Ethnic identity, national identity and intergroup conflict:

The significance of personal experiences

Thomas Hylland Eriksen

In Ashmore, Jussim, Wilder (eds.): Social identity, intergroup conflict, and conflict reduction, pp. 42-70. Oxford: Oxford University Press'. 2001

There exists by now an enormous literature on contemporary ethnicity and nationalism in its various forms, whether as North American multiculturalism or indigenous rights movements, as post-Soviet ethnonationalism in Central and Eastern Europe, as urban minority dilemmas or Islamic revivalism in Western Europe, as indigenista movements in Latin America, or as processes of political fission and fusion in contemporary Africa. The analytical focus of this chapter will be on the concept of identity. The naïve question to be asked at the outset it: What is it about identity politics that makes it such a formidable force in the contemporary world? Some important issues will have to be omitted, notably the relationship between identity politics, globalisation and reflexive modernity (e.g. Bauman 1993, Friedman 1994, Giddens 1994). Instead, what is offered amounts to an anthropological perspective on the relationship between personal identity and political identity, using conflicts based on identity politics as empirical examples.

This chapter sets out to do three things: First, a brief overview of the standard social anthropological perspective on the politics of identity is provided. Identity politics should be taken to mean political ideology, organisation and action which openly represents the interests of designated groups based on "essential" characteristics such as ethnic origin or religion, and whose legitimacy lies in the support of important segments of such groups. Membership in such groups is generally ascribed, unlike membership in other political groups (socialists, liberals, trade unions etc.). Then, by way of examples from India, Fiji and Yugoslavia, parallels and differences between some such conflicts are highlighted. The examples are chosen mainly for their differences: What they have in common is, apart from their temporal location to the 1990s, apparently only that they are based on ideologies of culture and identity. The Yugoslav conflict has been extremely brutal and tragic; the Fijian one has only involved a few casualties, but has led to important constitutional changes; while the Indian conflict, although occasionally bursting into violence, is largely contained within the framework of institutional Indian politics. If it can persuasively be argued that these conflicts have important features in common, it is likely that those features will also be present in other settings where identity politics has a major impact. The third and final part draws on classic political anthropology in order to suggest some general features of the relationship between personal identity and politics.

The Anthropology of Identity

The last decades of the 20th century saw a dramatic re-conceptualisation of core concepts, including culture and society, within the social sciences. Until the 1960s, the close overlap between culture and ethnicity or even culture and nationhood was generally taken for granted in the scholarly community. During the past thirty years, however, hardly a single serious contribution to the field has failed to point out that there is no one-to-one relationship between culture and ethnicity (the seminal text here is Barth 1969); that cultural differences cut across ethnic boundaries, and that ethnic identity is based on socially sanctioned notions of cultural differences, not "real" ones. While ethnic identity should be taken to refer to a notion of shared ancestry (a kind of fictive kinship), culture refers to shared representations, norms and practices. One can have deep ethnic differences without correspondingly important cultural differences (as in the Bosnian example below); and one can have cultural variation without ethnic boundaries (as, for example, between the English middle class and the English working class).

Several recent debates in anthropology and neighbouring disciplines pull in the same direction: away from notions of integrated societies or cultures towards a vision of a more fragmented, paradoxical and ambiguous world. The currently bustling academic industry around the notion of globalisation (see Featherstone 1990 for an early, influential contribution) represents an empirically oriented take on these issues, focusing on the largely technology-driven processes that contribute to increasing contact across boundaries and diminished importance of space. This focus on unbounded processes rather than isolated communities has contributed to a reconceptualisation of the social which is radically opposed to that of classic Durkheimian sociology and anthropology; where flux, movement and change become the rule and not the exception in social life (Strathern 1991, Hannerz 1992, Lash and Urry 1994).

Particularly in North America, the classic concept of culture has been used for what it is worth in domestic identity politics, leading in some cases to controversial policies of multiculturalism, where individuals have been endowed with special rights in accordance with their ethnic origins. Critics might point out that multiculturalism in some of its versions resembles apartheid; and also, that by positing a simple one-to-one relationship between ethnic origin and culture, it not only encouraged a "disuniting of America" (Schlesinger 1992), but also contributed to reifying misleading notions of culture seen as the commonalities of a bounded set of individuals, like so many nationalisms writ small.

A further disruptive tendency has been the so-called postcolonial movement in literary studies, spilling into anthropology and other disciplines, which has raised the question of who has the right to identify whom; a standard text in this field of discourse is Edward Said’s Orientalism (Said 1978), although Frantz Fanon developed similar insights two decades earlier. Said and others argued, briefly, that ethnocentrism was deeply embedded in Western scholarship dealing with non-Western peoples. Postcolonial critics also tend to call attention to the multiplicity of voices (an academic cliché by the late 1990s) present in any society and the general unwillingness of academic researchers to give all of them the attention they deserve.

Two related debates defined the field for many years. First, there was the controversy over primordialism and instrumentalism. Was ethnic identity "primordial", that is profoundly rooted in – and generative of – collective experiences; or did it arise as an ad hoc supplement to political strategies? An early, powerful defense of the instrumentalist view was represented in Abner Cohen’s work on urban ethnicity in Africa (1969, 1974), showing the conscious manipulation of kinship and cultural symbols by political entrepreneurs seeking political gain. This perspective is still used with considerable success in studies of identity politics. Who, then, were (or are) the primordialists? Clifford Geertz is often associated with this view, arguing as he does along hermeneutic lines that cultural systems are more or less self-sustaining and were thus not subject to the wilful manipulation of individuals (Geertz 1973), a perspective he retained when writing about nationhood in the Third World (Geertz 1967). Typically, however, ethnicity studies were – and are – instrumentalist in their basic orientation (Rex 1997).

The second debate, usually framed as the opposition between constructivism and essentialism, concerns the question whether ethnic or national communities are created more or less consciously, or whether they grow organically, as it were, out of pre-existing cultural communities. In nationalism studies, the most highly profiled antagonists regarding this issue have been the late Ernest Gellner (1983, 1997) and Anthony D. Smith. Smith (1986, 1991) has developed an intermediate position in arguing the importance of pre-existing ethnies for the development of nationalism while acknowledging its essential modernity. Gellner, on the other hand, champions the view that nations are entirely modern creations – the progeny of industrialism and the state – which more or less fraudulently invent their past to gain a semblance of antiquity and deep roots; his final statement on the issue is reproduced in the posthumously published Nationalism (Gellner 1997) under the heading "Do nations have navels?", a pun on the Biblical enigma relating to Adam’s navel. Regarding definitions, there are also important differences between theorists. While Gellner holds nations to be ethnic groups who either control a state or who have leaders who wish to do so, Anderson (1983) sees no necessary link between the abstract "imagined community" of the nation and particular ethnic groups; indeed, several of the main examples in his famous book on "imagined communities", including The Philipphines and Indonesia, are multiethnic countries. Yet others have distinguished between ethnic nations and "civic" ones (Smith 1991). There is nonetheless general agreement that nations are by definition linked with states, whether they are based on a common ethnic identity or not.

In anthropology at least, the recent shift towards the study of identities rather than cultures has entailed an intense focus on conscious agency and reflexivity; and for many anthropologists, essentialism and primordialism appear as dated as pre-Darwinian biology. In addition, there seems to be good political sense in discarding the old, static view of culture, which is being used for many political purposes which are difficult to endorse by academics committed to democratic values, ranging from the Balkan war to discrimination against ethnic minorities in Western Europe. Further, this is an age when the informants talk back. It could, perhaps, be said that a main purpose of an earlier anthropology consisted in identifying other cultures. Representatives of these so-called other cultures are now perfectly able to identify themselves, which leaves the scholars either out of a job or with a new mission, that is to identify their identifications; in other words, to study reflexive identity politics.

Not so many years ago, anthropology was still a discipline fueled by a programmatic love of cultural variation for its own sake, and anthropologists involved in advocacy tended to defend indigenous peoples’ or other minorities’ traditional way of life against the onslaught of modernity. A main tendency in recent years has, on the contrary, consisted in deconstructing instrumentalist uses of notions of authenticity and tradition, and showing not only that the internal variation within a group is much greater than one would expect, but also that traditionalist ideologies are, paradoxically, direct results of modernisation (e.g. Roosens 1989). This theoretical shift is a very significant one. It offers a method for investigating strategic action, the politics of symbols, and contemporary processes of identity politics extremely well within a uniform comparative framework.

