For this reason, we need from the start explicitly to conceptualize our “ideal type” of “empire” or “nation”.^[1] These models enable us then to derive operational hypotheses, or (to be less brutally scientific) sets of assumptions and leading questions, which guide our selection, organization and interpretation of the primary materials.^[2] Of course, it doesn’t really matter how we define our “ideal types”, so long as they are of instrumental value in our research and so long as we make these models explicit. As such, criticism that any model or concept is intrinsically wrong misses the point, since, in the words of anthropologist Evans-Pritchard, “a theory may have heuristic value without actually being sound”.^[3] Or, as Sovietologist Alex Inkeles put it, “there is no such thing as a right or wrong sociological model. There are richer and poorer ones”.^[4] Indeed, abstract models are necessarily false when empirically tested: they can never be anything but sources of “fertile error”.^[5]
This approach, of course, is relativistic, but relativism, as the historian and philosopher R. G. Collingwood understood better than many of his own commentators, “is not an argument for historical scepticism. It is only the discovery of a second dimension of historical thought, the history of history.”^[6] The fact that history-writing is so intimately linked to its contemporary context, that it is so “subjective” and dependent upon the imagination of the historian, does not invalidate its cognitive potential.^[7] Indeed, to believe that history and its concepts could in any way be “objective” is not only self-deluding (as one writer has phrased it, “one does not doff an idiosyncrasy as easily as one dons a white coat”) but also dangerous, since it gives “scientific” status and credibility to ideas (not least, those of “empire” and “nation”) which are contingent and best treated with scepticism.^[8] Popular or political discourse might choose to speak of the “British Empire” or “British nation” (or, for that matter, of “liberal democracy” or “the market”) as if these phenomena existed a priori as real objects; scholars will not (especially if they are Scottish or play with stocks and shares).
I’ve undertaken this methodological detour to bolster my argument that the question as posed above itself needs to be “problematised”. A punchier answer would have been: it all depends how you define “empire” and “nation”. However, it might be more interesting to construe the question from another angle. Empirically to compare the political practices of two model systems, it may be more revealing to employ “categories and approaches” derived from a third concept. We might, for example, usefully approach the question from the point of view of “governmentality”.^[9] In this case, we are very likely to find that systems that we would otherwise consider as very different employ similar strategies of domination and some systems we are used to grouping under the same typology in fact uphold very different “regimes of truth”. Or we may find continuities in histories of one system where traditionally scholars, fixated on regime “type”, saw “transition” or “transformation”. This, in turn, may prompt us to reconsider the way in which we describe and conceptualise these systems in the first place.
The editors’ second question permits me to pick up on these reflections, as it asks “what are the discursive challenges in studies of political history of empires?” My answer, again, will be irritatingly evasive: it depends, of course, on what is meant by “discursive”. Scholars will always operate within one or more discourses, which constrain and delimit what they can say, but are at the same time necessary preconditions for saying anything comprehensible to the normative community. One problem, which I’ve touched on already, is the relationship between scholarly discourse and popular or political discourses, which manifest greater resistance to integrating new discursive terms or to “renegotiating” discursive relations (embodying, it could be argued, relations of power).^[10] There is also the question of distinguishing between the discourse(s) in which the researcher is implicated and the discourse(s) that might be the object of study, or which might construct the object of study. If we study “empire”, for example, do we study what scholars now conceive of as constituting “empire”? Do we study what is popularly thought of as “empire”? Do we study what contemporaries in the period we are studying considered to be “empire”? Or do we study the ways in which this concept of “empire” was itself constructed and reproduced?
The editors’ third question concerns “differences between the colonial policy/colonialism/expansionism of empire and that of nation state (in their ideological legitimation, nature and limits of politics of integration, the degree of economic motivation of certain policies)?”
To start with, I’d reiterate my main point above – that if we compare “empires” and “nations” on the basis of their policies or practices, we must be confident that these policies and practices have not been used already to construct the typology. I’m aware, however, that in repeatedly making this point, I’m to a large extent dodging the substantive questions which the editors are asking. As Blair Worden has noted,

research involves the shedding, not the confirmation, of our preconceptions. If historians go to the archives expecting certain answers to their questions, careful study of the evidence will almost invariably change their minds. It will alter not merely their answers, but their questions.^[11]

