Claiming the High Ground "nsd0e294"

1 Sherpa Country

1. Geographically the Himalaya is usually distinguished from adjacent mountain ranges such as the Karakorum, the Zanskar and Ladakh ranges, and the Hengduan Shan. It is generally considered to extend from Nanga Parbat

(8,125 m) in the great bend of the Indus river to Namcha Barwa (7,775 m) in the great bend of the Tsangpo-Bramaputra river and is bordered on the south by the Ganges plain and on the north by the Indus-Tsangpo depression and the Tibetan (Qinghai-Xizang) Plateau. The name of the range is ancient Sanskrit and means the Abode of Snow.

2. All of the world's peaks of 8,000 meters or more in height are located in Asia. Ten of the fourteen 8,000-meter peaks are located in the Himalaya and eight of these are located wholly in Nepal or on its borders with Tibet and Sikkim.

3. In some regions of eastern Nepal, including the Mount Everest area, the Great Himalaya also forms the watershed between Nepal and Tibet, with southern drainage to the Ganges and northern drainage to the Tsangpo. Elsewhere the watershed lies well into Tibetan territory and transverse rivers such as the Sun Kosi and the Arun cross the Himalaya en route to the Ganges. In western Nepal the Great Himalaya runs far south of the Tibetan border and does not form the watershed.

4. The mountains of the Mount Everest region, like the rest of the Great Himalaya, primarily consist of highly metamorphized Precambrian gneisses that have been overthrust along the Main Central Thrust Fault upon older formations. In the highest reaches of the Khumbu Himal, and particularly on the upper heights of Lhotse and Mount Everest, the black gneisses (dark-colored paragneisses) are overlaid by sedimentary rock formed of sediments laid down in the Tethys Sea, the ocean that was displaced by the collision of India and Eurasia and the subduction of the Indian plate. These rocks are either unmetamorphized or only slightly metamorphized. The Everest metasediments include biotite schists, marbles and phyllites, and grey crystalline limestones (Vuichard 1986:44-47).

6. Mount Everest and Lhotse are foreigners' names for Khumbu peaks. So too are the names of a number of other famous peaks such as Nuptse and Pumori. In some cases these peaks had no traditional Sherpa names. Mount Everest was called Chomolungma by Khumbu Sherpas, as it was by Tibetans, and was considered to be the home of Miyolangsangma, a goddess of nourishment and wealth. The meaning of Chomolungma is not certain, though it is often translated as Goddess Mother of the World, or Goddess of the Wind (for a more detailed discussion see Bernbaum 1990:7). The Nepali name for the mountain, Sagarmatha, means Forehead in the Sky, or Forehead Touching the Sky. Sherpas often now refer to the mountain as Everest, a name which honors Sir George Everest, the early surveyor general of the Survey of India, whose work helped establish the baseline from which the height of the mountain was first determined in 1852.

The only 8,000-meter peak in the Mahalangur range located outside of Khumbu is Makalu (8,463m), just beyond the high ridge which defines the eastern border of Khumbu. Lhotse Shar (8,383m), a subsidiary peak of Lhotse, is sometimes counted as a fourth 8,000-meter Khumbu Peak.

7. This range was also uplifted by the same process which produced the Khumhu Himal crest to the north and consists mainly of migmatitic and granitic rocks (Vuichard 1986:48).

8. The northern stream is also known locally as the Pheriche Chu. I follow the Lobuche Khola usage established by the National Geographic Society in its 1988 map of the Mount Everest region. The names in Western languages and Nepali used today for Khumbu rivers were bestowed by outsiders. Kosi and Khola, Bhote and Dudh are Nepali not Sherpa words. Kosi and Khola are Nepali terms for rivers. Dudh is Nepali for milk and Bhote for Tibetan . How long ago these names were bestowed on local features is unknown. Two of the earliest European maps of Nepal, the 1811 Kirkpatrick map and the 1819 Buchanan-Hamilton map (Gurung 1983:18-21) show the "Dud Kosi." Khumbu Sherpas conceive of rivers in a rather different way than Western geographers. They do not refer to the entire length of a river as a single entity. Instead they refer to a river ( tsambu ) or stream ( chu , which also means water or spring) in terms of particular reaches of water named according to the settlements and regions along the bank. An individual river may thus be referred to locally by many different names.

9. Hagen et al. (1963) and Vuichard (1986) suggest that the Bhote Kosi may predate the rise of the Great Himalaya. Hagen et al. note that the present valley appears to be far too large for the river flowing through it today to have fashioned it and suggest that previously it was the watercourse of the Dzakar Chu that now flows north from the Nangpa La into the Arun river in Tibet. Vuichard also suggests that this Bhote Kosi-Dzakar Chu system may have originally flowed northward, draining the slopes of a pre-Himalayan range located in the vicinity of the present Numbur-Kantega range.

10. There is some disagreement over the naming of the reach of river between the confluence of the Dudh Kosi and the Imja Khola and the confluence of this new stream with the Bhote Kosi. Some authorities refer to this as part of the Imja Khola (e.g., Schneider 1963). Sherpas, however, today refer to the same reach of river as part of the Dudh Kosi. I have adopted the Sherpa approach.

