Persistence and Change in Golog Social Structure
Nancy E. Levine
University of California, Los Angeles
This paper considers past circumstances and patterns of change among the ethnic Tibetan pastoralists known as Golog during the second half of the twentieth century. It focuses on three features of sociopolitical structure: the encampment, the larger territorial groups which form a segmentary system, and the nature of political leadership. Reconstructing traditional social arrangements grows increasingly difficult with every passing year. Golog were incorporated in the Chinese nation-state nearly half a century ago, and only a few elderly people are able to provide clear accounts of former times drawn from personal experience. Even for these elders, the tumultuous changes of ensuing years complicate recollection. Turning to written sources is no solution. Early traveler’s accounts are superficial. Scholarly publications are few in number, nearly all of them have been compiled recently, and the information they present has been distorted by the politicization of modern Tibetan history. Despite these difficulties, such reconstructions have great value. First, they can offer insights into processes of social change, a subject of theoretical and practical relevance. Traditional cultural ideas and values have influenced and continue to influence individual responses to ongoing social and economic reforms and thus the shape of an evolving, modern Golog society. Put more simply, knowledge of the past aids in understanding the present. Second, such reconstructions make a major contribution to cross-cultural studies of social systems. Golog stand as a rare instance of pastoralists who lived independent of central states and modern markets well into the twentieth century; thus parallels with and differences from other societies in comparable circumstances can tell us much about the range of variation and potential for such societies.
The term Golog is an ethnonym used by dispersed populations living in Qinghai, Sichuan, and Gansu Provinces. These peoples share a common culture and speak distinctive dialects of Tibetan. Most are transhumant pastoralists who migrate between seasonal grazing sites, although some of the lower-lying regions include substantial agricultural populations, and certain Golog legends cite origins in farming areas. In many eastern Tibetan areas rainfall is abundant and the core of the pastoral economy is the yak, although in higher altitude and dryer regions, the pastoralists keep large herds of sheep and goats. These were subsistence economies, in which nearly all food and the materials of daily life were provided by the products of domestic animals, supplemented by limited trade with agricultural communities and easily accessible towns. In the 1960s, Golog were collectivized and in the 1980s, they were integrated in the market economy of a modernizing China (see Levine 1999).
This paper relies on field research among several pastoralist groups and on recent published materials compiled by Golog scholars. I carried out the first period of research, between July and October 1994, in Serthar County in Garze Tibetan Autonomous Prefecture and the second period of research, from August to September 1997, in Chigdral, Machen, and Matod Counties in Golog Tibetan Autonomous Prefecture.  The purpose of these projects was to trace changes in family structure, household economy, and pastoral management over the last several decades.
The first element of sociopolitical structure to be discussed is the rukor (Tib. ru skor), or encampment, which was commonplace among Tibetan nomadic pastoralists. The term referred to the people who lived alongside one another and herded their animals on collectively held tracts of grassland. Ekvall translates the term as “tent circle” (1968: 28), highlighting the fact that pastoralists traditionally lived in (yak-hair) tents and characteristically camped in a circular pattern. This is well illustrated in a photo taken in 1926 by Joseph Rock and reproduced in Clarke (1998: 48). The literal meaning is closer to group circle, however, having a sense similar to the idea of family circle in English. Rukor varied greatly in size; people recall memberships of three to ten and sometimes even more households. The number of households which camped together was due to the mix of local environmental, demographic, political, and social circumstances. Features of topography and the condition of the grasslands set an upper limit to how many animals could be based in a particular site. Because these conditions varied at different seasons and at different camping grounds, rukor might change in size over the course of a year. They would segment in smaller or less productive valleys and expand in more favorable environmental conditions, leading to annual cycles of fission and fusion. Demographic variations led to lasting changes in group size. A group would segment when it grew too larger for its resource base or when it lost prime grazing lands to an aggressive, stronger group. It might adopt new members when too many of its own members had died or when it was enriched by the takeover of more desirable regions. Social insecurity--fears of warfare and theft--encouraged people to form larger rukor. By contrast, a single household might camp alone in very secure conditions. A rukor usually consisted of close relatives, typically parents and some or all of their married children, so that the breadth and durability of a household’s social networks also affected the size of its rukor.