In other words: Cracks in the edifice of mainstream social and cultural anthropology, some of them directly inspired by events well beyond the confines of academia, have led to a widespread reconceptualisation of society and culture. Reification and essentialism have become central terms of denounciation; multiple voices, situational identification and cultural flows are some of the keywords delineating the current intellectual agenda. It has become difficult if not impossible to talk of, say, Nuer culture, Hopi culture, Dutch culture and so on, since such terms immediately invite critical questions of whose Nuer culture, Hopi culture and so on, intimating that there is an infinite number of versions of each culture, none of which is more "true" than the others (Holland 1997). Ethnicity and nationalism, then, become the political reifications or constructions of a particular authorised version of a culture; freezing that which naturally flows, erecting artificial boundaries where they did not exist before; trimming and shaping the past to fit present needs, inventing traditions where no organic traditions exist, or are adequate, to ensure a sense of continuity with the past.

A new kind of political responsibility has entered academia in acute ways during the last decades. Academic or semi-academic statements about nations, ethnic groups or cultures may now immediately be picked up, or assimilated more or less subconsciously, by ideologists and politicians wishing to build their reputation on national chauvinism, ethnic antagonism, enemy images and so on. The liberal academic establishment thus wags a warning finger at those who dare to talk of culture as the cause of conflicts, shaking their heads sadly over those lost souls who have not yet heeded the words of leading theorists such as Barth (1969) and Gellner (1983), criticising those who do not realise that culture is chimerical and fleeting, and that reified culture is a dangerous tool. It is, thus, not only intellectually correct, but also politically correct to reject all forms of essentialism.

The current scholarly orthodoxy on ethnicity and the politics of identity can be summed up as follows:

• Although ethnicity is widely believed to express cultural differences, there is a variable and complex relationship between ethnicity and culture; and there is certainly no one-to-one relationship between ethnic differences and cultural ones.

• Ethnicity is a property of a relationship between two or several groups, not a property of a group; it exists between and not within groups.

• Ethnicity is the enduring and systematic communication of cultural differences between groups considering themselves to be distinct. It appears whenever cultural differences are made relevant in social interaction, and it should thus be studied at the level of social life, not at the level of symbolic culture.

Ethnicity is thus relational, and also situational: The ethnic character of a social encounter is contingent on the situation. It is not, in other words, inherent.

This instrumentalist framing of ethnicity, which may appear simply as a set of methodological guidelines, has firm, although usually untheorised, philosophical foundations and is, as I have tried to show elsewhere (Eriksen 1998b), deeply embedded in empiricist thought. I shall argue that this approach, notwithstanding its strengths, is limited in overemphasising choice and strategy (instrumental aspects) when analysing identity politics. As a result, the self is taken for granted (A.P. Cohen 1994), and it is therefore not shown how it can be possible to mobilise particular aspects of personal identity for antagonistic identity politics. Yet, anthropology as a discipline is in a privileged situation to study the dynamics of identity politics, precisely because of its focus on the ongoing flow of social interaction.

Although an enormous amount of anthropological research has been carried out on ethnicity and nationalism since around 1970, surprisingly few studies have dealt with violent conflicts and conflict resolution (see, however, Tambiah 1994, Turton 1997). The dominant approaches to ethnicity have been instrumentalist (with a focus on politics) or constructivist (with a focus on ideology), and research questions have concentrated on the establishment and reproduction of ethnically incorporated groups, not on the circumstances under which ethnicity may become politically less important. While my examples (below) and the ensuing discussion will indicate the fruitfulness of these approaches, it is also necessary to point out the need for a phenomenological understanding of social identity, which sees it as emerging from experiences, not as a mere construct of ideology. In this, I follow scholars such as Cohen (1994) and Jenkins (1996), who have called for an anthropology of identity which does not concentrate exclusively on its political and ideological aspects, but also strives to understand the self.

As noted by Holland (1997), anthropologists are generally associated with a culturalist view of the self, arguing its cultural specificity as against psychologists, who have been more prone to a universalist view – the self as something proper not to particular cultures but to humanity, with universal characteristics lurking below a thin veneer of culture. For the purposes of the present argument, it is not necessary to take a stance on this controversy, partly since it can be presupposed that modernity creates a particular kind of selves with important shared characteristics everywhere (Giddens 1991), but also because this chapter restricts its scope to the relationship between personal experiences, ideology and political mobilisation. As the examples will hopefully show, the similarities are more striking than the differences here.

Culture and the breakup of Yugoslavia

No other recent ethnic conflict has been more intensely studied, discussed and moralised over than the breakup of Yugoslavia in 1991. This was followed by three major wars and a number of smaller skirmishes, and the situation in many parts of ex-Yugoslavia remains unstable and tense. In Europe, the outbreak of war in Yugoslavia has been interpreted by hundreds of commentators. In the press, it was occasionally argued, along sociobiological or Hobbesian lines, that humans are driven by aggressive instincts which emerge when the social fabric falls apart, in this case organised as factions ultimately based on kinship. Another view, popular in the European nationalist right, implied that ethnic conflict was inevitable when different groups are forcibly integrated into one state. Most scholars have, on the contrary, tended to focus on the cultural logic of feuding in Balkan society, the deep economic crisis underlying the conflict, or the rise of Serbian supremacy during the 1980, looking for contradictions within Yugoslav society rather than into human nature for an explanation of the conflict. A widespread view nevertheless sees the cultural differences between the constituent groups as a basic cause of the conflict (cf. the influential analysis by Ignatieff, 1994, or Huntington’s controversial model, 1996).

Yugoslavia was a state that came into being twice: after the First World War and after the Second World War. The first Yugoslav republic (1918—41) was for all practical purposes the Serb monarchy writ large, and was riddled with continuous internal tension. Croats and Slovenes reluctantly supported King Alexander’s regime, seeing it as a possible defense against Italian, Austrian and Hungarian aggressors. It was a precarious state, periodically dictatorial, which had been on the verge of collapse several times before it fell apart as Germany invaded the country in 1941. While Serbs generally resisted the Germans, Croats collaborated and saw the German intervention as an opportunity to create their own state.

Before the formation of the second Yugoslav republic in 1945, both of the largest constituent groups suffered large-scale massacres – Serbs at the hands of Croatian fascists (Ustasa) in 1941, while large numbers of Croats were killed by Serbian communists (Partisans) in 1945. The new Yugoslav state (1945-91) was a non-aligned socialist federation led by a pro-Serbian Croat, Josip Broz Tito, until his death in 1980. Ethnicity was officially declared a non-issue in socialist Yugoslavia. This official blindness stems from the Marxist view that class is a more objective and more authentic vessel of social identity than ethnicity or nationality, which was officially seen as an expression of false consciousness. This does not, however, imply that Yugoslav policies were particularly repressive regarding expressions of cultural distinctiveness symbolising ethnic identity – on the contrary, in this area Yugoslavia was more liberal than many Western European countries whose leadership feared separatism and social fission. Ethnic identity was seen as politically irrelevant, and partly for this reason, the use of various languages and the practice of different religions were tolerated in civil society. It is true that the merging of Serbian and Croatian into one language, Serbo-Croatian, in the 1950s signalled an attempt at building a unitary Yugoslav identity, but the two languages were so closely related that few, except Croat intellectuals reacting against the relegation of specific Croat variants as "dialect", seem to have taken offence (Schöpflin 1993). Albanian remained an official language in Kosovo, as did Slovene in Slovenia.

There were nearly twice as many Serbs as Croats in Yugoslavia, and twice as many Croats as Slovenes or Muslims. Regarding the territorial dimension, Slovenes were – and are – largely confined to the nearly mono-ethnic republic of Slovenia. Both Croatia and Serbia had large minorities of Serbs and Croats, respectively, as well as Gypsies, while Serbia also included nearly a million Albanians (in Kosovo) as well as smaller numbers of Hungarians and others. In Macedonia, most of the population were (and are) Slav-speaking Macedonians, while the Montenegrins of Montenegro are culturally close to the Serbs. Bosnia-Herzegovina was the most throroughly mixed republic, with roughly equal numbers of Serbs, Croats and Bosnian Muslims often living in mixed areas. As is well known, the conflicts of the 1990s have modified this picture somewhat, creating large monoethnic territories in formerly mixed areas.

