CHAPTER 2

COLLAPSE OF EMPIRES AND THE NATIONALIST THREAT

THE NARRATIVE OF CHINA THAT MOST HISTORIANS ARE FAMILiar with recounts that following the death of President Yuan Shikai in 1916, “warlords” and “ideas” take center stage. In such tellings, the warlords, charged with having “made national politics look like a game of musical chairs with guns,”1 provide the backdrop against which purveyors of progressive ideological agendas begin to lay the foundations for a new China. More often than not, the landmark political events of this decade are scrutinized less for their contemporary geopolitical relevance than for the ostensible influence they were believed to exert in stimulating a new revolutionary movement, the realization of which lay far in the future.² Thus the Treaty of Versailles is linked to the May Fourth Movement, the Bolshevik revolution to the establishment of two Leninist party-states, and the May 30th Incident to the Northern Expedition. For those whose interest in China is limited to the lives and livelihoods of its ethnocultural majority—Han resident in the inner provinces—this framework of China as a nation beset by imperialists has long proven serviceable. In fact, even when historians look to complicate or otherwise interrogate this narrative, they still tend to proceed from a conceptual premise that takes the Han narrative of the nation-state as its normative focus.

When we adopt the perspective of a Han governor in Xinjiang, however, it quickly becomes apparent just how inadequate the concepts of “revolution” and “nationalism” are to an understanding of developments outside the Han heartland. As we saw in chapter 1, Governor Yang Zengxin (1912–28) repeatedly justified his policies in terms of their suitability to “empire,” understood to be a type of state that privileges the institutionalization of ethnic and spatial difference. Seen from this point of view, the era of the warlords in China (1916–28) and the prolonged course of the Bolshevik revolution in Russia (1917–23) were marked less by ideological ferment than by the contest for geopolitical power in Eurasia. In other words, the fate of Han rule in Xinjiang after 1916 allows us to reframe the following four decades in China not in terms of the success or failure of the revolution—often viewed as a vehicle for Chinese nationalism—but rather as the collapse and reconstitution of imperial authority across Eurasia, as brought about through competing strategies of difference. When the Bolsheviks, free from the entrenched colonial interests of outside powers, managed to reconstruct their empire first, the stage was set for the exportation of a new form of difference to the non-Han frontier of China, with which the former czarist empire shared an extensive border.

In Xinjiang, the collapse of imperial authority both in China and in Russia was experienced almost simultaneously. The stage on which the initial collapse played out was the Kazak and Kyrgyz refugee crisis of 1916. Long exempt from forcible integration into the political projects of European Russia, the Turkic-speaking nomads of Russian Turkestan responded to calls for their conscription to the European war front by organizing armed resistance throughout the Karakul region. The decision to renounce a long-standing czarist policy of ethnic privilege for the empire’s non-Russian peoples was a repudiation of the politics of difference, implemented in response to a geopolitical crisis in the metropole: manpower shortages along the Russian front during the Great War. When the consequences of Saint Petersburg’s actions spilled over into Xinjiang, Governor Yang was able to obtain his first glimpse of the deterioration of central government authority in the Russian empire. In attempting to manage the fallout from the Russian collapse within his own province, however, Yang would also come to realize just how fragmented his own Chinese republic had become in the brief time since President Yuan passed.

RUSSIAN COLLAPSE AND RECONSTITUTION

“Countless Kyrgyz lie dead in the passes, piled one on top of the other. Their frozen corpses, stuck solid to one another, obstruct the mountain roads. It is a sight too horrible to describe.”3 This was the gruesome report forwarded to Governor Yang’s desk on October 2, 1916. It described the fates of thousands of Kyrgyz and Kazak nomads who had fled Russian reprisals and ventured up into the Pamir Mountains of western Xinjiang. Informed that Russian authorities had no intention of sparing the remaining 300,000 Kazak and Kyrgyz refugees from a similar fate, Yang Zengxin reflected on the depths to which the Russian state had suddenly plunged. “The mountains are littered with corpses,” he wrote. “Russia is supposed to be a great civilized country, with respect for human life. How can they let this happen?”⁴ The answer to the governor’s question was to be found in the shivering bodies of innocent bystanders, snatched as strategic collateral by fleeing nomads. “When the Kyrgyz fled from Russia,” Yang later learned, “they took with them a countless number of white-skinned women and children as hostages. These women and children are now howling from starvation and cold in the passes, their plaintive cries for help unanswered.” Before long, Russians armed with machine guns and bayonets were streaming over the border at will, in frantic search of wives and children, descriptions of whom were posted as “missing persons” signs throughout the province.⁵

For Yang, the crux of the crisis was its timing. Just a few months before the first refugees and their heavily armed pursuers crossed into his jurisdiction, President Yuan Shikai passed away in Beijing. With his passing went all semblance of central authority in China, with no clear successor in his wake. In the absence of Yuan’s unifying presence, Han officials had only discredited institutions to work with. Whereas the revolutionaries had undermined the authority of the monarchy, Yuan’s abolition of the republic, suppression of parliamentary bodies, and attempt to establish his own monarchy had all done irreparable damage to the integrity of the republic. Both of these processes were ultimately rooted in the foreign presence in China, in the face of which neither the Manchu court nor its Han successors had proven able to protect domestic interests. As a result, after Yuan’s passing, Chinese officials lacked a clear consensus regarding whom they could entrust to coordinate the affairs of the former empire. In Xinjiang, the loss of support from the Chinese metropole spelled trouble for Governor Yang, who could no longer look to a higher authority to enforce his writ throughout the province. The Kazak and Kyrgyz refugee crisis thus constituted the geopolitical context through which Yang would witness the collapse of imperial authority in both China and Russia.

Outside the provincial capital of Urumchi, the governor’s political authority was severely circumscribed. Among those regions of Xinjiang still retaining their own military units and the right of direct communication with Beijing were the Altay minister (zhangguan), Tarbagatai councilor (canzan), Ili general (jiangjun), and Kashgar commander (titai). As any glimpse at a map of contemporary Xinjiang will show, the jurisdictions of these four semiautonomous offices were all situated in strategic regions sharing a border with the Russian empire. As a result, the vast majority of Kazak and Kyrgyz refugees fled into mountain strongholds directly under their administration. Over the next year, Yang engaged in a three-way struggle among himself, Beijing, and the members of this semiautonomous quartet to determine who held ultimate authority for coordinating government policy toward the Russian refugees. To Yang, the threat to provincial security was self-evident. “For each chief, there are roughly one thousand subordinates under his rule,” Yang explained. “Even the lesser chiefs claim the loyalties of several hundred men. If Russian troops cross the border and begin to hunt down these chiefs, untold numbers of their subordinates may rise up in resistance.”6

One by one, each member of the semiautonomous quartet proceeded to adopt a martial solution to the refugee crisis, deploying troops or other coercive measures to round up Kazaks and Kyrgyz and deliver them to the nearest Russian consulate. When Yang’s provincial rivals began to justify their actions via recourse to directives issued in Beijing, the governor responded by insisting on the superiority of policies bred from long experience in China’s northwest—in other words, his own. “Even though these are the Ministry [of Foreign Affairs]’s orders,” Yang wrote on one occasion, “it appears that they simply do not understand the situation on the borderlands. We must act in accordance with local conditions and avoid rash actions that will lead to incidents.”7 As heavily armed Russians tramped across the land and Xinjiang’s officials prodded cornered nomads like so much cattle, the governor made frantic efforts to regain control of the situation. When some twenty to thirty thousand Russian Kyrgyz holed up in the mountains of Aksu—the only region in Xinjiang sharing a border with Russia but not subject to the rule of one of the governor’s provincial rivals—Yang decided to make a stand.

Aksu was the only region within Xinjiang to which Yang had ever been posted outside the capital, having served as daoyin (circuit intendant) there for three years following his transfer from neighboring Gansu. Just as important, its current daoyin, Zhu Ruichi, was, like Yang, a veteran Qing official who had spent the greater part of his career in the non-Han borderlands. The two men implicitly trusted one another, viewing themselves as a distinct class of frontier official whose areas of expertise were increasingly impinged upon by carpetbagging Han sent from Beijing. When a telegram from Semireche informed Yang that a Russian punitive expedition to Aksu was imminent, he and Zhu coordinated a united front in response. Much to the surprise of both men, Beijing had secretly authorized the Russian expedition. “When the Russians arrived here,” Zhu reported, “the first thing they told me was that our government had already approved the capture of Russian criminals.”⁸ Zhu, firmly in Yang’s camp, continued to drag his feet, prompting a formal note of protest. “The unwillingness of your officials to assist us in the apprehension of these criminals resulted in a host of problems,” the Russian consul in Urumchi later complained to Yang. “Please order the daoyin of Aksu to help us in capturing these bandits.”⁹ With the Russians frustrated in Aksu, Yang sent Zhu to take up the post of daoyin in Kashgar, where Zhu and Yang’s brother coordinated a pacifist campaign in contravention of the forward approach advanced by Ma Fuxing, the Kashgar commander.

FIGURE 2.1. The Kazak refugee crisis and the autonomous quartet, 1916–18. Cartography by Debbie Newell.

