Usage of the term “tribe” in anthropological literature has been both dominating and problematic. 1

In Middle Eastern studies the term tribe is not used as a political evolutionary category (as in the band-tribe-chiefdom-state evolutionary narrative), nor is it regarded as a mere construction of the Western colonial administration (as it is the case for some groups in sub-Saharan Africa). Groups that could be described as “tribes” existed in the context of pre-colonial and pre-modern states (integrated in the state structure, or opposing it), and other tribal groupings.

Compared with the rural and urban sectors that comprise the predominant demographic majority, pastoral nomadic tribes have received a disproportionate attention in contemporary studies of Middle Eastern societies (Abu-Loghud 1990).

Anthropologists do have a tendency to study marginal and exotic populations, but, historically, pastoran nomadic tribal groups in many areas in the Middle East have played a political role that outweighs their demographic magnitude.

It is estimated that during the nineteenth century between one-fourth to one-third of the Iranian population was tribal and nomadic.

Fars was one of those regions heavily populated by tribal groups. Tribal or pastoral nomadic groups were participants, though in different degrees and forms, in a larger regional system of changing social, economic, political and cultural interrelations.

Qashqai political role in the region as a tribal polity has dissolved in the last thirty or forty years. Their "traditional" or premodern forms of rule and rebellion have transformed. Their lives have been deeply integrated with the larger economy and political field.

Study of civil society issues in contemporary Iran includes discussion of its ethnic groups. For many members of larger and smaller Iranian ethnic groups with a dominant tribal background (such as the Kurds, the Baluch, the Turkmen, or the Qashqai and the Lurs of Fars) tribal identities, though transformed, still continue to be relevant.

Qashqai flexible and branching tribal identities still are relied upon to form individual and group networks in the civil society and in the state apparatus. Reconstructed in daily encounters, branching tribal identities constitute only one set of identities for a Qashqai individual, formed alongside, and in combination with, those related to the nation, ethnic group, gender, age, class, religion, political orientation, occupation, kinship group, locality, region, and people with a pastoral nomadic/tribal (ashayeri ) background.

Branching tribal identities are still reconstructed during social gatherings, economic transactions, narrating past affairs, handling of alliances or conflicts, and contesting or affirming local-level political changes brought about in the post-Islamic revolution period.

With all its universalistic claims, the modern state still continues to reproduce and reconstruct tribal identities--for instance, in texts and operations related to census taking, allocating resources, or maintaining order.

Tribe 2
In nineteenth century Fars most of the tribal groups were composed of pastoral nomads and transhumants. Besides pastoralism, other main economic resources included agriculture, gathering, handicrafts, exchange, providing protection and transportation facilities, and banditry and raiding.

The anthropological term "tribe" has been applied, by different writers, to a variety of populations with diverse political structures.

Terms translatable to “tribe” (tayfeh, il, el,) were used in Fars with multiple, flexible, though related connotations. Part of the population of the region was "tribal" (iliati, ili, or ashayeri). Among tribes (in contrast to the non-tribal, mostly Tajik population) patrilineages were part of larger, more inclusive, and branching tribal unit(s) with a hierarchy of political offices. A tribe was a named political and administrative group among the Qashqai, and to a large extent throughout the region, with the following set of characteristics.

At the local level a strong patrilineal ideology was the dominant principle of organization. Other major social principles operating at this level were those related to exchange of labor and service for production shares and products, and those related to rights to animals and land, as well as those pertaining to maternal kinship, exchange of women, and cooperation.

The tribe was basically a branching political unit, a segmental/hierarchical structure composed of segments or sub-units that were, more or less, united under a centralized leadership, but not always a single leader. The leaders of a tribe came from the leading lineage of that tribe, but, over time, there could be power shifts between influential lineages of a given tribe. I will call such a branching, centralized, and stratified tribal structure--such as that of the Qashqai--“segmental” or “segmental/hierarchical” to distinguish it from the branching, but acephalous or egalitarian model known as the “segmentary.”

Region’s tribal groups were not all of the same size. Some groups were very large, such as the Qashqai confederacy, or any of its larger tribes. There were also minor “independent” tribes that were not part of a larger tribe or confederacy. Qashqai local patrilineal descent groups (smaller or larger lineages or "clans") were integrated into larger groups that could be called tribal sections (or sub-tribes). Sections were united to form tribes, and tribes to form a confederacy of tribes. Each level of the official (tax-related) branching structure (section, tribe, and confederacy) had its own formal leadership positions.