Although the ruling Communist party seems to have believed that a common Yugoslav identity would eventually supersede the national identities based on ethnic membership, ethnic identity remained strong in most parts of the country throughout the postwar era. There were nevertheless important exceptions, particularly in cities such as Belgrade, Sarajevo and Zagreb, where many people increasingly identified themselves primarily as Yugoslavs, and where mixed marriages were common.

Ethnic identities did, in other words, not disappear during the existence of Yugoslavia. In some urban areas they were arguably weakened, but it could be — and has been — argued that the non-ethnic character of Yugoslav politics actually led to its strengthening as a vehicle for the political opposition and made it possible for Serbs to gain control over the armed forces and state bureaucracy: since political ethnicity officially did not exist (only cultural ethnicity did), there were no institutionalised ways of preventing one group from dominating the public sector.

The wars in ex-Yugoslavia have bequeathed to the world the neologism ethnic cleansing. It is nevertheless easy to show that the conflicts involving Serbs, Croats, Bosnian Muslims, Slovenes and Albanians were never conflicts over the right to assert one’s ethnic or cultural identity, but were based on competing claims to rights such as employment, welfare and political influence. What needs to be explained is the fact that the conflicts over these resources were framed in ethnic terms rather than being seen as, say, regional, class-based, or even ideological.

The relevant questions are therefore: What is the stuff of ethnic identity in ex-Yugoslavia; in which ways do the groups differ from one another, and why did group allegiances turn out to be so strong? Bosnia-Herzegovina may be considered as an example. There are three large ethnic groups inhabiting Bosnia; Serbs, Croats and Muslims. The main difference between the groups is religion; Serbs are Orthodox and Croats are Catholics. (Ironically, religious fervour was not particularly widespread in pre-war Bosnia.) They all have common origins: Slav immigration into Illyria from the north took place between c. AD 400 and 700, and the cultural differences between Croats and Serbs are perhaps comparable to those between Norwegians and Swedes. The "objective" difference between Bosnian Christians and Bosnian Muslims, further, has been compared to the difference between English Protestants and English Catholics (Cornell & Hartmann 1998). Unlike the impression sometimes given in Serbian and Croatian propaganda, Bosnian Muslims are not the descendants of alien invaders, but of locally residing converts. Although each group has its numerical stronghold, many Bosnian regions and villages were mixed before the war. This implies, among other things, that they went to school together, worked together and took part in various leisure activities together. A Serbian villager in Bosnia had more in common, culturally speaking, with a Muslim co-villager than with a Serb from Belgrade. This would hold true of both dialect spoken and way of life in general. However, since religion turned out to be the central marker of collective identity in the Bosnian conflict, the effective boundary was drawn not between villagers and city-dwellers, but between religious categories.

The boundaries between the groups may seem arbitrary. However, the large, "national" groups are clearly embedded in smaller, local networks based on kinship and informal interaction, as well as being culturally founded in religious schisms, collective myths or memories of treason and resistance under Ottoman rule, massacres, deception and humiliations. Although it is tempting to argue that any so-called cultural trait can be exploited in the formation of national or ethnic groups, it is obvious that not just anything will do. Nothing comes out of nothing, and strong collective identities – such as the ones revealed during the war in Bosnia – are always embedded in personal experiences. In one of the most detailed accounts of ethnicity at the village level in pre-war Bosnia, Bringa (1996) shows that although cultural differences between the groups were perhaps negligible, and although relations between Serbs, Muslims and Croats might be cordial at the local level, there were nevertheless important social practices of affiliation that created boundaries between them – not in the cosmopolitan Sarajevo middle class embraced by Western commentators, perhaps, but elsewhere. Intermarriage was rare, the close informal networks of friends tended to be monoethnic, and the discrete groups maintained different, sometimes conflicting myths of origin. The intimate sphere, in other words, seems to have been largely monoethnic and by this token, Bosnia was a plural society in the classic sense (Furnivall 1948); the public arenas were shared, but the private ones were discrete.

One may choose not to speak of such features of social reality and everyday life in terms of "culture", but they are no more "invented" than any other social fact. People do not choose their relatives, they cannot choose to do away with their childhood and everything they learnt at a tender age. These are aspects of identity which are not chosen, which are incorporated and implicit. People relate to them as reflexive agents, but they do so within limitations that are not chosen. Such limitations form the objective foundations of social identification. When analysts such as Cornell and Hartmann (1998) argue that ethnic identities before the war were weak, sometimes socially irrelevant, and in many cases ambiguous (many had parents from different groups), they refer to particular segments of society – a main example is the Yugoslav basketball team – whose members were active participants in pan-Yugoslav contexts.

The maintenance of ethnic boundaries in socialisation and the private sphere reveals a main cause of the failure of Yugoslav social engineering in doing away with ethnic identification. It does not explain the outbreak of war in the early 1990s, but it indicates why the groups that emerged were so strong, and why they were based on ethnicity (seen as fictive kinship) instead of, say, class or region. Their foundation must be sought not in the biology of kinship, as some might want to argue, but in the phenomenology of social experience, the raw material of personal identity. This argument will be elaborated further after an examination of two very different examples of intergroup conflict in polyethnic societies.

The Fijian coup-d’etat

The Pacific island-state of Fiji, located at the ethnographic crossroads between Melanesia and Polynesia, is perhaps less heterogenous than Yugoslavia, but it is scarcely less ethnically divided. Its population of about 800,000 is largely composed of two ethnic categories: Fijians and Indians. The Fijians are indigenous, largely Christian, speak a Polynesian language and make up slightly less than half the population. The Indians are uprooted "overseas Indians" whose ancestors were brought to Fiji during colonialism under the British indentureship system described, probably a trifle too grimly, as "a new system of slavery" by Tinker (1974). They are overwhelmingly Hindus (with a Muslim minority), speak a locally modified dialect of Hindi, and were slightly more numerous than the indigenous Fijians until the political changes in the late 1980s leading to mass migration of Indo-Fijians. The small minorities of Europeans and Chinese are politically insignificant, but the economically powerful Europeans, representing the former colonial régime, have in no small measure shaped the Fijian public sphere, notably through establishing English as the national lingua franca.

The relatively brief history of democratic Fijian politics up to 1987 has been described with the metaphors of balance and power-sharing (Premdas 1993, Kelly 1998). While Indians in practice wielded disproportionate economic power, it was tacitly agreed that Fijians should be politically paramount. However, there were indications that the "equilibrium" model was under severe stress in the early- to mid-1980s, and in the 1987 elections, a coalition supported by most Indians and only a few Fijians won. Although the new Prime Minister was a Fijian, many saw his government as a vehicle for Indian communal politics. In May and September 1987, the military seized power through two successive coups-d’etat, explicitly doing so to protect "native" Fijian interests. In the period following the coups, thousands of Indians emigrated.

The tension between Fijians and Indians had been evident throughout Fijian history. Unlike the situation in Yugoslavia and especially Bosnia, nobody would question the view that there are deep cultural differences between Indians and Fijians. Their languages, cultural traditions, religions and gender relations differ markedly — indeed, one of their recent ethnographers has poetically described striking, and culturally potent, differences in body language between Fijians and Indians (Williksen-Bakker 1991). There seems to be little informal interaction between the groups, most rural areas are dominated by one or the other, and intermarriage has always been nearly non-existent. The effective separation of Fijians from Indians has always been much deeper than that obtaining between the major groups (or "nationalities") in rural Bosnia after the Second World War. Until 1987, policies of compromise had nonetheless ensured political stability and had made Fiji a remarkably liberal and relaxed society.

Invoking the slogan "Fiji for Fijians" (which had originally been launched before the 1977 elections), the writers of the new constitution, promulgated in 1990, ensured continued indigenous Fijian dominance of the political sphere, by according that group disproportionate representation in parliament, ruling that only Fijians can become Prime Ministers, and giving Fijians preferential treatment in other areas as well, such as religion. In addition, Fijians are guaranteed control over most of the arable land, about 82 per cent of which has been communally owned by Fijian kin groups since the beginning of colonialism in the 19th century.

