DEVELOPMENTALISM IN LATE QING CHINA, 1874–1911

Abstract

This article explores the changing historical referents that Qing officials used in arguing for the extension of direct governance to the empire's frontier regions from the 1870s to the 1900s. In the 1870s, Shen Baozhen and Zuo Zongtang made the case for a change of governance on Taiwan and in Xinjiang respectively by reference to past Chinese frontier management. However, in the first decade of the twentieth century Cen Chunxuan and Yao Xiguang both referred to the European past, and specifically the history of European colonialism, to argue for reform of frontier policy in Mongolia. I argue that this shift was a result of both the empire's altered political circumstances and a growing belief in the inevitability of an evolutionary fate which awaited nomadic peoples, who were destined to be colonized. Yet this was not a case of Chinese thinkers simply adopting European ideas and perspectives wholesale. The adoption of European historical referents was entangled with Han Chinese perceptions of Mongolian populations which had been carefully cultivated by the Manchu Qing state.

In the mid-nineteenth century the 200-year-old Qing empire operated administrative governance systems in its ‘frontier’ regions which were distinct from those deployed in the eighteen provinces of the former Ming empire. ‘Frontier’ (bianjiang 邊疆) is a relative term because it imposes a Han Chinese perspective on the Manchu Qing rulers who identified part of this area, Manchuria, as their homeland rather than as a distant frontier.¹ Nevertheless these regions, two-thirds of the territory of the current People's Republic, were not ruled through the typical system of administrative prefectures and counties (junxian 郡縣) employed in China proper. This was the system under which Han Chinese empires had ruled their territory since the Qin dynasty (221–206 bce). They were placed under the separate jurisdiction of the Lifan Yuan 理藩院, the Office for the Administration of Outlying Regions. Areas outside the junxian system were, from a Han Chinese perspective, not considered an integral part of the state. Government was light touch compared to China proper. Imperial residents were appointed to co-opt indigenous leaders, who managed their populations according to their own traditions.² Han Chinese immigration into frontier regions was heavily discouraged because officials felt that rowdy migrants would upset the indigenous populations, thus creating military conflict in strategically vital regions, and running up unnecessary expenses for the empire.³ Beginning in the early nineteenth century, this calculation began to change, as some, predominantly Han Chinese, officials felt that sending Han migrants to frontier regions would solve both the emerging crises of the empire's overpopulation and its chronic lack of funds. They argued that the junxian administrative system should follow migrants into the frontier regions.

This article suggests that Qing officials’ move from using Chinese to European historical examples to argue for the extension of the junxian system into the frontier from the 1870s to 1911 is indicative of a new sense of global evolutionary time. In this period, officials arguing in favour of reform of the frontiers all favoured a set of policies that Joseph Lawson has described as a form of ‘developmentalism’, which included a focus on ‘opening wastelands’ (kaipi huangdi 開闢荒地).⁴ In borrowing this language I am not making anachronistic links to post-Second World War economic theory, or to normative language deployed by imperial actors in the early twentieth century to justify imperial rule.⁵ Instead, I refer to calculations of political economy which proposed opening land up to agricultural production, or developing it, as the best means to fund its defence through taxation of its produce. Shen Baozhen 沈葆楨 (1820–79) and Zuo Zongtang 左宗棠 (1812–85) respectively argued in favour of developing Taiwan and Xinjiang by making them provinces in the 1870s and 1880s. In the 1900s, Cen Chunxuan 岑春煊 (1861–1933), and Yao Xiguang 姚錫光 (1806–?) advanced similar arguments in favour of extending the provincial system to Mongolia. While the concrete policy proposals of these men were broadly similar, their choices of historical referents in support of their arguments were not. Whereas Shen and Zuo compared their policies with past practice on Chinese frontiers, Cen and Yao turned to the European past, comparing Qing strategies in Mongolia with the management of European colonies.

The significance of this shift should not be placed in the context of a move from rejecting ‘traditional’ Chinese culture to embracing a universal – Western – modernity. Instead, as Charlotte Furth has pointed out, Chinese reformers could and did continue to draw on a multiplicity of indigenous sources to make their case for seemingly foreign-inspired reforms.⁶ The global historical time which Yao and Cen espoused was one which was as much, if not more, a product of Chinese understandings as it was of European sources. Not only was the understanding of Darwinian thought adopted by Yao and Cen heavily mediated by Yan Fu's 嚴復 (1854–1921) translations, which emphasized the inevitability of progress, but their application of this to Mongols was also mediated by earlier Qing dynasty constructions of Mongol identity. These mediations highlight the fact that the movement of ideas in this context is not best understood using Mary Louise Pratt's conception of a ‘contact zone’ which ‘emphasizes how subjects are constituted in their relations to each other’.⁷ The moment of contact – in this case, of Darwinian ideas – was not as significant as the continued circulation of these ideas within the Chinese context. A ‘history from between’, as outlined in the introduction to this special issue, opens up the possibility of exploring knowledge which is produced through exchange in an imperfectly connected world and which is equally, if not more, influenced by subsequent circulation within more localized spheres of exchange.

