Navigation – Plan du site

AccueilNuméros184Tribe and state in Iran and Afgha...

Tribe and state in Iran and Afghanistan: an Update

Tribu et État en Iran et en Afghanistan : une mise à jour
Richard Tapper
p. 33-46


L’auteur reprend ici l’analyse de « la tribu et l’État en Iran et en Afghanistan » issue d’une conférence qu’il avait organisée en 1979. Il constate que les thèmes abordés à cette occasion sont toujours d’actualité compte tenu des bouleversements politiques et sociaux des trois dernières décennies consécutifs aux révolutions intervenues dans ces deux pays. S’il y a une résurgence du tribalisme, les formes et fonctions des « tribus » ont considérablement évolué. L’auteur réitère sa conclusion : « Si le tribalisme a ses défauts et ses limites, il n’apparaît nullement comme un anachronisme dans la mesure où il pourvoit à la sécurité de ses membres et leur assure une survie à long terme » dans le monde contemporain.

Haut de page

Entrées d’index

Haut de page

Texte intégral

1IN SUMMER 1979, the late David Brooks and I convened a conference in London on the theme of ‘Tribe and state in Iran and Afghanistan, 1800-1980’. Each of us had conducted ethnographic fieldwork during the 1960s in one of the major tribal groups in Iran, David among the Bakhtiari and myself among the Shahsevan, and we had both made extended studies of the histories of these groups. In 1970-1972, I had also worked with Durrani tribes people in northern Afghanistan.

2The idea for the conference was born in Tehran in late 1977, as we discussed our reactions to the Festival and Seminar on Popular Traditions that we had just attended in Isfahan. The tribes had featured prominently in the Festival, and it was clear that the Iranian government now considered them to be ‘safe’ enough to promote as cultural curiosities, as a tourist attraction as well as grist to the ethnographers’ mill. It was no longer mentioned that some tribes had once posed a political threat to the state. David and I felt that, in view of the considerable amount of research that had been done over the last two decades on the ethnography and history of the tribes of Iran and Afghanistan in the 19th and 20th centuries, the time was now ripe for stock-taking and generalization, and an attempt at systematic comparison within a historical perspective, both between Afghanistan and Iran and also with other areas of the world.

3By the time of the conference political upheavals in both countries had given the topic added interest and contemporary relevance. There had been the sort of tribal resurgence that so often in the past had accompanied such upheavals. Among the initial problems faced by the new Islamic Republic of Iran was resistance on the part of regional, ethnic and tribal minorities. Within the country, substantial numbers of pastoral nomads settled over the previous decades had resumed their former way of life, and tribal leaders long used to exile in the West had been welcomed back. In Afghanistan, following the socialist coup in April 1978, successive governments had failed to win popular support, and an insurgency of Islamic and tribal elements was spreading.

Tribes, Nomads, Pastoralism and the State

4In planning the conference, we were aware of two major questions that would need to be addressed: the definition of ‘tribe’, and whether tribes could be defined or even discussed apart from the state – we had both participated in debates over tribe-state relations, and were convinced that – at least in these two countries – any discussion of the nature of either tribe or state, now or in the past, must include a discussion of the relationship between them.

5The old anthropological question, ‘What is a tribe?’ could not be avoided: there was considerable diversity of opinion, both among anthropologists and historians, and among those studying Iran and Afghanistan, not to mention their subjects, as to what constituted a ‘tribe’. Moreover, standard anthropological notions of ‘tribe’ bore little relation to the groups so labeled in these two countries, or indeed to the Middle East generally. The classical model of tribal society in the Middle East, conforming with Durkheim’s notion of ’mechanical solidarity’, was of egalitarian descent groups. This criterion best fits Arab tribal societies, where genealogies are particularly extensive; a classic example is the Rwala, a ‘tribe’ of some 250,000 souls, though some even larger non-Arab groups such as the Bakhtiari (500,000) of Iran or the Durrani Pashtuns (over 2 million) of Afghanistan have been called ‘tribes’ on the same grounds. Many proponents of this view would deny the term ‘tribe’ to any group without a descent ideology.

  • 1 Other writers (such as myself), however, are unwilling to take either extreme position, and refer t (...)

6Others, however, saw tribes as essentially territorially distinct political groups and expected them to be led by chiefs; they applied the term ‘tribe’ to almost equally large groups that lacked unifying descent ideologies and were heterogeneous in origins and composition, such as the Qashqa’i, the Khamseh or the Shahsevan in Iran. At this level of major cultural-political groups of 100,000 or more people, then, there was disagreement as to whether the term ‘tribe’ is applicable on the grounds of culture (a descent ideology) or political structure (chiefship and/or political-territorial unity). 1

  • 2 This notion is held by numerous historians and other writers, who also assume tribes to be descent (...)

