The discussion published in Etnograficheskoe obozrenie (issue 1/2005) on the Ethnic Atlas of Uzbekistan in general, and Alisher Il’khamov’s “Archeology of Uzbek Identity” in particular, reveals a sharp scholarly divide. By interpreting the formation of Uzbek identity through a theoretical perspective that is familiar to Western scholars, while drawing on ethnographic research that is familiar to those whose scholarly legacy is Soviet, Il’khamov has laid the groundwork for a complex and difficult scholarly conversation. As Il’khamov notes in his own response to commentary, his work draws criticism from two sides: many of the Western scholars argue that social constructivism is built on deconstruction, and that Il’khamov has not engaged in enough deconstruction or exercised enough skepticism about the “reality” of Uzbek identity, while his Soviet-trained critics see his work as utterly rejecting the findings of decades of scholarship on Uzbek ethnogenesis.
One of the main complaints coming from the Uzbek scholarly establishment about this work, the Ethnographic Atlas of Uzbekistan, concerns its claims to authority. Dilarom Alimova, Zoia Arifkhanova, and Shamsutdin Kamoliddin, all associated with Uzbekistan’s Academy of Sciences, charged in a published critique that the collection did not draw on available scholarship, did not represent the understandings of Uzbekistan’s experts, and should not have presented itself as an authoritative piece of scholarship.^[1]
Regardless of my own opinion about Il’khamov’s social constructivist approach to Uzbek ethnic identity, I believe that freedom of expression is essential both to the development of democratic society and to the vitality and progress of scholarship. When parties to an intellectual dispute call into question not the ideas, but the character of their opponents, then scholarship itself is smothered. And when one party to a scholarly dispute is forced to seek refuge abroad (in this case, Il’khamov), that creates an atmosphere of fear, so that scholars on all sides are tempted to censor their own ideas. In particular, one critic charged Il’khamov with tailoring his approach to the demands of an international funder. I would urge all scholars to recognize that we all take money (salaries, grants) from government entities and from other funders, and that all of us can equally be accused of serving a master rather than thinking and writing independently. I hope that, nonetheless, our primary commitment is to the pursuit of knowledge and to open scholarly debate.
The Ethnic Atlas itself is an uneven work. The first section, a collection of brief articles about “ethnic minorities” in Uzbekistan, attracts attention for two reasons: first, its recasting of what were called “nationalities” in Soviet publications as “ethnic minorities,” and second, the lack of comparability among the articles themselves. The designation of “ethnic minorities” in the first section seems to be driven by the conventions of Soviet census taking. The Soviet census enumerated members of “nationalities” present in each union republic, listing them in alphabetical order, or grouping them according to linguistic similarities. Hence, the handful of British, French, or American born immigrants to Soviet Uzbekistan showed up in the nationalities list among the hundred or more other national identities, such as Tatars and Ukrainians. Without really defining the term “ethnic group,” the Atlas follows the list of Soviet nationalities, and includes rather interesting discussions about why there were Austrians in Turkestan, for example. But is there actually any ethnic group to be discussed if there are a handful of descendents of Hungarian prisoners of war in Uzbekistan, or if a few French made Uzbekistan their home? The editor’s choice to follow the national identity charts of census takers in deciding which “ethnic” groups to include makes it difficult for the reader to differentiate between the Soviet concept of nationality and the sociological concept of “ethnic group” or “ethnic minority,” which carries numerous implications in international relations. In addition, the authors of these articles seem to have been given no template, so that the kind of information that shows up in the article about Koreans, concerning food and family, does not appear at all in other articles, which instead concentrate on particular historical moments or roles in society.
