Introduction
Contemporary China rarely features in a discussion of gun violence. The country seems to be devoid of a gun culture, if not gun ownership among its civilian population. Contrary to outside inattention to a ‘Chinese gun problem’, however, is the Chinese state’s intense surveillance trained on any sign of manufacturing, ownership, trade, and transportation of guns by non-state entities and private persons.Footnote 1 Firearms-related violations do exist, and the state punishes violators severely. For the first eight months of 2023, for example, Chinese authorities claimed to have raided and destroyed 258 dens where caches of guns and explosives were found, while arresting a total of 14,000 offenders.Footnote 2 Although murders by guns are rarely reported and mass shootings almost unheard of, the large number of gun-related crimes, as indicated above, points to a legal code and political culture in which unauthorized possession of firearms is construed as a dire threat to public safety and state security.
Firearms in private hands do exist but are limited to a tiny fraction of the population, and are found primarily among members of commercial firing ranges scattered around the outskirts of a few major cities.Footnote 3 The attributes of a recreational hunting market and extremely low frequencies of gun violence seemingly suggest that post-revolutionary China has achieved an enviable level of public safety and state security domestically, rendering a discourse on violent guns superfluous. Nonetheless, such a discourse is well and alive. In addition to the state media’s indefatigable reporting on mass shootings in the United States, a specifically Chinese gun problem discourse also regularly chastises unlawful gun manufacturing in the country’s western and often ethnically heterogenous frontier regions. State media reports are at the forefront of this discourse, lending certain toponyms catchy nicknames such as heiqiang jidi (hometown of ghost guns). The state security apparatus occasionally fuels this practice of naming and shaming by providing sketchy and decontextualized evidence. The notoriety of being labelled as a hometown of ghost guns attracts national attention when quasi-historical narratives conflate violence, lawlessness, and national security threat with ethnic difference, geographical peripherality, and the frontier region’s pre-PRC (People’s Republic of China) history in its own state-making endeavours.
This article treats the morass of quasi-historical narratives around one so-called ‘hometown of ghost guns’ as the entry point to uncover a piece of faded history in modern China—how arms made and unmade the state-making history of the Muslims in frontier Qinghai province. Centring the desirability and mobility of firearms in the twentieth century, this article argues that the history of modern China can be read as a history of establishing a mono-ethnic control of arms. Furthermore, the article shows how propagating quasi-historical narratives and casting different historical actors in the roles of heroes and villains in state-building reveal an ethnocratic regime’s cultural means in regulating political inclusion and exclusion.
The heroization of one’s own ethnic kin to strengthen technological nationalism and the villainization of other ethnic people’s possession and pursuit of firearms as anti-nationalist form a symbiotic duo in promoting ethnocracy, defined by political geographer Oren Yiftachel as a regime structure that ‘facilitates the expansion, ethnicization and control of a dominant ethnic nation over contested territory and polity’.Footnote 4 As such, skewed historiography plays a role in satisfying a populist demand for ethnically exclusive control of ‘the principal means of coercion in a territory’ by dichotomizing armed historical actors into ethnonationalism’s heroes and its enemies.Footnote 5 With this secondary focus on narratology, this article also illuminates how historiography intersects with racial/ethnic hatred in a non-Western context.
As will be demonstrated in what follows, Hualong’s Muslims earned this special attention from the state even though it was not a particularly noticeable hotbed for gun-related violence that warranted the state’s discriminatory and disproportionate deployment of law enforcement. Rampant feuds involving gun use are arguably more frequently reported in parts of Guangdong, Guangxi, and Sichuan provinces. Hualong’s special threat to national peace lies in the state’s perception of its capacity to produce the means for organized violence on a broad social scale. Gun manufacturing in a remote, less bureaucratized, and multi-ethnic borderland—especially when undertaken by an ethnocultural group distinct from the national majority—can easily be construed by the state in resurrectionist terms. Although this capacity poses no real threat to the state’s integrity in contemporary China, its unchecked development can still place significant strain on state resources.
By contrast, guns had a much more prominent role in state-making and state-unmaking in an earlier era. The importance of guns in the creation of political power in the pre-weapons-of-mass-destruction age was underscored by figures such as Lenin, who in his State and Revolution defines the state as ‘special bodies of armed men’.Footnote 6 Mao Zedong, founder of the PRC state, also famously appreciated the value of the gun in giving birth to a state: ‘Political power grows out of the barrel of a gun.’Footnote 7 Befitting Charles Tilly’s erudite thesis of state-making as organized crime and, one may add, armed crime, in the 1930s, Sun Dianying (1889–1947), one of the most resourceful warlords in North China, traced his enterprise back to 15 rifles he acquired to arm his drug trafficking gang.Footnote 8
Unlike parts of Sun’s home region of North China, where control shifted at head-spinning speed in the early twentieth century, Hualong was situated within arguably the longest-surviving and most stable territorialized political regime of the period—the Qinghai Muslim state. Following the school of Charles Tilly, this article takes a capacious definition of state-making as ‘a long and violent struggle pitting the agents of state centralization against myriad local and regional opponents’.Footnote 9 In this formulation, a hierarchy of polities is not a given, as centralizing agents all vie to assert claims to statehood. The nominal authority of a self-styled central state is therefore contingent. It is a fluid and continuously negotiated configuration in which everything remains open to redefinition. As in the classic European context of the nineteenth century, ‘Monarchs, princes, lords, bishops, municipal oligarchs, and regional parliaments recurrently and violently confronted one another in a struggle for control over the means of administration and coercion.’Footnote 10
The Qinghai Muslim state, or, as some may argue, a Muslim-governed proto-state, fits this definition of statehood because, as this article demonstrates, it outlasted two nominally central regimes—the Qing empire in its twilight years and the Beiyang Republic—and preceded two later contenders for national authority: the Kuomintang (KMT) government proclaimed in Nanjing in 1928 and the PRC in 1949. Throughout this turbulent period, it remained consistently focused on building its own state capacity in industry, finance, trade, and access to arms.Footnote 11 Crucially, this article shows that Qinghai skilfully exploited the mobility and fluidity of arms enabled by a trans-imperial infrastructure. This infrastructure encompassed Western missionary networks, treaty ports and foreign concessions inherited from the late Qing, a revitalized maritime hajj route, Japanese imperial ambitions, and an expansionist Chinese nationalism struggling to gain a foothold in the former empire’s frontier regions. The contemporary vilification of Hualong as the ‘hometown of ghost guns’ must be understood as a historical residue of its real and fictionalized involvement in Qinghai’s early twentieth-century state-building enterprise.