7Another notion that is no part of standard anthropological definitions but is strongly entrenched in both academic and administrative discourses in many parts of the Middle East, is of ‘tribe’ as the political and sociocultural dimension of pastoral nomadism, such that the category of ‘the tribes’ is conventionally synonymous with ‘the nomads’. 2 But there is nothing in either pastoralism as a system of production or nomadism as a mobile way of living that necessarily leads to organization in tribes, whether defined politically in terms of territory and chiefship, or culturally in terms of common descent.

8Numerous observers have noted how the geography and ecology of most Middle Eastern countries favour pastoral nomadism. The terrain and climate made large areas uncultivable under preindustrial conditions, and suitable only for seasonal grazing; and as only a small proportion of such pasture could be used by village-based livestock, vast ranges of steppe, semi-desert and mountain were left to be exploited by nomadic pastoralists. Such nomads until very recently numbered tens of millions, and almost all were organized politically into tribes under chiefs. Equally, tribes (defined in political terms) have commonly also had a pastoral economic base and led a nomadic way of life.

9But an insistence that tribes in the Middle East were necessarily pastoral nomads, organized in descent groups, ignores major tribal groups in Anatolia, Iran, Afghanistan and indeed several Arab countries, which often included both settled cultivators and pastoral nomads and were complex and heterogeneous in composition. Thus, most of the Pashtuns of Afghanistan are (and have always been) farmers or traders, with little or no leaning to pastoralism or nomadism, and well-known groups in Iran such as the Qashqa’i, Bakhtiari, Kurds, Baluch, Turkmen and Shahsevan have been at least partly settled agriculturalists. Of course, by conventional anthropological definitions, many of these groups were not tribes at all, but ‘chiefdoms’, or even ‘protostates’.

10Any coincidence between nomads and tribes (whether descent-based, or led by chiefs) was not so much a causal relation as a function of relations of both with central states. Settled state administrations intent on registering and taxing the inhabitants of territories which they claimed to control have classically had ambivalent attitudes to both tribespeople, with their personal allegiance to each other or to chiefs, and nomads, with their shifting residence. Many earlier states, however, were themselves founded on military forces drawn from pastoral nomadic tribes, often organized in military units of tens, hundreds and thousands. Rulers have fostered pastoral nomadism in strategic parts of their territories, and have frequently created tribes, tribal organization and tribal chiefs.

Explaining Variation in Tribal Organization

  • 3 See my ‘Introduction’ [Tapper ed. 1983: 1-82]. I have elaborated some of the main arguments in vari (...)

11The conference amply justified our hopes for a diversity of perspectives and fruitful discussion. In the long ‘Introduction’ to the resultant volume I drew some comparative conclusions based on a reading of the papers in the light of developments in Afghanistan and Iran since the conference. 3

12In both Iran and Afghanistan, tribal groups had been notorious as makers and breakers of kings, and dynasties of tribal origins had ruled both countries until well into the 20th century. Both states had ‘tribal problems’; and the tribes in each had ‘state problems’. Historically, tribes and states formed a single system: until recently, no state was without tribal elements, no tribe existed without relations to at least one state.

13Empirical tribal groups in Afghanistan and Iran conformed to no single pattern of organization. None of the following were universal features: pastoral economy, nomadic or semi-nomadic movements, descent group organization, centralized chiefship, egalitarian ideology. Nor did conventional images of tribes conform to a single stereotype. Afghan tribes were renowned as hardy, independent, warlike mountain villagers, farming barren fields, and rigorous if not fanatical in their devotion to Islam. The tribes of Iran by contrast were understood to be pastoral nomads, organized into strong centralized confederacies under powerful and aristocratic chiefs, and notorious for their ignorance of or indifference to Islam. Both stereotypes contained some truth, but they disguised a wide range of empirical variation between the extremes of the powerful, centralized tribal chiefdoms of Western Iran with their broad pastoral bases, and the independent, hardy, egalitarian Pakhtun mountain villagers of the eastern borderlands of Afghanistan.

14There had been several attempts to identify ‘internal’ and ‘external’ factors in these varied patterns of tribal organization and leadership, and to apply a range of theories of evolutionary and cyclical change to the tribes of Iran and Afghanistan.

15For instance, in his classic study of the Basiri tribe of Southwest Iran, Fredrik Barth made an important analytical distinction between, on the one hand, the local pastoral communities (of some 20-30 families) whose organization and structure were generated by internal factors such as the pastoral ecology and economy; and, on the other hand, the tribe and the Khamseh confederacy of which it formed a part, a political superstructure organized by or in response to external factors such as settled society, neighbouring tribes, and agents of the Iranian state [1961].

16Others showed how competition for pastoral and agricultural resources both within and between local communities led to the formation of rival blocs. Rivalry and conflict were commonly mediated by independent religious leaders. Formal political chiefship emerged only in conditions of surplus production, or where a dominant chiefly group could be supported by external resources, for instance by raids on vulnerable trade routes or wealthy urban centres, or across international borders.

