Ethnic Exclusion in Nationalizing States

Andreas Wimmer. The Sage Handbook of Nations and Nationalism. Editor: Gerard Delanty & Krishan Kumar. 2006. Sage Publishing.

The most prominent theories of nationalism disagree on a number of crucial questions: whether nations are modern inventions or rather rest on pre-modern ethnic foundations; whether state modernization and warfare are the cause or the consequence of nationalism; whether nationalism flourishes exclusively on the soil of industrial capitalism or everywhere where the model of the nation-state has been implanted; if nationalism has indeed already passed its greatest moment and we thus find ourselves at the threshold of a post-national age. All these major debates rest, however, on an underlying consensus. Nationalism and nationhood are portrayed as integrative political and social forces transforming older, exclusionary and hierarchical societies. Nationalism and the nation-state provided the basis for the democratic inclusion of large sections of the population that were hitherto held at arm’s length from the centres of power. Nationalism also changed the power relations in the cultural domain by raising the status of the despised culture and language of the lower classes.

The terms in which the integrative power of the national community is conceived vary from author to author. For Karl Deutsch, a nation was constituted by a shared communicative space, enhanced by similar cultural codes (not necessarily a language) and dynamized by the uprooting and mobility that urbanization and modernization had brought about. The state is notably absent from this picture and seems to play no role in shaping communicative spaces, assimilating minorities or enforcing legal discrimination against them (Deutsch 1953). This improved with Ernest Gellner’s account.

He emphasized the role of the educational system of nation-states in bringing about cultural and especially linguistic homogeneity. The state played the role of a servant to an industrial capitalism in need of flexible, mobile workers who can quickly assume new roles in an ever-changing division of labour (Gellner 1983). He did observe that being governed by bureaucrats of foreign language and culture may stimulate national awakenings and conflicts, such as in the Czech example (his ‘Ruritania’) used to illustrate the industrialization argument. And he did consider the durable inequalities that groups may suffer if their culture or physical appearance made assimilation into the nation difficult—having in mind the Jewish and African American experience. But such domination and exclusion did not, in his view, represent a major feature of the new world order of national states but rather one of its rare and deplorable pathological permutations.

According to Benedict Anderson, imagining the national community was made possible by the delegitimation of dynastic rule, the disenchantment of universal religions and the rise of vernacular languages through the combined influence of Protestantism, the modernization of absolutist state administrations and the development of a market for printed materials (Anderson 1991). The national community was held together by common language or the experience of restricted social mobility within the territory of a colonial province. It was imagined as a community of shared origin and history that would live through all the historical changes and secular developments that the newly discovered emptiness of time made it possible to think of. While Anderson noted, en passant, that the nationalist leaders of Latin America, including Simon de Bolivar, were not prepared to count the black population of Venezuela or the Indian peasants of Colombia in the national ‘we’, he did not draw any consequences from this observation. In general, there is little room for the more conflictive, warlike aspects in his analysis of the rise of nationalism as the Zivilreligion of the modern world.

Anthony Smith goes furthest in blinding out the struggles for and against domination that accompany the establishment of national states. Although he is aware that the solidarity of ethnic and national communities is often the result of conflict and war (Smith 1981), his main intellectual project went in an entirely different direction: to challenge the modernist account of his former teacher Ernest Gellner by pointing to the continuity between modern nationalism (and nation-states) and pre-modern ethnic communities (and ethnically defined polities). Nationalism thus reframes already existing ethnic myths, historical memories and symbols of identity (Smith 1986). In a neo-Herderian fashion, then, history breathes through the body of immortal ethnic groups that grow to full blossom in the age of nationalism. From this perspective, nations appear as historically stable communities of solidarity which provide human beings with dignity, cultural meaning and a sense of belonging.

Finally, authors like Michael Mann or Charles Tilly emphasized the role of warfare between competing absolutist states in the generation of nationalism—which is seen as the response of an emerging civil society against the increasing pressures of taxation and conscription. But they did not look at the new forms of exclusion along ethnic lines that the nation-state brought about, emphasizing instead its capacity to create solidarity and loyalty and thus to mobilize the population for external conflict (Tilly 1975; Mann 1995).