Mongolia was ‘between’ hinterland and homeland for the Qing dynasty, occupying a very different place in the Qing geostrategic outlook from that held by Taiwan and Xinjiang. Yao and Cen's proposals for frontier reform there were thus far more radical. When the Manchus had conquered Ming China to found the Qing empire in 1644, they had done so in alliance with Mongol federations. The Mongols had therefore been part of the ruling coalition and enjoyed privileges including local autonomy and their own quota of positions within Qing officialdom. When Yao and Cen proposed making Mongolia a province and encouraging Han migration there, they were effectively recommending that this alliance be abandoned, and that Mongol elites be stripped of their privileges and autonomy. This was part of a process that had started in the early nineteenth century, when Han Chinese officials, often censored from commenting on affairs in areas of Manchu conquest, began to take a proprietary attitude to the entire Qing empire, rather than just to the former Ming territories.⁸ Yao bluntly stated that nomadic herders were destined to be replaced by sedentary agriculturalists. By comparing their policies to European colonial practices, he and Cen were placing their reforms in what they perceived to be global historical time, a time when a state was either colonizer or colonized. In doing so, they suggested that their radical reforms were both essential and inevitable, as evolutionary progress and continued foreign aggression both appeared to them to be.

Yao and Cen's proposals emerged at an uncertain point in Chinese intellectuals’ ongoing efforts to come to terms with the apparent superiority of the West. Immediately after the Second Opium War, thinkers such as Feng Guifen 馮桂芬 (1809–74) and Prince Gong 恭 (Yixin 奕訢, 1833–98) suggested that European learning had Chinese origins.⁹ However, by the time of the May Fourth protests in 1919 (so-called because of the student demonstrations on that date regarding China's treatment in the treaty of Versailles), key intellectuals in the New Culture Movement, such as Chen Duxiu 陳獨秀 (1879–1942), argued that China was not bound by its cultural and institutional past and could simply construct a Western future irrespective of it.¹⁰ The difference between the relative conservatism of the ‘Chinese origins’ thesis and the radical proposals of ‘May Fourth’ intellectuals has led to suggestions that the intellectual move from one viewpoint to the other in the sixty-year gap between them was simply a period of transitional uncertainty. Yan Fu's early writings are seen as representative of this ‘pervasive instability within modern Chinese cultural discourse’.¹¹ Yet, on the question of frontiers at least, the evolutionary model of change proposed by Yao Xiguang was radical and provided a degree of certainty. It suggested that, regardless of any cultural differences between China and the West, the fate of nomads on China's frontiers would inevitably follow that of nomads in other colonial situations. In this sense at least, China's future was determined by the European, and not a Chinese, past. The changing representation of Mongols was not purely European, however, but the product of ‘entanglement’ between Han Chinese and European perspectives.¹² Late Qing thinkers were able to present the Mongols as essentialized nomads as much because of the Qing construction of Mongol identity as nomadic as because of European ideas about the fate of nomadic peoples.

In order to explore the shift in temporal representations of Qing frontiers in arguments for their reform, I will begin by elaborating on the position of Han Chinese with regard to the frontiers in the early nineteenth century. Emerging Han discussions of the frontiers will be placed in the context of the principle forum for Qing policy debates, the palace memorial system, and the tradition of memorialists drawing on the past in making their arguments. I will then explore Shen's and Zuo's proposals for frontier reform in the 1870s and 1880s to highlight the repertoire of Chinese referents available to discuss frontier policy and the similarity of Shen's and Zuo's policies with those proposed for Mongolia in the 1900s. Finally, Cen and Yao's use of the European past is examined. I suggest that the European examples they deployed suited their arguments but that, more significantly, they were influenced by using evolutionary thinking as a lens to conceptualize historical time, projecting the inevitability of progressive change into the future.

I

Under the literary inquisition of the latter part of the Qianlong reign (1735–96 ce), Han officials were largely proscribed from discussing inner Asian frontier regions whose addition to the empire had been a product of Manchu conquests and alliances.¹³ Thus, their knowledge of these regions was gleaned from Ming dynasty works which described regions such as Mongolia as inherently foreign, or from the essentialized visions curated by the Qianlong emperor and his Manchu officials. As we will see, these visions appear to have played as much of a role in shaping attitudes to frontier regions and peoples as European exemplars would do. In the 1790s the ban on Han discussion of the frontiers and military affairs was largely lifted because of rebellions which required the direct involvement of Han officials in military affairs, creating a need for greater circulation of knowledge of military tactics and affairs of state.¹⁴ This led to a resurgence of statecraft (jingshi 經世) scholarship which emphasized the exhaustive use of agricultural land through increasing yields and developing new cash crops.¹⁵ In 1820, Gong Zizhen 龔自珍 (1792–1841), a prominent statecraft thinker, began to reconsider the Qing approach to the frontiers, and in 1820 set out ‘A proposal to establish the western regions as a province’ (西域置行省议 ‘Xiyu zhi xingsheng yi’), in which he suggested settling Han Chinese in Xinjiang to enlarge its tax base, fully exploit mineral reserves, and secure the frontier.¹⁶ While the proposal was never submitted, it attracted the attention of scholars when it was included in The imperial dynasty's collection of essays on statecraft (皇朝经世文编 Huangchao jingshi wenbian), compiled by Wei Yuan 魏源 (1794–1857) and He Changling 賀長齡 (1785–1848) in 1827.¹⁷ Gong's ideas were thus taken up by a subsequent generation of officials who, after 1860, were tasked with dealing with the twin threat of internal rebellions and external pressure.

Gong's successors continued these debates, most notably through the Gu Yanwu shrine association, which both emphasized the importance of the Han literati commenting on affairs of state and circulated knowledge of frontier regions.¹⁸ From the 1870s, officials took up their arguments for frontier reform through the palace memorial system. Palace memorials were initially secret documents, designed for the emperor's inner court to prevent the outer court bureaucrats of the Yongzheng emperor (r. 1723–735 ce) from limiting the emperor's autonomy.¹⁹ By the late Qing, however, these memorials had largely lost their secretive character, with their contents often being reprinted in the Capital Gazette.²⁰ Emperors would frequently quote from submitted memorials verbatim and reissue them asking for commentary from the empire's high officials. Thus, memorials were not just modes of correspondence with the central government, but were sites of important policy debates over, for example, the advisability of introducing telegraphs and railways to China, or whether to prioritize defence of the coast versus the north-western frontier.²¹ In this context, officials’ choices of historical referents were significant because they indicated not just their own perspectives, but also those which they felt might be convincing to wider officialdom. This is in part why Zuo Zongtang in the 1880s had to at least refer to Xinjiang's past, but Cen Chunxuan and Yao Xiguang in the 1900s could reject Mongolian historical precedents outright, after the defeats of 1895 and 1900.