The conflict culminating in the military coups can, at one level of analysis, be seen as a clear case of group competition. Indians have done better economically than Fijians. Ironically, this may partly be explained through the British colonial policy of indirect rule relating to Fijians, who were allowed to retain important traditional institutions, such as chieftainship and the rudiments of a caste system, making them in consequence unprepared to compete with Europeans and Indians in a capitalist economy later. The Indians, by contrast, had shed important aspects of their traditional social organisation in the process of migration, and were accustomed to economic individualism through the indentureship system whereby they were made to work on the European-owned plantations.

The demographic growth rate among Indians has been higher than among Fijians (just as the Muslims in Bosnia were more prolific than the other groups), and there was generally a growing sentiment among Fijian leaders that they were becoming a minority in the land of their ancestors. The inegalitarian measures introduced by the military regime, discriminating between categories of citizens on ethnic grounds, were condemned by the international community, but less strongly than one might have expected in a different setting. What is remarkable about the Fijian case is the nativist quality of the supremacist rhetoric; how they brought "sons-of-the-soil" arguments to bear on a national legislation creating in effect a two-tier society where non-Fijians were in effect relegated to the status of second-class citizens. They argued that their culture, like that of Maoris in New Zealand and Aborigines in Australia, was threatened with marginalisation from outside forces. In contrast, nobody in countries like Mauritius, Trinidad & Tobago and Guyana – which are in many ways similar to Fiji, but lack substantial indigenous populations – would have been able to invoke arguments of cultural authenticity and preservation of traditional cultures in a bid to introduce differential treatment for different ethnic groups (see Eriksen 1992, 1998a for details on Trinidad and Mauritius).

It may well be asked whether contemporary Fijians, being Methodists and proficient English-speakers, are any less culturally uprooted than, say, Trinidadians of African descent or, for that matter, Fijians of Indian descent. The question of cultural authenticity is outside the scope of this chapter; let me now make a few more pertinent points relating to this example.

Fiji had developed an informal formula for interethnic accomodation where the largest ethnic groups divided societal sectors between them. In addition there were – and still are, at least to some extent – developing fields of shared meaning and cross-cutting alignments, such as the common use of English as a national language and a shared educational system. Nevertheless the segregation between the groups in both social and cultural domains is more striking to the outside observer than tendencies towards assimilation. Conflict avoidance would thus have to rely chiefly on group compromise rather than the development of a hybridised, shared identity – an option that has occasionally been proposed by politicians and intellectuals in other insular, postcolonial plantation societies such as Mauritius and Trinidad. Finally, due to historical circumstances and cultural differences, Indians and Fijians have participated in different ways and have succeeded to varying degrees in the modern sectors of politics and the economy. It could indeed be argued that processes of modernisation in Fiji, far from reducing cultural differences, have deepened them, at least at the socially operational level. Unlike in Bosnia, it is possible to refer to differences in local organisation, cosmology and traditional economical practices when accounting for the ethnic conflict in Fiji, which is nevertheless much less violent than the Bosnian one. This is a reminder, against cultural determinists à la Huntington (1996), of the relative unimportance of cultural differences for ethnic conflict – and it also indicates one of the main strengths of the constructivist/instrumentalist perspective on identity politics: Cultural differences do not in themselves lead to intergroup conflict, but are invoked strategically to mobilise support. At the same time, it must also be conceded that the differences in life-worlds and personal identities in Fiji, like in Bosnia, explain why the political cleavages were given an ethnic expression. The differences were already there before they were exploited for particular political ends.

Hindutva: an apparent anomaly in contemporary Indian society

My third and final example again differs from the two previous ones in significant ways. India is a tough case for any scholar trying to develop a general theory of ethnicity or nationalism, and with few exceptions, it does not figure in general introductory texts on the field. India is hardly a state based on cultural similarity or even equality in the Western sense; it is a country with deeply embedded hierarchies and a very considerable degree of internal cultural variation. Its population of nearly a billion is divided by language, religion, caste and culture, and it has often been argued that India is culturally more complex than continents such as sub-Saharan Africa or Europe. Although eighty per cent of the population are Hindus in one meaning of the word or another, India also has the second largest Muslim population in the world (after Indonesia) and more Christians than all the Scandinavian countries put together. Since independence (and partition) in 1947, India has been defined in Gandhian/Nehruvian terms as a secular, federal country using English and Hindi as national languages, but with another dozen or so of official regional languages.

Since the early 1980s – but particularly forcefully during the 1990s – a formerly marginal political movement has steadily increased its influence in India, culminating in its victories in the successive General Elections of 1998 and 1999. This is the movement often referred to as hindutva, meaning roughly "Hindu-ness", which rallies behind slogans to the effect that India should be redefined as a Hindu country. The hindutva movement, led by an organisation called the RSS (Rashtriya Swayamsevak Sangh) began modestly in the interwar years, and its more recent parliamentary wing, the BJP (Bharatiya Janata Party, "The Indian People’s Party") is now in power not only federally, in New Delhi, but in several of the states as well.

The rhetoric of hindutva is strongly reminiscent of European ethnic nationalism. It invokes ancient myths of bitter defeats and noble sacrifices, re-framing them to fit a contemporary political scene. It quotes liberally from nineteenth-century poets and sacred texts, and it re-defines history to make the past conform to a re-defined present. It advocates a return to the roots, condemns Westernisation and its adverse moral effects on the young, praises the family as the key institution of society and seeks to promote the vision of India as a hindu rashtra – a Hindu nation. While the late Rajiv Gandhi allowed himself to be photographed wearing a Lacoste shirt and khaki shorts, BJP leaders always wear traditional Indian clothes. The main enemy image is nevertheless not the West but Islam, which is depicted as a martial and cruel religion alien to the subcontinent, and Indian Muslims (the descendants of converts, like Bosnian Muslims) are represented partly as traitors to Hinduism, partly as foreign invaders. The demolition of a mosque in the northern town of Ayodhya in December 1992, the ensuing riots in several Indian cities and the call for the rebuilding of a Hindu temple allegedly destroyed by a Mughal ruler four centuries ago, marked a climax of sorts in this respect (see van der Veer 1994 for details).

The phenomenal rise of this traditionalist movement is a result of several connected processes of sociocultural change or modernisation. First, the very notion of hindutva, Hindu-ness, is a modern one. Hinduism is not a "religion of the Book". It is an uncentralised religion with scores of holy scriptures, thousands of avatars (incarnations of divinities), and very many ways of worshipping them. The idea of the Hindu identity as an imagined community based on cultural similarity is alien to Hinduism as such, which is a religion based on complementarity, difference and hierarchy. Regarding political Hinduism, some Indian commentators actually speak of a Semitization of Hinduism whereby it takes on structural characteristics from the great religions of West Asia.

Secondly, the hindutva movement is explicitly modelled on European nationalism – some early hindutva ideologists were even warm admirers of Hitler – which has been, for a hundred and fifty years, an attempt to reconcile change and continuity by talking of roots and traditions in a situation of industrialisation and urbanisation. This is obvious in hindutva practice, whereby issues regarding national anthems, dress and foreign foods are given prominence, while profound social changes continue to affect everyday life as before. There is a clear connection between the rise of the BJP and the liberalisation of the Indian economy, the rise of a substantial new middle class with a strong consumerist orientation, and the rapid spread of new mass media including the Rupert Murdoch-controlled Star TV Network. While liberalisation of this kind stimulates consumerism (perceived as Westernisation), it also indirectly boosts traditionalism since the new patterns of consumption and the new media scene may indicate that cherished traditions are under threat.

Thirdly, the "contagious" influence from political Islam is obvious; hindutva is the assertion of Hindu identity as opposed to Muslim identity both in Pakistan and in India itself. Doubly ironically, hindutva has double origins in European Romanticism and West Asian political Islam. When its first ideologist, Dr. Veer Savarkar, wrote in the 1920s that "Hindutva is not the same thing as Hinduism", he was therefore right, but not for the reasons he believed. Savarkar saw hindutva as a wide-ranging social movement emanating from Hindu faith and practices, while a more historically correct account sees it as the result of cultural diffusion from Europe and West Asia.

