Pirie art

Since 1958 the Tibetan pastoralists of Amdo, on the north-eastern part of the Tibet plateau, have been subject to Chinese rule. Initially there was a period of conflict and violent confrontation during which many Tibetans lost their lives or were imprisoned. There followed, during the Cultural Revolution, a crack-down on religious activities, the destruction of monasteries, the deposition of Tibetan leaders and the collectivisation of agricultural and pastoral activities. After Mao’s death in 1976 came a period of reform and liberalisation which saw the return of livestock to the ownership of the pastoralists’ tents and their reorganisation into population groups around the new urban centres on the grasslands. These herders have become relatively autonomous again in their pastoral activities but the Chinese state continues to exercise a form of highly centralised administrative control, regulating trade, economic development, property rights and family size and the police continue to take responsibility for law and order. Few escape their grasp and it is accepted by Tibetans that imprisonment is likely to result from their involvement in a serious fight or a killing.

The pastoralists are, nevertheless, adamant that mediation by one of their own specialists, ultimately a Buddhist lama, is necessary before a conflict between them can be properly settled. Otherwise, violence will be met by counter-violence in escalating degrees as the communities of the protagonists are drawn into the conflict on the basis of group loyalty. Even if a man responsible for a death is imprisoned, his family must pay blood money in an agreed amount in order to avoid reprisals. Two forms of law, therefore, operate on this part of the Tibetan plateau. Government agents control many aspects of pastoral life and mete out criminal punishments to those who do not act, in their eyes, in the interests of the Chinese state. At the same time, the Tibetans maintain relations amongst themselves based on principles of loyalty and revenge, culminating in processes of mediation over which religious leaders exercise authority. This would appear to be a straightforward case of legal pluralism, involving two different legal systems, Chinese and Tibetan, each with its own norms, processes and rationales, one based on the notion of state control and the other on ideas of loyalty and [page 78] revenge. It is only relatively recently that the former has been superimposed on the latter.

There has been a tendency in studies of legal pluralism, as Roberts describes (1998: 105), to analyse legal phenomena in terms of distinct legal or normative ‘orders’, or ‘plural fields/domains/discourses/systems’.^{^[1]} Santos, for example, refers to different scales of law in Mozambique – local, national and global – as having their own legal norms and rationales (Santos 2006: 45). He suggests a concept of ‘hybridisation’ to make sense of the complexity that has resulted from the imposition of a succession of political or legal cultures – colonialist, socialist and democratic – one upon the other (Santos 2006: 47). However, this is not, I would suggest, the best way to make sense of the legal phenomena I am describing here.^{^[2]} There was never a monolithic Tibetan or ‘traditional’ legal system on this part of the Tibetan plateau. Before the arrival of the Chinese order was maintained through overlapping spheres of political, administrative, religious and judicial power. The control that the government’s agents now seek to exercise has added to this complex but does not represent a distinct legal ‘system’ with its own norms and rationales. There are some areas in which the Tibetans resist state norms but others in which they are adapting to new social values; there are cases in which the authority of the police is rejected but others in which pastoralists actively engage with the government’s representatives.

This, I would suggest, is important because the analysis of legal pluralism in terms of distinct systems tends to lead to one of two approaches, either of which might appear to explain the Tibetan reactions to Chinese government control. The first of these is to regard legal relationships in terms of domination and resistance, a type of model that [page 79] has become widespread in political anthropology. This is the way in which Santos develops his analysis of the role of traditional legal authorities in Mozambique, suggesting that it represents a ‘claim to an alternative legal modernity’ (Santos 2006: 61). The notion of resistance has some but, I would suggest, only very limited application in Amdo.

The second approach is to regard the use of distinct forms of law in terms of ‘forum-shopping’, a term coined by K. von Benda-Beckmann (1981) to indicate the strategic use of law and legal arenas and the political interests that may motivate the choices people make between them. In Amdo the room for such manoeuvre is, however, limited and the appeal to different norms and legal authorities is often simply a case of adaptation to an unavoidable form of control. More importantly, however, as I seek to show in this article, such choices on the part of the Tibetan are determined by ideas and concepts about what order is and the sources from which it can be imposed, more than by any strategic or political interests. Indeed, the choices made by the pastoralists sometimes go against their own best interests.

My suggestion is that rather than thinking in terms of different legal systems, the issue of norms and that of the exercise of power – political, judicial or administrative – must be kept analytically separate if one is to make sense of the relations between the Tibetan pastoralists and the Chinese government. There are cases in which the Tibetans resist the authority of the government’s representative on the basis that they seek to enforce inappropriate norms, for example restricting pastoral movements; there are others where, by contrast, they welcome the possibility of an external authority, for example to restrain the outbreak of violence; however, there are others where they only accept a very specific form of authority, that of a Buddhist lama, to resolve conflict.^{^[3]}

The term ‘Amdo’ is used by Tibetans to refer to parts of the Chinese provinces of [page 80] Qinghai and Gansu, an area roughly the size of France. The valleys of the northern and eastern borderlands are home to farmers who have long mingled with the neighbouring Han Chinese and Hui Muslim populations. However, the Amdo heartlands, which extend west towards largely uninhabited desert and southwards to the valleys of Kham, consist mostly of rolling pastureland inhabited by nomadic pastoralists who depend for subsistence on large herds of yaks and sheep. Formerly these pastoralists undertook long journeys in order to trade their animals for barley, tea, and domestic and religious goods in the markets found at monastic centres. The establishment of roads and towns has now brought market centres into the grasslands but the production and trade of livestock remains the backbone of the pastoralists’ economy.

