Ethnicity versus nationalism

Thomas Hylland Eriksen

Journal of Peace Research, vol. 28, no. 3, 1991

Abstract

The aim of this article is to identify some conditions for peaceful coexistence between the state and populations in multi-cultural societies. Initially, the concepts of ideology, nationalism and ethnicity are examined briefly. It is argued that a successful ideology, such as a nationalist or an ethnic one, must simultaneously legitimize a social order, i.e. a power structure, and provide a meaningful frame for the articulation of important, perceived needs and wishes of its adherents. A few empirical cases are then considered. These examples, ranging from the multi-cultural island-states of Mauritius [treated at greater length here] and Trinidad & Tobago [dealt with in a different vein here] to the Saami (Lappish) minority situation in northern Norway, involve conflicts between nation-states and ethnic groups, and between different ethnic groups within the nation-state. Some conflicts and the methods employed to resolve them, are compared. The uniqueness of nationalism as a modern, abstract "binary" ideology of exclusion and inclusion, and its powerful symbolic as well as practical aspects, are stressed and contrasted with "segmentary" ethnic ideologies. Finally, the article proposes a list of necessary conditions for the peaceful coexistence of culturally diverse groups within the framework of a modern nation-state.

The conclusion is that the main responsibility lies with the state insofar as it possesses a monopoly of political power and the legitimate use of force. State policies should genuinely attempt to decentralize power while at the same time recognizing the right of being cultural distinctive, even in matters relevant for political discourse. State nationalism should not be symbolically linked with the collective identity of only one of the populations.The culturally homogenizing tendencies of nationalism must in other words be counteracted through institutional arrangements which secure some form of ethnic autonomy and encourage cultural pluralism. The alternatives are violent suppression and the enforced assimilation of culturally distinctive groups.

1. Aims and concepts

Virtually every modern nation-state is to a greater or lesser extent ethnically divided. This frequently implies a potential for various forms of conflict - from armed conflicts to autonomist movements and political segregation along ethnic lines (see e.g. Horowitz, 1985; Wallensteen et al., 1986, for overviews).

Two central aspects of the contemporary global situation indicate that ethnic conflicts may be of increasing relative importance. First, the East-West conflict is presently on the wane. The recent changes in the global political system both call the attention of scholars and policy-makers to conflicts which cannot be understood within the idiom of the Cold War, and further directly stimulate the growth of a wide range of new ideological movements in the former Eastern Bloc, many of them drawing explicitly on nationalist and ethnic rhetoric. Secondly, processes of modernization in the Third World lead to ever more encompassing confrontations between dominant nationalisms and other ideologies in many countries.

Ethnic ideologies are at odds with dominant nationalist ideologies, since the latter tend to promote cultural similarity and wide-ranging integration of all the inhabitants of the nation-state, regardless of their ethnic membership. It can therefore be instructive to contrast ethnic ideologies with nationalism in contemporary nation-states. Through examples from ethnically complex nation-states, the variable content and social impact of such different ideologies are explored. The purpose is to identify some conditions under which culturally justified conflicts may arise within modern nation-states,1 and to suggest conditions for their resolution or avoidance. The general perspective is from within; that is, ideologies and practices are considered largely from the point of view of their adherents. It will be argued, further, that the multi-ethnic nation-state is no contradiction in terms - that it may indeed be a viable and stable political entity.

1.1. Ideology
The central concept of ideology is treated throughout as a double concept. On the one hand, ideology serves to legitimize a particular power structure and in this respect conforms to a conventional marxist view. On the other hand, ideologies necessarily derive their popular, potentially mobilizing force from their ability to organize and make sense of the immediate experiences of their adherents; they cannot, therefore, be regarded simply as forms of false consciousness2. The term ideology can profitably be used in the plural insofar as people evaluate available ideologies critically and compare them through choosing their strategies and practices. The final outcome of a competitive situation involving two or several ideologies depends on their respective persuasive power among their frequently ambivalent audiences. It follows from this that an analysis of particular ideologies, in this case nationalist and ethnic ones, demands an understanding of the lives of the followers of the ideologies in question. An analysis of ideology cannot solely consider the properties of the political system and the ideational content of the ideologies themselves, since beliefs and other forms of knowledge contribute to the reproduction of society only to the extent that they are embedded in interaction.

1.2. Nationalism and ethnicity
In its most basic sense, ethnicity refers to the social reproduction of basic classificatory differences between categories of people and to aspects of gain and loss in social interaction. Ethnicity is fundamentally dual, encompassing both aspects of meaning and of politics. Ethnicity is, however, a concept which refers to a multitude of socio-cultural phenomena. It may appear at our doorstep any time and vanish in a matter of seconds: for instance, my relationship with foreign students at the university has ethnic connotations and can thus be viewed as an ethnic relationship. They enter my office and go away; the duration of such an ethnic relationship can be less than half an hour. Similarly, my Pakistani-Norwegian grocer enters my life to a very limited degree, and the ethnic aspect of our relationship is nearly negligible (although never entirely absent). On the other hand, the term ethnicity can also refer to large-scale, long-term political processes such as the relationship between blacks in the United States and the US nation-state; it can refer to intricate trade networks throughout the United Kingdom or to the religious sentiments of individuals; sometimes ethnicity becomes nationalism historically, sometimes it vanishes altogether, and so on. In a certain sense, ethnicity is created by the analyst through the questions she poses in her research. What makes ethnicity a more interesting concept in the contexts considered below than say, class, is its empirically pervasive nature: Ethnicity can, if sufficiently powerful, provide individuals with most of their social statuses, and their entire cultural identity can be couched in an ethnic idiom.

In social anthropology and urban sociology, ethnicity has been analysed extensively at the level of interpersonal action, at the level of the township, at the level of factioning and riots, etc. In this restricted context, I focus on ethnic phenomena which involve nation-states directly or indirectly, and where ethnicity is manifest through political organization3.

