The Paper Road

The Eyes of Others

A man is suspended over the river. Leather straps bind him to a half-cylinder of bamboo that slides on a rope of twisted bamboo strands greased with yak butter. Having plummeted to a point over the river's center, he will haul himself to the opposite bank, hand over hand. The edges of the strands are sharp, so the crossing is painful as well as physically demanding. The photograph was included in Acting Consul G. Litton's secret report to the British Foreign Office on his journey to the Upper Salween River.1 It is filed in the archives of the Royal Geographical Society with many other images of suspension bridges over the great rivers that run in parallel gorges through northwest Yunnan: bridges of cane, ropes and planks, timbers and iron chains. The back is stamped with the photographer's name, George Forrest, and the date, 1905. At that time, in the confident opinion of the Royal Geographical Society president, the Salween remained the last great riverine puzzle on earth, its northern reaches yet unexplored by any European.2

The drawing looks almost like a cartoon-strip version of the photograph. The rope is stretched between two trees; the bamboo runner and inverted human figure are clearly depicted; a simple tree stands to the right; the whole is surrounded by a square frame. The drawing is from the title page of a hand-written manuscript called Lònv. The composition within the frame is this word in what is now known as dongba script. The left side is the word lò, meaning "to cross over," the first syllable of lòk'ö, to cross a rope bridge. The tree on the right is the word nv, which designates a ritual effigy. The manuscript titled Lònv was written to be read at a funeral for a child. Joseph Rock bought it in the early 1930s in the Lijiang valley. It now resides with 1,114 other dongba manuscripts in the Staatsbibliothek in Marburgh, Germany.3

G. Litton walked up the Salween, or Nu, intending to draw a line between two empires. He had recently been appointed the British representative to a joint boundary commission charged with fixing the northern border between China and Burma. His Chinese counterpart was Shi Hongshao, the new trade commissioner (daotai) for west Yunnan. An 1897 treaty, in which the Qing court granted the British a trading post in the border city of Tengyue, had left this issue unresolved, specifying only that a fixed boundary would be surveyed in the future. Over the next two years, cartographers for both sides proposed five different boundaries, synthesized in a "five-coloured map" for the south, in the Wa states, and another "five coloured map" for the north, west of the Salween.4 The British Foreign Office had declared that it would be "for the advantage of both countries and of their mutual commerce that British jurisdiction should be established over the whole of the upper basin of the Irrawaddy."5 This in mind, Litton argued that the huge, unsurveyed mountain range rising from the west bank of the Salween, the Gaoligong range, was a "natural and ethnological divide" between the two empires.6 In response, Shi Hongshao restated the Qing court's position that the local rulers of the peoples Chinese ethnologists called Nu, Lisu, and Qiu, who occupied that range and many territories west of it within the Irrawaddy basin, were tusi, native hereditary officials enfeoffed by Ming and Qing emperors, or fuyi, hereditary officials granted titles by the Ministry of War. Formally, these rulers were under the jurisdiction of the district government in Weixi, a town on the Mekong, and their polities were Chinese territories.7 Litton believed to the contrary that no actual relationships existed between the district government and the Nu, Lisu, and Qiu chiefs in the region. It was to prove this that, in October 1905, accompanied by two soldiers from the Indian Army and tens of Lisu porters, each carrying sixty to seventy pounds of rice, he walked up the gorge of the Salween.8

His other companion was George Forrest, on an expedition to Yunnan for the British seed firm Bees Ltd. Forrest had already been in Yunnan nearly a year. He had made two extensive tours of Yunnan's northwest, the first with Litton, the second on his own-with a cook, a groom, three servants, and two muleteers ("I cannot travel with less," he wrote his fiancée.)9 His second journey had been particularly eventful: he had barely survived a Tibetan attack on the French mission at Cigu on the Upper Mekong River, losing all his possessions, including an extensive collection of specimens. Two months later, though barely recovered, he happily accepted Litton's offer to make him the naturalist and photographer of this expedition to a new and promising territory. [Figure 8 about here]

The two British men began the expedition with a sense of excitement tempered by sober purpose. They were performing a vital task for the empire-even if in this rather unimportant periphery. To them, their collaboration was perfectly natural: naturalists had taken part in the military and administrative surveys of each new territory acquired by the empire in India.10 They investigated the gorge's climate, botany, zoology, and geography, and they made ethnological observations of its human inhabitants. The latter, crucially for Litton, were "wild Lisoos with poisoned arrows" under "no control of any sort ... by any Chinese or other chief."11 At the bamboo rope bridge, Forrest took a photograph, and they discharged their firearms, subduing into "awestruck silence" the crowds of Lisu villagers who seemed to be vying to control their crossing. They then climbed up the east bank of the gorge to the summit of the Mekong-Salween divide, where they accomplished their journey's main objective. "Here a surprise awaited us," Litton wrote the Foreign Office, "for the view to the west was perfectly clear, and the whole of the great Salwin-Irawadi divide was spread out before us. From a little below the pass, the range could be followed to the north as far as the eye could reach, until at a distance of about 100 miles from where we stood, and in approximate lat. 28° 30' N., it was merged into a huge range of dazzling snow peaks trending westwards."12 "Not the least doubt remained in my mind that here was the true frontier between India and China," he added: it was a "vast wall."13 Litton did not live to press his case, dying of malaria only weeks after his return, but his description was repeatedly quoted in insistent British proposals that the range dividing the Salween and Irrawaddy watersheds be the border. Though Qing negotiators vigorously opposed these proposals, the British treated the divide as a "provisional border." It became the de facto border in 1913, when British troops occupied the village of Pianma (Hpimaw), west of the divide, despite furious nationalist protests in China.14 The border was not finally established until long after Burmese independence in 1960.15

As for George Forrest, exhausted by the journey he let Litton's words take the place of his own in letters home and in a paper to the Royal Geographical Society.16 He was accustomed to being struck silent by views of glistening snow covered peaks, "beyond conception" or "beyond description." They signaled to him the limits of thought and language. He would forget about that distant, shining range for many years as he explored other parts of Yunnan. But eventually, inspired by his indigenous collaborators, he would remember it: as the sign of something always beyond his grasp; as a solution to the great phytogeographical puzzle of the genus Rhododendron; as a transcendent Eden, where the difficulties of engagement and representation that this landscape presented might finally melt away.