Zhu’s transfer to Kashgar coincided with the February Revolution in European Russia, followed just nine months later by the Bolshevik revolution. In Xinjiang, the impact of the former was felt immediately. The transitional government in Saint Petersburg, now consumed with far more serious matters closer to home, no longer had the time or resources to devote to the situation in Xinjiang. Before long, a general amnesty was issued, and both sides agreed to peacefully repatriate all remaining nomads. As this was what Yang had lobbied for all along, he jumped at the chance to usher fugitive Kazaks and Kyrgyz back over the border. To his surprise, however, the governor soon came face-to-face with an unfamiliar feature of the new political landscape. In short, while the Russian central government in Saint Petersburg may have lost all appetite for confrontation, vigilante Russians stranded on the ground in Central Asia most certainly had not.

The first tangible signs that Russian imperial authority was in freefall came in June 1917, in the form of a report to Kashgar daoyin Zhu Ruichi. Eyewitnesses recounted a chilling scene. Approximately seven hundred Russian Kyrgyz, armed with little more than consulate-issued amnesty papers from Kashgar, had been gunned down as they passed through an alpine valley. Vigilante Russians equipped with machine guns were almost certainly responsible for the massacre. “What is the meaning of all this?” Yang thundered to Beijing. “First they issue a general amnesty, then they execute every last one of them!” Getting these nomads to return in the first place had been a nigh-impossible task, Yang observed, one that required vast amounts of provincial funds and labor. “Not only has this inhumane action resulted in the deaths of those who ventured to return, but it also ensures that those still remaining will not risk a similar fate.” Throughout the summer of 1917, additional reports of massacres and poisonings continued to reach an exasperated Yang. The Russian ambassador in Beijing, Prince Nikolai Kudashev, could not have cared less. “Before they fled to Xinjiang,” the prince patiently explained, “the Kazaks pillaged and ravaged the people of Russian Turkestan. It is only natural that the people will enact their revenge upon their return.”10

The response from the Russian ambassador, while certainly cold, was also a tacit admission that Saint Petersburg was now powerless to put a stop to the carnage in Xinjiang. For Yang, this was uncharted territory. Never before had a Chinese official of his generation witnessed the wholesale collapse of the authority of a European power in China, unless the vacuum created by its retreat was immediately filled by the advance of a rival. But no one stepped in to fill the Russian void in Xinjiang. Similarly, Yang was not accustomed to a Chinese metropole that proved unable or unwilling to help him coordinate policy among his nominal subordinates in Xinjiang, as had been the case while President Yuan Shikai was in office (1912–16). As such, the grim reality of the situation soon dawned on the governor: even prior to the outbreak of the Bolshevik revolution in late 1917, the political authority of both the Russian and Chinese empires throughout Eurasia had already suffered irreparable damage. What partial structures of authority still remained would be swept away entirely by the Russian civil war, whose bloody partisan battles everywhere spilled over into China’s northern borderlands.

Both developments are on remarkable and candid display in a telegram issued by Governor Yang during the dark days of November 1919. The catalyst was the sudden appearance of White partisans in the vicinity of Kashgar, heretofore a relatively quiescent front in the civil war. “I see that these Russian troops are equipped with machine guns,” Yang observed. “Yet my own government has not only never issued a single machine gun for the defense of Xinjiang, but likewise has never even provided a single gun or bullet for the most fundamental of tasks.” Reminding Beijing of his innumerable requests for weapons that had gone unanswered over the past two years, the governor then questioned the patriotic credentials of his own central government. “I do not know whose land you think Xinjiang belongs to, nor whose government rules over its people. Every time I reflect on this matter, it pains me deeply.” Yang addressed this telegram to then president Xu Shichang, who, like Yang, had served under Yuan Shikai. “Back when President Yuan was still in office, you were then state councilor. In those days, we all worked in tandem to address the affairs of Xinjiang, like the pulse of a body in synchronized rhythm. You were not yet afflicted by your current state of apathy and indifference.” Yang signed off with a plea to look beyond warlord politics. “I beg you to look upon Xinjiang as part of the Republic of China and under the jurisdiction of its central government. Do not look upon Xinjiang as simply the domain of one man, Yang Zengxin, and thus beyond your purview.”11

But that was exactly what it had become. Moreover, it could not be said that the governor was entirely displeased with this state of affairs. After all, it was precisely this volatile political environment that allowed Yang to spin an elaborate web of isolation about his province. At every step of the way, Yang justified his retreat from the outside world in terms of the “difference” he—and he alone—knew best how to manage in Xinjiang. Already in 1917, Xie Bin reported that Yang maintained regular communication almost exclusively with the Ministries of Foreign Affairs and Finance, only rarely corresponding with the Office of the President.¹² For Yang, the world outside Xinjiang offered nothing but sabotage for the “sanctuary” of which he now saw himself as the sole and wisest steward. “In Kashgar alone,” Yang warned Commander Ma Fuxing in late 1919, “we must be on guard against the British and Russian consuls, Japanese intelligence agents, and investigative teams from China’s Military Advisory and Border Defense departments. Every little thing we do is being monitored by foreigners and Chinese alike.”13 In 1920, China’s postmaster-general complained to the governor that “inspection procedures in Xinjiang have become quite unnecessary. Some items, such as newspapers, have not seen a single copy delivered in three years.”¹⁴

Not content merely to censor all forms of media, Yang also began to thoroughly vet all officials who ventured outside Xinjiang in the course of their duties. In 1927, Owen Lattimore crossed paths with several such men, including one who had served as the governor’s parliamentary representative in Beijing for some time. “He was now returning to Urumchi, to take up a new post as Chief of Police,” Lattimore learned. “But so strict is the control that the old Governor lays on all subordinates that he had been delayed for days at Chuguchak [Tacheng], while his application to reënter the province and proceed to the capital was being considered; in other words, while other equally confidential agents of the Governor were ascertaining that he had not returned from Peking tainted with unsuitable ideas, or contaminated by extra-provincial affiliations.” In addressing those rivals whom Yang imagined intent on recruiting such intermediaries, the governor did his best to advertise his province in the grimmest of hues. “Some ignorant people like to wax poetic about how lush and fertile Xinjiang is,” the Mongol and Muslim-dominated Provincial Assembly once wrote on Yang’s behalf. “Perhaps they do not know how emaciated our oxen are, how their skin clings to their bones, and how even this meager ration is fought over both tooth and nail. Perhaps they do not know how catastrophic is the state of our finances.”¹⁵ In the event anyone remained in doubt as to the identity of these so-called ignorant people, Yang was elsewhere more explicit. “If [Xinjiang] were to fall into the hands of some eminent person from the inner provinces, he would find that it is not possible to maintain even a single brigade of troops here.”¹⁶

By 1921, it would not be an exaggeration to say that Yang stood among the ashes of Russian and Chinese imperial authority in Central Asia. And since no one in either metropole had yet demonstrated the ability to put the pieces of the former empires back together again, Yang had little incentive to relax his control over provincial security, which remained tight. At a minimum, the governor could take solace in the fact that, during the past five years, he had managed to cull Xinjiang’s semiautonomous quartet down to one remaining member: Kashgar commander Ma Fuxing. On the other hand, Yang’s long-standing commitment to an ethno-elitist alliance composed of Muslim and Mongol nobles presented an institutional legacy of difference that foreign rivals could attempt to reorient in alignment with their own agendas. As it turned out, White partisans fleeing the battlefields of the Russian civil war were the first to try this. In 1919, Viktor Lyuba, the Russian consul in Ili, “printed several hundred pamphlets in the Muslim script and disseminated them all over the place,” according to a report delivered to the governor. “He is trying to incite the Muslims and Turbans so he can form a new army.” Not to be outdone, Red agents on the other side of the border adopted the same strategic approach, reportedly paying a group of Hui agents 300,000 rubles to return to Ili and inform the local Muslims that their day of “liberation” was at hand, and that they should rise up in response.17

These overtures to the non-Han peoples of Xinjiang posed a sort of threat qualitatively different from anything Yang had encountered before. He was, of course, intimately familiar with various incarnations of national determination platforms that had long circulated throughout the region. “At present,” Yang observed in 1920, “the issue that most requires our attention is this doctrine of Muslim independence.”¹⁸ Initially, however, Yang had less reason to worry about such doctrines in Xinjiang than he did those aimed at the Russian colonial presence across the border, whose hand fell much heavier on their Turkic subjects than did the Chinese on theirs. In fact, many of those who criticized Russian rule in Central Asia professed a sympathetic attitude toward Chinese rule in Xinjiang, which was viewed as more “hands-off” than that of the Russians.19 Thus, from Yang’s standpoint, the goal was not so much to suppress widespread Muslim discontent with Han rule in Xinjiang as it was simply to prevent nationalist agitations incubating elsewhere in the Muslim world from taking root in his province. “Calls to throw off the foreign yoke grow stronger day by day,” he wrote in 1920, assessing global politics. “Whether in China or elsewhere, everybody now harbors such thoughts.” To Yang, the implications of such trends were clear and disturbing. “The Turban people are gradually gaining more knowledge about the world. In the future, there is little doubt that those in Russian Turkestan … will drive out the Slavic race of Russians, and restore their own Khoqand Khanate.”²⁰

To prevent the Han race from being driven out of its non-Han lands, Yang did everything in his power to prevent nationalist propaganda from infiltrating Xinjiang. In 1918, when a British tract urging Arabs to throw off the yoke of Ottoman rule in the Middle East suddenly showed up in Xinjiang, Yang praised “its originally benevolent intent” in furthering the aims of one of China’s wartime allies. At the same time, however, he ordered any copies of the pamphlet that might surface in Xinjiang to be confiscated and destroyed. After all, he wrote, “the tract encourages people to extricate themselves from alien rule.” For Yang, there was nothing wrong with “the Arabs seeking independence from Turkish rule.” What he was afraid of, however, was that “Muslims in other countries will then be misled by their actions, and seek to overthrow the ruling government in their states.”²¹ Here we see Yang dealing with the double-edged sword of nation-building efforts harnessed for offensive purposes. In other words, Yang had finally come face-to-face with the propensity since the late nineteenth century for empires to create nations not on their own territory but “preferably on another empire’s territory.”²² Assuming such a fate did not befall the former Chinese empire, Yang was perfectly happy to watch it befall the Ottoman Empire.