I will treat nationalism and ethnicity as ideologies which stress the cultural similarity of their adherents. By implication, nationalists and ethnicists will, in a situation of conflict, stress cultural differences vis-à-vis their adversaries.4 The distinction between the two may therefore appear to be one of degree, not one of kind - particularly since many political movements are commonly perceived as being both nationalist and ethnic in character. What to make, then, of say, autonomist movements in Soviet Central Asia, proclaiming Azeri or Armenian nations, insofar as their official status is that of ethnic minority groups? The difference, in this case, is in the eye of the beholder. A self-proclaimed nationalist holds that state boundaries should be identical with cultural boundaries (see Gellner, 1983, for an excellent discussion of the concept). If such claims are not acknowledged as legitimate by the political authorities of the state in which she resides, they will perceive her, and define her, as an ethnic revivalist. In other words, the major difference between ethnicity and nationalism lies, as they are delineated here for convenience, in their relationship to the state. Unsuccessful nationalisms therefore tend to become transformed into ethnicities whose members reside more or less uncomfortably under the aegis of a state which they do not identify with their own nationality or ethnic category. This has happened to certain indigenous peoples of autonomist persuasion, to many of the "one hundred and four peoples" of the Soviet Union, and to some extent, to the white minority of Zimbabwe, whose variety of nationalism in the end lost the battle for political and cultural hegemony.5 Many of the ethnics6 condemned to such a fate eventually vanish through migration, extermination or cultural assimilation. On the other hand, there are ethnicities and ethnic movements whose ultimate aim is not - and can never be - full statehood. Urban minorities in Europe and North America are obvious examples; such groups are in many respects integrated in ways radically different from groups who claim territorial rights. Finally, we need to distinguish provisionally between those indigenous "Fourth World" peoples favouring autonomy but not full statehood, and those ethnic minorities (or nations without a nation-state of their own) whose legitimized leaders or spokespersons work for total political independence. Ethnic minority situations are frequently ambiguous in this regard. Greenlanders make up an ethnic category to the extent that their destiny is intertwined with that of metropolitan Denmark, but they constitute a potential nation-state to the extent that they collectively vie for full political autonomy. Their identity as Greenlanders can therefore be regarded both as an ethnic and as a national one, depending on the analytical perspective. This contradiction is naturally manifest also in the experience of many Greenlanders. The widespread switching between ethnic and national identities in Poland and other Central European countries in the 1920s and '30s further exemplifies the contextual character - and empirical interrelatedness - of ethnicity and nationalism as popular ideologies (see Neumann, 1990).

Nationalism entails the ideological justification of a state, actual or potential. Judged on this criterion, ethnicity can sometimes be interpreted as a form of stagnant nationalism which may eventually, or periodically, become manifest as nationalism.

The social importance, the "semantic density", of such ideologies varies immensely historically, geographically, contextually and situationally - both at the level of the individual and at the level of the political system. The mere fact that "nationalism exists in country X" or "ethnic minority groups live in state Y" does not necessarily imply that such ideologies play an important part in the lives and/or political processes encompassed by the system. The relative importance of nationalism and ethnicity is an empirical question, and the cases discussed below suggest the circumstances under which they can assume importance.

2. Nationalism vs. ethnicity

Viewed geopolitically, nationalism is an ambiguous type of ideology. It can be aggressive and expansionist - within and outside of state boundaries; and it can serve as a truly peacekeeping and culturally integrating force in a nation-state or a region. Nationalism is frequently regarded by liberal theorists as a universalist kind of ideology emphasizing equality and human rights within its polity, but it can just as plausibly be seen as a kind of particularism denying non-citizens or culturally deviant citizens full human rights and, in extreme cases, even denying them membership in the community of humans (see Giddens, 1987, pp. 177 ff. for a critical discussion of these aspects of nationalism). Depending on the social context, then, nationalism may have socio-culturally integrating as well as disintegrating effects; it sometimes serves to identify a large number of people as outsiders, but it may also define an ever increasing number of people as insiders and thereby encourage social integration on a higher level than that which is current. There is nothing natural or historically inevitable in this. For the nation is an invention and a recent one at that; to paraphrase Anderson (1983), it is an imagined community; it is not a natural phenomenon, despite the fact that the object of every nationalism is to present a particular image of society as natural. Nationalism is ever emergent and must be defended and justified ideologically, perhaps particularly in new states, where alternative modes of social integration, usually on a lower systemic level, remain immediately relevant to a large number of people. The "multi-ethnic" or "plural" state is the rule rather than the exception (Smith, 1981); however, cultural plurality can evaporate historically, it can lead to the formation of new nation-states, it can lead to conflict between ethnics or between state and ethnic, or it can be reconciled with nationhood and nationalism.

2.1. The emergence of nationalism
Historically, an important part played by nationalist ideologies in many contemporary nation-states has been to integrate an ever larger number of people culturally, politically and economically. The French could not be meaningfully described as a "people" before the French revolution, which brought the Ile-de-France (Parisian) language, notions of liberal political rights, uniform primary education and not least, the self-consciousness of being French, to remote areas - first to the local bourgeoisies, later to the bulk of the population. Similar large-scale processes took place in all European countries during the 19th century, and the modern state, as well as nationalist ideology, is historically and logically linked with the spread of literacy (Goody, 1986), the quantification of time and the growth of industrial capitalism. The model of the nation-state as the supreme political unit has spread throughout the 20th century. Not least due to the increasing importance of international relations (trade, warfare, etc.), the nation-state has played an extremely important part in the making of the contemporary world. Social integration on a large scale through the imposition of a uniform system of education, the introduction of universal contractual wagework, standardization of language etc., is accordingly the explicit aim of nationalists in e.g. contemporary Africa. It is, of course, possible to achieve this end through contrasting the nation with a different nation or a minority residing in the state, which is then depicted as inferior or threatening. This strategy for cohesion is extremely widespread and is not a peculiar characteristic of the nation-state as such: similar ideologies and practices are found in tribal societies and among urban minorities alike. Insofar as enemy projections are dealt with in the present context, they are regarded as means to achieve internal, national cohesion, since international conflicts are not considered.

Nationalism as a mode of social organization represents a qualitative leap from earlier forms of integration. Within a national state, all men and women are citizens, and they participate in a system of relationships where they depend upon, and contribute to, the existence of a vast number of individuals whom they will never know personally. The main social distinction appears as that between insiders and outsiders; between citizens and non-citizens. The total system appears abstract and impenetrable to the citizen, who must nevertheless trust that it serves his needs. The seeming contradiction between the individual's immediate concerns and the large-scale machinations of the nation-state is bridged through nationalist ideology proposing to accord each individual citizen particular value. The ideology simultaneously depicts the nation metaphorically as an enormous system of blood relatives or as a religious community, and as a benefactor satisfying immediate needs (education, jobs, health, security, etc.). Through this kind of ideological techniques, nationalism can serve to open and close former boundaries of social systems. Some become brothers metaphorically; others, whose citizenship (and consequently, loyalty) is dubitable, become outsiders. In Figure 1 below, the peculiar communicational features of nationalism and the nation-state are depicted crudely and juxtaposed with the Gemeinschaft-like kinship or locality-based organizations they seek to replace and imitate in their symbolism. The major difference is that nationalism communicates through impersonal media (written laws, newspapers, mass meetings etc.) whereas kinship ideology is communicated in face-to-face interaction. The former presupposes the latter as a metaphoric model.

