Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Anthropology. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 01 December 2020


  • Philip Carl SalzmanPhilip Carl SalzmanMcGill University


A tribe is a regional security organization. It ties together a number of local primary face-to-face groups. It is charged with control of territory, defense against outside intruders, and protection of humans, livestock, and productive resources, such as wells and cultivation. Whatever productive activity tribesmen are primarily engaged in, such as pastoralism or cultivation, each male, with the exception of holy men, serves also as a warrior.

Tribes are usually defined by a symbolic idiom that asserts a primordial connection among tribesmen. Descent from a common ancestor is an idiom used to define many tribes. Tribal names are often those of the ancestor that all members share. Internal divisions may also be defined by ancestry; a descent idiom allows group divisions at every level of the genealogy.

Tribal subgroups are also charged with security and are defined as having “collective responsibility”; that is, the moral norm is that each member is responsible for what other members do and, as a consequence, all members are seen by outsiders as equivalent. There is also a moral norm to aid fellow tribesmen, the obligation stronger for close kin, weaker for more distant kin. Internal tribal relations among subgroups are based on what anthropologists call “balanced opposition” or “complementary opposition.” Each tribal subgroup is “balanced” against other subgroups of the same genealogical order, which in principle, and often in practice, serves as a deterrent against hostile acts.

Tribal leadership can take the form of primus inter pares. However, in tribes in contact with states, more formal leadership roles, with at least the trappings of authority and power, can develop. Whatever the role of the tribal leader, he depends upon consent of the tribesmen. In tribal subgroups, political process tends to be highly democratic, and leaders are those who can elicit agreement among the members and then carry out the will of the community.

Tribes are social organizations that are not static and do not always maintain form. They respond to environmental opportunities and constraints. If a state nearby is in trouble, with failing leadership and an unruly population, a tribe may mount a campaign to invade and conquer the state, setting itself up as a ruling dynasty. In these cases, tribes lose their tribal characteristics and become a ruling elite. However, if a nearby state gains strength and expands its territorial control, it may overrun and defeat the tribe, encapsulating it, incorporating it, and even assimilating it.

Like all important terms of sociopolitical classification, such as “family,” “race,” “nation,” and “democracy,” the term “tribe” is multivocal; that is, different people use the term to mean different things, as is spelled out in detail in the Oxford English Dictionary.

The origin of the term is the Latin of ancient Rome, where the three real or purported descent groups, perhaps the descendants of the Latins, Sabines, and Etruscans, were labeled “tribes,” “tri” meaning three, and bhu or bu meaning “there are.” Thus, the first OED meaning is “A group of persons forming a community and claiming descent from a common ancestor.” Another example is the twelve tribes of ancient Israel. Later in Roman history, tribe was the label used for the thirty political divisions of the Roman people.

Some contemporary Jews refer to the Jewish people as “the tribe,” which fits with the broader meaning in the OED, “A particular race of recognized ancestry; a family.” A narrower definition was used in Ireland where it applied to “the families or communities of persons having the same surname.”

Relevant to international anthropology studies is the application of tribe to non-Western peoples in the OED: “A race of people; now applied esp. to a primary aggregate of people in a primitive or barbarous condition, under a headman or chief.” Large West African ethnic groups, some highly developed kingdoms, were referred to by some observers as tribes.

There is no “true” or “correct” meaning to any term, including “tribe”; meaning is arbitrary and in natural languages assigned by common usage of a term. Specialist technical languages are designed to do a particular job. Anthropologists have tried to develop terminologies to allow a better understanding of social and political life. In this article, a “tribe” is defined as a regional defense organization because this meaning aids in understanding what tribes do, how they operate, and why they are important.

A tribe is a regional security organization. It ties together a number of local primary face-to-face groups into a larger political entity. Local groups may be villages, herding camps, or even urban quarters. Tribal organization orders both internal social control and external defense and aggression. The symbolic idiom in which tribal affiliation is expressed may be descent, territorial residence, or cultural commonality.

Not all stateless societies are tribes. Anthropologists often refer to peoples whose largest political unit is a face-to-face group as a “band” (Service 1962; Steward 1963). This is a common form of organization among peoples depending upon nomadic hunting for their subsistence (Lee and DeVore 1976; Turnbull 1965). Some horticultural villagers live in communities not part of large political entities. One example is the Kuikuru of the Amazon Basin (Carneiro 1961). Another is the Plateau Tonga people of Zambia, whose villages were independent from one another (Colson 1951, 95). Even in the fact of relentless slave and cattle raiding, the villagers did not attempt to unite to defend their country (Colson 1951, 100).

In contrast to independent bands and villages, tribes are political units that encompass many local groups, whether villages, herding camps, or urban quarters. The subgroups are tied together by means of a symbolic idiom, such as common descent (Lancaster 1997; Lewis 1961), a shared age grade system (Spencer 1965), or acknowledgment of a leader (Barth 1964). Members of tribes have obligations to the tribe at large, as well as to their constituent groups, often of the same nature, such as defense. The boundaries of tribes are generally clear, defined by descent, territory, or both.

Tribes are participatory democracies. All men and all families are jurally equal. Everyone has a right to speak and express an opinion about any decision that is made. Discussions are often lengthy, and the ideal is a consensus on the decision to be taken. Mature males and female elders represent their families in public discussion.

In regard to almost all actions, tribespeople rely on collective responsibility and self-help. Collective responsibility means that each member of a group is responsible for whatever any other member does, and all members of a group are responsible for what any member does. The tribal motto could be “all for one, one for all.” Self-help means that whatever needs doing, they do it themselves.

