Introduction
Duan Pengrui 段鵬瑞 had had enough. It was only the beginning of summer, and yet he and his little surveying party were sweating like rain,Footnote 1 plodding along the interminable riverbank, cursing the heat and humidity of the river valley. Following the course of the lower reaches of the winding DzayulFootnote 2 on foot, his itching, red-welted ankles were as much a record of the route they were traversing as they were a ledger of the local arthropods. His men complained about the buzzing swarms of mosquitos and scorpions, lying in wait in the scramble and brush they waded bare-handed through, leaving them so swollen and bloody they could barely pay attention to the measurements and observations they were supposed to be taking. Duan concluded in his notes of the expedition that the miasma of this particular part of the valley was so severe that despite the fertile soils, the land around remained uncultivated and unsettled because survival, let alone establishing livelihoods, was near impossible in these conditions. He noted that Tibetan pastoralists from Batang Province had migrated here about ten years prior, intending to open up new farmland. That project had been seemingly abandoned and the settlers had moved elsewhere, leaving behind only a few households in the area.Footnote 3
Shaped by the acquisitive desires of imperial state-making and capital, modern-day Zayul County today borders India and Myanmar (Burma) to the south, and China’s Yunnan Province to the southeast. Arising in a series of glaciers in the Tibetan Gangri Karpo mountain range (an eastern trans-Himalayan range that transitions into the Indian Mishmi hills or Qilinggong mountains in China), the Dzayul Chu (རྫ་ཡུལ་ཆུ་) or Zayü he (察隅河) courses from Zayul County in the Tibet Autonomous Region to merge with the northeastern Gangri Garpo Chu (also called the Rongto Chu) below Rima, a town sitting right on the boundary between India and China in modern-day Arunachal Pradesh. From here the Dzayul enters the Assamese plains as the Lohit, a major tributary of the Brahmaputra River in India. Historically, the Dzayul region was marked by the central basin of the Dzayul River with its two branches, and the Salween River (Nagchu ནག་ཆུ་ or Nujiang 怒江), flowing eastward into Burma. Through the eighteenth, nineteenth, and twentieth centuries, these Eastern Himalayas were framed as the “wild” frontiers of British, Chinese, and Tibetan state-building efforts—a fertile paradise whose productive potential could transform imperial destinies, as much as its treacherous terrain and hostile inhabitants would surely thwart the “natural” march of these ambitions.
For pastoralists, peasants, merchants, pilgrims, and other inhabitants of the Brahmaputra or Tsangpo valleys, the river formed an integral part of their everyday life and practice. Their intimate knowledge of the river, its course and the routes leading to, from, and along it, was widely disseminated throughout local audiences. The production of knowledge about the Dzayul/Lohit–Brahmaputra/Tsangpo landscape was the outcome of a long process of productive activities that supported and reproduced life on its banks and its valleys. At the same time, these river systems were also turned into objects of sustained study by (colonial) state-building efforts, which sought to chronicle human stories about navigation, agriculture, and urbanization. Who could be considered the “expert,” or the authoritative voice that extrapolated these riverine environments? What kinds of skills and training were validated in the recognizing of such expertise? Whose or what kinds of knowledge were rejected? These are the central questions that I explore in this article.
Situated forms of knowledge about these river systems were often present exclusively as metaphysical representations of the river, accessible to largely local audiences, not always recorded in written formats, and often indistinguishable from community practice. By contrast, Tibetan, British Indian and Qing attempts to study the river relied heavily upon translating the knowledge possessed by locals into registers that would be intelligible to non-local audiences. Although all three of these projects of knowledge dissemination had extensive reach at the time, the focus of the latter two on conversations surrounding the “universal” nature of so-called “modern” map knowledge are more comprehensible to twenty-first-century readers.
This article disentangles the strands of various state knowledge projects that sought to interpret this mountainous region, thereby reconstructing the historical processes that constituted the corpus of Qing “colonial” geographical knowledge about the region and the “experts” who rendered legible this information to broader imperial, non-specialist audiences. I argue that the relationship between “knowledge” and “expertise” was a complicated one: “expertise” did not clearly draw from possessing a deep knowledge of the land and terrain or how to navigate it. Rather, “expertise” arose from the creation of hierarchies around asymmetrical interactions between individuals, their diverse knowledge practices, and the land upon which these practices were enacted.
To illustrate my arguments, I focus on selected Chinese and Tibetan mediators of the situated histories of the trans-Himalayan and sino-Tibetan borderlands. I reconstruct the processes by which the geo-knowledge of these regions was regularized by the Qing state by tracing the interactions and participations between imperial producers, objects, and consumers of that knowledge, against evolving hierarchies of value that restricted the definition of “expert” to particular systems of knowledge. This approach follows anthropologists of expertise who understand expertise as fundamentally interactional and thereby fundamentally ideological,Footnote 4 and I will demonstrate that the so-called expert was necessarily implicated in ascribing and organizing the objects of investigation in terms of their perceived value, thereby creating uneven relationships. These asymmetries would be further institutionalized and naturalized as technical protocol by academies that sought to set field standards. In the process, the epistemic regimes that granted recognition to such “expertise” routinely appropriated, decontextualized, and erased the contributions of “local” interlocutors from the corpus of knowledge about the mountains, the river, and its environs.
Through the nineteenth and twentieth centuries, the production of “modern” frontier spaces in Asia was shaped by complex processes and interactions located within the drives of imperial expansion and uneven practices of knowledge production about geographical spaces.Footnote 5 These “frontiers” were thus closely linked to the very process of empire-making itself, as new geographical entities that did not map on to pre-colonial landscapes. The attempts to explore and produce “new” knowledge about frontier regions were linked both to the expansion of imperial territory, as well as the embedding of imperial rule along those frontiers. The diverse geographies that marked the Himalayan and trans-Himalayan landscape were gradually absorbed into a framework of frontiers and borderlands that linked colonial territories possessed by imperial states. The historical attempts by these states to impose artificial political boundaries in the trans-Himalayan and Tibetan regions transformed these vibrant pre-colonial landscapes into British and Chinese imperial objects, and reorganized connected geographies into oppositional political borders that constantly required constant delineation, justification, and military defense, a legacy inherited by the modern nation-states of India and China.