Rukor were mutual aid associations, and one of the synonyms for them was rurog (Tib ru rogs), “households of friends.” The aid most often cited by Golog today occurred in the event of a death, when neighbors would help with domestic chores and bring lamas to conduct the funeral. In everyday contexts, rukor households would help one another search for lost animals, guard the animals against thieves, conduct seasonal moves, provide workers to substitute for someone who was ill, and so forth. Members of rukor households would travel together to distant markets to buy grain and other commodities and trade on behalf of a household unable to send a representative along with them. A poor household might borrow animals from rukor neighbors. Gelek (1998: 50) describes rukor households as complementing one another’s inadequacies, so that a household rich in livestock but poor in workforce would cooperate with one having surplus workers but inadequate herds. Similar sorts of aid also might be sought from close relatives outside the encampment. Those households in the most severe distress--due to the premature deaths of key adults--would link themselves a very close relative. Living in one tent or separate tents, their members would conduct all economic activities jointly. In these ways, the ties of kinship and of neighborhood reinforced one another.
During the collective era, the traditional rukor was replaced by more formal systems of labor cooperation, involving production teams. In Serthar these were known by various neologisms which combine Tibetan and Chinese words. For example there was the cun (Ch.) chhung (Tib.), literally "small village," and cun shog (Tib. shog), or “village division.” There also was the cun chog (Tib. phyogs), or “village section” and dzi chog (Tib. rdzi phyogs), a wholly Tibetan term meaning “herding group.” These groups were responsible for joint, coordinated productive activities and they characteristically included ten households, although their organization appears to have differed in the different communes. In Serthar, these hybrid terms still are used to describe groups which conduct various administrative functions and have rights to certain sectors of pastureland. They tend to be named after some feature of the territories they occupy.
Nowadays in Serthar, and to a greater extent in Golog Prefecture, the idea of the traditional mutual aid rukor has nearly vanished. When asked about it, even middle-aged people look perplexed, although a fuller description elicits some recognition. The closest parallels in Golog today can be found in concepts such as nye skor, “friends and neighbors,” ru khag, referring to the old production brigade of the collective period, and khyim khag or sbra khag, (pronounced wa khag), meaning house or tent section. In Serthar, people still speak of cun chhung and cun shog, and also use the term ru shog to refer to households that live in close proximity and jointly pasture their animals on the same sectors of grassland.
This raises a major question: why has the rukor diminished in importance rather than being revitalized under the modern household responsibility system, as has happened with so many traditional institutions and practices. I can suggest three possible reasons. The first is the increasing value given to autonomy: the view that in the current market economy households are responsible for their own well being and that assistance by neighbors is non-obligatory or depends on the strength of kin ties. From what people say, it seems that households that are economically successful under the new regime feel no special obligation to aid those less fortunate. There is, moreover, no sense of benefit to be gained from working together with a neighboring household whose labor force complements one’s own. The result is that every household is thrown on its own resources.
A second reason may be traced to the ongoing privatization and division of pasturelands. This had advanced further in Golog, where the field research was conducted at a later date. In 1997 Golog people had been assigned separate grazing lands in all seasons and were indeed grazing their animals separately. In 1994 in Serthar, only winter pasturelands had been divided, but even then people continued to graze their animals together, reconstituting what (by law) were separate grazing areas as collectively managed tracts of land.