11. Sherpas did not name glaciers and the first cartographers to produce detailed Khumbu maps bestowed the names which are now in use (Schneider 1963). Often the names of local grazing areas and herding settlements were applied to adjacent glaciers. One of these, the Lungsampa Glacier in the upper Dudh Kosi valley, is misnamed, as the grazing area by that name is actually located in the next valley to the west alongside what is now known as the Sumna Glacier.

12. A more detailed treatment of altitudinal zonation of vegetation is provided in chapter 5, while soils are taken up in chapter 3.

13. Climatic data have been collected by the government of Nepal in Nauje since 1948 (Joshi 1982:399) and for shorter periods of time at Shyangboche, Khumjung, Tengboche monastery, and Lhajung (Byers 1987b; Muller 1959). Data for Nauje and Tengboche monastery, the two stations where they have been collected for the longest period, may have sometimes been based on haphazard measurement and recording.

14. The relatively dry conditions of these valleys enable agriculture to be carried out at higher altitudes than in wetter regions of the Himalaya. High rainfall and a high percentage of cloud cover during the summer encourages the development of cloud forest at altitudes above 2,500 meters on windward slopes in many areas of Nepal (Metz 1989:161).

15. In Nauje, for example, 89 percent of the annual mean rainfall (1949-1977) of 1,048 millimeters fell between May and October and 79 percent of it fell between June and September (Joshi 1982:400).

16. Occasionally autumn hurricanes in the Bay of Bengal cause torrential rains in the Himalaya which can fall as heavy snows. In October 1987 such a storm dropped more than 100 millimeters of rain on Nauje in less than twenty-four hours and heavy snowfalls occurred above 3,500 meters which blanketed much of Khumbu for many weeks.

17. The settlement of Milingo is also sometimes referred to as a village, although today only six families regard their dwellings there as their main house.

18. Occasionally Khumbu Sherpas say that Pharak begins at the first Pharak Sherpa village, Thumbug (Nepali Jorsale), rather than at the confluence slightly upstream. Similarly some people consider other Sherpa regions to begin with the first settlement that is culturally distinct from neighboring ones.

19. According to legend the Amphu Laptsa was crossed long ago by Rais from the upper Hinku valley to the southeast, who unsuccessfully attempted to settle Khumbu before the arrival of the Sherpas. There are no traditions of Khumbu Sherpas using the Amphu Laptsa or the Mingbo La other than when accompanying foreign exploring, surveying, mountaineering, and trekking tourist parties.

20. Another road has now reached a similar distance southwest of Khumbu and may well ultimately be extended to Salleri.

21. For an eyewitness account of the construction of the Lukla airstrip and a discussion of its consequences see James Fisher (1990). The Lukla airstrip has become the main conduit for mass tourism into the region. Shyangboche airstrip was used very little during the 1980s and is only suitable for the very small Pilateus Porter planes. In autumn 1990, however, the Everest View Hotel began running a regular Pilateus Porter air service to Shyangboche and helicopter charters to the hotel also became common. In recent years the Royal Nepal Air Corporation has kept Lukla airport open for limited service during the monsoon months.

22. The sacred peaks, however, are not places which are set aside from all resource use as has sometimes been suggested (BjØnness 1986). Khumbu Sherpas object to attempts to climb Khumbila and certain mountains sacred to particular clans, but they do not consider the slopes of these mountains off-limits to forest use, grazing, or agriculture.

23. See Berreman (1963 b ) and Karan and Mather (1987) for discussions and maps of Himalayan culture regions. The distinctively long-sleeved Tibetan and Sherpa cloak style, with sleeves extending a foot and more longer than the arm length, seems likely to have been adopted from Tang dynasty Chinese styles. Zhi stones may be a rare form of agate and are so valued that a single stone may cost three thousand dollars or more today in Khumbu.

24. The Bhotia peoples of the high-altitude Tibetan border regions of the Great Himalaya and the trans-Himalayan valleys of the Inner Himalaya (including the Sherpas) comprise less than 1 percent of Nepal's total population. According to the 1981 census (Nepal. Central Bureau of Statistics. 1984) there were 73,589 speakers of "Bhote-Sherpa" (a bureaucratic category rather than an actual language), accounting for only 0.5 percent of the total population of the country. If peoples such as the Tamangs (522,416) and Gurungs (174,464), some of whose ancestors according to legend came from Tibet (Fricke 1986:29-30; Holmberg 1989:14, n. 8; Messerschmidt 1976 b ), are included, the percentage of Nepal's population descended from immigrants from Tibet rises to at least 5 percent. It has been suggested that the 1981 census considerably underestimated the number of Tamangs and may also have underestimated Bhotias. The reported decline in the number of Sherpas in the Solu-Khumbu district between 1971 and 1981, for example, appears much too high.

25. The Nepali word Bhote and the anthropological term Bhotia are derived from Bhot , an old name for Tibet. It is used by a number of lower-altitude, nonethnically Tibetan peoples in India and Nepal to refer to the ethnically Tibetan peoples living on the border of Tibet.

26. While some of the inhabitants of the Taplejung region of northeastern Nepal (and particularly the village of Gunsa and other parts of the Kambachen valley near Kachenjunga) are often referred to as Sherpas (Bremer-Kamp 1987; Sagant 1976), their relationship culturally and historically to the other groups who call themselves Sherpa has not yet been established. The complex multi-cultural settlement pattern of Helambu and some other areas in which Sherpas form a component of the population is discussed below.