The anthropological hierarchy of terms lineage, clan, tribe, and tribal confederacy does not necessarily represent the flexibility of Qashqai branching system and its terminology. The "branching structure" is rather a formal abstraction of historical processes regarding politics of inclusion and exclusion, domination and autonomy, and alliance and opposition. It had to be perpetually reconstructed, imposed, negated, or negotiated. It was a flexible structure, so was the terminology used to refer to its units. The same term that could be translated as tribe, tribal unit, or tribal group (e.g., tayfeh, or tayefeh used throughout Fars by speakers of various languages) could be used to refer to smaller (lineage, clan, section) or larger (tribe, tribal confederacy) units. This terminological flexibility was represented by the tribal people, as well as in different state-initiated documents about tribal matters of the time.

The English phrase "tribal confederacy" is used to refer to the Qashqai as well as the Khamseh of Fars. The Khamseh, as a large tribal group, was formed in the second half of the nineteenth century (later than the Qashqai), was much less centralized, and did not have the linguistic and ethnic homogeneity of the Qashqai.

No belief in a common tribal ancestor was expressed for the Qashqai, or any of its constituting tribes and sections. Each tribe was constructed as an administrative, political, historical, and cultural unit--various cultural markers were used to distinguish each. There were territories associated to each tribe, though each tribe’s winter and summer territories were subject to change over time.

Shift of tribal affiliation, similar to ethnic change, was also common. Tribal groups, whether large, such as the Qashqai or any of its major constituting tribes, or smaller, such as a section or a minor tribe, were formed by fusion of groups of various origins around, or under domination of, a leading lineage.

Tribal units were flexible, they could rapidly enlarge, or break apart, and join other tribal groups. During the nineteenth century, expansion characterized the Qashqai as a whole, its Amaleh tribe (formed by groups directly associated to the paramount leaders), and its two large and newly formed tribes (the Dareshori and the Kashkoli). But, then there were times when some tribes, including the latter two, tried to leave the confederacy. Of large Turki-speaking tribes that existed in Fars prior to the expansion of the Qashqai in the nineteenth century, some were defeated and dispersed, some were included in the Qashqai relatively intact, and some were later officially incorporated in the Khamseh confederacy.

Region's tribes were internally stratified in socio-economic terms, but degree of internal differentiation varied over time and from one tribe to another. Various tribes were not similar in terms of degree of articulation of their political hierarchies with the state structure, and integration of their economies with regional economy. Generally, however, the processes of “internal” socio-economic stratification, and articulation with “external” economies, were more complex and intensified among the Qashqai, than the Lur tribes of western, or the Arab tribes of eastern Fars.

The Qashqai constituted a complex, centralized, and class-based tribal community. It was a multiply constructed, integrative, yet contested political formation. Its nineteenth century history was characterized by internal and external strife. As such, one should refrain from a unitary, functionalist, or adaptational understanding of the Qashqai political structure. Internal cooperation, as well as domination and conflict characterized the Qashqai, or any of its constituting tribes.

Rather than one overall or totalizing model of Qashqai tribal structure, there were a variety of discourses on, and models or aspects of, tribal relations. The official political hierarchy was only one discourse on tribal relations. There also existed parallel (and at times cross-cutting) status and socioeconomic hierarchies. Exchange of women was a primary mechanism reconstructing status groups (one gave wives to groups of equal or higher status, but not to those from a lower status).

Socioeconomic differentiation was formed in the processes of production, control of productive factors, exchange and distribution of products and resources, appropriation and distribution of tribute, as well as banditry and raiding.

Though the Qashqai composed a highly stratified population, economic, political, and status mobility was also common. Banditry and raiding constituted an important channel of upward economic and political mobility. A host of sudden external and internal factors, such as an act of punishment by the state or a tribal leader, and natural factors, such as localized draughts, could also affect dynamics of socioeconomic differentiation in a tribal community. Hierarchical relations had their relative fluidity. They, too, had to be constantly proclaimed, recognized, or contested, in short, reconstructed. Tribal status and socioeconomic hierarchies, too, have rapidly dissolved and transformed in the last three or four decades.


The Segmental-Hierarchical Structure and Its Alternatives
The model of social structure presented in the literature on Middle Eastern tribal populations has been mainly either a segmentary or a segmental-hierarchical one. Larger Iranian tribes during the nineteenth and twentieth centuries have usually been described as hierarchical and segmental (branching).