For the British state in India, surveying and mapping Tibet as a whole or a part, or geodetically locating places like Lhasa, always included the brief of tracing the course of the river Tsangpo, as well as where the river either drained into Indian territory or linked with the Brahmaputra. Significantly, between 1760 to 1920, almost every hydrographic survey of the Tsangpo-Brahmaputra led by British surveyors was performed as part of a military expedition. Further, between 1840 and 1903, the British state’s knowledge of this river and the surrounding Himalayan regions was almost exclusively built out by native surveyors employed by the Survey of India.Footnote 6 While the link between the production of geographical knowledge about the Himalayas and British colonial governance has been well-established,Footnote 7 the true extent to which the British Indian state depended on local subaltern agents moving through and past the Himalayas remains insufficiently analyzed.Footnote 8 Similarly, the role and place of Tibetan and other “borderland” agents or populations has also been relatively underdeveloped in the histories of Qing-era and Qing-state geographical knowledge production in the nineteenth and twentieth centuries.Footnote 9
My arguments are explicated in the course of three sections: the first two sections investigate how the Qing state validated particular presentations of information as useful geographical knowledge when it followed established practices of collection that were referential and citational. Over the course of the nineteenth and early twentieth centuries the imperial state organized these practices as disciplinary protocol and specialized knowledge, through targeted processes of institution-building. These institutions in turn mediated, enforced, and normalized the new kinds of interpretative frameworks through which spatial formations like the Himalayas or Tibet would be viewed, and who thereby could be deemed the “expert” knower of those regions. Because the epistemic regime granted recognition only to selected enactments of “expertise,” the organization of the Qing imperial archive(s) of geo-knowledge about the Himalayas effectively overwrote situated community histories and the spatial relationships of local populations with these mountains. The third section attempts to unpack and move beyond the dominant interpretive frameworks of colonial knowledge production, reading past the silences of the Chinese archive’s scant references to local knowledge practices. Here the focus shifts to Tibetan and local histories of the Himalayas, foregrounding in particular the role of gter ston གཏེར་སྟོན་ (hereafter tertön) or “treasure-discoverers” in exploring new regions for settlement in the mountains.
The universality of contemporary maps can tend to present the construction of imperial archives as totalizing projects, where the “scientific modern” becomes a pre-developed global “truth” that relentlessly sweeps aside all other systems of knowledge in its rampaging path.Footnote 10 However, the cartographic discourses and technologies that resulted in the generation of new maps about the Himalayan and trans-Himalayan frontier regions in the late nineteenth and early twentieth centuries were not the result of a teleological, total, subsuming by a hegemonic European knowledge system. Although the work of circulation theory has moved us away from the language of “encounters,” the notion that certain knowledge spheres had parallel as opposed to intertwined trajectories still holds ground. However, the assumption that “indigenous knowledge” is a wholly separate sphere from so-called “Western” knowledge production merely roots the idea of the “indigenous” in the colony, whereas the “colonial,” while directed towards the colony, is understood to originate in the metropole, with the latter ultimately diffusing into and completely transforming the former. By disassembling the practices that undergird the enactment of “expertise” about the geo-knowledge of the Eastern Himalayas, I question the binary categories of “colonial” and “indigenous” knowledge about the region.
Importantly, “colonial” epistemology in Qing China (or British India) was not a top-down, hegemonic system that upended all other knowledge systems in its path, and nor is it easily explained through processes of transculturation, hybridity, or horizontal exchanges with those traditions of knowledge. Theses that argue for totalizing transformations seriously undercut the power of local forms of knowledge production and transmission against the weight of colonial states and their imperial agents, while at the same time underplaying the extent of competition and accommodation within situated forms and their extrapolations. The “scientific modern” was an uncomfortable site of negotiation where the British and Qing states were forced to reckon with indigenous, community-based, religious, or even mystical traditions of knowledge in the production of geographical knowledge about the lands that defined these empires at their borders. These negotiations depended on deeply fraught and uneasy processes of accommodation between different forms of expertise and skill, where extrapolators and analysts of geographical knowledge about the Himalayas jostled to establish themselves against the constraints of imperial systems predicated on coercive technologies to identify territory. Therefore, I argue that geographical “expertise” was a performance, rather than the inherent possession of knowledge(s) about the region.
Part 1: The Vocabulary of Qing and Chinese Place-making in Tibet, 1720–1911
In the final decades of Qing rule, the Li-fan Bu 理藩部 (Ministry of Colonial Affairs)Footnote 11 initiated a series of administrative experiments as part of state attempts to standardize imperial policy regarding frontier management.Footnote 12 A renewed attention to the Tibetan frontier in the latter half of the first decade of the twentieth century included the establishment of new liaison offices, under which Zhao Erfeng 趙爾豐 (1845–1911)Footnote 13 was granted the title of Border Affairs Minister at the newly established Ministry of Border Affairs in Sichuan and Yunnan (川滇邊務大臣衙門), stationed at Garze.
Between 1910–1911, Qing efforts to formalize the status of Tibetan regions within its imperial dominions included commissioning new surveys of the Dzayul river basin to demarcate “Chinese” territory in the foothills of the Himalayas. Zhao Erfeng implemented a series of interventionist policies intended to forward the administrative mandate of bringing Tibet and Tibetan affairs closer to the purview of the imperial center in Beijing. He charged three of his subordinate officials, Cheng Fengxian 程鳳翔, Duan Pengrui, and Xia Hu 夏蘋, with surveying on foot the route of the Dzayul River, which flowed through the Mishmi hills into (British-Indian held) Assam as the Lohit.
The Eastern Himalayas had not been surveyed by any agents employed by the Qing court since at least 1720.Footnote 14 Therefore, not only were the 1911 surveys new material at the time, but they can also be accurately characterized as the very first in-person Qing surveys of this region, particularly of the route of the Dzayul River. The Jesuits surveying the lengths of the Qing dominions for the Kangxi atlas did not travel to Tibet. As with other regions that they did not or were unable to personally travel to, they relied on local elites to conduct independent surveys, and included those maps in the grand atlas.Footnote 15 As a result, the geodetical information compiled into the atlas was pulled from a variety of surveying techniques and epistemologies, since although the Jesuits attempted to use triangulation (as far as possible) for their own surveys, they did not insist upon it when receiving surveys or reports from others. Kangxi reportedly commissioned a team led by a Tibetan Buddhist lama by the name of Chu-er-qin Zang-bu La-mu-zhan-ba 楚爾沁 藏布喇木占巴, Tsultrim Zangpo Rab Jampa (tshul khrims bzang po rab ’byams pa ཚུལ་ཁྲིམས་བཟང་པོ་རབ་འབྱམས་པ་) who had previously studied geometry and arithmetic in Beijing to conduct independent land surveys of Tibet on foot.Footnote 16 He limited his surveys to the main roads connecting Qing-controlled territories to Lhasa, while depicting other areas on the basis of second-hand data.Footnote 17 He delivered his findings to the Jesuits in 1718, and his surveys were reproduced in both the Yongzheng and Qianlong atlases. Although the latter two included piecemeal additions to the Kangxi edition, new surveys of the frontiers were only ordered for Xinjiang, not Tibet.