27. The Solu-Khumbu region (including Pharak) constitutes a large part of the district of Solu-Khumbu, a district shared also with substantial Rai and Nepali populations who actually outnumber the Sherpas. J. Fisher gives the number of Sherpas in the Solu-Khumbu district as 17,000 (1990:55), whereas the 1981 census puts the total at 15,166 out of a total district population of 88,245. Estimates of the number of Sherpas outside of Solu-Khumbu are far rougher. Hutt (1986) suggests that there may be approximately 20,000 Sherpas in Nepal. There may be another 7,000 in the West Bengal district of India, most of them in the Darjeeling area (J. Fisher 1990:55). Twenty thousand may be low for the total in Nepal. There may be as many as 25,000 or even 30,000 depending on how "Sherpa" is defined and whether or not the "Sherpa" population of northeasternmost Nepal in the Taplejung-Gunsa region or part of the population of Helambu are included. In Solu-Khumbu and adjacent areas (Olkadunga, Khotang, Ramechap, and Dolakha districts), however, it is likely that all or nearly all of the population classified "Bhote-Sherpa" by the Nepal government in recent censuses is Sherpa. The total number of people recorded in the 1981 census (Nepal. Central Bureau of Statistics. 1984) as native speakers of ''Bhote-Sherpa" in these districts alone is 26,473. (In a number of other districts it is impossible to use census figures as even a rough guide to count Sherpas since these areas are inhabited by both Sherpas and other "Bhotia'' groups).

28. It is somewhat ironic that Tenzing Norgay's achievement on Mount Everest made Sherpa a household word around the world. Tenzing Norgay grew up in the Khumbu village of Thami Og and married a woman from nearby Chanekpa before he emigrated to Darjeeling to seek his fortune in mountaineering. But he was not a Sherpa in the strictest sense, having been born in Tibet of non-Sherpa parents, and would be considered to be a "Khamba" by Khumbu Sherpas. This important social distinction is discussed later in this chapter.

29. All the pre-World War II Everest expeditions were forced to recruit their Sherpa high-altitude porters in Darjeeling and to approach the mountian via Tibet.

30. Among the most notable anthropological contributors to this still-continuing exploration of Khumbu society and culture have been Ortner (1978, 1989), Pawson, Stanford, and Adams (1984), Pawson et al. (1984), J. Fisher (1990), and Adams (1989). Fürer-Haimendorf's work (1964, 1979, 1984) is worth special note as still constituting the finest, broad ethnographic treatment of the Khumbu Sherpas. Ortner's analysis of the beliefs, values, and cultural patterns which underlie Sherpa social behavior, ritual, and history (1989) is also exceptional, although it should be read while keeping in mind that her earlier work (1978) is based on fieldwork in Shorung and that these observations and insights are not always fully applicable to Khumbu.

31. For a fuller review of the geographical literature and pertinent reports by physical scientists and national park administrators see Byers (1987 b ).

32. That these basic characteristics are widespread and have been relatively enduring, however, should not obscure historical and regional differences in Sherpa culture—a point which Ortner has also recently raised in discussing the difficulties of identifying a Sherpa cultural style or ethos (Ortner 1989:4-6).

33. I base these brief remarks on Sherpa identity and the discussion later in this chapter on Sherpa regions on my experiences discussing regional geography with Sherpas in the Solu-Khumbu district and adjacent areas. The term by which Sherpas refer to Nepalis, rongba , is also used by Tibetans and signifies people of the lower valleys.

34. In many cases there are clear differences between Sherpas and non-Sherpas on all four of these points. Tibetans and Tamangs share with Sherpas a faith in Tibetan Buddhism and are therefore considered, at least by Khumbu Sherpas, to be in a special class of peoples who are also "believers." But distinctions among Sherpas, Tibetans, and Tamangs are easily drawn on the basis of the other three as well as many additional points.

35. Many Khumbu people today speak of the Sherpas as a people who mountaineer and work for trekking companies and who raise yak and potatoes. Some of these features may indeed set contemporary Khumbu Sherpas apart from non-Sherpas or from some other Sherpa groups, but, like the characterization of some recently popular styles of dress or architecture as distinctively "Khumbu Sherpa," even such long-standing traits as mountaineering prowess and potato cultivation have not always set Khumbu Sherpas apart from other Sherpas.

36. If Sherpas did indeed migrate to Nepal from the eastern Tibetan area of

Kham (see below) it seems likely that their language may have even closer linguistic similarities to eastern Tibetan dialects.

37. Most Sherpa marriages are monogamous, although like Tibetans, Sherpas also tolerate polyandrous and polygynous arrangements. These, however, are very rare in Khumbu today and for at least the last several generations polyandry was not nearly as common in Khumbu as it was in Tibet.

38. Much of Tibetan society, including that of the central region of Tibet, is apparently not organized along a clan basis. Fürer-Haimendorf (1984:187), however, reports that Tibetans from the Kham region, the reputed original homeland of the Sherpas, do have clan affiliation.

39. The Nyingmapa is the oldest of the four major sects of Tibetan Buddhism (the name itself refers to "old ones"). It developed out of the introduction of Tantric practices by the Indian adept Padmasambhava, called Guru Rinpoche (Precious Teacher) by Tibetans and Sherpas. Nyingmapa is noted among the Tibetan sects for its rites of exorcism and other magical practices and for its married village lamas who are caretakers of village temples and preside over local life-cycle ceremonies.