‘Made in Hualong’—Villainizing the detractors of an ethnocratic state
Hualong County is a Muslim-majority territory on the northeastern edge of the Tibetan Plateau. According to the 2020 national census, the 1,058 square-mile district supports a total population of 200,474 people.Footnote 12 Muslims are in a sure majority of 67.7 per cent in Hualong, with the Chinese-speaking Hui taking up 61.2 per cent and another Muslim group, the Turkic Salars, making up another 6.5 per cent.Footnote 13 In 2005, a murder case in eastern China’s Hebei province involving the victim being shot in the head spurred a wave of investigative journalism to find out the source of the weapon. The Party-state’s mouthpiece ran a 45-minute prime-time TV story featuring 12 officials, including the highest public security chief of Qinghai province. The officials publicly testified to tracing the origin of the bloodied gun to villages in Hualong County.Footnote 14 Subsequently, various national news outlets flocked to this territory to report on its underground gun-making workshops, which the state security apparatus confirmed as ‘the hometown of the best-quality ghost guns’ in China.Footnote 15
There is very little doubt that Hualong produced numerous guns in the early 2000s and even earlier.Footnote 16 In a 2007 press conference the Ministry of Public Security named Hualong as a centre for the production, sale, and stockpiling of guns and explosives, while disclosing that during a three-month special campaign launched the previous year it confiscated about 178,000 ghost guns, 3,900 tons of explosives, and 4.75 million bullets.Footnote 17 Situated in the multicultural contact zone between Chinese agriculturalists and Tibetan and Mongol pastoralists, it should be pointed out that Hualong’s thriving underground arms production and trade in the 1990s and 2000s were in part attributable to this specific geographic and ethnographic milieu.
Tibetan and Mongol pastoralists in Qinghai had maintained a long-standing love for gun ownership in the modern era, as observed by Ma Hetian (1987–1962), a prominent university president and a notable figure in China’s frontier research in the 1940s.Footnote 18 In the late 1980s and 1990s, a revived craze for gun ownership was fuelled by the windfall profits of the gold rush and poaching of antelopes in the Tibetan Plateau. On 25 May 1989, 42 Chinese gold rushers perished in a snowstorm in western Qinghai’s Golmud. The news became a household sensation all over China because Golmud’s vice-mayor was involved in extorting private gold rushers.Footnote 19 Throughout the 1990s, annual arrivals of gold rushers numbered between 50,000 and 60,000 in the mountains of Hoh Xil. In time, a fast-growing international market for Shahtoosh, so-called ‘soft gold’, diverted the attention of some gold rushers, who went on to poach the endangered Tibetan antelopes for their precious wool.Footnote 20
Those clandestine ventures in a way necessitated the possession of firearms to guarantee business safety. Being at the centre of the firearms trade put Hualong at a unique advantage because it was a sparsely populated county removed from China’s heavily policed eastern cities and agricultural regions, not only by its semi-pastoralism, but also by distance and elevation.Footnote 21 More importantly, beginning in the mid-1980s, China embarked on updating its defence industry.Footnote 22 Many former armament manufacturers became a fiscal burden for the state and were asked to either merge with bigger and more solvent state-owned enterprises or transform into manufacturers of civilian products. Some arms manufacturers were simply left to disintegrate, allowing workers to sell inventories of gun components and other equipment as scrap metal on the private market.Footnote 23
Hualong’s underground gunmakers were a direct beneficiary of this industrial shakeup. In nearby Gansu, the Tianshui Minshan Machinery, a once renowned armaments manufacturer, saw its business transformed into a subsidiary function of a Beijing-based steel company, at one time betting its future on making motorcycles and refrigerators.Footnote 24 The phased-out production capacity included the 56 Type Automatic Rifles (an imitation of the Soviet AK 47) and the 64 Type Pistols, providing Hualong with crucial components and equipment.Footnote 25 Hualong itself was home to an ammonium nitrate plant that began producing dynamite in 1972, but was bankrupt by 1982.Footnote 26 It is no coincidence that Hualong’s guerrilla-style gun-making workshops emerged around this time. While the 56 Type Automatic Rifles were the weapon of choice for gold rushers and the Tibetan antelope poachers in those years, the electrification of the rural areas of Hualong also began around this time. It was not until 1984 that all 621 hamlets in the county were plugged into the grid, providing the gun workshops with much-needed power supply.Footnote 27
Lawless Muslims making military-grade rifles in subterranean workshops made the right kind of sensational headlines news outlets wanted to carry.Footnote 28 This state-sanctioned coverage of renegade Hualong gunmakers subsequently spawned a litany of reports pointing at Hualong’s historical role as the armament manufacturing centre of a bygone Muslim-controlled regime—the militarist Ma family regime that lasted the entire first half of the twentieth century. The state’s mouthpiece, China Central Television, interviewed one elderly suspect who self-identified as an ‘engineer’ who had worked in the Qinghai Muslim state’s arsenal.Footnote 29 Thus, the instant notoriety of Hualong was aided by the gravity of its alleged historical role in a Muslim state- building project, which suffers from a dearth of credible research due to its denigrated status as a ‘reactionary warlord government’ in PRC orthodoxy.Footnote 30 The opacity of Hualong’s pre-PRC history fans the public’s fascination with it by providing the imaginative space to cast Hualong as the ultimate enabler of Muslim power and prowess in the region.