Or, put more succinctly, the researcher must be the “servant of his evidence”.^[12] In this case, the test of experience warns me to be extremely wary of making any generalisations about the policies and practices of “empires” and “nation states”. More useful concepts, in my opinion, cut across categories of “empire” and “nation”.
For example, if we consider population displacement in the space of the former Russian empire, are the policies of the newly-independent East European nation-states fundamentally different from those of the new, multinational Soviet polity?^[13] Or if we compare the state-building practices of “nationalizing” and “revolutionary” states, focussing, perhaps as suggested, on their integrationist policies, legitimizing ideologies and economic motives? All defined their normative communities and insisted that some categories of population could never be part of the collective, motivated by what has been termed the “impetus to homogenize.”^[14]. Soviet ideology upheld the pre-eminence of a normative socio-economic community and the states of eastern Europe strove to a lesser or greater degree for the socio-ethnic homogeneity of their population. Both distinguished “core” from “peripheral” or “alien” populations in social terms, and to some extent sought to distribute them spatially in corresponding manner. Their methods could be equally brutal, albeit varying in scale, their motives were equally “ideological” (though the ideologies, of course, differed) and they pursued common goals of self-definition, “security” and economic modernization.^[15]
Looking for antecedents of these practices, can we not identify many continuities in state practice between the imperial tsarist regime and the new multinational, and resolutely anti-imperial, socialist state (and even the new nation-states that emerged from its frontier lands)?^[16] Does this make the Russian empire any less “imperial”, or Soviet Russia any more so? Do post-colonial nation-states inevitably employ “imperial” practices against their own social or spatial peripheries? Or are these categories just heuristically inappropriate?
To take another angle, if we focus on states’ politics of integration, we find practices in late nineteenth century France which could commonly (which I use as a synonym of “erroneously”) be considered “imperial”.^[17] Or, for that matter, in the development of Great Britain.^[18] Yet in these cases we are not studying these states as empires, but as nations: we are not considering the expansion of their state power beyond their borders, but the expansion of “central” power over the “periphery”.
Does distinguishing between “imperial” France and “national” France, or between “imperial” Britain and “national” England (this is where the problems really begin!) help us to understand the history and character of their colonial policies? Perhaps it’s better to start without such typologies. At times, empires have been essentially non-national (the “barbarian” domination of the late Roman Empire is a case in point) or multinational. At other historical conjunctures, nation-states believed that it was their “natural” destiny to expand their power beyond their borders. At other times, imperial expansion has been co-terminous with nation-building and it is hard to distinguish one from the other.
Can we begin by categorize “empires” and “nation-states” as spatial systems? Usually, empires are represented according to the center-periphery model, in which the metropolis dominates and exploits the colonial lands.^[19] And yet, it is quite possible to study nation-states according to this model without claiming that nations are empires (although when they are constituted from different ethnic or cultural groups, as they usually are, they might indeed manifest “imperial” or “colonial” tendencies, at least from the perspective of the periphery).^[20] In any case, spatial concepts are analytical not descriptive, and cannot themselves account for the phenomenon under study: as geographer John Agnew wrote, “[t]he use of the spatial couplet ‘core-periphery’ carries a number of misleading and dangerous implications … (including) the hypostatization and ascription of causality to spatial categories.”^[21] In other words, a political system might be characterized by a certain form of spatial organization, but it is not defined, much less determined, by that form.
At this point, I’ll turn to another of the editors’ questions, concerning the possibility of drawing analytical (or empirical?) distinctions between political, social, economic, and cultural processes in the situation of evolving modern and mass society? Is there any specificity (they ask) to these processes given the context of “empire”/multinational polity and nation state? How far it is possible to analytically expand the concept of “political space” in the context of empire and nation state.
I’d like to ask whether it would be analytically useful, i.e. if it would illuminate more than it would obscure, to consider Soviet repressive practices, and especially the Gulag's operations in its period of territorial expansion (1929-1936) in the context of European “governmentality”, specifically practices associated with overseas “imperial” expansion and domestic penal and population policies. Firstly, I am thinking of the way in which a modernizing center, in the course of redefining itself and its constituent “elements”, seeks to establish its power by redefining its “peripheries”, not just territorially, but discursively, for example in terms of central versus peripheral social, ethnic, sexual etc. categories. (Of course, in the modernizing West European states those categories defined as socially peripheral often found themselves forcibly transported to the spatial peripheries, as the center strove to establish an ordered socio-spatial hierarchy). In this process, we witness the transformation of the early modern “body politic” into the modern “body social” and the emergence of modern forms of “governmentality” in both discourse and practice.^[22]
In terms of discourse, recent research suggests the idea that the Soviet political police in its monitoring of popular mood conceived of society as a living organism and of its own role in medical or epidemiological terms.^[23] In this account, social health is governed by regulating society's constituent “humours”, which require periodic intervention, extraction (iz'yatie) and isolation of pathogenic “elements” (a form, we might say, of social bleeding) to preclude contamination and to restore the dominance of healthy elements (which is inevitably disturbed again as these very interventions catalyze the activities of hostile elements).
As noted earlier, there are evident parallels between the socio-economic reordering of the multinational Soviet polity and the “nationalizing” aspirations at work in the East European successor states to the tsarist empire, which similarly identified their healthy citizens, distinguishing them from those deemed less than “healthy”.^[24] In part, this epidemiological rhetoric can be traced back to wartime and post-war concerns about hygiene among refugee populations.^[25] It operates as a ‘regime of truth’ across many states, imperial, multinational and national, and distinguishing among these categories does little to help us understand these processes and practices.
In terms of practice, the notion of “iz'yatie” prompts me to think of West European states’ deportations of their convicts and socially marginal (e.g. the “chronically unemployed”) to their colonies, which was similarly motivated by security concerns and economic ambitions, and similarly legitimized by notions of rehabilitation on the “frontier”. As early as 1584, Richard Hakluyt, whose Discourse on Planting did much to promote transportation as a means of populating English colonies, wrote that “the fry of wandering beggars .. burdenous to this realm, may there be unladen, better bred up and may people waste countries.”^[26] Proponents of transportation for vagrants, convicts and dissenters also argued that it offered greater domestic security, relief of overcrowding in penitentiaries, cheap labour in colonial domains and the possibility that deportees would find, in the words of one vigorous mid-nineteenth century advocate of the policy, “space for exertion, reward for industry and a comparative oblivion for the past” through their pioneering labours.^[27] Transportation, he claimed, “[a]s a means of making men outwardly honest, of converting vagabonds, the most useless men in one country into active citizens of another and thus giving birth to a new and splendid centre of civilisation, it has succeeded to a degree unequalled in history”.^[28] Are these pronouncements best understood as “imperial” or “national”? Obviously it makes little sense to consider them in these terms at all: to do so would hide from view essential commonalities of practice among modernizing West European states which were developing as empires and nation-states at the same time.
In imperial Russia, too, there were supporters, and equally vigorous opponents, of transportation. Russia already had long traditions of political and religious deportation and convict transportation, sometimes combined with forced labour (katorga, first introduced as a penal measure in the late 1600’s). The motives behind sustaining this policy were as complex as those of European colonial powers. In the words of a Russian jurist of the late nineteenth century,