Up to the late nineteenth century, palace memorialists made frequent references to the Chinese past when constructing their policy arguments. Sometimes these references were formulaic nods to heroics which had become idioms in their own right, such as officials’ frequent calls to ‘lie on brushwood and taste gall’ to deal with the challenges facing the empire from 1860 onwards.²² This is a reference to Yue Goujian, king of the state of Yue in the Spring and Autumn period (722–481 bce), who allegedly endured great hardships to ensure the eventual triumph of his state against its enemies.²³ Often officials referred to past policies which they saw as applicable to a contemporary policy debate. When the Qing general Zeng Guofan 曾國藩 (1811–72) wanted to argue against using foreign troops to put down the Taiping rebels outside Shanghai in the 1860s, he drew on a precedent from the Tang dynasty (618–907 ce). The Tang, he pointed out, had borrowed foreign Uyghur troops to help put down the An Lushan rebellion, only to be weakened afterwards by their constant demands for recompense.²⁴ Zeng's use of this history indicates that other officials would have been aware of this example and that he believed that they might find it convincing. It also suggests that, at this point, Zeng and at least some of his interlocutors did not differentiate the current foreign threat from that posed by other foreign groups throughout Chinese history.

This embedded culture of using the Chinese past was upset by defeat to the West, not in 1842 but after the Arrow War (1856–60).²⁵ After 1860, some officials, in a dramatic break with past practice, began seeking lessons for China in the European past, as well as that of China. In 1861, one Wei Muting 魏目庭, an imperial censor for the provinces of Hubei and Hunan, memorialized the emperor in favour of purchasing European ships and guns. He first drew on Chinese precedent, arguing that the Shunzhi and Kangxi emperors had relied on Jesuit missionaries to cast European cannon for them in their wars of conquest and that the Qianlong emperor had relied on foreign rifles to achieve his ‘ten complete military victories’.²⁶ He then turned to Europe, pointing out that Peter the Great had travelled to the Netherlands to learn about foreign naval tactics before returning home to construct his own navy. Now, Wei suggested, ‘Europe regards the Russian army as the strongest’.²⁷ Despite the novelty of his argument, there was a logic to Wei's choice of historical referents because his policy proposals referred directly to relations with European powers. Using European referents to discuss China's policy on its own frontiers was a far more radical step. Wei's argument was representative of a generation of ‘self-strengthening’ (ziqiang 自強) reformers after 1860 who drew a distinction between European techniques (yong 用) and Chinese substance (ti 體). These thinkers argued there was only a need to adopt European skills, while preserving Chinese society and institutions. It is in this context that the arguments of Shen and Zuo, both active in the post-1860 reform movement, should be read. Both men looked to the Chinese past to make the case for broadly the same reforms of frontier policy as those later proposed by Yao and Cen.

II

Shen Baozhen and Zuo Zongtang agreed that making Taiwan and Xinjiang provinces was necessitated by new circumstances, but, in doing so, they did not depart from established Chinese reformist thought. Up until the end of the Ming dynasty (1368–1644 ce) Taiwan was relatively untouched by the Han Chinese. Han Chinese colonization commenced rapidly from the 1630s after the Dutch East India company established a trading post and encouraged Chinese immigration because they required agricultural workers.²⁸ The Dutch were supplanted in 1662 by the Ming loyalist regime of Zheng Chenggong 鄭成功 (1624–62), which was itself displaced by Qing forces under General Shi Lang 施琅 (1621–96) in 1683. After the occupation, Qing policy was non-interventionist, focusing on quarantining Han settlers and ‘cooked savages’ (shufan 熟番) – that is, pacified aborigines – on the west coast, while leaving the east coast to the ‘raw’ savages (shengfan 生番) – aborigines who had not submitted to Qing rule.²⁹ This strategy collapsed with the Mudan incident of 1874, when nominally Japanese sailors were shipwrecked on the island's east coast and some of the survivors were murdered by Taiwanese aborigines.³⁰ The Meiji government demanded compensation, but the Qing court refused to pay on the grounds that the aborigines were not their subjects.³¹ Acting on the advice of an American lawyer, Charles Le Gendre, the Japanese took this as an admission that the east coast of Taiwan was unclaimed land and launched an expedition to the island to punish the aborigines and scout out its potential for colonization.³² Within six months of the arrival of the expedition, the Qing court agreed to pay an indemnity.