Hualong’s gun manufacturing history, if proven true, would transform our understanding of state-making by one group of China’s Muslims, the level and depth of their technological development, and the multiple sources for China’s military modernity. However, exhaustive research by previous scholars and myself have failed to yield any evidence of such.Footnote 31 This study shows conclusive evidence to the contrary: that is, Hualong was, at most, home to an arms depot in the 1920s. The alleged gun manufacturing in Hualong did not go beyond repair shops manned by a number of low-skilled blacksmiths. The Chinese official source documents a total of 39 ironsmiths in Hualong in 1943.Footnote 32 Even if there is some ambiguity regarding those ironsmiths, that does not make them a convincing number of gunmakers given that there were 14,098 working age (15 to 64) males in the county in 1947.Footnote 33
One may argue that the specialized skills of gunmakers could have made them a mobile and concealed section integral to the ‘Huihui army’—the revered military forces controlled by Qinghai’s Muslim generals—thus making them invisible to Hualong’s population census.Footnote 34 However, both Chinese intelligence sponsored by the KMT government (1928–1949) and those who served at the nerve centre of the Muslim state noted that the entire Qinghai province, including its better developed capital city, Xining, lacked the structural enablers for a firearms industry.Footnote 35
A detailed study of contemporaneous firearms production in Sichuan identified three industries that were indispensable to the making of firearms in the 1930s and 1940s: coal-fired power stations, copper and steel smelters, and chemical plants capable of processing sulphur and producing ammonium nitrate for gunpowder.Footnote 36 Sichuan’s firearms industry benefitted from an inheritance of a local copper cash mint, which was opened in 1905 using expensive machinery imported from Great Britain and Germany. That mint was easily repurposed to mass produce bullets. In 1919, the province’s authorities procured two 500-kilowatt-per-hour generators, a 250-millimeter rolling mill, and a two-ton steam hammer from the United States. This equipment was installed in Chongqing in 1934, making the factory the largest electrified smelter and rolling mill in southwestern China. Chongqing’s coal mining also had a jumpstart when the British opened a machine-operated coal mine in 1904, becoming a Sino-British joint venture in 1908.Footnote 37
By contrast, Qinghai had none of this crucial industrial capacity. Lacking a local mint, the region had stopped using copper cash by 1931. Instead, it circulated monetary certificates (a kind of paper notes) backed by the gold bullion deposited at the Xining central bank.Footnote 38 For extra-regional trade, Xining mostly accepted silver coins minted in Gansu’s Lanzhou in exchange for its exports.Footnote 39 A rare source published in 1938 comprehensively examines Qinghai’s various economic sectors. Totalling 68 pages and including detailed statistics, the source, however, reveals that government-owned coal mines were still relying on coolies to transport coal.Footnote 40 While the territory also had iron ore, lead, and copper, all mining activities were strictly a pre-industrial affair. The investigator noted that while local ironsmiths were engaged in smelting and making steel, the small-scale operations were only capable of producing artisanal knives and daggers.Footnote 41
This outsider’s observation aligns with that made by Chen Bingyuan (1909–1991), a county magistrate in eastern Qinghai and one-time news chief for Ma Bufang (1903–1975), the paramount authority of Qinghai between 1933 and 1949. Chen notes that Ma Bufang opened eight different factories to supply his army and calvary in the late 1930s. Although gunpowder, dynamite, and bayonets were produced under the euphemized name of a ‘machine repairment institute’, the plant had no more than 19 electrified machine tools and a single welding machine.Footnote 42 This equipment was only sufficient to produce ‘woks, pans and knives for the army, in addition to repairing guns and military vehicles’. The other factories, which were unelectrified, supplied the army with uniforms and undertook repair work.Footnote 43
It bears mentioning that all eight military-affiliated factories were built in and around Xining, the capital city, leaving Hualong nearly 70 miles removed from these productive activities. In fact, pre-communist history reveals no evidence of gun-making practice in Hualong. The county only began making some firearms in 1959, when a commune-organized gunpowder factory and another sulphur processing plant were established. As mentioned earlier, a sizeable ammonium nitrate plant was operational between 1972 and 1982.Footnote 44 Such a plant was part of Mao’s ‘third front’ strategy amid the high tension that existed with the Soviet Union beginning in 1963.Footnote 45 Although it has been proven that the dramatic hypothesis that Hualong was the armaments manufacturing centre of the Muslim state is unfounded, the conclusion invites a bigger question—how then did the Muslim state acquire its firearms, which decided the life and death of any state-building project in the first half of twentieth-century China? The next section will address this question.
Guns of a Muslim state—Securing arms in the trans-imperial age
Travelling across northwestern China in 1947, the Shanghai-born American journalist and political scientist Arthur Doak Barnett had this to say about the Qinghai Muslim state: ‘Ma Pu-fang’s regime is one of the most efficient in China, and one of the most energetic.’ Observing that Ma Bufang was ‘the third member of a local family dynasty’, Barnett was most impressed by the military muscle that enabled the Mas’ control of that vast territory for three generations:
At present, General Ma Pu-fang’s troops are among the best soldiers in the country. The bivouacs and camps of these troops are spotless, and the soldiers themselves are well dressed and disciplined. One sees these troops throughout the eastern districts of the province, and they help to create a general atmosphere that is very martial.Footnote 46
To Barnett’s observation about the efficiency and energy of the regime, one may also add the even rarer attributes of longevity and stability. The Qinghai Muslim state was the longest-lasting and most stable political entity after the fall of the Manchu Qing court in 1911 and before the inauguration of Mao’s PRC state in 1949, an extraordinary feat in comparison to all other China-based political regimes, both on the frontier and in the centre, that rose and fell during those decades.