Segmental-hierarchical model can be regarded as the anthropological representations of the "official" or "formal" tribal model, presented by state and tribal elites. This model is presented in "official" documents regarding the administration and taxation of the tribal groups, and also by tribal (usually elite) informants to outsiders, including anthropologists. It is, however, only one of the discourses on tribal politics presented by the tribal elites and masses. The segmentary model has been criticized by two “alternative” approaches. One was initiated by Bourdieu (1977), in his study of an Algerian tribe, and the other developed by Geertz and others influenced by his approach (for instance Eickelman 1976 and Rosen 1979) in their studies of Moroccan communities.

Here, I present the segmental-hierarchical model as one (set) of several discourses on, or models of, Qashqai society. Similar to the non-orthodox or alternative models, I stress flexibility, variability, practical context, and negotiative aspects of tribal relations, groupings, identities, and networks.

Alongside dissolution of “traditional” official hierarchical relations, individual and group networks have become even more important in daily interactions.

The "network" model on social relations is presented in life-histories and narratives of previous political encounters, and in the way people interact with others and discuss such interactions. However, in order to limit the individualistic and market-model biases of the some versions of the alternative outlook, I point to social consequences and contexts of personal networks and negotiations.

Personal networks are, or become, social networks. One’s association with an urban merchant, a state official, or another individual from another tribal group or village could be used as a base for constructing further networks by one’s relatives. Furthermore, negotiations of social relations are carried out in contexts where individuals and groups have differential access to force, wealth, and symbolic capital. Tribal relations and identities were negotiated, but they were also imposed.


Flexibility of Alliances and Counter-Alliances
A pertinent, persistent, and pervasive feature of tribal politics among the Qashqai--and among other tribal groups in Fars--has been a flexible and nesting system of alliances and counter-alliances. This somewhat perplexing, but, nevertheless important feature of internal and external tribal politics has remained mainly understudied. It is, however, a main aspect of political relations as they were reconstructed by the region's tribal populations, and as they were retold in oral and written narratives.

Flexible system of alliances was also a main aspect of relations of tribal groups to non-tribal political forces in the region--such as the state, and the rural and urban civil society political forces. In the course of reconstructing (flexible) alliances individual networks and group relations were brought together.

Flexible alliances and counter-alliances are related to another major focus of this study, that of social dramas. During these ritual-like processes of political conflict individuals and groups represent and, at the same time, reconstruct their political relations, alliances, and identities. Sudden shifts of political alliance do happen during social dramas. Flexibility of alliances is related to the constructive potentialities of social dramas. Improvisation is a defining facet of dramas of political conflict.

Flexibility of alliances is not a feature confined to tribal groups or to pre-modern period. It is characteristic of large or small scale politics, of nomadic, urban, and rural communities, and of premodern as well as modern political fields.

Elastic factionalism has puzzled interpreters of Iranian politics. Flexible politics could also be viewed in the light of the fact that political relations are subject to constant reconstruction, whether in ordinary practices, or in extra-ordinary social dramas. It is the negotiative and improvisatory aspects of political relations, manifested particularly during social dramas, that gives them their elastic form.


Bandits, Rebels and Raiders
Political leadership among the Qashqai was not confined to positions in the formal official political hierarchy. Heads of bandits and raiding groups were also men of power. Tapper (1983), in his overview of nineteenth century tribe-state relations in Iran and Afghanistan, pointed to the difference between two types of tribal leadership, namely the brigand and the chief.

He also distinguished between two situations concerning tribes, each situation with its own type of leadership. In the "tribal situation" brigands and their followers challenge state officials and state power. In the "government situation," where the state structure has incorporated the tribal hierarchy, hereditary chiefs rule over tribal followers, represent them, and mediate between them and other communities and powers. Actual tribal leaders combine varying characteristics of the two mentioned leadership types. These two related aspects of tribal leadership has also pointed out by van Bruinessen among Kurds (1983:374-375).

In Fars, the lines differentiating a tribal situation from a government situation were elastic, subject to constant reconstruction. The two situations were characteristic of the same region, and even of the same tribe during a given short period. The two forms of power were complementary as well as contradictory. The same person could shift his power from one form to another. A head brigand-raider could become a tribal leader with an official state position, while a tribal chief could challenge the central government and/or its provincial representatives by acting as a rebellious brigand-chief, or by supporting such groups. Official tribal chiefs, paradoxically, to further their power, could even support banditry and raiding.