The incident led Qing officials to realize the need to strengthen their control over eastern Taiwan and prompted Shen Baozhen, who had handled the Qing response to the Japanese invasion, to call for the island to be made a province. Shen's policy recommendations foreshadowed the developmentalist arguments of Zuo, Cen, and Yao. He argued that, ‘once the thorns have been cut on a daily basis, settlements and production will rise and we will gradually be able to civilize (hua 化) the savages, and then we might be able to add their land to the common supply’.³³ Shen's references to civilizing the indigenous population such that they would effectively become Chinese illustrates his cultural, rather than racial, figuration of their difference from the Han. This distinction was long-standing in Chinese frontier thought: when Cen and Yao later discussed Mongolia, they did not acknowledge any possibility of indigenous transformation. Shen also warned of the European threat facing Taiwan, suggesting that, just as the British and French had made Singapore and Annam colonies, so they might turn to Taiwan. He pointed out that the Europeans’ ‘many tools are sufficient to make full use of Taiwan's land. What is today our so-called defensive barrier (outuo 歐脫) can on some other day become a metropolis; the roots [of this] are already deep.’³⁴

Although Shen was pointing out the threat of European colonialism and proposing a developmentalist response, his argument remained firmly rooted in Chinese historical referents. His use of outuo evoked a strategic concept which had been used in the Records of the Grand Historian (Shiji 史記, c. 94 bce) to indicate the vacant land lying between the Han empire and the nomadic Xiongnu federation which threatened its northern frontier.³⁵ This reference appears to have had two functions. First, it was a direct critique of past policy on Taiwan, because many officials had argued that Taiwan should serve as a defensive buffer or ‘hedgerow’ (fanli 藩篱) or ‘screen’ for China, but that this did not require the expense of developing the island.³⁶ Second, Shen was drawing attention to the shifting of frontiers over time since outuo referred specifically to territory which had been outside Chinese domains when Sima Qian was writing, but which by Shen's time was an established part of the Qing state. For Shen, empty defensive barriers would not and could not remain and, just as the northern frontiers were no longer vacant, Taiwan's east coast would eventually be colonized.

He was not alone in drawing attention to historical change to discuss new policy on Taiwan. Even an author critical of full-scale colonization projects recommended that Taiwanese aborigines who did not submit to some sort of Qing rule should be put to the sword because ‘now is really not their time!’ (jin qie fei qi shi ye 今且非其时也).³⁷ Yet, while this thinking emphasizes the policy similarities between reformers in Taiwan and later frontier thinking which would argue that Mongolian herders would, and should, inevitably decline, it is devoid of the latter's evolutionary influences. First, the same memorialist references the Zhuangzi, a classical Chinese source, to emphasize that aboriginal populations would ‘naturally, over a long time … gradually obey our beneficent sovereign’.³⁸ The end of the aborigines’ time was thus a reference to the belief in a natural process of acculturation to Han Chinese norms given the power of Han culture, rather than evidence of evolutionary processes which favoured some races over others. Second, the memorialist emphasized the need to ‘guide matters according to the circumstances’.³⁹ This principle was firmly rooted in Chinese reformist thinking, as Zuo Zongtang made explicit in his arguments in favour of making Xinjiang a province.

The conquest of Xinjiang was completed by 1760, after which time the region was incorporated into the Qing empire.⁴⁰ Despite this, Han Chinese remained relatively ignorant of the areas of Manchu conquest until after 1800. Unlike Taiwan, however, this was not the first instance of Han engagement with the region because both the Han (202 bce–220 ce) and Tang (618–907 ce) dynasties claimed protectorates over parts of it. In the eighteenth and nineteenth centuries, Han Chinese scholars writing geographies and travelogues on ‘the western regions’ uncovered Han Chinese toponyms for local place names to emphasize the long-standing connection of the Chinese to the area.⁴¹ This encouraged early nineteenth-century statecraft thinkers such as Gong Zizhen to call for greater incorporation of the region into the empire through promoting migration. Crucially, it also affected Qing policy decisions from 1864, when Yakub Beg (1820–77), an official within the khanate of Khoqand, took advantage of Muslim unrest in Shanxi province to seize control of much of Xinjiang.⁴² To make matters worse, Russia used the turmoil as a pretext to occupy the Yili region in 1871. While officials such as Li Hongzhang 李鴻章 (1823–1901) argued in favour of abandoning Xinjiang because, even were it recaptured, it could not be held against Russia for long, Zuo took up the arguments of earlier statecraft thinkers, in some cases directly restating Gong's work, and argued in favour of reconquest and the incorporation of Xinjiang as a province.⁴³ Between 1877 and 1882, during his reconquest of the region, Zuo submitted five memorials to the throne on the subject.⁴⁴

Zuo's thinking on Xinjiang developed significantly over the course of his reconquest campaign. He began his first memorial on the subject by emphasizing continuity with the past, pointing out that ‘when a state is founded it has frontiers, and in ancient times as in the present this has the same significance’.⁴⁵ He thus proposed cultivation through military colonies (tuntian 屯田), a tactic, and terminology, borrowed from the Han dynasty. His proposals drew heavily on his education at Changsha's Yuelu Academy, which championed statecraft thinking.⁴⁶ However, he later revised his initially environmentally determinist conception of Xinjiang, arguing in favour of developmentalist policies such as introducing irrigation to transform the landscape and to enable the introduction of sericulture.⁴⁷ This was not his only innovation: Zuo also inverted his past insistence on the similarities between the Han frontier and that of the Qing. His argument drew on the past precedents applied by his interlocutors only to reject them:

The Zuozhuan says ‘The son of heaven's purpose is to defend against the barbarians to the east.’ Prior to the Zhou and Qin dynasties, [we] did not have the ability to discuss this. Since the Han dynasty, the opening of the road [to the west] began with Zhang Qian, who was unable to obtain the support of the Yuezhi [against the Xiongnu], [and] though he requested horses for vast profit, he could not enter that small state's fortified cities. The Han policy towards the western territories exhausted the strength of heaven in order to pursue them, [but] suddenly faced regrets at Luntai. Therefore Ban Gu believed that obtaining these areas would bring no profit, and abandoning them would bring no harm! Now those who support abandoning Xinjiang refer to him. I have to say this is correct, [but] I turn to wrangling with those who say ‘the army cannot be stopped, the territory cannot be abandoned’, and, although the topography is roughly the same now as it was in ancient times, differences have been caused by differing standards. [We must] thoroughly act according to the situation, and rule appropriately according to the times.⁴⁸