40. Shrines and temples are constructed on Tibetan models as described in Tibetan texts. Some use may also be made of Tibetan geomantic principles, that are recorded in certain other texts. House styles differ in some important respects (particularly in Sherpa use of pitched roofs and noncourtyard design). Houses more similar to Sherpa design, however, are found in eastern and southeastern Tibet.

41. MacDonald (1980:141) also notes oral traditions of early administrative control of Khumbu by Tibet, Tibetan claims to which were given up after Nepali authorities sent troops to the region to rectify the situation. He relates this story, however, in the context of an era in Khumbu history which he dates to 400 years ago. This would put the Tibetan administrative presence in Khumbu long before the unification of Nepal in its current sense in the late eighteenth century. Suggestions of Tibetan administrative activity in Khumbu in the nineteenth century are still more startling, for Khumbu was supposedly incorporated into the Nepali state in the late eighteenth century and was visited as early as 1805 by a Nepalese official (Stiller 1973:265). If there is any substance to these legends it would testify not only to the ambiguous allegiances of Khumbu Sherpas of that era but also to the remoteness of the region from the centers of administration of both Nepal and Tibet. According to Khumbu traditions Sherpas paid tax simultaneously to both Nepal and Tibet in the early nineteenth century and discontinued this only after the defeat of the Tibetans in this area by Jung Bahadur Rana's forces in the war of 1855-1856.

42. G. Clarke (1980 a , 1980 b ) has observed that many of the inhabitants of Helambu (Yelmo) who have often been called "Helambu Sherpas" or who call themselves Sherpas, are actually not related to the Sherpas of Solu-Khumbu. He traces their descent to intermarriage between Tibetan immigrants from the Kyirong area and Tamangs from the lower country just to the south. There are also a number of families, especially in the large village of Tarkyagang, who are descended even more directly from more recent Tibetan immigrants and in the

. Some of the Sherpa settlers in the Arun region left Shorung around 1825 (Fürer-Haimendorf 1975:117). Others emigrated there from Shorung and Pharak later in the century. In the villages of Navagaon and Tashigaon there are families of several Solu-Khumbu clans (including Salaka, Chawa, Pinasa, Gole, Lamaserwa, and Goparma), and some elderly individuals trace their origins to their ancestors' migration from Solu-Khumbu settlements four or five generations ago. The Chyangma area was settled by a group of Shorung migrants between 1725 and 1750 (Oppitz 1968). Rolwaling was settled by families from western Khumbu around 1860 (Sacherer 1975; 1981:157). Sacherer notes that early Rolwaling settlers were poor families, some of which included people who were escaping from bad debts or fleeing after having committed crimes. Aziz (1978) has noted that some of the Tibetans who came to Khumbu in the nineteenth and early twentieth centuries were also people fleeing exactly the same kinds of situations. Similar stories go back as far as the tale of Dzongnangpa (see below) and the early days of Khumbu settlement. Some Sherpa emigrants to Darjeeling also may have migrated not so much because of the allure of career opportunities there as to leave behind awkward situations in Solu-Khumbu.

44. Sherpas began settling in Darjeeling in the nineteenth century, drawn by wage labor and trade opportunities in the then summer capital of the British Raj. By 1901 there were already 3,450 Sherpas in Darjeeling according to A. J. Dash ( Darjeeling 1947, cited in Ortner 1989:160). Beginning in 1907 Sherpas became noted for high-altitude mountaineering and this became a major source of work during the period before World War II when Darjeeling was a major mountaineering center. By 1951 the number of Sherpas in the region had reached 7,539 (B. Miller, Ph.D. diss., University of Washington, 1958; cited in Ortner 1989: 217 n. 4). These Sherpas included emigrants from Khumbu as well as from Shorung and Pharak. The relative percentage of Khumbu migrants, the degree to which they hailed from different parts of the region, or the effect that this emigration had on Khumbu population density and land use are not known. In 1984 I was told by Sherpas in Darjeeling that their numbers had declined since the 1950s. A number of families were said to have moved back to Nepal following the eclipse of the mountaineering and tourism industry in Darjeeling by tourism development in Nepal.

45. Farther up the Arun are communities of non-Sherpa Bhotias known as Shingsawa.

46. Such cultural diversity within a population often considered to be a single "people" is not unique in the Himalaya. Berreman, for example, has noted considerable cultural variation among the Pahari of northern Uttar Pradesh (1960). The Rais are perhaps an even more striking example. Anthropologists

have discussed a dozen different Rai groups that are said to speak mutually unintelligible dialects (McDougal 1979) and among whom religious practices, architecture, and other facets of life are by no means uniform. Ethnic diversity among the Tamang has also been noted (Fricke 1986; Holmberg 1989).

47. Such marriages may have been forbidden, however, in Shorung.

48. In many other spheres, however, cultural variation within Khumbu is minor. Dialects, for example, do not differ nor do any customs of building houses or wearing a distinctive village style of dress.

