It is because subjects do not, strictly speaking, know what they are doing that what they do has more meaning than they know.

Pierre Bourdieu^[1]

In official Soviet parlance, the USSR was a multinational federation, not a nation-state or empire. The Soviet people were, supposedly, a historically unprecedented kind of “community,” but never a “nation.” Yet these discursive realities should not cause us to overlook fundamental features of the Soviet project that seem “national.” Nor, for that matter, should its official anti-colonial rhetoric mean that the Soviet Union might not appear to be a variety of empire. Indeed, much recent work on the history of the Soviet state purports to reveal its imperial or colonial dimension. This development is not entirely new. An older, Cold War era literature labeled the Soviet Union an empire as an expression of political condemnation.^[2] If that older tradition implicitly wrote from the standpoint of non-Russian nationalisms, the newer literature has taken “modernist” suspicions of nations and nationalism to heart.^[3] Some of it operates in the tradition of comparative historical sociology, juxtaposing the Soviet experience to that of dynastic multinational states such as the Romanov, Ottoman, and Habsburg.^[4] Other work is informed by post-colonial theory, explicitly or implicitly comparing the Soviet experience to that of European colonial empires of the late 19th and early 20th century.^[5] This article is a critique of the growing body of literature on the “Soviet empire.” I argue that viewing the USSR primarily through an imperial or colonial lens blinds us to critical features of the USSR as a multiethnic state and society. Rather, we need to take the effort by the Soviet state to craft a culturally uniform Soviet national community seriously and should not overemphasize similarities between the Soviet Union and its imperial predecessors or colonial contemporaries. My essay has five sections. First, I propose a way of distinguishing between empire and national state that emphasizes the practices of rulers and seeks to move beyond objective definitions and subjective perceptions. Then, I apply these concepts briefly to the Romanov Empire, principally for purposes of contrast with the Soviet experience. I then describe the two principal aspects of the Soviet approach to cultural diversity – indigenization and Sovietization – so as to highlight the distinctiveness of the Soviet experience. Finally, I conclude with some reflections on the consequences of my approach for understanding the Soviet collapse.

EMPIRES AND NATIONAL STATES

What should we mean by “empire”? Comparative historical research on empire has tended to emphasize objective, structural relationships. In his book Empires, Michael Doyle defines empire as “a relationship of political control imposed by some political societies over the effective sovereignty of other political societies.”^[6] Ronald Suny likewise describes empire as “a particular form of domination or control between two units set apart in a hierarchical, inequitable relationship, more precisely a composite state in which a metropole dominates a periphery to the disadvantage of the periphery.”^[7] As Mark Beissinger has argued, however, such objective approaches overlook the fact that the very use of the term empire is “a claim and a stance.” He proposes that “The most important dimension of any imperial situation is perception: whether politics and policies are accepted as ‘ours’ or rejected as ‘theirs.’”^[8] Terry Martin, too, has called for historians of the Soviet Union to adopt a “subjective approach to empire.”^[9]
Both of these approaches, the objective and the subjective, however, can easily apply to practically every state in modern history, almost all of which have had a variety of “territories” or “peoples” under a single political sovereignty. Indeed, most contemporary states are characterized by metropole and periphery, by regional economic and political inequality, and by cultural difference; many suffer from the perception of inequality, oppression, or alien rule in some part of their territory.^[10] That is, neither objective structures nor subjective perceptions reveal significant distinctions among states with diverse populations and territories. We can grasp the distinctiveness of empire – and thus its utility as a concept – and move beyond the dichotomy between objective structures and subjective perceptions by focusing on what rulers do. And, as Pierre Bourdieu emphasizes, the subjects of historical investigation often do not really know what they are doing. This applies equally to rulers and ruled, the powerful and the powerless, and the colonizers and the colonized. A focus on practices helps us to understand nation and empire as a range of possibilities rather than as static relationships, on the one hand, or the subjective perception of the oppressed, on the other.