From the mid-thirteenth century the whole region, along with the majority of the Tibetan plateau, was dominated by Mongols, whose power over the western part of China only came to an end in 1724. Having been converted to Buddhism by Tibetan lamas, they supported the establishment of the Dalai Lamas’ regime in central Tibet in the mid-seventeenth century and a student of the Fifth Dalai Lama, Jamyang Zhepa, established the monastery of Labrang, which obtained considerable economic and administrative, as well as religious, power in Amdo. However, the Dalai Lamas’ government never succeeded in establishing political, military or fiscal control over Amdo and the region was not politically unified with central Tibet. After defeating the Qoshot Mongols in 1724, the Manchu Qing dynasty established Xining as the administrative centre of Qinghai, an area which roughly corresponded with what the Tibetans regard as Amdo. Their Amban exercised more of a supervisory than an administrative role, however, and even this was resisted by the Golok tribes^{^[4]} to the south. The Amban allowed the major monasteries, such as Labrang, Repkong, Kurdi and Taktsang Lhamo, to retain considerable power over the Tibetan populations in their areas. A number of secular leaders also retained the power they had established over large groups of tribes, sometimes amounting to small kingdoms.^{^[5]}

After the Chinese revolution of 1911 the Hui Muslims, under their leaders Ma Qi and Ma Bufang, came to exercise considerable influence over the northern part of Amdo, in particular the agricultural communities that had fallen more closely under Manchu control. They pushed south to challenge the authority of the Golok tribal chiefs, but never succeeded in dominating these areas (Clarke 1989). The Golok tribes, along with those of Sokwo, under their own king, and the tribes under the control of monasteries and secular rulers remained largely independent of their influence. Carrasco, drawing on a large number of early sources from the early 20^th century, states that:

These remote pastoral areas have been more loosely integrated into the political structures of the various states than the agricultural centres. (...) Feuding and arbitration, that is, direct action of the people concerned, are instead the main legal mechanisms. When trade routes go through these areas, the danger of robbers is ever present. The authority of Tibetan or Chinese state officials is at times enforced, usually through local chiefs, but at other times these areas relapse into complete independence. These areas form, then, Tibet’s ‘land of insolence’, regions always ready to fight for independence, which only occasionally are brought under the authority of the state. (Carrasco 1959: 221)

Until the Chinese occupation in 1958, therefore, there was no form of centralised political control or administration in Amdo. To the north the Manchu and, later, the Hui Muslims exercised a certain fiscal and administrative control over some of the (mostly agricultural) Tibetan populations. Further south, successive Mongol, Manchu and Muslim forces allowed Labrang, Repkong and other powerful monasteries, kings and tribal leaders to exercise relatively autonomous administrative power, appointing headmen to the tribes in their areas.

Robert Ekvall, who lived in Amdo in the 1930s and 40s, states that in some areas the monastery acted as an ‘autocratic ruler’ over an entire group of tribes (Ekvall 1939: 68-69). Typically, this involved appointing headmen to the local tribes. My informants in Labrang agreed that ‘before the Chinese came’ the monastery controlled ‘everything’: it was the jösa (rgyai sa) (seat of the king). Outside these areas some tribes were headed by an elder man, selected for his ability, in others a group of elders exercised authority, while others had hereditary rulers (Ekvall 1939: 67-70; Hermanns [page 82] 1959: 231). Golok, the largest tribal confederacy,^{^[6]} was famed for the ferocity and independence of its people, who were a constant source of fear for the surrounding populations (D’Ollone 1912; Rock 1956). Political and administrative control over the region was, therefore, exercised simultaneously by a number of rulers with overlapping spheres of military power, religious and political influence. Alliances between them could lead to a time of peace and stability, but competition for power could just as easily lead to war. Several of the tribes under Labrang, for example, were known as lharde (lha sde), protectors of the gods, a title conferred on those who had defended the monastery against threats from surrounding forces.

My own informants in Machu, an area formerly under the influence of Labrang, confirmed that headmen, gowa (’go ba), were appointed by the monastery to all the dewa (sde ba), tribes, of the surrounding area into which they were organised. These were monks, who held office for a term of three years, rotating between different dewa. The gowa himself approved the appointment of the headmen of smaller groups and was primarily responsible for settling ‘problems’ in the area. Intractable conflict that the respective gowa were not able to solve themselves was referred to the monastery, but in serious cases Jamyang Zhepa (in successive incarnations) or another senior lama would act as mediator. The monastery, therefore, exercised a certain amount of legal control in its areas. In the tribes beyond Labrang’s control either another monastery would appoint the gowa or they would be under the control of a hereditary ruler, such as the kings of Sokwo and Ngawa. These would perform similar functions to those exercised by the Labrang gowa.

Within the Golok confederacy different groups were ruled by hereditary families, between whom there were often relations of conflict, but who could also unite their forces to face external aggression. These ruling families traced their lineages back to common ancestors. They carried out similar functions to the gowa appointed by the monasteries elsewhere, determining pastoral movements and settling conflict. Nevertheless, even here senior lamas could be called upon to mediate serious feuds. These could be from monasteries outside their own areas. Khongtang lama, one of the most senior attached to the Gelukpa monastery of Labrang, was, for example, recently called upon to mediate a feud between different Golok tribes, even though these groups pay their primary allegiance to the Nyingma monastery of Pelyul.

Ekvall (1964, 1968) analyses the types of conflict that occurred amongst these pastoralists in the period prior to the Chinese occupation and their elaborate methods of mediation. He talks of the pastoralists’ propensity to ‘aggressive personal decision-making, quick and dramatic responses to exigencies and the willingness to take calculated risks’. He refers to their frequent recourse to reprisal and the ‘effort in persuasion, mediation and the good offices of mutual friends made ... in the attempt to keep antisocial attitudes from developing into antisocial actions.’ Ultimately, mediation required securing the protagonists’ agreement to the acceptance of indemnity as a substitute for reprisal. This might be achieved by the pressure of community consensus but when conflict developed into warfare, or threatened to do so, mediators had to be ‘important men from other polities – chiefs, famous lamas, noted orators, and men of affairs – who can only meet in competitive ostentation after considerable delay,’ (1968: 75-80). Similar forms of mediation are found on the plateau today.