Thus, while the Qing surveys of 1716–1718 in Tibet comprehensively mapped the central Himalayas, if there had been any planned surveys of roads in Kham or Eastern Himalayas, in particular of the eastern end of the Gangri Garpo range, those were hastily abandoned when the Dzungar Khanate invaded Tibet.Footnote 18 Barely escaping from Lhasa with their lives, the surveyors withdrew to Kham to wait out the war first at Qamdo (昌都寺) and then at Chaya Monastery (察雅寺) in modern-day Chamdo Prefecture.Footnote 19 The work of the incomplete surveys was not taken up again even after 1720 when the Qing had established themselves in Tibet. And although the production of geographical knowledge about the Tibetan frontier remained a robust scholarly tradition throughout the late eighteenth and nineteenth centuries through the publication of meticulously detailed, locally organized gazetteers, geographical treatises, and memoirs, the absence of imperially commissioned surveys after 1719 meant that most of these works largely tended to rely on older, unverified sources.Footnote 20 The cultural authority of these geographical works did not depend on whether their authors(s) had actually been present in person in the areas they were writing about. Rather, through the eighteenth and nineteenth centuries, “expertise” was performed by the literati as a very particular presentation of knowledge about the locale: the semiotic enactment of lengthy citational practices.
The 1718 surveys continued to principally inform the official archive of Chinese geo-knowledge about Tibetan regions well into the late nineteenth century, such as the Xizang Quantu 西藏全圖 (“Complete Map of Tibet”) copied by Han Daling 韩大令 for the Xizang Tukao 西藏圖考 (Atlas of Tibet) compiled by Huang Peiqiao 黃沛翹 between 1885 and 1886 and eventually published in 1894. The Xizang Tukao also reproduced maps included in the Xizhao tulue 西招圖略 (A description of Tibet accompanied by maps), and although those were based on inspection tours personally undertaken by Song Yun 松筠 (1752–1835),Footnote 21 the Dzayul watershed was one of the regions the amban had not toured himself. Huang Maocai’s 黃楙材 (1843–1890) Longchuanjiang Kao 龍川江考 (Note on the Longchuan River), written during his travels in 1878, recorded the local names of the two branches of the Dzayul river; but since his journey only took him through Burma rather than Assam, the geodetical information regarding the names and location of those rivers did not come from route surveys that he had conducted himself.Footnote 22
The scope and scale of the survey work produced under the command of Cheng Fengxian, Duan Pengrui, and Xia Hu was, at least in this regard, fairly unprecedented. By this point in the early decades of the twentieth century, the basis of the map’s cultural authority had gradually shifted from the literatus in his study to the surveyor in the field, largely contemporaneous with developments in mapping practices in British India.Footnote 23 And much like the surveys of the Himalayan and trans-Himalayan regions conducted under the aegis of the Survey of India, the 1910–1911 Qing surveys of the Dzayul region were also part of a military expedition.
Zhao Erfeng’s punitive invasion of Kham in response to the Khampa rebellion has been well-documented, the exceptional violence of which had earned him the moniker of “Butcher Zhao” even before his appointment at Garze.Footnote 24 Following the occupation of Batang (in modern-day Sichuan) in 1906 and the appointment of Lian Yu 廉隅 (1866–1972) and Zhao Erfeng as liason officials by 1908, Zhao launched a series of military campaigns intended to establish the Qing as the sole political authority in the region. The surveys commissioned by him were intended to include important ethnographic and demographic information with reference to the local inhabitants, the cultivable lands and resources naturally occurring therein, and geodetical information regarding the natural features of the landscape. The agents who gathered this information were professionalized to different degrees as military surveyors, since although surveying had recently become an institutionalized part of the training offered in military academies, individuals had different degrees of technical skill.
In 1909, Cheng Fengxian, then rear battalion commander of the border army stationed at the salt wells at Yanjing in Batang Province, was ordered to lead troops southward to garrison Sang’ang Quzong 桑阿卻宗 and its surrounding towns in southwest Kham (today these towns all form part of modern-day Chayu County).Footnote 25 In his 1910 travelogue titled Kamu xinan jicheng 喀木西南紀程 (Travels in Southwest Kham), Cheng Fengxian stated in the section titled Zayu Dili 察隅地理 (The Geography of Zayu) that he personally travelled to the mouth of the Yabiqugong 壓必曲龔 or Yepak River, a tributary of the Lohit, to follow the course of the river as far possible on foot. As detailed in the section titled Zhu Lucheng Zhan 諸路程站 (Travel Stations), Cheng Fengxian’s surveys covered both the upper and lower Dzayul river valleys, moving through Monpa and Yi territory in the historical Tibetan kingdoms of mon yul མོན་ཡུལ་ and spo yul སྤོ་ཡུལ་ in particular. His reports focused on the roadways that connected the major Tibetan towns of Tawang (transcribed in Chinese as Chawa), Qamdo (Chamuduo) and Kyirong Tsegong (Sang’ang Quzong) to Batang.Footnote 26 Both his travelogue and additional reports sent to Zhao Erfeng between 1909–1910 make substantial mention of British presence (and British flags) past the Dzayul watershed.Footnote 27 In particular, Cheng Fengxian pointed out that the British appeared interested in ascertaining the flows of local tribute through Monyul and Poyul, the same regions and revenue networks where Qing officials themselves were anxious to secure local loyalty.Footnote 28 In compiling Kamu xinan jicheng, Cheng Fengxian consulted prior works detailing the geography of the region, and in the process updated works such as Huang Peiqiao’s Xizang Tukao. Footnote 29
Zhao Erfeng wanted Cheng Fengxian to pay special attention not simply to where the features of the earth lay, but to the potential they might bear. He directed the surveying party to assess the cultivability and habitability of the lands they were passing through, with the intent of establishing new settlements and markets in the region.Footnote 30 While the Qing state so far had not pursued a developmentalist policy in Kham, the vein of such correspondence between Zhao Erfeng and his subordinates indicates a new interest in developing these Tibetan regions as resource frontiers.