Zuo rejected his peers’ use of Han dynasty precedent by placing it in a wider context. His initial reference to the Zuozhuan 左傳, a classical text believed to date to the fourth century bce, highlighted the importance of the emperor's role in pacifying the west and thus the classical roots of his own military campaign. He then stressed the passage of time to dispute the relevance of his opponents’ historical precedents. While the Zhou and Qin dynasties had not been able to advance into the west, progress had been made by the pioneering state agent Zhang Qian 張騫 (164–113 bce). Zhang had been tasked by the emperor Han Wudi with venturing to the kingdom of the Yuezhi in order to seek an alliance with them against the Xiongnu confederation of the Steppe. While Zhang's mission was ultimately unsuccessful, he brought back valuable information about the western regions which helped to shape Han strategy.⁴⁹ The Han were thus able to advance, until they suffered setbacks at Luntai, a present-day county in Bayingolin Mongol Autonomous Prefecture in Xinjiang. Zuo then agreed with his critics that as Sima Qian 司馬遷 (c. 145–c. 86 bce) – in the Records of the Grand Historian – and Ban Gu 班固(32–92 ce) – in the Hanshu – had argued, these events served to highlight the dangers of adventurism.⁵⁰ However, by placing this argument after his reflections on Han achievements when compared with those of the Qin and the Zhou, Zuo implied that reaching Luntai at all represented a relative advance. Finally, his appeal to ‘rule appropriately according to the times’ underlined his argument by appealing to a different classical source, the early legalist text known as the Shangjun shu 商君書, in which Lord Shang opined that ‘there is no single way to order the generation; to benefit the state, one need not imitate antiquity’.⁵¹

As memorials submitted to the throne, both Shen's and Zuo's proposals for colonization would have been read by a relatively small official audience. Yet they illustrate the deep roots of colonization strategies in late Qing political thought and the range of Chinese referents which could be used to argue in favour of them. Although both men addressed their memorials towards a specific policy problem, they were also both involved in controversial ‘self-strengthening’ projects, notably the Fuzhou dockyard, which drew on Western expertise and technology, and their calls for reform should be read in this context.⁵² They were part of a generation of reformist thinkers who used the Chinese past as a resource because, as Leigh Jenco has suggested, ‘the constant negotiation of the gaps between past and present would enable future development’.⁵³ This was in contrast to their conservative opponents who sought to preserve present practice into the future and did so by obscuring the multiplicity of past practices which might serve as a model for the present. With these Chinese referents available, Cen Chunxuan and Yao Xiguang did not have to use the European past to advance very similar policies. They did so because after 1895 the political and intellectual climate had changed dramatically, and their new referents reflected what they saw as a new age. Yet, in order to do this, they had to rely on tropes about Mongols which were much more locally sourced.

III

Mongolia, like other frontier regions, had been ruled by military governors and not by civilian governor generals as in China proper. There were also prohibitions on Han Chinese settling in the region to protect the delicate alliance between the Manchu Qing and Mongol interests.⁵⁴ The first calls to amend this situation came in the 1870s and the 1880s from Li Hongzhang and Zhang Zhidong.⁵⁵ However, their proposals were pragmatic responses to particular incidents rather than demands for wholesale reform. Li acted in response to incidents of Mongols selling their land to foreign missionaries. He suggested that the Qing state should not ‘uniformly prohibit [Han migration] but only allow [Mongols] to rent their land but not to sell it, and also to fix a limit on the number of tenants, and then foreigners will have no way of lying in wait to seize the space’.⁵⁶ This argument for piecemeal reform was confined to a brief memorial and drew on no historical precedents either through narrative examples or through proverbial references in four-character expressions. After 1900, however, officials set out grander visions for reform of the frontiers. These visions must be read in the context of changed political circumstances and the new intellectual trends they fostered.

These proposals were linked to the ‘new policies’ (xinzheng 新政) promoted to repair the damage after the Qing court's decision to back the Boxers. After the Qing defeat at the hands of eight foreign armies, the empress dowager Cixi called for reform, declaring that ‘what suffocates all under heaven is precedent (li 例)’.⁵⁷ She asked her officials to seek out not just the superficial causes of Western wealth and power, but rather to closely examine the core institutions of European states and how they diverged from Qing practice in order to identify the underlying causes of the power differential between them.⁵⁸

The empress was borrowing from a new generation of thinkers, including Yan Fu and Liang Qichao 梁啓超 (1873–1929), who argued that past reforms had failed because it was not possible to study European skills without acknowledging their embeddedness in European institutions and substance. This led them to call for ‘institutional reform’ or ‘changed referents’ (bianfa 變法), by basing reform on European institutional models.⁵⁹ While earlier self-strengthening thinkers had adopted a utilitarian approach to the adoption of European ideas, Yan saw their adoption as a necessary step in the survival of the Chinese people. This thought carried an intellectual tension between historical inevitability and the need to act which was unresolved. Later, ‘May Fourth’ generation thinkers would resolve this by regarding culture as historicized and thus rejecting the inevitability of the evolution of Chinese society along European models.⁶⁰ The emphasis on the inevitability of development in Yan's thought, however, was driven by his understanding and use of Darwinian thought.