------------------------------
Fig. 1 about here -- sorry about this, but it seems to have been lost during conversion. It will be inserted when it has been excavated (-THE)
------------------------------

Nationalism is ideally based on abstract norms, not on personal loyalty. Viewed as a popular ideology, nationalism is inextricably intertwined with the destiny of the nation-state. Where the nation-state is ideologically successful, its inhabitants become nationalists; that is, their identities and ways of life gradually grow compatible with the demands of the nation-state and support its growth. Where nationalism fails to convince, the state may use violence or the threat of violence to prevent fission (that is, in the modern world, the potential formation of new nation-states on its former territory). The monopoly on the use of legitimate violence is, together with its monopoly of taxation, one of the most important characteristics of the modern state; however, violence is usually seen as a last resort. More common are ideological strategies aiming to integrate hitherto distinctive categories of people culturally. Since national boundaries change historically, and since nations can be seen as shifting collectivities of people conceiving of their culture and history as shared, this is an ongoing process. Ethnic groups can vanish through annihilation or more commonly, through assimilation. They may also continue to exist, and may pose a threat to the dominant nationalism in two main ways, either as agents of subversion (they do, after all, represent alternative cultural idioms and values - this was how the Jews of Nazi Germany were depicted) or as agents of fission (which is evidently the case with Baltic nationalists).

Nationalist strategies are truly successful only when the state simultaneously increases its sphere of influence, and responds credibly to popular demands. It is tautologically true that if the nation-state and its agencies can satisfy perceived needs in ways acknowledged by the citizens, then its inhabitants become nationalists. The main threats to national integration are therefore alternative social relationships which can also satisfy perceived needs. There are potential conflicts between the nation-state and non-state modes of organization, which may follow normative principles incompatible with those represented by the state. This kind of conflict is evident in every country in the world, and it can be studied as ideological conflict, provided ideology is not seen as a system of ideas, but as sets of ideological practices. Typical examples are African countries, where "tribalism" or organization along ethnic lines is perceived as a threat (by the nation-state), or as an alternative (by the citizens), to the universalist rhetoric and practices of nationalism. From the citizen's point of view, nationalism may or may not be a viable alternative to kinship or ethnic ideology (or there may be two nationalisms to choose between, e.g. a Soviet and a Lithuanian one) - and she will choose the option best suited to satisfy her needs, be they of a metaphysical, economic or political nature. The success or failure of attempts at national integration must therefore be studied not only at the level of political strategies or systemic imperatives; it must equally be understood at the level of the everyday life-world. In a word, the ideological struggles and the intra-state conflicts, as well as the context-specific options for "the good life", shape and are simultaneously rooted in the immediate experiences of its citizens, - and the analysis must begin there.

2.2. Binary and segmentary ideologies
Nationalism, as the ideology of the modern nation-state, ostensibly represents universalist norms domestically, as opposed to particularist norms. A common type of conflict entailed by this opposition occurs in the labour markets of many countries. According to kinship-based and ethnic ideologies of the kind prevalent in many African countries, employment should normally be provided by members of the extended lineage (or the ethnic). According to nationalist ideology, employment should be allocated democratically and bureaucratically, according to formal qualifications, regardless of the personal relationship between employer and applicant. These contradicting norms pervade labour markets in many parts of the world. The example further indicates that an individual who perceives the differences, will adhere to the ideology whose implications are more beneficial to himself (see Helle-Valle, 1989; Eriksen, 1988, for fuller discussions). The general point to be made here is that whenever nationalism is ideologically opposed to ethnic and kinship ideology, it will strive to present itself as just and fair according to abstract principles. Whether or not it succeeds in this respect depends on its ability to persuade people that it is beneficial to themselves (in some respect or other) that they subscribe to impartial justice of the kind represented by the state.

Contradictions between abstract norms of justice and concrete norms of loyalty occur in virtually every realm of social life in modern nation-states. In most states, variations on this theme form a central part of the discourse on ideology; the question concerns which type of social identity is relevant and ultimately, how the social world is constituted (see Larsen, 1987). A relevant question while considering different forms of incorporation and integration in some modern states, is therefore this: Under which circumstances are social identities, specifically ethnic identities, made relevant in conflicts in modern states, how do such conflicts arise, and how can they be resolved?

The general answer to this question, as will be evident from the examples and subsequent discussion, is that such conflicts evolve when agents act according to particularistic systems of segmentary oppositions, which either contribute to inequality or are justified by perceptions of inequality, and where invocations of cultural differences can serve to account for such strategies. Let me elaborate briefly. Segmentary oppositions in social integration function according to the general scheme first developed in Evans-Pritchard's analysis of mechanisms for the articulation and solution of conflicts among the Nuer of the Sudan (see Evans-Pritchard, 1940, particularly chapter 4). The general formula is: "It's I against my brother, my brother and I against our cousins; my cousins, my brother and myself against our more distant relatives, etc." In a modern, multi-ethnic society, segmentary oppositions could be expressed thus by a member of the X'es in country N: "It's I against my family, my extended lineage and myself against the rest of the X'es; further, it's all of us X'es against the other people and the state of N; but it's all of us citizens of N against the people of the country M." The pattern of competition and potential conflicts could be envisaged as one consisting of concentric circles; the general model is analog, for degrees of difference are made relevant. Unlike the digital model advocated by nationalism, dividing people into only two, mutually exclusive categories (insiders and outsiders), segmentary ideologies entail degrees of inside- and outsideness.

Through its official policies, the state will normally favour forms of organization incompatible with corporate action along ethnic or lineage lines; its way of classifying is different (digital or binary) and the system of segmentary opposition suggested is therefore incompatible with the organization of most nation-states. On the other hand, the state may itself represent a form of "lineage organization", if it is controlled by a dominant ethnic.

One of the examples below describes a society where the nation-state skilfully mediates between the two conflicting principles of social organization.

