheless, intended or not, such comments—and examples could be multiplied manyfold—suggest pervasive Han prejudice regarding Xinjiang’s minorities. This prejudice is reinforced by state policies, the practice, widely-resented by Han, of admitting Xinjiang minorities to university with lower exam scores than Han; and the rhetoric of development, which stresses the benefits afforded poor, backward Xinjiang by the centre. In the 1950s propaganda regularly referred to Han as the ‘big brother’ and minorities as ‘little brother’ minzu. Although these terms are no longer publicly used, their patronising message still permeates private and official discourse, as in the encounter between Jiang Zemin and the Kazak herdsman quoted above. Wang Lequan’s recent comments that Chinese-language education was necessary to ‘improve the quality of ethnic minorities’ and bring them ‘in step with the twenty-first century’ are similar examples.

Faced with an ‘autonomy’ system that leaves real power in the hands of Han party secretaries, rapid economic development that seemingly benefits Han migrants more than Uyghur residents, a strenuous crackdown on anything resembling political organisation or criticism of the government, tightening state restrictions on religion, muscular enforcement of a Sino-centric historical narrative, a widespread sense that the government and judicial system favour Han, and pervasive chauvinism on the part of Han neighbours, Uyghurs have responded in a variety of ways. The open protests detailed above have been relatively rare; more common are quiet forms of eyeryday resistance or dissent. These too can take the form of ethnic slurs, this time directed at the Han. For example, I’ve been told by Uyghur informants that Han are calculating, dirty and promiscuous; they like to live with pigs and strew garbage all over their own neighbourhoods. Uyghurs will privately refer to Han as Khitay, a word centuries old. It originally referred to the Khitan people, is the source of the Russian word for China and the word ‘Cathay’ in other European languages and is the standard term for China in Central Asian Turkic languages today. Nevertheless, Chinese authorities perceive it as derogatory and ban it—thereby assuring its continued use by, among others, Uyghur school-children. While investigating the situation in Khotan in the late 1990s a ‘stability work group’ was chagrined to discover that Mao Zedong’s portrait had been torn out of 3, 722 lower school textbooks in the area—a literally separatist act! Worse still was the explanation one student gave: it’s a Khitay head, so he ripped it out of his new textbook, same as he’d done four years in a row. To this Uyghur schoolboy, Chairman Mao was just another Chinaman.

One scholar from Hong Kong has attempted to quantify the state of Han-Uyghur relations in Xinjiang. Despite evident reluctance of Uyghur respondents to answer sensitive questions candidly, his surveys of almost four hundred Hans and Uyghurs in Urumchi in the year 2000 reveal a fairly deep rift between the two communities. Although half of the Uyghurs in the sample spoke good Chinese, only 32 per cent believed Han and Uyghurs should marry. (Indeed, actual rates of intermarriage are much lower than this, and Han and Uyghur socialise little together.) A higher percentage of Uyghur respondents expressed pride in being a Uyghur national (91%) and Xinjiang resident (95%) than in being a Chinese citizen (88%); only 43 per cent of Uyghurs (vs 72 per cent of Han) said they ‘strongly believed’ that Xinjiang has been part of China since ancient times. Although half of Han and Uyghur respondents agreed that the open door reforms had raised standards of living for both groups ‘about the same’, nearly 40 per cent of Uyghurs believed that Uyghur standards of living had risen slower than those of Hans, and a majority of Uyghurs thought that there was a significant disparity of income between Han and Uyghur. On a major point of PRC propaganda, 53 per cent of Uyghurs either disagreed or found it ‘hard to tell’ whether national separatism was the ‘main danger to Xinjiang stability’. Not surprisingly, 81 per cent of the Han respondents agreed with this statement. Finally, fairly large percentages of each community believed their own ethnic group to be cleverer and more hygienic than the other.

Other scholars working in Xinjiang have examined Uyghur contemporary culture to uncover a variety of ways in which Uyghurs express dissent through alternative versions of history, political jokes, folk-sayings and allegorical songs. In one joke told in Xinjiang in the mid-1990s, Jiang Zemin meets with Zhao Ziyang (the Chinese leader disgraced for supporting the students on Tiananmen Square in May—June 1989). Jiang offers to rehabilitate Zhao by giving him a position as a vice-chairman. Zhao angrily shouts back, ’I’m not a Uyghur!’ The joke turns on the common knowledge that Uyghurs in the ‘Xinjiang Uyghur Autonomous Region’ never occupy principal positions, but rather fill the number-two seats.

Uyghur singers have used carefully chosen metaphors in their lyrics to make powerful political statements while eluding censorship. For example, the Uyghur folk-singer Omärjan Alim has sung about a guest overstaying his welcome, and about barren chickens occupying the roost and getting the grain while the fertile hens have to scratch for subsistence outside. Multiple interpretations of such songs are possible; in one reading, the Han with their strict limits on childbirth are the barren hens who have displaced the fecund Uyghurs. Such songs were very popular, circulating widely through cheaply reproduced cassette tapes. Some veiled political expression is critical not of the Han or the government but of Uyghurs themselves and aspects of Uyghur culture. For example, the saying ‘the axe-handle is always made of wood’ slyly points out that PRC rule in Xinjiang has always relied on a Uyghur support staff.