49. Khamba is generally considered to mean "people of Kham", the region presently divided between the easternmost part of the Tibetan Autonomous Region and the Chinese province of Sichuan. In the sense in which Khumbu Sherpas use it the word means an immigrant from Tibet or from a Bhotiainhabited region of the Himalaya who arrived in Khumbu before the great influx of refugees in 1959. Khumbu Sherpas even speak of Thakali, Nupri, and Lowi Khambas, Bhotia people from Thak Khola, Nupri, and Mustang. The very similar word Khyampa means "wandering" in the Tibetan dialect of the Humli-Khyampas of northwestern Nepal (H. Jasehko, A Tibetan-English Dictionary 1968; cited in Rauber 1980:59), and perhaps the Khumbu use of the word carries a little of that connotation.

50. Khumbu Sherpa society is much more ethnically varied than is the Shorung Sherpa population. There are very few Khambas in Shorung.

51. The number of lowland Nepalis stationed in Khumbu has increased from 84 in 1970 to 339 in 1982 (Fürer-Haimendorf 1984:32). Pawson counted 187 Tibetan refugees in Khumbu in 1982, a significant decline from the 274 counted in 1970 by the Langs (Pawson, Stanford, and Adams 1984:75). I know of fourteen Tibetan refugee families in Nauje, six in Khumjung, and one in Thamicho. There may be a few others.

52. Adams (1989:175-176) found in 1987 that only 45 percent (64 of 141) Khumjung household heads were considered to be Sherpa on the basis of their clan membership, while 43 percent (60) were Khamba or Khumbuwa. Presumably the "Sherpa" status here referred to "old" clan members. The remaining 12 percent of Khumjung households were Tibetans, a blacksmith family, a Tamang family, and several lowland Hindus who mostly came to the area as schoolteachers.

53. They may, not, however, have fully tallied the Bhote Kosi valley population since they reported an astonishingly high 32.5 percent of the houses there empty and presumably abandoned (Pawson et al. 1984:244).

54. The number of dwellings in most of these places is larger than the figures suggest since some families own houses in more than one village and I have tallied each family only once. Monks and nuns are not included in the count, nor are people who have retired to religious hermitages ( tsamkhang ). Some families who maintain their main residence in a community other than one of the main villages, such as the gunsa settlement of Thami Og in the Bhote Kosi valley or the secondary, high-altitude agricultural sites of Dingboche and Tarnga, may also have been missed in the count.

55. The full, multialtitudinal settlement pattern is most highly developed in

the Thamicho region. In this twenty-two-kilometer stretch of the Bhote Kosi valley and its tributaries there are forty-seven settlements situated at altitudes ranging from 3,400 to 5,000 meters. Only three of these are main villages.

56. Yul is used by Khumbu Sherpas to refer to village, valley, and region. Main villages are distinguished locally from other settlements in terms of function rather than simply by size, for some high-altitude secondary settlements (Dingboche, Tarnga) are larger than some main villages in terms of the number of houses in them. Main village residence defines one's participation in such important festivals as Dumje, Losar, and Pangyi; eligibility and responsibility to hold certain types of community offices; the site of ritual practices such as the placing of prayer flags on dwelling roofs three or four times per year; and the site of life-cycle ceremonies such as weddings ( zendi ). Main village houses are larger than those owned in secondary settlements and families usually spend more months per year living in them.

57. The sandy surfaces and level expanses of these three sites suggest that all three may once have been glacial lake beds.

58. The origins of both the names Nauje and Namche Bazar, the Nepali name, have been lost. Some Sherpas suggest that Nauje might be derived from the phrase nating uk che , which refers to a big forested "corner."

59. There were no bazaar or shops in Nauje prior to 1965. Tibetans and Rais instead visited Sherpas in their homes in order to trade.

60. Although lodges and shops are rare in these villages their economies today nevertheless are entwined with tourism. Most families depend on income from mountaineering and trekking work to purchase grain at the Nauje weekly market. Some also operate lodges in the high-altitude herding settlements on the approach to Mount Everest.

61. The origin of the name Khumjung is also uncertain. According to Ortner's reading of the nineteenth-century account of the Indian Survey pandit Hari Ram the name comes from Khumbu Dzong and was the home of the leading Khumbu administrator (Ortner 1989:23, 92-93). Fisher disagrees with this interpretation and suggests instead that the name signifies Khum Jung or Khum valley (J. Fisher 1990:xv-xvi). Perhaps this is short for Khumbu Jung, Khumbu valley. I can cast no further light on the question other than to suggest that Ortner has misinterpreted Hari Ram's account of his 1885 trip up the Bhote Kosi river and that his Khumbu Dzong is not at Khumjung but rather in Thamicho (the home of the gembu , the foremost officials of the nineteenth century), perhaps in the village of Thami Og. I have not heard of any oral tradition that makes direct reference to the site of the Khumbu Dzong which Hari Ram reported in 1885. This was the era when Shangup Dorje of Thami Og was gembu and it seems likely that the dzong was located in or nearby that village.