Brigand form of power among Iranian tribal groups has remained largely understudied. Banditry was considered a rebellious act. The formal or the state discourse downplayed this alternative form of leadership, and pushed aside the alternative discourse on brigandage.


Premodern or Traditional Social Structure
To picture the multiplicity of discourses on tribal relations, or the diverse aspects of Qashqai tribal society, one needs to examine a variety of texts. These are texts of formal and informal nature, of periods of conflicts and social dramas, as well as of years of solidarity or compliance. Such texts are those produced by various Qashqai, of different backgrounds and in different contexts, as group or as individual practices, as well as those presented by the non-Qashqai about the Qashqai. Rather than representing the tribe (or the tribal confederacy) in terms of a model of tribal structure, one should look at a set of multiple discourses on tribal relations, or a set of models of the Qashqai social ties. Official political, status, and socioeconomic hierarchies, the segmental or branching system, the tribal chief/brigand dialectic, individual and group networks, flexible system of alliances and counter-alliances, were all aspects of the Qashqai tribal social world.

I use the term "traditional" (an unfortunate choice) to refer to what existed in the nineteenth century, particularly the second half of that century. Yet, the nineteenth century, could be better viewed as a transitional period between the pre-modern and modern periods. With a political field dominated by the nation-state and modern discourses on the nation, the modern period brought gradual dissolution and transformation of traditional tribal relations

I view the Qashqai "premodern," or "traditional" tribal social structure as characterized by a variety of interrelated "aspects."

We may call these aspects, “structures,” and “models” of, or “discourses” on, the Qashqai “traditional tribal social (or socio-political) relations" --

i) There was a hierarchical formal administrative structure which integrated the Qashqai with the state structure (for taxing and “order” purposes).

There were also the hierarchies of (ii) the status groups and of (iii) the socio-economic classes.

There was also (iv) a segmental or branching structure to the whole confederacy and its divisions.

Heads of bandits and raiding groups were also men of power, or, to use Tapper’s (1983) phrase for tribal societies of Iran and Afghanistan, there existed (v) a tribal chief/bandit-rebel dialectic.

Qashqai heads of groups of bandits, rebels, and raiders, similar to the tribal chiefs, were engaged in appropriation and distribution of surplus, and punishment. Typically, tribal rebels could become chiefs; the reverse was also the case.

(vi) Individual and group networks constituted another major aspect of social life. These networks brought individuals and groups across tribal groupings together. They also brought individuals and groups from pastoral, rural, and urban background in social contact. Regional alliances included urban, rural, and nomadic/tribal forces.

(vii) The political field at local, regional, and national levels was characterized by a flexible system of alliances and counter-alliances. Over time, many individuals and groups changed their political and even tribal and ethnic affiliation.

Flexibility of alliances was even practiced during the course of social dramas, sometimes suddenly in the climax of dramas. Narratives of so many social dramas point to moments of impromptu change of political side, not only by main actors, but also by the viewers. Regional alliances normally contained urban, nomadic, and rural forces.

The Qashqai were active participants in the country-side political space. Their migratory routes pass by the region’s capital, Shiraz. In many regional and national upheavals, both during the nineteenth and modern (twentieth century) eras, the Qashqai also took part in the dramas staged in urban political space.




1 - Discussions of theoretical and empirical problems related to the usage of the concept “tribe” in anthropological literature date back more than two decades, for instance in Helm (1968), Sahlins (1968) and Fried (1975). Various writers have expressed serious reservations about the usage and vagueness of the term “tribe,” and related concepts. Among others, Godelier (1977) has proposed to discard the term as an anthropological category. The term, however, continues to enjoy widespread usage in anthropological literature, especially in introductory text books and ethnographic accounts of many populations in different parts of the world, including the Middle East.

2 - Anthropological studies of Middle Eastern tribal groups are numerous. Majority of anthropological works on Iranian communities are on its tribal groups, e.g. Afshar Naderi (1968), Barth (1961), Beck (1980, 1986), Black-Michaud (1972, 1976), Bradburd (1980, 1990), Digard (1973, 1979), Fazel (1971, 1979), Irons (1975), Loeffler (1976, 1978), Salzer (1974), Salzman (1971, 1972), Shahshahani (1980), Spooner (1969), and Tapper (1979a, 1979b).