As rulers confront the cultural diversity that generates senses of “ours” and “theirs” among the ruled, they implicitly or explicitly orient themselves to difference and similarity; they develop an interest in maintaining one or the other. With this idea in mind, I offer ideal types of empire and national state based on the practices of their rulers.^[11] The rulers of empire, an imperial elite, maintain their power – political, social, cultural, and/or economic – by preserving or inventing differences among the people they rule. Assimilation is not a characteristic policy of empires, which tend to be conservative and inclusive. In an empire, there is rarely a clear distinction between “us” and “them,” because all human beings are, potentially, the “us” to be included in the imperial community. Ultimately, the power of imperial elites depends on the maintenance of political stability, for which the apparent preservation of traditional authority is critical.^[12]
The ruling elite of a national state, by contrast, maintains its power by promoting or creating sameness among its subject populations and by bounding that sameness through conscious policies of exclusion of aliens, foreigners, or enemies. Whether on “civic” or “ethnic” criteria, the national state typically distinguishes among its population using binary dichotomies of “us” and “them.”^[13] This bounding is engineered by practices of assimilation and discrimination. Just as empires are inclusive and conservative, national states are exclusionary and transformational, consciously manipulating and managing their subject populations. They do so because creating a “common conceptual currency” and a culturally standardized population facilitates modernization and resource extraction in political conditions where human beings expect economic growth and popular sovereignty.^[14] In both empires and national states, of course, rulers’ practices are generated by an interest in maintaining power, but that interest is mediated by ideas and discourses about difference and similarity. Political and economic liberalism, notions of popular sovereignty, and ideas of progress, for example, have all acted like “switchmen” (to borrow Max Weber’s metaphor) and determined a new track along which rulers’ interests would be pushed: from imperial differentiation to national assimilation, from empire to national state.^[15] That is, modern political and economic ideas brought with them new notions of the political and economic consequences of difference and similarity among human beings; cultural diversity, in short, became a problem to be solved.
Recognizing the distinction between imperial and national practices requires going beyond superficial similarities of techniques of rule and focusing on the goal-orientation of practices. The fact that a technique originated in a colonial or imperial setting does not make it “imperial” or “colonial.” The use of ethnic or tribal census categories for classifying populations, to take one example, frequently migrated back and forth between metropole and colony.^[16] Likewise, Eugen Weber wrote many years ago that the nation-making practices of the Third Republic, which had taken on a “white man’s burden of Francophony” in the periphery of France, were “akin to colonization.” Parisian officials compared that periphery to their overseas colonies. This “colonial empire” ultimately became a nation because, “given time and skins of the same color, assimilation worked.”^[17] Weber’s analogy vitiated fundamental differences between national and colonial practices because it equated the metropolitan periphery with the colony. What mattered, ultimately, was what those officials did in Brittany as compared to Vietnam. On the periphery of France, they sought to make people French; in Vietnam they were trying to keep people Vietnamese. By the 1930s, Herman Lebovics writes, there was a “remarkable intensification of the trend to Vietnamize Vietnamese education” to the point that French was no longer taught in primary schools of the colony by 1936.^[18] That is, cultural policy in Brittany was national (inclusive and transformative), while that in Vietnam was imperial (differentiating and conservative).
This brief consideration of French policy in Vietnam indicates that a single ideal type of empire will not suffice and that we need to distinguish between traditional dynastic empires (“empires of nations,” to borrow Francine Hirsch’s metaphor) and modern colonial empires (“empires of a nation”).^[19] The former, such as the Habsburg, Ottoman, and Romanov, operated for centuries with a tradition of rule-by-difference. Theirs was a pragmatic strategy of elite cooptation and indirect rule. Rulers of such states did not see cultural diversity as a problem to be solved, at least not until political and economic developments – the spread of liberalism and nationalism – in the nineteenth century made their practices obsolete.^[20] Ottoman strategies of Ottomanism and Pan-Islamism and Romanov policies of “Russification” – that is, the nascent introduction of national practices – were responses to that obsolescence.