Ekvall also notes that ‘folk law’ generally involved reprisals, while ‘canon law’ (Buddhist law) embodied ideals of non-killing (Ekvall 1968: 76). However, it was impossible for the pastoralists, dependent on their livestock for subsistence, to comply with such religious norms. He might also have pointed out that the Buddhist condemnation of anger is directly at odds with the norms of revenge maintained by the Amdo pastoralists. Nevertheless Buddhist lamas were, and still are, regarded as the most potent mediators and the settlements they achieve are invariably based on the negotiation of compensation as a substitution for (legitimate) reprisal. There was, therefore, no uniform set of norms on the Tibetan plateau any more than there was a centralised judicial power.

In the absence of a state-like structure and a uniform, or centralised, legal system, therefore, the social order of these pastoralists was the product of a number of different configurations of power. Peace was largely dependent on the existence of truces and alliances between the various rulers. Conflict between tribes, or groups within them, needed to be settled by skilled mediators before it escalated into war-like proportions. Informal mediation or community consensus was employed to restrain and settle conflict at lower levels but skilled orators and reincarnate lamas were needed in the most serious cases. A set of common norms governed practices of feuding and mediation, but enforcing restraint and agreement to compensation was a difficult achievement. The authority of even the highest mediator could be contested, as is graphically described by Ekvall (1964). Individuals were expected to be defiant [page 84] and it was accepted that order might only be reestablished by calling on the authority of a mediator from another area. Was this, then, a case of legal pluralism? It would be difficult, I would suggest, to identify any element that could properly be labelled ‘law’ here. There were norms, but the extent to which they were accepted and followed was very much dependent on the authority of those who sought to enforce them.

These administrative and religious systems were substantially overthrown by the arrival of Chinese forces in the 1950s. Initially there was little change but an uprising in eastern Tibet in 1956 led to a tightening of Chinese rule and, eventually, to the flight into exile of the Dalai Lama in 1959. In Amdo the people refer to 1958 as the date when the Chinese ‘came’ to the region, when they crushed final resistance and began to exercise a close control over the pastoralists’ economic and religious affairs. Over a number of years they were organised into groups for collective herding, all the livestock and its produce being taken into state ownership. The older generations who lived through this period talk of the closure of the monasteries, the persecution of the monks, many of whom were either imprisoned or forced to take wives, the battles during which hundreds of Tibetans perished and the hunger they endured during the collective period. The Cultural Revolution, which began in 1966, saw the most consistent persecution of former leaders and the elevation of those deemed to be from the poorest families into superior positions. One old woman told me that this had been a good time for her, but the vast majority of my informants shuddered at the deprivations they had endured.

After the death of Mao in 1976 a period of reforms swept through China, reaching the Tibetan plateau in the early 1980s. The ‘Household Responsibility System’ was a policy introduced to try to promote economic development, which had singularly failed to occur under the collective system (Goldstein 1994: 98-99). The animals were returned to the private ownership of nomad families, being distributed according to family size, and along with this went infrastructure developments and land reforms. After a few years all restrictions on the sale and disposal of animals were removed and the nomads became almost fully autonomous again in their pastoral activities. At around the same time most of the monasteries were allowed to re-open and monks and senior lamas were released from gaol. The redistribution of livestock has done much to balance out what were previously great disparities in wealth. Formerly many families had no livestock and had to work as servants for the rich, although the class system, which saw aristocratic families elevated to a higher social status in central [page 85] Tibet, did not extend into Amdo. Initially every family was given enough animals for subsistence, although imbalances are now reappearing.

For administrative purposes the authorities have divided the population into xiang, units of one to two thousand people. These have largely been mapped onto the previously existing dewa, in terms of both population and territory. There has, however, been some reorganisation of nomadic groups. In Machu county, where there are ten xiang, Ngulra, the largest dewa, has been divided into two xiang and two smaller dewa have been combined into one. A defunct horse-breeding station has also evolved into a new xiang of nomadic families, named Matang. The Tibetans still use the old names and regard Ngulra as a single dewa, but also recognise Matang as a new dewa. The xiang are governed, as far as the authorities are concerned, by representatives who join the Communist party. These are responsible for transmitting orders issued by the governmental authorities at county level, collecting taxes, organising livestock vaccinations, and so on. As far as the pastoralists are concerned, however, organisation within the dewa is undertaken by a council made up of the gowa of the repkor (ru skor), the smaller groups into which each dewa is organised. The repkor consist of thirty or forty tents, that is around two hundred people, and they often have an identity which pre-dates the Chinese occupation. Their gowa are selected in the repkor meeting, normally holding office for a number of years. It is two or three of these who are put forward as xiang leaders to the authorities (and they have to join the Communist party in order to take up their appointments) but within the dewa they are not regarded as having any particular authority. The council, as a whole, takes the place of the gowa formerly sent by the monastery or supplied by a hereditary ruling family. Their tasks involve making arrangements for the visit of senior lamas, dealing with government authorities and dealing with the ‘problems’ that arise in the dewa, by which they mean cases of theft and conflict. Chinese rule has, thus, resulted in a certain (unintended) democratisation of dewa organisation.

The position is rather more complex in Golok, where the tribes formerly made up a larger confederacy. Here I found that local control over the leadership of the xiang is much less strong and there is no reference to repkor (as also noted by Levine, n.d.). My informants insisted that the new leaders are government appointees, rather than their own gowa, here known as xhombo (dpon po). Given the previously strong leadership of the larger tribal groups here the repkor were, in any event, probably of less significance, while members of the old ruling families were systematically persecuted and imprisoned during the Cultural Revolution. Now the xhombo just do the work of mediators, they told me, but former tribal loyalties are still strong and [page 86] emerge at times of conflict. Moreover, the xhombo still play a crucial role as respected and effective mediators.