Fourthly, and perhaps most importantly, the hindutva movement can be seen as a reaction against a growing egalitarianism in Indian society. Already in the 1950s, policies attempting to improve the conditions of the "Untouchables", the lowest castes, were introduced, and during the 1990s, very radical measures have been proposed to this effect – and in some cases carried out. About half of India’s population are now defined as being either Dalits ("Untouchables"), tribals, low-caste people or "OBCs" (Other Backward Classes), and in theory, all of 49.5 per cent of jobs in the public sectors should be reserved for these groups, following the recommendations of the government-appointed Mandal commission. Since the early 1990s, this principle has been enforced in many areas. Naturally, many members of the "twice-born", upper castes feel their inherited privileges eroding away, and hindutva is largely a movement representing the interests of the disenchanted upper castes. It is largely a reaction against the movement towards greater equality in Indian society. Although hindutva seems to promote equality among Hindus, an implication of its traditionalist Hinduism is the reinvigoration of the caste system, which in effect benefits only the "twice-born" castes.

This analysis of hindutva must by necessity be a superficial one (see Hansen 1999 for a full treatment). It must be remarked, however, that inter-religious marriages (and, indeed, intercaste marriages) are rare outside certain elite groups. Casual interaction between Hindus and Muslims is far from unusual, but like in the Bosnian and Fijian cases, the intimate (family) sphere as well as personal networks of close friends rarely cross religious boundaries. Social classification in India is nonetheless complex, and as will be indicated later, the Hindu/Muslim divide is only one of several possible social dichotomies – unlike in Bosnia and Fiji, where religious or ethnic contrasts tend to be paramount.

Some comparisons

Previous sections have outlined three contemporary conflicts involving collective identity as a political resource. The differences are obvious; the focus will therefore be on the similarities for now.

The conflicts have three important sociological features in common.

First, there is in all three cases competition over scarce resources. As Horowitz (1985) and many others writing about group conflict in contemporary societies have shown, such conflicts invariably involve perceptions of scarcity and struggles to retain or attain hegemony or equality. Successful mobilisation on the basis of collective identities presupposes a widespread belief that resources are unequally distributed along group lines. "Resources" should be interpreted in the widest sense possible, and could in principle be taken to mean economic wealth or political power, recognition or symbolic power – although what is usually at stake are either economic or political resources. This feature is easy to identify in all three examples described above: Fijians and Indians compete over relative political and economic power; the constituent groups of Bosnia compete over political power and/or sovereignty; hindutva is an attempt to defend the political and economic interests of "Hindus" in secular India.

Secondly, modernisation actualises differences and triggers conflict. With the integration of formerly discrete groups into shared economic and political systems, inequalities are made visible, as comparison between the groups becomes possible. In a certain sense, ethnicity can be described as the process of making cultural differences comparable, and to that extent, it is a modern phenomenon. The Fijian example, where the increasing integration of Fijians into the modern sphere made it apparent that Indians were doing better economically, illustrates this point. In India, the rise of the Dalit movement struggling for recognition and equal rights on behalf of "Untouchable" groups is an expression of the modern value of equality, and the counterreaction from the Hindu right is an attempt to stop egalitarianism from spreading, as well as reflecting – almost with the accuracy of a mirror-image – symbolic competition with Muslims within and (especially) outside India. The Bosnian example, admittedly, seems less straightforward, as socialist Yugoslavia was in many ways no less modern than its successor countries (some would indeed argue that at least at the level of ideology, it was infinitely more modern than them). What is clear, and which also holds true for other East and Central European countries, is that the sudden introduction of liberal political rights and a capitalist economy around 1990 – core characteristics of non-socialist modernisation – created new dimension of comparison between individuals and new arenas of competition.

Thirdly, the groups are largely self-recruiting. Intermarriage is rare in all three cases (excepting urban Yugoslavia). Although biological self-reproduction is by no means necessary for a strong collective identity to come about, it should be kept in mind that kinship remains an important organising principle for most societies in the world, and a lot of what passes for ethnicity at the local level is really kinship. Kinship has an important social dimension in addition to its symbolic side which is highlighted in ideologies of fictive or metaphoric kinship. Symbolic boundaries are never effective unless underpinned by social organisation.

Further, there are also several important ideological similarities.

First, at the level of ideology, cultural similarity overrules social equality. Ethnic nationalism in Yugoslavia, political Hinduism in India and the "sons-of-the-soil" rhetoric of Fiji all depict the in-group as homogeneous, as people "of the same kind". Internal differences are undercommunicated, and moreover, in the wider political context, equality values are discarded for ostensible cultural reasons. (Although it could be argued that hindutva is a Trojan horse concealing upper-caste interests with all-Hindu rhetoric, the point is that it stresses the commonalities of all Hindus irrespective of caste or language.)

Secondly, images of past suffering and injustice are invoked. Serbs bemoan the defeat at the hands of the Turks in Kosovo in 1389; Hindu leaders have taken great pains to depict Mughal (Muslim) rule in India from the 1500s as bloody and authoritarian, and indigenous Fijian leaders compare their plight to that of other indigenous peoples that have suffered foreign invations. Violence targeting the descendants of the invaders can therefore be framed as legitimate revenge. Even hindutva leaders, who claim to represent eighty per cent of India’s population, complain that Hinduism is under siege and needs to defend itself with all means available.

Thirdly, the political symbolism and rhetoric evokes personal experiences. This is perhaps the most important ideological feature of identity politics in general. Using myths, cultural symbols and kinship terminology in addressing their supporters, promoters of identity politics try to downplay the difference between personal experiences and group history. In this way, it becomes perfectly sensible for a Serb to talk about the legendary battle of Kosovo in the first person ("We lost in 1389"), and the logic of revenge is extended to include metaphorical kin, in many cases millions of people. The intimate experiences associated with locality and family are thereby projected onto a national screen. This general feature of social integration has been noted by Handelman (1990), analysing national rituals, and much earlier in Turner’s (1967) studies of ritual among the Ndembu of Zambia. In showing that rituals both have an instrumental and an emotional (or sensory) dimension – one socially integrating, the other metaphorical and personally meaningful – Turner actually made a point crucial to the present analysis, namely that loyalty to a larger collectivity (such as a tribe or a nation) is contingent on its imagery being personally meaningful.

Fourthly, first-comers are contrasted with invaders. Although this ideological feature is by no means universal in identity politics, it tends to be invoked whenever possible, and in the process, historical facts are frequently stretched. In Fiji, the Fijian population – although genetically a Polynesian/Melanesian mix – has a strong case here, although it is less obvious that Indo-Fijians can be immigrants to a country in which they were born, and therefore legitimately deprived of equal rights. Regarding Bosnia and India, as mentioned above, there is nothing to suggest that the ancestors of Muslims in the respective countries were more recent arrivals than the ancestors of Christians or Hindus, although Islam is a relatively recent import. What is interesting here, is how the varying depth of cultural genealogies ("roots") is used to justify differential treatment. The historical location of the self along the dimensions of descent and place is thereby invested with political significance.

Fifthly and finally, the actual social complexity in society is reduced to a set of simple contrasts. As Adolf Hitler already wrote in Mein Kampf, the truly national leader concentrates the attention of his people on one enemy at the time. Since cross-cutting ties reduce the chances of violent conflict, the collective identity must be based on relatively unambiguous criteria (such as place, religion, mother-tongue, kinship). Again, internal differences are undercommunicated in the act of delineating boundaries towards the demonised Other. This mechanism is familiar from a wide range of interethnic situations, from social classification in Zambian mining towns (Epstein 1992) to Norwegian—Sami relations in sub-Arctic Scandinavia (Eidheim 1971), the Sinhalese—Tamil conflict (Kapferer 1988) and Quebecois nationalism (Handler 1988): The Other is reduced to a minimal set of "traits", and so is the collective Self.

These similarities do not necessarily indicate that there are universal mechanisms linking personal selves and larger collectivities, but they do suggest that there is a universal "grammar" common to contemporary identity politics everywhere. In the final sections of this chapter, this argument will be pursued slightly further, and it will be suggested that universal connections between the self and the collective exist, which must be understood not only to account for traditional societies, but in order to make sense of the present. Far from being an "atavistic" or "primitive" counterreaction to globalisation or modernisation, identity politics is a special case of something more general, namely collective identity anchored in personal experiences.