2.3. Compromise and hegemony: Mauritius and Trinidad
Nowhere is the notion of the nation as an imagined community more evidently true than in the colonially created states. Commonly invoked as examples of this are the new African nation-states (see e.g. Smith, 1983; Hobsbawm & Ranger, 1983), whose boundaries were randomly drawn a century ago, and whose nationalisms are of very recent origin. Even more striking are the culturally constructed nationalisms of societies which were never pre-colonial. Mauritius and Trinidad & Tobago are examples of such emergent nations. Both of these island-states, one in the Indian Ocean, one in the Caribbean, are ethnically heterogeneous and have always been; the very societies were created through the mass import of slaves and indentured labourers during the modern era, and they have been independent less than thirty years7. Until the 1960s, then, the wider identities of the inhabitants of these islands were colonial; the people knew that they were British subjects and that, to some extent, they were ruled from Britain.

Mauritius and Trinidad, demographically similar, have followed different courses in creating their respective nationalisms. Let us consider Mauritius first.

Mauritians are as a rule very conscious of problems related to ethnic differences. Their society is made up by groups originating from three continents and four major religions; there is no clear majority, and yet, the Mauritian nation-state has hitherto avoided systematic inter-ethnic violence (the one notable exception to this is the series of minor riots around independence in 1967-8). Yet Mauritians are, regardless of ethnic membership, determined to retain their ethnic distinctiveness. Rituals celebrating particular religions are widely attended, there is little intermarriage between groups, and there is currently an upsurge in popular interest in cultural origins: Hindi courses are held for Indo-Mauritians who have never learnt their ancestral tongue, Arabic is being introduced as the language of the mosque, an Organization of Afro-Mauritians was set up in the mid-eighties, etc. Simultaneously, there are strong "centrifugal" forces at work encouraging a Mauritian nationalism which is identified with uniformity in cultural practices: the emergent industrial system of production demands uniformly qualified, mobile labour, which in turn requires a standardization of education. National radio, TV and newspapers increasingly influence the form and topic of discourse about society, and the political system takes little account of ethnic differences. Although parties tend to be ethnically based, their rhetoric is nationalistic, and public political discourse is issue-oriented. The Mauritian state, recognizing the immanent dangers of the potential dominance of one ethnic, has taken great pains to develop a set of national symbols which can be endorsed by anybody, and which are thus not associated with one particular ethnic.8 Caught between different, sometimes conflicting ideological orientations, Mauritians choose situationally between the universalist ethics of nationalism, and the particularist ethics of ethnicity. In matters relating to employment and marriage, ethnicity is still a major variable, but it is constantly counteracted by discourse proclaiming the superiority of abstract justice and non-particularism. The openness of Mauritian discourse, public and private - in particular, the fact that ethnic conflicts and cultural differences are acknowledged everywhere as facts of social life, coupled with the absence of a hegemonic ethnic - indicate the kind of inter-ethnic compromise realized in Mauritius. Although there are important contradictions between ideologies of ethnicity and ideologies of nationalism at the level of individual action, the contradictions are to a great extent reconciled on the national political level, where compromise, justice, equal rights and tolerance are emphasized. Ethnically based systems of segmentary oppositions are encouraged outside of the educational, political and economic systems, where the virtues of meritocracy are continuously stressed. Current economic growth certainly contributes to accounting for the stable political situation, but it is by no means certain that recession would automatically lead to the breakdown of the currently shared rules for inter-ethnic relations. Processes of national integration stressing the necessity of inter-ethnic compromise were evident over a decade before the current economic boom, which began in the mid-eighties. The ethnic equilibrium may be fragile, but the political system has repeatedly proven capable of coping with conflict.

Strategies of compromise, characteristic of Mauritian society, are - as we are painfully aware - by no means the inevitable outcome of ethnic plurality. In Trinidad, ethnicity takes on a different meaning. Like in Mauritius, ethnicity is important in many situations in daily life as well as in politics, but it is not always acknowledged as such. Strategies of playing down ethnicity as a relevant topic are frequently employed in public discourse; this kind of strategy is typical of dominant groups in many societies. The symbolic content of Trinidadian nationalism is a good example of this.

1956 may be said to have been the year when Trinidadian nationalism emerged. For the first time, a pro-independence nationalist political party (PNM; People's National Movement) won the general elections. What was the content of its nationalism? The main slogan was Massa Day Done; a reasonable translation would be "our era as colonial servants is over". Notions of self-reliance and self-determination were in themselves powerful official national symbols. To the average urban Trinidadian, these ideas were extremely attractive, and nationalism was a strong and intoxicating force in Trinidadian public life throughout the sixties. But to whom? Who were the Trinidadians whose community was created imaginatively by the PNM leaders? Looking more closely, we find several social schisms implicit in Trinidadian nationalism, the most important of which runs between blacks and Indians. The blacks are the larger group (but only slightly larger than the Indo-Trinidadian) and have held the political power since before independence. Indians were politically and economically marginal, largely confined to the canefields. The towns were dominated by blacks; the radio played black music, and the national heroes, the calypsonians, were nearly invariably black or brown Creoles. The core electorate of the PNM were the urban black. So what to make of the part played by Indians in early Trinidadian nationalism? - It is a fact that they were for generations alienated from power and influence; only since around 1960 have the majority of Indo-Trinidadians taken part in the national project of Trinidad & Tobago to the extent that they have received compulsory elementary schooling and extensive career opportunities in the national political and economic system. During the last 20 years, and particularly during the 1980s, there has been a strong wave of Indian ethnic revitalization in Trinidad. Culturally self-conscious Indians claim that Trinidadian nationalism is a black ideology with which they cannot identify without losing their identity as Indians. A question frequently raised critically by blacks as a reply to this accusation, has been whether it is possible to be simultaneously Indian and Trinidadian. Here it should be noted that it would be absurd to ask whether it is possible to be simultaneously black and Trinidadian, since black culture is identified with national culture. In other words, the issue deals with responses to state-monitored attempts at cultural assimilation. Defining Indian culture as anti-national, blacks confirm their own as that of the Trinidadian nation. Less powerful than the blacks politically and in public culture, but still a large category of people now well integrated economically and politically, Indians react partly through declaring their status as that of an oppressed minority, partly through allowing themselves to become assimilated, and partly through arguing that their customs and notions, too, form part of national Trinidadian culture. The latter line of argument recalls the official policies of the Mauritian state, where the desirability of cultural pluralism is emphasized (provided it does not conflict with bureaucratic and capitalistic values). In Trinidad, the legitimacy of ethnic systems of segmentary oppositions is rejected in official discourse, but there is also a systematic inequality of power between ethnic groups. Stressing an ideology of equality in an environment of inequality is characteristic of dominant groups.9 The unequal distribution of power thus seems to account for the significant variations in the techniques used for handling ethnic differences in Trinidad and Mauritius.