As others have noted, the contrast between the first section and the second, on Uzbek identity, is striking. In the first section, the only article that mirrors Il’khamov’s social constuctivist approach is Peter Finke’s article on Tajiks. The rest of the articles in the first section seem to take ethnic identities, or perhaps nationalities, as objective categories of analysis, and proceed from the assumption that these identities are fixed. Il’khamov’s effort to examine the archeology of Uzbek identity (employing Foucault’s metaphorical use of the term archeology, to mean an examination of layers of knowledge, Il’khamov’s critics seem oblivious to this distinction and thus condemn him for not studying archeology per se) thus poses a striking contrast to the first section of the Atlas. The decision to follow a highly theoretical piece on Uzbek identity with two lengthy articles on Uzbek language and on ulak^[2] again adds to the inconsistency of the volume as a whole; not only do the first section articles lack any parallelism, but the third section seems whimsical, not closely related to the brief descriptive articles in section one or to the theoretical approach of section two. Had Il’khamov entitled this work something like “Ethnic Groups in Uzbekistan,” and presented it as a series of vaguely connected articles all dealing with one theme (like so many published scholarly collections of articles), then perhaps his critics would have less to complain about. It is indeed the word “atlas” that gives the reader the expectation of consistency in scholarly approach, and that leaves the reader dissatisfied.
But Il’khamov’s critics also charge him with ignoring scholarly work on Uzbek ethnogenesis. This charge arises from their mistaken literal reading of the term “archeology” in Il’khamov’s title, and from confusion about the definitions of ethnicity, nation, ethnic group, as well as a dispute over theory. The theory dispute pits the ethnogenesis school that developed in Soviet scholarship against the social constructionist approaches that have become predominant in Western scholarship. (A scan of the literature on ethnogenesis shows that there are endless articles on the ethnogenesis of particular peoples produced by scholars in Eastern Europe and the former Soviet Union, and almost none by scholars from any other place.)
Anthropologist Bruce Kohl, in reviewing the history of Soviet anthropology and archeology, observes that the focus on ethnogenesis “became one of the central tasks of Soviet archaeology when the discipline switched from a Marxist inspired internationalism... to one concerned principally with the ethnogenetic history of the early Slavs, i.e., when Great Russian chauvinism and the buildup to the Great Patriotic War replaced internationalism.” This gave every nationality within the USSR an “ethnogenetic mandate or search for its origins. Competition over the remote past was intimately tied to the very structure of Soviet administration.” In the Soviet school, ethnogenesis was a determination of when an ethnic group marked by certain cultural traits came into being, and that is a question that can be answered with objective reference to particular documents or artifacts from the remote past. By contrast, Kohl notes, Western anthropologists view ethnicity as group self-identification, something that simply cannot be ascertained for the past masses, and in addition, “ethnic groups are malleable and constantly changing as the historical situation in which they exist unfolds; ethnicity, like culture, is never made but is always ‘in the making.’”^[3] Hence, Western studies of ethnicity are more concerned with ethnomorphosis (changing identities) than with ethnogenesis.
The Soviet anthropological concept of ethnogenesis formed in a context cut off from interaction with Western scholarship in the 20th century (social constructivism is not the only concept foreign to Soviet trained scholars). The theory of structural-functionalism, still influential in Western social sciences, is also absent in Soviet scholarship.^[4] Social constructivism was to a certain extent a response to the static analysis of structural-functionalism, the latter a theory that is useful for explaining social relationships, but not good at explaining change. Soviet ethnogenesis theory took its own path, and it is hardly a wonder that its practitioners find constructivism so foreign; they would find its predecessors equally foreign.
Building on Gramsci’s understanding of hegemony, as well as Foucault’s call for an archeology of knowledge (a deconstruction that allows the researcher to see the accumulated layers of verities that present to us a seemingly complete and unquestionable reality), Edward Said argued that Orientalism, or the study and objectification of “the East” is “a discourse... that is produced and exists in an uneven exchange with various kinds of power, …political, …intellectual… cultural... moral.”^[5] For followers of the social construction school, what we observe, and the way that we build our knowledge of the world, is inextricably linked to power, and there is no knowledge that is simply a neutral representation of reality.