It is only in the work of those authors most critical to the ideology of nationalism—not accidentally the first authors to write book-long essays on the subject—that the exclusionary character of nationalism and the nation-state are discussed more systematically. The conservative historian Eddie Kedourie wrote a book (Kedourie 1960), against which much of Ernest Gellner’s writing was implicitly directed, that portrayed nationalism as an ill-guided modern ideology, born out of German Romanticism and prone to bring destruction and violence to historically grown multicultural societies such as those of Eastern Europe or the Middle East. Hans Kohn deplored the totalitarian character of nationalism when it was combined with East European traditions of authoritarian rule (Kohn 1944)—swiftly overlooking the history of nationalist wars, forced assimilation and ‘religious cleansings’ that characterized the development of nation-states in the democratic, ‘civil’ West. Hanna Arendt decried the fate—which included her own—of those who fell between the grids of a world order of states that defined citizenship in national terms (Arendt 1951).

However, these remained marginal voices in a scholarly choir that praised the inclusionary character of nationalism and the nation-state. Little attention was given to the making of the boundaries of this egalitarian and inclusive community: the struggles over who belongs to the nation and thus should enjoy equal rights before the law, be called upon to participate in politics and be granted the privilege of having one’s own culture and language valued and legitimated by school and state. Thus, the fate of those who end up on the other side of the boundary went almost unnoticed: those not treated as equals before the law but as aliens or second-class citizens; who’s political voice will be disregarded as that of ‘minorities’; who’s culture will be excluded from the national sanctuary of museums and school curricula; who’s language will not be understood by administrators, university professors, policemen or judges. The horizontal inequality between the estates of agrarian empires, so vividly portrayed by Ernest Gellner, is replaced by a vertical inequality. The privileged access to the modern state that some ethnic groups—turned into nations—enjoy is mirrored in the exclusion of those who are being declared aliens, ethnic minorities, or immigrants with no such privileged relationship to the state.

There are three reasons for the orthodox focus on the process of inclusion within rather than the exclusion at the borders of the national community. First, most authors take it for granted that nationalism leads eventually to the establishment of ethnonationally more or less homogeneous states, such as the French, German or Italian proto-types. This obscures that fact that most nation-states in the contemporary world are ethnically much more heterogeneous. Accordingly, questions of dominance and subordination along ethnic lines play a crucial role in most modern nation-states, especially in the postcolonial world. With the partial exception of Anderson and Michael Mann, however, most of the well-known authors took the ideal envisioned by nationalists—the congruence of nation and citizenry—as the average case to be explained by a theory of nationalism and the nation-state. Secondly, the warlike process of achieving this ideal state is, more often than not, taken as a by-product of specific, accidental historical developments and not given an analytically central place in the portrait of nationalism and the rise of the nation-state. Whatever the traumatic and deplorable historical circumstances, what matters, according to the orthodox account, is that the historical train arrives at the final station, the homogeneous nation-state. Finally, the analytical horizon is often reduced to what happens within the borders of a would-be national state, thus obscuring the process of boundary-making and its exclusionary nature. All three mechanisms—confounding the ideal for the average, the teleological reasoning and the caging of the analytical perspective—establish and support what has been termed ‘methodological nationalism’, a characteristic of much social science thinking in the post-war period (Wimmer and Glick Schiller 2002).

The Boundary-Making Perspective

There is, however, a respectable tradition of research which looks at the interplay between national inclusion and exclusion along ethnic lines, between democratic participation of co-nationals and the authoritarian domination of ethnic others (Young 1976; Williams 1989; Verdery 1994; Grillo 1998; Wimmer 2002; Kaufman 2004a; Mann 2005). This tradition aims to understand how the imagining of a national community is intertwined with the creation of ethnic or immigrant minorities and how these boundaries are reinforced and reproduced subsequently. The emphasis lies less on explaining the rise of the nation-state in the West, as in most of the classic accounts referred to above, than on the consequences that the spread of the nation-state had outside the area of its original development—thus building on a perspective that had been established by Kohn, Kedourie, Anderson and Meyer (Meyer 1997).