A third reason is that rukor probably always were flexible and even superficial in a structural sense. This was recognized by Ekvall who wrote on his experiences as a missionary among Gansu Golog for eight years early in the twentieth century and subsequent interviews with Tibetan refugees in the 1960s. As he states:
The encampment itself is not a prime level of social structure in pastoral society, but is a by-product of security requirements or topography. In many districts it does not exist.... Where it does exist it varies in size from a small circle of five or six tents to a circle of as many as eighty tents. During the seasonal displacement, it may change according to topography, coalescing into existence on a plain or in a wide valley, and breaking up into individual tent sites in rough country. (Ekvall 1968: 28)
Cross-cultural comparisons may shed additional light on why rukor among Tibetan pastoralists no longer serve as a reliable source of mutual aid, and for this purpose Mongolia would seem an apt choice. Mongolia is a predominantly pastoralist society which has recently undergone transformations similar to China: from a centrally planned to a market economy. However there we find exactly the opposite response--despite the fact that the traditional encampment (khot ail) seemingly disappeared during the collective period. Nowadays the encampment has reemerged. These groups include from two to fifteen households depending on grazing conditions. Their members work together, either pooling their labor, subdividing pastoral work amongst themselves, or rotating various herding tasks (Cooper 1993: 159). Working together provides a means both of complementing labor imbalances among encampment households and of providing risk insurance, in that households experiencing temporary crises, due perhaps to illness in the family, can rely on the assistance of their neighbors. The comparison with Mongolia thus magnifies the puzzle and raises further questions. How important was the Golog rukor in the past? What determined greater or lesser cooperation between neighboring households? Are Golog neighbors today actually cooperating more than they admit and, if so, why? One is left only with the common prescription: more intensive field research--and quickly, before it is too late to answer these questions.
Like the rukor, the larger tshowa (tsho ba) were widely found among ethnic Tibetan pastoralists. Unlike the rukor, however, the tshowa was a unit of sociopolitical structure that has maintained its importance through the present day. The term is translated by Clarke (1992: 399, 1998: 198) as “local group” and “wider camp group” and by Ekvall (1968: 29) as “amorphous lineage grouping” and “one of two or more segments of a tribe.” Tshowa traditionally included several hundred households and controlled large tracts of land, which were independently utilized by constituent encampments. While the tshowa generally included members of different clans, one such clan tended to have prominence in the group, ideally the clan which was believed to have founded it which also should be the clan of its leader.
These tshowa appear to have been segments in what anthropologists term segmentary political systems, which often are associated with segmentary lineage systems. Anthropologists describe the largest such groups as tribes. In a segmentary system tribes are divided into two or more subgroups, each of which may, in turn, be subdivided into smaller subgroups. Members of tribes recognize a common identity, but because they tend to include so many people and be spread over such large territories, they are not especially cohesive. And although many tribal members trace descent from a common ancestor, this linkage tends to be long past, so social ties are relatively superficial. The component subgroups are smaller in population and because they occupy a segment of the tribal land, their members often live in closer proximity than members of the entire tribe. Also, because they trace descent to a more recent ancestor--a branch of the founding ancestor’s clan--their relationships are based on closer kinship ties and are more cohesive. Not all the members of a tshowa will have the same clan affiliations, however. There were rules against marrying close relatives, and people had to choose spouses from other clans who often lived in different groups. Therefore every individual had kin and in-laws in an array of groups. Households that experienced economic difficulties and were not receiving enough help at home or that quarreled with their neighbors could join up with relatives in another tshowa. This is how new clan segments came to be introduced into the tshowa.
These generalizations are drawn from societies around the world and they apply to the Golog as well. Golog histories include genealogies which offer unambiguous examples of nesting layers of groups, specify main lines and branch lines of clans, and make conceptual linkages between territorial and lineage groups. I include the following quote with supporting material to clarify what might otherwise be an obscure rendition of people and places. The Tibetan terms are given in parentheses and explanatory material from elsewhere in the same text is included in brackets:
The descendants of dBang Chen ‘Bum subdivided into the Upper and Lower dBang Chen River [lines]. The upper river group subdivided into the rGyu and gSang sections (ru). The rGyu subdivided into dPon sKor and sGar sKor. Other Golog [lines descended from dBang Chhen Bum] are Thang ba, Gangs pa, mDa’ ba. [At the edge of the plain are five Shog Chung. ] The lower river tshowa number six: the two rDo ru, the two sKung dGon pa, the two Khra gLing, the two gTsang ‘Jam tshang, the two Wa sKor sGrung, and the two Ab Sogpo.