Where is the identity of identity politics?

Social scientists have proposed many typologies of ethnic conflict, dividing the groups involved into categories such as majority, minority, irridentist and separatist, using variables such as division of labour, relative political power and historical intergroup relations as criteria of classification. In my view, this kind of exercise can at best generate a limited understanding of the dynamics of group conflict. To begin with, the very predicate "ethnic" is hardly appropriate to describe all conflicts based on identity politics. Indian Hindus are not an ethnic group in any meaningful sense, and it is a matter of definition whether Serbs, Croats and Muslims in Bosnia should be considered ethnic groups (they have shared origins only a few centuries back). Many contemporary conflicts displaying some or all of the features listed above cannot be seen as ethnic. To mention a few African examples: The Sudanese civil war is partly fought over religion (Northeners are Muslims trying to Islamicise the south), partly over culture and language – neither northern nor southern Sudanese are ethnic groups. Hutus and Tutsis in Rwanda and Burundi, like the constituent groups of Bosnia, are culturally very close; they speak the same language and have the same religion. The Somali civil war presents an even more puzzling case, as Somalia is one of the few sub-Saharan states that are truly ethnically homogeneous and so far the only one that seems to have relinquished the trappings of statehood completely, having dissolved into warring clans (an intermediate level of social organisation, between the family and the ethnic group) since the early 1990s. To the north-west of Somalia, one of the great forgotten wars of Africa is being fought over a contested border area between Eritrea and Ethiopia. Now, Eritrea, which seceded from Ethiopia in 1991, has never been based on religion nor ethnic identity, but has a vague legitimacy as a historical nation in the brief period of Italian colonialism before the Second World War. The current war is being fought between Tigrinya speakers on both sides of the borders, who are united through religion, language, customs, history and even kinship ties, but they are no less bitterly divided politically. This conflict in turn creates its strange bedfellows in the alliance between Ethiopian Tigrinyas and Amhara-speakers from the highlands, who are traditional enemies.

In other words, the concept "ethnic conflict" is misleading, whether it is used to classify phenomena or to explain hostilities. Several of the alternative terms one might consider are, however, no less misleading: "Cultural conflict" will clearly not do, as it is obviously not what is usually thought of as cultural differences that lie at the heart of the conflicts. At the village level, even Hindus and Muslims in India hold many of the same beliefs and worship in similar ways. The low-intensity conflict in Fiji involves groups that are by any criterion more culturally different than, say, the Bosnians groups. All the conflicts considered here are conflicts over resources perceived as scarce – territory, political power, economic gain, employment, recognition; rights in a wide sense. What they have in common is their successful appeal to collective identities perceived locally as imperative and primordial, identities associated with a deep moral commitment, whether ethnic (based on notions of kinship and descent), regional (based on place) or religious (based on beliefs and forms of worship). For these reasons, the term "identity politics" is preferable as a generic term for all such political movements, whether nonviolent or violent.

This final section will therefore amount to an attempt to unravel the identity of identity politics. What is it that makes it so powerful? What is the "identity" that such political movements can draw upon?

Benedict Anderson proposes an answer in the Introduction to his seminal Imagined Communities (Anderson 1983), where he points out that nationalism has more in common with phenomena such as religion and kinship than with ideologies like liberalism and socialism. He argues that nationalism (and, one might add, any form of identity politics) expropriates personal identity, transforming intimate experiences into the raw material of politics. I owe my existence to my parents, and by metonymical extension they represent the larger, abstract collective. I harbour tender feelings for my childhood, which by extension becomes my group’s glorious and tragic history. I feel attached to the place where I grew up, which was not just any arbitrary place, but the nation (or, as the case might be, the sacred land of Hinduism, the traditional territory of the Fijians, the tormented country of the brave, but sadly misunderstood Serbs). Indeed, this argument can profitably be seen as echoing Turner’s aforementioned argument on the instrumental and emotional dimensions (or "poles") of ritual. In both cases, the integrative strength of the imagined community (be it a tribe or a nation) depends on its ability to mobilise emotions proper to the intimate sphere of kinship and personal experience.

The conditions for this transformation to take place – the move from an interpersonally anchored identity to an abstract national, ethnic or religious identity – are usually tantamount to certain general conditions of modernity (cf. Gellner 1997). It is through school and mass media that people are taught to identify with an abstract, mythically rooted community of people "of the same kind". Through the replacement of traditional economies with an abstract labour market, they become participants in a large-scale system of subsistence. Through the implementation of a bureaucratic system of political management, their allegiances are at least partly moved from the concrete to the abstract community.

It is important to remember, as theorists of nationalism and ethnicity have pointed out time and again, that identification is relational, situational and flexible, and that each person carries a number of potential identities, only a few of which become socially significant, making a difference in everyday life. Even fewer gain political importance, forming the basis of power struggles and group competition. This is not, however, to say that collective identities can be created out of thin air. They have to be connected, in credible ways, to people’s personal experiences. These experiences in turn are flexible – not only historians, but everybody else as well selects and interprets events to make a particular kind of sense of the past – but not indefinitely so. Regarding our main examples, in Fiji, virtually nobody doubts whether they are Indian or Fijian, and politics – whether based on compromise or conflict – will have to take this into account for the foreseeable future. In Yugoslavia after the breakup, cross-cutting ties and cultural hybridity were undercommunicated. Cosmopolitanism was increasingly seen as a suspect, unpatriotic attitude, and people of mixed ancestry were forced to choose a bounded, unambiguous identity: They had to select past experiences that made them either Serbian, Croatian or Muslim, more or less like the proverbial North African mule, who speaks incessantly about his uncle, the horse, but never mentions his father, the donkey. In India, finally, some of the strongest scholarly arguments against the lasting influence of hindutva have actually pointed towards people’s personal experiences (Frøystad 1999). Since Indian everyday life is still permeated by caste distinctions, and caste continues to define the very fabric of social integration, these scholars argue that hindutva – the idea that all Hindus have something profound in common – is so counter-intuitive to most Indians that it can only unite Hindus as long as the enemy image of Muslims can be kept ablaze.

All the basic components of political identity familiar from classic political anthropology can be identified in contemporary identity politics: It is based on a sometimes ambiguous mix of kinship and locality; it has well developed myths of origin and myths of past suffering; and it distinguishes clearly between "us" and "them". The main difference between, say, the nomadic Nuer society studied by E.E. Evans-Pritchard (1940) in the 1930s and Serbian (or Croatian) nationalism today, is probably that of scale: While the Nuer rarely imagined themselves as members of larger groups than the clan, a Serb in Vojvodina can readily identify him/herself with a Serb in Kosovo. The act of transformation from personal, concrete social experiences to the abstract community is naturally much more demanding in a large-scale society than in a village-based one, hence the importance of modern institutions of communication, economic transactions and political rule for the growth of abstract communities.

Some final lessons from political anthropology

Having long ago abandoned the early ambition of becoming "a natural science of society", social and cultural anthropology has for decades been reluctant to formulate lawlike propositions about the functioning of society. The constructivist turn of recent years seems to confirm that contemporary anthropology is less concerned with absolute truths than with the analysis of local cultural constructions. This need not be so, and the study of current identity politics may illustrate the power of comparative anthropology in generating general hypotheses.

Early instrumentalist research on ethnic groups, particularly in Africa (as in A. Cohen 1974), searched for the logic of group cohesion, which they assumed to be roughly the same everywhere. The related, actor-based perspective developed by Barth (1969) and his colleagues, assumed the logic of action to be quite universal – people act to maximize benefits. Later analyses of the constructedness of ethnic and national ideologies (the seminal text is Hobsbawm and Ranger 1983; see Chapman et al. 1989 for an overview) also emphasise universal characteristics of a particular kind of societal formation (the modern state) and its relationship to group identities based on notions of culture. The canonical texts on nationalism (such as Gellner 1983, Anderson 1983, Smith 1991) also have clear universalistic ambitions. As this chapter has made clear, these approaches have obvious strengths, but they need to be supplemented by detailed research on the experiential world of the everyday; the Lebenswelt (life-world) of the actors. A renewed focus on the informal, intimate and often non-instrumental dimensions of everyday life reveals that terms such as "ethnicity" by themselves explain little. The parallels between a supporter of the BJP and a supporter of Serbian supremacists should not be located to their respective "ethnic" identities or "civilizational" membership, but to the fact that their everyday life, social networks and personal obligations connect them to particular groups that may be exploited politically, given the right circumstances. It should also be kept in mind that class politics can sometimes be a form of identity politics (Shore 1993), which can profitably be understood along the same lines as ethnic or religious identity politics. The cause of group allegiance lies in the everyday, not in the overarching ideology.