Once a state apparatus has been taken over by a nationalist movement, a process of nationalizing its basic principles of exclusion and inclusion is set in motion. This politics of ethnic boundary-making by nationalizing state elites can take on different forms, depending on power relationships, the nature of the ethnic mosaic on the territory of a new state and the relationship with the nation-building projects of neighbouring states who may host similar ethnic populations on their territory. Across all these variations, however, we can discern a common pattern: the new elites try to establish a distinction between a dominant ethnonational core, ‘the people’ considered to represent the legitimate foundation of the new state, and those who are seen as not belonging to that core and thus to the legitimate ‘owners’ of the state.

Why should the modern state be propelled to distinguish between national core and ethnic outsiders? Why does it try to nationalize criteria of access and membership? We should first mention its heightened capacity, compared to pre-modern governments, to exercise control over the population—through a unified administration, an integrated school system, a coordinated and centralized military and judiciary apparatus—and enforce its rules even in remote parts of the state’s territory (Hechter 2000: chs 3 and 4; Mann 1995). The modern state uses this technical power to enforce certain ethnic-national boundaries because it derives its legitimacy from the nationalist doctrine according to which it represents the will of ‘the people’ (Wimmer 2002: ch. 3). Defining the ethnic boundaries of this people is of utmost political importance because these boundaries now also determine who will and who will not be included by the legal, political, welfare and military institutions of the state. According to the nationalist doctrine, only full members of the nation have the right to be treated equally before the law, to participate in national politics, to be taken care of in case of illness or old age, to be defended against outside aggressors.

However, it would be exaggerating to maintain that pre-modern, non-national states were not interested in ethnic boundaries at all. Many of them were, in at least three ways. Some pre-modern states were based on some, albeit comparatively loose notion of ethnic homogeneity (for examples see Smith 1986: Part I). Secondly, the hierarchical strata of agrarian empires were, contrary to what Gellner assumed (Gellner 1983), sometimes defined in ethnic terms. The Spanish empire, to give an example, distinguished between peninsulares (Spanish-born settlers and administrators), criollos (New World-born individuals of Spanish descent), indios and negros. Thirdly, the early modern principle of cuius regio eius religio led to the first systematic attempts at homogenizing a population in religious terms—as seen in the expulsion of Jews from Spain under Isabella and Ferdinand, of Huguenots from France through the Edict of Nantes and the countless similar episodes after the principle had been adopted by the Treaty of Westphalia.

The change from empires to modern nation-states, however, implied three fundamental differences in the politics of ethnic boundary-making. First, the principle of ethnonational homogeneity and of the ethnic-national repre-sentativity of the ruling elite became de rigueur for the legitimization of authority. Thus, state elites now attempted to systematically homogenize their subjects in cultural and ethnic terms, usually by declaring their own ethnic background, culture and language as the ‘national’ core into which everyone else should aspire to melt. Secondly, stratifying ethnicity, dividing rulers from ruled, privileged groups from less privileged ones, was replaced by a vertical ethnic boundary that separated foreigners from nationals, national majority from ethnic minority. The trans-ethnic, universal principles of imperial rule—in the name of Allah, of the spread of civilization, of revolutionary progress—had to give way to the particularist ideal of national self-rule. Third, for all these reasons, the state apparatus now embarked upon an active politics of diversity management that pre-modern empires were neither interested in nor capable of.

Several variants of this politics of diversity management have been studied (Young 1976; McGarry and O’Leary 1993; Young 1994; Esman 2004; Mann 2005), of which I will discuss only two: the creation of national communities through the policies of assimilation of ethnic others who are seen as potential members of the nation; and the enforcement of boundaries between national majorities and ethnic minorities in cases where assimilation is not seen as an option.