By April 1910, Zhao Erfeng had commissioned Duan Pengrui, then a member of the Yanjing Salt Bureau, to report to Cheng Fengxian and extend the work of the survey.Footnote 31 Duan Pengrui was native to Yunnan Province, but aside from his hailing from Jianchuan County it is unclear why he was hand-picked for this task. It is also unclear whether Duan Pengrui had demonstrated technical capabilities that his superior did not possess, since Duan Pengrui and his surveying party appear to have largely retraced Cheng Fengxian’s traverse and did not survey new areas. The results of these surveys were ultimately published in the Batang Yanjing Xiangtu Zhi 巴塘 鹽井鄉土志 (Local gazetteer of Yanjing township in Batang), which also included three new maps, titled Menkong Quanjing yutu 闷空全境舆图 (“Map of Monyul territory”), Zayu Quanjing yutu 杂瑜全境舆图 (“Map of Zayü territory”), and Sangangquzong dajiang ximian yutu 桑昂曲宗大江西面舆图 (“Map of the western Yangzi river through Sangang Quzong”).
In line with the 1891 directives issued by the Huatuchu 畫圖處 (Office for the Preparation of Illustrations), these three maps were drawn to a prescribed scale, employed north as the reference direction, and marked latitude and longitude using a square graticule. Legends were displayed on all three in the left bottom corner, detailing the placement of mountains and passes, rivers and waterbodies, cliffs, fields, villages and human settlements, trails and plank roads, bridges, and garrisons in particular. They also followed the set format of appended explanatory text on the right side of the map, focusing on topographical features, the number of households, population, and resources in each of the three spatial projections of Monyul, Zayü, and the western Yangzi. They do not employ conical projection, and most of the mountains of the region are depicted in a style adjacent to traditional landscape painting. An apparent absence of the Yutu Zongju 輿圖總局 (General Mapping Offices)Footnote 32 or their equivalent in Batang and elsewhere along the Tibetan frontier makes it difficult to conclude where Duan Pengrui would have produced the final drafts of the maps, and whether he relied on his own team or had additional help. However, despite the lack of dedicated offices, by the nineteenth century the vocabulary of surveying and map-making had already normalized a great deal of so-called “hybrid” or intercultural practices and techniques,Footnote 33 enough that Cheng Fengxian’s surveys and Duan’s maps are not exceptional for their time. Instead, they are representative of general trends in late-Qing mapping, including the notion that physical aspects of the landscape such as rivers and mountains represented the natural limits of that space. Indeed as Zhao Erfeng directed Cheng Fengxian, “whether it is a mountain or a river, there must be a boundary. Once you have investigated each one by one, report back to me.”Footnote 34
Furthermore, since the General Mapping Offices usually provided personnel for agriculture and revenue surveys, it is also possible that an adaptable office such as the Salt Bureau took upon itself the brief of a mapping office, as is suggested by the appointment of Duan Pengrui, who was granted a juren 舉人 title in 1891, as Diaocha weiyuan zhoupan 調查委員州判 (Survey Commissioner and assistant magistrate) for the duration of the survey.Footnote 35 The employment of staff like Duan Pengrui is also evidence of the nature of institutional support for planned future surveying activities along the Tibetan frontier between 1905–1911.
That Support could not be sustained beyond 1911, however, and the proposed re-drawing of provincial boundaries by Zhao Erfeng on the basis of his subordinates’ survey work was never substantially translated into administrative policy. Among other proposals, Zhao Erfeng requested that Chayu 察隅 and Kemai 科麥 be established as new counties, both under the jurisdiction of the Border Affairs Ministery 邊務大臣衙門.Footnote 36 By April 1911 he had appointed Xia Hu as the county governor in Kemai and ordered further surveys of the regions lying west of the Dzayul, in addition to authorizing him with establishing identifiable chains of command among the local populace and issuing new door plaques to householders in order to regularize the new provincial boundaries.Footnote 37
Part 2: Organizing “Expert” Knowledge
Accounts such as Duan Pengrui’s were always notable for the level of itemized detail contained within them, organizing and enumerating as they did all possible aspects of the physical landscape: its topographical features, the flora and fauna, the mineral resources and local agricultural and market specialties, and the roads and networks of human settlement. The wealth of detail offered an invitation to the reader—at the twentieth-century Qing court or in any other temporal and spatial sphere—to marvel at the author’s immense knowledge of the landscape across all its attributes. Despite the fact that these surveyors were rarely native to the regions and landscapes they described in their accounts, they could nevertheless establish themselves as the local experts by demonstrating a near total knowledge of all things that lay above the earth and below it. Was this expertise always gleaned through first-person observation or were there other agents and intermediaries involved in the process?
There is generally scant mention of any actors other than the surveyors themselves and their immediate circle(s) in these accounts. The particulars regarding local involvement and assistance on these missions is hard to come by. Xia Hu’s records are the only ones from our three Chinese surveyors that make any mention of local intermediaries by name. Although Duan Pengrui himself was native to Yunnan, there is no evidence that he or the other surveyors held any kind of fluency with regard to Tibetan and that family of local languages. Interactions with the local populations or residents are mentioned selectively and chiefly in the course of extraordinary events rather than in the everyday mechanics of travel and exploration. And yet, in order for these non-local surveyors to have conducted the actual work of surveying new landscapes on foot, they would have required extensive local help. Unfamiliar as they were to the terrain and tongues around them, they could only carry out the missives they were charged with through the help of intermediaries. In turn, these intermediaries performed a gamut of duties crucial to the progress of a surveying expedition—ranging from the reconnaissance of lands to be surveyed, carting equipment and belongings, cleaning, handling and assembling technical equipment, cooking, setting up and clearing camp, basic caregiving when expedition “leaders” were indisposed or otherwise unable to continue their progress, translating and forming necessary connections with locals, their markets and their officials, arranging permits, recording scientific observations, defending equipment and personnel from idle human or animal curiosity and hostile intent, and other active mediations between the earth and the ledgers upon which information about it was being recorded.