When a child died in a Naxi home in the Lijiang valley, her parents sent for a ritualist, a dongba, who helped wash the corpse, dress it in new clothing, and place it in the coffin. A little later, the parents held a funeral in their courtyard, attended by a few kin and friends. The dongba read several manuscripts. Only Lònv was written specifically to send a child off on her final journey. It seems to have been rare: many funerals for children probably did without it. It described the child's soul, unable to eat, unable to see and hear, unable to use hands or feet, unable to clothe itself. It spoke of the child's journey across a rope bridge separating the world of the living from the world of the dead. It told of how a dongba cut the bridge with a knife, severing the link between the two worlds in order to prevent parents or siblings from following. The manuscript specified the materials to be used: sacrificial animals, black clothing, a strip of blackened hemp for the rope bridge, a fir branch to which the rope was tied. Other texts described the soul's journey, guiding it north through the mountains of northwest Yunnan, toward the lands from which the ancestors had come, and eventually to the great mountain Ngyùná Shílo Ngyù, origin of all things, where the worlds of mountains and forests met the worlds of the gods.

In the early twentieth century the upper Nu (Salween) was about two months travel from the Lijiang valley. There were many rope bridges in closer places, along the Lancang (Mekong) and Yalong rivers. But all these had two ropes, each with one end higher, so the traveler could slide all the way across without resorting to the painful inverted crawl depicted in the Lònv manuscript. The distant upper Nu, with its single-rope bridges, lived in the ritual imagination of Naxi farmers and traders.17 Many Naxi from the Lijiang valley traveled long distances for trade: a late Qing gazetteer mentions that Naxi came over one thousand li to sell items to the Nu, who inhabited the far upper reaches of the river that bears their name.18 The Nu, whom Naxi called Nun, their land on the far upper Nu (Nundǜ), and the Lisu (Lissuu), who lived in the part of the gorge Litton and Forrest investigated, were all represented in the written lexicon of Naxi ritualists. In particular, a small ceremony to perform rain was performed frequently around Lijiang in the early twentieth century. In this rite, held by a spring, dongba invited the spirits who brought rain to descend from all the great rivers and lakes in four directions. In the west, they came from the lands of Nundǜ descended to the town of Weixi on the Mekong, circled the mountains to the north and east, and then entered the Lijiang valley. The origin of this journey, and its highest point, was the great snow mountain Nundǜ Gkyinvlv, the center of that "huge range of dazzling snow peaks" that Litton and Forrest spotted at the apogee of their expedition, shining from a hundred miles further north.19

A photograph and a drawing. Each might be said to participate in a separate archival regime. They are archival in the most obvious sense. They are both fragments of large collections that came together in focused historical periods, in projects that followed quite precisely defined rules. I find the word "regime" appropriate because of its use in physical geography to connote all the factors that create the quality of a river's flow: the amount of rainfall, the type of bedrock, the shape of the channel-everything that makes the river meander, braid, flood, disperse into deltas or join with other streams. In a similar way, an archival regime includes all the rules, habits, and ideologies that decide how perceptions are transferred to paper, collected, and stored. It includes habits and ideologies of seeing and walking, of inscribing and photographing, of collecting, shipping, storing, organizing, and reading.

The British expedition up the Salween was guided by what Bernard Cohn called a "survey modality" of investigation. This mode of summing up a landscape and its inhabitants in a series of comprehensive reports had been central to the colonization and administration of India. Litton's report was added to the heaps of documents written by colonial explorers and administrators in India that had been piling up in the archives of the British Museum, the Royal Geographical Society and related institutions for over two centuries. The British consul who replaced Litton sent all Forrest's photographs from the expedition to the Royal Geographical Society. His images of bridges were filed in drawers marked "bridges," of people in drawers labeled "types," of crossbows in drawers stamped "weapons." Thomas Richards has argued that this kind of archival activity, more than any process of administration or governance, was the central project of the British Empire during the late nineteenth and early twentieth centuries. After the mid-nineteenth century, British imperial power found unity and coherence in an archival myth: the idea that institutions of knowledge production could forge the millions of facts flowing in from all corners of the world into a coherent whole. In an empire where administrative institutions were thinly spread, this myth of archive took the place of actual civil governance in many places. Litton and Forrest participated wholeheartedly in this myth as they walked up the Salween. Despite the complete absence of any British civil authority in the region, the failure of numerous plans for a railway linking Burma to the Yangtze, and the Foreign Office's stated abhorrence for any further expansionist adventures, they both looked forward to the immanent British annexation of Yunnan. Even so, this would be the last time Forrest would imagine himself directly involved in the project of imperial administration. After Litton's death, his activity would be guided by a more limited regime within the survey modality of investigation.