There are thirteen branch tshowa (ma lag): Kha Tsher, Yar lKog, mGon sKor, ‘Phas ‘Dzi, Drug Shog, Che bo, sGo ‘Dzi, mGar sKor, Khra rGyud, A ‘Bum, Khe sKor, ‘Ja’ sKor, and Rong sKor. Together with the dMu ba, there are twenty five further tshowa branches (yang lag). (Don-grub dBang-rgyal and Nor sDe. 1992: 34)
Clearly Golog understand their social structure in ways that resemble the segmentary lineage systems described for pastoral populations in the Middle East, East Africa, and elsewhere. Their traditional system involved nested groups, which were based on closeness of patrilineal descent and well defined prescriptions for alliance in conflicts. suggest that these sorts of systems arise in the absence of permanent, higher-level political units and that they enhance the likelihood of success in military conflict (Sahlins 1961: 323, 328). Anthropologists also argue that such a system is well suited to conditions of population mobility or instability, because it:
defines relationships between individuals, provides a basis for group formation, and specifies the relationships between groups in terms of a genealogical idiom and thus without reference to spatial contiguity or residence. By so doing, it provides individuals who are not stable spatially and are continually coming into contact with different people...with a stable social and political framework in which rights and obligations are specified, expectations are relatively clear, and various kinds of support are guaranteed (Salzman 1978:627).
Tshowa have maintained their importance through the present day, and one must question why their fate has been so different from that of the rukor. One could point out that many tshowa became collectives (Tib. mnyam las khang) in the 1960s and 70s, with the largest tshowa subdivided into two or more collectives. Under the present governmental system they are recognized administrative units, which are known as cun in Chinese. However this simply leads to another question: did tshowa maintain themselves because they were coopted by the new bureaucracy--possibly because they were an obvious analog to the agrarian villages with which administrators were more familiar? Or did they maintain their importance because they were so central to pastoralist thinking about their social world? Available evidence suggests that the persistence of tshowa is not simply a function of their administrative role. First, one can contrast ideas about the tshowa with ideas about the xiang, or rural township, the most important level of the modern administrative structure. Xiang have little significance in people’s self-identification or everyday discourse, while people’s tshowa membership--together with their clan membership--profoundly define their identity, even in the modern social world. Second, as soon as the collectives were disbanded, they were forgotten and expunged from people’s concepts of group affiliation. Although the larger tshowa were divided into several collectives (and mature people remember which collective their household belonged to) these memberships have no relevance in the modern day. Cross cultural comparison is helpful here, for we find that in many societies in the world segmentary frameworks of social organization and systems based on clanship have proven to be extremely persistent, even under conditions of radical social change (Salzman 1978:629 - 631).
Feudal Leaders or First Among Equals?
All tshowa had leaders. The primary leader was known as the ponpo (Tib. dpon-po, also tsho dpon). At this point in time it becomes very difficult to assess how much authority and power these leaders held and what the closest analogs were in a cross-cultural sense. There is a wide range of variation in the extent of stratification in segmentary political systems in pastoral societies (Khazanov 1994: 146 - 47).
One pole of comparison is to be found among Middle Eastern pastoralists, who recognize a similar segmentary lineage system and see membership in known clans as a prerequisite for status in the group. These groups displayed a fairly egalitarian system of leadership in which the leader could be described as in which the leaders hold few important advantages over their followers. As one anthropologist describes the situation:
Though sheikhs were normally recruited from only a minority of a tribe’s lineages, this never resulted in the evolution of a social hierarchy because each Bedouin lineage considered itself to be the equal of any other. In fact Bedouins rarely displayed much formal respect toward their leaders and felt free to argue with them and dispute their decisions. The office of sheikh had little inherent power because the ability to command without consent was severely limited. Leaders in such egalitarian tribal organizations gained their positions by displaying special skills in mediating problems within the tribe or successfully organizing raids and wars against other tribes. A sheikh also required access to considerable wealth in order to provide hospitality to visitors and to aid destitute members of his own tribe (Barfield 1993: 82).