Nation-building: From assimilation to ethnic cleansing

Many new state elites have embarked upon a project of nation-building—or ‘nation-destroying’ if seen through the eyes of the objects of such policies (Connor 1972). Even fervent nationalists were often conscious of the limited reach of their own vision of the world. Massimo D’Azeglio, Cavour’s predecessor as prime minister of Piedmont, famously suggested, in the first meeting of the parliament of the newly united Italian kingdom, that ‘Italy is made. We still have to make Italians’… the same held true for Arabs, Turks, Germans, Nigerians, Mongolians and Frenchmen. Not only peasants were turned into Frenchman by the nationalizing state, to paraphrase a famous book title (Weber 1979), but also Aquitanians, Provençales, Occitanians and other linguistic groups that had maintained, at least to a certain degree, a sense of regional identity, sometimes based on or connected to histories of previous political independence, such as in the case of Savoy. In Poland around 1919, the state asked its population about their national background—an eminently political question in this region of disputed boundaries. In the East, which was later annexed by Russia, three-quarters of a million people answered by identifying as tutejsi, which roughly translates as ‘locals’. They spoke white-Russian dialects and adhered to the Orthodox faith, but did not see themselves as Russians or white-Russians. They certainly do today (Hroch 1985: 166). Other examples abound: the various Slavic and Albanian groups in the Peloponnese became Greeks; Wendish peasants in Eastern Germany, Germans; Copt-speaking communities, Arab-speaking Egyptians (Deutsch 1953: 94).

To be sure, not all attempts at nation-building were successful. In Somalia, the idea of a Somali nation as a community of political destiny has not had much success in overarching and erasing clan and regional identities (Rothchild 1995). The failed nationalizing projects of Czechoslovakia or Yugoslavia (Sekulic et al. 1994) are other examples. The unsuccessful attempt by the Kemalist state to declare Kurdish-speakers ‘mountain Turks’ adds to this list.

Nation-builders may employ various strategies to overcome existing ethnic divisions (cf. the similar typology of McGarry and O’Leary 1993). They may simply push them aside and propagate the new idea of a national community of solidarity, using the classical tools of school, army and administration to teach and propagate this vision. ‘Invented’ traditions, flags, symbols, anthems may help in achieving this end, as an enormous literature in the wake of the seminal volume edited by Hobsbawm and Ranger has shown (Hobsbawm and Ranger 1983).

Secondly, a nationalizing state may actively encourage ‘mixture’ and ‘amalgamation’ of various ethnic groups and their cultures into the melting pot of the grand nation. Mexico’s ideology of mestizaje, which was supposed to create, in the words of Mexican philosopher and longtime minister of education José Vasconcelos, a ‘cosmic race’, is one variant of this strategy (cf. Wimmer 2002: ch. 6); the Brazilian state’s ideology of ‘whitening’ to deepen Brazil’s ‘racial democracy’is another (Skidmore 1993 [1974]).

Thirdly, forced assimilation may be the means of transforming the mosaic of local ethnic and religious identities into the national picture of a homogeneous population—from a Kokoschka to a Mondrian, to play with a gellnerian metaphor. The Bulgarization of Turkish names in the 1980s may serve as a rather dramatic recent example. Another is the successful absorption and total assimilation of a large group of mixed Dutch-Indonesian descent that had fled to Holland after the archipelago gained independence. Despite a ‘racial’ difference, the policies of cultural assimilation through special education, dispersed settlement all over the country and controlled absorption into the labour market resulted in the disappearance of the group and the corresponding boundary—a formidable demonstration of the power of nation-building (Willems et al. 1990).

Other examples are the sedentarization of gypsies and the forced adoption of their children by majority parents, framed as policies of re-educating the deviant and degenerated race of the itinerants, a common practice throughout Europe. In Switzerland, a state-sponsored programme that forced gypsy children into foster families and asylums ran from 1926 to 1972. The Australian state, committed to its ‘white Australia’ policy, aimed at annihilating the aborigine population by a forced adoption programme which lasted from World War II to 1967 (Wolfe 2001: 872-3).

Fourthly, we may describe the various forms of ethnic cleansing as a strategy of homogenization. Ethnic cleansing by nationalizing states can be traced back to the two Balkan wars and from there to the ‘population exchange’ between Greece and Turkey, the extermination of large sections of the indigenous population of El Salvador during la matanza, the Holocaust, the mass massacres and evictions during the partition of India and straight up to the recent events in Bosnia and Darfur. Such ‘final solutions’ to the ‘problem’ of ethnic heterogeneity are a typically modern phenomenon pursued by a state apparatus that is dedicated to realize the ideal of ethnonational homogeneity by means of force and violence. The ethnic heterophobia of nationalizing states is, together with other more precise contextual factors that Michael Mann (2005) has identified, responsible for these moral nadirs of modern history—whether liberal nationalists like it or not (cf. O’Leary 1998).