In a summary of Hari Ram's report Rawat (1973) describes the location of the dzong in the context of what appears to be the Bhote Kosi valley: "From a few miles north of Jubang [Jubing, a Rai village in the lower Dudh Kosi valley south of Pharak and Kharikhola] to Khumbu Dzong, the lower parts of the mountain sides are thickly wooded. . . . For the four or five miles beyond

Khumbu [Dzong], a solitary pine, rhododendron, or a Tibetan furze may be seen. After this not a tree is to be seen, and till the suburbs of the Ting-ri [Ganggar] are reached the only vegetation met with is a short grass." (Rawat 1973:165). This describes a trip from Jubing to Tibet which would have passed Nauje and then continued up the Bhote Kosi valley to the Nangpa La. Hari Ram notes that Khumbu Dzong is "about two miles north of Nabjia [Nauje] and on a flat part of a spur" (Survey of India, Exploration in Tibet and Neighboring Regions, 1879-1892 , 1915 vol. 8, pt. 2; cited in Ortner 1989:23). Nabjia is clearly Nauje from his detailed description of the place. Two miles north might indicate the Thamo-Mende area, two major gunsa which might have been a seasonal home of the Thami Og gembu. Thami Og, the home of the gembu in those days, is about four miles north of Nauje and otherwise fits Hari Ram's description fairly well.

The only dzong about which legends still circulate in Khumbu, however, is the ruin of one building locally called a dzong that could be seen as recently as 1990 at Top Data (Cannon Hill) not far from Nauje. This site is believed to have been a Tibetan post in the nineteenth century, but in spring 1991 a Magar work crew dismantled the ruins of the two-room structure to supply building stone for a nearby teashop. There was also said to be a large ruin early in the twentieth century at Tarnga which some people think may have been a dzong and attribute to Dzongnangpa.

62. Sherpas consider that this name was probably earlier Khum Te, or upper Khum in contrast to Khum Jung.

63. Upper village families do, however, have land in lower Pangboche, as do two families who are based in nearby Milingo.

64. Gunsa are often occupied only for a very few weeks each year during times when crops must be tended or when they serve as a herding base. In the Bhote Kosi valley, however, many families move to houses in settlements such as Pare, Thamo Og, Thamo Teng, Samde, and Mende for the winter. In the Thamicho region many families formerly only lived in their houses in the main villages for a few weeks each year, primarily in spring and autumn. These families often had quite large houses in the gunsa settlements and moved many of their possessions there with them in winter. In recent years these moves have been decreasing somewhat and the main villages are becoming more important winter bases.

65. The familiar name Sherpa is actually a distortion by outsiders.

The accuracy of the clan records that Oppitz discovered is by no means certain. Nor is it known when the texts were actually written (although Oppitz suggests that the most important, the Ruyi or The Report on the Clans , is "probably four hundred years old" (1968:143). MacDonald has challenged this claim and has even suggested that the apparently old text may be a recent prank (1987:58).

66. Both the date of the emigration from Kham and the arrival in Khumbu remain, however, very tentative.

67. It is entirely possible that the pass may already have been known as a route into the southern lands and it could conceivably have been used by earlier

migrants, pilgrims, and traders. This, however, is not clear from surviving legends. Ortner (1989:26) suggests that Tibetan hermits were already making use of some of the caves in eastern Khumbu before the Sherpa arrival. There is an oral tradition that early Sherpas may have arrived in the Bhote Kosi valley before the main group crossed the Nangpa La, and that they came to Khumbu via Rongshar valley, Rolwaling, and the Tashi Laptsa pass. But some Khumbu elders do not agree with this idea, and the version which I have stressed is the one most widely accepted today in Khumbu.

68. Some Khumbu Sherpa clans such as Mendewa, for example, trace origins not to Kham but to very nearby regions of Tibet. Presumably they came to Khumbu at some time after the first group of settlers. Other lineages trace their Khumbu ancestors to eighteenth- or early nineteenth-century Sherpa immigrants from Shorung and adjacent areas.

69. One variant Khumbu legend has it instead that the original Khumbu settlers were seven brothers, each of whose lines became a clan.

70. This is a common Sherpa interpretation of the Tibetan concept of beyul or hidden valley. Khumbu Sherpas believe that Guru Rinpoche, the great spiritual hero of Nyingmapa Buddhism and the man who did so much to establish Buddhism in Tibet in the eighth century, himself visited Khumbu and through his powers established it as a beyul. Guru Rinpoche is said to have flown north to Khumbu after having obtained special spiritual power during meditation at Mara Tika (Halashe), a cave in a ridge saddle near the confluence of the Dudh Kosi and Sun Kosi which is now both a Sherpa and a Hindu pilgrimage place. His stay at Arka Phuk, a cave on the slopes of Khumbila above Khumjung, is said to be marked by several of his handprints on the cave wall.

71. They could, however, also be relics of the activities of more recent Gurung shepherds or even former generations of Sherpa herders. These particular ruins predate the arrival of the Tibetan refugees who left behind many similar abandoned structures in the Bhote Kosi valley after their sojourn there in the early 1960s.

72. The Kulunge Rai inhabit the Hinku and Hongu Khola valleys southeast of Khumbu. One elderly Sherpa relates that Kulunge Rai have told him that they honor Ma Pe with special rites.

73. Unfortunately important details such as the type of grain pollen discovered are not given.

74. Arrangements between Sherpas and Rais in a number of areas appear to reflect Sherpa recognition of earlier Rai territorial rights. Some parts of Pharak, for example, paid a herding tax to Rais earlier in this century and families who use summer grazing grounds to the east of Pharak in the Mera area pay grazing taxes today. Arun Sherpas continue to pay a tax to nearby Rais in order to occupy their village lands and this may also be true for Sherpas in the Kulung and Salpa areas.