So long as calls for national liberation in Xinjiang produced little more than a paper trail, Yang’s countermeasures were relatively straightforward and effective. When they entered on the backs of heavily armed White Russians, however, the governor faced a crisis of unprecedented proportions. In 1920–21, Yang’s worst fears came to pass. First came General Andrei Bakich (1878–1922), a Montenegrin Serb who made camp at Tacheng with eight thousand soldiers and another five thousand refugees in tow. One month later, Cossack Ataman Boris Annenkov (1889–1927) followed him into the Ili valley, still commanding a force of about one thousand battle-tested troops. Through equal parts diplomacy, subterfuge, and military coercion, Yang managed to imprison Annenkov and his men in the desert wastes of Dunhuang, where they proceeded to deface ancient murals in the Thousand-Buddha Caves. Bakich, however, heartened by the arrival of one General Novikov and his men in Tacheng, began to prepare for a counterattack on Red positions just over the border. In May 1921, the Reds delivered an ultimatum to Yang outlining the consequences in store for him should he continue to harbor an enemy combatant in Chinese territory. The governor, unable to call on Beijing’s support or confront a militarily superior Russian foe, informed the daoyin of Tacheng that “[I] must row my boat with the current.” Heaven, he declared, “has presented us with an opportunity to destroy the Whites by borrowing the strength of the Reds. We cannot go against the will of Heaven.”23

On May 24, the Reds, with Yang’s blessing, burst into Tacheng and opened fire on the Whites. Much to the governor’s dismay, nearly two-thirds of the White force managed to escape and flee toward Altay, putting the men in close proximity to Outer Mongolia, itself a major battleground in the Russian civil war. Here General Bakich first deployed the national idea via the barrel of a gun. “For too long now you have been wantonly abused and violently oppressed by your Han overlords,” he announced in July. “Now I have driven them out and taken control of Altay.” Bakich’s goal was join forces with Baron Roman von Ungern-Sternberg, who had just captured Urga, the capital of Outer Mongolia. Bakich outlined his preferred denouement in Outer Mongolia while calling for Mongol and Kazak assistance in Altay. “The Boghd Khan will soon lead forth an army of Mongols to help you loosen your shackles of oppression,” he promised, referring to the spiritual leader of Outer Mongolia. “Since all of you stand to benefit from such developments, I am requesting each and every person to provide cattle, sheep, and horses, as well as rice and grain rations.”²⁴

FIGURE 2.2. Annenkov and Bakich in Xinjiang, 1920–21. Cartography by Debbie Newell.

His calls appear to have fallen on deaf ears. Prince Ailin, one of the most important members of Yang’s ethno-elitist alliance in Xinjiang, remained loyal to the governor and promised to steer his Kazaks away from Bakich. And once Red armies invaded Urga and succeeded in capturing Ungern-Sternberg —over the futile protests of Beijing and other northern Chinese warlords, who had promised to drive him out themselves—Yang knew what he had to do. “The ancients teach us: cure poison with poison.”25 On September 12, representatives of the governor met with Red agents in Tacheng and negotiated new terms for a Red assault on Altay. A second attack flushed Bakich out of Altay, and he was captured in Outer Mongolia soon thereafter. The Bolsheviks then kept to the agreement they had signed with Governor Yang and withdrew their soldiers from Xinjiang. In contrast, they elected to remain in Outer Mongolia, where they cast themselves as stewards of an independent Mongol nation, free at last from Chinese “oppression.” Though this new Mongol “nation-state” posed a perennial geopolitical threat to Yang, the fact that his Han contemporaries in the heartland had failed where he had succeeded provided the governor with one of his most cherished platforms of strategic difference. “If we were to implement the rapacious policies of inner China in Xinjiang,” Yang warned in 1927, “the people would surely rebel, and Xinjiang would go the way of Outer Mongolia and Tibet.”26

Yet Yang still had to live with the “poison” of the Bolsheviks, who now surrounded his province in some form on three sides. Rumors regarding their tactics for soliciting political support, both among Russian Turks and Chinese expatriate Turks, had been trickling into Xinjiang for several years. “The strength of the Reds is growing fast in Tashkent,” came a February 1920 report from Kashgar. “Without regard to nationality or ethnicity, they are holding elections that are open to Hui, Turbans, Kazaks, and Russians.” An uprising in Bukhara emboldened a group of expatriate Turkic workers from Kashgar to declare that “the Turbans in Kashgar have been the victims of unbearable exploitation by Han officials, and that an armed battalion should be sent to drive the Chinese officials out of Kashgar so that they may enjoy the same freedoms” as did the Turkic peoples of Russian Turkestan. In Andijan, émigré Turkic laborers from Xinjiang set up a committee and “issued pronouncements concerning independence.” Though the situation across the border was far more complex than the governor could have imagined, for Yang, the bottom line was clear. Such slogans, he lamented, “cannot be suppressed by force, nor can we neutralize them with countermeasures.” By May 1920, Yang’s spies informed him that “formal autonomy has already been implemented in Turkestan. They have established a republic, with organized elections, and a seventy-four-member senate. All power rests with the Turkestan government.” The governor’s commentary on this report is telling. “It is frightening to see these Muslims and Turbans handling their own affairs of state during this time of chaos.”²⁷

Initially, Yang found that there was very little he could do about these alarming developments across the border. Mostly, he seems merely to have seized upon the threat of a national liberation movement among Turkic expatriates across the border to justify his own attempts to further consolidate his power in Xinjiang. In 1924, Yang learned that the fourth and only remaining member of the semiautonomous quartet, Ma Fuxing, the Kashgar commander, had initiated a suspicious correspondence with the Cao warlord faction in Beijing. The governor’s response was to march a Hui army on Kashgar and execute Ma.28 In justifying Ma’s death to the outside world, however, Yang said nothing about the commander’s shady dealings with other Han warlords. “He abused his power and deprived the people of their wealth and women,” announced the Xinjiang Provincial Assembly on Yang’s behalf. “Knowing no restraints, he styled himself a pasha and incurred the deep enmity of the people, who longed for his execution.”²⁹ But it was not just any “people” who longed for the commander’s death: it was “Chinese and Russian Turbans,” who had “gathered in Andijan, proclaiming their intent to lead an army on Kashgar and get rid of Ma Fuxing.” In other words, the threat of a national liberation movement from abroad—a pretext that only the Han governor of a non-Han land could invoke—had motivated Yang to kill one of his own officials. “Now that Ma Fuxing and his son have been eliminated,” he reasoned after the affair, in a telegram later printed in his Records, “the outrage of the Turbans has been appeased, and there will be no further cause for any incidents.”³⁰

It is, of course, possible that Yang truly believed the continued presence of Ma Fuxing in Kashgar would invite the unwelcome attentions of radicalized expatriates operating in what had once been Russian Turkestan. After all, consolidation of the new Soviet state in Central Asia certainly gave Yang much to reflect upon. “Ever since Russia reorganized its polity, it has become a federation of autonomous parts,” the governor observed. “There are now over twenty republics within this federation, which looks quite different from the Russia of the imperial era.” Within each republic, representatives of the majority ethnic group had been installed in prominent government offices, in an attempt to make each republic look as though it “belonged” to the majority of people who lived within its borders. “The governor of Semireche is a Kazak,” Yang explained to the daoyin of Altay. “Though still a part of Russian territory, officials in this ‘Kazakstan’ are drawn mainly from the Kazaks themselves. Similarly, they let Kazaks serve in their own army.” According to Yang, such a novel state of affairs was born not of compassion but of pragmatism. “It is not as though the Soviet government is reluctant to use military force to suppress their Russian Muslims and Kazaks. But the situation in which it finds itself makes it impossible to do certain things [qi shi you bu neng].” That is why, according to Yang, they allowed “the Russian Muslims in Andijan and Tashkent to set up an autonomous Muslim state, and the Russian Kazaks in Semipalatinsk and Zaysan to set up an autonomous Kazakstan.”31