Significant material changes introduced by the government have been the establishment of new towns and the construction of a network of roads, a programme which has received new impetus with Jiang Zemin’s ‘Develop the West’ campaign initiated in 1999. These have resulted in major material improvements in nomadic life, as they bring trading points and markets for food and household goods within easy reach of most families. Most nomadic movements can now be conducted by truck, which has significantly eased this laborious process. Schools, health care, post offices and telephones are readily available in the towns and the headman of the group I stayed with in Machu described his plans to improve the educational possibilities for the children by asking the local authorities for assistance with accommodation in town. Many educated Tibetans are able to get salaried employment there, within government offices or the police force, as teachers, medical assistants and so on. Indeed, many of the local government officials are now Tibetans, although the higher level regional administrators are still predominantly Han Chinese. In many ways, therefore, the pastoralists have come to accept the norms of modernity: they value education and respect those who obtain positions as teachers or in government. They have adopted new standards of dress, cleanliness and material wealth and are taking advantage of new economic opportunities.

The authorities exercise strict control in many areas of nomad life, however. As well as levying animal taxes, they oblige the nomads to vaccinate their animals, a process that the pastoralists regard with suspicion and resentment, not least because they have to pay for vaccines they regard as useless. They also enforce a population control policy, requiring women to undergo sterilisation after their third child. My informants were never very clear about what sanctions the authorities would impose for non-compliance: ‘they force the women to have the operation’, they said, giving the impression that physical coercion was employed at times. The imposition of heavy fines on a family was also mentioned. This policy is deeply resented and some families evade these rules by giving a child to a relative with a smaller family, but most of the women of child-bearing age that I knew about had had the operation, or were resigned to having it when the time came.

Most controversially, the authorities are implementing a policy of fencing the pastureland, as part of the official regulation of pastoral practices, designed to promote ecological sustainability through private land ownership (Goldstein 1996: 12). Land is now officially held by individual families on 50-year leases (Banks et al. 2001), but [page 87] the authorities have also made it a requirement that the nomads should erect fences between their individual plots, instead of letting their livestock graze together with others in their encampment, as they would prefer. This is much resented by the pastoralists who complain that they can no longer ‘go anywhere’ with their animals.^{^[7]} Boundaries mean possession of land and scope for disputes between neighbouring groups and have been implicated in recent instances of violent conflict (TIN 1999). Some groups are able to evade these requirements by erecting temporary fences which make it look as if the pastureland has been divided. However most are resigned to the fact that sooner or later permanent fences will have to be erected. The view of my educated informants in Machu is that this is part of the government’s attempts to pin down and control the mobile and elusive pastoralists. These are cases in which the government’s actions are consciously resisted, although the scope for evasion is very limited. New sets of norms have, therefore, been introduced, some of which are resisted, others accepted. Very new structures of power and control have also been established, but without entirely undermining older forms of authority.

While the former dewa have been substantially recreated as xiang under the modern administrative system, the new boundaries and fencing policies have not allowed the pastoralists to revert fully to their formerly flexible land tenure arrangements. A whole repkor used to move as a single body, its pasture being used as common property. Only the winter camping grounds were occupied by the same families on a recurrent basis, building winter shelters of a semi-permanent nature and harvesting a crop of oats as winter fodder. Yeh argues that there has been a significant increase in conflict amongst the Amdo nomads since the area’s incorporation into the PRC, which can be traced to these territorial policies (Yeh 2003). They contradict old forms of socio-territorial identity, she says, precipitating boundary conflicts and giving rise to new forms of domination and resistance. There is much force in this but, as I have discussed elsewhere, the conflict she describes can largely be regarded as a continuation of former practices (Pirie 2005). The new territorial arrangements have occasioned new conflicts but the forms of retributive violence according to which they [page 88] are played out represent a historic tradition, based on social norms which have persisted into the period of modern state control.^{^[8]}

The Amdo Tibetans talk frequently and readily about both actual and potential violence and the men display self-consciously masculine qualities in their daily lives: most carry knives and the lead weighted ropes that they use to round up the animals are also potentially lethal weapons. They talk as if their neighbours are always just about to attack or steal from them and everyone could tell me stories of fighting and killings that had occurred within their families or villages within the last few years. One of the Ngulra gowa, for example, described a fight that had occurred between two men, one from Ngulra and one from the neighbouring dewa of Chocomama. The Chocomama man had been killed and men from his dewa came almost immediately to take revenge on two men of the killer’s repkor. The gowa told me that they had come right into the encampment to attack the people and this had made the men of Ngulra ‘angry’ and determined, as a body, to retaliate. A number of men explained to me that they ‘have’ to get angry if a member of their family has been killed and they must take revenge on a member of the murderer’s family. In the event of a serious theft of livestock it is also expected that men of the victim’s tent will get angry. Although they may be restrained from initiating a fight, if they happen to meet a member of the thief’s family, they would ‘have’ to fight him.

When a conflict occurs within one dewa, or two neighbouring repkor who normally maintain good relations, procedures for resolving it are relatively informal. Mediation is carried out by the gowa and there is considerable social pressure on the disputants to agree to a settlement. When 30 sheep were stolen from the tent in which I stayed, for example, the elder son, Jamku, went, with some friends, to identify the thief, on the basis of divination clues given by a monk. When they found him, in another dewa, Jamku declared his intention to fight immediately, but his friends restrained him. On their return to the tent members of his family and relatives persuaded him to let the gowa intervene and to arrange a compensation payment. The whole matter was settled in this way over the course of a few weeks with an agreement to return the animals and make an additional apology payment. This recalls the informal processes of mediation, and ‘community consensus’ in favour of peace which took place in the 1930s and 40s (Ekvall 1968).