17Alternatively, state rulers appointed or approved tribal chiefs. Where the state was weak, it might have to buy the chiefs’ loyalty with subsidies; where it was strong, it could demand taxes and tribal military levies. State control over the tribes fluctuated over time. Tribal groups, in their turn, could adapt and integrate, or practice avoidance and – in Ernest Gellner’s terms [1983] – ‘divide that ye be not ruled’. A strong state cannot tolerate such diffusely-organized ‘jellyfish tribes’ as Malcolm Yapp called them [1983]; in order to control or destroy them it must first provide them with a backbone in the form of chiefs.

18At the local community level, the common form of leadership was that of greybeard elders, with little power or authority of their own. Above that level, leaders of tribes and confederacies displayed some combination of elements of two ‘ideal types’, ‘the brigand’ and ‘the chief’. The former depended for authority over his followers on being able to ensure a flow of booty; the latter was more likely to inherit his position or to rely on government appointment. A third type of leader emerged at different historical junctures: the mullah or Sufi sheikh or pir, whose authority over his followers was primarily charismatic and spiritual, though undoubtedly some followers were motivated by expectations of booty, as well as respect for the leader’s holy lineage. Actual leaders were likely to combine elements of these ideal types of authority.

19The history of tribe-state relations in Iran and Afghanistan showed evidence of processes of both evolutionary and cyclical or alternating change. There were instances of state formation among tribes, but also of devolution or retribalization, for example following the dissolution of the Kurdish emirates in the 19th century. Cyclical processes too were evident in the history of tribal chiefly and royal dynasties in these two countries, conforming to the well-known and much-discussed model derived from Ibn Khaldun.

20In summarizing the material from Iran and Afghanistan, I suggested that the most useful way of conceptualizing ‘tribe-state relations’ was not as an opposition of substantive social, economic and political structures so much as an opposition of tendencies, modes or models of organization, not just analytically distinct but consciously experienced as a tension within the tribal groups and states discussed. As bases of identity and political allegiance and behaviour, ‘tribe’ gives primacy to ties of kinship and patrilineal descent, while ‘state’ insists on the loyalty of all persons dwelling within a defined territory, whatever else their relation to each other. ‘Tribe’ stresses personal, moral and ascriptive factors in status, while ‘state’ is impersonal and recognises contract, transaction and achievement. The division of labour in the tribal model is ‘natural’; in the state model it is complex. The tribal mode is socially homogeneous, egalitarian and segmentary; the state is heterogeneous, inegalitarian and hierarchical.

21The opposition between these two models, their confrontation with each other and with social reality, creates a tension, a dialectic with varying resolutions. Thus, whether because of ecological limitations, state pressures or inherent contradictions, the ‘pure’ tribe is an empirical impossibility. Most groups that were termed tribes had some form of segmentary descent ideology as the basis for political loyalties, but all used other principles too, to guide action and association at different levels of organization. All tribal groups in Iran and Afghanistan had a territorial dimension, though they tended to ascribe common descent to all those who, by whatever means, had acquired rights in their territory. But there were several tribes, especially in Iran, which did not even pretend to an ideology of common descent, organizing as explicitly political local groups with a common leadership; in these terms, they were proto- or ministates within larger, empire-like states. The strength of egalitarian ideals varied widely, as did the extent of inequality in practice; even the most egalitarian tribal groups displayed some inequalities of wealth (however narrow) and leadership roles (however lacking in authority). Conflict over material interests, between rivals or between rich and poor, was endemic to all tribal groups; and ‘anti-segmentary’, territorially-based blocs and factions were regular tribal phenomena.

22The ‘pure’ state is similarly impossible. Citizenship (that is, in premodern Iran and Afghanistan, subjection to the ruler) is acquired through a mixture of territoriality and descent. In other words, there is state within every tribe, and tribe within every state; state is partly defined in terms of tribe, tribe in terms of state. Most empirical tribes and states are various forms of hybrid, such as chiefdoms, confederacies or tribal states. Tribal states take different forms. Every state must boost its legitimacy, claim the moral allegiance of its citizens, by promoting elements of a common national culture and way of life. The most powerful symbols in a nationalist ideology are shared religion and a concept of mother/fatherland, but insofar as the ideology stresses common descent or origins (real or fictive, plausible or otherwise) it resembles a tribal one. Some states have claimed common descent or origins for all citizens and sought to deny or eliminate any internal ethnic differentiation; but most contemporary states have to recognize minorities, which may be tribal in culture or organization, though today they are often termed ‘ethnic’ or ‘regional’ groups or ‘nationalities’. Other states, controlled by one ‘tribal’ (descent-based) elite, may make no attempt to disguise cultural differences under a national ideology, rather reserving privileges and power for the dominant tribal group.

23Both these modes, the ‘tribe’ and the ‘state’, exist as opposed cultural categories within the experience of individuals, as well as in the structure of systems: dyads from Afghanistan such as khan and khel, qawm and gund, nang and qalang, resonate with classics in the anthropological literature such as siba and makhzen from the Maghreb, gumsa and gumlao from the Burma Kachin. They are not descriptive but cultural categories, idioms that are inadequate to explain the fluid and complex workings of actual tribal societies, let alone the relations of tribe and state. Account must be taken of the formation of ‘anti-segmentary’ communities at certain levels of organization, of the patterns of bloc alliances among them, and of increasing centralisation involving hierarchical relations of patronage.