There are some differences: unlike Golog, these leaders did not automatically inherit their father’s position; and, as will be discussed below, some Golog leaders may have held more power than their Bedouin counterparts.
More authoritarian leadership was found in Southwest Asia. Here, however, unlike Golog, these groups were extremely large, including several hundred thousand people who made no pretense of common ancestry and sometimes even spoke different languages. Their leaders were more authoritarian and “had the power to compel their followers into action and to enforce their decisions by coercion if necessary” (Barfield 1993: 107). Their main role lay in representing their group to powerful outside states. Consequently they often lived in the cities of those states, where they were important figures in state politics (Barfield 1993: 111-112). In Serthar, there was a confederacy of different tshowa, who claimed no common clan ancestor. However all believed that they were Golog and even the confederacy leaders lived as pastoral nomads within their own tshowa.
Nonetheless the view of entrenched inequality is supported by a number of historians in China. Their view may derive from politics in the collective period, when nomadic society was seen in terms of a class model.  The ponpo were set apart as feudal leaders (brog bdag) who controlled productive resources, here the grazing lands. As a recently published Golog history puts it:
Speaking of individual districts, people said that the land belonged to the general community, but really the leaders had all power of dividing and using the land. The leaders also had the power to rent and sell the land to other districts (tshowa). Generally nomadic people were the subjects of the leaders. So, they were definitely responsible for taxes and corvee labor. After the land ... had been divided by the leaders, they had the right to graze their animals on it.... The leaders were the owners of the four corners of the land. The ordinary nomadic subjects were able to make use of the grass. As in the proverb, the essence of the land belongs to the leader. There were two different ways of dividing the land. First, from the leader, the land was divided into small districts. Second, it was divided for monastic territories (sde) and individual households. Those who were politically powerful, those who were economically well off in productive resources and the monasteries directly took possession of the lands divided by the leaders. Ordinary nomadic subject households--with the exception of land used collectively by the tshowa--did not have long-term rights over the land. The tshowa leaders had power and could take back the land divided to each group (shog kha) and allocate it back and forth, as they wished. Nomadic subjects who crossed the border’s of another group’s (shog kha) territory to graze their animals would be punished. For other grasslands, the leaders also controlled the rents for lands they had rented out. In short, while the general community says that it has power over the grasslands, except in name, the leader had use-control over the land. That is the main characteristic of feudal power (Don-grub dBang-rgyal and Nor sDe. 1992: 51 - 52).
There are various arguments to be made on each side of the argument. On the one hand, modern-day recollections of the past stress that tshowa leaders were no richer than anyone else. Some people say that these leaders had no means of accumulating property. On the other hand they had various kinds of power and authority which could be used to economic advantage, such as determining where their own households would camp and graze their animals. In some cases, people recollected leaders selecting the best sites for themselves. The kind of power held by the paramount leader (dpon chhen) of the Serthar confederacy is another matter. Without a doubt he was extremely wealthy; without question his household and his group controlled the richest grazing lands in Serthar. Because such leadership was hereditary and ordinarily passed from father to son within one clan, these advantages were fairly secure. One gets the sense that only major mismanagement would result in a leader being deposed or abandoned by his tshowa. While there are recollections of such events, they seem to have been uncommon. Another indication of entrenched, structural inequality is that recompense (wergild) for the death of a tshowa leader was double that of a poor individual and the death of a member of his clan was intermediate between that of the poor person and the leader himself. Recompense for the death of members of the paramount Washul (dBal Shul) clan also was more costly. There was no distinction between males and females in this--one indication of the relatively high status of Golog women.