There is little in the way of specialized institutions in tribal organization. That is why anthropologists call tribes “segmentary” because each subunit takes the same form and function as every other subunit, and larger units encompassing more people have the same form and function as smaller units. Tribal families carry the responsibility of educating the young, producing and distributing goods and services, supporting the ill and impoverished, and engaging in military defense and vengeance, and in aggressive predation. The two sectors in which there is specialization are religion, with small numbers of religious specialists, and external relations in which a few tribal officials—chiefs or sheikhs—serve as intermediaries between the tribe and external entities.

The phrase “self-help” is used most commonly in the anthropological literature in reference to physical conflict. Groups charged with self-defense will respond to threats or, in case of injury or loss, will engage in retaliation on their own. There are no police or similar agencies to help them. Thus, they rely on self-help, which means acting in their own interests as they see them. This is common in societies with balanced or complementary opposition in which small groups are balanced against other small groups and large groups are balanced against other large groups.

In genealogically based tribes, descent groups, lineages, vary in size according to the distance of the apical ancestor: the more distant the ancestor, the larger the group; the closer the ancestor, the smaller the group. In case of a conflict, adversaries distant from one another activate large groups while adversaries close to one another activate small groups. Each tribesman in a descent-based system is conceptually a member of a nesting set of lineages, from the nuclear family up through small face-to-face lineages, to medium-sized lineages, to tribal sections, and the tribe itself. Which level lineage is activated depends on who the opposition is.

Thus, all tribesmen are warriors, whatever other occupations they have, with the exception of religious figures, who commonly do not fight. The status of men, then, depends on their willingness to fight to support their fellow group members. Group solidarity, especially in conflicts, is highly valued. Group solidarity and willingness to fight acts as a deterrent to other groups which may have aggressive designs. At the same time, collective responsibility is a constraint on individual action, as all members of the group are implicated in acts by individuals. There is pressure for members to respect his group and to act within tribal norms, and there is, at least implicitly, the threat of being expelled from the group and its protection, which would leave the expelled individual a free target.

Independent tribes, that is, tribes sovereign over their members and territory, have a distinct organizational pattern: They are egalitarian, all adult males being jurally equal. They are decentralized, with most decisions being made by local groups in pursuing their own interests. There is a wide latitude of freedom, or agency, for individual actions.

This article focuses on independent and quasi-independent tribes. As state-integrated or assimilated tribes, ex-tribes, post-tribes, tribal ethnicities, and ex-tribals do not function as tribes, they are not considered here.

Tribal social and political organization contrasts with and is opposite to state organization, particularly preindustrial state organization. States are hierarchical, centralized, have specialized institutions, such as armies, tax collectors, and scribes, and tend to be despotic. Whenever possible, states “pacify” tribes, integrating and assimilating them. Studies of tribes under state control are properly studies of state polities as they integrate diverse populations.

Tribes have leaders but not rulers. Leaders in local and more encompassing groups have as their primary function to elicit a consensus from group members about any matter that must be decided. The basis of public decisions is public opinion. Tribesmen must agree with policies and decisions or else their control of the means of coercion (i.e., fighting skill and weaponry) and their ease of mobility (particularly in nomadic societies) make dissent relatively easy, effective, and ultimately decisive. It is difficult and unwise to bully mobile warriors.

Should a tribe lose full sovereignty to another political entity, such as a state, and become encapsulated to a degree by the state, the political situation changes. Leaders become intermediaries between agents of the state and the tribesmen: the state wants the leader to impose policies favored by the state and holds him responsible for tribal actions; the tribesmen want the leader to protect them from the agents of the state and to extract from the state resources desired by them. In encapsulated tribes, the leader faces the foreman’s dilemma, being pulled in contradictory directions, but also is able to gain power and resources and thus bolster his status.

State-encapsulated tribes maintain their structures and continue to operate through them, but their ability to use coercive power is greatly restricted. In particular, predatory raiding outside of the tribe is suppressed. In full encapsulation, weapons, such as firearms, may be confiscated.

State-incorporated tribes lose control of their territory and leadership. They thus become dependent on access to land controlled by others and upon agents of the state. At this point, the ex-tribesmen can be regarded as peasants, but in some cases, tribal identity persists, and ex-tribesmen can be mobilized within state contexts, for example, in voting.

State-assimilated tribes lose many aspects of their internal structure and much of their culture. The ex-tribesmen adopt the culture of the state and the dominant ethnic group. This is reflected in language, clothing, food, dwellings, and other material culture. Assimilated individuals with tribal backgrounds may lose their tribal identities.

Until recent times, the states with which tribes have had to contend were preindustrial states, which did not so much provide services to its subjects as draw its resources from its subjects. Tribal values of jural equality, individual autonomy and agency, decentralized decision-making, and control of resources are always compromised by the tribe falling under the control of a state. Generally, tribes followed various strategies to remain independent of states, which included armed defense, mobile escape, and alliance with other tribes or opponent states.

Segmentary Societies

Segmentary societies played a large part in Emile Durkheim’s vision of social evolution: segmentary societies based on mechanical solidarity (likeness) came to be replaced by complex societies based on organic solidarity (interdependence). Segmentary societies themselves were an evolution from hordes (bands, in current anthropological parlance):

We give the name clan to the horde which has ceased to be independent by becoming an element in a more extensive group, and that of segmental societies with a clan-base to peoples who are constituted through an association of clans. We say of these societies they that are segmental in order to indicate their formation by the repetition of like aggregates in them. . ., and we say of this elementary aggregate that it is a clan, because this word well expressed its mixed nature, at once familial and political.