Colonial empires, by contrast, are inconceivable apart from the nation to which they belonged. For these states, “‘nation-building’ and ‘empire-building’ were mutually constitutive projects.”^[21] In such empires of a nation, the power of ruling elites depended upon colonial practices of racial and ethnic differentiation, designed to distinguish the national community from the colonized and the colonized from one another. Such practices limited the universalizing promise of the liberal national state. The construction of the colonial state as a “juridically distinct sphere, clearly separated from the domestic realm” was directly connected to the emergence of national citizenship in the metropole, which required a clear and stark legal differentiation from colonized populations.^[22] Colonial practices of ethnic and racial differentiation served to demonstrate why economic and political sovereignty were appropriate for the colonizer but not the colonized. This was a constant process whereby “the most basic tension of empire” – the necessary limit to liberal universality – was worked through by the colonizer as the colonized “refashioned and contested European claims to superiority.”^[23] Authority in the colonial empire was legitimized in modern terms in the metropole and in traditional terms in the periphery.
This last point is what differentiates the national state from the colonial empire. Both are fundamentally modern inasmuch as they share the recognition that cultural diversity is a political problem. Both, importantly, rely on scientific knowledge about that diversity to rule diverse populations. Yet they offer fundamentally different solutions to that diversity. The colonial state, Partha Chatterjee writes, “was destined never to fulfill the normalizing mission of the modern state because the premise of its power was a rule of colonial difference, namely the preservation of the alienness of the ruling group.”^[24] Colonial difference, in Mahmood Mamdani’s words, served to “stabilize racial domination” by “ground[ing] it in a politically enforced system of ethnic pluralism (institutional segregation), so that everyone, victims no less than beneficiaries, may appear as minorities.”^[25] The rulers of the traditional multiethnic empire based their practices on the cultural and legal traditions they found among its various subject peoples, holding them to a universal particularity. The colonial empire, by contrast, juxtaposed metropolitan universality to colonial particularity because “traditional” differences worked for maintaining its domination in the periphery in the context of the universalism it upheld in the metropole. The universalism of the traditional empire, designed outside the context of modern political discourses, needed no boundaries. In the context of such discourses, colonial empires did and provided them by attention to difference itself.
Subjectively glimpsed from the periphery, colonial and national practices may seem similar, however, because both result in cultural and political domination of “them” over “us.” Both abandon the pragmatic traditions of the “empire of nations” and generate senses of “ours” and “theirs.” But they look very different from the (also subjective) standpoint of the center, precisely because of the differing orientation of state practices – to make people more alike or to keep them different. And those orientations are ultimately inscribed in the very practices that rule the periphery – for example, in the judiciary, in school curricula, military conscription, police procedure, and in local administration. We cannot fully understand the perceptions of those on the periphery, the extent to which a kind of rule is perceived as “alien,” without attention to the practices of rulers which frame those perceptions. To refer back to my earlier example, early-twentieth-century Bretons and Vietnamese were faced with very different choices by their rulers. For Bretons, the choice was to become French or, perhaps, to attempt to secede. For Vietnamese, the choice was to attempt to secede or to remain subjugated. Becoming French was not an option for them, precisely because of the orientation of the metropolitan center. In this respect, metropolitan practices matter in the periphery because they limit the periphery’s perceptions and thus potential historical outcomes. A focus on the periphery alone does not allow us to see all those possibilities. A brief examination of the Romanov imperial state in the late nineteenth century illustrates these ideal types of empire and national state and demonstrates how they help us understand its efforts at managing cultural diversity.

IMPERIAL RUSSIA: FROM EMPIRE TO NATION

Russia’s imperial practices from the conquest of the Khanate of Kazan in the sixteenth century through the 1830s, when the regime first proclaimed a program of “Official Nationality,” were based on a pragmatic tradition of elite co-optation and administrative pluralism.^[26] In the east this was done by claiming the mantle of the Golden Horde and continuing a tradition of “Steppe diplomacy”; in the west this entailed confirming the authority of the estates. Intermittent efforts at administrative rationalization in the eighteenth century, inspired by cameralism and enlightenment doctrines, consistently faltered on the shoals of this pragmatism, because the concern for stability outweighed the perceived need for change. Such policies did not seek to culturally transform entire populations, in any event.^[27] Imperial pragmatism gradually met its end between the 1830s and the 1860s, which saw both the Great Reforms and two Polish revolts, as the state grappled with the impact of nationalism and liberalism at home and abroad.