Deconstruction, followed to the extremes that Derrida suggested, leads to a dead end, quite as much as Soviet ethnogenesis theory does, in that the deconstruction school ultimately denies that there is a “real” subject that can be known. The social constructionists run into the same problem; in the struggle against positivism and essentialism, they argue that what is knowable is not “das ding an sich,” but representations of it – and the politics of those representations.
Il’khamov defends his work by rejecting “fundamentalism,” the extreme logic of either essentialist or social constructionist positions. He claims, and I agree, that there is a Uzbek nation whose development and identity can be studied; approaching that study with the methods of social constructivism can be productive only if the researcher builds on assumptions that there are facts that can be known (however contested and partial that knowledge is), and that there exists a collectivity of people whose common identity, however formed, is significant to them and to others. And yet Il’khamov’s argument that such study shows us that knowledge is indeed political, and that the construction of a nation is a political and not an objective-natural process, is disputed by some of our colleagues in Russia and Central Eurasia, a response that proves the point.
As others have noted, the ethnogenesis of Uzbeks described by several generations of Soviet scholars (whatever their differences over the relative significance of the Shaybanid invasion) is teleological. According to the ethnogenesis school, people who lived in Central Asia centuries ago can be described as Uzbeks, and their language as “old Uzbek” because they and their descendents are linked by blood, culture, and destiny. Modern claims can be made to ancient ancestors. Those who inhabited the soil of Mawarannahr are the ancestors of those who live there today, i.e. those ancients are proto-Uzbeks. If some of them spoke a Turkic language, that is simply more proof of their identity and connection to present day Uzbeks. Similar stories make sense of the past throughout the world, but constructivism points out that narrative coherence should not be mistaken for a careful presentation of complex evidence from the past. The historian Prasenjit Duara shows that the dominant accounts of Chinese history absorb narratives about the more successful kingdoms that flourished in East Asia, joining them in a fictive sequence as pearls on a thread, while ignoring the stories of the less successful polities that existed on the same soil and contended against the winning kingdoms. An orderly, linear past, one that leads to contemporary greatness, is always the creation of those who in modern times seek to establish a unitary and exclusive heritage for their modern nation.^[6]
Let me take the liberty of proposing a counterfactual argument, one that has nothing to do with what really happened, but can clarify thinking. What if, instead of invading Mawarannahr, Shaybani and his tribe (who claimed Uzbek Khan as their founder) had invaded and conquered Siberia? Today Tuvans, Sakha, and others might be calling themselves Uzbeks, while Russia might never have conquered Siberia. Mawrannahr might, in turn, have continued under the political control of Timurids, or might have come under the control of Safavids and become more heavily Shia and Persianate. In any case, no one in this part of Central Asia would now call themselves Uzbeks, and no one would now teleologically refer to their pre-16th century forebears as proto-Uzbeks, or their language as early Uzbek. But a group associated with the name Uzbek did invade and that group seized politically dominant positions. The invading group was relatively small. Both Soviet ethnography and Il’khamov offer explanations, grounded in the same scholarship but with different theoretical framing, for why there is now an Uzbek nation. The Soviet ethnogenesis school sees a natural process, while Il’khamov argues for a political one, with the early Soviet state as the most important factor in consolidating a variety of groups under the Uzbek name.
Structural functionalism in anthropology, a school of analysis that was based on the study of tribes, raises questions about the reproduction and stability of social structures and argues that structures continue because they serve social purposes and fulfill needs. Much of Il’khamov’s analysis concerns the demise of tribal identities and the rising importance of ethnic identities in Turkestan. Il’khamov offers a critical examination of Sart, and addresses why Sart did not become the name of the new national identity in the early 1920s, even though it was (arguably, and with ambiguous data) the predominant identification of sedentary Turkic speakers in many regions. But why was “Uzbek” a widely accepted alternative? In order to answer this, we might first consider the purposes that ethnic identity serves.