What Governor Yang was attempting to describe has been the subject of much research by Soviet historians over the past two decades. One scholar characterizes the Soviet transformation of its non-Russian borderlands as “a strategy aimed at disarming nationalism by granting what were called the ‘forms’ of nationhood.” In other words, it was a preemptive strike against the ideal of national determination, an attempt to make nationalism work for the empire, not against it. In much the same way as a vaccine protects its host from a fatal disease by injecting a less potent version of the same disease into the bloodstream, Soviet policymakers similarly “midwifed” silhouette versions of independent nationhood, hoping to dissuade non-Russian elites from seeking national independence outside the umbrella of Soviet stewardship.³² Governor Yang knew exactly what the Soviets were up to, and he cared for it not one bit. “Bolshevik promises of independence for Muslims do not respect national borders,” the governor noted with alarm. “The effect in Xinjiang cannot be underestimated…. The independence of their Muslims will spark an unstoppable tide of radicalism, with a political crisis sure to follow.”³³

The problem for Governor Yang was essentially one of geography. “We have no choice but to maintain an active relationship with the Soviet Union,” he conceded in 1925. “But their policies of governance are not welcome in Xinjiang.” Soviet ideology, Yang wrote on another occasion, was “a lurking menace to the entire world. How can we expect Xinjiang, which everywhere shares a border [with the Soviet Union], to be able to resist the gradual infiltration of extremism?” Yet no matter how vigilant the governor may have been against Soviet affirmative action policies, he could not silence its most enthusiastic promoters. “The Kazaks of Russian Kazakstan will collude with Chinese Kazaks,” Yang wrote with resignation in 1924. “The Mongols of Russian Mongolia will collude with Chinese Mongols. The Muslims of Russian Turkestan will collude with Chinese Muslims.” And on and on. The governor, rarely one to admit defeat, finally conceded the brilliance of Soviet policy. “It is now a matter of Russian policy to make aggressive use of their Mongols, Kazaks, Hui, and Turbans [liyong Meng Ha Hui Chan wei zhuyi]. In response, Chinese policy must now turn to defending against Mongols, Kazaks, Hui, and Turbans. The Russian position, however, is far superior. All our counter-measures are destined to fail.”³⁴

FIGURE 2.3. Soviet national republics: the view from Xinjiang, c. 1924. By 1924, Governor Yang was surrounded by Soviet nationality projects. Responding to the new republics on his western, northern, and eastern borders, he predicted that “the Kazaks of Russian Kazakstan will collude with Chinese Kazaks,” the Mongols of “Russian Mongolia will collude with Chinese Mongols,” and the “Muslims of Russian Turkestan will collude with Chinese Muslims.” As a countermeasure, he bargained for the establishment of five Xinjiang consulates in Soviet Central Asia, intending to monitor the dissemination of Soviet affirmative action ideas among his own subjects. “If I let Xinjiang’s Turbans cross the Soviet border with impunity and without any restrictions whatsoever,” he wrote, “the ten thousand seasonal expatriate laborers of today will become the ten thousand agitating returnees of tomorrow.” Source: Yang, Buguozhai wendu sanbian, vol. 2, 15–16; and vol. 1, 33. Cartography by Debbie Newell.

Although Yang rarely provided specifics regarding the nature of the “collusion” (goujie) he envisioned, Owen Lattimore, who traveled throughout Xinjiang in 1927, was quite adept at pinpointing the source of the governor’s malaise. “The Qazaqs on the Russian side of the frontier,” Lattimore wrote, “are able to purchase modern arms, a liberty not allowed by the Chinese,” who confined most Chinese Kazaks—the Altay tribes, bordering Outer Mongolia, being an exception—to lightweight paramilitary duties. “Moreover, that part of the old province of Semirechensk adjoining the Ili territory is now the Autonomous Soviet Socialist Republic of Qazaqistan. The Russian Qazaqs therefore have a dash and initiative to which their kinsmen who are subject to the Chinese cannot aspire. Consequently the triumph of the Russian Qazaqs in all encounters of border thieves and cattle lifters has a galling effect of attrition on the Chinese prestige.”35 In 1924, reports streamed in to Urumchi that Mongols from the new Mongolian People’s Republic, under Comintern direction, were encouraging Xinjiang’s Mongols to take up arms and “unite” with their “motherland.” Yang read these reports with interest. “They are telling all Mongols under our jurisdiction to shed the shackles of abusive government and endless toil, and join the blissful path of development.” Before long, Yang found that the Mongols were expanding their irredentist claims beyond the Altay region, declaring, on the basis of historical demographics, that Ili, Tacheng, and even Urumchi should also “return” to the Mongol embrace. “Ever since the Outer Mongols began to receive Soviet financial and military backing,” Yang lamented, “their ambition knows no bounds.”36

Governor Yang could not similarly promise his Kazaks or Mongols their own “motherland,” nor had his career as a late imperial official taught him the virtues of sponsoring the development of mass nationalisms. “If ethnic boundaries are excessively distinguished from one another [zhongzu jiexian guoyu fenming],” Yang argued in 1920, “this will cause fish to be separated from water and birds from their flocks.”³⁷ Instead, all Yang could think to do was to reduce the tax burden of the nomads in question and broadcast stock Confucian platitudes. “We can no longer sanction the treatment of our Altay nomads as so much fish and meat,” Yang wrote to one of his officials. “Otherwise, their hearts will open up to foreign intervention, and our borders will never know a day of peace.”³⁸ Elsewhere, he addressed the threat of “self-government” more directly:

If we do not fix our own internal affairs first and effect a fundamental resolution, then it will prove impossible to make the Muslims and Turbans forever endure the rule of Han officials, and the compact among our peoples will break. The tide of calls for self-government rises day by day. It is not something that machine guns nor artillery can stop. If good government and noble instruction cannot be upheld, if virtue is absent and principles are not just, then all I can do is eradicate corrupt governance, punish greedy officials with extreme severity, strengthen the resolve of the people, and eliminate lurking threats.³⁹

For Yang and other Han officials educated during the imperial era, the very idea of national determination represented a perversion of the natural social order. Though all Chinese officials were exhorted to “love” or otherwise “cherish” the common people, there was to be no question of who was in charge. In 1913, Yang Zuanxu, a Han general sympathetic to the revolutionary movement, told the Muslims of Kashgar that “the age of ignorance has ended” and anyone could now aspire to become governor of the province.⁴⁰ In response, Yang Zengxin shut down the “new method” Muslim schools of Kashgar and arranged for Yang Zuanxu’s transfer out of the province. He was fortunate to leave Xinjiang with his life, a luxury not accorded some of his allies. The discourse of social mobility promised by such an ethnopopulist platform was simply not welcome in Governor Yang’s Xinjiang. From his perspective, the job of those in charge was to ensure that a pretext for political agitation—or even mere political engagement—did not percolate down to the commoners.

As Yang must have been painfully aware in the face of Soviet financial and moral support for national determination across the border, his actions came off as decidedly feeble. Though the governor regarded Soviet affirmative action institutions with strong distaste, even he was ready to acknowledge that they represented a “superior” (yousheng) adaptation to the storehouse of imperial repertoires with which he had been long familiar. Now that the Soviets were “making aggressive use” (liyong … wei zhuyi) of the non-Russian peoples of Central Asia, Yang needed to be equally proactive or risk the extinction of the conservative order he so fervently championed. He was not about to cast himself as a sponsor of mass ethnic nationalisms. What he was prepared to do, however, was search for a way of neutralizing the Soviet ethnopopulist “poison” at its source. To do so, he would first need to grab hold of one of the most potent institutions of Western political subterfuge in China—the consulate—and attempt to turn it back on its creators.

“THEY CALL THEM CONSULATES …”

The Bolsheviks were newcomers to Xinjiang, and it took them some time to discover that Ili and Tacheng were not cities in Outer Mongolia.41 On the other hand, Governor Yang, despite getting on in years, did everything he could to educate himself about socialism and its doctrines, even going so far as to attend weekly tutoring sessions with a Soviet consular aide. “Wherever there are poor people,” Yang reflected in early 1923, “this so-called doctrine of common property [junchan zhuyi] will find ready converts.”⁴² His marked distaste for Soviet economic and ethnic policies notwithstanding, Yang made it a point to attend to each and every diplomatic courtesy expected of an official in his position. “The days of the Russian empire are over,” Yang wrote in 1920. “In the realm of international diplomacy, we are starting anew on equal footing, leaving behind the aggressive policies of the autocratic era.”43 As such, the governor paid lip service to the new relationship by extending personal condolences to Foreign Commissar Chicherin upon “the early death of the great leader, genius, and scientist V. I. Lenin,” while making it a point to congratulate the Soviet consul in Urumchi upon the anniversary of the October Revolution.⁴⁴

For their part, the Soviets, while privately disparaging the Han governor of Xinjiang in their internal reports, were actually quite willing to engage Yang on terms to his liking. Perhaps most importantly for the governor, the Soviets promised not to propagate socialist ideologies within the borders of Xinjiang. Though it is still unclear precisely to what degree they managed to uphold this promise during the time Yang was in office, at least in principle, the Soviets do appear to have been committed to this end. That is, while the Turkic expatriates of Xinjiang could expect sympathetic treatment and perhaps even an “autonomous” Uighur political unit of their own while living within the boundaries of Soviet Central Asia, they could not expect Soviet sponsorship of a Uighur nationalist platform to accompany the return of expatriates to Xinjiang.⁴⁵ In other words, the moment Soviet affirmative action threatened to make the leap from domestic to international policy, Moscow’s geopolitical interests would ultimately determine whether that leap would be made. In the final analysis, pragmatism trumped ideology.