When the Romanov imperial elite turned to practices that promote sameness, they did so in response to foreign and domestic threats.^[28] The Tsarist regime’s policies of cultural Russification in the late imperial period followed from its understanding that cultural diversity was now a problem to be solved.^[29] In the Western regions, it sought to integrate some groups (e.g., Ukrainians, Belorussians) into a Russian nation; others (e.g., Poles, Jews) were excluded by national policies of discrimination.^[30] In the Volga region and the Urals, too, incorporation of peoples into a Russian nation was the state’s eventual goal; there it adopted the “Il’minskii system” of native-language instruction in order to assimilate baptized and “pagan” non-Russians to a religiously-defined Russian nation. It mandated that traditional Muslim schools incorporate Russian-language instruction.^[31] New policies were accompanied by an official re-envisioning of the empire and the dynasty as culturally Russian, rather than foreign or European, in the reigns of Alexander III and Nicholas II.^[32]
Insofar as policies of Russification in the western borderlands were not matched in Central Asia or the Caucasus, however, Imperial Russia seems to have lacked a coherent nationality policy. Practices in Central Asia, and to a lesser extent in the Caucasus, sought to enforce boundaries between colonizer and colonized.^[33] Il’minskii’s system, for example, was never introduced into Turkestan; neither were Russian law or administration. Late imperial Russia’s center thus ruled through both national and colonial practices. The goal in the Western borderlands and the Volga region was to assimilate and to repress non-Russian national aspirations and thereby create a kind of nation. The goal in Central Asia and the Caucasus was to differentiate and thereby build a modern colonial empire. Both of these tendencies were part of a transition away from the traditional, pragmatic practices of differentiation that characterized imperial policy from the sixteenth to the mid-nineteenth centuries.^[34] There remained ambiguity about where some parts of the empire might fit into this new schema; there was certainly disagreement as to what would constitute the Russianness at the base of the new national state – language, religion, and citizenship were all possibilities.^[35] Recent work shows how military intellectuals and orientalists both were beginning to imagine something like an all-Russian nation based on citizenship in the late imperial period.^[36] But such debates should not distract us from the overall direction of ruling practices.
In this respect, the assertion that Imperial Russia was a “contiguous” empire rather than an “overseas” one has tended to cloud understanding more than to enable it.^[37] Certainly early-modern Russian empire-building was distinctive, at least in contrast to European patterns, with porous frontiers rather than vast oceans and no firm distinction between center and periphery.^[38] Yet in the second half of the nineteenth century, Russia’s rulers sought to overcome this blurring of colony and metropole. The direction in which the late imperial Russian state’s practices were moving was very similar to “overseas” colonial empires such as Britain and France. Indeed, the interest shown by the regime in researching and ruling through native “customs” in the Northern Caucasus reveals Russia to be a European pioneer in colonial policy.^[39] Russia was looking more and more like a colonial “empire of a nation.” Cultural Russification and colonialism were distinct, but consistent, policies aimed at that goal, operating in different parts of the empire.
World War and resulting collapse caught the Tsarist regime before the transition to a modern national state and colonial empire was complete. Indeed, perhaps because that transition had not been completed, it was possible to reconstitute the empire into something new; this the Bolsheviks sought to do. Like its late imperial predecessor, the new Soviet elite sought through its nationality policy to create difference and sameness simultaneously. However, the nascent nomenklatura implemented both traditional imperial and national practices throughout the Soviet space. That is, it understood the Soviet Union to be both a traditional multiethnic state and a potential unitary national state at one and the same time. In this respect, Soviet policy represented both a break with the immediate past and continuity with a more distant one. Whereas Imperial Russia’s practices were moving in a common direction with colonial empires of the day, Soviet practices were sui generis and, ultimately, contradictory. I demonstrate this through an examination of the two principal aspects of Soviet nationality policy from the 1920s to the 1950s: korenizatsiia (indigenization) and Sovietization, which represent examples of imperial and national practices, respectively.