These tasks all required interactions and relationships with a diverse group of intermediaries who could provide the information and (often literal) direction required by our surveyors. As with contemporaneous British colonial archives, these contributions are rarely recorded by Chinese officials, and if they do find mention, it is within the framework of performing “social” roles such as providing practical intelligence. These texts, therefore, organized and emplotted information, events, and interactions as narrative decisions that provided the cohesive framework for the presentation of the surveyor account itself as “expert knowledge.” Included in such narrative decisions were the codification of various interactions as the intellectual work of first-person observation, with the simultaneous relegation of the other agents of those interactions as non-intellectual, “practical” information. Their “expertise” thus emerged from a dynamic process through the claims that these surveyors made about the different types of people they encountered, the relative knowledges the locals possessed, and the networks within which those people and knowledges were embedded. And thus, the expert authority of the imperial surveyor could be bolstered through either the absolute omission or a more general characterization of the local intermediary as the non-intellectual Other.Footnote 38
Since these Qing-era frontier surveys did not keep or submit field diaries to their offices upon their return, it is unclear whether the extent of their dependence on local manpower and capital would have been fleshed out in written mediums other than official reports and gazetteers. At any rate, the genre of the difangzhi 地方志 (local gazetteer) involved the processing of information through several registers in order to establish the author as the arbiter and provider of authoritative knowledge about frontier spaces. Importantly, the preface to these locally organized gazetteers, treatises, and memoirs included historiographical information on previous works published about that place, including the names of their authors. Particularly by the nineteenth century, local gazetteers were compiled by and published under the aegis of local elites and literati rather than through the initiative of the Qing court itself,Footnote 39 while the prefaces would continue to discursively emphasize the links between local officials and the imperial center in Beijing. It was therefore not uncommon for gazetteers of even small Chinese counties to have multiple lengthy prefaces, demonstrating to the reader that the place was well-incorporated within the imperial order and that the knowledge of and about that place had a long literary history. The practice of including such prefaces also shows that these authors demonstrated their knowledge of these regions not simply as knowledge inherently possessed, but as validated through extensive citational reference to past gazetteers and other productions of geographical import.
So although Xia Hu’s survey reports could only be constructed through the specific information provided by local “leaders” (土官) such as Juegen 覺根 of Yuantigongla 原梯龔拉 Village (near modern-day Garzê Prefecture),Footnote 40 for example, it is ultimately Xia Hu, and Liu Zanting 劉贊廷—the Qing-appointed magistrate who compiled most of those reports into the county annals—whose singular voices constitute the official archive of the region.Footnote 41 The passing mention of Juegen is rare in that no other individual Tibetan, official or otherwise, is alluded to in the work of surveying. Duan Pengrui and Cheng Fengxian’s monographs are altogether devoid of locals except as tribal groups that hold the land and resources the imperial center desired to acquire.
The Qing boundary markers below in Figures 1 and 2 were photographed in 1911 by Major C. P. Gunter and William Michael Dundas during the course of the Abor-Mishmi mission.Footnote 42 British attempts to annex the entirety of the Brahmaputra valley since the early nineteenth century had manifested as successive military campaigns to subdue the local communities and inhabitants of the Assamese and South Tibetan culture regions. The Abor and Mishmi missions formed part of an escalated attempt by the colonial state to push the borders (or the “outer line”) of British India as far into Tibetan territory as possible—these crystallized in 1914 as the McMahon line.Footnote 43 Importantly, this and contemporaneous British-led missions in that region noisily claimed to survey previously unexplored territory, while directly following the traverse routes and survey maps that had been previously compiled by Tibetan and Himalayan surveyors in the employ of the Survey of India.Footnote 44
Figure 1. Qing boundary marker at Menilkrai (modern-day Walong town in Arunachal Pradesh) (Photo by C.P. Gunter R.E./Royal Geographical Society).
Figure 2. Boundary marker and photographer William Michael Dundas, 1911 (Photo by C.P. Gunter R.E./Royal Geographical Society).
The boundary markers were inscribed in Chinese and dbu-med (དབུ་མེད་ or “headless”) Tibetan, and informed the beholder in both languages that they were at the southern limit of the Dzayu region in Qing imperial territory.Footnote 45 In sharp contrast to the Chinese side, the right side is so messily painted that the top part of the marker where 大清國 has been inscribed into the Tibetan is nearly incomprehensible. The distortion of the Tibetan script (top to bottom instead of left to right) clearly suggests that the inscription was painted on after the marker had been planted into the ground so as to fit its shape, and that this particular inscriber was not habituated to adapting Tibetan to this format. As an “unpracticed” hand to such adaptations, then, the inscriber may have been a local, native speaker of Tibetan attached to the surveying party or a Chinese speaker attempting to reproduce written Tibetan. As I have shown, none of the Chinese surveyors were native to the region they were investigating. Were it not for artifacts such as this marker, it would be easy to conclude that these surveying missions progressed through the region with little to no interaction with the Tibetan and other local populations.
These absences in the archive are crucial to the mechanics of how our Qing surveyors established their own eyewitness accounts as the authoritative knowledge of the region. The enactment of their expertise depended on how efficiently these non-local investigators could establish, maintain, and communicate publicly the relationships they held with the kinds of culturally valuable objects and practices that informed the regime of geographical information valued and institutionalized by the imperial Qing state. As anthropologists of expertise have postulated, “becoming” an expert consistently relies on the creation and hierarchized curation of cultural objects (whether they be peoples, knowledges, or networks), as either more or less valuable depending on a discretionary judgment of how relatively inaccessible or obscure those objects might be to the non-specialist audience.Footnote 46 The emphasis on local intermediaries and assistants would make the work of exploration and surveying more accessible to the reader, suggesting that knowing the region was not the a priori possession of erudite or even embodied knowledges, but the act of meticulously collecting local and situated networks. This incentivized the creation of distinctions among the people encountered by our surveyors, as well as asymmetries between people and objects, prioritizing observations about relatively obscure objects (such as the landscape), and the absences of the locals except as holders of economic objects that would bolster imperial power in the region.
To render legible the knowledge of the Dzayul region to, in this case, other non-local literati and court audiences, the work of surveying and mapping chiefly created a new interpretive framework through which to view this landscape. The institution of new counties, administrative units, and the standardization of place-names were thus all fundamental to determining how this area should be read and interacted with. The facets of being non-native and distant to the space that was being inscribed as the academic object in Qing maps, atlases, and gazetteers, here, became crucial to the recognition of the erudite bureaucratic official as the local expert. So much so that the rendering of local names and places into Chinese has now become the only point of access for the historian into the past of these regions, and the Tibetans, their place-names and people, are lost.