Perhaps the simplest but most ubiquitous assertion of private independence in Xinjiang consists of setting one’s watch to Xinjiang time, two hours behind that of Beijing. Despite its vast west-east girth, China has only one time zone—an echo of the imperial era when Chinese emperors assumed ritual authority over matters calendrical and chronological. The official world and most Han in the Xinjiang, as throughout China, follow Beijing time. However, non-Han residents in the region have maintained a defacto Xinjiang time zone in the face of official uniformity. This is an act of defiance, but also arguably one of convenience, for by so doing Xinjiang residents can get up at 6 or 7 a.m. (instead of 8 or 9 a.m.), have their lunch (chüshlük tamaq — ’noon meal’) around noon (not 2-4 p.m.) and so on.

Chinese Turkestan in the Early Twenty-First Century

As an emblem of globalisation, fast-food chains are now a soggy cliché. Nevertheless, their advent in a new corner of the world is noteworthy, not so much as a harbinger of globalisation, but as an indicator of how integrated a place has already become. The Kendeji-Kentakiy-KFC outlet in Urumchi was Xinjiang’s first international fast-food franchise. Before the KFC parent company, Yum Brands, could plant its flag in Xinjiang, problems of licensing, supply, demand, labour management, accounting standards, quality control and so forth all had to be resolved to its satisfaction. Besides these prerequisites, the arrival of KFC marked still other changes on the local level. The new Grand Bazaar pedestrian mall, of which KFC became an anchor tenant, replaced the old Erdaoqiao Bazaar, until recently the heart of the Uyghur quarter, the commercial centre of the city and a favourite place for lunch. While vibrant, Erdaoqiao bazaar had also been a noisy, messy firetrap.

The Kentuky-isation of Urumchi underscores a central fact about turn-of-the-century Xinjiang: integration with the world and with China were parallel, largely inextricable processes. For good or ill, at the turn of the century Xinjiang was in the throes of both.

This chapter has attempted to encapsulate many still-unfolding events and trends of the decades spanning the turn of the twenty-first century. This is somewhat like trying to sketch, from a stance hard by the tracks, the train hurtling past on its way to Kashgar.

However, the broad lines of the picture are clear enough. In some respects the changes in Xinjiang since the mid-1980s have followed directions already defined during the PRC period, some even introduced during the eighteenth and nineteenth centuries by the Qing authorities. The pace of change has, however, accelerated. Xinjiang’s economy has expanded, although governance of the region from Beijing continues to require expensive subsidies. The region’s population has grown rapidly, largely through migration, though this migration has been voluntary, not the mandated population transfers of the past. Exploitation of Xinjiang’s resources proceeds at an ever faster rate as land is reclaimed, water channelled, oil and gas extracted.

In some ways many developments of the 1990s and 2000s are new. First, Xinjiang’s environment may have reached carrying capacity. Although more efficient use of available water will permit some ecological rehabilitation and even some further growth, and there may still be fossil water in untapped aquifers, development in the region cannot continue in the unbridled fashion of the past fifty years. Xinjiang inhabitants may have to cope with steep reductions in water supplies as glacial sources disappear.

Second, China’s hold on Xinjiang is now more secure than ever. The disappearance of the Soviet Union and reduced circumstances of Russia eliminated a long-term imperial rival in Central Asia, one once both willing and able to dabble in Xinjiang affairs for economic or political gain. New US bases in Central Asia notwithstanding, there is no new counterpole just across the Pamirs from Xinjiang. In fact China is the emergent economic and diplomatic leader throughout the Central Asian region. A Chinese pipeline spans Kazakstan; Chinese goods fill Central Asian bazaars; one can even find Chinese-language software on computers in Almaty internet cafés.

Third, in tandem with changes in the international context, Xinjiang’s own situation vis-à-vis China, the region and the world has changed. Over the long term historically, the region has been more often a crossroads than a cul-de-sac . Though relatively remote for much of the latter half of the last millennium, and particularly during the decades of frosty Sino-Soviet relations, the region is a backwater no longer. The Soviet break-up gave Chinese planners a green light for intensified integration of Xinjiang with China through investment in infrastructure. Reform policies and improved communications likewise opened the region to easier contacts with foreign countries. Moreover, these contacts were not limited to trade: despite state efforts to monitor and control the flow of information in and out of Xinjiang, people there entered into increasingly robust communications with people elsewhere in Central Asia and the world. Expatriate dissidents, international media and human rights groups began to receive and publish regular reports from Xinjiang. Information from these sources on such events as the Ghulja Incident successfully challenges the official PRC versions.

The resentment and resistance of some non-Han groups to Chinese rule during this period was not in itself new. But official approaches to it did present some novel developments. It remains to be seen to what extent Uyghur separatism is linked with groups espousing terrorist methods and a radical Islamist agenda, as Chinese propaganda argues and US acquiescence to it implies. But the new representation of Xinjiang as a battleground in the ‘global war on terror’ was a departure from previous PRC practice, as was the correspondence of a crackdown in Xinjiang with modest expansion of free expression elsewhere in China. Likewise, early in the twenty-first century Chinese leaders seemed to be departing from the minzu autonomy system as the CCP had erected it in the early 1950s. Though this ‘autonomy’ had always been closely hedged, in the early 2000s the party seemed to be questioning its own earlier theory and the premises underlying special status for non-Han groups and ethnic-majority regions in the PRC. The start of the new century thus saw Xinjiang occupying a new position both in China and in the world.