24In effect, varying articulations of all three processes – segmentarity, community and bloc formation, and centralisation/patronage – produce the transformations of tribal society that have been observed. The major variable is the influence of the state, both as an external force and as an idea in opposition to the idea of tribe. The essence of the latter is indeed kinship and egalitarianism (the basis of a segmentary lineage system), while the state is essentially territorial (the basis of communities and opposed blocs) and centralized (the basis of patronage). It is in these terms that we can understand both the variations in actual tribal forms and the changes that have occurred, whether we adopt a cyclical (or oscillation) model of change or acknowledge the apparently irreversible (evolutionary) changes that have now taken place in the transition from tribe to state.

The Return and the Transformation of the Tribes

25I reproduce these arguments here, because they seem to me to have lost little of their validity in the intervening thirty years. Yet prevailing notions of Middle Eastern tribes continue to suffer from some of the misconceptions that my analysis sought to dispel.

26In the mid-20th century, social scientists of all persuasions expected tribal and ethnic minorities within contemporary nation-states to succumb sooner or later to policies of modernization and national integration, and many were confident that class would replace ethnicity as the major dimension of social and political identity. Many anthropologists began to regard the study of their traditional subject-matter – tribal peoples – as an antiquarian irrelevance, turning instead to the newly fashionable subdisciplines of urban anthropology and the anthropology of the state.

27These expectations and trends were confounded towards the end of the century by the persistence or creative revival of tribal and ethnic minority identities in virtually all countries of the world, and by increasing academic and popular perception of violent intercommunity conflicts as tribal or ethnic in nature. Sociologists, political scientists, historians, geographers and others showed renewed interest in the study of ethnic and tribal minorities of the ‘Fourth World’ – no longer the sole preserve of anthropologists.

28Not least of the problems involved is the question of the categories and terms of description and analysis to be employed. Prominent among such terms have been ‘tribe’ and ‘tribalism’. These refer to a category of human society whose study was once regarded as largely the prerogative of anthropology, yet anthropologists themselves have notoriously been unable to agree on how to define them. Small wonder then if historians too, and for that matter political scientists and others interested in tribalism, have differed widely in their understandings of the terms.

29The initial problem, then – one that is central to the discipline of anthropology, even more central than it is to philosophy – is that, in our study of human society and culture, we are constantly faced with the inadequacy of our technical terms: either they are neologisms, technical jargon; or we try to give technical, jargonistic definitions to words borrowed from popular discourse. The first case, jargon, by its nature, alienates us from popular discourse. In the second case, our terms have to fill the impossible dual role of being both tools and objects of analysis. Yet successful jargon eventually finds its way into popular discourse. ‘Tribe’, ‘tribalism’, ‘ethnicity’ (and one could add ‘state’ and ‘race’) are problematic cases in point.

30Let me illustrate by quoting reports from three countries that have never been far from the front pages of western newspapers since 2001:

  • 4 Amatzia Baram, New York Times, 28 October 2003.

311) ’Attacks on coalition troops [in Iraq] should be viewed through the prism of tribal warfare. This is a world defined in large measure by avenging the blood of a relative (al-thā’r); demonstrating one’s manly courage in battle (al-muruwwah); generally upholding one’s manly honor (al-sharaf).’ 4

  • 5 Anonymous, Iran Daily, 20 October 2004.

322) ’Tribal tourism is a means of eradicating poverty and expanding social justice for it would benefit one of the most deprived social groups. Deputy head of Iran Tribal Affairs Organization for Social and Communications Affairs, Mohsen Ahmadi, elaborated that selling local handicrafts, artistic works and dairy products, and renting out tribal tents and lodges would help diversify the tribesmen’s jobs and boost their economy... Ahmadi pointed to migrating tribes who are scattered almost all across the country as one of the most fascinating draws for foreign visitors and anthropologists.’ 5

  • 6 Shoib Najafizada, Agence France Presse Report, 26 October 2004.

333) ’At first glance the election commission’s map of Afghanistan appears to be a colour-coded layout of the country’s patchwork ethnic groups. On closer inspection, it turns out to be a display of voting results.’ 6

34Among other things, these reports indicate, first, the continuing importance of tribes, tribalism and ethnicity. But they also show continuing disagreement as to what these phenomena are. In Iraq tribalism and tribal identity (as marked by violence and masculine honour) are viewed as at root cultural phenomena; in Iran, the tribes are seen as peaceful, exotic, pastoral nomads, a tourist attraction; in Afghanistan, ethnic groups and tribes display their political nature in voting patterns.