Trinidad and Mauritius were chosen as examples because they are in many ways similar, yet display two very different solutions to the problem of multiculturalism versus nationalism. Both maintain ethnic peace on the national level; neither has currently an ethnic problem involving systematic physical violence, whether between individuals or between state and individual.10 However, the Trinidadian model structurally resembles that of less successful multi-cultural societies. The United States is an example of such a society, where all citizens, regardless of race and religion, have the same basic rights, but where rules of social mobility favour some but not all, and where nationalism is identified with cultural symbols of the hegemonic group. Thus, blacks and Hispanics are disqualified in a way structurally similar to that of Indians in Trinidad. Ideologies of equality in this way serve to justify inequality whenever they fail to account for cultural differences. Additionally, the US nation contains - or encapsulates - ethnic minorities whose cultural distinctiveness is in important ways incompatible with the requirements of national society. This is clearly the case with Amerindian groups, who more obviously than blacks and Hispanics suffer culturally from the intrusion of nationalistically justified imperatives. Participation in the capitalist economy, the schooling system etc. may contradict important features of their way of life. In the case of such groups, the problem is not only one of inequal distribution of power; it is perhaps chiefly a problem of cultural and political autonomy. In this kind of state/ethnic relationship, the powerless, "muted" group may demand the right to be culturally different in confrontation with the state, in a context of overwhelming power asymmetry.

We now turn to a description a conflict of this type, which is nevertheless atypical - and therefore interesting analytically - because this state is in principle willing to take part in dialogue with the minority.

2.4. Indigenous peoples and state penetration: The example of northern Norway
The relationship between the Norwegian state and the Saami (Lappish) minority in Northern Norway is complex, and a brief outline of some aspects of the contemporary relationship will have to suffice.

Since the start of the post-war wave of ethnic revitalization among the Saami (roughly since the 1950s), the Saami organizations' demand for cultural and political self-determination has grown in intensity. The ethnic processes taking place in territories settled by Saami are similar to nationalist movements. There is a current resurgence in popular interest in the recodification and glorification of their stigmatized cultural tradition, and there has consequently been an increasingly articulated dichotomization in interaction with Norwegians and mainstream Norwegian culture and society (Eidheim, 1971). These processes are similar to those of the burgeoning Norwegian nationalism of the mid-19th century (see Østerud, 1984). There is one major difference, however, between indigenous rights groups such as the Saami, and classical nationalist movements. The Saami do not presently demand full sovereignty; they do not intend to set up a Saami nation-state. Orienting themselves towards international law, the Saami nevertheless fight for self-determination in matters considered vital to their survival as a culture-bearing group. In this they have aims comparable to those of indigenous groups in the Americas, in Australia and elsewhere. This would have to include an institutionalization of the relationship between the state and themselves built on an official recognition of their right to self-determination as an indigenous people and a recognition of the state's duty to grant these special rights.

A profound dilemma for the Saami movement, then, is rooted in the rather paradoxical situation that the state against which they fight for self-determination must also, in the last instance, be accepted as an ultimate guarantor for the very same rights that it threatens. Norwegian policies vis-à-vis the Saami, insofar as they have acknowledged the Saami as a culturally distinctive minority, have until recently focused on questions of juridical rights defined within the national Norwegian idiom. The Saami movement was not successful until it was able to present itself effectively as the representative of a Fourth World people and present its case in the idiom of international law, although an institutionalized division of power between the nation-state and the newly elected Saami parliament (1989) is now emerging. Unlike the situation in Mauritius and Trinidad, where negotiation takes place in a shared idiom of discourse, the State-Saami context is still one where there is not always agreement regarding the very rules of the game (see Eidheim, 1985, for a full discussion).

This dilemma goes to the core of a central problem of nationalism: the nationalist tendency towards cultural homogenization, and the accompanying tendency to frame every political question in the state's legalistic, bureaucratic form of discourse. This disqualifies culturally distinctive groups from full participation, and simultaneously promotes their assimilation. The process taking part in the northernmost part of Europe is an interesting one from this point of view, since the state is here in principle sympathetically inclined to a dialogue with a well articulated, culturally distinctive group. The recent founding of an elected Saami parliament (with limited power) may enable Saami to articulate their political demands in their own terms. Such an attempt may, however, be unsuccessful for two reasons: First, the structure of the Saami parliament is modelled on Norwegian political institutions - it resembles a county council - which may result in an internationalization of the form of Saami politics. Secondly, the necessary discourse with the Norwegian state must probably be kept within a Norwegian idiom focussing on juridical rights and duties.

The ideological situation of contemporary Saami is a difficult one. Simultaneously a Norwegian citizen and member of the modern world on the one hand, and a member of a cultural minority on the other, the average Saami is faced with a number of difficult choices. He is culturally and ideologically opposed to, and yet economically and structurally dependent on, the Norwegian state. It is relatively easy for Saami to assimilate, to become Norwegian, and many do. This should not be taken as an indication of Norwegian nationalism among the indigenes - there is little in their history and contemporary situation encouraging such an ideology, - it should rather be seen as a tangible indication of the division of power and opportunities in a modern state society. Unless a truly ingenious model of autonomy within the national state is developed, the structural imperatives for Saami to assimilate will probably work in favour of assimilation in the long run, and the Saami ethnic may eventually vanish. The dominant Norwegian nationalism will in that case emerge victorious; not primarily as a belief system, however, but as a power structure and a set of unified, integrating political, economic and domestic practices. Ethnically based systems of segmentary oppositions (Saami values/principles against Norwegian values/principles) will in this case be invalidated: if they eventually cease to be relevant in all kinds of interaction, then the Saami minority will have been been fully assimilated.

On the other hand, if the principles of international law concerning the rights of indigenous peoples are fully acknowledged in the practices of the Norwegian state, then the Saami may survive as a culture-bearing group within the territory of the Norwegian state, which may thereby avoid otherwise inevitable accusations of cultural genocide.

It should be noted, finally, that the Saami movement draws much of its legitimacy from political entities not constituted by the state or by a system of states (such as the UN or the Common Market), but from international Fourth World organizations and informal networks, and through transnational public support. Fourth World politics thus serves as a countervailing influence - however modest - to the state's monopoly of political power in the contemporary world.