As was the case for Darwin's reception in the Arab world, in China a ‘matrix’ of readership and translation resulted in his work being understood through his readers and translators, none of whom were more influential than Yan.⁶¹ Yan was recruited to the school attached to the French-run naval dockyard at Fuzhou by Shen Baozhen and sent to England to continue his studies in 1877. By 1881 he had read Herbert Spencer's A study of sociology, but it was not until the Sino-Japanese War (1894–5) that he applied these ideas in a series of articles, including ‘In search of strength’ in a Tianjin newspaper.⁶² Yan transformed Spencer's description of an impersonal process whereby social groups compete for supremacy, with the fittest surviving, itself already an inexact rendering of Darwin's ideas, into a prescription for the transformation of Chinese society.⁶³ Yan's use of Darwinian thought to suit his understanding of the crisis China faced after defeat to Japan is evident in his selective translation of Thomas Huxley's Evolution and ethics. For example, he was interested in Huxley's metaphor of plant colonization, hinting at China's own colonization, but he omitted passages which suggested that evolution could result in retrogression or stasis.⁶⁴ Yan's ideas were further perpetuated by Liang Qichao, who extended the idea of competition between nations to competition between peoples within nations.⁶⁵ This would be refracted in Yao and Cen's discussion of the differing fates of Mongol and Han peoples.

The proposals for reforming Mongolia were stimulated by the crisis of 1900 because, as part of its alleged efforts to suppress the Boxers, Russia had occupied neighbouring Manchuria and had refused to comply with its obligations to withdraw troops.⁶⁶ The governor of Shanxi Cen Chunxuan's 1901 memorial proposing to make Mongolia a province echoed the proposals of Shen and Zuo. He began by arguing that ‘formerly military colonies were regarded as the best policy for the defence of our remote borderlands … [but] supposing we regard opening up (tuo 托) as that which is most pressing?’⁶⁷ This distinction, between the establishing of military camps and the opening up of land to support frontier populations whose taxation could pay for their own defence, was one which both Shen and Zuo would have recognized. He then followed their tactic of emphasizing past changes in the frontier which justified present policy. He noted that, in the past, inner Mongolia had been regarded as beyond the frontier; its people, moreover – variously the Xiongnu, the Tujue, and tartars – were regarded as the enemy. Now its land and people had been incorporated into the empire and Russia was the enemy.⁶⁸ Finally, he shared Shen and Zuo's concern about the dangers of a passive policy of simply regarding a frontier as a ‘screen’ (fanyuan 藩垣) defending the capital which required military surveillance and little else.⁶⁹ He even drew on the same metaphor as Zuo for military complacency, with both claiming that, since the founding of the capital, ‘the beacon fires have not been lit’ to warn of an advancing enemy, suggesting either that he was directly influenced by Zuo or that they were both drawing on the same source.⁷⁰

Given the extent to which Cen's thought was embedded in prior frontier thinking in suggesting that using the people to cultivate land could pay for frontier defence, it is surprising that he should go on to claim that ‘[I] request a plan which has no antecedents (qing ji wu xian yu ci zhe 请計無先於此者)’.⁷¹ In fact, his claim, albeit a tenuous one, was that his policy had no Chinese precursors. Rather, he had found a model which could be followed in European colonial practices:

[I] consider that recently since the Russians have managed Heilongjiang [they have] opened more than 200 farming villages, [and] the Americans and the English opened San Francisco and Melbourne respectively for cultivation. They were all vast, bleak, and desolate places, with hard, infertile, and impoverished land. They exhausted their efforts to manage recruiting people to plough the frontier. Once real prosperity was achieved they became popular and the fees extracted by the state from the profit from cities and opening the land were used to support [defending] the borders. Now there are canals and rivers spreading into Mongol border territory which can attract [people], there are people who know [this] renting land and tending sheep, [and] those refined people go through many difficulties [lit. ‘hack through brambles and thorns’]. [If they] work together to industriously build, they can imitate those unfathomable lands.⁷²

Cen's claim that his frontier policy was novel is a political claim about the sources of inspiration for his strategy rather than a factual one. Given that the memorial was written soon after the empress dowager's call to learn from European examples, this is in some ways unsurprising. Claiming that a policy was foreign-inspired was beneficial in this context. However, after the crises of 1895 and 1900, comparing Chinese frontiers with those in Europe and America served another purpose. By placing them in what he regarded as global – though European-determined – historical time, Cen reminded his readers of the urgency of the foreign threat and suggested that his proposals were part of a wider, inevitable, global process. In this he appears to have been drawing on Yan's and Liang's influential presentations of Darwinian thought as entailing competition between peoples, with radical consequences. By opening up the land to the ‘good people’ – by which Cen meant Han Chinese agriculturalists – he was implying the upending of Mongol privileges. This effectively placed Mongols on a par with past objects of frontier reform including Uighurs and, worse in terms of the empire's racial hierarchies, even the Taiwanese shengfan. In effect, he also wrote them out of history by arguing that opening up was essential because ‘if there are no people, how can [we] put [our] plans into effect’.⁷³ The only ‘people’ (ren 人) in this schema were Han and not Mongol. The evolutionary implications of this proposal were explicitly set out in the proposals of Yao Xiguang.

As with Cen, although Yao's policy-making was deeply influenced by policies familiar to Chinese statecraft thinking, he turned to Europe for his historical models. In the 1890s, while working as a magistrate in Anhui, he had stressed the importance of agricultural cultivation and complained that only 40 per cent of cultivatable land was being farmed.⁷⁴ Despite the statecraft origins of his views, in his proposals for making Mongolia a province he drew on the example of European colonialism. He began by saying that the ancients had known that solid frontiers were the most important part of frontier policy, but swiftly moved on to argue that frontiers were strengthened by what Westerners called ‘expansion power’ (pengzhangli 膨脹力). This vocabulary appears to be directly borrowed from the evolutionary model of inter-state relations proposed by Kang Youwei 康有為 (1858–1927) and Liang Qichao. In April 1898, in a lecture to the Society for Preserving the Nation, Kang Youwei asserted that the ‘thermal power’ (reli 熱力) of the people's hearts could increase the ‘expansive force’ (zhangli 漲力) of the nation.⁷⁵ This vocabulary was further developed in an article in Liang Qichao's Qingyi bao 清議報. The author warned that Western powers planned to use their infrastructure projects, notably the Russian Trans-Siberian Railway and a proposed US canal across Nicaragua, to export their ‘expansion power’ (pengzhangli 膨脹力) to dominate Asia.⁷⁶ As Martin Dusinberre's article in this issue highlights, the neologism arose from Japanese attempts to understand the driving force underpinning European expansion.