In order to complete this analysis, it is necessary to go a few decades back, to classic political anthropology, in order to see how the perspectives developed earlier in the chapter can be enriched by the work of previous generations.

The integration of persons into groups can be described as the work of an inverted refrigerator: The function of a refrigerator is to generate inward coldness, but in order to do so, it more or less inadvertently, as a side-effect, creates outward warmth. Conversely, groups form to create warmth for their members, but they necessarily create some outward coldness in order to be able to do so. Under particular circumstances, the outward coldness is more readily perceptible than the inward warmth. A sociological principle originally formulated by Georg Simmel, known as "Simmel’s Rule", simply states that the internal cohesion of a group is contingent on the strength of external pressure. This principle may explain why group integration generally is so much stronger in small groups, especially if they are oppressed, than in large ones – why, for example, Scots seem to have fewer difficulties in defining who they are than the English.

An interesting corollary of Simmel’s Rule is the fact that what kind of group that emerges depends on where the perceived pressure comes from. Both gender-based and class-based social movements have periodically been successful, given that the perceived threat was seen not as alien religion or foreign ethnic groups, but as male supremacy and ruling classes, respectively. In accordance with this, some inhabitants of Sarajevo during the war felt that the conflict was really an urban—rural one, since city-dwellers had a lot in common, irrespective of religion, that they did not share with rural people. Strong opposition groups in India, similarly, argue against a view of Indian politics as divided between Hindu communalists and liberals, as they see the main problem of Indian society as one of poverty and distribution of resources, which neither of the parties seems to give priority to. In Fiji, finally, the immediate reason for the 1987 coup was the establishment of a government of national unity promising to address issues of social welfare and economic development rather than intergroup issues. In other words, re-definitions of societal cleavages are entirely possible in so far as they do not contradict people’s everyday experiences too obviously.

In the course of this chapter, cross-cutting ties and conflicting loyalties have already been mentioned as mitigating forces in situations of intergroup conflict. Phrased within the terminology used here, one might say that shared experiences across boundaries reduce the risk of conflict. In Max Gluckman’s re-interpretation of Evans-Pritchards Nuer material from the 1930s (Gluckman 1956), this point was made forcefully. The Nuer were organised along kinship lines across villages, but they were also locally integrated in villages. The women married out of the village and the lineage, so that everybody had affines (in-laws) in other villages. Furthermore, men were tied to non-relatives through trade, initiation rituals and friendship. All of these factors led to a reduction in the incidence of violence among the Nuer. In contemporary identity politics, it can easily be seen – and has been remarked above – how political leaders emphasise inwards similarity and outwards boundaries in order to reduce the potentially mitigating impact of cross-cutting ties. They are true to people’s everyday life, but try to emphasise certain experiences at the expense of others (inward solidarity and similarity, outward conflict and difference).

Another, even more time-honoured principle from political anthropology is the twin notion of fusion and fission in tribal societies. When the sole organising principle for a group lacking hierarchies and formal political office is kinship, there are limits to the group’s growth; at a certain point, it splits into two. Without such a fission, internal conflicts would soon become overwhelming given the simple social organisation of such societies, and the effects would be destructive. Fusion of discrete groups has also been studied extensively, but in many acephalous societies it is seasonal (nomadic groups fuse in the dry season or in winter) and fragile.

A more dynamic view of contraction and expansion of tribal groups was developed, especially by Africanists, from the 1940s (Fortes and Evans-Pritchard 1940). In studies of feuding and political competition, they showed how two or several local groups who might be periodically involved in mutual feuding, united temporarily when faced with an external enemy. This form of organisation, described as segmentary by Evans-Pritchard, follows the proverb often cited in recent years to explain the logic of the Somali civil war: "It’s me against my brother, my brother and I against our cousins, and our cousins, my brother and I against everybody else." A form of segmentary logic is apparent in politics nearly everywhere; a distinguishing mark of modern nation-building has nevertheless been its attempt to channel loyalties away from various sub-national levels of identity in order to monopolize the political loyalty of individual citizens.

The segmentary logic creates a fluid, relational political organisation which,, in its pure form, is impracticable in modern state societies given their requirements for stability, centralised power and reified systems of political representation. This does not, however, mean that segmentary identification does not continue to exist, and one of the causes of oppositional identity politics in modern nation-states is caused by their not providing subnational identity groups appropriate political arenas, thereby encouraging counterreactions in the form of identity politics directed against the state.

The formation of identity based political groups generally entails both an expansion and a contraction of the focus for identification. At the time of the breakout of conflict in Bosnia, the federal or even state level was increasingly seen as irrelevant – the process was one of fission. At the same time, internal conflicts and schisms within each constituent group were minimised, and as a result, each group became more coherent and united than before. In the cases of Fiji and India, this is even more obvious: Among Fijians, rivalry between chiefs and clans has diminished in importance as Fijian politics has grown increasingly ethnic; similarly, rifts within the Indian population on the basis of regional origin, which could formerly lead to Indian sub-groups supporting Fijian-dominated governments, have become much less important since the military coup. In the Indian case, the very idea of hindutva implies an enormous expansion of the in-group for Hindus. Trying to bridge differences based on language, caste, region and culture, hindutva tries to create a morally committing all-Indian Hindu identity based on symbolic equality. This is, in the Indian context, a very radical move. Simultaneously, Indian citizenship and (over-arching, supra-religious) national identity become less important since the federal Indian state includes millions of non-Hindus, who are depicted as internal enemies by hindutva spokespersons.

In all the examples considered here, group segmentation at a higher level and the ensuing formation of imagined communities larger than the locality but smaller than the state, is immediately related to the need for a firm boundary in a situation of conflict within the state. The alternative identity of national citizenship, which encompasses the other as well, no longer functions. Interestingly, contemporary identity politics is very similar to nationalism, for example in its appeal to mythical foundations, its abstract postulation of similarity and equality, its rejection of segmentary identity formation and its attempt to reduce a world of many small differences to a world of only a few, major ones. In many cases, it is more successful than nationalism, particularly in postcolonial, multiethnic states. This is not only because identity politicians promise its adherents that they will win zero-sum games against political competitors, but also because they are able to represent themselves as natural extensions of people’s personal, experience-based identities.

A challenge for modern states, thus, consists in coping with the fact that personal identity can be exploited politically not only by the state itself but by others as well, not because the self is infinitely multifaceted (it is not), but because the experiences and relationships that make up the self can be expanded symbolically in several, often conflicting ways. Processes of segmentary fusion and fission, the formation of different kinds of groups involving overlapping personnel due to the functioning of "Simmel’s Rule", group-based antagonism and competition: these ways of expressing political interests, underpinned by shared meaning within the in-group, are no more eradicated by modernity than personal identity is. The challenge thus consists in laying the foundations for "a sense of belonging to a community larger than each of the particular groups in question" (Laclau 1995:105), and this can only be done by first acknowledging both the richness and the variability of personal identities.

In the face of violent identity politics, "ethnic cleansing" and the strong attention to "roots" and historically based identities (ungenerously described as "the narcissism of small differences" by Michael Ignatieff) characterising many societies in recent years, it is not surprising that intellectuals have recently tried to think essentialism away, emphasising the endlessly flexible and fluid character of human identification. A typical expression of this position can be found in a recent text by the influential sociologist Zygmunt Bauman, where he states: "If the modern ‘problem of identity’ is how to construct an identity and keep it solid and stable, the postmodern ‘problem of identity’ is primarily how to avoid fixation and keep the options open" (Bauman 1996, p. 18).