The Lònv text was a fragment of another archival regime. Naxi ritualists probably began writing manuscripts in pictographic script sometime during the Ming dynasty, as armies from Lijiang swept through the province's western regions. During the eighteenth and nineteenth centuries, they wrote and copied these books with great assiduity. In the early twentieth century, tens of thousands of manuscripts were stored in the attic libraries of dongba in and around the Lijiang valley. Most were written to be recited during ritual performances; some were for divination and astrology. Many rituals took the form of driving or guiding inhuman entities over the landscapes of northwest Yunnan, and many texts contained long lists of places, some extant, some once extant, some belonging to other worlds. As a whole, the archive of dongba texts contained an extensive geography of northwest Yunnan, far more detailed, at the opening of the century, than any other representation in maps, reports, or herbaria. This archive too was underlain by rules, habits, and ideologies of seeing and walking, writing and collecting, storing, organizing, and reading. Dongba did not create and organize knowledge for purposes of rule, though perhaps they once might have. They wrote and copied texts to regulate the social well-being of people and communities by manipulating relations between human and nonhuman social entities. Their mode of investigation did not survey the landscape. It auscultated its depths, threw out lines of communication to its hidden presences, and divined traces of a past that had vanished from its surfaces.

One might think that each of these two regimes would be incomprehensible from the point of view of the other. And indeed, there is no evidence that either directly incorporated knowledge from the other. Forrest did not find routes, place names, or plant names in dongba texts: he was barely aware of their existence. And no dongba altered any text to include botanical or geographical knowledge created by Zhao Chengzhang and the twenty-five to thirty others from Nvlvk'ö who worked with Forrest. Nevertheless, the two regimes came into close contact for nearly three decades, and at some points of friction they very likely shaped or deformed each other.

By the late 1920s, Zhao Chengzhang, Zhao Tangguang, Li Wanyun, He Nüli, and the rest of the senior generation of explorers from Nvlvk'ö were masters of the taxonomical botany of northwest Yunnan, in large part their own creation. They were expert at keying out plants; they knew the slight differences that separated some species and varieties; they knew far more than anyone else in the world about the geographical distribution of alpine species in the region. By 1921, Forrest and Zhao were having "very heated arguments" about particular species. In one case Zhao and several others maintained, against Forrest's strong objection, that the rare tree-sized Rhododendrons fastidium, giganteum, and protestum were all distinct species. Despite his pride in his knowledge of the complex genus, his stubbornness in argument, and his tendency to tyrannize his employees, Forrest had learned to have great respect for their taxonomical expertise. In his report to his sponsors, he left the issue open, reporting only the differences of opinion.20

But Zhao and his fellows were experts in other realms as well. In many ways, dongba culture had been in decline since the eighteenth century. It had largely moved out of the towns of the Lijiang valley into the surrounding mountainous countryside. Located just above the edge of the valley, their village belonged to this drier, poorer, mountainous hinterland. But it was also in the orbit of the market town of Baisha, long the most important center of dongba learning. Nvlvk'ö and the other villages in the southern foothills of the Yulong range, though small and poor, contained the greatest concentrations of dongba and libraries of dongba manuscripts remaining within the loop of the Yangtze. It was from Nvlvk'ö that the dongba archive moved out larger world: when Joseph Rock began to collect dongba books there, he had a library of hundreds in within days and thousands in a few months.

The money they earned from their quest for flowers made Zhao Chengzhang and his colleagues their village's most prosperous and important citizens. As elders, they officiated at village-wide events like the yearly muanbpò ritual to regulate the village's health; they were the central guests at funerals; and they were in a position to sponsor rituals to cleanse their houses, cure their kin, defeat slander, and overcome drought. Dongba books were central to all these activities: from just a few to as many as seventy were read aloud at each. There is no evidence that Zhao or any of his colleagues were trained as dongba, but it is certain that they listened to many key books being recited many times: they all had a general knowledge of the dongba corpus. Despite the slow general decline of the dongba cult, it was still, for these men, a deep source of knowledge, attitudes, and orientations, particularly about its central protagonist, the living, spirit-inhabited landscape of northwest Yunnan.

In Part 1, I place these two archival regimes side by side in the context of Forrest's and Zhao's explorations of northwest Yunnan. For the men from Nvlvk'ö these regimes were very much coexistent. But to understand the specific ways they coexisted and, perhaps, intersected, can only be a speculative exercise. As I mentioned in the Introduction, almost nothing written by the first generation of explorers from Nvlvk'ö remains: there are only Zhao's four maps and two scraps of ledger, a few labels on bird specimens, and one brief, crudely written thank-you note. Any ideas about their relations with the gorges, mountains, and meadows they wandered have to be pieced together from two very different sources. On the one hand, there are Forrest's letters, reports, notes, and photographs, tightly focused on the botanical enterprise. On the other, there is the massive archive of dongba texts, collected within the Yangtze loop, supplemented by a small descriptive and ethnological literature about the Lijiang valley and the Baisha region.

Tacking between these sources, I make several conjectures about how the dongba archive might have rubbed up against, and created new and compelling possibilities for, the phytogeography of northwest Yunnan. In particular, I show how the dongba corpus might have shaped the decade-long quest of Forrest and his mentor Isaac Bailey Balfour for the center of origin of the genus Rhododendron. I speculate about how the encounter between the dreams of geographical botany and the words of dongba texts might have tilted that search north and west, focusing it finally on that gigantic range that Forrest saw glowing in the distance during his walk up the Salween river with Litton in 1905.

These arguments are the path I have chosen to trace through the geographies of these two archival regimes. Route maps such as these, even if limited in scope, are perhaps the best way to come to know such a geography. The alternative-maps that pretend to a comprehensive domain view-would fail this vast, varied, complex, and interesting landscape in important ways. Along this path, I open up some of the questions that animate the rest of this book. How do rules about how we must perceive inflect how we walk and see? How do perceptions inscribed on paper become interleaved with the substance of the earth, to inflect other perceptions? How in this process does the earth emerge into social being? In what ways might this social being serve as a resource for experiences that circumvent established ways of thinking and living the divides we make between the social and the natural?