(Durkheim 1933, 175)

Attention to segmentary societies was given considerable emphasis by British social anthropologists in the mid 20th century. Two works describing segmentary societies which became influential in political anthropology were African Political Systems (Fortes and Evans-Pritchard 1940) and Tribes Without Rulers (Middleton and Tait 1958).

But perhaps no author had more impact on thinking about segmentary societies than E. E. Evans-Pritchard (1940). In his ethnography of a southern Sudanese people, The Nuer, he described a model of social opposition and alliance that he called “balanced opposition” and others have called “complementary opposition.” Evans-Pritchard applied this model to the Bedouin of Cyrenaica in his brilliant, macro-historical work, The Sanusi of Cyrenaica:

Qabila is the word generally used to denote a tribe or primary tribal division. ‘Ailat are the lineages into which a clan is divided and hence the sections of a tribe of various sizes in which these lineages are found and after which they take their names. Biyut are small lineages, or extended families, with a depth of five or six generations from the present day to their founders.

Though sections of a tribe may be opposed to one another they regard themselves as an undivided group in opposition to neighbouring tribes, and are so regarded by their neighbours.

Each tribe in this commonly accepted sense has its watan, its homeland, its soil, its arable, its pastures, and its wells; and each has its camel-brand which is also carved on the tombstones of its dead. The tribal lands are vested in the tribe, which has residual rights to them. . . .A tribe is conceived of as a huge family descended from a common ancestor, from which the tribe generally takes its name.

(Evans-Pritchard 1949, 54–57)

Segmentary societies can manage large-scale conflict, spatial displacement, and loss of sections because each segment is self-supporting and self-sufficient. There are no specialized institutions without which the society cannot operate. In this sense, segmentary societies are well adapted to conflict and change. An illustration is the lengthy guerilla war carried out by the Bedouin of Cyrenaica against the far better-armed Italians (Evans-Pritchard 1949). The Italians did not get the upper hand until they ruthlessly decided to exterminate all of the Bedouin and their herds.

Local Organization in Tribal Societies

Tribal organization, based on genealogy, territory, and leadership, is relatively stable, consistent, and robust. Tribal organization is a kind of constitution, a cultural umbrella under which tribesmen can live protected. In contrast, local organization must be flexible in order to deal with day-to-day problems and with seasonal and annual variations. Thus, local organization must be able to work within the framework of the constitutional tribal organization but be able to respond to changes in circumstances.

Among the Yarahmadzai Baluch, local groups, herding camps, halk, are constituted by means of a contract, karar, between herd owners, wasildar, and a shepherd, schwaneg, for one year of herding (Salzman 2000a). In other words, the herding camp is reconstituted every year. This allows the increase or decrease in the number of herd owners and the number of livestock, depending upon how good or poor the pasture is and will be. The herding group is in this way adjusted to cope with the limitations or opportunities of changing environmental conditions.

As well, the annual reconstitution of the local group allows the rejection of members from the previous year if it is felt that they did not do their share of the collective labor, were uncooperative, or were undesirable in some way. Most local groups include members from a range of lineages, but it is common that there is a core from one lineage, tribesmen who make up a plurality of the camp, and whose longevity in the camp gives them particular weight.

In case of conflicts, the lineages that make up the tribal system are activated, with lineage mates from different local groups coming together, jam kon, to present a united front and to act collectively.

Local groups in tribes can be stationary or nomadic, depending upon the subsistence pattern. Among the Nuer of the southern Sudan, during the wet season they live on mounds and engage in horticultural cultivation as well as herd their livestock (Evans-Pritchard 1940). During the dry season, the Nuer migrate to the riverbanks in order to fish and sustain their cattle. This is a reliable transhumance, and village locations tend to be stable. But individuals and small groups do not always stay in the same villages. Individuals are free to move out to another village or to start their own, with or without other members of the village. The result is that many Nuer individuals live in villages in which members of other lineages are dominant. The Nuer tribal system continues to serve as a security framework for these individuals even as their residence shifts. If need be, they can call upon their lineage mates for assistance.

Not all pastoral peoples keep their herds with them, as do the Nuer. In East Africa, for example, among the Maasai (Spear and Waller 1993) and Samburu (Spencer 1965, 7), it is common for the main cattle herd, other than some milch cows, to be far from the stable settlement where the elders, women, and children reside. The herd is kept in far-ranging cattle camps, watched over by young male “warriors” who live in the camps. So too among the Karimojong, where stable settlements are found in the center of the territory and with cattle camps scattered throughout the territory and beyond (Dyson-Hudson 1966, chap. 2).

In East Africa, a number of tribes have members who are stationary agriculturalists and others who are nomadic pastoralists. Edgerton 1971) studied four of them to compare the lives of the villagers with the nomads. The Hehe farming settlement is described:

The farming neighborhood of Ngelewala was quiet. It seemed even to lack the expected sounds of children and birds. . . .The lack of bustle and noise came in part from the great dispersion of homesteads. No matter where one stood, the houses of these Hehe farmers remained hidden from view by their wide separation, their distance from the walking paths, and by the fringe of trees and crops that surrounded them. Much was hidden within these houses too, for these massive, thick-walled enclosures seemed to stand as fortresses separating their occupants from the world outside.

(Edgerton 1971, 93)

Hehe pastoralists are also described:

Pawaga was oppressively hot, flat, and open. . . .Unlike Ngelewala, one could see through the thornbush for great distances, and along the paths, both people and livestock seemed to exist in droves. . . .The houses were packed together into nucleated settlements. But the houses were more open, with their thin walls, partial enclosure, and flimsy construction. Around these villages there were people, cattle, and much shouting and laughter.