More specifically, state terror and violence against minorities often serve the aim, as Appadurai (1998) has argued, of making clear, in a complex situation of overlapping membership, where the boundaries to the dangerous enemies lie. Violence thus cuts the tumour from the flesh of the nation’s body, to paraphrase language often used by the intellectual fathers and organizational masterminds of genocides. Gathering ‘Jews’ into the camps and ghettos of Nazi Europe, driving ‘Armenians’ onto the mountain roads of Anatolia, forcing ‘Tutsis’ into the churches and schoolhouses of the land of a Thousand Hills makes unambiguously clear who ‘Jews’, Armenians’ and ‘Tutsis’ are where intermarriage, assimilation and conversion have previously blurred boundaries.

The creation and management of ethnic minorities

Not all ethnic minorities are singled out, however, for a policy of assimilation or expulsion and not all such policies succeed in erasing all marks of ethnic difference from the landscape of identities. In many cases, minorities are meant to remain permanently outside of the sphere of national imagination but inside the state’s territory. Whether or not an ethnic group is envisioned as being a potential member of the national family depends on the structure of political alliance in the crucial early phases of nation-state formation, as recent research has shown. Anthony Marx explains how different constellations of political conflict and alliance led to inclusion of the black population into Brazil’s nation-building project and to their exclusion and domination in the United States and South Africa (Marx 1999). Similarly, I have tried to show that it depends on the reach of elite political networks which groups will be considered part of a nation to be. Thus, the trans-ethnic character of political networks in Switzerland explains the exceptional history of multi-ethnic nation-building. Those networks were limited to a Creole-mestizo elite in newly independent Mexico, which accounts for the exclusion of the vast majority of the indigenous populations from their nation-building project up to the Mexican revolution. The segregation of political networks along ethno-religious lines in pre-independent Iraq inhibited an Iraqi nationalism from emerging as a politically dominant force once the country was released from the colonial leash (Wimmer 2002).

What happens to those who remain outside of the national community, who are not meant to assimilate into it and are not driven from the territory through forced expulsion or relocation? We can postulate a certain pattern of how nationalizing states deal with permanent minorities on their territory. The first step often consists in creating or re-arranging ethnic categories to describe and administer those local groups that are perceived as not fitting into the national picture. Various local communities, peasant villages and urban communities organized along lines of neighbourhoods, local churches, or guilds, are grouped into larger ethnic entities. This helps to administer them more easily and to exercise some form of control over them by naming ‘representatives’ for these newly forged entities and co-opting them into the bureaucratic-administrative system. Over time, these newly created categories become inscribed into the administrative routines of the state. A recent example of such ethno-genesis is the emergence of the ethnic group of the Comanche (Hagan 1976:133) out of a variety of bands of different ethnic origin.

A major technique for minority creation is the census. A small literature on the politics of boundary-making through national censuses has emerged (Alonso and Starr 1987; Nobles 2000; Arel 2002). Recent examples from the United States are the creation and growing acceptance of the categories of ‘hispanic’ (Padilla 1986) or Asians’ (Espiritu 1992; Okamoto 2003), which originally made little sense from the point of view of those that were designated as such. Much earlier, the boundary between ‘black’ and ‘white’ was imposed by state agencies on a more diverse and complex system of classifications that had previously been recognised in the South (Lee 1999). However, attempts at imposing new ethnoymes by modern nation-states may also fail. The authorities of the homeland of Ciskei in Apartheid South Africa created the ethnonation of the ‘Ciskeians’ but remained the only ones to find the new category meaningful (Anonymous 1989).

The creation of ethnic categories is a first step in the process of singling out and ‘managing the problem’ of ethnic minorities—‘die jüdische Frage’, ‘the Negro problem’, ‘la cuestión indígena’, and so on. It is often followed by enforcement of the distinction between national majority and ethnic minority through the three related strategies of segregation, legalization and discrimination (see also, with regard to racialized groups, Wacquant 1997). By tying the distribution of life chances to membership in ethnic categories, segregation, legalization and discrimination powerfully affect the way individuals define themselves and are formidable tools to enforce the distinction between national majority and ethnic minority (Forsyth 1999).