75. According to a Khumbu legend the Rais were asked to help restore to power the sons of a woman who had escaped the assassination of Dzongnangpa, an early Khumbu political leader and his associates (see below) and made her way south to settle in Dongbu country. The Rai king obliged and sent an army

up the Dudh Kosi. For six months they were stymied in upper Pharak below the confluence of the Bhote Kosi and Dudh Kosi and were unable to advance into Khumbu. The Rai forces were victorious, however, after the sons' mother helped them outflank the Sherpas by guiding them along a high route on the slopes west of the Bhote Kosi.

Ortner also discusses these events, and gives the Dongbu king's name (or the name of his descendant who was conquered by Prithivi Narayan Shah's forces) as Makwan Sher. She suggests that Makwan Sher was a Sen ruler (Ortner 1989:89), whereas Konchok Chombi testifies that he heard that Makwan Sher was the Kiranti ruler of Kirtipur in the Kathmandu Valley. He further notes that the troops who invaded Nepal were Rai from the lower Dudh Kosi valley region, not forces from the Kathmandu valley. But Karna Sena, ruler of Vijaypur (with a capital at that place in the hills north of Dharan in eastern Nepal) was in control of the lower Dudh Kosi valley at the time of Prithivi Narayan Shah's conquest of that region in 1772 and the Rai leaders there had recognized his sovereignty (Pradhan 1991:67, 113-115). Indeed, the current name Rai may derive from the Sen practice of giving the tital raja to the Rai leaders, which was then also conferred on them by Prithivi Narayan Shah. This was spelled raya and latter was changed to Rai (ibid.:52) for the groups that called themselves Khambus. The seventeenth- and eighteenth-century political status of the upper Dudh Kosi valley, including Khumbu, however, is still far from clear.

76. This conflict is said to have occurred because a Sherpa yak was killed in the Golila area (just south of the Lamjura pass) by Rais who were unfamiliar with yak and believed that it was a wild animal of a new type. The Sherpas and Rais agreed to meet on an appointed day and fight to determine territorial control. The Sherpas won and as a result came to exclusively settle the areas they inhabit today.

77. Ortner (1989:26) believes that the Pangboche and Dingboche areas of eastern Khumbu were the center of early settlement. It is conceivable that the arriving immigrants from Kham might have established themselves early on in this region, but from the oral traditions it appears that no permanent settlements developed there until several generations later following the establishment of a temple at Pangboche by Lama Sanga Dorje. Stories of the early settlement of Dingboche and the idea that Dingboche was the first Khumbu village are probably later Sherpa speculations based largely on observations of the area's suitability for yak and barley. The Bhote Kosi valley at the foot of the Nangpa La would have provided equally fine grazing and excellent opportunities at Tarnga and other sites for the cultivation of irrigated barley.

78. Several versions of the stories about Dzongnangpa's misrule and attempts on Lama Sanga Dorje's life are told. All accounts agree that Dzongnangpa was unsuccessful in having Lama Sanga Doric killed and was eventually assassinated himself by Sherpas near Tarnga at Chakuparteng after he had killed the Zamde lama whose hermitage was above the Langmoche Chu just south of Tarnga. Dzongnangpa was killed during a party in a tent. Sherpas collapsed the tent and beat the trapped Dzongnangpa to death. Some people relate that a number of his followers were also killed.

79. Temples were established at Thami Og and Kerok in Lama Sanga Dorje's lifetime, apparently in the mid-seventeenth century. But it is unclear whether the Thami Teng and Thami Og areas were already settled or whether the temples were built in then isolated sites. Through much of the nineteenth century the gembu were members of a Thami Og lineage. The office passed to a Nauje resident (Tsepal, who was a Golila-Gepchua Sherpa who had married a Nauje woman) in 1895 following complaints against the previous gembu's (Shangup Dorje of Thami Og) handling of the office (Fürer-Haimendorf 1979:120-121).

80. Ortner describes this as a political conflict between two pembu (1989:51). My impression from the Khumbu accounts is that the pembu system had not yet developed at this time, and that while Dzongnangpa was a political leader who had built autocratic power the lama was a spiritual leader who was interested in pursuing his meditations in a series of hermitages and in instructing followers. From this perspective the conflict resulted from Dzongnangpa's jealousy over Lama Sanga Dorje's spiritual power and popularity among Sherpas and his disciples and his resentment of lamas, their disciples, and religion in general. This same attitude underlay Dzongnangpa's later murder of the Zamde lama (see Ortner 1989:85-86) which he accomplished by destroying his spiritual power. Sherpa horror at these excesses (and specifically the anger of some of the Zamde lama's followers) led them to assassinate Dzongnangpa and end his terrorizing of religion and the religious. There may also have been an element of political competition as well, however, if it is true that the Zamde lama was the local political leader in the Tarnga area at the time Dzongnangpa arrived. Lama Sanga Dorje may also have had some political aspirations. There is a tradition that he had been advised by a high Tibetan lama to be a spiritual and temporal leader in the style of powerful Tibetan lamas. This desire may have been behind his emigration from the Bhote Kosi valley with his followers and their establishment of a new community in the wilds of eastern Khumbu.

81. Sherpas sometimes talk of a four-hundred-year tradition and note that the temple has been rebuilt three times after the initial structure was destroyed by an avalanche.