Russia, in fact, has never considered [transportation] merely as a punishment, but has used it as a means to resolve problems of internal and external politics, for example to occupy foreign regions [contrees etrangeres], to populate and to pacify provinces recently joined to the Empire and, not least, as a standard administrative measure. It has practised it without interruption for three centuries, often in opposition to the general principles of penal law … it is one of those rare creations of Russian penal law, born entirely of the needs and conditions of Russian life.^[29]

In Russia, therefore, transportation was a practice also bound up with both imperial policy and domestic state-building practices. But can we even differentiate between the two? The German historian Werner Markert proposed that Russia’s acquisition of a contiguous territorial empire through centuries of outwards expansion brought about the increasing centralization of the central state, the leveling of social differentiation, the elimination of social and corporatist interests and produced, in effect, a state of internal as well as external colonialism.^[30]
In the course of the nineteenth century, over half a million convicts, some sentenced judicially, others subject to administrative penalties, were transported to Siberia, many accompanied by their families. Are we to understand the settlement of Siberia as an imperial practice or a strategy of nation-building (American transcontinental expansion raises the same question)? From 1859, Russian convict settlers were directed further eastwards because of fears concerning potential “contagion” of the growing free population on the “frontier”.^[31] At the same time, transportation became a subject of increasing debate among Russian, as well as international, officials and jurists. At the 1872 International Prison Congress in London, Professor Vladimirov defended the policy as “a powerful agent of colonization,” while at the Stockholm Congress a few years later, Kokovtsev declared himself opposed to penal resettlement (ssylka na poselenie), while voicing support for katorga.^[32] Within Russia, the imperial government established a series of commissions to discuss transportation, the last of which, chaired by Minister of Justice Nabokov, proposed in the 1880’s to end administrative deportation, except for ecclesiastical “crimes”, but to continue katorga within prisons, with exile of convicts to Siberia after completion of their term.
The turn of sentiment against transportation coincided with greater efforts on the part of the state to promote free “emigration” to the eastern regions. In 1895, the Russian jurist Ivan Foinitski declared (concluding a survey of the English experience) that:

Transportation perverts the colony, imposing on it a sort of servitude and encouraging to a large extent a return to barbarity. Emigration, on the contrary, develops there the spirit of enterprise and gives it a new strength to maintain free institutions.^[33]