In order to better resolve this issue one can examine some of the other responsibilities of leadership. For one thing, a tshowa leader was expected to guide his group in disputes with other groups over land and animal thefts. He was expected to sponsor providing aid to a household that was in need, help a family in mourning, give generously to monastic institutions, and organize armed parties to fight against enemies or retrieve stolen animals. Leaders furthermore were expected to settle and assess fines for murder or animal thefts and to decide when it was acceptable to hunt or uproot the soil to look for herbs, these being religiously proscribed activities. In Serthar, the ultimate recourse in legal disputes was the leader of the confederacy.
As noted above, leadership was fairly stable within certain clans and ordinarily would pass from father to son. If there were no sons, it would pass to someone of the same clan. If there were no representatives of that clan or if there was a truly exceptional person of another clan, it could go to a new clan. Conversely, if there was more than one son, they would lead jointly, although it might happen that one brother would demonstrate greater leadership abilities and become the sole leader. If the tshowa was growing, it also might segment on the line of those brothers and their descendants, with each new tshowa taking control of separate valleys.
However the leader’s power was circumscribed. Each tshowa also contained several persons labeled “ministers” or those who were “second-in-command” (blon-po) and also leaders of smaller subgroups, which were commonly known as bchu dpon, “leaders of ten.” The latter categories of leader were selected by ability and were supposed to come from a different clan than the tshowa leader. People to whom I spoke said that a competent second-in-command could compensate for an incapable leader, who otherwise might cause irrevocable damage to his tshowa. The second-in-command was expected to be a persuasive speaker who could settle disputes between contending parties within and with other tshowa. He and the leader were responsible for determining the times of seasonal moves. The subgroup leaders, or “leaders of ten,” were selected from the most competent herdsmen. It was their responsibility to consult with individual households about the condition of the grazing lands and when it would be good to move camp and to convey that information to the other leaders.
Judging from the information at hand, it seems that the traditional system of leadership was not heavily oppressive and that the tshowa leaders held only minor economic advantages over their fellows. Since pastureland is used collectively, it would seem inapt to say that the leader owned the land, as a landowner might have exclusive use-rights among an agricultural group. The leader could determine which households in which encampments had grazing rights to which plots of land, and this power certainly could be used to advantage his own family and the families of his relatives and friends. However he could not exclude anyone from the land or take it all for himself. The presence of other leaders in the tshowa and local mediators (jo tha, gzu ba mkhan, or jo dpon) also circumscribed his power. A system of checks and balances also affected the decisions of the confederacy leader. Another factor to consider is that many nomadic tshowa stood outside the confederacy. They were termed “outside groups” (phyi sde) and governed themselves, but probably allied with the “inner groups” (nang sde) for protection against mutual enemies.
NEEDS SUBSTANTIAL REWRITING
How does one conclude a paper which consists of unresolved questions about the nature of traditional nomadic social structure? One can hardly be optimistic about obtaining satisfactory answers, since the necessary data will become increasingly more elusive with each passing year, as the memories of elders fade and as those elders pass on. Thus it would seem highly worthwhile to try to acquire the necessary data as soon as possible, in order to paint a fuller picture of the traditional society before it becomes too late.
Barfield, Thomas J. 1993. The Nomadic Alternative. Englewood Cliffs: Prentice Hall.
Clarke, Graham E. 1992. “Aspects of the Social Organisation of Tibetan Pastoral Communities. Proceedings of the 5th Seminar of the International Association of Tibetan Studies, Narita, Vol. II, pp. 393 - 411. Narita: Naritasan Shinshoji.
Clarke, Graham, ed. 1998. Development, Society, and Environment in Tibet. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.
Cooper, Louise. 1993. “Patterns of Mutual Assistance in the Mongolian Pastoral Economy.” Nomadic Peoples 33: 153-62.