The problem for Yang, however, lay in his realization, as early as 1920, that “Bolshevik promises of independence for Muslims do not respect national borders.” After all, the very definition of expatriate is someone who resides outside the borders of their country of citizenship but retains cultural, social, economic, and sometimes political ties to their homeland. Thus, even if the Soviets refused to formally sanction the designs of some of Xinjiang’s Turkic expatriate community toward their homeland, they had little control over what these same Chinese subjects might do once they returned to Xinjiang filled with new ideas about nation and state. During his fourth expedition to Xinjiang in 1930, archaeologist Aurel Stein summarized his discussions about this problem with a Han official in Kashgar: “Difficulty from Soviet propaganda and how to meet it. Encouragement to be given to those who after visit to Farghana report true facts.”46 Yet the Soviets had nothing to hide: their institutions of affirmative action were intended to be publicly visible and rhetorically omnipresent. No one who left Xinjiang as a “Turban Head” could fail to notice that such ethnonyms were now regarded as racist and derogatory by Soviet intellectuals, and that, as newly christened Uighurs, they were entitled to some form of political autonomy such as that bestowed upon their ethnic brethren across the border. All in all, the number of Uighur expatriates who resided in Soviet Central Asia—variously estimated at anywhere from fifty to several hundred thousand—was considerable, and their potential collective political power certainly posed a threat to the governor of Xinjiang.⁴⁷

It should therefore come as no surprise that Yang’s next demand, right after his insistence that socialism not be propagated within Xinjiang, was that the Soviet state refrain from transmitting its ideologies among “Chinese subjects traveling in Russian territory.” To this the Bolsheviks could not agree. Characterizing Yang’s request as compelling the Soviets to “prohibit Chinese expatriates from reading the newspapers,” the Bolshevik representative responded with a simple rebuttal. “The newspaper offices in socialist countries propagate socialist ideas as a matter of course.” As their content involved merely “the guiding principles of science and socialism,” Yang had nothing to fear from any incendiary political content he may have imagined to exist within their pages. As for the governor’s specific request that the Soviets “prohibit Chinese citizens from joining various Russian organizations, where they will be exposed to speeches and publications, while also forbidding the sons of Chinese citizens from enrolling” in Soviet schools, Yang was roundly castigated on these points. The Soviet response was that to act thus “would be to deprive expatriate Chinese citizens of their rights and freedoms, including their right to receive an education while in Russia. Not only would this prove difficult to enforce, but it would also infringe upon the laws of our country.”⁴⁸

This was unacceptable to the governor. He knew what his Turkic subjects were exposed to while abroad, and he knew that it would be his responsibility to deal with them if and when they returned to Xinjiang. By then, however, it would already be too late, obliging Yang to resort to coercive measures against his indigenous subjects, the likes of which—incarceration and execution—he was not eager to associate with his isolated and vulnerable Han administration. What Yang needed was a forward base of operations in Soviet Central Asia to head off the threat before it crossed the border. He got his first chance in March 1920, when Bolshevik agents agreed to meet Han officials in secret at Khorgas Pass near Ili. The result, two months later, was the Ili Provisional Trade Accord, the first equal treaty to be signed by a Chinese official and a foreign government in over seventy years. The accord was, first and foremost, an acknowledgment of the extensive economic interests merchants on both sides of the border had in free passage, the obstruction of which during the long years of the civil war had meant famine and hardship for large portions of the population.49

Even more important from the governor’s perspective, however, was the stipulation that the Bolsheviks and the Chinese would each be permitted to install a “commercial representative” within the other’s border. For the Bolsheviks, who chose to establish their commercial office in Ili, this was merely a stepping-stone toward the recovery of all five former czarist consulates in Xinjiang. For Governor Yang, who chose to send Ili County magistrate Zhao Guoliang to Semipalatinsk, this was an unprecedented concession. Prior to Zhao’s appointment, the Xinjiang expatriate community across the border had been overseen by wealthy Turkic merchants and notable Muslim religious figures who had been formally recognized as aqsaqal by the governor in Urumchi.⁵⁰ In the face of ubiquitous Soviet propaganda, Yang could no longer trust these men. Now, however, for the first time, Yang was permitted to station abroad a permanent Han official who would send back regular reports about “commercial affairs” in Semipalatinsk.

Not surprisingly, Zhao Guoliang seems to have been far more interested in keeping tabs on the radicalization of Chinese citizens in Russian Turkestan than he was in facilitating the movement of trade caravans. In one of his very first reports to Urumchi, Zhao noted that the Soviets had already begun to make overtures to émigré laborers and merchants from Xinjiang, hoping to train them for clandestine missions back into Xinjiang.51 The governor’s suspicions appeared to be vindicated. “More than half of these men,” Yang wrote, referring to the thousands of itinerant Turkic laborers who crossed the border in search of seasonal work each year, “are involved in hard labor. They are an uneducated floating population, and are extremely susceptible to instigation by foreigners.”⁵² Convinced of expatriate radicalization abroad, Zhao’s first cable from Semipalatinsk advised Xinjiang’s local officials “to conduct thorough background checks on any future military conscripts, so as to avoid allowing any man who has performed seasonal hard labor abroad from filling our ranks.”⁵³

Yang was thrilled to have finally established a forward base in Soviet territory, modest though it was. From the Bolshevik perspective, however, their one commercial representative office in Ili was still a far cry from the generous political representation their czarist predecessors had once enjoyed. Thus it was only a matter of time before the Soviets lodged a formal request with the governor allowing them to reopen the five consulates of the imperial era. Yang surely expected this move, and likely knew that he could not hope to keep the Soviets at bay forever. And yet, in contrast to the troubled decades of the late Qing, this time Yang was determined to get something of substantive value in return for the concession of so powerful an institution. “They call them ‘consulates,’” observed Aksu daoyin Zhu Ruichi in 1926, “but in reality they are bases from which a monopoly on all commercial transactions is forcibly imposed.”⁵⁴ It is unlikely that Yang thought he would be able to impose his own economic monopoly on commercial transactions anywhere in Soviet territory. But if he could manage to set up a consulate or two of his own, he might just be able to impose a political monopoly on suspect émigrés who hoped to return to Xinjiang.

In the end, Yang reached for the stars: he insisted on a regime of absolute parity, proclaiming that the Soviets would get the keys to the five czarist consulates only if he was allowed five sister consulates in Soviet Central Asia, all managed by Han appointees from Urumchi. In acting thus, Yang was attempting to export to virgin territory the institutional embodiment of a foreign ideal: reciprocal diplomatic representation among ostensibly equal states. Though some late Qing statesmen had eagerly lobbied for the liberal establishment of consulates abroad and could point to a brief but encouraging precedent set by the first Qing officials to be stationed permanently in Korea, Yang’s initiative represented the most innovative and aggressive adaptation of all.55 The Soviets were clearly taken aback. “The five consulates that the old Russian government maintained in Xinjiang,” came the reply, “are not to be compared with the sudden establishment of [Chinese] consulates in Soviet territory, in a region that has never before hosted such consulates.”⁵⁶ Yang turned such logic right back on its head: Xinjiang, the governor pointed out, had likewise never hosted a single foreign consulate until the Russians had bullied their way into the province half a century earlier. Initiative had a way of forging its own precedent.