2.5. National attitudes to ethnic minorities
Ethnic minorities pose a problem to the national state to the extent that they communicate their distinctiveness in contexts where this distinctiveness is incompatible with requirements of the nation-state, notably those referring to formal equality and uniform practices. The minorities, as is evident from the example of the Saami, are faced with threats of more or less enforced assimilation. The intensity of such pressures to assimilate is generally linked to the degree of modernization and the level of state integration in national society. Where ethnic minorities could formerly be ignored and left alone, they are, in the modern world, defined from the outside as citizens of the national state, and are thus given equal rights by an administrative apparatus unable to - or at least unwilling to - grant its subjects unequal rights on grounds of cultural distinctiveness. Indigenes or other ethnically distinctive populations may, too, serve as negative symbols of the nation, in which case the relationship is chiefly one of conflict or oppression, not one of possible compromise. This was clearly the case in Nazi Germany, where Germanness was defined in contrast to the un-Germanness of Jews, Gypsies and Slavs (and this still holds to some extent in modern Germany; see Forsythe, 1989). On the other hand, ethnic minority populations can also be used symbolically in an apparently opposite way, as metaphors of the nation. This seems to be the case in Australia, where aboriginals "have become so close to the centre of nationalist thought that they have suffered from it" (Kapferer, 1988, p. 142). In emphasizing the purity and ancientness of aboriginal society, official Australia prevents their assimilation in a manner not dissimilar from policies of apartheid; that is, they are given differential treatment due to differences in culture (or race). That Aboriginals are not treated as equals by the Australian state is evident (Kapferer, ibid.), and Australian prejudices against people of non-Northern European descent indicate that Australian egalitarianism applies only to those perceived as the same kind of people(see Kapferer, 1988, pp. 183 ff.).

2.6. Autonomy or assimilation?
On the one hand, ethnic minorities may demand specific rights because of their distinctive culture and way of life. On the other hand, they may suffer systematic discrimination if they are granted such rights by the state. South African apartheid is an even more obvious example of this than the Australian policies vis-à-vis Aboriginals. When the "Bantustans" or "homelands" were created, black South Africans were formally allowed to refuse to contribute to the white economic system to which they were, inextricably, structurally tied. The teaching of African languages among blacks has also been encouraged in apartheid policy. This has enabled blacks to retain parts of their cultural heritage, and it has equally efficiently debarred them from political participation in South African society.11 Their systems of segmentary oppositions have been isolated from the wider social context of which they potentially form part.

It may seem, then, that neither solution is viable. If all citizens are to be treated equally, then cultural minorities are disqualified because their particular skills are ignored. But if citizens are treated unequally on the basis of cultural difference, then cultural minorities suffer discrimination because they lack certain rights granted the rest of the population. It may seem, then, that ethnic minorities are bound to lose any conceivable battle with the state.

The dilemma is easier to resolve - at least in theory - than it may seem. If we consider the Trinidadian situation again, the crucial factor in the cultural predicament of Indo-Trinidadians clearly consists in the official definition of nationalism. If Trinidadian nationalism is to be defined as coterminous with black culture, then Indians have to choose between evils, as it were; either they assimilate and become "Creoles", or they retain their Indianness at the risk of being ostracized and disqualified. If the definition of Trinidadianness on the contrary is extended to include Hindus, and if India is officially recognized as an ancestral Trinidadian land, then it may be possible to be Indian and Trinidadian without more ado. Similarly, multi-cultural nations such as Australia, the United States and South Africa could conceivably extend the idiom of nationalism to include non-white people, creating compromises and tolerating differences in a "Mauritian" fashion.

The more fundamental problem is, however, not yet resolved. For nationalism, intimately linked with the state and large-scale organization, entails specific principles of social organization not necessarily compatible with those of ethnic minorities. The success of Mauritian nationalism seems to depend on the containment of such differences to contexts where ethnic segmentary oppositions do not interfere with the principles of the state. Cultural minorities, apparently, are thus forced to adapt to some of the demands of the modern state in order to be able to articulate their interests. This will to a greater or lesser extent entail cultural change. If they refuse, they run the risk of witnessing the purchase of their ancestral land for a handful of coloured glass beads. For the key variable in the understanding of relationships between nations and ethnics is power. The power invariably lies with the state, which officially represents the nation, which possesses the monopoly of legitimate violence, which contains the culturally hegemonic group, - and which thereby defines the terms of negotiation and the form of discourse. Powerless groups must therefore learn to master the language of the powerful, and in this process they may have to alter their cultural identity substantially. This applies equally to aboriginal populations and to urban minorities, although emphases may differ; for one thing, urban minorities, unlike many indigenes, usually engage in wagework and in this conform to a central requirement of nationalist ideology.

3. The justification of nationalism: Symbols, power, integrating practices

In order to function successfully, nationalism must legitimize the power of the state, and it must simultaneously make the lives of citizens seem inherently meaningful. The partial failure of Norwegian nationalism to make sense to the Saami in this dual fashion has led to negotiations, where the Norwegian nation-state nevertheless sets the terms by ignoring and tacitly disapproving of Saami identity and selfhood. Indeed, in all the examples mentioned except that of Mauritius, which is in this respect considered a success, conflicts between nation-states and ethnics can be understood along this dimension. If the state fails to persuade its citizens that it represents the realization of (some of) their dreams and aspirations, then its power may appear illegitimate. The result may be revolt, and in such cases the state may well resort to violence. This is well known from many countries, past and present. My point has been that there are also powerful non-violent means available for the nation-state to secure its monopoly of political power, even if nationalist ideology fails. The most important is the state's exclusive right to define the terms of discourse, including its right to collect taxes. In well integrated states, these terms of discourse take on the form of doxa (Bourdieu, 1980); that is, they are perceived as unquestionable. In states including groups which are not integrated in the state through a shared education, participation in the same economic system, etc., this form of statal power is perceived as a form of coercion; as enforced "acculturation", as it were.