The rise of Muslim power in Qinghai must be traced back to Ma Bufang’s grandfather Ma Haiyan (1837–1900) and his fin-de-siècle death in the act of saving the Manchu imperial family from the retribution of the Western Eight-Nation Alliance, who lashed out against the Qing following the violent anti-foreign Boxer Rebellion (1899–1901). As a reward for his bravery and loyalty, the Qing court appointed his son, Ma Qi (1869–1931), to succeed Ma Haiyan’s military position and settle their troops in the semi-nomadic district of Xunhua near Xining.Footnote 47 The Ma family’s military power in the region began assuming a monopolistic character in 1912, when the Beiyang central government appointed Ma Qi the garrison commander of Xining. Staffing his officer ranks with members of a reformist sect of Islam, the forces became the family dynasty’s private army, earning the sobriquet the ‘Huihui Army’.Footnote 48
In 1929, a year after the northbound KMT forces defeated the Beiyang central government, Ma Qi wrested control of Xining and the vast Tibetan and Mongol territories to its west from Sun Lianzhong (1893–1990), the Han appointee of the newly proclaimed Qinghai province, and installed himself as the government chairman. After his death in 1931, power was transferred to his younger brother, Ma Lin (1873–1945), who in turn retired due to illness five years later and put his nephew and Ma Qi’s son, Ma Bufang, in charge in 1936 amid a grand inauguration-ceremony-cum-military-parade.Footnote 49
The ascendant family fortune and uninterrupted succession of power is worth noting because of the highly volatile geopolitical situation in all other legacy regions of the Qing empire until the coming into being of the PRC state. While the Ma family was energetically building a Muslim state in Qinghai, one is reminded that during the same period, Beijing saw the abdication of the last Manchu emperor in 1911 and the highest office of the central government changing hands 15 times from 1912 to 1928, only to be replaced entirely by a newly rising centralizing state from the far south. While the Nanjing-based KMT had a good run of ten years in China’s affluent seaboard provinces, by 1938, it had to abandon its capital to Japan’s imperial army and found its territory reduced to a few southwestern provinces. Even the famed Chinese communists led by Mao were faring badly in those years, only able to credibly claim a territoriality for its renegade state as late as 1940 in the arid Shaan-Gan-Ning border region, five years after the Long Marchers found their uneasy terminus in Yan’an.Footnote 50
As a ‘trans-frontier’ polity without unfettered overland access to any strong foreign ally, the longevity and stability of Qinghai’s Muslim state is truly puzzling, especially with regard to the issue of securing arms.Footnote 51 Among the four Qing legacy frontiers of Mongolia, Xinjiang, Tibet, and Qinghai, Qinghai’s disadvantage is stark, due as much to geographic as political reasons. It was locked out of easy overland access to British India via Tibet, which was a direct competitor over the territorial control of Amdo, the ethnographic Tibetan region constituting the majority of Qinghai’s land area. The Muslim state’s staunch anti-communism meant that it could neither access direct Soviet help nor benefit from any of the Soviet proxies to its north and northwest—the Chinese communist base in Shaanbei, the Mongolian People’s Republic further north, and the Xinjiang-based Sheng Shicai regime, which was another Soviet dependency.Footnote 52 In fact, the Muslim state went to war with all three better-positioned neighbours—Lhasa, Urumqi, and Yan’an—and had participated in attacking the Mongols both near and far.Footnote 53
Landlocked Qinghai, therefore, had to rest its military and economic security on lifelines to the east, building rapport with whoever was controlling China’s affluent seaboard provinces while strenuously maintaining some direct access to port cities to purchase foreign goods and especially firearms. The Ma family dynasty consistently showed political acumen by maintaining good relationships with claimants to central power in eastern China. Ma Qi, following the example of his father, tried to save the Manchu empire in its last days in 1911 by violently suppressing the Chinese Republican revolutionaries in nearby Gansu and Shaanxi, only to decide to attach himself as a regional powerholder in the Beiyang regime when the last Qing emperor abdicated.Footnote 54 After the KMT routed the Beiyang regime in 1928, Ma Qi quickly switched alliance to acknowledge Chiang Kai-shek’s Nanjing government and in turn received Nanjing’s endorsement of him as the government chairman of Qinghai.
New evidence has emerged to buttress the case that the Muslim state never reduced itself to the status of a client state of the Nanjing, and later Chongqing, regime, even as the Ma patriarchs enlisted under KMT auspices. The exact nature of that relationship is better characterized as a quid pro quo alliance, especially when it concerned core interests. In 1932, nominally at the behest of Nanjing, for example, Xining went to war with Lhasa in Yushu in southern Qinghai. Fully aware of Nanjing’s desire to exert military pressure on Lhasa from both Qinghai and Sichuan to force a political concession on Sino-Tibetan territorial disputes, the Ma patriarchs requested 2,000 rifles, 20,000 rounds of ammunition, funds of 40,000 silver yuan, and five wireless telegraphs from Nanjing in return for their attacks on Lhasa’s troops.Footnote 55
In 1936, at the height of Ma Bufang’s successful campaign against the Chinese communists who passed through northern Sichuan and southeastern Qinghai to seek aid from the ‘socialist motherland’ (that is, the Soviet Union), Ma Bufang made sure that he was not just fighting to gain Chiang Kai-shek’s goodwill.Footnote 56 A declassified document dated 20 October 1936 from the Republic of China (ROC) Defense Ministry shows that Chiang Kai-shek promised to deliver 100,000 rounds of bullets for machine guns and 10,000 bullets for regular rifles to Ma Bufang as a show of ‘extraordinary reward’ (yu ge jiangli).Footnote 57 A few months later, Ma Bufang boastfully telegraphed Chiang Kai-shek with the news that he had annihilated the communist forces who took to the Hexi Corridor in western Gansu to reach the Soviet Union via Xinjiang. Ma was forthcoming about his spoils as well—his forces had gained two wireless radios, several crates of medicine, and over 200 rifles, in addition to many valuable gun components.Footnote 58
Those examples could lead one to conclude that Xining’s ‘Huihui army’ was merely a mercenary for Nanjing. However, as William Haas’s study convincingly shows, the Ma patriarchs and their close allies in neighbouring Ningxia, another Hui Muslim-controlled region, were confident and ready to fight against Nanjing when the latter schemed to encroach on Xining’s territoriality.Footnote 59 In the famous ‘the battle of three Mas repelling Sun Dianying’ incident, the long-standing Muslim alliance worked effectively to shut off their territories to any armed intruders, even those acting at the behest of Chiang Kai-shek.Footnote 60 What gave the Muslim alliance confidence and sustained their autonomy from Chiang was not unfounded arrogance, but a hitherto unexplored dimension of Muslim state-building in northwestern China—their access to international firearms that were out of Chiang’s purview.