35Since the invasions of Afghanistan and Iraq such references to tribes and tribalism, by journalists, NGO researchers and seasoned academics, have multiplied. Viewed overall, they display the same ambivalences that have dominated attitudes towards tribes on the part of both foreign observers and Middle Eastern urbanites for a hundred years or more. Thus, the more romantically inclined, who perhaps idealize the tribes, regard them as the archetypal ‘civil society’, in the sense of a refuge from a failed or oppressive state. Most politicians and political scientists, on the other hand, view tribes and tribalism as obstacles to the proper functioning of the state, and in particular, to the development of ‘civil society’. Analysts, moreover, are divided between those who see tribes as inherently egalitarian, if not democratic, and those for whom tribes are necessarily feudal, tyrannical and patriarchal, led by powerful chiefs who impose their will ruthlessly through armed thugs and threaten the security of the state.

36Any notion that there is a necessary historical association between tribal organization and pastoral nomadism is doubtful, either in Iran or Afghanistan. For these two countries – and others in the region between the Atlantic and the Himalayas – there is no necessary association between pastoralism, nomadism, and tribalism. Many known (and probably many unknown) groups of pastoral nomads have avoided tribal organization, as a strategy to avoid state control. Major tribal societies numbering millions consist of peasant farmers and urbanites, with little or no pastoral or nomadic connections or history. The first question proposed for discussion – ‘To what extent [does] the evolution of pastoral economic activities in the context of globalization allow nomadic practices to still occur and if this way of life, profoundly shattered during the past decades [...] eventually contribute to perpetuate some tribal orders?’ – would seem of limited relevance to such examples.

37Similarly, the notion of tribes as stateless descent groups may have some validity in North Africa, but it cannot apply to most historical tribal groups in Afghanistan, Iran and beyond.

38The ambiguity in discourses about ‘tribes’ – whether they are pastoral nomads or settled farmers, whether they are chiefdoms or descent groups, cultural or political, even pro-state or anti-state – not only divides academics but obscures (or fuels) current political debates at national level about the future role of tribes and tribalism.

39Officials – and many academics – have taken a highly positivist view of tribes, expecting them to be mappable, bounded groups, with little membership change, and wanting an exact terminology for classificatory and comparative purposes. From a government perspective, even the most autonomous rural populations should have identifiable patterns of organization, and leaders who may be treated as representatives; if they do not have these patterns or leaders, they may be encouraged to produce them. Some rural and nomadic populations have avoided government control and exploitation, and even the attention of historians, by failing to produce such leaders or recognizable forms of ‘tribal’ organization. But government-created ‘tribes’, whose names may appear in the records as such, may exist only on paper. Further, tribal names found in official sources imply a uniformity of sociopolitical structure which, in so far as it exists, may be entirely due to administrative action, and may disguise fundamental disparities in culture and in forms of social organization.

  • 7 The same applies to the English terms ‘state’ and ‘government’, and their local equivalents (dowlat (...)

40A desire to establish a consistent and stable terminology for political groups has too often obscured the nature of indigenous concepts and terms, which are no more or less precise or specific as regards socioeconomic or cultural features, degree of political centralization, level of organization and so forth, than their English language equivalents such as ‘tribe’, ‘tribal group’, ‘clan’, ‘lineage’, ‘family’. 7 Even in the most apparently consistent segmentary terminology, individual terms are ambiguous, not merely about level, but in their connotations of functions or facets of identity: ‘economic’, ‘political’, ‘kinship’, ‘cultural’. However, as with equivalents in English practice, the ambiguity of the terms and the flexibility of the system are of the essence in everyday negotiations of meaning and significance. Most of the terms that have been translated ‘tribe’, such as il, cashīra, oymaq, qabīla, qawm, tayfa (and their plurals), contain such ambiguities, and attempts to give them – or ‘tribe’ – precision as to either level, function or essence, are liable to be misdirected. A precise terminology may aid comparison, but is unlikely to explain behaviour or to provide an adequate translation of local categories and perceptions.

41Such a perspective identifies the categories ‘tribe’ and ‘tribal’ as ideological concepts and disqualifies them from analytical usage. Any use of these terms, or any ascription of a tribal identity, is a political statement that defines the speaker and their relation to their audience as much as it defines the group or individuals so labelled. The same disqualification applies to the use of any particular name for a population. There are indeed strong continuities represented in tribal names; but their content (what it means, for example, to be a Kurd or a Pashtun) changes and differs from context to context. For the theorist this is a fascinating field of investigation. For the policy-maker it is a minefield.

42In modern times, the nature of the state in the Middle East has been transformed, and new nation-states have been created, with clearly defined territorial borders and citizenship, centralized administrations, powerful standing armies; state institutions have pervaded whole areas of social, economic and religious life that were previously not the state’s concern. In many new nation-states, such as Republican Turkey, Pahlavi Iran, Afghanistan since 1880, and Baathist Iraq, new regimes sought ‘national integration’ by imposing a national language and sometimes religion, usually those of the ruling elite and not always that of the majority of the citizens. The state inevitably treated as enemies of such ‘national integration’ and state-formation all minority identities (tribal, ethnic, religious) that claimed the primary loyalties of their members. Many such tribal, ethnic and religious identities also span state frontiers drawn in colonial times; notably the large stateless tribal ‘nationalities’: Kurds, Pashtuns, Baluches and Turkmens.