Strategies of segregation aim at reducing the interaction between members of different ethnic categories. This greatly supports the plausibility of the categorization, since it creates or reinforces group boundaries and closure and thus makes the division of the social world appear natural and self-evident. Examples are the residential segregation rules imposed on minorities, the paradigmatic case being the creation of Jewish ghettos in early modern Europe—albeit this is an example of a medieval policy of community segregation rather than of modern minority management—and of black ghettos in North America after the First World War (Massey and Denton 1994). Another example is the marriage rules such as those prohibiting ‘white’ and ‘non-white’ marriages in the United States; the first laws were passed in 1661 in Maryland (Frankenberg 1993) and the last corresponding constitutional provisions were abolished in Alabama more than three hundred years later.

As this last example indicates, law represents a powerful mechanism for enforcing ethnic boundaries and the different statuses ascribed to the various ethnonational categories. The most important tool in the legal arsenal of boundary enforcement is citizenship laws (Brubaker 1992; Wimmer 2002: ch. 3). They tie universal human rights to a specific ethnonational community, as Hannah Arendt was the first to remark (Arendt 1951). And they made membership in such communities a matter of birth and inheritance. Once acquired, one’s citizenship becomes a permanent, ‘deep-seated’ characteristic to be transmitted to the next generation—born from ‘French parents, one would be and remain ‘French’ even if one had never set foot on the Hexagon. In contrast, ethnic minorities that were not considered part of the national majority were often relegated to the status of second-class citizens—such as African Americans in the South or Jews in pre-war Eastern Europe—or sometimes even completely deprived of all citizenship rights. Examples of the latter include the so-called Faili Kurds in Iraq, who in the 1970s were deprived of Iraqi citizenship and then driven over the border to Iran (McDowall 1996: 30), or of the Banyarwanda in Zaire, who were denaturalized in 1980 following a retroactive nationality law (Lemarchand 2004). The struggles over the citizenship status of Russians in the newly independent Baltic states are well known.

A final strategy of ethnic boundary enforcement is institutionalized discrimination: the unequal treatment of persons of different ethno-racial background in the day-to-day workings of the state administration—even when no restrictions are placed on formal citizenship rights. One of the most dramatic examples of negative discrimination is again provided by the American South before the civil rights movement. Discrimination by state authorities against ethnic minorities is widely reported from the newly independent Soviet successor states (Grodeland et al. 2000) and the developing world (Horowitz 1985: 194; Hyden and Williams 1994). Positive discrimination may also reinforce and institutionalize ethnic boundaries between a national majority and ‘underrepresented’ or, to the contrary ‘overachieving’ ethnic minorities, the most prominent examples being the minority quotas in the US educational system (Bowen and Bok 2000), in the Soviet bureaucracy (Vujacic and Zaslavsky 1991; Martin 2001), and in Malaysia, Nigeria, India and Sri Lanka (a critical view on these policies is provided by Sowell 2004; for other examples, see Horowitz 1985: 655f.).

Once the distinction between national majority and ethnic minority is established and enforced, members of the dominant ethnic group with a privileged relationship to the nationalizing state share a common interest in controlling the boundary (Rothschild 1981: ch. 5). Various strategies are known from the literature. One is establishing a ‘moving cultural target’ for assimilating groups, thus recreating a boundary with new diacritical markers when previous assimilation by minority groups has threatened to make it permeable or fuzzy. Examples such as the assimilating Jews in nineteenth-century Europe or Sanskritizing caste-less groups in India are discussed by David Laitin (Laitin 1995). Other authors have observed that in Guatemala and highland Mexico, the ethnic boundary persists despite considerable cultural assimilation, mostly due to the boundary policing strategies of ladinos and mestizos (Tax and Hinshaw 1970; Colby and van den Berghe 1969: 173; Smith 1975: 228; Reina 1966: 31f). Tellingly enough, those fully assimilated may be rejected as indios revestidos (‘disguised Indios’) or, in South America, as cholos (Aguirre Beltrán 1967: 301-11). In Northern Ireland Catholics were recognized by their gestures, body language and idiosyncrasies of grammar (Easthope 1976, cited in Banton 1983: 180; Burton 1978, cited in Jenkins 1997). In all OECD countries, citizens with ‘foreign’ names are confronted with very substantial forms of discrimination on the housing and labour markets, as a series of studies using the ILO methodology has shown (Taran et al. 2004). More imposing markers of identity to ensure non-ambiguity of boundaries include the star of Davis in Nazi Germany, the ethnic labels in Rwandan identity documents (Longman 2001), or the percentage of Indian ‘blood’ certified by government agencies in the United States (Meyer 1999). They are all formidable instruments to police the boundary and prevent its blurring through strategies of assimilation and passing.