Don-grub dBang-rgyal and Nor sDe. 1992. Yul mGo-log gi Lo-rgyus Deb-ther Pad-ma dKar-po’i Chun-po Zhes Bya-ba bZhugs-so. Xining: mTsho sNgon People’s Publishing House.
Ekvall, Robert B. 1968. Fields on the Hoof. Nexus of Tibetan Nomadic Pastoralism. New York: Holt, Rinehart and Winston.
Gelek. 1998. “The Washu Serthar: A Nomadic Community of Eastern Tibet.” Development, Society, and Environment in Tibet. G. Clarke, ed., pp. 47 - 58. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.
Khazanov, Anatoly M. 1994. Nomads and the Outside World. Madison: University of Wisconsin Press.
Levine, Nancy E. 1998. “From Nomads to Ranchers: Managing Pasture Among Ethnic Tibetans in Sichuan.” Development, Society, and Environment in Tibet. G. Clarke, ed., pp. 69 - 76. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.
-----. 1999. “Cattle and the Cash Economy: Responses to Change Among Tibetan Nomadic Pastoralists in Sichuan, China.” Human Organization (1999) 58:2:161 - 172.
Rock, Joseph F. 1956. The Amnye Ma-chhen Range and Adjacent Regions. Rome: Istituto Italiano per il Medio Ed Estremo Oriente.
Salzman, Philip Carl. 1978. “Ideology and Change in Tribal Society.” Man 13: 618 - 37.
Sahlins, Marshall. 1961. “The Segmentary Lineage : An Organization of Predatory Expansion.” American Anthropologist 63: 322 - 345.
 These two periods of research were supported by the Committee on Scholarly Communication with China (1994) and the National Science Foundation (1997), in both instances supplemented by funds from the University of California, Los Angeles. Thanks also are due to the National Center for Tibetan Studies in Beijing, for sponsoring the research and to Dr. Gelek and Mr. Dawa Tshering who aided me during field research. A special note of thanks is due to Khenpo Chonam, who is from Golog Autonomous Prefecture, and assisted me in translations of literary material on the area.
 One person said, "a household could stay by itself if there were many strong sons or lots of fierce dogs to protect against thieves."
 Ekvall states that in situations where rukor were permanent, they were “organized with a headman--usually the wealthiest or most able individual--and sometimes the encampment is known as his Ru sKor “ (1968: 28).) He does not, however, indicate whether he is talking about Golog and/or other Tibetan pastoralist populations.
 One person said, “the position of leader goes to the oldest legitimate son first, an illegitimate son second, or, if there are no sons, to a man selected by the ‘second in command’ (blon po) or a respected lama. It has to stay in the correct clan. But if there is someone exceptional or the proper clan has no members, then the leader’s lineage can change.”
 There also were shog dpon --some of the larger tshowa in Golog were called shogpa (also known as shog chen). This term literally means wing, and these wings were internally subdivided into shog chung, or “little shog.” At the highest level were
leaders of overarching confederacies, which linked many tshowa, like the former leader of the Washul (dBal Shul) Serthar, Rigdzin Dondrup.
 At that time pastoralist families were assigned to class identifications. These included nomadic leaders, (brog bdag), rich nomads (‘brog phyug), three middle level groups, (collectively called ‘brog ‘bangs, nonetheless pronounced drogdrang), and poor nomads (known as ‘brogpa dbul ‘phongs).
 In addition, as is typical for uncentralized pastoralists peoples, the recompense for a death would be settled more easily and at lower cost within the group. It would probably be uncollectable for very distant unrelated groups.
 There also are accounts of bchu-lnga dpon or bchu-drug dpon, leaders of fifteen or sixteen households, and leaders of five households, or lnga dpon. As Clarke has noted for Hainan nomads (1992: 398-99), it is common to find units of social organization assessed in terms of groups of tens, hundreds and thousands of households, which seemingly derives from the Mongol system. For another example, one of the tshowa I visited in 1997--Hor sKor ma in Ma sTod county-- was described as having traditionally included 1,000 households.