Il’khamov, following Russian colonial period ethnographers and government agents (many of the ethnographers in the Russian colonial period were Russian government agents, paid to collect and construct knowledge for the purposes of extending state power), discusses the seeming proliferation and then diminishing of nomadic tribal names. The basis for tribal cooperation was not that various tribes sensed a common ethnicity; the basis of tribal cooperation was loyalty to a leader, largely because the leader could enable tribe members to obtain the goods that they wanted and needed. When a confederation leader or a tribe leader could no longer provide, his leadership ended. In the pre-colonial period, small tribes played important roles in the Central Asian polity. An emir or khan could expand the territory under his rule only by relying on the military strength provided by tribes, and in return the ruler gave a loyal tribe the right to certain lands, to grazing places, and to revenues gained from dominating sedentary peoples (taxes, trade, loot). Members of the tribe benefited from membership; their participation in tribal military action brought them rewards and basic sustenance.^[7] But what became of these tribal functions when Russia’s armies invaded and conquered Central Asia? They met their demise. No longer needed for the military support of emirs or khans, tribes no longer had much purpose, and membership in a tribal organization no longer was the primary provision of livelihood. Well before the 1917 revolution, tribalism lost most of its significance, and although many people could still name their tribal affiliation, it would not have been a dominant factor in ordering life or the basis for an emerging ethnic identity.
Structural functionalism helps to explain the demise of tribal identity in Turkestan and Bukhara, but does it also help to explain why “Uzbek,” rather than “Sart,” became a much larger group identity? “Uzbek” was a name associated with political power after the 16th century. By the 19th century some nomads used Uzbek as their larger group name (beyond their tribe), but many sedentary people also identified as Uzbeks (explained perhaps by intermarriage and other forms of alliance making). Perhaps most important, those who formed the military-political elite (as opposed to the cultural elite Sarts whom Il’khamov discusses) were Uzbeks. Uzbek emirs and khans gave land grants to Uzbek beks, who settled and become wealthy and powerful leaders of communities. After the Russian conquest, the Russian colonial administrators used and favored some of these elite Uzbeks, thus allowing them to continue building their social capital even under a changing system. The name Uzbek was far more closely associated with political power than the name Sart; is it then surprising that Turkestanis did not fight to continue to be identified as Sart, when they were given the opportunity to call themselves Uzbek?
In exploring this line of argumentation, I place myself in the social constructionist camp; identities are malleable, changing, subject to circumstance, and are formed and reformed in opposition to others. But I do not think they are formed from nothing, or that they are infinitely malleable; the real, lived past, with several millennia of societal formation in Central Asia, is an important and inescapable part of the picture of identity formation. But at what historical junctures is it appropriate to pose questions about ethnicity and nation? Il’khamov’s critics charge that his choice to begin the discussion in the 16th century is a choice far too late, and that Uzbek identity, though not necessarily under that name, can be traced much further back. Il’khamov, they say, should have read archeological studies. To be sure, there are abundant scholarly studies of archeology in Uzbekistan, but archeologist Bruce Trigger warns:

“While often wrongly equated solely with language or culture, ethnicity refers specifically to group self-identification, which is not subject to direct archeological observation... Historically or ethnographically established ethnicity can be securely projected backward in the archeological record only when cultural (and ideally also physical) continuity can be demonstrated not merely at the regional, but more specifically at the local and community levels.”^[8]

While there is evidence that one of the written languages of pre-16th century is closely linked to present day Uzbek, that does not mean that we should project Uzbek ethnic identity into the past, into periods in which we cannot know what forms of identity were important to people, and in which we can be fairly certain that factors other than ethnicity formed the basis for their social and political interactions. It is quite clear that in the dynastic and tribal states and empires of pre-modern Central Asia, there were numerous linguistic, cultural, and tribal groups, but ethnicity played little role in interactions. If we think of Jews in pre-modern Central Asia as an ethnic group (rather than a religious group), we then see distinction and discrimination, but the basis of distinction and discrimination was religious; Islam, the religion supported by the state, prescribed certain laws and attitudes toward Jews. But the state’s basis for favor toward other groups was instrumentalist, not ethnic; a tribe that could provide support to the emir or khan would be favored, regardless of characteristics that we now call ethnicity.