Disagreements on conceptualization and analysis of Middle Eastern tribal structures are also characteristic of the literature. As an example, Brooks, Digard, and Garthwaite, in the above mentioned volume by Tapper (1983), expressed basically dissimilar ideas on some of the basic socio-political characteristics of the Bakhtiari confederacy, neighbors of the Qashqa’i. Previously, Fazel (1979) and Leoffler (1978) had presented different interpretation of the tribal system of the Boir Ahmad, neighbors of the Bakhtiari and the Qashqa’i. Examples on populations outside Iran are Barth (1959) as opposed to Ahmad (1976) on the Pashtun, and Gellner (1969) and Hart (1976) in contrast to Eickelman (1976), among others, on Moroccan groups.

- Bharier 1977, Gilbar, G. 1976

- Though having its roots in the term ashireh (tribe), it could connotes pastoral nomads. Pastoral nomads in Fars were also mainly tribally organized. A better translation for our case is perhaps the couplet pastoral nomadic/tribal.

- Local communities were the site of appropriation of surplus labor and reproduction of relations of production, exchange, and distribution. Multiple relations of domination and cooperation characterized the local level nomadic communities. The Qashqa’i included a population of laborers (households of shepherds, dependents, and retainers). But, in contrast to the majority of rural population, it also included a large population (class, strata, or classes) of small and medium-size property owners. Before the transformations of the last three decades, a large population of the laborers was working for small and medium-size property owners. Besides the elite, small and medium-size property owners also had direct economic, social, and cultural ties with urban merchants.

- Analysis of the political characteristics of some polities in other parts of Asia and Africa in terms of the “segmentary”-state model (that is a combination of branching and centralized state-like or state-incorporated structures) has been provided, for instance, by Leach (1956) and Southhall (1988). In the classic segmentary model the opposing units would unite at a higher level of segmentation to oppose another unit at that level. Among the Qashqa’i, as it will be described later, this idea held generally true in many cases, but there were also numerous other cases where internal political conflict (among political leaders who were cousins and half-brothers--and their tribal followers--was perpetuated and even promoted by conflicts in the larger field.

Similar to Leach’s model (1956) on the highland tribes of Burma, I have portrayed historical data on the region’s tribal groups as oscillations along continuums between dispersed-centralized poles and egalitarian-hierarchical poles (Shiva 1973). Gellner (1981:38), Gathwaite (1983:12-16), van Bruinessen (1983:369), and Tapper (1983) have pointed to such shifts among Middle Eastern tribes.

- Such a terminological flexibility is generally characteristic of other tribal groups in Iran and other parts of the Middle East.

- Among some smaller tribes in the region the idiom of over-all intra-tribal patrilineal kinship (especially myths referring to a common ancestor) was sometimes expressed, together with other themes such as common history, leadership and territory, to signify tribal unity and identity. See, for example, Afshar Naderi (1968) on the Bahmai tribe of Kohgiluyeh.

- I have emphasized the class nature of the Qashqa’i society and criticized functional and adaptational interpretations of Qashqa’i tribal structure, because of the tendency to overlook the society’s internal contradictions (Shiva 1978).

- Beck’s valuable study (1986) includes discussion of some major hierarchical and segmental aspects of Qashqa’i society.

- The Yomut Turkmen in the nineteenth century, described by Irons as non-hierarchical or non-stratified segmentary (1975), is a main exception. Fazel (1979) has also provided an interesting (and “atypical”) analysis of the Boir Ahamdi political structure, by adopting a model similar to that used by anthropologists to describe Pacific Islands' chiefs and chiefdoms.

- A somewhat similar system of a checkerboard of alliances and counter-alliances, known in the literature as the liff system, has been noted for the Moroccan tribal groups--for instance by Hart (1976). My particular emphasis is on their flexibility.

- Flexible alliances are inclusive of an "individual-based" model of social structure, similar to that presented by Eickelman (1976) and Rosen (1979), with an orthodox or "group-based" model, similar to that of Gellner (1969) and Hart (1976) for Morocco.

- Bill (1975).

- Some commentators have gone as far as attributing features such as compulsive factionalism and politics of distrust to the Iranian "national character" (e.g. Westwood 1956:122-123, quoted in Abrahamian 1982:171), while some, such as Abrahamian (1982:171-172), have stressed the role of social and political issues in political conflicts of contemporary Iranian history. In the last several decades, new social and political discourses have increasingly played a larger role in the way Qashqa’i groups and individuals politically express themselves in the civil society, and also in the state apparatus.