Yao, heeding Cixi's call to understand underlying causes of European strength, tried to explain the nature of the ‘expansion power’ which drove successful European colonialism. He determined that three inter-related forces were required to support expansion: demography, government, and wealth produced through agriculture, commerce, and industry.⁷⁷ As a result of these factors, Western populations overflowed and they were ‘able to travel thousands of li from their own countries to invade other people's countries, like water flowing onto the earth’.⁷⁸ Yao's open acknowledgement of the violation of foreign sovereignty inherent in this expansion makes his subsequent comparison between these processes and proposed Han expansion in Mongolia all the more striking:

Without exception those skilled in agriculture are the first to come into contact with this newly opened land. For it is necessary for agriculture to flourish and then the undeveloped grasslands (caolai 草萊) are opened up and the abundant vegetation is mown and like this people start to stay in those places, [this is the] way that things are handled overall, and smoke and houses begin to assemble, and industry and commerce then attach themselves to this. This is naturally not easy to manage. America and Australia both started by trying their best to manage agriculture, and this is clinching proof of its effectiveness. How is this different from our Mongolian lands? Therefore those who have the ability should exhaust their efforts on expansion, [and] those who do not have the ability to expand should exhaust their efforts promoting [this expansion], this is really the correct way [to manage] the borders!⁷⁹

Yao, influenced by evolutionary thinking, thus argued that those who did not have the capacity to expand – the Mongolian herders of the grassland – should be complicit in, and even encourage, their own colonization. He deployed more recently imported scientific language from Europe, via Japan, to underline the inevitability of this process. He argued that nomadic peoples were like fluids which could only take on the shape of the container they were placed in, while agriculturalists were like solids, who defined the shape of the utensil itself; ‘this being the case, nomadic herders cannot grow, whereas agriculturalists rapidly solidify [their position]’.⁸⁰ He evidenced this with the historical example of Russian eastern Siberia, where ‘the nomads are daily declining and the agriculturalists are daily expanding’.⁸¹

By using the language of ‘expansion’ and European historical referents, Yao sought to demonstrate that his policy was essential to a new time. He suggested that Ban Gu's and Sima Qian's complaints about military adventurism on the frontier were misguided because they did not ‘know or see that the powerful expansion of Westerners was not fuelled primarily by military strength, but rather military strength came from it’.⁸² They were, in other words, cut off from the knowledge which Yao possessed: knowledge of a world in which expansion was vital for survival and in which that expansion could only follow a European model of colonization. He argued that Westerners had used their expansion power to ‘harm our national strength’.⁸³ Others were more direct in explicitly stating that European imperialism had brought about a new historical era. By 1908, one Manchu thinker, Rong Sheng 榮陞, labelled this an ‘era of national imperialism’ (guomin diguo zhuyi shidai 國民帝國主義時代), in which states conquered foreign territory and assimilated foreign races.⁸⁴ The European colonial past allowed them to threaten China in the present and, if the Chinese did not adapt frontier practices by learning from this past, the state would not survive. The global historical time which Yao saw China as joining was a form of inevitable, evolutionary time, where certain groups had a future, while others did not. Here he appears to have followed Yan Fu in implying that humans could intervene in evolutionary forces, but in suggesting that only the Han, and not Mongol nomads, could do so.

The image of the Mongolian population that Yao, and to a lesser degree Cen, presented was not one which was simply borrowed from European tropes about nomadic populations. Rather, it was firmly entangled with indigenous conceptions of Mongolia which were a product of the Mongols’ special status within the Qing empire. Yao's insistence on the radically nomadic nature of Mongol peoples does not accord with what he himself must have experienced when travelling in Mongolia. Land use was carefully managed by the noble or Buddhist ecclesiastical authorities, such that herders had a right to move their herds only within a given locality.⁸⁵ This was not the radical, literal, fluidity of populations which Yao described. His description was persuasive for a Han audience because this closely resembled the essentialized image of Mongol populations which had been passed on to Han Chinese by Manchu officials since the Qianlong reign.⁸⁶ This image was also a direct result of tensions caused by increasing numbers of Han Chinese violating immigration bans and entering Mongolia from the early nineteenth century to engage in activities such as mushroom picking. In response, local ambans (headmen responsible for the Mongol population) warned of the need to ‘purify’ the Mongol banners. Despite a long-standing hybridity between Han and Mongol, particularly in border areas, fears of undocumented migration led to a more radical vision of a pure Mongolia.⁸⁷ This in turn led Mongols to increasingly derive their identity from their banners, an identity in part created by the categorizing efforts of the Manchu Qing.⁸⁸ This vision, of a pure and differentiated Mongolia, allowed Han thinkers like Yao to merge the historiographical tropes of Mongol nomadism with that of the nomadic populations colonized by Europeans.