Obviously enough, the degree to which nationalizing states enforce ethnic boundaries and discriminate between national majority and ethnic minorities varies from one state to another and from one historical period to another. Along a continuum from more inclusive to more exclusive constellations, we find, at one end, extreme cases of ethnocratic domination, such as Iraq under Saddam Hussein, which was effectively controlled by the members of his own clan and tribe (Baram 1997). Shi’as and Kurds were systematically excluded from higher ranks in the bureaucracy, party and the army, which did not hesitate to declare war against the civilian populations of the Kurdish North (Wimmer 2002: ch. 6). At the other end of the spectrum, we find the contemporary United States, which has officially abandoned the exclusive ethnocracy of Anglo-Saxon Protestantism and embarked upon a remarkable programme of multi-ethnic nation-building through the official recognition (albeit not social inclusion) of an ever-greater number of ethnic and racial minorities (Kaufman 2004b).

However widely the boundaries of the national community are imagined, however, it remains a bounded community, with the large majority of the world’s population on the outside. The institution of citizenship is the legal tool to enforce social closure along national lines even in cases where the doors of assimilation or of recognition as minorities are held widely open. In some highly integrative nation-states, such as Switzerland, which managed to build up a multi-ethnic, multi-religious national community and where ethnic domination between citizens is largely unknown, one-fifth of the resident population are systematically and legally discriminated against as foreign nationals—and thus denied the right to vote, the right to choose one’s profession and place of residence freely and the right to be taken care of by the welfare state in case of lifelong dependency (Wimmer 2002: ch. 8). Characteristically, whenever the legal discrimination between citizens and non-citizens was reduced, access to the national territory for immigrants became more selective and restrictive. The more the nation opens its gates for established immigrants, the more it closes its borders to those left on the outside (cf. also Lucassen 1995). We should therefore be careful in taking the recent decline in legal discrimination of immigrants as a sign that the inclusive logic of the modern nation-state has finally won over its more shadowy, exclusionary sides (for such a view, see for example Joppke 2004).

In less domestically inclusive nation-states, we find many of the same mechanisms of exclusion between different ethnic segments of the state’s citizens rather than between the latter and immigrants. Where the political networks of the nationalizing state elite did exclude large sections of the population, an overarching, inclusive mode of imagining the nation à la Suisse could not develop. The new state classes then use ethnicity as a basis for mobilizing a political following and in turn favour their co-ethnics when it comes to deciding who gets a government job, where to build a hospital or a bridge, whom to give justice to in a trial or whom to admit to the newly founded universities. The state administration, the school system, the army are thus compartmentalized along ethnic lines and ethnic discrimination and favouritism flourish. Depending on power relationships and the waxing and waning of political alliances, larger or smaller groups may gain control of the state apparatus and successfully drive others from the sources of power. In extreme cases such as Syria, Iraq or Burundi, a demographic minority may be in total control of the state and its repressive apparatus and thinly veil the authoritarian ethnocracy with a nationalist discourse appealing to all citizens of the country.

Wherever a society is situated on this continuum of variations—and other dimensions could easily be added—they are characterized by some form of closure and exclusion along ethnonational lines. These shadowy sides of the modern nation-state have remained largely unexplored by the classic works that have shaped the historical sociology of nationalism. Whether or not such exclusion can be defended on normative grounds, as the political philosophy of liberal nationalism maintains (Miller 1995), is an entirely different matter—as is the more general question of whether the exclusionary features of the nation-state highlighted in this chapter will fade away as the universalizing logic of the rule of law further unfolds. It may suffice to note here that the ‘de-nationalization of the modern state that many social scientists have noted during the 1990s and some have interpreted as signs the coming of a ‘post-national’ age (e.g. Soysal 1994) has been reversed in some noticeable cases—from the United States to the Netherlands.