(Edgerton 1971, 96)

With these differences in local settings, the people, although all Hehe, were different too. Here are the Hehe farmers:

Even the most superficial observation of Hehe interaction was sufficient to call into question the sincerity of these polite, differential encounters. The Hehe themselves readily admitted that they were insincere, and said that these expressions of concern and caution in their everyday relations were essential to avoid serious afront or insult. The Hehe were astonishingly sensitive to insult; they lived in continual dread of affront, either given or received. It was for this reason, they said, that men feared each other and took such care not to give offense by the least breach of decorum. . . .When the Hehe became angry, they did so impulsively, without apparent control over the resulting verbal abuse or physical assaults.

(Edgerton 1971, 94)

Here are the Hehe pastoralists:

It was obvious that the veil of constraint and secrecy had been lifted. People were clearly more open, more spontaneous. They laughed and they quarreled openly. Everything they did was more animated. . . .There was drumming, dancing, and singing. . . .In Pawaga, men displayed a constant readiness for combat. They lived in a dangerous world of large animals and human enemies. Lions, elephants, and hyenas were a continual threat to livestock, children, and adults. Hostile tribesmen, Gogo or Baraguyu, posed a violent threat as cattle thieves. In Pawaga, Hehe men displayed a constant readiness to fight—against animals, against men from other tribes, or against one another. . . .And if they fought [among themselves], as they frequently did, there were no signs of lingering acrimony. . . .Here men fought casually, without hesitation, but with control. They fought fairly and sub-lethally.

(Edgerton 1971, 96)

Edgerton suggests that these different local settings and the constraints that they offer influence the differential attitudes, values, and behavior of their inhabitants:

Where the farmers are constrained in their emotions and indirect in their actions, the pastoralists freely express emotions, both positive emotions such as affection, sexuality, or bravery, and dysphoric emotions such as guilt, depression, brutality, and fear of death. Where the farmers are indirect in their actions, especially their actions in relation to conflict, the pastoralists act independently and aggress openly. . . .Pastoralists respect authority, particularly of senior persons and of prophets, and they are self-controlled where the farmers are impulsive. . . .The pastoralists are concerned with the consequences of wrongdoing, and they value cooperation, industriousness, the clan and other kinsmen. There is no comparable, socially cohesive focus among the farmers.

(Edgerton 1971, chap. 11)

Edgerton finds the members of each of the four tribes different from the members of the other three. He also finds diversity between the agriculturalists and the pastoralists within each tribe. Nevertheless, across the four tribes there are certain commonalities distinguishing the settled agriculturalists from the pastoralists. Local communities in local settings make a difference in tribal life.

Tribal Leadership

Tribesmen will follow their leaders as long as their leaders are taking them where they want to go. Tribal leaders gain their support not from coercion, patronage, or status, but from their tribesmen, from public opinion. Given that tribesmen are all warriors, all armed, trained, and experienced in combat, in many cases mounted, and in all pastoral tribal societies, mobile in their persons, family, household, and livestock, they are not willing to follow leaders against what they see as their interests. Consequently, tribal leaders are not rulers who could impose their wills on their tribes.

According to Evans-Pritchard (1949, 59–60), “In such segmentary systems there is no state and no government.” The social position of a shaikh “must in no sense be regarded as a ruler or administrator. . . .The precarious authority of the Shaikhs rests not on force but on the renown and esteem they enjoy in the tribe.” Similarly, according to Lancaster, among the Rwala Bedouin, “No individual has political power, no groups has political power and no family has political power; power is restricted to the workings of public opinion. Even public opinion has no formal coercive power; co-operation can be withdrawn and that is all” (Lancaster 1997, 77). Lancaster reports that “Emir Sattam ibn Hamad Sha’alan . . . is quoted as saying ‘I can only do what my people wish’” (Lancaster 1997, 96).

Tribal leaders often give hospitality to their tribesmen, and even gifts. The social process taking place is the leader tapping into public opinion, trying to discover what his tribesmen are thinking. This process is also manifested in judgments about conflicts among tribesmen. According to Beck,

A primary feature of the ilkani’s [confederacy chief] style of leadership, which reflected political processes among the Qashqa’i in general, was the large open gathering in the reception tent, where he presided but did not explicitly dominate the discussion. . . .Men talk . . . a consensus slowly emerges without any single individual’s having made the decision unilaterally. . . .The ilkani possessed information . . . that others present did not, but he tended to introduce it in general conversation to allow others to consider and discuss it. People presented their problems in this open forum, presided over by the ilkhani and his advisors, and decisions could be reached in the course of group discusion without the ilkhani’s having made any kind of a definitive statement.

(Beck 1986, 218)

If tribal leaders are not rulers with power and authority, what are they, and what distinguishes them? Among the Rwala,

The difference between sheikh and tribesman is not so much in reputation as in reputation range. The Emir and the sheikhs are mediators on behalf of the tribes with other tribes or with national governments. With other tribes they negotiate with the sheikhs not because they are hierarchical equals (they are equals anyway of “free Bedu”), but because their counterparts in other tribes have the same sort of reputation range as themselves. As for governments, the sheikhs negotiate with senior officials or the heads of government direct.

(Lancaster 1997, 81)

The role of the Sardar, chief, in each of the Sarhadi Baluchi tribes, is constrained by the fact that these are segmentary tribes in which political authority and military capacity rest with the segments, patrilineages of various sizes (Salzman 2000, chap. 11). Within the tribal system, the Sardar has no assets he can use to coerce the behavior of his tribesmen. He can offer no economic rewards, can threaten no punishment. The role of the Sardar within the tribal system is that of mediator between and among the lineages. The Sardar can use persuasion, referring to moral principles held by the tribesmen. For example, he would say that “we are one,” “we are all brothers,” “we are all of the same line,” as a way of convincing conflicting segments to compromise and come to a settlement of any dispute.