Yao's proposals were also addressed to internal Chinese political debates about the new structure of the empire, which placed the Chinese in a new historical time, while leaving the Mongols ‘behind’ the times. At the end of his memorial he referenced the debate between fengjian and junxian as systems of government, suggesting that nomads were suited to the fengjian system of devolved local power, while agriculturalists fared best under the centralized junxian system.⁸⁹ This debate had been ongoing since the formation of the Qin empire (221–206 bce), with many thinkers expressing a preference for the pre-Qin fengjian system, with its decentralization and limited role for government. However, in most dynasties, thinkers had to overcome this ideological preference because, from the Qin onwards, political authorities relied on the junxian system of prefectures, over which the government had more control.⁹⁰ To overcome this paradox, many Chinese scholars had argued that, while the fengjian system was admirable, it was necessary to rule according to the times.

Yao probably returned to this debate in response to Liang Qichao's revival of it to argue in favour of fengjian-style localism.⁹¹ For Yao, localized control was not suitable in an age of agricultural colonists. Yao rejected Liang's vision, reasserting the temporal priority of the junxian system, which suggests that, while influenced by the evolutionary paradigm, he did not simply take on Liang's ideas wholesale. He argued that the fengjian system could not be restored after the Qin empire had imposed prefectures and counties, but nevertheless he thought that the continued existence of the fengjian system in Mongolian territory could help explain the backwardness of Mongol populations. In this sense, the idea of a time of centralized empires was not simply imported from the West and projected onto a Chinese tabula rasa. Rather, evolutionary thinking about the development of the state blended with pre-existing Chinese debates about statecraft, as well as with essentialized ideas about the Mongolian people.

IV

The distinction between Shen and Zuo, writing in the 1870s and 1880s, and Yao and Cen, writing in the first decade of the twentieth century, cannot be drawn along policy lines, but rather through their divergent understandings of historical time. All of these thinkers shared a developmentalist policy platform which proposed settling agriculturalists on frontier ‘wasteland’, displacing its current occupants if necessary, to strengthen central government control over threatened border regions. Where they differed was in their choice of historical referents to justify their proposals, and, more significantly, in their attitude to historical time. Shen and Zuo drew on Lord Shang's admonition to ‘rule appropriately according to the times’, drawing attention to the differences between past and present time in order to make the case for change. Although Cen and Yao continued to make use of Chinese tradition, they were almost certainly influenced by the evolutionary thinking of Yan Fu and Liang Qichao, and therefore concomitantly attempted to place China's frontiers in the context of global historical time. At a time when reformers were warning that European ‘expansion power’ might overwhelm Asia, they drew on the European past to understand what these past changes meant for China's future. They were not determinist in their thinking, because they were suggesting policy options which might alter China's place in this new world. In this they followed Liang's conversion of evolution from an impersonal force to one which could be shaped by human intervention. Yet Yao in particular took a determinist stance on the present occupants of the frontier. They would be colonized either by Europeans or by the Chinese, because in an age of empires some form of central control by a large state was inevitable, and nomadic life no longer tenable.

Yao's and Cen's attitude to historical time reflected that of the reform-era generation of thinkers. The ‘May Fourth’ generation which succeeded them would reject evolutionary models. They suggested that China could not simply evolve along Western lines towards a universal future because the evolution of Western states was a product of Western culture, and thus change in the future required effacing their own, Chinese, past.⁹² Yao's and Cen's evolutionary models represent an important watershed between this perspective and that of Shen and Zuo, who looked to the Chinese past to justify change in the present. They suggested an inevitability in the process of evolution from nomadism to sedentarism but they did not see this as culturally specific. Both thought that Chinese settlers could follow the examples of Europeans and Americans on a universal path of development. In practice, both Yao and Cen and the May Fourth generation believed that activist policies were necessary to change China. Their intellectual underpinnings, however, in the form of their differing interpretations of evolutionary process, were quite distinct. Additionally, Yao and Cen were not introducing new policies, just old policies to new regions. In contrast, the May Fourth generation emphasized the wholesale overhaul of Chinese society.

One important function of the use of contingent European and American historical examples for Yao and Cen was to create a sense of historical inevitability about their policy proposals. For the Qing state, even in the 1900s when it was attempting to reform its institutions and practices at an empire-wide level, reforming Mongolia remained highly controversial. By linking his proposals to observable empirical phenomena, both in the course of European history and in the natural world, Yao could argue that doing nothing was not an option. It was inevitable that herders would be replaced by settlers, so the only question remaining was which settlers would do the replacing. It was vital to understand what Yao and Cen suggested were the facts about European expansion in order to facilitate effective Chinese expansion, which would be fundamentally different from past Chinese efforts to rule the frontier. Neither explicitly referred to an ‘age of empires’, but they acknowledged a break with the past and the arrival of a new time which was distinct from simply promoting change through reference to Chinese precedents, such as the legalist thinking of Lord Shang.

Yao's and Cen's understanding of the place of Chinese frontiers in historical time reminds us of the connection between evolutionary, teleological models of time and colonialism on a global scale. They drew similar conclusions to European thinkers who, influenced by social Darwinian thinking, placed indigenous populations on a different time stream – one without a future. This was succinctly put by a British Chinese imperial maritime customs official who, in the 1880s, compared Taiwanese shengfan to ‘the last of the Mohicans … the impression left on my mind was a confused and rather sad one. I had been amongst a people whose days are numbered.’⁹³ As Sebastian Conrad has recently argued, cultural and intellectual historians have overemphasized the connections which linked these shared understandings of temporal change without paying enough attention to their contexts. In doing so, they have implied a diffusionist model which suggests that the diffusion of ideas themselves could be a catalyst for change.⁹⁴ Yao and Cen were, of course, influenced by Liang Qichao, Yan Fu, and their Japanese and European sources. This differentiates them from Shen and Zuo. Yet their understanding of a new time was not just transferred from Europe but was entangled with a combination of their perception of Mongolians, a contingent product of Manchu frontier policy, and long-standing statecraft thinking.