Learning Race

He had never quite found his footing in the British class system. His parents were of the ranks of workers, clerks, shop assistants, displaced tradespeople, and aspiring, mobile entrepreneurs who had wreaked massive transformations in class relations during the late nineteenth century. His own opportunities to move within the class hierarchy-to gain a secondary education, to try immigrating to Australia, to become a colonial explorer-were due to unforeseen gifts or patronage from others of superior class position. In India and Burma, colonial travelers and functionaries reoriented themselves to accord with an immediately apparent cartography of difference. As Partha Chatterjee has put it, colonial states operated through a "rule of difference," and in India the salient differences were racial.21 Ann Stoler suggests that the work of drawing and redrawing maps of racial difference produced new middle-class sensibilities, organized around images of racial purity, sexual virtue, and proper masculinity.22 This cartography gave imperial men who moved between metropolitan and colonial societies while occupying underprivileged class positions in Britain new opportunities to recraft their class identities. "White Englishmen were able to use the power of the colonial stage to disrupt the traditional class relations of their country and enjoy new forms of direct power of subject peoples," Catherine Hall writes. "At the same time ... their own identities were ruptured, changed, and differently articulated by place."23

As he began his first expedition in 1904, traveling through India and Burma towards the western entrance to China, he was unsettled, in a way an upper-class Englishman might not have been, by the brutal class distinctions of race and class on display. Upon arriving at the Watson Esplanade Hotel in Bombay, he set out to see the sights. He did not walk; "no European ever walks"; he rode in a gharry, pulled by "gharry-wallas." All the servants were servile; they kept salaaming him: "it makes one feel uncomfortable at first." But the ideologies of race were quickly enforced and quickly absorbed; the servants worked for almost nothing, as he had, but "apparently they are quite pleased with such payment, and to give them more, as I was inclined to do, only makes them lose respect for you." From the gharry, he looked down on the streets: men and women carrying away waste on their backs, men naked but for turbans and loincloths, beggars "afflicted with some disabling and generally loathsome disease." "I saw the [British] captain kick and hammer one of the porters until I thought he intended killing him, for a most trivial offense." He watched himself reshape his attitude and comportment toward the lowest classes: "I can swear at them and order them about now, [but I don't think] that I shall ever reach the kicking stage."24

Above all, he discovered opportunities to reaffirm his middle-class faith. This faith emphasized that the core of true masculinity was individual integrity and freedom from subjugation to the will of others. The most fundamental demonstration of masculinity was the capacity to establish, protect, provide for, and control a home.25 He had established the foundations for his future domestic life before leaving Scotland, promising himself to Clementina Trail. He would spend the rest of his life working to build and provide for a home over the distance between Yunnan and Edinburgh. In India and Burma, he quickly learned to reframe his masculine sexual virtue along racial lines. At night, a new acquaintance took him to Bombay's brothel district. There, the filth, nakedness, and wretchedness of racially other bodies that had assaulted him in the daytime streets were concentrated in experiences of unregulated transgress of boundaries, moral and physical: "In the native quarter ... the stench is indescribable in places ... a mixture of sweaty bodies and all sorts of reeking abominations.... I never thought it possible for vice to be paraded so openly anywhere. We were continually being tackled by the women, some of whom even went the length of trying to get into the carriage beside us.... I pity the poor wretches." In Bhamo, the last steamer port on the Irrawaddy before the frontier with China, he was advised to buy a Burmese girl to take with him. "All the officers in the regiments stationed here ... keep them.... I could get a dozen tomorrow if I wanted them .... but I wouldn't touch any of them with a tarry [that is, tar-covered] stick. There is only one woman in the world for me and that is Clem, and she is white all through.... I have kept straight all my life and I have every reason in the world to keep straighter than ever."26 It was a full and satisfying conflation of racial and sexual purity.

He was reorienting his sense of self in relation to different others, making decisions, some fleeting, others enduring, about how to negotiate the lines of difference that fissured colonial society. As he approached the border with China, he discovered new uncertainties and new opportunities. European discourses about China and the Chinese had long been more multivocal than discourses about race in India or Africa. On the one hand, since the eighteenth century an influential strand of writing about China had described it as an advanced civilization with a venerable history, literate populace, and wise government. On the other hand, during the nineteenth century China had been incorporated into European racial theories, giving rise to a discourse about "racial character," centering on images of Chinese as a degenerate race, enfeebled, corrupt, deceptive, filth-ridden, and prone to disabling disease.27 After the Boxer rebellion and the punitive military operations that followed (1898-1901), florid racial invective became a commonplace feature of traveler's accounts, ethnographic descriptions, and fictional treatments of China. Yet older visions of China as an advanced and enlightened civilization were not entirely displaced. George Steinmetz has shown how some European colonists in China indulged in imaginary identifications with the Chinese as they negotiated these differences. "The prevailing European fantasy involved projecting oneself into the role of a Chinese mandarin or philosopher-king ruling over a literate and civilized people."28 The most popular way to play with such fantasies was to make photographic portraits of oneself in the dress of a Chinese official. As he neared the border, Forrest tried on such imaginary cross-identification. "Once I get right into China," he wrote his mother and sisters, "I shall put on the regulation Chinese dress, big baggy trousers reaching to the calf and a loose blouse with a big hat and Chinese shoes."29

He experimented with attitudes towards officials and soldiers he met along the road. He regarded them all as "Chinese"; later he would learn to make finer distinctions: on both sides of the border he passed through what he would come to know as Shan states. Near the border, the party waited several days while the commander of a Chinese battalion stationed in the region requisitioned men from surrounding villages to build a bridge across the Namsa (Namwan) river for the caravan.30 He entertained the region's elite, including the district military commander and a "rather curious old cove" who was probably the hereditary ruler (sawbwa) of a Shan state. They exchanged calling cards, eggs, whisky, biscuits, chickens, cigarettes, tinned plums, and cherry brandy. "The soldier one ... is a most kindly and polite little fellow, and I liked him best of all. He at once started to try to learn me Chinese."31 It was among the most genuinely good natured exchanges with local elites he would ever record.