If Jamku had initiated a fight, however, the violence could have escalated to include his friends, family and, ultimately, the whole dewa. Once retaliatory killings have begun, such a feud can only be resolved through more elaborate procedures of mediation. Injuries must be compensated for by payment for the value of the damage and death with blood money, mnyö rtong (mi stong). Acceptance indicates an agreement by the victim not to fight over the matter. Conflict between persons or groups from different dewas, which might easily result from a theft or a drunken fight in the county town, is always difficult to resolve, people told me. The process of mediation, in such cases, follows a set pattern and closely mirrors the accounts given by Ekvall for the pre-1958 period (Ekvall 1954, 1964, 1968). Outsiders, such as monks, initially intervene to establish a temporary truce so the parties can talk. Then it is the task of the mediators, zowa (gzu ba), to carry out the mediation. These are likely to be the gowa from neighbouring dewa or senior monks from a local monastery. A meeting is set up in a neutral place, now often the local town, with different groups in different rooms and the mediators acting as go-betweens. Their task, they told me, (and this is confirmed by Ekvall 1964) is not to determine the facts or apportion blame but to determine the appropriate level and nature of compensation. Blood money is calculated after the deaths of each side have been reckoned and set off against each other and depends on the identity of the victims and the nature of the killings. A settlement may also involve apologies, other payments and offerings of religious books. In the case of the Ngulra-Chocomama fight, for example, Ngulra would not accept the normal blood money because men had been killed within their own encampment, but the case was settled by the intervention of a senior lama, who ordered a small, but symbolic, extra payment by way of apology. The mediators may also establish where boundaries should run, deciding on a just division of pastures depending on the local history of land use. Crucially, the mediators have to convince the parties that the proposed compensation takes into account all these relevant factors.

Mediation is seen as a difficult and not always successful undertaking. In Golok the xhombo, members of the former ruling families, have a special status and are often renowned for their oratorical skills. Even here, however, the lamas are regularly appealed to in the most problematic cases. The most senior lamas at Labrang, Jamyang Zhepa and Khongtang (until his death in 2000), were very active, travelling throughout Amdo to settle major disputes. It was Khongtang who, as already mentioned, settled a long-running feud between two Golok tribes in the late 1990s, a conflict which had its roots in the period before 1958. They are seen as able to resolve conflict that is beyond the capabilities of local mediators. Tibetans always tell the truth in front of them, my informants told me, and the lamas suggest just solutions, jömdri [page 90] (rgyu ’bras), taking into account the history of the case. This was a very important concept and often attributed to the lamas’ wisdom and knowledge of religious texts, history and precedent. The Tibetans always made the point to me that the lamas were successful because everyone ‘believes in’ and ‘has faith in’ them. They are seen as capable of persuading reluctant and obstinate parties by appealing to their own religious authority. In such cases, therefore, although a conflict may have been caused by a new pasture division, or occasioned by the availability of cheap alcohol in the new towns, the course of retaliation and mediation follow patterns already described by Ekvall.

Now, when there has been a killing or a serious fight, however, the police invariably get involved. Some pastoralists successfully elude their grasp but culprits are generally caught and punished. Some years ago, for example, some animals strayed from the family with whom I subsequently stayed onto the pastures of the neighbouring repkor. When one of the boys went to fetch them he was badly beaten by the neighbours. The result was a fight between men of the two repkor, during which one was killed by the boy’s elder brother. Revenge killings would almost certainly have been carried out had the gowa of the dewa not intervened collectively to secure a truce. In the mean time the boy’s two elder brothers were arrested by the police and imprisoned. The younger was just held for a couple of months but the elder was incarcerated for eight years. Nevertheless, the mediators decided that their family should also pay blood money in an amount which amounted to more than the family’s entire wealth and the whole repkor had to combine to raise the money. The mediators also decided that the family should be expelled from the area for three years, to remove them from contact with their neighbours. Following this, good relations were restored between the repkor and the family has now returned from exile without any lingering animosity.

The Chinese state system of criminal punishment is seen by the Tibetans, in such cases, as completely inadequate for achieving a final settlement to their disputes. ‘They cannot produce just solutions they do not care about the history’, my fieldwork assistant said dismissively. By treating violence and murder as crimes to be punished by the state, the police fail completely to engage with the nomads’ norms of retaliation and compensation. Once fighting has erupted, the norms of retaliation rule supreme and the police, refusing to order compensation in accordance with these norms, are unable to settle them. This would, therefore, appear to be a case of resistance to Chinese authority.

If one focussed solely on the norms and processes of conflict resolution it would appear that there are two different legal orders on this part of the Tibetan plateau. The state sees itself as responsible for order and regards violence as a disruption to that order, representing a crime against the state and requiring punishment. For the Tibetans, on the other hand, this type of retaliation represents a form of relations between groups which follows its own logic, governed by expectations of retaliation, loyalty and compensation. In many cases the Tibetans are submitting themselves to a form of double punishment, adding their own compensation payments to the criminal punishment imposed by the representatives of the state. In doing so they are denying legitimacy to the government’s system of criminal punishments. Does this amount to the assertion of an alternative form of law by the Tibetan pastoralists in the way Santos describes the role of traditional legal authorities in Mozambique? He suggests that the issue of legal pluralism often involves contests over the legitimacy of the power to exercise political and administrative control:

What is at stake is ... the relationship between the political control and the administrative control of populations and their territories, and particularly the question of the legitimacy of the power needed to secure either form of control. (Santos 2006: 68).

He describes the continuing role played in dispute settlement by traditional authorities as a case of resistance to the legal order of the state and to a global (Western) modernity. Their activities represent ‘a claim to an alternative modernity’ (Santos 2006: 61).

In Amdo, while there is a normative line that can be drawn between Tibetan and governmental attitudes to conflict, I would suggest that it is not so clear that the pastoralists are claiming some alternative to the government’s system. The government’s representatives are not, in fact, regarded by them as entirely without authority. When violence breaks out, or threatens to break out, for example, they are seen as offering an additional source of decision-making, which could, at least potentially, be effective in resolving problems between dewa.