  • 8 See A. Baram [1997] ; see also R.L. Taylor, ‘Tribal Alliances: Ways, Means, and Ends to Successful (...)

43In Iraq, until the mid-20th century tribalism and ethnicity were the dominant sources of identity and organization. The major ethnolinguistic identities were Arab and Kurd, each comprising a range of different tribal groups. Religious allegiances (Shi’a and Sunni and the numerous small minorities) often cut across tribal boundaries. Under the UK-backed monarchy, then the Baath, Iraq underwent ‘modernization’ and nation building in the conventional pattern: tribal and ethnic identities were suppressed and denied. Saddam Hussein however promoted his own tribal group, and arming and subsidizing the Sunni tribes was a significant element in his post-1991 reorganization, 8 neglected or misunderstood by the occupiers long after the 2003 invasion. Tribal groups have largely reemerged, with sheiks (usually best regarded as elders, not chiefs) expected to represent their followers to the new authorities. In the absence of a credible or legitimate state, especially in the countryside, tribes have offered identity and solidarity and all the functions of ‘civil society’ and the state, not least law and order. The tribes have been well organized, well led, cutting across religious and regional differences. They had well-tried forms of organization that had operated efficiently (and to an extent democratically) in previous times of weak state structures.

44In the Afghan context, ‘the tribes’ usually refers to the Pashtuns, who form up to 50% of the population (some say more, some say less) and at the same time have a similar or greater number of fellow tribespeople across the frontier in Pakistan. Other ethnic groups in the country have ‘tribal’ traditions, notably the Shi’i Hazaras and the Sunni Turkmens and Uzbeks of the North, while the main minority, the Tajiks, are non-tribal almost by definition.

45From 1747 until the 1970s, Afghanistan was ruled by Amirs and Shahs from the Durrani Pashtuns, who based their power on paying subsidies to tribal elders, especially on the frontier with India/Pakistan. If the tribes supported the government, they were left to run the villages their own way – this was broadly true in most of the country, where, despite an administrative structure of provinces and provincial and district governors, feudal lords, Pashtun and non-Pashtun, controlled local affairs.

  • 9 See for example B. Glatzer [1998] and A. Giustozzi [2005].

46The importance of ethnic and tribal identities in Afghan politics and society has undoubtedly increased since late 1970s. After the Soviet invasion, much of the countryside was controlled by resistance forces, until the victory of the mujaheddin in 1989; the subsequent disastrous civil war of the early 1990s led to the rise of the Taliban. During the years of revolution, invasion, occupation, resistance and civil war, both Afghan governments and foreign political analysts insisted that ethnic and tribal identity was no longer the key factor in determining political behaviour: rather, we should look to ideologies and programmes. Since the late 1990s ethnic analysis has returned to fashion. The strong identification of the Taliban with the Pashtun (especially Durrani) tribes, the frequent occurrence of apparent ‘ethnic’ massacres by and of the Taliban, and bitter conflicts between Uzbeks and Pashtuns, persuaded analysts to dust off the old maps and once again to discuss ethnic and tribal identities as determinants of political loyalties and conflicts. Since the US-led invasion in late 2001, political analysis and government and NGO reconstruction efforts have sought to take account of ethnic identities and tribal forms of organization. 9

47Thus, Scott Baldauf, one of the most perceptive journalists in Afghanistan at the time, wrote after the 2001 invasion by US-led forces:

  • 10 ’Feudal Lords Key to Afghan Peace’, Christian Science Monitor, December 21, 2001 (http://www.csmoni (...)

During two decades of war, in the absence of a functioning central government, its duties fell to ethnic or tribal leaders. Rural Afghans have long looked to such leaders to settle land disputes, punish crimes, and get their voices heard. 10

48The Taliban ranks were drawn from Durrani and other Pashtuns: but they were openly and ideologically ‘detribalized’ by fifteen years of civil war, exile, orphanage, and religious education. The Taliban saw the tribalism of the 1980s and 1990s mujaheddin groups as corruption and lawlessness, and sought to replace it with ‘pure’ tribal values, which they equated with Islam. A CIA officer who had worked with Afghan fighters during the 1980s was reported as saying:

  • 11 S. Baldauf, ‘Key to Governing Afghans: The Clans’, Christian Science Monitor, June 24, 2004 (http:/ (...)

[Local Taliban leaders] are tribal chiefs, who give themselves Islamist credentials for foreign consumption, but the real source of their power is their tribe [...] Their power does not extend beyond the influence of their tribe. 11

49When the Taliban leaders were removed, tribal government was resumed in much of the country, especially the Pashtun areas:

  • 12 S. Baldauf, ‘Feudal Lords Key to Afghan Peace’.