The ideological power of nationalism is often (but not always) expressed in the official identification of enemies, and as has been noted many times by analysts, warfare can serve as a nationally integrating force. Any segmentary opposition (or other forms of conflict) within the polity may be postponed and "forgotten" when an external enemy encourages the realization of the highest, unambiguously binary level of the system of oppositions. The Falklands/ Malvinas war between Britain and Argentina (1982) is a recent example of this familiar mechanism, at least if seen from a British point of view. Similarly, the identification and prosecution of internal enemies has been a familiar technique of integration for centuries. Contemporary witchhunts include the Kenyan police-state's "internal war" against the partly mythical opposition group Mwakenya and, emerging from popular (not state-monitored) nationalism, French nationalists' designation of North Africans as the main threat to Frenchness. In order to understand the persuasive power of nationalism on the one hand, and its oppressive aspects on the other, it must be conceded that nationalism is, ultimately, a particularist form of ideology: it defines cultural and social boundaries on behalf of a community, and it excludes those who do not fit in. I have argued that these boundaries are flexible, but have also indicated that they are not indefinitely so. Notably, nationalism - as the ideology which holds that the boundaries of the state should be coterminous with the boundaries of the cultural community12 - requires cultural uniformity in certain respects. Nationalism represents a simple binary opposition (between citizens and non-citizens), whereas other ideologies differentiate between people in segmentary terms.

The state, which by the late 20th century necessarily represents a successful nationalism (i.e., it is a nation-state), possesses a monopoly of violence and has exclusive rights to extract tribute in the form of taxes. It is therefore in the immediate interest of a successful nationalism to promote cultural homogeneity as regards law and order, and economic activity. Conflicts between pastoralists and the new states in Africa typically exemplify this problem. Pastoralists do not acknowledge the laws pertaining to private property (nor, for that matter, national borders), and since their economy is not chiefly a monetary one, they do not contribute financially to the state. Therefore, they are by definition anti-nationalists insofar as they reside within the state which, as ideology has it, should be coterminous with the cultural community. In a very fundamental sense, then, every human being in the late 20th century is encouraged - or forced - to take on an identity as a citizen. As indicated, those who do not tend to lose. The battle between nationalist and ethnic ideologies is most frequently won by the dominant nationalism, which is already represented in the state. However, as I have suggested, there are possible compromises between the ideology of the nation-state and ethnicity - even if the inherently aggressive assimilating drive of state nationalism is acknowledged. Let us therefore consider some conditions for the resolution - or avoidance - of conflicts between state nationalism and ethnicity.

4. Conditions for multi-cultural peace

Two main types of conflict involve nationalist ideology. Many conflicts arise between states or potential states. Every international conflict involving states - as well as civil wars such as the one in Sri Lanka, where one party fights for political secession - are varieties of this kind of conflict. The ideologies activated are all explicitly nationalist in character.

This discussion has focussed on the second type of conflict. This kind of conflict unfolds within a state where neither party favours political secession. Such conflicts can involve the state and one or several ethnics; ideologically, they are ambiguous as several of the combatants may claim to represent universalism and nationalism on behalf of all of the groups involved in the conflict, notwithstanding that some other group may (or may not) form the majority and/or be in charge of the state administration. This category of conflicts is the most complex, empirically and ideologically.

By way of conclusion, we can now indicate some necessary (although not sufficient) conditions for the resolution of types of conflicts involving categories of people where their stressing mutual cultural differences forms an important part of the ideological justification of the conflict, and where the boundaries of the state are not challenged. In other words, this is an attempt to delineate conditions for peaceful cultural plurality within a modern state.

4.1. Necessary conditions for peaceful multiculturalism
· Equal access to the educational system, the labour market and/or other shared facilities should be deemed as desirable. This must also entail the right to be different, the right not to participate in national society in certain respects, the right to enact systems of segmentary oppositions not sanctioned by the state. The judiciary system will normally limit the extent of the articulation of such differences. Laws are changeable.
· National identity should be available to all citizens regardless of their cultural differences.
· State policies pertaining to multiculturalism should take account of possible culturally contingent differences in their definitions of situations.
· By implication, the state cannot be identified with a set of symbols exclusively representing one or a few component populations.
· Political power should be decentralized, and different principles for local political organization should be accepted.

Differences between nation-states as regards modes of integration, political systems and economic circumstances are enormous. Since I have throughout this article treated the nation-state as an analytical concept, I am now compelled to mention some of the relevant differences between actual, historically situated nation-states.

First, the differences in degree of incorporation within the state are crucial. For instance, many African and Melanesian societies are hardly at all integrated on a national level; their members hardly participate in national society. The problems discussed in this article do not apply to them yet (although they are faced with different problems).

Secondly, the degree of cultural uniformity within nation-states varies. Even in Mauritius, where the absence of cultural uniformity seems to have been turned into a blessing for nationalism, cultural homogeneity is very high in important respects; there is consensus as regards the political system, there is uniform participation in the educational system as well as the capitalist economy. Conflicts between state and ethnic are more difficult to resolve when representatives of the ethnic demand participation on their own terms, which need not be those of the nation-state.

Thirdly, it is empirically significant whether a particular nation-state and its accompanying ideology has emerged out of feudalism or out of colonialism (or both at once, as seems to be the case with some of the post-1989 East European nationalisms). The former societies tend to be better integrated, socially and culturally, than the latter.

Fourthly, specific political traditions or histories influence the nature of inter-ethnic relations. The history of slavery contributes to shaping the contemporary relationship between blacks and the US nation-state and seems to prevent constructive dialogue. On the other hand, the moderate success of independent Zimbabwe as regards ethnic relations shows that there is nothing inevitable in this kind of historical process.

Fifthly, and perhaps most fundamentally, the actual division of political and economic power (and thereby, the division of discursive power) constitutes, in an important sense, the social structure of a society. In a word, groups which are oppressed, poor and stigmatized have little opportunity to articulate their claims convincingly. The remarkable success of North American Jews in retaining their ethnic identity, governing their own destiny and yet being recognized as good Americans, a striking success compared with other immigrant groups, has been possible only because their economic power has been considerable.

In sum, if violence or other serious conflicts between nation-state and ethnicity are to be avoided, then the state must reduce its demands as regards the degree of cultural integration of its citizens. Since it is virtually second nature of modern, bureaucratic states (unlike earlier, pre-nationalist states) to promote cultural integration at any cost, this is extremely difficult to achieve. It remains an indubitable fact, nevertheless, that the responsibility lies largely with the state so long as it insists on retaining its monopolies of political power and the use of legitimate violence.

NOTES

Acknowledgements. Several persons have read and commented on an earlier version of the article. Hülya Demirdirek provided stimulating criticism and comments on both substantial and theoretical issues. Iver B. Neumann contributed many valuable insights and suggestions from the field of International Relations. Georg Henriksen and Jo Helle-Valle raised important issues concerning the treatment of nationalism and ethnicity. Nils-Petter Gleditsch suggested several improvements of both form and content. Last but not least; thanks to Harald Eidheim's suggestions, the exposition has been considerably clarified - particularly in the section dealing with the Saami. The ongoing research project on ideologies in Trinidad and Mauritius has been funded by a grant and a subsequent fellowship from the Norwegian Research Council for the Sciences and the Humanities (NAVF).