The establishment of the new dewa (or xiang) at Matang, for example, has given rise to a continuing boundary dispute with Ngulra, the adjoining dewa. During much of my fieldwork I was staying in the Ngulra repkor which immediately adjoined Matang and [page 92] whose summer pastures were the subject of this dispute. The Matang tents were visible less than a mile away across the pastures. As my field assistant put it, Matang had asked for some land, which Ngulra had given to them, but then they had started demanding more so the Ngulra gowa had had to tell them to stop. Members of his encampment were going around canvassing opinion on the possibility of a fight, asking if other men were ‘angry’ about the situation. At the same time, the gowa was constantly worrying about violence breaking out. At one stage, early in my fieldwork, he announced that he was going to have a meeting with the ‘chief gowa’ to try to resolve the dispute once and for all. I assumed that he meant senior gowa in the dewa, but it turned out that he was going to meet with government officers in the local county town. When this did not work he then took a party to petition the authorities in the district head-quarters at Gannan. Why, I wondered, should these Tibetans be asking the government’s officials to make decisions on what was, effectively, an inter- dewa dispute, especially when they had such elaborate processes of mediation at their disposal? Why were they not calling upon other gowa or lamas and engaging in a process of mediation?

I realised later that there was a significant difference between settling a boundary, something which is now seen to be under the jurisdiction of the state authorities, and settling a feud, which must be done through a proper process of mediation. The government had created this boundary problem and so should resolve it, was their attitude. In this case violent conflict had not yet broken out, although if it had, mediation and compensation payments would almost certainly have been required and ultimately they might have had to call in a specialist mediator. Many Tibetans, in fact, complained to me, as they did to Yeh, that the authorities could, and should, do much more to prevent and control nomad violence (Yeh 2003: 520). Even if they cannot produce ‘justice’ in mediation the Tibetans see them as having the power to put a temporary stop to violence so that their own mediators can step in. They are, in fact, resentful when the government agents do not use their power in this way. Calling a truce requires an authority which could be exercised by the police, given the resources at their disposal, even if they seek to enforce an alien set of norms.

The 50-year leases that officially define property relations in pastureland are, therefore, meaningless for the pastoralists and they resent the interference of these authorities in the definition of property boundaries. However, it is, in practice, accepted that the government’s bureaucrats now have the authority to made decisions in the case of boundary disputes. This is an area in which the pastoralists have always come into conflict with one another and, in the interest of avoiding escalating feuds, the gowa now turn to these officers to make the decisions that could pre-empt [page 93] violence. To an extent this involves recognising their authority to regulate relations between pastoralist groups. This is not just a matter of political expediency, however. The pastoralists’ history is one of warfare, feuds and uncertainty in inter-tribal relations. It is not, therefore, a wholly alien notion for them to turn to a powerful, and even distant, external authority, such as the state now represents, in order to regulate these relations.

There are certain cases in which the introduction of new laws by the Chinese state has led to resistance. In the management of livestock and family relations, in particular, the nomads resent governmental interference. The requirement that they should define and respect the boundaries between the property of individual tents impinges severely on the management of their pastoral activities, as well as the relations within and between their smaller groups. There is a clash of norms here and a certain resistance to the implementation of the government’s rule. However, the scope for this is limited and the state is able to prevail without much difficulty. The Tibetans have, on the other hand, adopted many of the norms of modernity, in matters of dress, education, health-care and even music. When it comes to the regulation of trade, education and health services, moreover, they have come to accept that governmental authorities are in control and that they have vastly expanded the functions formerly undertaken by monastic authorities and those in the larger urban centres.

These attitudes, I would suggest, are largely to be explained by the pastoralists’ ideas of what order is and how it should be imposed. There would be little disagreement between the pastoralists and government officials over the basic concept of order and the idea that an ideal society is free of violence and conflict. This is a clearly articulated norm among the nomads, even if those caught up in a spiral of retaliation also express an obligation to undertake violent reprisals (Pirie n.d.a). There are quite different ideas, however, concerning the question of the proper response to violence. The state is officially the guardian of order and its representatives are supposed to treat instances of conflict as crimes against the state, requiring punishment. The Tibetan pastoralists, on the other hand, regard violence as setting up a set of relations between and within groups, based on loyalty and retaliation. Nevertheless, this does not lead to a straightforward resistance to governmental legal control, rather to a selective submission to the authority of its agents.

The Tibetans’ reactions to the state’s attempts to impose order on them are, thus, complex and, I would suggest, shaped more by a set of internal pluralities than either resistance or the straightforward pursuit of their political interests. Although the norms of revenge govern the relations between groups, when it occurs within the repkor, the smallest social group above the tent, violence and theft are treated as crimes to be punished. When a number of adolescent boys were caught stealing some yaks from other tents in the repkor in which I stayed, for example, other members of the village expressed shock and outrage. They were caught and beaten severely by the gowa, with the help of their elder relations and the approval of the repkor meeting. People told me that if a killing occurred within a repkor the perpetrator’s tent would be permanently expelled. The expulsion of the family from its repkor, mentioned earlier, involved a similar element of punishment. In the case of conflict and violence that occurs between dewa, on the other hand, the idea is not to punish but to negotiate compensation. Moreover retaliation, as long as it does not involve excessive escalation of violence, is generally regarded as inevitable, even a necessary response on the part of the victim. There are quite different ideas here about the meaning of conflict and violence and about the proper response to it, ideas which shape the pastoralists’ relations with the state authorities.

A certain normative ambivalence also characterises their relations with Buddhist leaders. As described elsewhere, the lamas have a charismatic religious authority, based on their status as reincarnates, which means they have the power to subjugate the untamed spirits of the environment and the undisciplined thoughts of their students (Pirie 2005). It is this that gives them the authority to impose the norms of order on angry and defiant individuals and is one that secular leaders, whether the Tibetans’ gowa or government officers, could never match. However, the Buddhist norms of compassion and peace and the condemnation of anger, which are articulated by the lamas in other contexts, are not the basis of their authority over conflict or the solutions they promote. The lamas invariably suggest compromise based on the pastoralists’ norms of retaliation, including the payment of blood money.