Some experts say this return to feudalism, where warlords gain power by exercising power – relying on weapons and pragmatism rather than ideologies or written laws – could endanger the fledgling government. But others argue that, at least at first, Afghanistan’s best opportunity for peace is to tap into a traditional infrastructure that may be unstable – even brutal – but works. 12

50In 2004 Baldauf wrote:

  • 13 S. Baldauf, ‘Key to Governing Afghans: The Clans’.

Two and a half years after the fall of the Taliban, the fight for control of Afghanistan continues tribe by tribe and village by village. It’s a battle of hearts and minds, where the enemy – Taliban and Al Qaeda – know the rules and nuances of tribal society better than the Americans, and perhaps better than some of the urbanized Afghan officials who now rule the country.
It’s a battle where alliances are made and broken over blood relationships and tribal feuds, rather than adherence to an extremist form of Islam preached by Osama bin Laden. The fluid power struggles present an opportunity for the US to cast its lot with tribal leaders to get intelligence and secure local protection for reconstruction projects. However, engaging in tribal politics also risks deepening feuds and undermining the country’s transition to a more modern, democratic system based on merit rather than blood.
‘Tribes are arguably more important than ever,’ says David Edwards, an anthropologist with extensive experience in Afghanistan based at Williams College in Massachusetts. ‘Given the fact that the present administration neither is very strong nor has a great deal of legitimacy, tribal structures have rebounded.’ 13

51In Iran, tribal affairs are very different. Although there have been recent reports of tribal and ethnic unrest in some remote parts of the country, such as Baluchistan and Kurdistan, the armed forces, notably the revolutionary guards, have in effect eliminated ethnic and tribal politics. Since the revolution the tribes have been officially redefined as cultural rather than political groups; as colourful pastoral nomads with kinship solidarity, rather than the political organizations led by powerful chiefs that they were until the 1920s [Tapper ed. 1983]. The situation now resembles that which obtained towards the end of the Pahlavi period, for example at the 1977 Festival of Popular Traditions in Isfahan, when the regime celebrated its defeat of the nomad tribes: tribal cultures were taken out of their social and especially political contexts and displayed in public as museum pieces – a ‘culture bazaar’, as one Iranian anthropologist described it [Shahshahani 1986]. Now – as then – the tribes are promoted as tourist attractions; one can find scores of ‘tribal tours’ websites, offering foreigners nights among the famous groups such as the Bakhtiari and the Qashqa’i.

52Assuming the establishment of strong, autonomous central governments, will this domestication of the tribes also be the future for Afghanistan and Iraq? A big assumption, perhaps. But what are the processes involved?

53In the premodern world, tribal groups in all these countries had contracts with their governments, with the state authorities. In return for a degree of autonomy, and possibly subsidy, they ‘protected’ main roads and trade routes; when strong, the tribes also ‘protected’ cities; the chiefs often settled there.

54Before the age of machine-guns, tanks and aircraft, tribal militias were the equal of anything the government could muster against them – indeed successful states employed tribal militias. The 20th century brought new military technologies which tipped the balance, it seemed for good, in favour of the modernizing state, leaving the tribes ‘backward’ technologically as well as culturally. But the Afghan resistance, with the Kalashnikov and the Stinger, restored something of the old balance; and they inspired tribal resistances in today’s Iraq.

55Generally, tribal forces have proved that they are still hard to beat on their own territory; but they are amenable to persuasion, to negotiations of loyalty by diplomatic outsiders. In the contemporary world of nation-states, whose priorities – whatever the effects of globalization – still appear to be state-formation, nation-building and national integration, tribal and other minority identities are clearly still perceived as at least potential threats to the state.

56Iran has dealt with these threats, first by violence (in the 1920s and 1930s), secondly by denial (in the 1950s and 1960s), and now by domestication (since the 1970s). That the domestication is not complete is shown by the preelection visit to Kurdistan in May 2009 by the Supreme Leader Ayatollah Khamane’i; in his speeches, while warning the (mainly Sunni) Kurdish tribes of ‘the plots of the enemies’, he was clearly seeking their support for his regime, and in particular for his favoured candidate for president, the incumbent Mahmoud Ahmadinejad.

57The condition of nation-states in the 21st century will be largely a function of the policies of the super-powers (USA, Europe, Russia, China, India). Where these fail to provide what people need: social, political and economic security, employment, services, cultural and political expression – and not necessarily ‘freedom’ and ‘democracy’ as defined in US terms – tribal and ethnic loyalties and organizations are likely to continue to meet these needs.

58I concluded my introduction to The Conflict of Tribe and State in Iran and Afghanistan: ‘Tribalism has its faults and limitations, but its provision of social security and its long-term survival value should recommend it as no anachronism in the last decades of the 20th century.’ [1983: 75] I suggest that the events of the first decade of the 21st century indicate that this was an accurate prognosis.

Haut de page


Baram, Amatzia — 1997, ‘Neotribalism in Iraq: Saddam Hussein’s Tribal Policies 1991-1996’, International Journal of Middle East Studies 29: 1-31.

Barth, Fredrik — 1961, Nomads of South Persia: The Basseri Tribe of the Khamseh Confederacy. Londres, Allen and Unwin.