SOVIET IMPERIAL PRACTICES: KORENIZATSIIA

Korenizatsiia entailed identifying, classifying, bounding, and in some cases inventing, ethnically defined administrative units that would be staffed by “natives.” In part it was, in Terry Martin’s words, “a radical strategy” to prevent “decolonization” and to “disarm nationalism.”^[40] The idea was, simply enough, that non-Russians would be more willing to accept Soviet power if familiar people speaking a familiar language represented that power. A diverse land needed diverse rulers and the institutions to train them. Soviet authorities represented the policy as a rejection of Tsarist Russification and colonialism, yet korenizatsiia bore a striking similarity to earlier Russian imperial practices of elite co-optation. And as with their imperial predecessors, political reality (as Lenin perceived it) dictated such a policy; it was necessary if the Bolsheviks were to attract support of the non-Russians in the wake of imperial collapse.
Although this collection of policies was chosen for very pragmatic reasons, Yuri Slezkine has pointed out a critical ideological-discursive aspect to them: dividing human beings according to ethnic identity was, or at least became, completely natural to the Soviet political leadership. As he aptly puts it, “Soviet nationality policy was devised and carried out by nationalists.”^[41] Understanding the adoption of korenizatsiia thus requires us to address both its pragmatic and discursive elements. The former appears to make the policy seem similar to that of traditional empires, while the latter resembles colonial discourse in its reification and sometimes invention of ethnic boundaries. We can make some headway in this direction by contrasting korenizatsiia with other kinds of differentiating state practices.
Dynastic empires typically recognized “traditional” law and authority and co-opted local elites. Colonial empires co-opted, and sometimes invented, local elites as well. The British, for example, spared no effort in searching for “tribes” and “chiefs” that would correspond to their expectations even where Africans lacked them.^[42] Scientific knowledge about the “content” of cultural difference was fundamental to colonial policy. In contrast to the traditional imperial practice, but similar to colonial ones, Soviet nationality policy was based on the understanding that cultural diversity was a political problem to be solved; as Francine Hirsch shows, ethnographic knowledge was critical in its establishment and implementation.^[43] But in contrast to the colonial empires, the Soviet state did not inject content into ethnic categories. There was no Soviet tribal or customary law. Special Muslim courts, for example, were permitted to exist only through 1927 and then were abolished with “a highhandedness” of which the Soviets’ “French colonial counterparts in the Maghrib and other parts of the colonial world would never have dreamed…”^[44] The curriculum and organization of non-Russian schools were, from the late 1920s, designed to be as close as possible to those in Russian schools. Language was understood to be merely a transparent medium rather than reflective of cultural “content.”^[45]
It is in this context that Soviet ideological nostrums about achieving unity through diversity and of the possibility of ethnic identity eventually withering away are often invoked. That is, Soviet ideology supposedly held that content-less ethnic diversity was a stage on the path of overcoming ethnicity itself. The irony here, of course, is that there is little evidence that Stalin – the principal architect of korenizatsiia – took such ideas seriously.^[46] Indeed, he consciously chose to miss the greatest opportunity to do so. Amid his regime’s centralizing and industrializing drive during the “Revolution from Above” of 1928-1932, Stalin rebuked those who understood Cultural Revolution to entail assimilation; he emphasized that multiple ethnic identities would be a part of Soviet civilization for a very long time.^[47] Contrary to Lenin, Stalin held that ethnically bounded “nations” would continue to develop under socialism, and invented the category of “socialist nation” as an elaboration of this theory. According to the architect of nationality policy, such socialist nations would not disappear until the establishment of a socialist economy on a global scale, the “second phase” of a projected worldwide dictatorship of the proletariat.^[48]
Concomitant with this development in official Marxism was the emergence of an official discourse of ethnogenesis, which became the basis for much research in the social sciences and history.^[49] Socialist nations were supposed to have ethnic roots that scholars could trace “from earliest times.” Both Stalin’s ideological innovations and ethnogenetic discourse tended to reinforce the reification of ethnic identity associated with the original adoption of korenizatsiia in the early 1920s. In his penultimate theoretical work, on linguistics, Stalin reasserted the key idea of ethnicity’s timelessness (understood linguistically), as non-Russian intellectuals themselves understood at that time.^[50] The consequence was the hardening of ethnic boundaries and an increase in the salience of ethnic identity for Soviet citizens in their daily life, even when the actual content of ethnic differences was diminishing.