Part 3: Tibetan Experts of their Himalayan Frontiers
The circulation and dissemination of geo-knowledge about the Tibetan plateau and its border-worlds between the seventeenth to early twentieth centuries continued to be dominated by the textual genres of the lam-yig (ལམ་ཡིག་) (records of the road/route), gnas-yik (གནས་ཡིག་) (records of the place), and the gnas-bshad (གནས་བཤད་) (explanation of the place).Footnote 47 These guidebooks contained navigational information for pilgrims travelling to one or more pilgrimage sites, narratively presenting a two-dimensional knowledge of place: one spatial and one atemporal.Footnote 48 The narrative construction of a site as “holy” in these Tibetan works drew on more than a simple focus on the situated nature of an otherworldly or transcendent power. By way of practices such as pilgrimage, the lineage of such sites becomes a focus for the various manifestations of human power.Footnote 49 The division between the planes of human and non-human existence is not particularly sharp in the folkloric and literary imagination of most Tibetan works, rather, the physical environment in both its organic and transcendent dimensions is imagined as inhabited by a multitude of deities and spirit forces, making every spatial plane at once sacred and political.Footnote 50
Most works of Tibetan geography presented information in a fashion that foregrounds and reiterates an unbreakable link with an antique Tibetan civilization and a Buddhist world order, a technique which resonated with Chinese works of geography. While these techniques of coding information are exactly what make Tibetan works notoriously difficult to date, they ensured that the texts could forward claims to authoritative knowledge. Such a literary device, where works opened as taking precent from and expanding the scope of prior knowledge—vigorously emphasizing continuity—allowed critical scholarship to circumvent extant intellectual orthodoxies within Chinese and Tibetan knowledge systems. This also ensured a wider readership and the sustained dissemination of new works of geographical interest.
Monasterial productions historically narrativized perceived threats to Buddhism in Tibet as the erosion of vitality in the mountains surrounding the plateau. In this way, the Himalayas were importantly linked in ecclesiastical literature to the sustenance of a Buddhist world order in Tibet.Footnote 51 The conventional representation of “hidden lands” (sbas yul སྦས་ཡུལ་, hereafter beyul), in both Tibetan and Western accounts, is of a utopian space that provides refuge in times of war and social disorder. These places are repositories of spiritual power and sites of physical and metaphysical pilgrimage, where practitioners might also retreat to meditate or otherwise engage in spiritual practice.Footnote 52 The various beyul were scattered throughout the Himalayas, woven through the chain of the southern borderlands of Tibet into modern-day Nepal and India. This was no accident, as we shall see.
The beyul trope chiefly inscribed the histories of the philosophical traditions that were marginalized by the Gelug (dge lugs དགེ་ལུགས་) school of Tibetan Buddhism.Footnote 53 These expansive lands were narratively imagined as secreted away in valleys blessed by PadmasambhavaFootnote 54 in the eighth century, refuges that could only be “opened” in times of historical upheaval, or similarly approximate turmoil, by specific tertön (gter ston གཏེར་སྟོན་).Footnote 55 The role of tertöns was central to the transmission of Nyingma (rnying ma རྙིང་མ་) knowledge productions, as these leaders alone could claim legitimacy as interpreters of terma (gter ma གཏེན་མ་),Footnote 56 hidden texts that would reveal the location of beyul at historically appropriate moments.Footnote 57
The practice of inscribing historical legitimacy into a text by claiming a legacy of continuous burying and revelation is common to both Indic and Tibetan religious literary traditions. In their capacity as tertöns, Tibetan mystics and other religious practitioners demonstrated their ancient historical legacy, and thus their authority, by performing memory triggers: engaging in a visionary treasure hunt across the Tibetan landscape for scripture and other talismans (terma) purportedly concealed by Padmasambhava, the discoverers “recalled” teachings and prophecies received directly from Padmasambhava in their previous incarnation(s).Footnote 58 After meditating on the unearthed treasure, usually in the form of statues or scrolls, the tertöns would then compose volumes of texts relating specific directions about spiritual practice, as well as other aspects of daily life. This particular enactment of expertise also rested on the demonstration of a mastery of the terrain as it may have been represented through both embodied and literary productions of knowledge.
Tertöns were never the sole composers of these texts—the process of composition could include oral dictations transcribed by one or more disciples, or the retrospective claim to receiving such dictations years and even centuries after the tertön had died. In this way, a nineteenth-century text could successfully masquerade as one composed centuries earlier, ostensibly “discovered” by a leader in a time of political or spiritual crisis, who could, on the basis of the “ancient wisdom” in this text, then lead communities of people to settle in new lands that the tertön himself would “open.” And indeed, the most prolific period of guidebook-writing and publication came between the late nineteenth and early twentieth centuries,Footnote 59 coinciding with the peak of British and Qing interest in Tibet and the Himalayan borderlands. These guidebooks laid claim to lands that had already been or were being demarcated by these imperial states as British Indian or Qing frontier territory.Footnote 60 And while the precise number and distribution of the hidden lands can somewhat differ depending on the guidebook source under analysis, these texts all locate the ring of the Himalayan borderlands—ranging from the upper reaches of eastern Tibet (historical Dokham), circling through the southern borders of the Tsang and Ü provinces into Ngari Prefecture—as the landscape where beyuls are embedded. These texts further coincide in reporting the regions of Pemakö (Padma-bkod པདྨ་བཀོད), Abor and Mishmi country (spanning parts of the historical Dzayul region), Sikkim (‘bras mo gshongs བྲས་མོ་གཤིངས་), and the Chumbi valley, as the most important beyul.
Tibetan geographies did not always have a linear or consistent delineation of the “borderlands.”Footnote 61 By the nineteenth century, guidebooks in the beyul genre explicitly identified the Himalayan mountains as the savage frontier where Tibetans—guided by tertöns—were required to settle, in order to expand Buddhist civilization. The Himalayas were not the uninhabitable limits of this Tibetan imperial imagination, rather, the mountains were the spaces that Tibetan populations were required to settle in order for Buddhism—interchangeable in these texts with the concept of civilization itself—to flourish.