82. This is also referred to in a 1919 document granting nearby local land to the Tengboche monastery (Ortner 1989, appen. 2).

83. MacDonald notes that most of the small, early Khumbu "temples" were "not within the limits of human settlements" (MacDonald 1980:141).

84. The settlement area of Khumjung and Kunde, a narrow valley between the cliffs of Khumbila and an area of tremendous boulders, is popularly likened to a horse. A rock which is said to be the horse's head is located on high ground southeast of Khumjung.

85. Ortner (1989:93) discusses the establishment of the Khumjung temple and its possible relation to increasing village political power, speculating that the office of the gembu may have become established in the village at that time. According to Khumbu oral traditions founding the temple was related instead to the conflicts at the Dumje festival between Khumjung and Thamicho villagers. It was also made possible by the increasing population of Khumjung and Kunde and the growing wealth and power of villagers such as Nam Chumbi, a man who

established several shrines ( chorten ) in the area and who was a powerful and at times feared and hated pembu prior to his assassination sometime after 1830 (and according to one account in about 1851-52). In any case it is incorrect to relate events in Khumjung to the gembu, since nineteenth-century gembu were Thami Og residents. There are no oral traditions of any Khumjung man ever having held the office.

86. They also had fields at Chorkem, a saddle just above Nauje, where today only Nauje families farm. Chorkem in that era was an important place in Khumbu, for here Sherpas from Khumjung, Kunde, and Thamicho met Rais hauling grain from the lower Dudh Kosi valley and bartered salt for it. There is some evidence of an old, direct trail from Chorkem to the upper Bhote Kosi valley and stretches of an old trail along what is said to have been the main route from Pharak to Thamicho and Tibet can still be followed along the Bhote Kosi valley slopes below the site of Nauje. This main route bypassed Nauje taking a more direct line up the valley. Chorkem is also infamous as the place where Nam Chumbi was assassinated.

87. According to some Sherpas Khumbu families in the early nineteenth century may have thus been paying taxes simultaneously to both Tibet and Kathmandu. One campaign of the 1855-56 conflict is said to have been conducted in the Dudh Kosi valley. Nepali troops outflanked the Tibetan fort at Top Dara near Dzong Chu Tok below Nauje and with the help of Sherpas temporarily pushed the Nepalese zone of control beyond the Nangpa La to the area of Kaprak. After the war the area north of the Nangpa La was returned to Tibetan administration, but Khumbu remained firmly in Nepali control.

88. It seems possible that this may have been an already existing local system of administration that was simply validated by the central government as a convenient way to implement its rule in a remote area. The Sherpa institution of gembu may well be derived from the similar Tibetan office of the Tingri region (Aziz 1978:199-200), as may have been the office of pembu. The gembu was the most powerful local official in the region until the early twentieth century when gembu Tsepal abandoned his office and fled to Tibet following a second failed attempt on his life during which he killed three men. (According to Ortner (1989:116) one of these was the former gembu, Dorje, but Fürer-Haimendorf (1964:121) and Konchok Chombi do not agree that Dorje was killed.) His son Pasang assumed the office and carried out the gembu's duties for some years, but eventually moved to the ancestral family home village in Golila-Gepchua and gave up Khumbu administration about 1933. The office thereafter lapsed.

89. Both the offices of gembu and pembu were regarded as hereditary, although either could be relieved of his authority by the Nepal government and the office given by the state to someone of its choosing. This, however, has occurred only twice in the last hundred years, once with a gembu and once with a pembu, both around the turn of the century.

90. A few years earlier, in 1810, there had been a royal decree which allowed Shorung traders to trade as far north and Khumbu traders as far south as they wished (Schrader 1988:245).

91. Khumbu Sherpas, for example, made use of Nepali courts for several

major cases in the nineteenth century, including a dispute between the gembu and villagers of Khumjung over herding regulations and a land dispute between Khumjung and Phurtse villagers which involved a sacred forest. In both these cases the government interestingly deferred the final decision to local Sherpa officials. In one twentieth-century case Khumbu-based Nepali officials were accused by Sherpas of hunting illegally in Sherpa protected forests.

92. A village panchayat often did not correspond to a single village, but was rather a governmental unit established to administer a population of approximately 2,000-4,000 people and might encompass a number of communities. This was the case in Khumbu. In some parts of Nepal lack of correspondence between the boundaries of the new local government's jurisdiction and the individual villages had important ramifications for resource-management policies, because forests and pastures previously owned and regulated by a single village sometimes now came under the joint control of a number of settlements.

93. The authority of the pembu remains substantial. Besides collecting taxes several pembu also continue to settle local disputes and carry out other roles that they formerly performed. Two have held the office of pradhan pancha as well.

94. In an effort to promote regional economic growth the central government encouraged the development of a country-wide system of periodic markets. Weekly markets were established in the mid-1960s in the Solu-Khumbu district near the district center of Salleri in Shorung (originally at Dorphu and later at Naya Bazar) and at Nauje as well as lower in the Dudh Kosi watershed at Olkadunga and Aislalukarka. Within a short time market transactions in Nauje for grain and other products virtually replaced the old barter exchanges that Sherpas had previously made with Rais and lower-altitude Sherpas. The establishment of the weekly market, however, has not transformed Khumbu subsistence crop production into a commercial activity.