Across the border in Tengyue, a tiny cadre of Europeans struggled to impose a regime of difference modeled on India's. Tengyue was a thriving town of about three thousand with a large merchant class involved in the cross-border trade. Forrest settled in at the consulate and held dinner parties "amongst ourselves of course" with the only other Europeans there: Acting Consul G. Litton and the British customs commissioner and his two assistants. When he went out, they insisted he take a soldier:

-He goes in front and by continual shouting and pushing gets the people to clear out of the way. He makes no bones about shoving some of them almost on their faces. It seems nasty, but it is really the only way and by doing this the people seem to respect one more.... Mr. Litton says that if it wasn't for the punishment which they know would be meted out to them our lives wouldn't be worth a moment's purchase, and I believe it from the look of some of them. Of course we always go armed and as they are great cowards, this keeps them in check. In coming home from the dinners or going anywhere at night we always had an escort of 4 soldiers and all [or most] of our servants carrying huge paper Chinese lanterns, quite a procession.32

Only a few years before, combined European forces had put down the Boxer rebellion, occupied the capital, forced the empress dowager to flee, ransacked the Summer Palace, and forced huge indemnity payments on China. The uprising had not spread to the southwest, however, and remote Tengyue had seen no violence. Even so, anonymous notices sometimes appeared on the streets excoriating the weak Qing government, decrying foreign influence, and agitating against British plans to build a railway from Burma. Two months after Forrest's arrival, Litton telegraphed to warn the Foreign Office that the Chinese authorities had received a rhyming pamphlet stating that foreigners "practice the violation of Chinese women and children and destroy tombs." In fact the pamphlet (a copy of which Litton sent with his report) was an announcement in somber verse by the Tengyue Militia Office (tuanlian ju) of reforms that would eradicate abuses and introduce more discipline into training, with the aim of discouraging the foreign powers pressing at Yunnan's borders.33 Forrest gradually came to understand that the townsfolk of Tengyue were more interested in peaceful trade than violent nationalism. In subsequent visits, he abandoned the shows of force to which the other Europeans in town seemed addicted.

Still, whenever pressed, he drew on the assessment of "Chinese character" he had begun to learn in Tengyue. After three weeks, Litton invited him on a journey east to Dali then north to Lijiang and to the Khampa town of Gyaltang (Zhongdian) in Yunnan's northwest. The route suited his purposes, traversing high ranges where he might find novel alpines. A caravan route took them to Dali, where Litton decided it was his business to investigate reports of tax corruption at a horse fair in Songgui, a day north. On the way, their servants, hired in Dali, balked at showing them a shortcut. "At last, we lost our tempers, and Litton started to bully them." That night, thieves took Forrest's pony and two mules from the temple where the party put up. The two British men gave chase, caught a man, and trussed him with rawhide ropes. In the morning they handed him over to the local magistrates, who ordered that he be given three thousand strokes with a bamboo rod while strung up by his hands, a fifteen-pound weight tied to his queue-a punishment he was unlikely to survive. Satisfied, they went to the fair: "I never saw such a beastly rabble in all my life. All the scum of China seem to have collected.... A crowd of ruffians ... followed us about making jeering remarks.... Then one of them started to pick up stones with the intention of stoning us, but when we saw this we drew our revolvers, and then you never saw such a scatter in all your life. The sight of weapons was sufficient without going even further." The next day, "we stayed at an inn at Hoching (Heqing) and were troubled by the usual crowd of gapers who followed us about wherever we went."34

In November and December of the same year, Forrest made another journey, this time with no European companion. With Litton, he had made calculated shows of force based upon the idea that the Chinese were "great cowards." On his own, he tended to lose control of such demonstrations, falling into rages that he would recall with remorse. In a village near Lijiang, "we also had a row with the people ... because they wouldn't unlock their temple door and let us spend the night there. Temper not improving you see! But I think there is some excuse as the average Chinaman would rile the heart of a wheelbarrow." He blew the lock off the door with his rifle, barricaded the door from the inside, and left early the next morning.35 In March, before the passes had entirely cleared of snow, he attempted to retrace most of the route he had traveled with Litton. From Dali, he went north to Jianchuan, where he commandeered a temple to avoid the county government seat, or yamen, and its magistrate-"for I despise the officials as much as the people." A crowd collected; some people began to examine and handle his baggage, and he ordered them out. One, a student, wouldn't move: "taking him by the scruff of the neck and the seat of his baggy breeches, I heaved him out of doors." The crowd yelled and threw stones at the temple windows. "Tired, hungry, and enraged as I was, this was more than I could stand and I let myself go ... and I must confess that I was most heartily ashamed of my conduct. I seized a stick, rushed out amongst them, and began laying out right and left."36

He recorded these confessions in a diary in the form of letters addressed to his mother, who passed them around to Clementina and his sisters. Letters were his sole means participating in domestic life, positioning himself as a proper husband for Clem, whose parents, from a higher class, disapproved of him. He was feeling his way. Here, the mapping of difference was far more fluid than in India, where it was supported by state institutions. Each outburst was touched off by the "insulting" behavior of people who appeared not to recognized distinctions he was anxious to assert. But there was more to it than his perception of insult or threat. In India, his interactions with different others had been accompanied by pity, loathing, and reaffirmations of racial and sexual purity. He had seized every opportunity to reassert moral and physical boundaries: he might gaze at vice, but he would never let it into his carriage. Here, confrontation with difference inspired rage and hatred instead: he had so quickly reached the "kicking stage," mixing it up with all the sweaty bodies.