Simply put, the road from Qinghai to North China and its international port cities opened earlier and lasted longer than Chiang Kai-shek’s eight-year control of this region. Nanjing first assumed jurisdiction over North China in 1928, only to lose it in 1937 when Japan’s imperial army rolled over Beijing and incorporated much of North China into its various allied regimes. Trade and communication ties cultivated by the northwestern Muslim militarists, first with the Manchu court and then the Beiyang government, did not simply evaporate overnight after Chiang provincialized the former imperial capital. Nor could Chiang Kai-shek undo the international nature of port cities such as Tianjin and Qingdao. What has not been examined by previous research on Qinghai is its access to a vast network of international traders from as far afield as Germany and British India, and as near as the fast-expanding Japanese empire and its client states in Manchuria, Inner Mongolia, and after 1940, a collaborationist Nanjing.
Germany loomed especially large in the export of firearms to China in the 1930s. The British Crown’s commercial councillor at Shanghai ranked Germany the biggest exporter of manufactured goods to China in that period. With a market dominance of 24 per cent among all foreign nations from April 1935 through to March 1937, Germany outperformed the United States at 20 per cent, and Japan and the United Kingdom at 17 per cent each.Footnote 61 The report noted that firearms and military supplies accounted for 14 per cent of Germany’s China-bound goods during the same period. Most of the German goods, both civilian and military, were disembarked and locally produced in the port city of Qingdao, which imperial Germany’s largest diaspora in Asia developed as a colony and naval base for its East Asian Squadron from 1898 to 1914.Footnote 62
The northwestern Hui Muslims found their foothold in Qingdao in 1929, when Ma Fuxiang (1876–1932), the patriarch of the Muslim militarists in Ningxia and an uneasy ally of Ma Qi, became the first Chinese mayor of Qingdao, backed up by his troops which were occupying part of Shandong during and after the KMT’s Northern Expedition. In Qingdao, Ma Fuxiang supported a Muslim initiative to buy three Western-style properties adjacent to each other and fashioned them into a mosque complex that could accommodate a congregation of over a hundred Muslims.Footnote 63 Ma also purchased another property and outfitted a halal restaurant there. These facilities greatly facilitated the proliferation of Muslim traders in the city while increasing their contact with powerful foreign traders from Germany and Japan. Notably, Ma Fuxiang also built ties with the Japanese consul in Qingdao after he helped Japanese businesses to put down a four-month strike organized by Chinese workers in 1930.Footnote 64
The Muslim state in Qinghai did not have to open official shops in Qingdao to trade for German arms, however. The Germans brought their shop to Xining. Chinese sources have observed the eccentricity of the Muslim state as Xining hosted the German Divine Word Father Hieronymus Haberstroh (1893–1969), prefect of the largest Catholic diocese in northwestern China at the time, and allowed his church to freely proselytize. More intriguingly, Ma Bufang invited Haberstroh’s congregation to pray for his troops during Mass, seemingly in contradiction to Ma’s religious faith.Footnote 65 According to Ma’s news chief, Father Hieronymus Haberstroh in fact played a much bigger role in supporting the Xining regime. He was responsible for maintaining Ma Bufang’s wireless telegraphs, running a hospital for the rehabilitation of Ma’s injured soldiers, and securing gasoline, Zeiss binoculars, radio receptors, handguns, and other military supplies from German sources, in addition to operating Qinghai’s first winery.Footnote 66
This multi-talented German priest fits the profile of German trade representatives described in Sir Louis Beale’s report on German business success in 1930s China—that the German advantage was attributable to ‘The generous provision of expert technical men on the spot who assist in planning enterprises … and representatives who are not only active and numerically adequate in large cities, but penetrate up-country areas, and often reside there’.Footnote 67 German communications technology and networks also provided Xining with crucial updates on both domestic and international affairs. In the episode of Xining repelling Sun Dianying’s assignment to northwestern Qinghai, Father Hieronymus Haberstroh regularly updated the Ma patriarchs on how things were unfolding in Nanjing, Ningxia, as well as Inner Mongolia, where Sun’s troops were stationed.Footnote 68
Yet, Germany was a smaller presence in North China than Japan was, especially in the port city of Tianjin, which Xining could more easily reach to sell its most valuable goods—wool and alluvial gold—in exchange for loans and military supplies.Footnote 69 In 1918, Xining’s prefectural magistrate Gong Qinglin (?–?) was elected to the Beiyang House of Representatives. A year later, he served as the chief of the Department of Government Affairs under the pro-Japan Beiyang Prime Minister Jin Yunpeng (1877 –1951). Using Gong as his frontman, Ma Qi was able to trade Qinghai’s precious wool for 850 rifles, over 500 cavalry guns, and eight light machine guns from Japanese armament dealers based in Tianjin. His army was also trained to Japanese standards in those years.Footnote 70
Japan was mostly seeking Qinghai’s wool. A 1940 report sponsored by the Japanese military noted that Qinghai’s wool, especially a variant named sawalu (さわる, meaning touch), bested those from all other regions of China in terms of its quality and valuation.Footnote 71 It also noted the absence of industrial capability in Qinghai to realize sawalu’s full value, requiring it to be exported first via Tianjin and later the Yellow River port city of Baotou in Inner Mongolia.Footnote 72 As early as 1918, Ma Qi opened branch shops of his family-owned companies De Yi Heng and De Shun Chang in Tianjin. Between 1924 and 1927, for example, a staggering 29 million jin’s (or close to 16,000 metric tons) worth of Qinghai wool left Tianjin.Footnote 73
A wool-for-loans scheme signed in 1934 further helped stabilize Japan’s import of Qinghai wool. Records from Japan’s Ministry of Foreign Affairs show that in 1934 Minister Hiroki Hirota met with Ma Qi, who used an encrypted name ‘qingyazhai’ (J. Masa Kiyosai), in Guangzhou. A deal signed between the two parties allowed Xining to take loans from Japan to be repaid in kind with wool and alluvial gold; and establishing a joint venture was discussed.Footnote 74 Trade statistics show that although the outbreak of war in 1937 drastically reduced Japan’s foreign trade due to Tokyo’s trade embargos to protect war supplies, the import of wool jumped 77.3 per cent from 1936.Footnote 75 A 1938 survey of Qinghai’s economy also found that Qinghai’s biggest imports were from Japan.Footnote 76