Towards the end of the nineteenth century, free population movement eastwards had massively intensified. Between 1871 and 1916, over nine million migrants, mostly Russian and Ukrainian peasants, “emigrated” within imperial territory, the majority to Siberia, the Kazakh steppe and the Far East.^[34] At the same time, official and scientific circles, and increasingly the wider educated obshchestvennost’, became concerned that the empire’s “colonising forces” should be controlled to ensure a “correct redistribution of population,” which, it was thought, would promote “correct progress”.^[35] However, this emerging discourse conceived the frontier not only as a symbol of progress and expanding civilization, but also as a site of danger and source of potential contamination for the central state.^[36] The answer was to call for government intervention to select the settlers in the center, to direct flows of settlement and to make provisions for correct discipline and education on the frontier.
Can we make better sense of these policy changes in terms of Russia’s imperial history or its history as a modernizing state? What about continuities in practice after the October revolution? Under the Soviet regime, there were renewed attempts to promote the free colonization of the North and East, often championed by the same officials who had worked in the tsarist resettlement administrations.^[37] At the same time, there was new interest in penal transportation as a means to meet, in the first instance, the political exigencies of the embattled revolutionary regime. As early as 1919 the Soviets had established concentration camps in Arkhangelsk oblast’ to isolate antagonistic groups and nefarious individuals “extracted” from the center. Soon, state agencies also considered reviving the use of penal resettlement as a means to colonize remote regions and to exploit their resources.
Many of the most prominent Bolsheviks had suffered exile or katorga, and bore the imprint of the experience when they assumed high office in the post-revolutionary state. Dzerzhinskii, for instance, had spent eleven years in exile, forced labour and prison before 1917, including a term in the European Far North. After the Revolution, he founded and led the Cheka, and was appointed to head first the Commissariat of the Interior (1919), then the Commissariat of Transport (1921) and later the Supreme Council of the National Economy (VSNKha, 1924).^[38] In 1923, Dzerzhinskii wrote:

The republic cannot be merciful towards criminals and cannot waste resources on them; they must cover the costs associated with their care with their own labor; they must be used to settle undeveloped areas in Pechora, in Obdorsk [Salekhard] … We will have to work to organise forced labour (penal servitude) at camps for colonising undeveloped areas that will be run with iron discipline. We have sufficient locations and space.^[39]

The idea that forced labour could offer a way to settle and develop the periphery at no cost to the center naturally also appealed to economic interests. In November 1925, Deputy Chairman of the VSNKha G. L. Piatakov (who had spent a year in exile in Irkutsk in 1914, before escaping via Japan to join Lenin in Switzerland) sent a report to Dzerzhinskii in the latter’s role as Head of VSNKha. In this, he stated:

My study of geographical factors affecting industrial issues has convinced me that in order to create the most elementary conditions for a work culture, compulsory labor settlements will have to be established in certain regions. Such settlements could also relieve overcrowding in places of incarceration. The GPU should be instructed to explore these issues.^[40]

In 1929, therefore, as collectivization boosted the supply of forced labour, Deputy Head of the OGPU G. G. Iagoda proposed a program for the penal colonization of vast territorial expanses on the basis of “existing laws and existing practice”. Initially he had in mind the Northern Krai (Ukhta and Pechora), but his vision ultimately stretched from Karelia and the Kola peninsula to the Far East.^[41] The immediate objective of this strategy, as stated in the 27 June Politburo decree was to “colonize these areas and exploit their natural resources by means of the use of prisoner labour”.^[42]
Is this best understood as a colonial expansion of the state’s economic frontier, or as economic exploitation of territory already within the state’s borders? Where settlement and colonization were directed to lands inhabited by non-Russian indigenous populations, should this be conceived as a new form of imperialism? Professor Schmidt, head of Glavsevmorput', which originally had authority over much of the Soviet Arctic, declared that his organization was “a modern socialist equivalent of the East India Company” (it would be important to understand what the word “socialist” implied to Schmidt in this context).^[43] Could the NKVD in the 1930’s, with its control over transport infrastructure (rail, roads and water-routes), over the forcible re-distribution of population, over the exploitation of essential resources in the North (after the demise of Glavsevmorput’), and over the surveying and mapping of territory, be considered a form of imperial agency?
To conclude this essay, in the light of the development of European “governmentality”, its forms of “modern” social knowledge and its associated representations, rhetoric, spatial practices and population policies - might we come to understand something new about the nature of the Russian empire, Soviet Russia and the Soviet Union? Might this be a better key to understanding important aspects of their histories than categories of “empire” or “nation”? Of course, there are doubtless other aspects of these histories, and of the histories of minority nationalities within their changing political space, which are suited to conceptualization within an explicit “imperial” or “national” framework. And there are doubtless further problems (historical “subjectivities”, “everyday life”, gender or sexual identities, population displacement, to cite just a few of the innumerable new historiographical sub-fields) which are best conceptualized and approached in yet other distinct ways.