The Sardar could also perform this negotiating function externally, for example, in conflicts arising between tribes. Occasionally the Yarahmadzai would come into conflict with the neighboring Gamshadzai, and the conflict would either have to be resolved or risk escalation. The process of negotiation was facilitated by the marriages and kin ties among chiefly families in different tribes.

The Sardar was also the war leader, leading the tribe in predatory raiding and in outright military conflicts. Jiand Han, the Yarahmadzai chief for the first decades of the 20th century, led annual raiding parties and also was the general in confrontation with other tribes.

Some accounts of tribal chiefs have described them as all-powerful. Fredrik Barth describes the Basseri Khan as “the central, autocratic leader of the tribe,” who had “great power and privilege.” This “power is conceived as emanating from him, rather than delegated to him by his subjects” (Barth 1964, chap. 5). Because the Basseri, consisting of sections of different origins, was not seen as a kin group, the Khan played the role of the central symbol of the tribe. And tribesmen would say that the Khan was their all-powerful leader.

The reality, however, was somewhat different. Drawing on evidence from Barth’s book, it is clear the Basseri Khan, far from being omnipotent, was heavily dependent upon the consent of his tribesmen (Salzman 2000b). First, the Khan has no independent sources of funding to buy support. Second, the Khan has no enforcement organization, police, or army to coerce his tribesmen. Third, the tribesmen and their capital assets, livestock, were mobile and could easily escape if they so desired. Barth records the fact that various Basseri sections had left and gone elsewhere. Fourth, skilled in their favorite pastimes, riding horses and hunting, the tribesmen were an irregular cavalry. In other words, the means of coercion—abilities and equipment—were in the hands of the tribesmen. Barth mentions that tribal Khans of the region have been assassinated.

The Basseri Khan was very skilled at tapping public opinion in his tribe. His hospitality brought tribesmen into his tent as guests, and guests were always obliged to share information with their hosts. The Khan knew what his tribesmen liked and did not like, what they wanted and did not want, and how they felt about various developments. The Khan was thus in a position to lead his tribesmen where they wanted to go and thus to reinforce his leadership position.

The image of the Basseri Khan as all powerful was a public relations ploy to portray the Basseri as cohesive and solidary behind their powerful leader and thus to give weight to any Basseri demands of other tribes or Persian officials, as well as to give pause to any thoughts by other tribes or Persian officials of adventures against the Basseri.

Balanced or Complementary Opposition

Evans-Pritchard’s model of balanced opposition posited that tribesmen would ally with closer kin against more distant kin. So while members of two biyut might be in conflict, whether over livestock, pasture, injury, or honor, any conflict with a member of another ‘ailat would bring those two biyut as members of their common ‘ailat in order to oppose the other ‘ailat. So too would members of different ‘ailat unite under their common qabila in opposition to a different qabila, and so on, through tribes and tribal confederacies and, in the Middle East and North Africa, up to Muslims versus infidels.

This model, as is often the case with anthropological fads and fashions, was quickly adopted and applied widely and then in the next generation widely rejected. The most general argument marshalled against it was that it was not a behavioral model that could predict every aspect of conflict among tribal peoples. For example, it was said, as a refutation, that close kin groups tend to live close to one another and conflict tends to arise among them rather than more distant kin, although in fact conflict between close kin groups is entirely consistent with the model. It was said that when members of two larger and more distant groups come into conflict, all of the members of each group do not mobilize in support. It was also argued that some tribal groups allied with outsiders to increase their advantage (Peters 1967). These two latter observations are correct and would be refutations of the complementary opposition model if it were a behavioral model.

The tribal complementary oppositions model, however, is not a behavioral model, but a moral model. It is a statement of what is ethical and thus what should be done. The model asserts that tribesmen should support closer kin against more distant kin or outsiders. This is the right thing to do, according to the model. It could be likened to the Ten Commandments, urging us to honor our parents, not to take God’s name in vain and not to murder or commit adultery or to covet what belongs to our neighbors. The Ten Commandments is not refuted if someone takes the Lord’s name in vain or covets his neighbor’s wife. The Ten Commandments are understood as moral guidance, and examples of the violation of its prescriptions and proscriptions are often seen to reinforce the validity and necessity of honoring the Ten Commandments. That is why violating the Ten Commandments would bring feelings of guilt to many violators.

So too with both collective responsibility and complementary opposition, which are best understood as models of moral action, the violation of which can result in a sense of disorder and guilt. These models place moral pressure on individuals and groups to conform and are the source of guilt about inappropriate behavior for those who deviate from the prescribed course.

For tribesmen, following the guidelines of collective responsibility and complementary opposition is the “natural” thing to do and the “normal” course of events. Because individual or group interests and attractive opportunities may lead tribesmen to violate the norms, the complementary opposition is not predicative in all cases. But in many cases, probably the majority of cases, tribesmen do follow the guidelines or are prepared to follow them if called upon.

Generally in tribal society, cohesion and solidarity of the kin group is highly cathected. Among the Baluch, the term of kin solidarity is topak; a kin group that was cohesive was called patopak, while one lacking solidarity was called beatopak (Salzman 2000a, 238). The prestige of a kin group and presumably its deterrence rest partly on the degree of its solidarity. A patopak kin group would be expected to quickly and decisively come to the aid of one of its members, primarily in a violent conflict but also in economic extremity. Kin groups will sometimes contribute livestock to a member who has lost a herd or contribute money to a member who has fallen into smothering debt.