Il’khamov focuses his discussion of Uzbek ethnicity on a much more recent time, the Russian colonial and early Soviet periods. In a structuralist analysis, as well as in a constructionist analysis, we need to pay attention to the purposes that ethnicity served, and these changed with colonialism. The Russian state, after conquering Kokand, Bukhara, and Khiva, put in place institutions and social dynamics that gave political and individual importance to ethnicity. Ethnicity was not a primary organizing factor under the khanates; for Russia ethnicity was important as a political tool. This changing significance of ethnicity in Russia was linked to the Russian state’s changing views on nationalism; in the post-reform return to autocracy under Alexander II, the Russian emperor saw a rising Russian nationalism as a base of support for autocratic rule against both democratic and other nationalist tendencies. After 1881, policies favoring Russians over the Empire’s other subjects (Poles, Jews, Ukrainians, etc.) gave new individual and group importance to these forms of identity (ethnic or national).^[9] In colonial Turkestan, the indigenous groups faced a new “other,” and identity formation under Russian rule tended toward emphasizing factors that separated indigenous people from Russians: self-identification was often expressed as “Muslim” – hardly a meaningful distinction until the arrival of a large group of non-Muslims, and certainly not an ethnic identity.
Russian and non-Russian colonists in Turkestan were citizens of the Russian Empire; the indigenous peoples of Turkestan, Bukhara, and Khiva were not granted citizenship in the Empire (with a few individual exceptions). The colonial administration in Central Asia differentiated among peoples so that citizens of Russia (whether they were Tatars, Russians, Ukrainians, or Germans) were subject to one set of institutions (courts, taxes, schools) while indigenous peoples were subject to a different set. Although the Russian state enumerated the inhabitants of Turkestan, categorizing them by names that may be seen as ethnic group names, the differing indigenous categories (named in the 1897 census as Uzbek, Tajik, Sart, Kara-Kirgiz, and Turko-Tatar, with the final group possibly also including immigrants) did not mean different access to state resources and institutions; all of these groups experienced the same non-citizen status. In sedentary communities, the indigenous people who were likely to be granted roles in the colonial system of control were the local political-military elite, the ming-boshi and the bek; my own guess is that these people would have associated themselves with Uzbek rather than Sart group identity. While Sarts, as Il’khamov argues, had cultural resources, it may be posited that the status of the name Uzbek grew under Russian domination as at least some Uzbeks managed to maintain, and perhaps increase, their social positions under colonial rule. The Jadids argued about identity terms in the pre-revolutionary period, but after the revolution, when the Soviet Nationalities Commissariat determined to call sedentary Turkic speakers in Central Asia all by the name Uzbek, there is little evidence that anyone newly named Uzbek (from Sart or Turk) strongly rejected their new identification. Perhaps their acceptance of the name was connected to the ongoing association between “Uzbek” and political-military leadership.
However, identity in the national republican structure of the Soviet Union carried different implications than it had in the Russian Empire. Belonging to the titular nationality in a union republic meant better access to jobs, for example. Under the policy of korenizatsia in the 1920s, Uzbeks were promoted “from below” to positions in government enterprises and the Party. This policy went some way toward reducing the heavy overrepresentation of Russians in Uzbekistan and allowing Uzbekistan to be a republic that at least partially belonged to its indigenous people, but it also meant that it was more advantageous to be an Uzbek in Uzbekistan than to be Tajik or Kyrgyz in Uzbekistan, and more advantageous to recognize oneself as Uzbek than to insist on being seen as Sart (a category that had no official recognition and thus provided no access to public goods). To point this out is to highlight that while ethnicity in the Soviet context built on elements of earlier identities, the Soviet state reconstructed those ethnicities by giving them new purposes.
What I have outlined here concerning Uzbek identity is far more speculative than what Il’khamov proposed concerning Sart identity; both structural-functionalism and social constructivism serve to raise theoretical questions about the changing importance of Uzbek identity (and hence, who identified themselves as Uzbek and why) in the early 20th century.