These contradictory places of serenity and utopian peacefulness were also regions of savagery and extreme adversity. Pemakö, for instance, was visualized as both a paradise of medicinal groves filled with rainbows and a gathering place of devilish spirits, murderous tribes, hungry tigers, bears, leopards, venomous snakes and leeches that prevented pilgrims and refugees from gaining access to the beyul.Footnote 62 The very journey to “open” a beyul, narratively framed as a journey through heaven and hell at the same time, effectively functioned as a hero’s quest for tertöns.
Guidebooks inscribed with the legends involving Pemakö likely spread in Tibet through the early to mid-seventeenth century.Footnote 63 As a result, tertöns of the Nyingma school systematically explored Pemakö in the late seventeenth and early eighteenth centuries, cementing its status among the most famous of the hidden lands.Footnote 64 For the adepts of the Nyingma and Kagyu orders, the journey to this valley situated in the Brahmaputra gorge became the ultimate obsession well into the twentieth century. The journey to Pemakö came to stand as an allegory to the path to enlightenment itself.
A series of locations along the Brahmaputra River were interpreted by successive tertönsFootnote 65 as the five chakras of the goddess Vajravārāhī, thusly implying that the beyul of Pemakö in its entirety represented her body.Footnote 66 Other versionsFootnote 67 also describe Pemakö as the body of a nāga or serpent deity and the goddess as the Nāga Tamer (klu ‘dul ma ཀླུ་འདུལ་མ་) suggesting that the association with Vajravārāhī was not simply to describe a bounded nature to holy ground, but rather an allusion to the land as domesticated or even “civilized” by the practice of Tantric Buddhism. This particular narrative version would appear to recall the legend of the supine demoness, whose body pinned down by Songstan Gampo’s temples became the terrestrial representation of a civilized Tibet itself.
Frederick Jackson Turner’s thesis of the frontier as the “meeting point between savagery and civilization”Footnote 68 was certainly resonant of both British and Qing imperial discourse in the Himalayas and Tibet. Contemporaneous Tibetan knowledge productions also traded in these familiar imaginations of the indigenous inhabitants of the Himalayan borderlands as brutish tribesmen that would, like the wild landscape itself, ferociously rebuff the overtures of civilizing influences of Buddhism and its followers. The ethnic matrix of indigenous Assamese, Nepali, Bhutanese, and Sikkimese peoples were collapsed into the groups of Mon-pa མོན་པ་ and Klo-pa ལྷོ་པ་ (or Lho-pa, literally “people of the south”), both terms originally meaning “barbarian” in the Tibetan symbolic world.Footnote 69 The so-called opening of the space of the beyul, then, was only possible through the conquest, submission, and eventual conversion of the local populace, the indigenous peoples of the Himalayan borderlands.Footnote 70
The “civilizing burden” was further fulfilled through the building of border-taming temples to foster communities that could propagate the practice of Buddhism, as well as the introduction of sedentary agriculture. Pemakö’s “head” and “throat” chakras in the lower Tsangpo valley were “opened” in late eighteenth or early nineteenth centuries by Chime Dorje (Kun bzang ‘od zer Gar bang Chi med rdo rje ཀུན་བཟང་འོད་ཟེར་གར་བང་ཆི་མེད་རྡོ་རྗེ་)Footnote 71 and the “heart” chakra, further down the Brahmaputra river, by another contemporary, Orgyan Lingpa (O rgyan ‘Gro ‘dul gling pa ཨོ་ནྱན་འགྲོ་འདུལ་གླིང་པ་ཨོ་).Footnote 72 The establishment of temples and monasteries in these locations ran parallel to the settlement of Tibetan Buddhist populations in the region, displacing the local Abor, Tani, and Mishmi peoples.Footnote 73
The trope of the beyul was the single most powerful device for Nyingma and Kagyu authors to write the history of their orders between the seventeenth and early twentieth centuries, during which time Gelug hegemony was consolidated over central Tibet and increasingly over Amdo. The spread of Gelug monasteries across Ü, Tsang, Amdo, and even Kham, especially during the reigns of the fifth Dalai Lama (1617–1682)Footnote 74 and his regent Desi Sanggye Gyatso (1653–1705),Footnote 75 created the first recognizable unified territorial iterations of “Tibet.”Footnote 76 Within this conception of Tibet, the Nyingma, and to a lesser extent, Kagyu schools became increasingly marginalized. While the Nyingma school never combined the ecclesiastic and temporal authority that the Gelug, or before them the Sakya, orders achieved, tertöns were able to preserve and propagate their knowledge traditions by gaining the patronage and alliance of powerful imperial agents.
The kingdom of Spo bo སྦོ་བོ་ (also known as Powo, Poyul, and Pomê, henceforth Pomê) in South-Eastern Tibet, a semi-autonomous principality that functioned largely independently of the Lhasa administration until the early decades of the twentieth century,Footnote 77 became a home for many tertöns seeking to settle communities in the Himalayan beyuls and Pemakö in particular. Situated in the Eastern Himalayas right at the entrance to the Brahmaputra gorge,Footnote 78 the Pomê government cultivated strategic alliances with the Nyingma and Kagyu schools.
Throughout the seventeenth and eighteenth centuries, the tertöns who travelled to and settled the region of Pemakö did so under the patronage of the rulers of Pomê and often at their behest. Tertöns such as Chorje Lingpa, Chime Dorje, and Orgyan Lingpa discussed above acted in effect as royal preceptors of the Pomê kings, ruling in their name and under their banners.Footnote 79 The anthropologist Geoff Childs has argued that beyul texts during this period essentially frame the hidden land itself as a small kingdom where royal lineages would be preserved under the guidance or rulership of the tertöns themselves.Footnote 80 In this manner, a key aspect of the beyul trope was its political use as a device not only to preserve the Nyingma and Kagyu lineages, but to articulate an alternate political landscape of Tibet itself.