To illustrate, here is a case of complementary opposition that I observed in 1973 in Iranian Baluchistan (Salzman 2000a, chap. 10). It took place at the Hamuni Mashkel, the large drainage basin on the Iran-Pakistan border. Mashkel was the site of date groves belonging to several tribes, including the Yarahmadzai (descendants of Yarahmad) with whom I was living. My home was with the Dadolzai lineage, a minimal political segment called a brasrend (a line of brothers); every lineage of whatever size, from small families to the tribe as a whole, was called a rend (line).

The Dadolzai occupied a settlement in Gorani, along with other brasrends that had their own settlements. A conflict arose between the Dadolzai and the Kamilhanzai, a cousin lineage to the Dadolzai, when date palm trunks that had been cut to serve as rafters for a mud brick hut had disappeared from where they were left in the grove. The Dadolzai who had cut the palms, and who was furious, was sure that a member of the neighboring Kamilhanzai had taken them. He approached the Dadolzai mastair, headman, to authorize a war party. Permission was given, but only to retrieve the trunks, avoiding direct conflict with the Kamilhanzai, if at all possible. A war party of eleven men was formed and went in search of the trunks; the trunks were located and retrieved, with no Kamilhanzai resistance, but the attempted theft remained a potential source of conflict between the lineages. Some negotiation, and perhaps compensation, was called for.

While the Dadolzai and Kamilhanzai were thinking about this, news came of another event: an elder and a young man of the Soherabzai tribal section, to which both the Dadolzai and Kamilhanzai belong, was beaten by a group of tribesmen from the Rahmatzai tribal section. This beating was retaliation for the beating of a Rahmatzai by a Soherabzai in a dispute over a camel that ate dates from the Soherabzai’s palms. The beating of the Soherabzai elder was regarded as a disrespectful and inappropriate response to the earlier incident regarding the date-eating camel; the reaction was anger among the Soherabzai tribesmen.

The Soherabzai mobilized. The Dadolzai and Kamilhanzai set aside their minor dispute, now thinking of themselves as Soherabzai, and came together to discuss their strategy for revenge against the Rahmatzai. The Rahmatzai in-law living with the Dadolzai skedaddled back to his lineage mates lest he become a target of other Soherabzai. Hearing that the Rahmatzai were massing to engage with the Soherabzai, the Dadolzai, Kamilhanzai, and other Soherabzai formed a war party of over one hundred men of all ages, most armed with primitive weapons (stocks, stones, brass knuckles, knives, swords), and headed, accompanied by the intrepid ethnographer, to the location of the putative engagement. They were stood up; the Rahmatzai did not show, but the Soherabzai could take satisfaction from the fact that they had successfully mobilized and were ready to engage in combat. The conflict between the Soherabzai and Rahmatzai continued, on and off, for five years, with many instances of confrontation and conflict.

Even this “normal” complementary opposition had “flaws,” if one views the model as a mechanical, behavioral one. In the Dadolzai war party there were two members of another tribe, the Gamshadzai, who resided at Gorani with the Dadolzai. Further, not every member of the Dadolzai participated; only those who were close neighbors took part in the war party. Similarly, with the Soherabzai war party, not all of the many hundreds of Soherabzai participated; only those who were nearby participated. However, if we accept complementary opposition as a model of morality, the actions of the tribesmen can be seen to be in the correct spirit.

Let us not lose sight of the fundamental importance of complementary opposition in the lives of tribesmen. Collective responsibility means that no tribesman is “fair game” to aggressors, but each tribesman is backed by the other members of his kin group. Similarly, no small group will be faced with a hostile large group because each group at every level is balanced by an opposing group of the same order. Collective responsibility and complementary opposition thus provide deterrence against aggressive adventurism, for any perpetrating aggressive act can expect retaliation from its complementary kin group. Where there is peace in tribal society, it is due to people following this moral model and the deterrence it provides.

Tribes and States

Tribes with no states in their environments, or remote from government centers of power and from government agencies, or engaged only with weak states are best able to be “tribes without rulers” and to maintain an egalitarian political system with great agency for individuals and kin groups. But many tribes resided in the same region as states or empires, and many had relations with states, although these may vary over time. When tribes and states were in contact, negotiations often took place. When states or empires were weak or reaching far beyond the areas of their control, they would sometimes pay tribute or bribes to the tribes; when states or empires were strong and could impose conditions, tribes sometimes paid tribute or suffered state imposition.

In addressing the relations between tribes and states, it is necessary to remember that tribes and states are based on different and largely contradictory principles: Tribes are decentralized and egalitarian, and based on collective responsibility. States are centralized and hierarchical, and rely on specialized administration and enforcement institution. In this sense, tribes and states are incompatible. Therefore, relations between tribes and states present serious challenges to all parties.

The relations between tribes and states can be set out as a series of relations on a continuum, from tribal political supremacy at one end to state supremacy at the other end. The continuum with descriptive labels may look something like this:


These are descriptive states ranging from tribal supremacy on the left to tribal destruction on the right. This is not a time sequence and there is no suggestion of necessary movement from one state to another, nor that a shift from one relation to another in any way, annihilation excepted, precludes return to a different relation. The intent of this continuum is to point out the many different relations that may exist between a tribe and a state. The following is a discussion of the stages of this continuum.