In today’s world it may be ethically imperative to endorse Bauman’s position, but it is equally important to keep in mind that humans are not free-floating signifiers, and no amount of benevolent intentions will be able to change people’s life-worlds overnight. Rather than trying to think them away, it is necessary to understand them and come to terms with their enduring power. Notwithstanding globalisation and the universalisation of modernity, cultural differences continue to exist, within and between places, within and between nations and ethnic groups. It is also, however, doubtless true that carbon can be turned into graphite as well as diamonds, and the ways in which cultural differences become socially relevant vary importantly. But to pretend they do not exist outside ethnic and nationalist ideologies would be intellectually indefensible; people’s personal experiences are the very raw material of such ideologies. Here lies an important limitation in constructivist models of identity. Collective identities are constructed, consciously or not, but nothing comes out of nothing. In locating the universal not in the workings of identity politics (it changes historically and varies geographically), nor in the eternal sovereignty of the state (the same objection applies), but rather in the social life-worlds in which individuals make sense of the world, we may have found a basis for comparison which will outlive academic fads and contemporary politics.

The author would like to express his thanks to the editors and three anonymous readers for their critical readings and useful comments on an earlier version of the chapter.

References

Anderson, Benedict (1983) Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso.

Barth, Fredrik, ed. (1969) Ethnic Groups and Boundaries. Oslo: Scandinavian University Press.

Bauman, Zygmunt (1993) Postmodern Ethics. Cambridge: Polity.

– (1996) From pilgrim to tourist; or A short history of identity. In Stuart Hall and Paul Du Gay, eds., Questions of Cultural Identity, pp. 18—36. London: Sage.

Bringa, Tone (1996) Being Muslim the Bosnian Way: Identity and Community in a Central Bosnian Village. Princeton, N.J.: Princeton University Press.

Chapman, Malcolm, Maryon McDonald & Elizabeth Tonkin (1989) Introduction – history and social anthropology. In Elizabeth Tonkin, Maryon McDonald and Malcolm Chapman, eds., History and Ethnicity, pp. 1-21. London: Routledge.

Cohen, Abner (1969) Custom and Politics in Urban Africa: A Study of Hausa Migrants in a Yoruba Town. London: Routledge & Kegan Paul.

– (1974) Two-Dimensional Man: An essay on power and symbolism in complex society. London: Routledge & Kegan Paul.

Cohen, Anthony P. (1994) Self Consciousness. An Alternative Anthropology of Identity. London: Routledge.

Cornell, Stephen and Douglas Hartmann: Ethnicity and Race: Making Identities in a Changing World. London: Pine Forge Press 1998.

Eidheim, Harald (1971) Aspects of the Lappish Minority Situation. Oslo: Scandinavian University Press.

Epstein, A.L. (1992) Scenes from African Urban Life: Collected Copperbelt Essays. Edinburgh: Edinburgh University Press.

Eriksen, Thomas Hylland (1992) Us and Them in Modern Societies: Ethnicity and Nationalism in Trinidad, Mauritius and Beyond. Oslo: Scandinavian University Press.

– (1998a) Common Denominators. Ethnicity, Nationalism and Compromise in Mauritius. Oxford: Berg.

– (1998b) Culture and ethnicity: A second look. Lecture at 6th SIEF conference, Amsterdam, 21 April 1998.

Evans-Pritchard, E.E. (1940) The Nuer. Oxford: Clarendon.

Featherstone, Mike, ed. (1990) Global Culture. London: Sage.

Fortes, Meyer and E.E. Evans-Pritchard, eds. (1940) African Political Systems. London: Oxford University Press.

Friedman, Jonathan (1994) Global Culture and Local Process. London: Sage.

Frøystad, Kathinka (1999) Personal communication.

Furnivall, J.S. (1948) Colonial Policy and Practice :A Comparative Study of Burma and Netherlands India. New York: New York University Press.

Geertz, Clifford, ed. (1967) Old Societies and New States: The Quest for Modernity in Africa and Asia. New York: The Free Press.

– (1973) The Interpretation of Cultures. New York: Basic Books.

Gellner, Ernest (1983) Nations and Nationalism. Oxford: Blackwell.

– (1997) Nationalism. London: Weidenfeld & Nicolson.

Giddens, Anthony (1991) Modernity and Self-Identity. Cambridge: Polity.

– (1994) Beyond Left and Right. Cambridge: Polity.

Gluckman, Max (1982/1956) Custom and Conflict in Africa. Oxford: Blackwell.

Handelman, Don (1990) Models and Mirrors: Towards an Anthropology of Public Events. Cambridge: Cambridge University Press.

Handler, Richard (1988) Nationalism and the Politics of Culture in Quebec. Madison: Wisconsin University Press.

Hannerz, Ulf (1992) Cultural Complexity: Studies in the Social Organization of Meaning. New York: Columbia University Press.

Hansen, Thomas Blom (1999) The Saffron Wave: Democracy and Hindu Nationalism in Modern India. Princeton, N.J.: Princeton University Press.

Hobsbawm, Eric and Terence Ranger, eds. (1983) The Invention of Tradition. Cambridge: Cambridge University Press.

Holland, Dorothy (1997) Selves as Cultured: As Told by an Anthropologist. In Richard Ashmore and Lee Jussim, eds., Self and Identity: Fundamental Issues, pp. 160-190. Oxford: Oxford University Press.

Horowitz, Donald L. (1985) Ethnic Groups in Conflict. Berkeley: University of California Press.

Huntington, Samuel (1996) The Clash of Civilizations and the Remaking of a World Order. New York: Simon and Schuster.

Ignatieff, Michael (1994) Blood and Belonging: Journeys into the New Nationalism. London: Vintage.

Jenkins, Richard (1996) Social Identity. London: Routledge.

Kapferer, Bruce (1988) Legends of People, Myths of State:

Kelly, John D. (1998) Aspiring to Minority and Other Tactics Against Violence. In Dru C. Gladney, ed., Making Majorities: Constituting the Nation in Japan, Korea, China, Malaysia, Fiji, Turkey and The United States, pp. 173-197. Stanford, CA: Stanford University Press.

Laclau, Ernesto (1995) Universalism, particularism and the question of identity. In J. Ratchmann, ed., The Identity in Question. London: Routledge.

Lash, Scott and James Urry (1994) Economies of Signs and Space. London: SAGE.

Premdas, Ralph (1993) Balance and ethnic conflict in Fiji. In John McGarry and Brendan O’Leary, eds., The Politics of Ethnic Conflict Regulation, pp. 251-274. London: Routledge.

Rex, John (1995) The Nature of Ethnicity in the Project of Migration, in Montserrat Guibernau and John Rex, eds., The Ethnicity Reader: Nationalism, Multiculturalism and Migration. Cambridge: Polity.

Roosens, Eugeen E. (1989) Creating Ethnicity. London: SAGE.

Said, Edward W. (1978) Orientalism. New York: Pantheon.

Schlesinger, Arthur M. jr. (1992) The Disuniting of America. New York: Norton.

Schöpflin, George (1993) The rise and fall of Yugoslavia. In John McGarry and Brendan O’Leary, eds., The Politics of Ethnic Conflict Regulation, pp. 172-203. London: Routledge.

Shore, Cris (1993) Ethnicity as Revolutionary Strategy: Communist Identity Construction in Italy. In Sharon Macdonald, ed., Inside European Identities, pp. 27-53. Oxford: Berg.

Smith, Anthony D. (1986) The Ethnic Origins of Nations. Oxford: Blackwell.

– (1991) National Identity. Harmondsworth: Penguin.

Strathern, Marilyn (1991) Partial Connections. Savage, MD: Rowman and Littlefield.

Tambiah, Stanley (1994)

Tinker, Hugh (1974) A New System of Slavery: The Export of Indian Labour Overseas 1830-1920. Oxford: Oxford University Press.

Turner, Victor (1967) The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca, NY: Cornell University Press.

Turton, David, ed. (1997) War and Ethnicity: Global Connections and Local Violence. Woodbridge, Suffolk: University of Rochester Press.

Van der Veer, Peter (1994) Religious Nationalism. Hindus and Muslims in India. Berkeley: University of California Press.

Williksen-Bakker, Solrun (1991) Fijians in business: A study in the transformation of symbols. Ph. D. thesis, University of Oslo.