Gellner, Ernest — 1983, ‘Tribal Society and Its Enemies’, in R. Tapper ed., The Conflict of Tribe and State in Iran and Afghanistan. Londres, Croom Helm: 436-448.

Giustozzi, Antonio — 2005, ‘The Ethnicization of an Afghan Faction: Junbesh-i-Milli from Its Origins to the Presidential Elections’. Crisis States Programme Working Paper no 67. Londres, Development Studies Institute, London School of Economics.

Glatzer, Bernt — 1998, ‘Is Afghanistan on the Brink of Ethnic and Tribal Disintegration?’, in W. Maley ed., Fundamentalism Reborn? Afghanistan and the Taliban. Londres, Hurst: 167-181.

Shahshahani, Soheila — 1986, ‘History of Anthropology in Iran’, Iranian Studies 19: 65-86.

Tapper, Richard — 1990, ‘Historians, Anthropologists and Tribespeople on Tribe and State Formation in the Middle East’, in P. Khoury and J. Kostiner eds., Tribes and State Formation in the Middle East. Berkeley, University of California Press: 48-73. — 1991 (1975), ‘The Tribes in 18th and 19th Century Iran’, in P. Avery, G. Hambly and C. Melville eds., From Nadir Shah to the Islamic Republic. Vol. 7 of The Cambridge History of Iran. Cambridge, Cambridge University Press: 506-541. — 1994, ‘Change, Cognition and Control: The Reconstruction of Nomadism in Iran’, in C. Hann ed., When History Accelerates. Londres, Athlone Press: 188-211. — 1997, Frontier Nomads of Iran: A Political and Social History of the Shahsevan. Cambridge, Cambridge University Press: 10-24. — 2002, ‘Introduction: The Nomads of Iran’, in R. Tapper and J. Thompson eds., The Nomadic Peoples of Iran. Londres, Azimuth: 10-39. — 2008, ‘Who Are the Kuchis? Nomad Self-Identities in Afghanistan’, Journal of the Royal Anthropological Institute 14: 97-116.

Tapper, Richard ed. — 1983, The Conflict of Tribe and State in Iran and Afghanistan. Londres, Croom Helm.

Yapp, Malcolm — 1983, ‘Tribes and States in the Khyber, 1838-1842’, in R. Tapper ed., The Conflict of Tribe and State in Iran and Afghanistan. Londres, Croom Helm: 150-191.

Haut de page


1 Other writers (such as myself), however, are unwilling to take either extreme position, and refer to these larger groups (whatever their apparent basis) as ‘confederacies’, locating ‘tribes’ at a lower level of political structure, that of first or second order components, numbering at most some thousands of individuals. Such tribes commonly (but still by no means always) combine territorial and political unity under a chief with an ideology of common descent.

2 This notion is held by numerous historians and other writers, who also assume tribes to be descent groups, often borrowing from anthropology the term ‘segmentary lineage’. See my comments [Tapper 1990].

3 See my ‘Introduction’ [Tapper ed. 1983: 1-82]. I have elaborated some of the main arguments in various publications [Tapper 1991], which despite its publication date, was the first draft of some of the argument, having been written and submitted [1975, 1994, 1997, 2002, 2008].

4 Amatzia Baram, New York Times, 28 October 2003.

5 Anonymous, Iran Daily, 20 October 2004.

6 Shoib Najafizada, Agence France Presse Report, 26 October 2004.

7 The same applies to the English terms ‘state’ and ‘government’, and their local equivalents (dowlat, hokumat).

8 See A. Baram [1997] ; see also R.L. Taylor, ‘Tribal Alliances: Ways, Means, and Ends to Successful Strategy’, Carlisle Papers in Security Strategy, Strategic Studies Institute, United States Army War College, 2005, available at http://www.strategicstudiesinstitute.

9 See for example B. Glatzer [1998] and A. Giustozzi [2005].

10 ’Feudal Lords Key to Afghan Peace’, Christian Science Monitor, December 21, 2001 (http://www.csmonitor.

11 S. Baldauf, ‘Key to Governing Afghans: The Clans’, Christian Science Monitor, June 24, 2004 (http://www.csmonitor.

12 S. Baldauf, ‘Feudal Lords Key to Afghan Peace’.

13 S. Baldauf, ‘Key to Governing Afghans: The Clans’.

Haut de page

Pour citer cet article

Référence papier

Richard Tapper, « Tribe and state in Iran and Afghanistan: an Update »Études rurales, 184 | 2009, 33-46.

Référence électronique

Richard Tapper, « Tribe and state in Iran and Afghanistan: an Update »Études rurales [En ligne], 184 | 2009, mis en ligne le 01 janvier 2011, consulté le 01 mars 2024. URL : ; DOI :

Haut de page

Droits d’auteur

Le texte et les autres éléments (illustrations, fichiers annexes importés), sont « Tous droits réservés », sauf mention contraire.

Haut de page
Search OpenEdition Search

You will be redirected to OpenEdition Search