How real was the threat he felt? Certainly antiforeign nationalism was rife in those years preceding the 1911 revolution. And just as surely his posture incited hostility. One can imagine the anger of townspeople thrown out of their temple by an arrogant, rifle-wielding foreigner. But for many in the crowds, courtesy rather than hostility may have been the prevalent mood. These were the same looks imperial officials expected to receive as they traveled through towns with retinues of servants and gaudy palanquins, and they took these looks to evince fear, respect, and obeisance to authority. For Forrest, however, the looks were the problem. Why would the "average Chinaman rile the heart of a wheelbarrow"? It was a matter of ocular posture-what he called "gaping." It was not that these open looks were always evidence of hostility, though they sometimes were. It was not that they always challenged his superiority, though they sometimes did. It was that in gaping crowds he had no means of dialog, no words to learn, no gifts to exchange, no calling cards to accept, no language at all, whether about difference or identity. Mute gaping stilled exchange, reducing relations to the purely visual, a terrible mirror. Kicking, laying about with a stick, grabbing people by the seats of their pants, were ways of breaking that mirror and establishing contact.

Gaping and Photography

Conditions of visibility were a persistent concern for colonial scientists and travelers in the nineteenth and early twentieth century. The power to frame what was seen and to regulate affect attached to seeing were at the heart of many imperial projects.37 Imperial men and women created cultures of display, in which exhibitions and photographs represented the colonized world as objectively visible.38 Historians of colonial photography have explored many ways in which photography structured fields of vision to emphasize the subjective, empirical gaze of the imperial viewer and deflect the power of returned looks.39 British colonial travelers reproduced the "world as exhibition" by following established itineraries, seeing famous sights in predictable ways, and structuring their visions of the landscape around specific aesthetic principles: the sublime, the picturesque, the romantic, the realistic.40 Relations of visibility were always entangled with relations of difference. Deborah Poole has pointed out that the production and circulation of millions of photographs of colonized peoples "lent support to the emerging idea of race as a material, historical, and biological fact."41 Theories of race created racial categories much the same way Linnaean taxonomy created species, selecting particular features to use as criteria. Racial theorists chose photographs to use like botanical type specimens as standards for comparison. And photographs of colonized peoples circulated through public and private archives, where they were categorized and organized through the principles of an archival regime. Exhibitions, photography, and science were machineries for framing and structuring fields of visibility-the gazes of "imperial eyes."42 In the ambiguously colonized peripheries of empire, however, these were cumbersome machineries, laborious to master and difficult to apply to daily experience.

During his first expedition, Forrest struggled to reconcile the landscape of China with the world that exhibitions and photographs of the colonized world had conditioned him to expect. Eventually, he learned to coordinate different modes of visual exchange with a typology of race and map them onto the aesthetic, affective, and botanical qualities of the landscape. The possibility of such a cartography began to dawn on him during his first journey with Litton. North of Lijiang, the travelers walked for three days through the deep gorge of the Jinsha, or upper Yangtze, in a "magnificent heat." They climbed out of the gorge to a boggy pass, where they camped in heavy rain. Forrest happily collected some of his first alpine flowers: Gentiana, Primula, and Saxifrage.43 The next day, the party descended to the Zhongdian plateau, a triangle of very high land about ninety miles long, surrounded by mountains and the Yangtze gorge to the south and bordering on the Tibetan region of Kham to the north.

This plateau, known in Tibetan as Gyaltang, was a strand in the network of roads linking Yunnan to Tibet and Southeast Asia, known as the Southern Silk Road; it was the main corridor for the trade in tea and horses between Yunnan and Tibet. It had been passed back and forth among polities for centuries.44 From the mid-seventeenth century, the plateau had been ruled by a hereditary nobility of two lineages who traced their ancestry to the ancient Tibetan kings. In 1723, in the context of a widespread rebellion against Qing rule in eastern Tibet, the Yongzheng emperor ordered armies from Sichuan and Yunnan to occupy the plateau. It was formally absorbed into the province of Yunnan as Zhongdian; appointed officials were sent to govern it; a garrison of some eight hundred imperial troops was installed in its largest town, and the hereditary rulers were demoted to district commanders.45 By the time of Forrest and Litton's visit, then, the plateau had been within the fold of imperial administration for a century and a quarter. Nevertheless, the hereditary officials had retained their armies and much of their influence, and imperial officials seemed to have great difficulty asserting administrative control. In 1905, the subprefect (tongpan) subordinate to the prefect of Lijiang controlled only forty soldiers, remnants of the garrison installed in 1723. This tiny army had recently been faced with rebellions in seven of the plateau's eight districts.46 Three miles north of the plateau's center of government was an important Gelugpa monastery, housing 1,226 monks, another competing source of political authority.

Forrest and Litton walked over the plateau in the rain. It was, Litton wrote in his report to the Foreign Office, "most picturesque ... the wide sweep of barley, wheat, and oat fields interspersed with plots of marsh, the dark pine woods on the lower slopes, the bare mountain tops above, and the sparkling mountain streams below reminded my companion, a Scotch botanist, of a cultivated highland valley in his native land."47 Rhododendron choked the foothills; blue Gentiana and red Euphorbia carpeted the pastures: to the "Scotch botanist" it was "one huge flower garden." In the late afternoon, they came to the southernmost of the plateau's two towns, called, in Chinese, Xiao Zhongdian. The barley harvest had begun, and the harvesters were holding a festival, eating boiled beef and piles of buckwheat cakes. The whole place seemed prosperous: shining fields, large white houses, well-fed people. While the land appeared fertile, most of the plateau's wealth came from trade. Caravans from the plateau moved through the length of Yunnan to bring tea, steamed and packed into large bricks, from the Tai Lü (or Shan) states in the far south of the province to Kham and central Tibet. Most households bred horses, and traders from the plateau supplied animals to Tengyue and Burma for the cross-border trade.