We can reconcile the pragmatic-political and ideological-discursive dimensions to korenizatsiia by looking at Soviet practices together with discourses. Korenizatsiia could be understood by the nomenklatura, and by many historians today, as a policy of ethnic nation-building. Indeed, the term itself (natsional’noe stroitel’stvo) was in use during the 1920s and 1930s; thereafter, the regime took credit for allowing “backward peoples” to develop into “nations.” And the long-term implications of korenizatsiia did indeed lead to independent national states, and perhaps to nations as well. But in the context of the Soviet Union as a whole – that is, if we look at its orientation from the standpoint of the central state authorities rather than from the non-Russian periphery – korenizatsiia was first and foremost a practice of differentiation of populations akin to that of pre-reform Russia. The power of the elite and the unity of the state depended on such differentiation, which sought to replicate a traditional imperial approach to diversity in modern conditions and thereby avoid institutionalized colonial racism.
In this context, it is useful to distinguish between the universalism of the liberal national state, which required colonial-style limitations to maintain its power in the periphery, and that of the Marxist Soviet state. The latter’s universalism resembled that of the earlier empires in not requiring such limitations because it was a universalism based on a generalized (albeit content-less) particularity. Perhaps this difference emerged because liberalism’s universalistic utopia of a society of autonomous individuals had already been achieved (its cost was class conflict at home and colonialism abroad), while that of Marxism remained in the future. And while a global Marxist universalism seemed very far away even to Stalin himself, Soviet national unity was a far more immediate (and pressing) matter.

A SOVIET NATIONALIZING STATE

If korenizatsiia emerged practically with the Soviet Union itself, practices of Sovietization emerged only later, in the mid-to-late 1930s, as the full implications of the commitment to “building socialism” (and thus a modern, industrialized society) became clear. Such practices were designed to forge a national community by creating a “common conceptual currency” shared by all Soviet people and molding standardized, culturally-interchangeable individuals suitable for modern politics and economics. Russian language and culture were the most obvious basis for Soviet uniformity because Russians were more numerous and because “Russianness was assumed to be largely transparent, meaningless, and therefore equal to modernity…”^[51] In language policy, principles of reform emphasized introducing Russian neologisms, rather than those from local sources, as well as using the Russian alphabet as the basis for most non-Russian languages. Furthermore, as of 1938, non-Russian schoolchildren were required to study the Russian language.^[52] This contrasted sharply with policy in the 1920s, which sought to distinguish local languages as much as possible from Russian and which tended to denigrate Russian language and culture.^[53]
No less important than the new emphasis on Russian language, official expectations of educated non-Russians were that they would become more like Russians in their lifestyle (“byt”). The native Siberian peoples “were supposed to actively desire to take part in socialist construction, take baths, study the Constitution, and read Stalin and Anatole France.”^[54] Uzbek men were supposed to unveil their wives.^[55] Not surprisingly, as Douglas Northrop has shown, the decision to judge loyalty to the Soviet order by the presence or absence of the veil subverted korenizatsiia policies in Uzbekistan.^[56] Attempting to make people the same, a national practice, contradicted efforts to keep them different, an imperial one. Further evidence of the disruptive potential of ethnic diversity became manifest in the Second World War, during which many Central Asians were mobilized into industry and sent to the Urals. Local authorities could not cope with their particular needs, especially for traditional foods.^[57] The results were high rates of desertion and illness. Had these workers been culturally closer to Russians, the regime would have been better able to use them for its tasks.