A major part of Qinghai’s imports from Japan was firearms. A report by the Tianjin Garrison to the Ministry of Army in December 1936 asked the Ministry to reconsider the sale of 1,000 38 Type Infantry Rifles to Ma Bufang’s brother, Ma Buqing, citing an intelligence report from western Inner Mongolia charging that Japanese arms were used by Ma Buqing in skirmishes with pro-Japan Inner Mongols.Footnote 77 However, a few months later, in May 1937, the China Garrison Headquarters approved another sale, which included 5,000 32 Type cavalry bayonets, among other items, and directed the purchase deal to the army’s Mukden Arsenal.Footnote 78
The KMT took note some of those dealings. An intelligence report received by the KMT Central Military Commission on 31 October 1936 offers some clue as to how Xining’s arms procurement worked in Tianjin. The report says that Ma Bufang first openly fell out with a Mongol noble named Yaling beile in Qinghai, expelling the Mongol and confiscating his properties. The Mongol beile went to Tianjin and procured a gift of 2,000 rifles from Japanese traders there. Ma Bufang then rescinded his earlier decision and restored Yaling beile’s name and property. Yaling subsequently served as Ma Bufang’s middleman to reach a deal with Japanese arms dealers to purchase weapons. Another Mongol from Chahar province in Western Inner Mongolia, which formed a pro-Japan regime in 1935, then arranged the exchange in the British Concession in Tianjin and trucked the firearms to Chahar before sending them onto Qinghai. The same report also noted that Japan set up an intelligence unit in Suzhou in western Gansu. The unit’s agents donned Mongol attire to enter Qinghai and engaged in trade talks, before being escorted out of Qinghai by Ma’s troops.Footnote 79 Apparently, in those years Mongols were able to traverse much of North and Northwest China as trans-imperial agents. Having Khoshut Mongols as one of the indigenous groups certainly helped Ma Bufang to make use of the Mongol’s mobility.
Japan’s occupation of North China, instead of cutting off Xining from Japanese firearms and other goods, made trade easier. In order to tap into the Muslim heartland of Ningxia and Qinghai for both trade and to form political alliances, the South Manchurian Railway Company (Mantetsu) began making plans to extend the Beijing-Baotou Railway to Ningxia (today’s Yinchuan) in 1936.Footnote 80 In the same year, Mantetsu also sent a reconnaissance aircraft to photograph the Baotou-Ningxia region for two days, stopping to discover the saturation of Japanese goods in local markets.Footnote 81 The pilot, Yoshiro Hayama, simultaneously the intelligence chief for Suiyuan province, met and dined with Ma Hongkui (1892–1970), the second-generation Muslim paramount authority of Ningxia, and was pleased to hear Ma’s officials expressing a pro-Japan stance.Footnote 82 These preparations were not just blueprints. After Japan occupied western Inner Mongolia in the second half of 1937, it soon created a new centre for trade out of Baotou, which lies directly downstream from Xining on the Yellow River.Footnote 83
The Hui Muslims’ shared interest in trading with Japan was affected by Chiang Kai-shek’s declaration of war on Japan following the Marco Polo Incident, but only to an extent. On 3 March 1938, Gansu’s newly appointed governor, Zhu Shaoliang (1891–1963), reported to Chiang Kai-shek that Xining’s wool was being trafficked by ‘treacherous merchants’ to Baotou and Hohhot, apparently because Ma Bufang had decided not to deal directly with the Japanese traders in Baotou due to the unfolding war.Footnote 84 However, given that Ma’s military had monopolized the valuable wool trade, the ‘treacherous merchants’ were surely not acting against Xining’s interests. As Chiang Kai-shek’s crisis deepened, the pro-Japan regime in Nanjing also found in Ma Bufang a potential ally. The liaison between Xining and Nanjing was so alarming that on 7 July 1939, Dai Li (1897–1946), Chiang’s intelligence chief, informed Chiang that Nanjing’s Wang Jingwei (1883–1944) had sent a team via Inner Mongolia, then Ningxia, to reach Xining, bearing ‘valuable gifts’ and 400,000 silver yuan.Footnote 85 Wang’s agents stayed in Xining for a while before being escorted to Hualong, where they set up a wireless radio to communicate with Xining, Tianjin, and Beijing, then the nerve centre of Japan’s occupation forces in North China.
Beyond Japan-occupied Baotou and Wang Jingwei’s cell in Hualong, Xining was also able to reach Japanese overseas. The Ma patriarchs maintained a continuous streak of making the hajj. Ma Qi’s younger brother, Ma Lin, for example, took a break from his government chairmanship of Qinghai in 1936 and stayed in Mecca for half a year.Footnote 86 While Chen Bingyuan frames Ma Lin’s trip as resulting from a power struggle between Ma Lin and his nephew Ma Bufang, Ma Lin was actually on a diplomatic mission. Chinese Muslim hajj records show that Ma Lin assumed the title of the ‘Chinese emir’ while in Egypt and the Hijaz.Footnote 87 A study of the pilgrimages of Japanese Muslims shows that Ma Lin reached out to Japanese agents in Mecca.Footnote 88 Suzuki Tsuyoshi, a Muslim convert and ‘a political maneuverer among the Muslim peoples’, promised to follow up with Ma Lin again in Shanghai, although that meeting did not materialize due to the outbreak of the second Sino-Japanese War.Footnote 89
The hajj network apparently also played a direct part in Xining’s armaments procurement. According to Chen Bingyuan, Ma Bufang’s De Xing Hai company had a representative office in Calcutta that sourced British arms and other manufactured goods for his troops.Footnote 90 His family also bought a property in Sri Lanka’s Columbo, another stopover on the maritime route to Mecca. The Columbo property served as a warehouse for Xining to store British and Indian supplies he had acquired.Footnote 91 While it is possible that all China-based regimes in the first half of the twentieth century possessed some extraordinary ability to acquire foreign arms, Xining seemed to stand out in that it utilized the Muslim hajj route for such purpose. It is also no mean feat to associate with imperial Japan from such a landlocked homebase. While further evidence is needed to corroborate Chen’s claims, Japanese government sources from the time indicate that the hajj network’s role in sustaining Qinghai’s state-making enterprise has been grossly underestimated. As previously shown, Ma Qi met Foreign Minister Hiroki Hirota to secure Japanese loans before the war, whereas Ma Lin chose to reach out to Japanese agents in Mecca after the war broke out.Footnote 92
The story of Xining’s arms acquisition explains the remarkable success of Muslim state-building in Qinghai, not only in terms of the regime’s longevity but also how such longevity was achieved. The Ma patriarchs were adept at a sophisticated balancing act—using firearms sourced from imperial powers who were the declared enemies of the Chiang Kai-shek regime while strategically acknowledging and making use of Chiang’s authority. Secondly, Qinghai joins its peer frontier regimes such as Tibet, Xinjiang, and Inner Mongolia in an emphatic ethnocratic tendency as it almost exclusively used Muslims in its military force, especially the upper ranks, even as it accommodated other ethnic groups sharing the territory to arm themselves.