Conquest: There is a long history of tribal conquests of states, occupation of their territory, and imposition of a new regime, a new dynasty as rulers of the conquered population. Ibn Khaldun’s model of tribe–state relations posits a cycle of state decay and tribal invasion and conquest (Khaldun 2015). The 7th-century Arab Islamic conquests from India to Morocco depended largely on Bedouin tribal armies (Karsh 2006; Lindholm 2002). The Muslim conquest of southern Iberia in the 8th century was the work of Bedouin and Berber tribal armies. Later invasions also relied on tribal warriors, such as the 11th-century invasions by the Banu Hilal and the Banu Sulaim (Evans-Prichard 1949). Iranian regimes were treated repeatedly to invasion and conquests by mountain and desert tribes (Britannica 1968).

Predation: Many tribes have raided on the peripheries of states, such as the nomadic tribes of the steppe to the north of China (Barfield 1989); the tribes on the periphery of the Roman Empire (Dijkstra and Fisher 2014); the Turkmen in the northeast of Persia (Irons 1975); and the Baluch in the southeast of Persia (Salzman 2000a). This provided wealth for the tribes while they maintained their independence and their fighting skills.

Alliance: Tribes within the near distance of states or empires would on occasion forge alliances with the state or empire, holding off on aggressive acts toward the state by themselves or other tribes. In return, financial benefits and sometimes honorific titles would flow to the tribe (Dijkstra 2014; Salzman 2014).

Clientage: With a balance of power favoring the state, tribes could be drawn into a patron–client relation, with the state dictating the terms of its expectations while withholding impositions and providing benefits to the tribe (Dijkstra and Fisher 2014).

Encapsulation: Tribes accept the suzerainty of the state, of which they are now a part. However, their tribal structures remain intact and internal political processes continue as if the tribe were independent. The tribal chief becomes a state functionary, providing the connection between the tribe and the state. Engagement of tribes with states required communication, mediation, and negotiation, and it fell to leaders to fill these requirements (Salzman 1973, 1974). Commonly, the leaders gained in stature and, to the degree that state resources flowed through them, power (Salzman 1974). The Sardar of the Yarahmadzai, under encapsulation by Iran, was the conduit for resources and benefits flowing from the state to the tribe, thus bolstering somewhat his power (Salzman 2000a). With a significantly crystallized leadership hierarchy, tribes without rulers become tribal chiefdoms. Beck argues, “the Qashqa’i and similar kinds of polities in Iran and elsewhere were in fact formed and sustained by broad external forces, and that the resulting [hierarchical] sociopolitical structures were ultimately ways of integrating the affiliated people into state structures” (Beck 1986, 349). In other words, chiefly hierarchies arose in tribes as a result of interactions between tribes and states.

Integration: Tribal structures are to a considerable extent made to conform with state structures. Tribal leadership is replaced by state leadership. For example, the Basseri Khan was replaced by an army colonel (Barth 1964) and Qashqai migrations were controlled by the Persian military (Beck 1986). With structural integration, the tribe as a political structure ceases to exist. The ex-tribesmen have become peasants, subject to the demands of the state apparatus. Pastoral tribes have become pastoral peasants (Salzman 2004, chap. 5). Lacking coercive power, many pastoral tribes no longer control territory and the pastures necessary for pastoral production. For example, the Yoruk of Turkey were decapitated, losing their political leader, and had to pay villagers for access to pasture (Bates 1973). The Komachi of Iran were integrated economically as well as politically, providing wool for the Kerman carpet trade, but had no political clout and depended upon wealthy patrons or government bureaucrats for access to pasture (Bradburd 1990). In other cases, land titles were granted to tribal notables, leaving ordinary tribesmen to serve as sharecroppers. Barbara Aswad describes this for the Hatay Plain under the Ottoman Empire:

When the long-range Turkmen Nomads of the Reyhanli Confederation were forced into permanent settlement, a governmental appointee was designated as the head of the confederacy and individual land titles were given to the notables of some forty clans which composed the confederacy. The notable, on is part, was ultimately subordinate to the head of the confederacy, the Pasha, who was appointed by the governor and was obligated to raise soldiers for wars and collect taxes.

(Aswad 1971, 23)

Some degree of tribal identity was maintained, but it was rather like ethnicity, a way of communicated commonality horizontally, among peers, and in seeking work (Aswad 1971, 24).

Assimilation: A post-tribal population can lose its cultural distinctiveness by adopting the language, dress, and customs of the larger state population. The Baluchi language of the Baluchi tribes has gradually migrated in vocabulary and pronunciation toward its cousin West Iranian language, Persian. This should not be a surprise, as men from the Baluchi tribes working in Iran or dealing with Persians have become bilingual. And while in their traditional territory, Baluchi tribesmen wear their customary wide, baggy trousers, long flowing shirt, and turban when they leave the territory, they wear more Persian or international clothes, even blue jeans. Among the Zagros tribes, such as the Basseri (Barth 1964), the dress of males is not highly distinctive in relation to Persian clothes, suits of Western-style tailored pants and jackets not being uncommon. Neither the Basseri nor the Baluch have totally assimilated to Persian culture, but there is a gradual shift in that direction. Other populations from tribes long gone have fully merged with the majority Persian population, their tribal ancestry little more than that.

Annihilation: Throughout history, in conflicts between tribes and states, tribes have sometimes been annihilated, their population killed or dispersed. We have recently seen attempts to annihilate tribes by the Islamic State (ISIS) (CNN 2014).