In October 1924, much to Yang’s delight, the Soviets relented, leading to the distinctly odd phenomenon of a Chinese province establishing and managing its own consulates entirely independent of the Chinese central government in Beijing.⁵⁷ Yang Zengxin touted his diplomatic victory as “an opportunity to look after our expatriate merchants and obtain compensation for their financial losses” during the long years of the Russian civil war.⁵⁸ No doubt there is some truth to this statement. Yet it is clear from the surviving documentary record that Yang was most intent on doing unto the Russians as they had long done unto him. If the Soviet consuls made it their business to extend the influence of their state into Xinjiang, then Yang would make it his business to export the most pressing items on his agenda into Soviet Central Asia. “If I let Xinjiang’s Turbans cross the Soviet border with impunity and without any restrictions whatsoever,” he concluded, “the ten thousand seasonal expatriate laborers of today will become the ten thousand agitating returnees of tomorrow.”⁵⁹

Alarmed at seeing the tools of imperial competition turned back against themselves, the Soviets hatched a plan to eliminate the Chinese expatriate community altogether, in order to deprive Xinjiang’s five consulates of their raison d’être. In July 1925, before any of Yang’s consulates had managed to open their doors, Xinjiang’s trade representative in Semipalatinsk informed the governor that the Soviets were forcing all Xinjiang émigrés to adopt Soviet citizenship unless they managed to obtain a consulate-issued passport within the next three months. At that time, the only consular posting even remotely prepared to issue such paperwork was the office of the former Semipalatinsk commercial representative, which also happened to claim jurisdiction over the smallest number of Xinjiang expatriates. Yang sensed an ulterior motive. “Our expatriate workers are mostly composed of Kazaks and Turbans, who share ethnic and religious ties with the people living all along the Soviet border,” Yang observed. “If the Soviets succeed in laying claim to our unregistered expatriates, they will be able to expand their network of socialist infiltration and avoid all claims for financial restitution by our aggrieved merchants.”60

By mobilizing the printing presses, eliminating the six-yuan application fee, and threatening to treat Xinjiang’s exiled White Russian population in similar fashion, Yang’s five consulates appear to have averted the wholesale elimination of their non-Han constituencies—both those on whose financial behalf they hoped to lobby and those who they hoped would remain vulnerable to Chinese surveillance and other political countermeasures, should the need arise.⁶¹ For our purposes here, what is most fascinating about Yang’s interactions with the post-czarist state is the way he adapted to Soviet innovations of the Russian imperial repertoire, carefully selecting only those tactics and institutions he deemed most suitable for deployment in Xinjiang and its immediate environs. For someone of Yang’s education and temperament, the Western consulate and its ability to facilitate political subterfuge abroad proved most useful in combating the institutions of Soviet affirmative action. Because Yang viewed himself as steward of the last bastion of conservative imperial authority in all of China, he found utility only in those repertoires that he could deploy to keep disruptive revolutionary innovations at bay.

With the collapse of the regime, however, future Han governors of Xinjiang would find far more to admire in the Soviet approach to empire.

THE CHINESE COLLAPSE IN XINJIANG

If the morning of July 7, 1928 was like any other for Governor Yang, then he most likely rose from bed at four o’clock in the morning, as was his routine. He probably ate his standard fare of plain tofu and cabbage for breakfast, meditated for an hour, perused a text on Daoism, and made his way to the governor’s office to tend to the day’s affairs of state.62 It was going to be a gorgeous summer day, complete with soaring July temperatures. On Yang’s desk was the usual pile of reports and telegrams requiring his attention. At some point in the morning he decided to reexamine a directive that he had written up five days prior, on July 2. The matter concerned a caravan of Turkic merchants from Xinjiang who, owing to bad weather and a host of other difficulties, had requested permission to make a detour from the usual border crossing at Nilka Pass and instead enter the Soviet Union by another route. For Yang, there could be no debate on this issue. “The appointed border crossing for trade transactions with the Soviet Union is at Nilka Pass,” Yang wrote, still determined to paper over widespread transgressions of the 1920 Ili Provisional Trade Accord. “We cannot make an exception to this long-standing precedent simply for the sake of a few merchants.” The governor had already sent this reply to his officials in Ili, but he had not yet forwarded his decision to the provincial commissioner of foreign affairs—a bureaucratic oversight he was about to correct. “Print and distribute this directive, and make sure your ministry follows up on this matter,” Yang wrote. “This is an order.”⁶³

It was the last order the governor would ever issue. After Yang completed the mundane paperwork of the morning, he retired to take his daily nap. That afternoon promised to offer a respite from the morning’s bureaucratic drudgery, for today graduation ceremonies were to take place at the Xinjiang Academy of Russian Law and Politics. After his nap, the governor likely saw to it that sufficient copies of his complete works, Records from the Studio of Rectification, had been prepared as gifts for that year’s graduating class, along with his favorite Daoist treatises, as was his habit. He then made his way several blocks to the academy. Pomp and circumstance proceeded without a hitch, and everyone took their seats. Lunch was about to be served. It was two o’clock in the afternoon.

“Suddenly we heard the sound of multiple gunshots,” remembered one witness. “Amid the ensuing chaos, someone shouted out, ‘The governor has been shot!’”64 Most of the guests thought they were hearing firecrackers set off in joyous celebration. But soon the truth emerged. Several waiters garbed in formal attire had approached the governor’s table with spirits in hand. Yang rose for the expected toast. Instead of opening their bottles, however, each waiter pulled out a revolver and fired repeatedly into the governor’s chest. He doubled over, gargled the words “What the—?!” (ganma), and collapsed onto the ground. Yang Zengxin was dead. He was sixty-four years old.

Once shrouded in mystery, the details surrounding the assassination of Governor Yang are now gradually coming to light. A growing body of scholarship in Chinese suggests the likelihood that Feng Yuxiang—the peripatetic “Christian warlord” in perennial search of a rearguard base from which to “save” the nation—somehow managed to orchestrate the dramatic events of the Triple Seven Coup, so named for its occurrence on the seventh day of the seventh month in the seventeenth year of the republic. How Feng’s agents could have plotted beneath the watchful eye of the governor—reputed to know of a snake’s fever in Hami—remains something of a mystery. Nevertheless, the broad outlines of the assassination are clear. The head of the academy where the shooting took place, Zhang Chunxi, apparently maintained covert connections with Feng. In 1927, with the establishment of a new Nationalist government under Chiang Kai-shek, warlords like Feng scrambled to shore up their defenses against the newly acknowledged central government in Nanjing. Feng needed an additional rearguard base from which to engage the Nationalists, and Xinjiang had long fit his geopolitical criteria for such a base.⁶⁵

By 1928, it had become widely known that Yang was about to retire. “I’ve worked my ‘stratagem of the empty citadel’ [kongcheng ji] in Xinjiang for seventeen years now,” Yang told a visitor in 1928, referring to an ancient strategy for deterring enemies through the art of illusion. “But it won’t work much longer.”66 The spectacular resurgence of the Nationalists in the south, helmed by conservative leaders hostile to socialist ideologies, initially appeared to be a most welcome development for the old governor. On March 17, four months before his passing, Yang told a visitor from the inner provinces that “if the southern armies were to take Beijing today, he would fly the [Nationalist] white sun and blue sky flag tomorrow.”⁶⁷ Just a few weeks earlier, Yang had sent three cables of support for Chiang Kai-shek’s new government at Nanjing. “The Nationalist revolution has now succeeded,” Yang wrote. “The good fortune of the rest of the country is also the good fortune of Xinjiang.” On July 5, just two days before the assassination, the Nationalists wrote back, expressing fulsome admiration for the preservation of peace in Xinjiang and optimism that the coming transition would proceed apace. Meanwhile, the governor’s designated envoy, Liu Wenlong, arrived in Nanjing and began to initiate concrete discussions regarding the handover.⁶⁸ On July 6, the governor let it be known that “he was very excited about national unification. As for his own exit, he said it would not be a problem.”⁶⁹

Unfortunately for Yang, that exit came the very next day, quite ahead of schedule. In a gesture of good will, the Russians provided free berth for his coffin on the Trans-Siberian Railway, destination Beijing. Today, Yang is buried in a handsome tomb in the northern suburbs of the national capital, bypassed unawares by millions of tourists every year en route to the Great Wall ticket booth at Badaling. More importantly, his death initiated a deadly chain reaction that would eventually tempt his successor, Jin Shuren, into a decisive breach of Yang’s ethno-elitist strategy of difference. The assassination represented much more than a simple change in leadership. Rather, it was a geopolitical crisis of the sort that is almost always responsible for repudiation of policies that privilege the institutionalization of ethnic and spatial difference. Though Yang had managed to insulate Xinjiang from rival Chinese warlords for a full twelve years following the death of President Yuan Shikai in 1916—a year generally acknowledged as marking the inauguration of warlord politics in China—the continuing inability of the warlords in the inner provinces to reconstitute a strong central authority almost guaranteed that sooner or later someone would manage to infiltrate Yang’s inner circle.