Throughout its existence, lasting over four decades, Xining allowed Mongol pastoralists, Tibetan monasteries, and Han agriculturalists to bear arms of their own, and even at times organized those groups into peacekeeping militias for their respective localities.Footnote 93 However, this tolerance for the right to own and use arms by ethnic peoples other than the dominant one was gradually replaced by the more hardline ethnocracy of the PRC state, especially in Qinghai. The next section will address how after 1949 the PRC state in Qinghai began disarming Qinghai’s minority population, especially the Muslims, while politicizing their arms-bearing history as anti-nationalist.
Practising Han Chinese ethnocracy and proscribing its detractors
The Qinghai Muslim state dissolved on 27 August 1949, when Ma Bufang, his brother Ma Buqing, and their respective families took flight to Chongqing, on board a plane arranged by Claire Lee Chennault (1893–1958), the famed American aviator who led the KMT’s air force during the Pacific War. Ma Bufang’s son Ma Jiyuan and another brother Ma Buluan repeated the journey four days later, after setting fire to the Huihui Army’s Xining arsenal, causing explosions that went on for three days. Eight days later, the Communist People’s Liberation Army led by Wang Zhen (1908–1993) arrived, accepting the surrender of over 2,000 officers and soldiers, together with 46 cannons, 570 machine guns, 1,137 rifles, and 1,192 cavalry horses.Footnote 94
However, remnants of Ma Bufang’s forces continued their fighting until 1953, making use of leftover inventories elsewhere and airdrops from the Taiwan-based KMT government.Footnote 95 The last considerable batch of non-communist-owned firearms appeared in Qinghai in the 1958 ‘Great Rebellion’ launched by Tibetans.Footnote 96 Subsequently, it could be said that the Chinese communists had finally eliminated significant armed resistance in the former Muslim state. Although one may speculate that here and there pastoralists and bandits might have tucked away some weapons, the Maoist regime had another strategy to make sure that all arms were under its ethnocratic control—it simply eliminated ethnic rebels either by execution or decades-long incarceration.
In 1981, 23 years after the fact, the Qinghai provincial government began to review and overturn the case of its ‘overzealous’ crackdown on ‘ethnic rebels’ and remnants of Ma Bufang’s forces, who were sentenced in 1958. Notably, the rehabilitation included those Ma-regime cadres whose ranks were below the township and platoon leaders and who remained incarcerated 在押的县团以下人员.Footnote 97 Although we do not have more detailed evidence about how the detractors of the Maoist ethnocratic state fared between the 1950s and the Reform Era, the rehabilitation of Ma Bufang’s former grassroots bureaucrats in the early 1980s indicates that the challenge and threat they could have posed to a new Han-led regime was liquidated by the new state’s disciplinary apparatuses. To put this into comparative perspective, a similar exoneration campaign in Ningxia overturned the cases of 17,779 wrongfully charged individuals involved in seven so-called ‘ethnic rebellions’ between 1950 and 1969. Those 17,779 affected individuals were drawn from 157 Hui-majority communes or 701 hamlets.Footnote 98
The situation in Qinghai confirms the larger trend of Maoist ethnocracy, as the top political leadership in Qinghai appointed by Beijing was predominantly Han, as in other ethnic regions. Yet, it was also a departure from the general policy of the PRC regime’s Leninist policy of titular nationality (minzu), a mechanism to co-opt local ethnic elites into the regime structure. In Qinghai, despite or because of the Hui Muslim’s political dominance over Qinghai for nearly half a century, the early PRC state only designated two Hui autonomous counties in Menyuan and Hualong, both sub-units of Han-dominated prefectures. It is entirely possible that the administrative structure of PRC-era Qinghai was shaped by local demographic realities. Yet the exclusion of Xining, a Hui Muslim demographic stronghold and former capital of the Muslim state, from any meaningful consideration of Muslim political influence is particularly striking. This decision, according to Susette Cooke, was ‘no doubt due to mistrust of residual Ma regime influence’.Footnote 99
Cooke’s reading aligns with Benno Weiner’s study of the early years in PRC state-building in Qinghai. It is clear that although on occasion top communist leaders implored their cadres to distinguish ‘good Hui Muslims’ from ‘bad ones’, the bureaucracy and the security forces disregarded such rhetoric, treating all Hui Muslims as a disgraced but still threatening political class while excluding Muslims from positions within the government and local militia.Footnote 100 In the same time period, Han cadres were allowed to carry guns for self-defence as well as to assist in the suppression of counter-revolutionaries.Footnote 101 Qinghai’s move to disarm the Hui was an explicitly ethnocratic and ethnopopulist move given the widespread hatred of the Han for the Hui and the brewing desire for a reset of what many Han viewed as an ‘inverted’ ethnopolitical hierarchy, through retributive violence if necessary.Footnote 102
The foreclosure of the Hui Muslims’ access to arms as early as 1950 is noteworthy compared with all other non-Han peoples who likewise were demographically significant in their ethnic homelands. The Tibetans, for example, maintained a 1,000-strong well-armed Bodyguard Regiment under the command of the fourteenth Dalai Lama until his flight to India in 1959.Footnote 103 The Inner Mongols had Ulanhu as the commander-in-chief of its regional security forces until he was purged and moved to Beijing in 1967.Footnote 104 In fact, the PRC’s founding document, Common Program of the Chinese People’s Political Consultative Conference, stated unequivocally that the non-Han peoples had the right to ‘organize their regional and local armed forces’, before that clause was eliminated by the amended constitution proclaimed in 1975.Footnote 105 Thus, it was not only a Han populist current that saw the new state as an instrument to realize Han dominance over the Hui Muslims in Qinghai, but also the state, by disarming the Hui while allowing the Han to control the means of violence in the early 1950s, that promoted Han ethnocratic rule in Qinghai just as much.Footnote 106
Yet, disarming the Muslims is only one means by which Han ethnocracy was maintained. A continuous politicization of ‘good’ Hui versus ‘bad’ Hui has also been at the core of ensuring Han authority in determining political (dis)enfranchisement. As Uradyn Bulag points out, once the ethnopolitical order was reset, with the Han at the top of the new ethnic hierarchy, any threat to undermine Han majoritarian chauvinism makes a non-Han person a ‘bad minority’.Footnote 107 This rings especially true when it comes to the Hui Muslims in Qinghai, who must be subject to continuous trials in the court of Han populist public opinion for their armed resistance against the Han-dominated socialist state in the 1950s, their unholy state-making project before 1949, and their engagement in gun-making from the mid-1980s onwards.