  • Aswad, Barbara C. 1971. Property Control and Social Strategies: Settlers on a Middle Eastern Plain Turkey. Ann Arbor: Museum of Anthropology, University of Michigan, Anthropology papers no. 44.
  • Barfield, Thomas J. 1989. The Perilous Frontier: Nomadic Empires and China, 221 BC to AD 1757. Cambridge, MA: Blackwell.
  • Barth, Fredrik. 1964. Nomads of South Persia. Oslo, Norway: Universitetsforlaget.
  • Bates, Daniel G. 1973. Nomads and Farmers: A Study of the Yoruk of Southeastern Turkey. Ann Arbor: Museum of Anthropology, University of Michigan, Anthropology papers no. 52.
  • Beck, Lois. 1986. The Qashqa’i of Iran. New Haven, CT: Yale University Press.
  • Bradburd, Daniel. 1990. Ambiguous Relations: Kin, Class, and Conflict among Komachi Pastoralists. Washington, DC: Smithsonian Institution Press.
  • Britannica. 1968. The Cambridge History of Iran in Eight Volumes. Cambridge, UK: Cambridge University Press, 1968.
  • Carneiro, Robert L. 1961. “Slash-and-Burn Cultivation among the Kuikuru and Its Implications for Cultural Development in the Amazon Basin.” Antropologica Supplement No. 2.
  • CNN. 2014. “ISIS Kills 300 Members of Sunni Tribe, Iraqi Government Says.” CNN, November 4.
  • Colson, Elizabeth. 1951. “The Plateau Tonga of Northern Rhodesia,” In Seven Tribes of British Central Africa, edited by Elizabeth Colson and Max Gluckman, 94–162. Manchester, UK: Manchester University Press.
  • Dijkstra, Jitse H. F. 2014. “‘I, Silko, Came to Talmis and Taphis’: Interactions between the Peoples beyond the Egyptian Frontier and Rome in Late Antiquity.” In Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, edited by Jitse H. F. Dijkstra and Greg Fisher, 299–330, Leuven, Belgium: Peeters.
  • Dijkstra, Jitse H. F., and Greg Fisher, eds. 2014. Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity. Leuven, Belgium: Peeters.
  • Durkheim, Emile. 1933. The Division of Labor in Society. Translated by George Simpson. Glencoe, IL: The Free Press.
  • Dyson-Hudson, Neville. 1966. Karimojong Politics. Oxford: Clarendon Press.
  • Edgerton, Robert B. 1971. The Individual in Cultural Adaptation: A Study of Four African Peoples. Berkeley, CA: University of California Press.
  • Evans-Pritchard, E. E. 1940. The Nuer. Oxford: Clarendon Press.
  • Evans-Pritchard, E. E. 1949. The Sanusi of Cyrenaica. Oxford: Clarendon Press.
  • Fortes, M., and E. E. Evans-Pritchard, eds. 1940. African Political Systems. London: Oxford University Press.
  • Ibn Khaldun. 2015. The Muqaddimah: An Introduction to History. Originally published in the 14th century; abridged edition, Princeton, NJ: Princeton University Press.
  • Irons, W. 1975. The Yomut Turkmen. Ann Arbor: Museum of Anthropology, University of Michigan, Anthropology papers no. 58.
  • Karsh, Efraim. 2006. Islamic Imperialism. New Haven, CT: Yale University Press, 2006.
  • Lancaster, W. (1981) 1997. The Rwala Bedouin Today, 2nd ed. Prospect Heights, IL: Waveland.
  • Lee, Richard B., and Irven DeVore, eds. 1976. Kalahari Hunter-Gatherers. Cambridge, MA: Harvard University Press.
  • Lewis, I. M. 1961. A Pastoral Democracy. London: Oxford University Press.
  • Lindholm, Charles. 2002. The Islamic Middle East, rev. ed. Malden, MA: Blackwell.
  • Middleton, John, and David Tait, eds. 1958. Tribes Without Rulers: Studies in African Segmentary Systems. London: Routledge & Kegan Paul.
  • Peters, E. L. 1967. “Some Structural Aspects of the Feud among the Camel- Herding Bedouin of Cyrenaica.” Africa 37: 261–282. Reprinted in Emrys L. Peters, 1990, The Bedouin of Cyrenaica, chap. 4. Cambridge, UK: Cambridge University Press.
  • Salzman, Philip Carl. 1973, “Continuity and Change in Baluchi Tribal Leadership.” International Journal of Middle East Studies 4: 428–439.
  • Salzman, Philip Carl. 1974. “Tribal Chiefs as Middlemen: The Politics of Encapsulation in the Middle East.” Anthropological Quarterly 47: 203–210.
  • Salzman, Philip Carl. 2000a. Black Tents of Baluchistan. Washington, DC: Smithsonian Institution Press.
  • Salzman, Philip Carl. 2000b. “Hierarchical Image and Reality: The Construction of a Tribal Chiefship.” Comparative Studies in Society and History 42: 49–66.
  • Salzman, Philip Carl. 2004. Pastoralists: Equality, Hierarchy, and the State. Boulder, CO: Westview.
  • Salzman, Philip Carl. 2014. “The Meeting of the Twain: Tribe and State.” In Inside and Out: Interactions between Rome and the Peoples on the Arabian and Egyptian Frontiers in Late Antiquity, edited by Jitse H. F. Dijkstra and Greg Fisher. Leuven, Belgium: Peeters.
  • Service, Elman R. 1962. Primitive Social Organization. New York: Random House.
  • Spear, Thomas, and Richard Waller, eds. 1993. Being Maasai. London: James Currey.
  • Spencer, Paul. 1965. The Samburu. London: Routledge & Kegan Paul.
  • Steward, Julian. 1963. Theory of Culture Change. Champaign: University of Illinois Press.
  • Turnbull, Colin M. 1965. Wayward Servants: The Two Worlds of the African Pygmies. London: Eyre & Spottiswood.