As they entered the town, a large man greeted them, finely dressed, with a sword. He was, Litton later learned, the occupant of a post called, in Chinese, huotou. This was an arrangement for local administration found in scattered parts of Yunnan at the margins of the imperial administrative system. Huotou systems developed mostly in larger villages lying on routes frequently traveled by traders and officials. They were a communal defense against the expensive obligation to host official travelers, the responsibility of local elites. Xiao Zhongdian was two stages south of the plateau's administrative center, Zhongdian, and two stages north of the nearest administrative center to the south in Judian, so it was a natural stopping place for every traveler on this route. Both Tibetan noblemen and Chinese officials had the right to requisition hospitality and transport from its populace. The elite families of the town elected one household of their member to undertake a series of ritual and administrative duties. Among these was the obligation to feed and house important travelers and to provide men and animals to transport them and their baggage to the next stage. Since, as Litton noted, the post of huotou was "far from lucrative" it was usually held for only one year.48

The huotou escorted the travelers to his house. They lived there three days while Litton recovered from a fever. It was a "common meeting place of the people," in Forrest's words, "practically a club" in Litton's. To his enormous surprise, Forrest found that he liked the people who gathered there. What distinguished them from "the Chinese," whom he emphatically did not like, he decided, was the way they used their eyes: "The people came riding in from miles around to see us. However they are much pleasanter than the Chinese, they do not stand and gape as those do. They are very curious regarding things they don't understand, but once you explain the article to them, they are satisfied and go away. On the other hand the Chinese simply stand and stare with the most ignorant expression imaginable on their faces, and will not clear unless you really chase them."49 The articles that most interested these traders and farmers were the travelers' guns. Almost as soon as they arrived, their host urged Forrest and Litton to go shooting with him. Wet and cold as they were, they put it off until morning. The next day, they walked out into the barley fields that surrounded the town and shot more than a hundred pheasants. After this, Forrest became attached to his host. He described him in unusual detail to his mother and sisters:

A regular savage beauty, about 6 ft. 6 ins. tall and 3 ft. 6 ins. broad, I should say, clad in a coarse scarlet Tibetan cloak ... open at the neck and strapped round the waist, he had scarlet putties on his legs and huge Tibetan top boots over these. Matted hair hanging down to below his shoulders (no pigtails), and dirt ad libitum, completed his costume. He looked like a pirate out of one of Gilbert and Sullivan's plays. However, he was a right cheery sort, and he and I got on like a house on fire.50

As the description makes clear, this was a wealthy man, dressed in finery. Forrest coveted his beautiful sword with a scabbard worked elaborately in silver and turquoise.

When Forrest returned to the plateau again in November and December of the same year, it was very cold. The oil in his shotgun froze, and he had to take it apart and clean it before he could shoot pheasants on his way up the plateau's gentle valley. He stayed again in the huotou's house. "I really like the fellow in spite of all his dirt," he wrote. "He is such a big man, and yet as simple, jolly, and kindly as a child."51 Again, their companionship centered on shooting. The huotou showed Forrest his Tibetan-made muzzle-loaded flintlock rifle. It was quality workmanship, the best rifle available in Tibet before the 1911 revolution.52 He took it outside, aimed at a tree on a cliff some eighty to one hundred yards away, and missed. Determined to "give him an eye-opener," Forrest loaded his Winchester with its full eleven rounds and blasted away at a limestone patch on the cliff. "There was a crowd around us by this time, and I can tell you that the group of faces would have made the fortune of any photographer. You never saw such astonishment depicted in all your life." The two then competed with Forrest's rifle, emptying some fifty rounds into the cliff.53

It is not surprising that Forrest imagined this as a photographic moment. The association between guns and cameras was commonplace in colonial situations. As Paul Landau observes, the technologies of gun and camera evolved in parallel after the 1860s. Chemicals developed for ready-made cartridges were adopted for use in dry-plate cameras by Eastman and Kodak; elements of the designs of the best cameras were based on the mechanism of the Colt revolver; both technologies moved toward the capacity to make clean, rapid, repeating shots.54 By the end of the century, game hunting had become a largely visual enterprise, incorporating cameras to create visual trophies. Later, Forrest would photograph long rows of Lady Amherst and white-eared pheasants hung up by their feet from the eves of his house in Nvlvk'ö-both hunting trophies and scientific specimens. In addition, travelers who ventured out of the reach of the armies and bureaucracies of colonial administrations had few resources with which to demonstrate their superiority. Forrest had his sense of personal hygiene: thus his repeated remarks about dirt even at a moment of profound camaraderie. But it was difficult to stay clean and even more difficult to show one's cleanliness to others in a convincing way, especially when they were far better dressed than he. So his gadgets-his Winchester, his camera, his telescope-had to bear the burden of demonstrating difference.

The posture of astonishment Forest elicited in this crowd of Tibetans was nearly identical to the "gaping" of which he accused "the Chinese." Yet while those looks made him miserable and violent, these delighted him-and made him think of photography. In other situations, he and his competitors would use their guns to similar effect. Clever Francis Kingdon Ward didn't even have to shoot his: in a village on the Mekong he showed a crowd "what wizards the English are" by making the cartridge pop out of his breech-loading shotgun in response to his whistle. "The men were so obviously taken aback that they merely stared incredulously." "My gun and my camera," he observed, "were always a source of great interest to the Chinese."55 Later, in Tibet, Kingdon Ward, frustrated in his desire to photograph the shy women of Jana, set up his gramophone in the street, played an operatic prelude, and whipped out his camera to photograph the "enchanted crowd."56 Chapter 5 shows how Joseph Rock resorted to the same cliché in moments of melancholy.