For a national elite, such as the Soviet nomenklatura, similarity means interchangeability and reliability. A key part of the effort to create a Soviet national community, therefore, was to make all Soviet people, regardless of their ethnic identity, the same in certain fundamental ways. Making people more alike eases the transformative tasks of the national state. For the Soviet national elite, and especially for the planners, the USSR was, or should have been, a uniform economic and cultural area that required a standardized population to function effectively. That official ideology never described the Soviet “people” as a nation should not blind us to the fact that many Soviet policies treated the Soviet people like a nation.^[58] Likewise, bombastic ideological pronouncements about Russians as “elder brothers” and the chauvinistic treatment of non-Russian cultures should not lead us to believe that what was really going on was Russian chauvinism for its own sake. It was not necessary to make non-Russians more like Russians because Russians were seen to be ethnically superior, though certainly some Russian officials at the center thought that way.^[59] Rather, Russians were more functional for the goals of the state.
Perhaps the most notorious aspect of Stalinist nationality policy was the persecution of diaspora minorities and “enemy nations.” From early efforts to integrate Germans, Poles, Koreans, and Jews (to take just a few examples) into the Soviet order through the pragmatic policies of korenizatsiia, the regime soon moved to policies of deportation and repression.^[60] It is not accidental that the deportation of ethnic groups was preceded by policies of korenizatsiia. This was true as much for Poles and Koreans in the 1930s as for the Chechens in the 1940s.^[61] Deportation and korenizatsiia were fundamental alternatives, national and imperial practices respectively, for dealing with the problem of difference. Repression or exclusion of ethnic minorities as aliens is a profoundly national policy aimed at those groups that cannot or will not be integrated into the dominant national community. They are the policies of a “nationalizing” state, to borrow Rogers Brubaker’s terminology.^[62] The wartime deportations of entire peoples, no less than the “ethnic cleansings” of the 1930s, followed from the determination that those groups could not be “made Soviet” and had so demonstrated by their “traitorous” behavior.
Postwar anti-Semitism may appear to follow the same logic as earlier repressions of diaspora minorities. As Amir Weiner aptly puts it, “A core message of the anti-cosmopolitan campaign was that the Jew remained a Jew, an eternal alien to the body national, no matter what the circumstances.” He further argues that the emerging official myth of the war precluded a “particularistic Jewish space.”^[63] That myth had to be unitary. Jews were a threat to the Soviet national order. Yet, at the same time, anti-Semitism served the interests of other ethnic groups; limiting Jewish admission to Ukrainian institutions of higher education, for example, facilitated korenizatsiia in Ukraine by increasing the proportion of Ukrainians among the student body.^[64] In this respect, Stalinist anti-Semitism was truly “over-determined” and was perhaps the only case where imperial and national policies actually coincided.
Although practices of Soviet nationalization were more prominent from the mid-1930s, this “revision” of the original Soviet nationalities policy did not mean an end to korenizatsiia, which continued until the end of the Soviet period in one form or another.^[65] As a result, national and imperial practices stood in stark contradiction to one another from the late 1930s onward. For example, at Stalin’s death non-Russian schoolchildren were studying in native-language schools in unprecedented numbers; they did so, however, under a fully standardized Soviet curriculum that effaced the content of ethnic differences as much as possible.

EMPIRE VERSUS NATION(S)?

The history of Soviet nationality policy has usually been told in one of two ways. The first is a story of repression of non-Russian national aspirations and of cynical Bolshevik “concessions” to non-Russian national “reality.”^[66] For this “nation-destroying” literature, the Soviet state was a recast Russian empire, facing similar dilemmas and with similar motives as its predecessor. Once the dominant narrative, and still common among histories of specific regions and peoples, this story has been nudged aside in recent years. A new narrative emphasizes the extent to which the Soviet regime built nations to its own detriment and takes seriously its effort to promote multiple national identities within the context of a Soviet empire.^[67] According to this story, Soviet imperial unity and ethnic particularity were in fundamental contradiction.