When the dust finally cleared, Jin Shuren, a career northwestern official whom Yang had taken under his wing during his tenure in Gansu, emerged as the new governor of Xinjiang. Since Jin was cut from the exact same geopolitical and ideological cloth as Yang—a former Qing official from Gansu—he, too, evinced little interest in implementing Soviet affirmative action policies in Xinjiang. So long as the land was at peace and isolated from outside disturbances, maintaining the status quo bequeathed to him by Yang was a viable strategy, as the latter’s seventeen-year tenure as governor had amply demonstrated. But that was no longer the case. Yang’s once formidable cocoon had been breached, and anyone who wished to avoid following him to his grave needed to find a way to raise an army capable of deterring warlords such as Feng Yuxiang from marching a full-fledged army on Xinjiang. As the Chinese archaeologist Huang Wenbi learned in 1928, just months before Yang’s assassination, there was no easy solution to this problem: “It is said that Yang harbors a mortal fear of Feng, and is terrified that he is plotting to come and take over his land. He wants to raise an army, but is afraid of the consequences that may bring. But if he doesn’t raise one, there is no way he can defend against an outside enemy. He knows only too well that his own military forces are insufficient to engage in battle.”70

To bring these forces up to par, the new governor had to extract more resources from the people of Xinjiang than Yang had ever dreamed possible or prudent. Jin embarked on a state modernization project, heavily weighted toward the military and its ancillary accoutrements. The problem faced by Governor Jin would be faced by the Chinese Communists some thirty years later: in the pursuit of wealth and power in Xinjiang, the politics of resource mobilization all too often end up resembling a Han-led nationalization project by default, if not always by design. In his understandable rush to extract capital from his subjects, Jin systematically repudiated the ethno-elitist alliance that had once undergirded his predecessor’s rule. Though not his original intent, such a result was scarcely avoidable. Had Yang Zengxin survived the Triple Seven Coup and remained in office to confront his warlord rivals head on, it was likely a blueprint he would have had to follow as well, though with considerable distaste and likely under a formidable rhetorical smokescreen. After all, in sparsely populated and fiscally anemic Xinjiang, bereft of outside support, there were only so many ways to build an army: tax the peasants, squeeze the wealthy, or initiate mass conscription. Jin pursued all three measures at once. As Yang had noted more than a decade before, the major handicap of a Han governor of Xinjiang during the republican era was the lack of a strong central government capable of financing the institutions of empire. “Since the establishment of the Republic of China,” Yang once wrote, “the disbursement of shared funds from the inner provinces to Xinjiang has been cut off entirely. We have not been issued a single penny with which to cover administrative and military costs.” As a result, Yang continued, “I have had little choice but to print unbacked paper currency simply to make ends meet. The dangers of such a practice are too great to speak of.”71

Jin would do this and more, extracting from the peasants in his province the long-absent fiscal subsidies once provided by Beijing. That the peasants of Xinjiang were Uighurs, not Han, only exacerbated the problem. As Yang Zengxin had been fond of pointing out, in the Han heartland, disgruntled peasants constituted little more than a domestic disturbance: either they were suppressed by their Han rulers or they replaced their Han rulers, but outside forces were unlikely to get involved. Discontent among non-Han peasants in the borderlands, however, would facilitate a form of political discourse largely unavailable to discontented Han within the pass. In other words, leaders of a Uighur uprising could advocate for the realization of exactly what Yang had warned against to justify his continued rule in Xinjiang: national liberation from Han rule. This could now be glossed—with more than a little encouragement by Soviet intellectuals—as “colonialist” or “imperialist” in nature.

When the British archaeologist Aurel Stein passed through the southern oases in the winter of 1930–31, he witnessed Jin’s creative exactions firsthand. After noting the “popular apprehension that a tax of 1M. is to be imposed on every tree cut,” Stein learned that “every sheep killed in honour of a guest is now taxed S. 3½!” Several days later, he talked at length with his old aqsaqal friend Ghulam Muhammad, who “tells of increased exactions of Chinese administration. Bribes extorted in a way which [Stein’s Chinese companion] Chiang could not have thought of.”72 In return, Jin had nothing to offer either by way of material compensation or ideological justification, such as an ethnopopulist discourse of development might have supplied. The Uighurs sat on agricultural wealth, and that was precisely what the governor targeted, in greater quantities than ever before. None of Jin’s predecessors had viewed the Uighurs as fit for military service, and precious few could be integrated into the Chinese bureaucracy. As Han peasants elsewhere in twentieth-century China knew only too well, the impoverished yet industrializing state offered its rural residents—Han or non-Han—but a single role: as producers of agricultural surplus for expensive modernization projects, in a process known as “primitive accumulation.”

Yet Jin did not stop at the Uighur peasantry. He also took aim at the Russians, who operated the only viable trade network in the region, albeit at the expense of the Chinese administration. In September 1928, the governor issued orders to deport illegal Russian traders in Tacheng and Altay, followed eight months later by the detention of Soviet merchants in Turfan, all accused of participating in illegal cotton transactions.⁷³“Recent directives from the provincial government have transgressed the long-standing policy of friendly relations between neighbors,” wrote the Soviet consul in Urumchi in late 1928. “As a result, the Soviet Union has been forced to scale back its commercial operations.” By “illegal,” of course, Jin meant any activity not permitted by the 1920 Ili Provisional Trade Accord drawn up by Yang Zengxin and the Bolsheviks, a document both Soviet and Chinese authorities had been winking at for years. The Russians called foul. “Today is nothing like the friendly policies of yesterday,” the consul lamented.⁷⁴ Though Jin knew full well the risks of alienating the Soviets, he deemed the bark from the west to be of far less consequence than the bite from the east. In the end, the governor got his fair share of the merchant cut but paid for it with far more ill will than the Russians alone could muster. After all, it was the Mongol and Turkic entrepreneurs of Xinjiang who composed the chief clientele of the Russians, and they suffered the most when the Russian market contracted.

FIGURE 2.4. Jin Shuren, governor of Xinjiang, 1928–33. Following the assassination of Governor Yang in 1928, Jin’s attempt to revisit some of the integrationist and extractive policies of the late Qing state without the support of the central government led to a devastating collapse of the ethno-elitist order. The political vacuum produced by Jin’s ouster in 1933 was filled by a host of belligerents peddling nationalist platforms of every sort. Sven Hedin Foundation collection, Museum of Ethnography, Stockholm.

Last but not least, there was the Mongol and Turkic nobility, long spared the rod of reformation under Yang. Initially, Jin gave every impression of continuity with the policies of his predecessor. “I promise to observe the principles of our departed governor,” he informed Tsetsen Puntsag Gegeen, the Incarnate Lama of Karashahr and regent for the khan-wang of the Torguts, “implement a just and fair administration, and maintain the peace.”75 In return, the khans, princes, and other non-Han nobles of Xinjiang had backed Jin’s accession to the governorship in the wake of Yang’s death and stood beside him in a united front against the warlords of inner China.⁷⁶ Once he obtained the governor’s seal, however, Jin gave free rein to his reforming zeal. He imposed a personal monopoly on lambskin, long a bartering staple of Xinjiang’s nomads, thereby forcing the Lama of Karashahr to close several processing factories operating under his wing. By infringing upon the time-honored perquisites of the Mongol nobility, Jin alienated one of the most important members of Yang’s ethno-elitist alliance. Several years later, when Jin called upon the lama to supply Mongol cavalry for the eastern front, Tsetsen Puntsag Gegeen instead laid siege to the Chinese garrison at Karashahr, a move that ultimately cost him his head.⁷⁷

The fiasco involving the Incarnate Lama of Karashahr just four years into Jin’s reign is a powerful testament to the swift deterioration of the ethnopolitical compact that had once existed between Yang Zengxin and the non-Han elite of Xinjiang. There was, however, a clear purpose behind it all: the creation of a military force capable of ensuring that Jin did not join Yang in his grave. And in that, the new governor had largely achieved his goal. Less than five months after Yang’s death, Jin had expanded his standing army to a mind-boggling thirty thousand troops, more than three times what Yang had grudgingly maintained and far more than Xinjiang could support on its own. Just one year later, the army had again doubled in size, with 74 percent of the provincial budget allotted to the military. To facilitate movement of soldiers throughout the province, along with a reliable communication network, Jin poured money into road construction and telephony services, with several steamrollers imported from—and a handful of students exported to—Germany. Most importantly, Jin undertook to finance, in fits and starts, a pair of massive arms shipments from British India and the Soviet Union.78

Jin’s new conscripts were mostly Han of Gansu and Shaanxi stock, the two provinces providing the most direct path of migration into Xinjiang. Yang Zengxin had always deplored the admission of Han migrants into Xinjiang, and he did his best to seal off the border at Xingxing Gorge, just east of Hami. But times had changed. For much of Yang’s tenure, the most destabilizing battles among Chinese warlords had occurred far from Xinjiang’s borders, with contestants setting their sights on strategic Beijing or affluent Shanghai. But now, with Chiang Kai-shek’s Nationalist Party in command of the Yangzi delta and presiding over a new capital in Nanjing, some of the most vicious wars, fought far from the eyes of Western diplomats and donors, were occurring right on Xinjiang’s borders. And that meant a flood of refugees, impossible to stop no matter what the governor’s stance on Han migration might be. Jin tried to make the best of the situation, employing many tens of thousands of refugees in his newly bloated armies. But what to do with the rest, especially those trailed by women and children? Jin’s solution owed much to a fateful turn of events that he could not have anticipated. Ultimately, it would serve as the stage upon which the governor managed to alienate the entrenched interests of the Turkic nobility at Hami—the gateway from inner China to Xinjiang that no governor could afford to antagonize. When combined with the loss of his Mongol allies at Karashahr, this move would eventually lead to the end of Jin’s brief and ignominious tenure as governor.