The pastoralists thus adopt different attitudes to conflict depending on the context. They invoke different norms depending on whether they are discussing matters of religion with monks, reviewing the ‘problems’ of conflict on the grasslands or they have been personally subject to aggression. Similarly, they appeal to different sources of authority depending on the nature of the conflict and context in which it has arisen. It is the gowa and mediators, whose positions require them to take responsibility for the ‘problems’ of conflict, who are more likely to recognise the need to appeal to a form of external authority to reimpose order and, thus, to draw governmental [page 95] authorities into the relations between their own groups.

When violence has broken out the government authorities have come to realise that they frequently cannot achieve a solution without employing a Tibetan mediator. Especially in Golok, they often call on the xhombo to mediate conflict, even setting up a meeting place in the town for them, or calling upon a senior lama. In one remarkable incident, which occurred in the Golok county of Darlag, the local police had shot a monk who was making a protest about the police treatment of a group of monks in detention. A subsequent stand-off between the police and a large crowd was only resolved by the intervention of senior lamas, who persuaded the police to pay compensation to the family of the monk (TIN 2004). Here the police were forced not just to recognise the Tibetan norms of compensation, but also to comply with them, themselves.

As Tamanaha points out, what are regarded as state law, on the one hand, and as an indigenous or traditional legal system, on the other, are generally of quite a different character (Tamanaha 1993: 210-211). State law is often invoked to justify the exercise of power by a handful of elites, and often only in certain areas, displaying indifference to more general societal normative ordering. It can, in such cases, be more an instrument of power than a mechanism for maintaining social order. As such, is plays quite a different role from the lived norms of social order, those that are often classified as ‘indigenous’ or ‘traditional’. When considering issues of legal pluralism, that is, the issue of norms must be distinguished from questions about the authority of those who maintain order.

While we should expect a certain resistance, among the Tibetan pastoralists, to the imposition of state control, therefore, we should not see the continuing assertion of indigenous norms concerning the management of conflict as the assertion of an alternative form of legal ordering. Elites have been exercising power in Amdo since well before the arrival of the Chinese in 1958 and in the case of the lamas it is a power that has been welcomed as a source of order. It should not, therefore, be surprising that the Tibetans appeal to the authority of the police and state agents when it comes to the basic control of violence. Moreover, there have long been pluralities of norms in Amdo, particularly concerning the management of conflict, which mean that they often pursue revenge, even when it is patently against their own interests, or [page 96] those of their own groups.

I would agree with Roberts’s suggestion that the issue of legal change should be thought about in terms of ‘internally shaped responses’ to a particular environment (Roberts 1998: 102). Historically there has been a host of competing norms, expectations of judicial authority and power struggles among the Tibetan pastoralists, many of which persist under the vastly different circumstances of control by the Chinese state. Sources of authority have shifted over time and new power-holders have been incorporated into the legal landscape. The arrival of the Chinese in 1958 and the development of their administrative and legal control over the Tibetan populations, has added to this complexity. The modern administration has introduced the idea of centralised control, but even its officers, many of whom are now ethnic Tibetans, are adapting to the practices of the pastoralists. Likewise, adjustment by the Tibetan pastoralists to the government of the Chinese state has resulted in considerable adaptation of their social practices and expectations, particularly concerning trade, services and the determination of boundaries. In many cases new social ideals have been accepted.

Legal relations among these Amdo pastoralists are dominated by the dynamics that exist between state and society (Pirie n.d.b) and between tribal organisation and state structures (Pirie 2005). However, the dichotomies of state and society, domination and resistance or the idea of an ‘alternative legal modernity’ cannot explain why the Tibetan pastoralists now rely upon government officials to determine boundaries and prevent conflict, while regarding them as without authority to resolve their feuds. The Tibetan pastoralists are not asserting an alternative modernity or uniformly pursuing their political interests, so much as adapting their social norms and expectations of authority to a new set of power relations and administrative structures. They are doing so within a legal landscape that has always been plural and uncertain on the Tibetan plateau.

2001 ‘Property rights and the environment in pastoral China: evidence from the field.’ Development and Change 32: 717-740.

2001 ‘Community-based grassland management in Western China: rationale, pilot [page 97] project experience and policy implications.’ Mountain Research and Development 23: 1321-1340.

1961 Nomads of South Persia: the Basseri Tribe of the Khamseh Confederacy. London: Allen and Unwin.

2002 ‘Who’s afraid of legal pluralism?’ Journal of Legal Pluralism 47: 37-82.

1981 ‘Forum shopping and shopping forums: dispute processing in a Minangkabau village.’ Journal of Legal Pluralism 19: 117-159.

1959 Land and Polity in Tibet. Seattle and London: University of Washington Press.

1989 ‘Aspects of social organisation of Tibetan pastoral communities.’ Pp. 393-409 in Shoren Ihara, and Zuiho Yamaguchi (eds.), Tibetan Studies, Proceedings of the 5th Seminar of the International Association for Tibetan Studies. Naritan Shinshoji.

1939 Cultural Relations on the Kansu-Tibetan Border. Chicago: University Press.

1954 ‘Mi sTong: the Tibetan custom of life indemnity.’ Sociologus Ns. 4: 2, 136.

1964 ‘Peace and war among the Tibetan nomads.’ American Anthropologist 66: 1119-1148.

1978 Late Ching, 1800-1911, Part II. (Cambridge History of China, Vol 10). Cambridge: University Press.

1994 ‘Legal anthropology, legal pluralism and legal thought.’ Anthropology Today 10: 9-12.

1994 ‘Change, conflict and continuity among a community of nomadic pastoralists.’ Pp. 76-111 in Robert Barnett, and Skirin Akiner (eds.), Resistence and Reform in Tibet. London: Hurst & Co.