Conclusion
With a population of less than 130,000, or 0.012 per cent of the total Hui population, Hualong’s Hui Muslims have garnered a greater share of the national discourse on Hui than any other Hui society in contemporary China.Footnote 108 This is a peculiar fact because Hualong pales against any number of Hui population centres in terms of demographic size or historical significance. It was associated with Ma Bufang only because the younger Ma Bufang cultivated his military credentials there as a district commander from 1920 to 1931. But that thin lustre was no comparison to Xining’s commanding heights, or Linxia’s reputation as the ‘Mecca of Chinese Islam’, or even Ningxia (today’s Yinchuan), capital city of another twentieth-century Muslim state. Yet, Hualong’s name in Chinese national airtime and print media since the mid-1980s is unrivalled, owing to its role as a palimpsest recalling Hui militarism, Muslim resistance against a centralizing Chinese state, and a debunked Hui technological self-sufficiency in producing firearms, in the past or present.
It is understandable why the post-Maoist authoritarian regime would fear gun-making more than gun ownership or gun trafficking. After all, the Leninist party-state is obsessed with monopolizing the production of everything and anything by default. But the spectre of Hui Muslims’ technological self-sufficiency in the production of firearms raises more ethnonational anxiety than the entirely manageable public security threat a few underground gun-making workshops would warrant. As this article has shown, the Muslim state in Qinghai was not only more stable than its peers in the first half of the twentieth century, but it also enjoyed a longer life than all three Han-dominated centralizing states—the Beiyang Republic, the nationalist KMT government, and the communists in Yan’an—during the same period. This inverted ethnopolitical hierarchy in Qinghai also detracts from the Han’s self-reverential claim to be the most advanced culture among China’s so-called 56 nationalities.
That history of an ‘inverted’ ethnic hierarchy fuels the desire to continuously maintain the ‘corrective reset’, so to speak. Because technological self-sufficiency in defence production is central to the project of state-making and nation-building, safeguarding the reset requires maintaining a constant contradiction between firearms policy and discourse. The Han ethnocratic state simultaneously disarms any competitor group while keeping in existence the (phantasmal) threat of such a competitor group, at least in historical discourses. For the Hualong Hui Muslims, this means that they have to play history’s villains as a constant object of disarming and political re-education by the ethnocratic state, regardless of whether their gun-making workshops possess any real expertise or steady supplies.
This villainization of the Hualong Muslims bears out the ethnonational anxiety of the rising Han populism since the 1980s because, while all China-based state-builders engaged in acquiring arms and developing their capacity for self-sufficiency in the first half of twentieth century, their stories all merged in a master-narrative of Chinese technological nationalism.Footnote 109 Both Han warlord regimes and Han individual engineers are eulogized today for their endeavours to make guns in Chinese hands more Chinese and more lethal.Footnote 110 However, Hualong’s imagined technological sufficiency in firearms production does not belong with the narrative of technological nationalism, not because Hualong Hui Muslims averted the virulent current of ethnonationalism in general, but because they are imagined to be beholden to a different ethnonationalism.
Through an exhaustive investigation into how the Muslim state in Qinghai acquired its arms, this article clarifies two things. First, it demonstrates that Hualong’s technological sufficiency in supplying the Qinghai Muslim state with firearms is pure fiction. The discursive insinuations to the contrary by Chinese print and airtime media is a political move to ensure the monopolization of firearms by a domineering majority while also discursively keeping the disarmed Hui Muslim minority in the perpetual state of suspect and regulation. By associating this community’s contemporary violation of China’s gun control regulations with Xining’s historical state-making project, the state’s media narratives subject Hui Muslims in this frontier region to a national security gaze as well as the watchful eyes of majoritarian ethno-populists.
The article also contextualizes the remarkable success of the Qinghai Muslim state. Departing from previous scholarship that portrays the Muslim state as an insular regime struggling to keep at arm’s length from a centralizing Chinese state, this article demonstrates that Xining was adept in taking advantage of the mobility and fluidity of arms afforded by a trans-imperial infrastructure. That infrastructure included Western missionary networks, treaty ports and foreign concessions inherited from the late Qing, a revitalized maritime hajj route, Japanese imperialism, as well as an expansionist Chinese nationalism struggling to find a foothold in the former empire’s legacy frontiers. Neither a facsimile of Chinese revolutionary partisans who threw in their lot with big foreign powers wholesale, nor a manifestation of the non-Han people’s allegedly centrifugal penchant, Qinghai’s Muslim state-building shows how history’s contingencies created multiple and contending offspring of empire, with varying degrees of competitiveness in different respects.
Acknowledgements
For advice and comments, I am very thankful to Professor Miwa Shimada at Keio University, my two anonymous readers, and the editor of Modern Asian Studies. Special thanks to Professor Tomoyasu Iiyama and librarian Osamu Shimada at Waseda University for help in accessing some of the primary materials.
Competing interests
The author declares none.
Open access