Michael Taussig argues that Europeans and Americans found in the astonishment they provoked with such antics a reflection of their own deep fascination with technologies of mechanical reproduction.57 It is certainly true that these men were fascinated with the wonder they cultivated in others with guns, cameras, and gramophones. But they hardly seemed awed at the mysteries of these technologies. They were all deeply troubled by the sensation of plunging into a sea of looks, where the eyes of others framed their sociality in alien terms. Their attempts, in practice or imagination, to generate astonished looks and freeze them with a camera were efforts to turn the power of "gaping" back on itself. Martin Jay observes that the eye of the camera, unreadable by those at which it gazes, holds the viewer still, "static, unblinking, and fixated."58 To generate and capture fascinated looks was to draw "gaping" into the familiar field of mechanical reproduction-the field of the "world as exhibition," through which the sights of colonized places were reproduced and exhibited in the metropole.

* * *

Not being gaped at unless at his command became for Forrest a precondition for social exchange without violence. After his adventurous excursion up the Salween gorge with Litton, he settled down to collect with increased determination. Though he had traveled with cooks, muleteers, and servants, he had until now done his collecting himself. In the summer of 1906, with the aid of missionaries of the China Inland Mission in Dali, he hired two men and three women to collect in the Cangshan range, just west of that city: this was the only time he would ever employ women as collectors. Leaving these five in the charge of one of his servants, he went north to spend the season on the great Yulong range. He took the southern route into the range, straight up the main street of Nvlvk'ö. His letters make no mention of the people he met there. But fourteen years later, he would recall that it was that summer that he first began to hire men from that village as collectors. Over the next thirty years, the men from Nvlvk'ö amassed one of the most extraordinary collections ever attributed to a single botanical explorer. They spent years at a time away from home, eating what they could, sleeping in the open, enduring all the physical and emotional hardships that travel over this immensely rugged landscape entailed. He tyrannized them; he fired them for insubordination; he fumed at them as "Bolsheviks" when they asked for raises; he scoffed at their ritual life when funerals drew them back to Nvlvk'ö. But for all this, it is clear that he developed profound camaraderie with many and sincere respect for a few.

They had learned as children many of the plants of the Yulong range. But there were many skilled herbalists among the "Chinese" population of the province who may have made better collectors, had he the imagination to look for them and the linguistic skills or social connections to find them. What made it possible for him to employ these farmers and live with them lay in how they used their eyes. He never once mentioned being stared at in Nvlvk'ö, and given his sensitivities, this is evidence enough. Like the Tibetans of the Gyaltang plateau, Naxi farmers and traders found staring profoundly impolite. Looking directly at a stranger, particularly one of high status, no matter how strangely attired, was deeply disrespectful. Meeting an important stranger on a path or village street, one looked away and went about one's own business, satisfying one's curiosity later from within a doorway or behind a fence. Habits of looking were distinctly gendered as well. For a young woman to look directly at a man not of her household out of doors was improper; older women could manage more direct looks, but again not at high-status strangers. These were widespread habits of ocular comportment, shared with most of the highland Tibeto-Burman-speaking peoples of the province.

As it happened, they were also familiar habits, close enough to those he had grown up with. Living amongst people who, however curious, did not stare, he immediately found ways to initiate social exchange even in the absence of a common language: trading cards, shooting his gun, producing specimens. When such deeply installed, tacit habits of ocular exchange were broken by "gaping," possibilities for social exchange were also broken, "blind rage" ensued, and physical violence became his only means of establishing social relations. In addition, "gaping" profoundly fractured his visual field. Under the unblinking eyes of different others, he found it difficult to look at anything but himself, and his nakedness often came to mind. In 1905, complaining to his mother of the "ignorant expressions" of "the Chinese," he commented, "I have become so callus that I can stand now and take a bath with a crowd around me." In company of people who turned their eyes away, he could refocus his own, away from himself, onto the landscape, to find in it the qualities that made seeing productive. In Nvlvk'ö, he learned to turn the eyes of others toward that common object as well, initiating the long and intensive process of ocular instruction through which others' eyes were trained to see in the ways demanded by Linnaean taxonomical science. He could only begin this process once he had broken that mute circuit of eye against eye that gaping entailed.

Scholarship on imperial travel has emphasized looking as a mode of knowing and possessing. "The idea of the 'the gaze,' especially as exercised by the white male explorer, missionary, administrator, or itinerant naturalist," writes David Arnold "has rapidly come to be seen as one of the principal expressions of the wider colonizing process. ... 'the gaze' is the prelude to possession in more material and institutional forms, just as travel is more about imposing upon, than learning from, the landscape subject to the itinerant gaze."59 Travel made the landscape a spectacle, replicated after the mid-nineteenth century in exhibitions for other eyes, in both metropole and colony. Photography provided a convenient model for this mode of knowledge production: a clear-seeing eye, unimpeded by the vagaries of subjectivity, rendering fields of sight into real objects, which could be classified, archived, and exhibited.

The temptation for scholarship has been to write as though imperial gazes could, in reality as well as in ideology, be modeled on the kind of vision commonly attributed to the camera. As though vision could ever simply be that of a sovereign master of two eyes, surveying the world and rendering it as representation. As though acts of vision are not always embedded in social fields of vision, composed of intersections of multiple pairs of eyes, intimate and distant, immediate and deferred, real and imagined. The visibility to others of one's own gazing body is constitutive of one's gaze; one can only see as a social being if one sees oneself as visible and one's own vision as including, reflecting, embracing, or deflecting the gazes of others. In the case of imperial scientific travelers like Forrest, this is easier to notice in places off the edge of the map of empire where travelers had fewer resources, material and ideological, at their disposal, to insulate their eyes from those of others.