This articulation was not merely literary and conceptual. By the late nineteenth century, the interest in Pemakö had shifted to finding the location of the goddess’s womb chakra,Footnote 81 in the very same regions where British India had already begun to draw up borders with occasional reference to the Lhasa administration. The Kanam Depa (the title of the kings of Pomê) further set up a series of local offices (rdzong རྫོང) throughout Pemakö and had begun to collect taxes from the local settlers in their own attempts to structure parts of the region administratively.Footnote 82 These taxes were paid in the form of butter or pelts to the Pomê king in exchange for soldiers to protect the settlements against attacks by the local Klo-pa.Footnote 83
Between 1902 and 1908, over two thousand Tibetans set out from Kham, Derge, Pomê, and other parts of Eastern Tibet, escaping Qing expansion into Kham under Zhao Erfeng, to settle further regions in Pemakö. The first wave of these refugees settled in the Dri, Matun, and Tsu valleys, where they bought land from the Chulikatta Mishmi and appeared to able to maintain fairly cordial relations.Footnote 84 Over the following years, the Kanam Depa sponsored the movement of further communities of settlers into the Chimdro valley in northern Pemakö, into what the British already considered Indian territory. Led by Jedrung Jampa Jungne, a Nyingma tertön from Riwoche in Kham, these settlers built a border-taming temple they named Karmoling, near Mipi in modern-day Arunachal Pradesh.Footnote 85
To the eye of the imperial ethnographer, there was little difference between the Mishmi and the settlers from Pomê, and conflicts between them were racialized as “petty tribes” squabbling over territory. Although the Tibetans come to us in colonial narratives as a harried and destitute people who have stumbled into an environment they have little control over,Footnote 86 they may well have seen themselves and the tertöns they followed as pioneers charting a brave new civilizational course through savage lands. The conflicts between the Mishmi and the Tibetans, then, were conflicts over land capture where indigenes fought displacement and the defense of their livelihoods against settlers who explicitly intended to wholly remake the local social orders.Footnote 87
In this context, Pemakö must be read as a political space within a Tibetan settler imagination, independent of and prior to Qing and British imperialism. Furthermore, in its function as a regenerative space that could purify the practice of Buddhism in order to restore the true faith to the center, the conquest of Pemakö and the desire to retain this land were fundamental to a non-Gelug conception of what a Tibetan state could look like. This alternate vision of a Tibetan state never truly had the opportunity for fruition in the face of British and Qing armies in the Himalayas, and the kingdom of Pomê itself was forcibly integrated into the ambit of the thirteenth Dalai Lama’s government in 1928.Footnote 88 Nevertheless, British and Chinese exploration of this region scrubbed the landscape clean of these layered histories of habitation, settlement, and conflict, flattening it instead as a terra nullius passively awaiting colonial intervention.
Conclusion: Masters of All They Surveyed
The archive of “authoritative” geo-knowledge about the trans-Himalayas wove through complicated processes of place-making in which a highly stratified matrix of agents circulated and interacted. Against the constraints of imperial systems predicated on coercive technologies to identify territory, the enactment of particular and selective forms of knowledge, skills, language, and social relationships ultimately framed the recognition of who could be judged an “expert” of the region or its representation. In turn, the endorsements of particular kinds of expertise oriented the practices of geography, cartography, and ethnography as they developed as imperial “sciences” through the nineteenth and twentieth centuries. The broader narrative of British and Qing imperial competition elides a much more complex erasure of the processes involved in the ultimate transformation of the Himalayan mountain landscape.
By the turn of the century, the production of geographical knowledge in China, while still a robust literary tradition, had also begun to be professionalized in ways that drew greater and greater authority from field-based survey work. An evolving vocabulary of skill-building for surveyors was produced in a new regime of education at military academies, and institutions like the Huatuchu and the Yutu Zongju structured ways of seeing and knowing into the technical language of disciplinary protocols. The imperial state created new kinds of interpretive frameworks for organizing spatial knowledge, institutionalizing selected enactments of expertise as specialized domain knowledge, and rejecting other registers of knowledge production. In 1911, Cheng Fengxian, Duan Pengrui, and Xia Hu also established themselves as experts of the Dzayul region by finding their voices in this new regime of citational practice by diminishing and erasing the provenance of embodied forms of knowledge, the situated histories, and the mastery of the terrain possessed by the local Tibetan, Abor, Mishmi, Tani, and other indigenous peoples.
The resultant imperial surveys largely claimed the limits of natural features as the “natural” or even “traditional” boundaries of the imperial state, against local knowledge productions that framed those same topographical features as connectors rather than dividers. The Qing archive of geo-knowledge could only be assembled through erasing the provenance of, and decontextualizing the information appropriated from, these alternate and frequently embodied knowledge systems, but it was nevertheless the erudite bureaucratic official alone who could claim to hold specialized regional knowledge, while the indigene was cast as the non-intellectual laboring body. At the same time, these systems of “indigenous” knowledge can now only be accessed through the colonial archive(s) and interpretive frameworks, and as mediated by different kinds of imperial agents.
Although the non-dualist coding of locational information that constructed the situated nature of sites included in Tibetan guidebooks has generally invited the tendency to dismiss the seeming empirical shortcomings of this genre of literature as exotica that is divorced from historical or material forces, imperial knowledge had to consistently reckon with the disseminated forms of this corpus. From the Zanskar range towards Kashmir down towards Nepal, through Sikkim and Bhutan into modern-day Arunachal Pradesh, the Himalayan borderlands that formed the “wild frontiers” of Tibetan civilizational ambition were equally the savage country that the British Indian and Qing Chinese states desired to acquire and domesticate. Regardless of when these guidebooks might have been written, they were, in the nineteenth and twentieth centuries, living texts that competed with British and Chinese territorialization in the Himalayas. The delineation and claiming of territory through the use of the beyul as a mapping practice, contemporaneous with colonial surveying expeditions, reveal a far more layered picture of the state-building projects that transformed the Himalayan mountain landscape. The frequently arbitrary placement of imperial borders produced the very condition of Tibet as a “peripheral” frontier zone caught in between the ambitions of state and capital. The Tsangpo-Brahmaputra and Dzayul-Lohit surveys would come to frame the contestations over the McMahon line in 1914, ultimately locking the independent nations of India and China into a renewed commitment to imperial frontier management.
Acknowledgements
My thanks to the many people who commented on and discussed with me at length the various drafts and ideas informing this article: Idriss Fofana, Nishita Trisal, Eloise Wright, Gray Tuttle, Cameron Foltz, and Palden Gyal. The invaluable editorial support of Qiao Yang, Sarah Schneewind, Gina Grzimek, Rebecca Schmitt, Spencer Forbes, as well as the anonymous reviewers at the Journal of Chinese History helped me make final and important improvements.
Competing interests
The author declares none.
Open access
