For Western scholars of former Soviet Central Asia, dealing with the Soviet academic legacy can be challenging. Soviet anthropologists and historians had privileged access to the region for many years and produced a large body of research, much of it empirically valuable yet underpinned by primordialist and evolutionary assumptions that are no longer accepted by Western social scientists. Western historians unfamiliar with anthropological debates about the salience of “clan” and “tribe” may adopt too uncritically Soviet descriptions of social organization in the region. David Sneath’s essay serves as a useful reminder to be judicious in our use of Soviet ethnographic literature. Sneath also reminds us of the crucial role of states in making societies and peoples, an idea that was late to arrive in the field of Russian and Soviet studies.^[1] While it is now a commonplace that the Soviet state constructed nations on its territory, historians have paid less attention to the possible role of the tsarist imperial state in creating tribes and tribal confederations in Central Asia.
For all the value of his arguments, Sneath makes his case in far too sweeping a fashion. He claims that anthropologists have rejected the tribal model because it represents “primitive society” along an evolutionary scale and that terms such as “tribe” and “tribal chief” are applied exclusively to “backward” non-Western societies. Yet Sneath’s argument appears to rest mainly on a reading of works by Africanists. In sub-Saharan Africa (and in North America, though he draws less on the literature about Native Americans), a “tribe” may indeed have been what would have been called an ethnic group in Europe, but this is not the case among Bedouin Arabs, Berbers, or Turkmen. Anthropologists of the Middle East and North Africa have by no means abandoned the notion of the tribe in the sense of social and/or conceptual organization based on genealogy. The continued use of the tribal model is not due to a lack of reflection, as Sneath implies; on the contrary, scholars of the Middle East and North Africa have vigorously debated the meaning and significance of kinship-based social structures.^[2] In the Middle East, “tribe” does not imply a pre-state or proto-ethnic formation, nor has nomadism been seen as an “evolutionary dead end.” Rather, tribes and states – like nomadic and sedentary populations – have long been understood to coexist in dynamic interaction.
On the impact of colonial rule, Sneath also seeks to apply a model developed by Africanists that is of dubious relevance to Central Asia. Colonial authorities in the Middle East clearly did institutionalize tribes for their own purposes (the French in North Africa are a particularly good example of this), yet tribes in the region were not colonial creations in the same sense that they were in sub-Saharan Africa. While colonial states may have reinforced, undermined, and exploited genealogical systems in the Middle East, it would be a mistake to think that these modes of identification and affinity were exclusively imposed from outside.
What has been discredited in Middle Eastern anthropology is the structural-functionalist model of the segmentary lineage system, in which the “balanced opposition” of kin groups helps to regulate a stateless society. From the 1940s to the 1970s, a vast ethnographic literature on segmentary theory appeared, analyzing the social structure of groups ranging from the Bedouin of Libya and the Arabian peninsula to the Berbers of Morocco. Segmentary lineage theory came under attack from various quarters beginning in the 1970s, first by interpretative or symbolic anthropologists and subsequently by postcolonial and “reflexive” anthropologists. Critics argued that segmentary lineage theory was an ideal or idiom that was used by tribes to describe their social structure, but that did not actually determine their political behavior. In the view of these revisionist critics, structuralist-functionalist anthropologists were too quick to accept local informants’ descriptions of their social structure as reflecting the true state of affairs. In reality, social organization was far more flexible, with many factors other than kinship determining people’s social behavior.^[3]
Scholars have also noted that genealogies are often manipulated, with groups remembering – or inventing – only those ancestors who define a currently existing social or political group. The notion that pastoral nomads and tribal populations are fundamentally egalitarian has also come under fire. While egalitarianism may function as an ideology among such groups, in reality tribal societies often featured elaborate social hierarchies and stratification. For example, Abdallah Hammoudi has argued that Moroccan history abounds with examples of assimilation and integration of some groups by others, leading to relations of dependence and clientage. Where this has happened, genealogy is used not just to promote solidarity but also to facilitate stratification by distinguishing between descendants of the conquered and descendants of the conquerors. Among the Turkmen, communities were similarly stratified according to genealogical and historical criteria, with descendants of conquerors possessing the most prestige and a distinction made between “pure” Turkmen and those of “less pure” or partially “slave” ancestry.^[4]
While the static and mechanical view of tribal affiliation as a determinant of political action is defunct, the notion of genealogy as an important conceptual map and mode of thought remains useful in the Middle East and Central Asia. Among the contemporary Jordanian Bedouin, anthropologist Andrew Shryock has noted, the sense of community and the transmission of cultural information are both legitimized by genealogy.^[5] Historical knowledge is only considered authoritative if it comes through an unbroken chain of transmission from respected sources, preferably thorough face-to-face oral transmission. Shryock sees this as a legacy of the tribal origins of many Islamic societies – a “tendency to think of society as a framework of discrete human linkages of power and intellectual authority reproduced genealogically.”^[6] In tsarist-era Central Asia, similarly, Adeeb Khalid has shown that the knowledge preserved and conveyed in Islamic schools was only authoritative if it was obtained through interaction between student and scholar in a “chain of transmission going back to the author.” Texts in themselves were inadequate as conveyors of cultural information.^[7] More controversially, Shryock argues that the historical legacy of tribalism has resulted in divisive societies characterized by “contentious multivocality and genealogical opposition” – and that this pattern is true not only of Jordanian Bedouin but also of nontribal societies in the Middle East.^[8]
Sneath’s desire to avoid the pernicuous “conceptual apartheid” that can result from the use of terms such as tribe and clan is laudable. His telling account of the substitution of the term “chief” for “duke” in descriptions of Mongol society underscores the reluctance of Westerners to apply the categories of European history to “uncivilized” Inner Asians and other non-Europeans. Yet Sneath’s contention that kinship-based forms of social organization were invented ex post facto by anthropologists and colonial administrators is simply not supported by a reading of the scholarly literature on the Middle East and Central Asia. I do not wish to belabor this point, since others have addressed it, but I will mention just one of the firsthand accounts of kinship-based social systems with which I am familiar, namely, William Irons’s ethnography of the Yomut Turkmen in Northern Iran carried out in the 1960s. Unless one believes that Irons simply invented the entire book, or that his informants constructed a highly detailed fictitious social system for his benefit, it is obvious that genealogy is an important organizing principle and conceptual framework – though not the only significant factor – in the social, political, and economic life of the population he studied.^[9]
Moreover, to insist that kinship structures have no indigenous roots or significance deprives scholars of a useful tool for understanding identities and societies in the Middle East and Central Asia. In Central Asia, the nature of the “nations” that emerged under Soviet rule can only be partially understood if we discount this aspect of the region’s history and culture. I have argued elsewhere that the Turkmen “nation” may be viewed as a dynamic interaction between indigenous, genealogical conceptions of identity and externally imposed Soviet ideas of nationality.^[10] Turkmenistan is not the only post-Soviet state to draw on genealogical traditions and myths to legitimate its national statehood; Kazakhstan and Kyrgyzstan do so as well. The fact that the European (and hence Soviet) understanding of nationhood was rooted at least partially in notions of ethnicity based on common descent made the ideology of nationalism all the more palatable to kinship-minded Central Asians. The modern nations of post-Soviet Central Asia arose out of the synthesis of Soviet understandings of identity (which became increasingly primordialist and focused on ethnogenesis in the final decades of the Soviet era) and indigenous understandings of identity rooted partly in genealogy.