“Found large party of Hajis en route via Chitral,” wrote Stein in late 1930, as he descended into Kashgar over the Pamirs. “Among them men from Kumul & Turfan. Report of Hami Wang’s recent death.”⁷⁹ The death of Shah Maqsut in March 1930 set off a concatenation of events that would ultimately end with a coup against Governor Jin in Urumchi, a briefly independent East Turkestan Republic in Khotan, and a Japanese-trained Han warlord as the new governor. The spark igniting the fuse was Jin’s decision to abolish the Hami khanate, reclaim the prince’s vast landholdings for the state, dispossess his heirs of their inheritance, and resettle some Gansu and Shaanxi migrants on the defunct khanate’s land. This bold move has often been misinterpreted by historians. To begin with, it is clear that the prince’s subjects widely chafed under his rule, twice taking up arms during the late Qing, the last time under Governor Yang’s watch. In addition to the words of Governor Yang on this point, we also have the testimony of two envoys from the Chinese central government, Xie Bin and Lin Jing, who visited Xinjiang in 1917 and 1919, respectively, and studied local conditions. Xie noted how the Turkic peasants of Hami, hearing of the abolition of the Muslim khanates in Shanshan and Turfan during the late Qing, “twice circled the city walls to protest the corvée obligations imposed upon them by the Muslim prince, and demanded to be placed under the jurisdiction of Han officials.” Xie thought it a “pity” that Yang refused to accede to their request. Two years later Lin Jing described much the same situation, adding further details about how Yang resisted the demands of both his own local Han officials and the Turkic peasants in Hami for the abolition of Shah Maqsut’s prerogatives.80 Yang not only refused to impinge on Shah Maqsut’s prerogatives, but even went so far as to aid him in brutal suppression campaigns, all the while continuing to claim the prince as a Muslim ally of the Chinese state. The Hami prince had his ethnicity and strategic geographic location to thank. Elsewhere in Xinjiang, Yang was perfectly happy to undertake the systematic dismantling of every other semblance of Han or Hui autonomy. The Tarbagatai councilor, the Altay minister, the Ili general, and the Kashgar commander had all been victims of Yang’s consolidation of power during the chaos of the Russian civil war and its immediate aftermath.

Each time Yang took out one of his provincial rivals, however, a spell of military tension and political uncertainty had always followed. Such periods of brief instability had proven manageable so long as the person deprived of his power remained geographically isolated from the southeastern passes connecting Xinjiang with warlords in the inner provinces. Shah Maqsut and his Hami khanate, however, straddled the border with Gansu, the sole gateway to inner China and its warlords after the loss of Outer Mongolia. Simply put, there was no way an invading army from the east could skirt around Hami and strike straight for Urumchi. Thus, if Hami was at peace, warlords from Gansu would be hard pressed to justify its occupation, but neither could they choose an alternate stepping-stone to target. If Hami was not at peace, however, it was an open invitation for aspiring warlords from the inner provinces to try their luck in Xinjiang, with Hami as a rhetorical and strategic foothold. Yang Zengxin, knowing this, opted to leave Shah Maqsut and his retainers in place, no matter how much trouble they brought him. Jin Shuren, pressed for funds and inundated by Han refugees, viewed the timely death of the prince as a convenient pretext to kill three birds with one stone. He could gain control of the khanate’s tax revenue, resettle destitute Han refugees, and appease the long-suffering Turkic peasants of Hami, who had never ceased to pine for the khanate’s departure.

There were problems of implementation, to be sure. Minor skirmishes resulted from the perception that legitimate Turkic landowners were being dispossessed of their land to make way for Han migrants. The surviving documentary record shows that these initial conflicts were easily redressed or suppressed and did not pose a serious challenge to Jin’s administration. Far more threatening, however, were the actions of those who had the most to lose from the abolition of the khanate. These men were Shah Maqsut’s longtime retainers. Much like the Incarnate Lama of the Torgut Mongols, who laid siege to the garrison at Karashahr in response to Jin’s intrusions into his lucrative lambskin trade, it was the most powerful members of the Hami prince’s inner court who took the lead in organizing armed resistance against the governor: Khoja Niyaz Haji, captain of the palace guard; Yolbars Khan, the palace ordabegi (major domo); and Beshir Wang, a claimant to the throne. According to Jin, these men, “realizing that land reform was not in their interests, spread malicious rumors among the people, saying that the government was going to give Turban wealth and women to the Han.”81 From 1931 to 1933, Khoja Niyaz and Yolbars solicited and received military assistance from Outer Mongolia, while variously allying or fighting with a Hui warlord from Gansu, a brash twenty-something-year-old named Ma Zhongying. They may even have colluded with insurgents as far afield as Karashahr, Khotan, and Kashgar, who needed no further inducement to take up arms against Jin’s relentless impositions.82

By the end of 1932, every corner of Xinjiang was aflame. “The bones of the dead lie heaped in piles, steaming blood is everywhere plastered, the earth is scorched dry, and no matter where you look, your heart aches,” wrote Ma Zhongying, observing some of his handiwork. “The celestial garden of refuge has become nothing more than the home of tearful gods and wailing spirits.”⁸³ The battlefield machinations of the various contestants during these years were decidedly byzantine, to say the least, and have been explored in much greater detail elsewhere.⁸⁴ More germane to the issue at hand is how the chief belligerents justified their claims to political authority in Xinjiang. Khoja Niyaz and Yolbars, both creatures of the non-Han nobility and its elite Han stewards, quickly mastered the discourse of the nation, invoking the prospect of a “republic” populated entirely by what even they referred to in Turkic as the chanto (“Turban”) peoples.⁸⁵ Even though they once represented some of the most important pillars in Yang Zengxin’s ethno-elitist alliance—Yolbars would speak highly of Yang for the rest of his life—they were now determined to forge new careers by borrowing from the discourse of mass ethnic nationalisms. In other words, they hoped to reclaim their former ethno-elitist class privileges under an ethnopopulist guise.

Insurgents in Khotan broadcast the shrillest pronouncements. “Henceforth, we will have no need to the employ the language or place names of outsiders,” declared one such call to arms. “The black and yellow filth [Hui and Han] have stained our land for far too long. We must cleanse ourselves of this filth, and drive out the yellow and black barbarians. Long live East Turkestan!” For his part, Ma Zhongying, a Hui, distributed leaflets urging Muslims and Kazaks to “rise up and cease being slaves of the Han.”⁸⁶ Such inflammatory rhetoric eventually reverberated all the way to Nanjing and Moscow, where the Chinese ambassador to the Soviet Union recorded his alarm. “The Muslim groups are all calling for national determination,” he wrote, “and saying they will not tolerate Han rule anymore.” In Nanjing, Zhang Fengjiu, the Xinjiang liaison to the Nationalist government, cast a wary eye over the new political discourse now saturating the non-Han lands of the northwest. “In recent years,” Zhang wrote in 1933, “the notion of a great independent Muslim unity has gained widespread currency throughout the northwestern regions.”87

By alienating the non-Han nobles and their peasant constituencies, Jin Shuren ensured the destruction of the old order bequeathed him by Yang Zengxin. Like Yang, Jin was a former Qing official steeped in the ideology of the Confucian classics, and words from his brush read much like those from Yang’s. Unlike Yang, however, Jin could not claim the slightest alignment of word and deed. Though he pursued a roughshod agenda of resource extraction and military modernization, Jin was not an intolerant racist, nor did he “hate” his Turkic or Mongol subjects, as is often alleged. It was simply that in Xinjiang the greatest concentration of pliable wealth happened to be in the hands of sedentary Turkic peasants, while the most glaring conservative political arrangements just happened to be draped around the shoulders of the Muslim and Mongol nobility. In a similar vein, the most plentiful and vulnerable prospects for mass military conscripts intimately familiar with modern warfare just happened to be destitute Han migrants from war-torn Gansu and Shaanxi. Unfortunately for Governor Jin, this particular constellation of initiatives lent itself all too easily to charges of a bulldozing Han assimilationist project. And that in turn facilitated the articulation of its mirror opposite: national platforms for the non-Han peoples of Xinjiang.

As the old imperial order everywhere disintegrated around Jin, replaced by a bewildering variety of nationalist platforms, it became clear that the eventual victor, whoever it might be, would no longer find it possible to rule Xinjiang as Yang and Jin once had. After the events of the early 1930s, the political imagination of any future ruler in Xinjiang would need to encompass tactics from an imperial repertoire already familiar with the appeasement of expectations for political autonomy from each and every putative “nation.” And in Xinjiang, that meant taking a page from the Soviet Union and its affirmative action regime, of which Yang and Jin had been so wary. Ultimately, if Xinjiang was not to separate from the Chinese state, its reconstruction would most likely fall to politicians born and bred in the inner provinces, men who did not feel ill at ease with the progressive rhetoric of revolution so characteristic of Han nationalist movements in the heartland. Such men, of course, would be denied recourse to a narrative of political legitimacy that stressed long-standing ties to, and intimate knowledge of, the northwest.

The battle for Xinjiang reached its apex in 1933. The pool of potential victors had shrunk to just three, all Han or Hui and all provincial outsiders: Sheng Shicai, a Japanese-trained Han general from the northeast; Ma Zhongying, the Hui warlord from neighboring Gansu; and the Nationalist government of Chiang Kai-shek and Wang Jingwei in distant Nanjing. In April, their efforts would gain an added impetus when a coup backed by White Russian soldiers in the capital drove Governor Jin Shuren out of the province. With that, the stage was set for a political showdown. Regardless of who prevailed, one thing was now certain: with Jin Shuren went any hope of reconstructing the old ethno-elitist order upon which two millennia of East Asian empires outside the Han heartland had once rested.