This new story makes sense in that unity and diversity were conscious, yet contradictory, goals embedded in the practices of the Soviet state. No less important, it replaces the earlier historiography’s valorization of non-Russian nationalism with an attempt to solve a key historical paradox: how, in Ronald Suny’s words, “A state that had set out to overcome nationalism and the differences between nations had in fact created a set of institutions and initiated processes that fostered the development of conscious, secular, politically mobilizable nationalities.”^[68] Yet the new story overlooks that which made Soviet unity really contradictory to ethnic particularity, namely the effort to create a Soviet national community. True, there was never an effort to create a Soviet “nation” (natsiia), yet we should not allow the peculiarities of Soviet discourse to blind us on this score. Natsii were ethnic collectivities with distant origins in Soviet discourse. But the term “Soviet People” (Sovetskii narod) was in use from the 1930s to describe the kind of community that most national states aspire to mold. By contrast, the ethnic “nation-building” practices of the regime followed from the traditional and pragmatic imperial policy of differentiation, a policy which only superficially resembled contemporary colonial practices.^[69] Indeed one might go so far as to describe korenizatsiia as a desperate attempt to recreate traditional imperial politics within a national state. Thus attention to the discursive-ideological level alone leads to two mistakes: first, by taking Soviet discourse of ethnic nation-building on its face, we overlook the extent to which this coincided with imperial differentiating practices; second, by over-emphasizing the absence of a Soviet nation in official discourse, we overlook the extent to which Sovietization bears striking resemblance to nationalizing practices.
When we recognize that the regime was engaged in incompatible policies of both empire maintenance and nation-building, we are more likely to understand the problems the regime encountered when confronting ethnic difference. For, ultimately, the regime was pursuing fundamentally incompatible goals by means of contradictory practices. Put another way, if the basic tension of colonialism was between liberal universalism expressed in national citizenship and the simultaneous imperative to limit liberalism’s reach, the basic tension inscribed in the Soviet multinational state was between the regime’s need to create a culturally uniform national community and its need to maintain ethnic diversity. Understanding this contradiction, as we have seen, requires taking the policy orientation of the regime seriously and, no less important, demands examination of how that orientation was inscribed in practices. Such an examination does not mean a return to a history of central policy alone, but nor does it mean a fixation on the periphery that, however unintentionally, will likely recreate the valorization of non-Russian nationalism of the Cold War era literature.^[70] As important as the subjective perceptions of non-Russians may be, we should be careful not to objectify their subjective perceptions as the defining characteristic of the Soviet experience of ethnic diversity. For, as Bourdieu emphasizes, the meaning subjects attribute to their actions – and, by extension, their world – may not be the deepest meaning that historians can uncover.
Further attention to the “essential tension” between cultural diversity and national uniformity has at least two consequences for the study of Soviet history. First, rather than describing the collapse of the Soviet Union either as a consequence of the repression of non-Russian national aspirations or as the unintended consequence of multiple nation-building, the answers that correspond to the two narratives noted above, we should focus on the tensions inscribed in Soviet nationality policy itself. That is, whether the incompatibility of national and imperial practices themselves created the conditions for the collapse of the multinational state is an important question worth pursuing.
Second, attention to the effort to craft a Soviet national community naturally raises the question of a distinctive kind of Soviet modernity, a subject that has garnered increasing attention in recent years.^[71] Here I have emphasized the modernity of both national and colonial practices and the traditional nature of dynastic (and Soviet) imperial practices. I have suggested that the Soviet multinational state was a combination of the modern and traditional; it was not merely a distinctive kind of modern state, owing to its political elites’ desire to construct socialism, but a unique hybrid. In emphasizing Soviet uniqueness, I do not mean to argue for Soviet (or Russian) exceptionalism or incomparability. On the contrary, what was unique about the Soviet experience was not merely its socialist ideology, but also the way that it combined commonplace historical practices in distinctive ways. Attention to the Soviet state’s ambivalent confrontation with multiethnicity highlights that very distinctiveness.