1996 Nomads of Golok: a Report. Available online: www.case.edu/affil/tibet/tibetanNomads/golok.htm

2002 ‘Legal pluralism.’ Pp. 289-310 in Reza Banakar and Max Travers (eds.), An [page 98] Introduction to Law and Social Theory. Oxford: Hart

1990 Tribes and State Formation in the Middle East. Berkeley and Los Angeles: University of California Press.

1995 ‘From nomads to ranchers: managing pasture among ethnic Tibetans in Sichuan.’ Pp. 69-76 in Graham E. Clarke (ed.), Development, Society and Environment in Tibet. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

n.d. Reconstructing Tradition: Persistence and Change in Golok Social Structure. Available online: www.cwru.edu/affil/tibet/booksAndPapers/Segpaper.html.

2000 ‘Tough times for Tibetan nomads in Western China: snowstorms, settling down, fences and the demise of traditional nomadic pastoralism.’ Nomadic Peoples 4: 83-109.

1912 In Forbidden China: the D’Ollone Mission 1906-1909. B. Miall (trans.). London: T. Fisher Unwin.

1950 China and Tibet in the Early 18^th Century: History of the Establishment of the Chinese Protectorate in Tibet. Leiden: E.J. Brill.

2002 The Ching Dynasty to 1800, Part I. (Cambridge History of China, vol 9). Cambridge: University Press.

2005 ‘Segmentation within the state: the reconfiguration of Tibetan tribes in China’s reform period.’ Nomadic Peoples 9: 83-102.

n.d.a ‘Order, individualism and responsibility: contrasting dynamics on the Tibetan plateau.’ In Keebet von Benda-Beckmann, and Fernanda Pirie (eds.), Order and Disorder: Anthropological Interpretations. Oxford: Berghahn (forthcoming).

n.d.b ‘Kings, monks, bureaucrats and the police: Tibetan responses to law and authority.’ In Franz von Benda-Beckmann, Keebet von Benda-Beckmann, and Anne. Griffiths (eds.), Law, Power, and Control. Oxford: Berghahn (forthcoming).

1998 ‘Against legal pluralism. Some reflections on the contemporary enlargement [page 99] of the legal domain.’ Journal of Legal Pluralism 42: 95-106.

1956 The Amnye Machen Range and Adjacent Regions: a Monographic Study. Roma: Is.M.E.O.

2006 ‘The heterogeneous state and legal pluralism in Mozambique.’ Law and Society Review 40: 39-75.

2000 Changing Inner Mongolia: Pastoral Mongolian Society and the Chinese State. Oxford: University Press.

1993 ‘The folly of the ‘social scientific’ concept of legal pluralism.’ Journal of Law and Society 20: 192-217.

2000 ‘A non-essentialist version of legal pluralism.’ Journal of Law and Society 27: 296-321.

1997 Frontier Nomads of Iran: A Political and Social History of the Shahsevan. Cambridge: Cambridge University Press.

1999 Nomads Killed in Pasture Fights. 21st June. Available online: www.tibetinfo.org.

2004 New Details on the Shooting of a Monk in Golog. 18th October. Available online: www.tibetinfo.org.

1998 ‘Ideological combat and social observation: recent debate about legal pluralism.’ Journal of Legal Pluralism, special issue, 42: 21-59.

1959 ‘A standard system of Tibetan transcription.’ Harvard Journal of Asiatic Studies 22: 261-276.

2003 ‘Tibetan range wars: spatial politics and authority on the grasslands of Amdo.’ Development and Change 34: 499-523.

^{^[1]}The literature on legal pluralism also contains extensive debates about the use of the terms ‘law’ and ‘legal’ to refer to non-state forms of normative ordering (Tamanaha 1993, 2000; Fuller 1994; Roberts 1998; Woodman 1998; F. von Benda-Beckmann 2002; Griffiths 2002). I do not engage with these debates here, using the term ‘legal’ in the loosest sense, to refer both to forms of administrative control and processes of conflict resolution.

^{^[2]}Nor is it the only way to develop the concept of legal pluralism, as F. and K. von Benda-Beckmann point out in this volume.

^{^[3]}This article is based on fieldwork carried out in Amdo between 2003 and 2005, which was funded by the Max Planck Institute for Social Anthropology. I am very grateful to Franz and Keebet von Benda-Beckmann for their encouragement and advice with this project and for their comments on this article, along with those very helpfully provided by Gordon Woodman.

^{^[4]}I use the term ‘tribe’, following Khoury and Kostiner (1990); Tapper (1997), to refer to these pastoralists’ groups in a sense which distinguishes their form of organisation from that of a state (see Pirie 2005: 84). Several different terms are used by the Tibetans, the most common of which is dewa (sde ba), a general term meaning ‘section’ or ‘group’, which is the one I adopt here. I transliterate Tibetan terms according to local pronunciation and add the Wylie (1959) transliteration in brackets.

^{^[5]}The historical details are largely drawn from Petech (1950: 85), Fairbank (1978: 36-37, 94) and Peterson (2002: 226-227).

^{^[6]}I use this term, following Barth (1961) and Tapper (1997) to refer to alliances of tribes characterised by strong leadership, often developing administrative organisation and which are generally shaped by political, economic and cultural relations with neighbouring groups and states. See Pirie 2005: 95-96.

^{^[7]}Goldstein (1996) describes the ecological problems in fact caused or likely to result from restrictions on movement and the less flexible allocation of pasture that this allows (and see also: Levine 1995; Miller 2000; Banks 2001; Banks et al. 2001; and cf. Sneath 2000: 130-131 on Inner Mongolia).

^{^[8]}These norms, as I have described elsewhere, involve expectations of aggressive individual (male) behaviour and the countervailing pressure to show restraint and compromise conflict in the interests of the wider community (Pirie n.d.a).