1Relationships of coercion and integration between and within states are not, of course, necessarily constituted on the principles of the sovereignty of the state. When, in 1968, the USSR invaded Czechoslovakia and when, a decade later, the Red Army invaded Afghanistan, the limits of the relevant polity were drawn outside of national boundaries. Conversely, to the extent that the USSR failed to use violence to suppress autonomists in the Baltic republics in 1989-90, the relevant limits of the polity were drawn inside the state. In neither case was the state unambiguously perceived as the relevant political unit.

2A good, topically relevant demonstration of this dual character of ideology, is Kapferer's (1988) analysis of the nationalisms of Sri Lanka and Australia.

3This does not mean that ethnicity can be reduced meaningfully to politics. I have argued earlier (Eriksen, 1988) that ethnic identity and ethnic organization are both irreducible aspects of the phenomenon.

4Whatever their "objective content", cultural differences are important as long as they make a difference to the people involved. In a given situation, the communicated cultural differences between say, Kikuyu and Kamba in Kenya (who are linguistically close) may be more important than those obtaining between Kikuyu and Luo (who are linguistically distant).

5The Zimbabwean example brings out some of the ambiguities of the matter: To the whites, who lost the civil war, Zimbabwean nationalism presented itself as a relevant option to be endorsed or rejected. Many failed to make up their mind unambiguously, and tend to oscillate situationally between Zimbabwean nationalist and Rhodesian supremacist ideologies.

6My use of ethnic as noun is inspired by the French word ethnie, which is semantically wider than the term "ethnic group", which connotes tight group integration.

7Trinidad & Tobago became intependent from Britain in 1962, Mauritius in 1968. Both are members of the Commonwealth. Research in Mauritius and Trinidad was carried out, respectively, in 1986 and 1989.

8This is dealt with more fully in Eriksen, 1988, pp. 166-213. The issue of language is discussed in Eriksen,1990.

9See Ardener (1989, pp. 129-30) on dominant and "muted" groups with particular reference to gender.

10The attempted coup d'etat in Trinidad in July, 1990, was not ethnically motivated. Although known as "Black Muslims", apparently an ethnic label, the rebels were a tiny group of politically frustrated radicals with little initial popular support. It is possible that the looting and burning taking place in Port of Spain for a few days during the drama did have an ethnic aspect in the targeting of wealthy Syrians, but this was no marked feature of the riots.

11The South African situation further exemplifies the connection between industrialism and nationalism: Business interests in South Africa favour universal nationalism because it will integrate a larger number of people into the economic system, while other whites continue to support the non-nationalist apartheid system.

12The related, but different ideology of federalism, is not considered here. It may provide solutions to some of the problems discussed. This also applies to the "consociational" state model discussed and advocated by Lijphart (1977). Forms of conflict not considered here are those emerging from "irridenta nationalism" (the most famous example of which is probably that of Alto Adige/Süd-Tirol in North-Eastern Italy), and forms of national integration not considered include diaspora nationalisms, where the nation is not strictly localized to a territory.

REFERENCES

Anderson, Benedict,1983. Imagined Communities. London: Verso
Ardener, Edwin,1989. "The 'Problem' Revisited", in Ardener, The Voice of Prophecy and other essays . Oxford: Basil Blackwell
Bourdieu, Pierre,1980. Le sens pratique. Paris: Minuit
Eidheim, Harald,1971. Aspects of the Lappish Minority Situation. Oslo: Universitetsforlaget
Eidheim, Harald,1985. "Indigenous Peoples and the State: The Saami Case in Norway", pp. 155-172 in Jens Brøsted et al. (ed.), Native Power. Oslo: Universitetsforlaget
Eriksen, Thomas Hylland,1988. Communicating Cultural Difference and Identity. Ethnicity and Nationalism in Mauritius. Oslo: Occasional Papers in Social Anthropology
Eriksen, Thomas Hylland,1990. "Linguistic Diversity and the Quest for National Identity: The Case of Mauritius". Ethnic and Racial Studies, vol. 13, no. 1.
Evans-Pritchard, E.E.,1940. The Nuer. Oxford: Clarendon Press
Forsythe, Diana,1989. "German Identity and the Problem of History", pp. 137-156 in Elizabeth Tonkin et al., ed., History and Ethnicity, ASA monograph no. 27. London: Routledge
Gellner, Ernest,1983. Nations and Nationalism. Oxford: Basil Blackwell
Giddens, Anthony ,1985. The Nation-State and Violence. Cambridge: Polity
Giddens, Anthony,1987. "Nation-states and violence", pp. 166-182 in Giddens, Social Theory and Modern Sociology, pp. 166-182. Cambridge: Polity
Goody, Jack,1986.The Logic of Writing and the Organization of Society. Cambridge: Cambridge University Press
Helle-Valle, Jo,1989. Banyaruanda in Uganda. Unpublished mag. art. thesis, Department of Social Anthropology, University of Oslo
Hobsbawm, Eric & Terence Ranger, eds., 1983. The Invention of Tradition. Cambridge: Cambridge University Press
Horowitz, Donald L., 1985. Ethnic Groups in Conflict. Berkeley: University of California Press
Kapferer, Bruce,1988. Legends of People; Myths of State. Washington: Smithsonian Institution Press
Larsen, Tord, 1987. "Action, Morality and Cultural Translation". Journal of Anthropological Research, vol. 43, no. 1, pp. 1-24.
Lijphart, Arend,1977. Democracy in Plural Societies. A Comparative Exploration. New Haven: Yale University Press
Neumann, Iver B. , 1990, forthcoming. "Poland as a Regional Great Power". In Iver B. Neumann, ed., Regional Great Powers in International Relations. Oslo: Norwegian University Press
Smith, Anthony D., 1981. The Ethnic Revival. Cambridge: Cambridge University Press
Smith, Anthony D., 1983. State and Nation in the Third World. Brighton: Wheatsheaf Books
Wallensteen, Peter et al., 1986. World Armaments and Disarmaments. SIPRI Yearbook 1986. Oxford: Oxford University Press.
Østerud, Øyvind, 1984. Nasjonenes selvbestemmelsesrett. Søkelys på en politisk doktrine. Oslo: Universitetsforlaget

©Thomas Hylland Eriksen 1991

Nexus